- Email: [email protected]
The novel MER transposon-derived miRNAs in human genome
Gene 512 (2013) 422–428
Contents lists available at SciVerse ScienceDirect
Gene journal homepage: www.elsevier.com/locate/gene
The novel MER transposon-derived miRNAs in human genome Kung Ahn a, Jeong-An Gim a, Hong-Seok Ha b, Kyudong Han c, Heui-Soo Kim a,⁎ a b c
Department of Biological Sciences, College of Natural Sciences, Pusan National University, Busan 609‐735, Republic of Korea Department of Genetics, Human Genetics Institute of New Jersey, Rutgers, the State University of New Jersey, 145 Bevier Rd, Piscataway, NJ 08854, USA Department of Nanobiomedical Science & WCU Research Center, Dankook University, Cheonan 330‐714, Republic of Korea
a r t i c l e
i n f o
Article history: Accepted 17 August 2012 Available online 25 August 2012 Keywords: MER transposon MicroRNA MER derived miRNA Palindromic structure AGO proteins
a b s t r a c t MicroRNAs (miRNAs) are small RNA molecules (~20–30 nucleotides) that generally act in gene silencing and translational repression through the RNA interference pathway. They generally originate from intergenic genomic regions, but some are found in genomic regions that have been characterized such as introns, exons, and transposable elements (TE). To identify the miRNAs that are derived from palindromic MERs, we analyzed MER paralogs in human genome. The structures of the palindromic MERs were similar to the hairpin structure of miRNA in humans. Three miRNAs derived from MER96 located on chromosome 3, and MER91C paralogs located on chromosome 8 and chromosome 17 were identified in HeLa, HCT116, and HEK293 cell lines. The interactions between these MER-derived miRNAs and AGO1, AGO2, and AGO3 proteins were validated by immunoprecipitation assays. The data suggest that miRNAs derived from transposable elements could widely affect various target genes in the human genome. Crown Copyright © 2012 Published by Elsevier B.V. All rights reserved.
1. Introduction MicroRNAs (miRNAs) are single-stranded RNAs that are ~ 20–30 nucleotides long in their mature forms. They are generally responsible for gene silencing by binding to target messenger RNA (mRNA) transcripts. miRNAs are produced in several steps: they are transcribed as long primary transcripts, which are referred to as primary miRNA (pri-miRNA) transcripts. Next, they are processed by 2 well-known proteins, namely, Drosha and Dicer. Drosha cleaves the pri-miRNA down to ~70 nucleotides in length with a short double-stranded hairpin structure (precursor miRNA) in the nucleus. The precursor miRNA is exported to the cytoplasm and cleaved by Dicer to generate the mature miRNA form (~20–30 nucleotides). The mature miRNA comprises an active RNA-induced silencing complex (RISC) along with argonaute protein (AGO), which binds to the mature miRNA and mediates the interaction between the miRNA and its target mRNA (Denli et al., 2004; Lee et al., 2004; Okamura et al., 2004, 2009; Yang and Lai, 2010). miRNAs have been highlighted as key regulators in numerous types of cells and organismal processes (Krichevsky et al., 2003; Piriyapongsa and Jordan, 2008; Xu et al., 2010). In addition, it was reported that alterations in miRNA expression are associated with the initiation, progression, and metastasis of human tumors (Shalgi et al., 2010). Although most miRNAs originate from intergenic genomic sequences, some of them originate from sequences that have been Abbreviations: TEs, transposable elements; MER, MEdium Reiteration frequency; AGO proteins, argonaute proteins; miRNA, microRNA. ⁎ Corresponding author. Tel.: +82 51 510 2259; fax: +82 51 581 2962. E-mail address: [email protected] (H.-S. Kim).
characterized such as genes and transposable elements (Piriyapongsa and Jordan, 2007, 2008; Smalheiser and Torvik, 2005; Zhang et al., 2007). Furthermore, retroviral transposable elements (TEs) such as human immunodeficiency virus and bovine leukemia virus encode viral miRNAs (Althaus et al., 2012; Kincaid et al., 2012; Klase et al., 2007). In particular, because of the wide distribution of the TEs, the miRNAs derived from them could harmfully affect host cells and their genomic stability. If a miRNA that is derived from TEs with a high copy number is involved in gene silencing, it would have a huge impact on an individual's survival. However, it is still unclear how miRNA is produced from a TE. In the case of Made 1, it has been reported that its palindromic structure (the origin of hsa-mir-548) is produced by either orientation (+/−) and that it can form itself into a hairpin structure. Then, it is inserted into transcriptionally active genomic regions, and it takes the form of pri-miRNA structures that can be processed by the RNA interference enzymatic machinery (Piriyapongsa et al., 2007). MEdium Reiteration frequency (MER) refers to interspersed repeats in the genomes of primates, rodentia, and lagomorpha. Age estimations place the origin of most MER repeats at the time of decline in Mammalian-wide Interspersed Repeats (MIR) retroposition and before the origin of the Alu family. Intriguingly, short palindromic sequences contain a hairpin structure, which is very similar to the structure of miRNA. The miR-1302 gene family was derived from the MER53 transposon that retains its palindromic structure (Yuan et al., 2010). In this study, we computationally identified miRNAs that originated from MER elements and experimentally confirmed them. To identify the miRNA, we compared palindromic MERs with miRNA precursors and mature miRNAs from miRBase (http://www.mirbase.org/). Then, we carried out miRNA
0378-1119/$ – see front matter. Crown Copyright © 2012 Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gene.2012.08.028
K. Ahn et al. / Gene 512 (2013) 422–428
prediction to identify potential miRNA derived from the paralogs of palindromic MER consensus sequences. Finally, we experimentally verified the existence of MERs that originated from miRNAs. 2. Materials and methods 2.1. Computational analysis MER consensus sequences containing palindromic sequences were obtained from the Repbase browser of the Genetic Information Research Institute database. The genomic locations and sequences of repetitive elements, including MERs, were extracted from the human genome by using UCSC Genome Browser hg18 (http://genome.cse. ucsc.edu/) and they were analyzed using RepeatMasker 3.27 (http:// www.repeatmasker.org/). The sequences and coordinates of human pre-miRNAs and mature miRNAs were downloaded from miRBase v13.0 (http://www.mirbase.org/) and mapped to the human genome (hg18). Using the UCSC Table Browser, the genomic locations of the miRNAs and the repetitive elements from the human genome were compared. To identify the potential MER-derived miRNA genes, we developed a three-step operational scheme. We blasted the human genome with the palindromic consensus sequences of MER families using the BLAT program. Then, we carried out predictions of the MiRNA that originated from MER on MER paralogs by using miPred [http:// www.bioinf.seu.edu.cn/miRNA/] and miRNAFinder [http://bioinfo3. noble.org/mirna/]. miPred is a search tool for which all pre-miRNAs have a characteristic stem-loop hairpin structure. Therefore, the hairpin structures provide key clues to the ab initio prediction of pre-miRNAs. In addition, miRNAFinder can predict potential intronic miRNA in intron regions of the expressed genes (ESTs/cDNAs), find possible miRNA in genomic sequences, or predict if the input small RNA is mature miRNA. A mature miRNA is processed from the left or right arm of a potential precursor miRNA sequence. We examined the secondary structures of single stranded RNA sequences by using the RNAfold web server, which has limits of 7500 nt for partition function calculations and 10,000 nt for minimum free energy. The TargetScan 6.2 [http:// www.targetscan.org/] was used to identify the potential target sites of miRNAs. To determine the functional categories to which the target genes belonged, we used GOmir [http://www.bioacademy. gr/bioinformatics/projects/GOmir/] and displayed the GO categories of the target genes. 2.2. Cell cultures and lysates preparation HEK293, HCT116, and HeLa cells were seeded in Dulbecco's modified Eagle's medium that was supplemented with 10% (v/v) heatinactivated fetal bovine serum and 1% (v/v) antibiotics–antimycotic solution, and then incubated at 37 °C in a 5% (v/v) CO2 incubator. Cells (5× 106) from each cell line were washed with DPBS and collected by trypsinization. The cells were incubated in 1 ml of lysis buffer (20 mM Tris, pH 7.4, 2.5 mM MgCl2, 200 mM NaCl, 0.05% NP40) for 10 min on ice, and the cell lysates were cleared by centrifugation at 15,000 ×g for 20 min at 4 °C. 2.3. Immunoprecipitation, poly (A) tailing, and reverse-transcription To extract small RNAs from the human cells, AGO proteins AGO1, AGO2, and AGO3 were immunoprecipitated. AGO1, 2, 3-Antibody Beads were purchased from Wako Pure Chemical Industry, Ltd. Fifty microliters of Anti-AGO1, AGO2, and AGO3 Antibody Bead solutions and human cell lysates were incubated at 4 °C for 2 h. After the antigen–antibody reactions, the incubated mixtures were washed twice, and to each was added 50 μl of the elution solution. The eluted microRNAs were purified using phenol:chloroform:isoamyl alcohol mixture (25:24:1) and visualized on a polyacrylamide gel (data not shown). The isolated miRNAs were subjected to poly (A) tailing;
423
4 μg of each was polyadenylated by 5 U of poly (A) polymerase (Ambion). The reactions were incubated at 37 °C for 30 min. After incubation, poly (A) tailed miRNAs were recovered by phenol/chloroform extraction and ethanol precipitation. Reverse transcription of the poly (A) tailed miRNAs was performed using M-MLV reverse transcriptase (RT, Invitrogen) following the manufacturer's instructions. A solution (9.5 μl) including 4 μg of the poly (A) tailed miRNAs and 1 μl of 10 mM RT linker sequence (CTGTGAATGCTGCGACTACGA-18 dTs) was incubated at 65 °C for 5 min to remove any RNA secondary structures. After incubation, 5 μl of 10 mM dNTP mix, 4 μl of 5× RT buffer (including 250 mM Tris–HCl, 375 mM KCl, 15 mM MgCl2, and 50 mM DTT), 0.5 μl of RNase inhibitor and 1 μl of M-MLV reverse transcriptase were added. The reaction was then incubated at 42 °C for 90 min. After incubation, the reaction was allowed to proceed at 4 °C for 60 min. 2.4. RT-PCR MER91C located on chromosome 8 was amplified using the sense primers 5′-TGAAGGGGTTACAATTGGCAT-3′ (Chr.8, MER91C derived mature miRNA 5P) and 5′-TGCGGATGGCACCTCCTGAG-3′ (Chr.8, MER91C derived mature miRNA 3P). MER91C located on chromosome 17 was amplified using the sense primers 5′-TACAACTGGAAGGATGTTCAT-3′ (Chr.17, MER91C derived mature miRNA 5P) and 5′-GTGACATCCCTTGA GTTGTGC-3′ (Chr.17, MER91C derived mature miRNA 3P). MER96 located on chromosome 3 was amplified using the sense primers 5′-TGCAGCATTT AAGGAAGCACC-3′ (Chr.3, MER96 derived mature miRNA 5P) and 5′-GAG TGCCTCCTTAAATGTTTT-3′ (Chr.3, MER96 derived mature miRNA 3P). All the amplifications were carried out using the antisense primer (Universal primer) 5′-CTGTGAATGCTGCGACTACGAT-3′. RT-PCRs were carried out for 30 cycles at 94 °C for 40 s, 56 °C for 40 s, and 72 °C for 7 min. 3. Results 3.1. Comparative analysis of palindromic MERs and miRNAs from miRBase Eight palindromic MER consensus sequences were found (Fig. 1). Among them, the consensus sequence of MER91C matched the precursor of hsa-miRNA-652 in the miRBase and the E-value of the match was 4.00E− 08. Through a UCSC blat search, 80 paralogs aligned with the consensus sequence of MER91C were identified in the human genome (hg18). In addition, the consensus sequence of MER96 was matched with a hsa-miR-3680 and the E-value was 5.00E− 18. Ninety-seven paralogs of the MER96 consensus sequence were found in the human genome (Table 1). The palindromic consensus MER sequences were matched with the precursor miRNA sequence of the miRBase. As shown in Fig. 2, (a) MER53 was matched with precursor miR-1302-1 and (b) MER91C was matched with precursor miR-652, while (c) MER96 was matched with precursor miR-3680/3680*. We aligned the palindromic consensus sequences of the MER elements with the precursor miRNA sequences in the miRBase (Fig. 2). The consensus sequence of MER53 had 83.6% similarity with the precursor of has-mir-1302-1. In addition, the consensus sequence of MER91C had 75.8% similarity with the precursor of hsa-mir-652, and the consensus sequence of MER96 had 85.1% similarity with the precursor of has-mir-3680. The target genes of hsa-mir-1302 and hsa-mir-652 seemed to be intricately involved in cellular processes, intracellular membrane, catalytic activity, and binding by gene ontology (GO) cluster analysis (Fig. S1). However, the consensus sequences of other MERs, including MER123, MER124, MER126, MER133B, and MER134 did not exactly match any of the precursor miRNAs in the miRBase. Although there were no matches with the precursor miRNAs, we could not exclude the possibility that
424
K. Ahn et al. / Gene 512 (2013) 422–428
Fig. 1. Hairpin structures of MER DNA transposon consensus sequences. Eight palindromic consensus MER secondary structures of single-stranded RNA sequences were predicted by RNAfold. The MER consensus sequences containing palindromic sequences were obtained from the Repbase browser of the Genetic Information Research Institute database. The consensus sequence sizes are presented as MER53—189 bp, MER91C—140 bp, MER96—175 bp, MER123—210 bp, MER124—397 bp, MER126—308 bp, MER133B—140 bp, and MER134—216 bp, respectively.
they were potential miRNAs. We examined the conservation of respective mature miRNAs among paralogs of MER53, MER91C, and MER96. As shown in Fig. 3, each mature miRNA sequence exhibited high conservation among the MER paralogs. 3.2. Potential miRNAs derived from the palindromic sequences of MER paralogs To detect the miRNA derived from MER paralogs, we predicted potential miRNAs in the palindromic sequences of MER paralogs by using the miPred web server tool (URL: http://www.bioinf.seu.edu. cn/miRNA/). The miPred program can be used to predict either pseudo or real miRNA precursors by using the structural characteristics of the MER paralogs. Among the MER91C paralogs, 5 paralogs contained miRNAs: 1 contained pseudo miRNA and 4 contained real miRNA precursors. Seven among the MER96 paralogs contained miRNAs, 2 of which were real miRNA precursors. In addition, MER126 located on chromosome 11 and MER134 located on chromosome 2 were determined to be real miRNA precursors (Table 2). The MER91C paralogs containing the real miRNA precursors had high sequence homology with the consensus sequence of MER91C. The paralogs located on chromosomes 8, 9, 12, and 17 showed
sequence identities of 86.92%, 85.62%, 75.18%, and 84.31% with the consensus sequence, respectively. The MER96 paralogs that contained the real miRNA precursors were located on chromosomes 3 and 12, and they showed sequence identities of 69.23% and 90.51% with the consensus sequence of MER96, respectively. The MER134 and MER126 paralogs containing the real miRNA precursors showed sequence homologies of 71.29% and 71.56% with their consensus sequences, respectively. We analyzed the secondary structures of the MER paralogs that were predicted to have pseudo miRNA or real miRNA precursors using the RNAfold web server (http://rna.tbi.univie.ac.at/cgi-bin/ RNAfold.cgi) (Fig. S2). Through the analysis, we found that the MER paralogs had palindromic structures. We further investigated the secondary structures of the MER paralogs by comparing them with the secondary structures of their consensus sequences. When miRNAs originating from MER91C were compared with the consensus sequence of MER91C, they were found to contain palindromic sequences that were similar to the sequences responsible for the central regions of their stem-loop structures. The MER96 paralog located on chromosome 3 had a deletion of 50 bp in the central region of the stem-loop structure of its consensus sequence. The stem-loop region of MER126 located on chromosome 11 consisted of the stem-loop region of its consensus
Table 1 Comparison between miRBase and palindromic MER consensus sequences. Repeat name
Repeat family
Repeat length (bp)
Structure
Precursor sequence (mirbase sequence)
E-value
n-Number of paralogs (consensus sequence matched paralogs)
MER53 MER91C MER96 MER123 MER124 MER126 MER133B MER134
DNA DNA DNA DNA DNA DNA DNA DNA
189 140 175 210 397 308 140 213
Palindrome Palindrome Palindrome Palindrome Palindrome Palindrome Palindrome Palindrome
mir-1302-1 mir-652 mir-3680 Δmir-3606 Δmir-1302-4/Δmir-1302-6/Δmir-1302-8/Δmir-592/Δmir-3012a Δmir-383/Δmir-3934 or new miRNA Δmir-1279/Δmir-1323 Δmir-129-1
2.00E−30 4.00E−08 1.00E−17 E>1 E>1 E>1 E>1 E>1
n = 178 n = 80 n = 97 n=6 n = 25 n = 30 n=4 n = 11
K. Ahn et al. / Gene 512 (2013) 422–428
425
Fig. 2. Palindromic consensus MER sequences matched with the precursor miRNA sequences of miRBase. (a) MER53 was matched with precursor miR-1302-1 and (b) MER91C was matched with the precursor miR-652. (c) MER96 was matched with precursor miR-3680/3680*. The dotted line box indicates mature miRNA sequences present in the MER transposon, and the blank box indicates mature miRNA sequences. The values in the lower panel indicate the minimum free energy.
sequence. Taken together, this suggests that the palindromic structure of the miRNA that originated from MER was derived from a specific region of the MER consensus sequence that is responsible for the formation of its secondary stem-loop structure. However, the stem-loop region of the MER 134 paralog did not consist of its consensus sequence although the stem-loop of the MER 134 paralog was very similar to the stem-loop region of the consensus sequence (Fig. S3). In order to detect mature miRNAs derived from the predicted real miRNA precursors, we scanned them using miRNAFinder. By doing this, we identified duplex mature miRNA sequences in the precursors (Fig. S4).
3.3. Identification of MER derived miRNAs loaded onto AGO proteins in human cells In order to identify mature miRNA derived from MER transposons, we first immunoprecipitated 3 AGO proteins, AGO1, AGO2, and AGO3, from human cell lines HeLa, HEK293, and HCT116 because these proteins are miRNA processors. Then, the miRNAs loaded on the AGO proteins were processed for RT-PCR: they were isolated from the proteins, a poly A tail was added to the 3′ end of each miRNA, and then the miRNAs with poly A tails were reverse transcribed
a) MER 53 hsa-mir-1302 MIMAT0005890 UUGGGACAUACUUAUGCUAAA
b) MER 91C hsa-mir-652 MIMAT0003322 AAUGGCGCCACUAGGGUUGUG
c) MER 96 hsa-mir-3680* MIMAT0018107 UUUUGCAUGACCCUGGGAGUAGG
hsa-mir-3680 MIMAT0018106 GACUCACUCACAGGAUUGUGCA
Fig. 3. MER paralogs matched with mature miRNA sequences of miRBase. The left part shows the alignments of paralogs matched with their respective mature miRNAs (a) hsa-mir-1302, (b) hsa-mir-652, (c) hsa-mir-3680/3680* in MER53, MER91C, and MER96. The right part shows the conservation of mature miRNA between all the paralogs from the consensus MER sequences.
426
K. Ahn et al. / Gene 512 (2013) 422–428
Table 2 Predicted miRNAs of paralogs in MER families. Repeat name
Chromosome location
Strand
Length (bp)
MFE
P-value
Prediction confidence
Prediction result
MER91 MER91B MER91B MER91C MER91C MER91C MER91C MER91C MER96 MER96 MER96 MER96 MER96 MER96 MER96 MER126 MER133B MER133B MER134
chr1: 55742345–55742407 chr11: 17429558–17429620 chr6: 155299368–155299438 chr17: 62446009–62446137 chr8:124653188–124653320 chr12: 99569167–99569246 chr9: 99335525–99335655 chrX:134505672–134505737 chr3: 137118795–137118923 chr6: 72046626–72046709 chr11: 113055940–113056044 chr12: 34274960–34275083 chr12: 107790727–107790838 chr14: 67339009–67339095 chr19: 12856963–12857048 chr11: 8744807–8744912 chr2: 145620438–145620542 chr2: 144824239–144824326 chr2: 65923586–65923699
− + + − − + + + − + − + + + + − + − −
63 63 71 129 133 80 131 66 129 84 105 124 112 87 86 106 105 88 114
−18.9 −22.1 −21.2 −73.6 −69.4 −40.9 −53 −16.4 −82.1 −21.8 −41.8 −58.8 −33.5 −17.34 −30.5 −30.8 −45.8 −23.3 −41.71
0.334 0.062 0.203 0.001 0.001 0.001 0.002 0.318 0.001 0.304 0.075 0.001 0.021 0.793 0.05 0.002 0.001 0.383 0.001
76.40% 67.70% 74.60% 81.50% 86.50% 75.70% 77.60% 70.80% 79.10% 64.40% 60.90% 73.80% 56.60% 80.30% 71.20% 70.50% 70.20% 77.50% 76.30%
Pseudo miRNA precursor Pseudo miRNA precursor Pseudo miRNA precursor Real miRNA precursor Real miRNA precursor Real miRNA precursor Real miRNA precursor Pseudo miRNA precursor Real miRNA precursor Pseudo miRNA precursor Pseudo miRNA precursor Real miRNA precursor Real miRNA precursor Pseudo miRNA precursor Pseudo miRNA precursor Real miRNA precursor Pseudo miRNA precursor Pseudo miRNA precursor Real miRNA precursor
using the oligo dT primer that had a capture linker appended to it. To examine whether the mature miRNAs of the predicted miRNAs that originated from MERs existed in the human cell lines, we amplified the reverse-transcribed miRNAs through RT-PCR and visualized the PCR products on agarose gels. As shown in Fig. 4a, 2 mature miRNAs, 5p and 3p, derived from MER91C located on the human chromosome 17, were highly enriched by the 3 AGO proteins in the 3 human cell lines. However, mature miRNAs derived from the MER91C located on human chromosome 8 were weakly enriched by the AGO proteins; the 5ps were detected in all the 3 cell lines but the 3ps were not
detected in the HeLa cell line (Fig. 4b). The left arm (5p) of the mature miRNAs derived from MER96 located on human chromosome 3 was ubiquitously enriched by the 3 AGO proteins in the 3 human cell lines (Fig. 4c). The 5p of the predicted mature miRNA duplex was more enriched by AGO proteins than 3p of the miRNA. 4. Discussion The several consensus sequences of the MER transposons had almost perfect palindromic structure. We found that the paralogs of
Fig. 4. miRNAs that originated from MERs loaded onto cell line AGO proteins. (a) The left shows miRNA derived from MER91C located on chromosome 17 that was loaded onto AGO proteins of cell lines HeLa, HCT116, and HEK293. The right shows the secondary structures of miRNA derived from MER91C located on chromosome 17. miRNA derived from MER91C located on chromosome 8 (b) and miRNA derived from MER6 located on chromosome 3 (c) were loaded onto AGO proteins of cell lines HeLa, HCT116 and HEK293, respectively. The red line and blue lines indicate the predicted mature miRNA sequences in the secondary structures.
K. Ahn et al. / Gene 512 (2013) 422–428
the MER transposons also had palindromic structures and that their structures were similar to the hairpin structures of the miRNAs. In cases where the paralogs contained the core sequences of miRNAs such as seed regions that recognized their target nucleotide sequence (~ 7 bp), they could have a high impact on individual survival by producing miRNAs that cause translational repression or gene silencing. Therefore, we predicted which potential miRNAs could be derived from the MER paralogs by using the Repbase of Genetic Information Research Institute database. miRNAs that originate from transposable elements were recently identified in mammals and plants (Piriyapongsa and Jordan, 2007, 2008; Smalheiser and Torvik, 2005). A previous study demonstrated that some of the hsa-mir-548 genes originate from Made1 transposable elements that comprise nearly perfect palindromes (Piriyapongsa et al., 2007). In addition, it was reported that MER53 is the origin of the miR-1302 gene (Yuan et al., 2010). In that study, the authors examined 5839 MER53 elements and identified 36 potential paralogs of the miR-1302 gene among them. More than 100,000 copies of MER transposons exist in the haploid human genome (Jurka, 2006; Jurka et al., 1997). Interestingly, the MER families that were examined in this study had palindromic structures that were similar to the hairpin structures of miRNAs. In addition, they maintained a relatively high sequence identity among the members of each MER family. For example, the sequence identities between each individual copy and the consensus sequence of MER96 averaged 80%. Thus, we suggest that MER repeats that form palindromic structures could potentially produce miRNA. AGO proteins are involved in gene silencing mediated by RNAinduced silencing complex (RISC) by acting as the catalytic component of the RISC. AGO proteins could inhibit the expression of their targets either by using endonuclease activity or by attracting additional proteins that can affect translation, RNA stability, or chromatin structure. The majority of eukaryotic organisms possess more than one type of AGO protein and the functions of individual members of the family are often non-redundant. To identify miRNAs that originated from MERs in this study, we used human AGO1, AGO2, and AGO3 proteins, whose functions have been well studied. The major function of human AGO1 is related to heterochromatin silencing, whereas AGO2 is associated with RNAi, miRNA mediated gene silencing, and heterochromatin silencing (Hutvágner and Simard, 2008; Yang and Lai, 2010) and AGO3 is involved in transposon silencing in Drosophila (Hutvágner and Simard, 2008). We identified 3 potential miRNAs that were derived from the paralogs of the palindromic MER families and loaded onto AGO1, AGO2, and AGO3 proteins. These miRNAs originating from MERs could affect human gene expression or individual human survival through various functional pathways. Our results showed that 5p of the predicted mature miRNAs duplex was more enriched by the AGO proteins in the human cells than 3p. It was reported that the 3p miRNA strand is either degraded as merely a carrier strand or expressed abundantly as a potential functional guide miRNA. Also, well-conserved 3p miRNA* strands, particularly in the seed sequences of miRNAs, may afford potential opportunities for contributing to the regulation network (Guo and Lu, 2010). The miRBase data include substantial numbers of computationally predicted miRNA genes. In the case of the computationally predicted miRNAs were orthologous sequences, which was experimentally characterized miRNAs in another species. The repeat element-related miRNA families were found in the human, rhesus and mouse genomes with the lineage-specific expansion (Yuan et al., 2011). In addition, a number of current computational methods used for miRNA prediction do not consider TE-derived miRNAs (Bentwich et al., 2005; Li et al., 2006; Lindow and Krogh, 2005; Nam et al., 2005). Thus, considering computationally identified miRNAs, we suggest that the precise number of miRNA genes in the human genome is substantially higher than is currently believed. Although none of the MER paralogs was perfectly matched with established miRNA precursors, some of them showed relatively high sequence similarity to the
427
miRNA precursors. It could explain that after transposition of MER elements on the other genomic regions, they have influenced by various mutation events during evolutionary radiation. Therefore, they could not be perfectly matched with established miRNAs. It is possible that MER paralogs could produce novel miRNA. As shown in Fig. 4, through the immunoprecipitation of AGO proteins, we identified 3 potential miRNAs that originated from MER that were loaded on the AGO proteins. Therefore, our study contributes to an improved understanding of the relationship between miRNAs and MER transposons. Supplementary data to this article can be found online at http:// dx.doi.org/10.1016/j.gene.2012.08.028. Author contributions KA conceived and designed the experiments, and KA and J-AG performed the experiments. KA performed the computational analysis, and KA and H-SH analyzed the data. H-SK contributed the reagents, materials, and analysis tools. Finally, KA, KH, and H-SK wrote the paper. Competing interests The authors declare that no competing interests exist. Acknowledgments We thank Dr. Jungnam Lee for her useful comments during the preparation of this manuscript. This work was supported by a grant from the Next generation BioGreen 21 Program (No. PJ0081062011), Rural Development Administration, Republic of Korea. References Althaus, C.F., et al., 2012. Tailored enrichment strategy detects low abundant small noncoding RNAs in HIV-1 infected cells. Retrovirology 9, 27. Bentwich, I., et al., 2005. Identification of hundreds of conserved and nonconserved human microRNAs. Nat. Genet. 37, 766–770. Denli, A.M., Tops, B.B.J., Plasterk, R.H.A., Ketting, R.F., Hannon, G.J., 2004. Processing of primary microRNAs by the microprocessor complex. Nature 432, 231–235. Guo, L., Lu, Z., 2010. The fate of miRNA* strand through evolutionary analysis: implication for degradation as merely carrier strand or potential regulatory molecule? PLoS One 5, e11387. Hutvágner, G., Simard, M.J., 2008. Argonaute proteins: key players in RNA silencing. Nat. Rev. Mol. Cell Biol. 9, 22–32. Jurka, J., 2006. MER135: conserved mammalian repeat, probably derived from a nonautonomous DNA transposon. Repbase Rep. 6, 388. Jurka, J., Kapitonov, V.V., Klonowski, P., Walichiewicz, J., Smit, A.F., 1997. Identification of new medium reiteration frequency repeats in the genomes of Primates, Rodentia and Lagomorpha. Genetica 98, 235–247. Kincaid, R.P., Burke, J.M., Sullivan, C.S., 2012. RNA virus microRNA that mimics a B-cell oncomiR. Proc. Natl. Acad. Sci. U. S. A. 109, 3077–3082 (Epub 2012, 30). Klase, Z., et al., 2007. HIV-1 TAR element is processed by Dicer to yield a viral microRNA involved in chromatin remodeling of the viral LTR. BMC Mol. Biol. 8, 63. Krichevsky, A.M., King, K.S., Donahue, C.P., Khrapko, K., Kosik, K.S., 2003. A microRNA array reveals extensive regulation of microRNAs during brain development. RNA 9, 1274–1281. Lee, Y.S., et al., 2004. Distinct roles for Drosophila Dicer-1 and Dicer-2 in the siRNA/miRNA silencing pathways. Cell 117, 69–81. Li, S.-C., Pan, C.-Y., Lin, W.-c., 2006. Bioinformatic discovery of microRNA precursors from human ESTs and introns. BMC Genomics 7, 164. Lindow, M., Krogh, A., 2005. Computational evidence for hundreds of non-conserved plant microRNAs. BMC Genomics 6, 119. Nam, J.-W., et al., 2005. Human microRNA prediction through a probabilistic co-learning model of sequence and structure. Nucleic Acids Res. 33, 3570–3581. Okamura, K., Ishizuka, A., Siomi, H., Siomi, M.C., 2004. Distinct roles for Argonaute proteins in small RNA-directed RNA cleavage pathways. Genes Dev. 18. Okamura, K., Liu, N., Lai, E.C., 2009. Distinct mechanisms for microRNA strand selection by Drosophila Argonautes. Cell 36, 431–444. Piriyapongsa, J., Jordan, I.K., 2007. A family of human microRNA genes from miniature inverted-repeat transposable elements. PLoS One 2, e203. Piriyapongsa, J., Jordan, I.K., 2008. Dual coding of siRNAs and miRNAs by plant transposable elements. RNA 14, 814–821. Piriyapongsa, J., Mariño-Ramírez, L., Jordan, I.K., 2007. Origin and evolution of human microRNAs from transposable elements. Genetics 176, 1323–1337. Shalgi, R., Pilpel, Y., Oren, M., 2010. Repression of transposable-elements—a microRNA anti-cancer defense mechanism? Trends Genet. 26, 253–259.
428
K. Ahn et al. / Gene 512 (2013) 422–428
Smalheiser, N.R., Torvik, V.I., 2005. Mammalian microRNAs derived from genomic repeats. Trends Genet. 21, 322–326. Xu, Q., et al., 2010. Discovery and comparative profiling of microRNAs in a sweet orange red-flesh mutant and its wild type. BMC Genomics 11, 246. Yang, J.-S., Lai, E.C., 2010. Dicer-independent, Ago2-mediated microRNA biogenesis in vertebrates. Cell Cycle 9, 4455–4460.
Yuan, Z., et al., 2010. Origin and evolution of a placental-specific microRNA family in the human genome. BMC Evol. Biol. 10, 346. Yuan, Z., Sun, X., Liu, H., Xie, J., 2011. MicroRNA genes derived from repetitive elements and expanded by segmental duplication events in mammalian genomes. PLoS One 6, e17666. Zhang, R., Peng, Y., Wang, W., Su, B., 2007. Rapid evolution of an X-linked microRNA cluster in primates. Genome Res. 17, 612–617.
Contents lists available at SciVerse ScienceDirect
Gene journal homepage: www.elsevier.com/locate/gene
The novel MER transposon-derived miRNAs in human genome Kung Ahn a, Jeong-An Gim a, Hong-Seok Ha b, Kyudong Han c, Heui-Soo Kim a,⁎ a b c
Department of Biological Sciences, College of Natural Sciences, Pusan National University, Busan 609‐735, Republic of Korea Department of Genetics, Human Genetics Institute of New Jersey, Rutgers, the State University of New Jersey, 145 Bevier Rd, Piscataway, NJ 08854, USA Department of Nanobiomedical Science & WCU Research Center, Dankook University, Cheonan 330‐714, Republic of Korea
a r t i c l e
i n f o
Article history: Accepted 17 August 2012 Available online 25 August 2012 Keywords: MER transposon MicroRNA MER derived miRNA Palindromic structure AGO proteins
a b s t r a c t MicroRNAs (miRNAs) are small RNA molecules (~20–30 nucleotides) that generally act in gene silencing and translational repression through the RNA interference pathway. They generally originate from intergenic genomic regions, but some are found in genomic regions that have been characterized such as introns, exons, and transposable elements (TE). To identify the miRNAs that are derived from palindromic MERs, we analyzed MER paralogs in human genome. The structures of the palindromic MERs were similar to the hairpin structure of miRNA in humans. Three miRNAs derived from MER96 located on chromosome 3, and MER91C paralogs located on chromosome 8 and chromosome 17 were identified in HeLa, HCT116, and HEK293 cell lines. The interactions between these MER-derived miRNAs and AGO1, AGO2, and AGO3 proteins were validated by immunoprecipitation assays. The data suggest that miRNAs derived from transposable elements could widely affect various target genes in the human genome. Crown Copyright © 2012 Published by Elsevier B.V. All rights reserved.
1. Introduction MicroRNAs (miRNAs) are single-stranded RNAs that are ~ 20–30 nucleotides long in their mature forms. They are generally responsible for gene silencing by binding to target messenger RNA (mRNA) transcripts. miRNAs are produced in several steps: they are transcribed as long primary transcripts, which are referred to as primary miRNA (pri-miRNA) transcripts. Next, they are processed by 2 well-known proteins, namely, Drosha and Dicer. Drosha cleaves the pri-miRNA down to ~70 nucleotides in length with a short double-stranded hairpin structure (precursor miRNA) in the nucleus. The precursor miRNA is exported to the cytoplasm and cleaved by Dicer to generate the mature miRNA form (~20–30 nucleotides). The mature miRNA comprises an active RNA-induced silencing complex (RISC) along with argonaute protein (AGO), which binds to the mature miRNA and mediates the interaction between the miRNA and its target mRNA (Denli et al., 2004; Lee et al., 2004; Okamura et al., 2004, 2009; Yang and Lai, 2010). miRNAs have been highlighted as key regulators in numerous types of cells and organismal processes (Krichevsky et al., 2003; Piriyapongsa and Jordan, 2008; Xu et al., 2010). In addition, it was reported that alterations in miRNA expression are associated with the initiation, progression, and metastasis of human tumors (Shalgi et al., 2010). Although most miRNAs originate from intergenic genomic sequences, some of them originate from sequences that have been Abbreviations: TEs, transposable elements; MER, MEdium Reiteration frequency; AGO proteins, argonaute proteins; miRNA, microRNA. ⁎ Corresponding author. Tel.: +82 51 510 2259; fax: +82 51 581 2962. E-mail address: [email protected] (H.-S. Kim).
characterized such as genes and transposable elements (Piriyapongsa and Jordan, 2007, 2008; Smalheiser and Torvik, 2005; Zhang et al., 2007). Furthermore, retroviral transposable elements (TEs) such as human immunodeficiency virus and bovine leukemia virus encode viral miRNAs (Althaus et al., 2012; Kincaid et al., 2012; Klase et al., 2007). In particular, because of the wide distribution of the TEs, the miRNAs derived from them could harmfully affect host cells and their genomic stability. If a miRNA that is derived from TEs with a high copy number is involved in gene silencing, it would have a huge impact on an individual's survival. However, it is still unclear how miRNA is produced from a TE. In the case of Made 1, it has been reported that its palindromic structure (the origin of hsa-mir-548) is produced by either orientation (+/−) and that it can form itself into a hairpin structure. Then, it is inserted into transcriptionally active genomic regions, and it takes the form of pri-miRNA structures that can be processed by the RNA interference enzymatic machinery (Piriyapongsa et al., 2007). MEdium Reiteration frequency (MER) refers to interspersed repeats in the genomes of primates, rodentia, and lagomorpha. Age estimations place the origin of most MER repeats at the time of decline in Mammalian-wide Interspersed Repeats (MIR) retroposition and before the origin of the Alu family. Intriguingly, short palindromic sequences contain a hairpin structure, which is very similar to the structure of miRNA. The miR-1302 gene family was derived from the MER53 transposon that retains its palindromic structure (Yuan et al., 2010). In this study, we computationally identified miRNAs that originated from MER elements and experimentally confirmed them. To identify the miRNA, we compared palindromic MERs with miRNA precursors and mature miRNAs from miRBase (http://www.mirbase.org/). Then, we carried out miRNA
0378-1119/$ – see front matter. Crown Copyright © 2012 Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gene.2012.08.028
K. Ahn et al. / Gene 512 (2013) 422–428
prediction to identify potential miRNA derived from the paralogs of palindromic MER consensus sequences. Finally, we experimentally verified the existence of MERs that originated from miRNAs. 2. Materials and methods 2.1. Computational analysis MER consensus sequences containing palindromic sequences were obtained from the Repbase browser of the Genetic Information Research Institute database. The genomic locations and sequences of repetitive elements, including MERs, were extracted from the human genome by using UCSC Genome Browser hg18 (http://genome.cse. ucsc.edu/) and they were analyzed using RepeatMasker 3.27 (http:// www.repeatmasker.org/). The sequences and coordinates of human pre-miRNAs and mature miRNAs were downloaded from miRBase v13.0 (http://www.mirbase.org/) and mapped to the human genome (hg18). Using the UCSC Table Browser, the genomic locations of the miRNAs and the repetitive elements from the human genome were compared. To identify the potential MER-derived miRNA genes, we developed a three-step operational scheme. We blasted the human genome with the palindromic consensus sequences of MER families using the BLAT program. Then, we carried out predictions of the MiRNA that originated from MER on MER paralogs by using miPred [http:// www.bioinf.seu.edu.cn/miRNA/] and miRNAFinder [http://bioinfo3. noble.org/mirna/]. miPred is a search tool for which all pre-miRNAs have a characteristic stem-loop hairpin structure. Therefore, the hairpin structures provide key clues to the ab initio prediction of pre-miRNAs. In addition, miRNAFinder can predict potential intronic miRNA in intron regions of the expressed genes (ESTs/cDNAs), find possible miRNA in genomic sequences, or predict if the input small RNA is mature miRNA. A mature miRNA is processed from the left or right arm of a potential precursor miRNA sequence. We examined the secondary structures of single stranded RNA sequences by using the RNAfold web server, which has limits of 7500 nt for partition function calculations and 10,000 nt for minimum free energy. The TargetScan 6.2 [http:// www.targetscan.org/] was used to identify the potential target sites of miRNAs. To determine the functional categories to which the target genes belonged, we used GOmir [http://www.bioacademy. gr/bioinformatics/projects/GOmir/] and displayed the GO categories of the target genes. 2.2. Cell cultures and lysates preparation HEK293, HCT116, and HeLa cells were seeded in Dulbecco's modified Eagle's medium that was supplemented with 10% (v/v) heatinactivated fetal bovine serum and 1% (v/v) antibiotics–antimycotic solution, and then incubated at 37 °C in a 5% (v/v) CO2 incubator. Cells (5× 106) from each cell line were washed with DPBS and collected by trypsinization. The cells were incubated in 1 ml of lysis buffer (20 mM Tris, pH 7.4, 2.5 mM MgCl2, 200 mM NaCl, 0.05% NP40) for 10 min on ice, and the cell lysates were cleared by centrifugation at 15,000 ×g for 20 min at 4 °C. 2.3. Immunoprecipitation, poly (A) tailing, and reverse-transcription To extract small RNAs from the human cells, AGO proteins AGO1, AGO2, and AGO3 were immunoprecipitated. AGO1, 2, 3-Antibody Beads were purchased from Wako Pure Chemical Industry, Ltd. Fifty microliters of Anti-AGO1, AGO2, and AGO3 Antibody Bead solutions and human cell lysates were incubated at 4 °C for 2 h. After the antigen–antibody reactions, the incubated mixtures were washed twice, and to each was added 50 μl of the elution solution. The eluted microRNAs were purified using phenol:chloroform:isoamyl alcohol mixture (25:24:1) and visualized on a polyacrylamide gel (data not shown). The isolated miRNAs were subjected to poly (A) tailing;
423
4 μg of each was polyadenylated by 5 U of poly (A) polymerase (Ambion). The reactions were incubated at 37 °C for 30 min. After incubation, poly (A) tailed miRNAs were recovered by phenol/chloroform extraction and ethanol precipitation. Reverse transcription of the poly (A) tailed miRNAs was performed using M-MLV reverse transcriptase (RT, Invitrogen) following the manufacturer's instructions. A solution (9.5 μl) including 4 μg of the poly (A) tailed miRNAs and 1 μl of 10 mM RT linker sequence (CTGTGAATGCTGCGACTACGA-18 dTs) was incubated at 65 °C for 5 min to remove any RNA secondary structures. After incubation, 5 μl of 10 mM dNTP mix, 4 μl of 5× RT buffer (including 250 mM Tris–HCl, 375 mM KCl, 15 mM MgCl2, and 50 mM DTT), 0.5 μl of RNase inhibitor and 1 μl of M-MLV reverse transcriptase were added. The reaction was then incubated at 42 °C for 90 min. After incubation, the reaction was allowed to proceed at 4 °C for 60 min. 2.4. RT-PCR MER91C located on chromosome 8 was amplified using the sense primers 5′-TGAAGGGGTTACAATTGGCAT-3′ (Chr.8, MER91C derived mature miRNA 5P) and 5′-TGCGGATGGCACCTCCTGAG-3′ (Chr.8, MER91C derived mature miRNA 3P). MER91C located on chromosome 17 was amplified using the sense primers 5′-TACAACTGGAAGGATGTTCAT-3′ (Chr.17, MER91C derived mature miRNA 5P) and 5′-GTGACATCCCTTGA GTTGTGC-3′ (Chr.17, MER91C derived mature miRNA 3P). MER96 located on chromosome 3 was amplified using the sense primers 5′-TGCAGCATTT AAGGAAGCACC-3′ (Chr.3, MER96 derived mature miRNA 5P) and 5′-GAG TGCCTCCTTAAATGTTTT-3′ (Chr.3, MER96 derived mature miRNA 3P). All the amplifications were carried out using the antisense primer (Universal primer) 5′-CTGTGAATGCTGCGACTACGAT-3′. RT-PCRs were carried out for 30 cycles at 94 °C for 40 s, 56 °C for 40 s, and 72 °C for 7 min. 3. Results 3.1. Comparative analysis of palindromic MERs and miRNAs from miRBase Eight palindromic MER consensus sequences were found (Fig. 1). Among them, the consensus sequence of MER91C matched the precursor of hsa-miRNA-652 in the miRBase and the E-value of the match was 4.00E− 08. Through a UCSC blat search, 80 paralogs aligned with the consensus sequence of MER91C were identified in the human genome (hg18). In addition, the consensus sequence of MER96 was matched with a hsa-miR-3680 and the E-value was 5.00E− 18. Ninety-seven paralogs of the MER96 consensus sequence were found in the human genome (Table 1). The palindromic consensus MER sequences were matched with the precursor miRNA sequence of the miRBase. As shown in Fig. 2, (a) MER53 was matched with precursor miR-1302-1 and (b) MER91C was matched with precursor miR-652, while (c) MER96 was matched with precursor miR-3680/3680*. We aligned the palindromic consensus sequences of the MER elements with the precursor miRNA sequences in the miRBase (Fig. 2). The consensus sequence of MER53 had 83.6% similarity with the precursor of has-mir-1302-1. In addition, the consensus sequence of MER91C had 75.8% similarity with the precursor of hsa-mir-652, and the consensus sequence of MER96 had 85.1% similarity with the precursor of has-mir-3680. The target genes of hsa-mir-1302 and hsa-mir-652 seemed to be intricately involved in cellular processes, intracellular membrane, catalytic activity, and binding by gene ontology (GO) cluster analysis (Fig. S1). However, the consensus sequences of other MERs, including MER123, MER124, MER126, MER133B, and MER134 did not exactly match any of the precursor miRNAs in the miRBase. Although there were no matches with the precursor miRNAs, we could not exclude the possibility that
424
K. Ahn et al. / Gene 512 (2013) 422–428
Fig. 1. Hairpin structures of MER DNA transposon consensus sequences. Eight palindromic consensus MER secondary structures of single-stranded RNA sequences were predicted by RNAfold. The MER consensus sequences containing palindromic sequences were obtained from the Repbase browser of the Genetic Information Research Institute database. The consensus sequence sizes are presented as MER53—189 bp, MER91C—140 bp, MER96—175 bp, MER123—210 bp, MER124—397 bp, MER126—308 bp, MER133B—140 bp, and MER134—216 bp, respectively.
they were potential miRNAs. We examined the conservation of respective mature miRNAs among paralogs of MER53, MER91C, and MER96. As shown in Fig. 3, each mature miRNA sequence exhibited high conservation among the MER paralogs. 3.2. Potential miRNAs derived from the palindromic sequences of MER paralogs To detect the miRNA derived from MER paralogs, we predicted potential miRNAs in the palindromic sequences of MER paralogs by using the miPred web server tool (URL: http://www.bioinf.seu.edu. cn/miRNA/). The miPred program can be used to predict either pseudo or real miRNA precursors by using the structural characteristics of the MER paralogs. Among the MER91C paralogs, 5 paralogs contained miRNAs: 1 contained pseudo miRNA and 4 contained real miRNA precursors. Seven among the MER96 paralogs contained miRNAs, 2 of which were real miRNA precursors. In addition, MER126 located on chromosome 11 and MER134 located on chromosome 2 were determined to be real miRNA precursors (Table 2). The MER91C paralogs containing the real miRNA precursors had high sequence homology with the consensus sequence of MER91C. The paralogs located on chromosomes 8, 9, 12, and 17 showed
sequence identities of 86.92%, 85.62%, 75.18%, and 84.31% with the consensus sequence, respectively. The MER96 paralogs that contained the real miRNA precursors were located on chromosomes 3 and 12, and they showed sequence identities of 69.23% and 90.51% with the consensus sequence of MER96, respectively. The MER134 and MER126 paralogs containing the real miRNA precursors showed sequence homologies of 71.29% and 71.56% with their consensus sequences, respectively. We analyzed the secondary structures of the MER paralogs that were predicted to have pseudo miRNA or real miRNA precursors using the RNAfold web server (http://rna.tbi.univie.ac.at/cgi-bin/ RNAfold.cgi) (Fig. S2). Through the analysis, we found that the MER paralogs had palindromic structures. We further investigated the secondary structures of the MER paralogs by comparing them with the secondary structures of their consensus sequences. When miRNAs originating from MER91C were compared with the consensus sequence of MER91C, they were found to contain palindromic sequences that were similar to the sequences responsible for the central regions of their stem-loop structures. The MER96 paralog located on chromosome 3 had a deletion of 50 bp in the central region of the stem-loop structure of its consensus sequence. The stem-loop region of MER126 located on chromosome 11 consisted of the stem-loop region of its consensus
Table 1 Comparison between miRBase and palindromic MER consensus sequences. Repeat name
Repeat family
Repeat length (bp)
Structure
Precursor sequence (mirbase sequence)
E-value
n-Number of paralogs (consensus sequence matched paralogs)
MER53 MER91C MER96 MER123 MER124 MER126 MER133B MER134
DNA DNA DNA DNA DNA DNA DNA DNA
189 140 175 210 397 308 140 213
Palindrome Palindrome Palindrome Palindrome Palindrome Palindrome Palindrome Palindrome
mir-1302-1 mir-652 mir-3680 Δmir-3606 Δmir-1302-4/Δmir-1302-6/Δmir-1302-8/Δmir-592/Δmir-3012a Δmir-383/Δmir-3934 or new miRNA Δmir-1279/Δmir-1323 Δmir-129-1
2.00E−30 4.00E−08 1.00E−17 E>1 E>1 E>1 E>1 E>1
n = 178 n = 80 n = 97 n=6 n = 25 n = 30 n=4 n = 11
K. Ahn et al. / Gene 512 (2013) 422–428
425
Fig. 2. Palindromic consensus MER sequences matched with the precursor miRNA sequences of miRBase. (a) MER53 was matched with precursor miR-1302-1 and (b) MER91C was matched with the precursor miR-652. (c) MER96 was matched with precursor miR-3680/3680*. The dotted line box indicates mature miRNA sequences present in the MER transposon, and the blank box indicates mature miRNA sequences. The values in the lower panel indicate the minimum free energy.
sequence. Taken together, this suggests that the palindromic structure of the miRNA that originated from MER was derived from a specific region of the MER consensus sequence that is responsible for the formation of its secondary stem-loop structure. However, the stem-loop region of the MER 134 paralog did not consist of its consensus sequence although the stem-loop of the MER 134 paralog was very similar to the stem-loop region of the consensus sequence (Fig. S3). In order to detect mature miRNAs derived from the predicted real miRNA precursors, we scanned them using miRNAFinder. By doing this, we identified duplex mature miRNA sequences in the precursors (Fig. S4).
3.3. Identification of MER derived miRNAs loaded onto AGO proteins in human cells In order to identify mature miRNA derived from MER transposons, we first immunoprecipitated 3 AGO proteins, AGO1, AGO2, and AGO3, from human cell lines HeLa, HEK293, and HCT116 because these proteins are miRNA processors. Then, the miRNAs loaded on the AGO proteins were processed for RT-PCR: they were isolated from the proteins, a poly A tail was added to the 3′ end of each miRNA, and then the miRNAs with poly A tails were reverse transcribed
a) MER 53 hsa-mir-1302 MIMAT0005890 UUGGGACAUACUUAUGCUAAA
b) MER 91C hsa-mir-652 MIMAT0003322 AAUGGCGCCACUAGGGUUGUG
c) MER 96 hsa-mir-3680* MIMAT0018107 UUUUGCAUGACCCUGGGAGUAGG
hsa-mir-3680 MIMAT0018106 GACUCACUCACAGGAUUGUGCA
Fig. 3. MER paralogs matched with mature miRNA sequences of miRBase. The left part shows the alignments of paralogs matched with their respective mature miRNAs (a) hsa-mir-1302, (b) hsa-mir-652, (c) hsa-mir-3680/3680* in MER53, MER91C, and MER96. The right part shows the conservation of mature miRNA between all the paralogs from the consensus MER sequences.
426
K. Ahn et al. / Gene 512 (2013) 422–428
Table 2 Predicted miRNAs of paralogs in MER families. Repeat name
Chromosome location
Strand
Length (bp)
MFE
P-value
Prediction confidence
Prediction result
MER91 MER91B MER91B MER91C MER91C MER91C MER91C MER91C MER96 MER96 MER96 MER96 MER96 MER96 MER96 MER126 MER133B MER133B MER134
chr1: 55742345–55742407 chr11: 17429558–17429620 chr6: 155299368–155299438 chr17: 62446009–62446137 chr8:124653188–124653320 chr12: 99569167–99569246 chr9: 99335525–99335655 chrX:134505672–134505737 chr3: 137118795–137118923 chr6: 72046626–72046709 chr11: 113055940–113056044 chr12: 34274960–34275083 chr12: 107790727–107790838 chr14: 67339009–67339095 chr19: 12856963–12857048 chr11: 8744807–8744912 chr2: 145620438–145620542 chr2: 144824239–144824326 chr2: 65923586–65923699
− + + − − + + + − + − + + + + − + − −
63 63 71 129 133 80 131 66 129 84 105 124 112 87 86 106 105 88 114
−18.9 −22.1 −21.2 −73.6 −69.4 −40.9 −53 −16.4 −82.1 −21.8 −41.8 −58.8 −33.5 −17.34 −30.5 −30.8 −45.8 −23.3 −41.71
0.334 0.062 0.203 0.001 0.001 0.001 0.002 0.318 0.001 0.304 0.075 0.001 0.021 0.793 0.05 0.002 0.001 0.383 0.001
76.40% 67.70% 74.60% 81.50% 86.50% 75.70% 77.60% 70.80% 79.10% 64.40% 60.90% 73.80% 56.60% 80.30% 71.20% 70.50% 70.20% 77.50% 76.30%
Pseudo miRNA precursor Pseudo miRNA precursor Pseudo miRNA precursor Real miRNA precursor Real miRNA precursor Real miRNA precursor Real miRNA precursor Pseudo miRNA precursor Real miRNA precursor Pseudo miRNA precursor Pseudo miRNA precursor Real miRNA precursor Real miRNA precursor Pseudo miRNA precursor Pseudo miRNA precursor Real miRNA precursor Pseudo miRNA precursor Pseudo miRNA precursor Real miRNA precursor
using the oligo dT primer that had a capture linker appended to it. To examine whether the mature miRNAs of the predicted miRNAs that originated from MERs existed in the human cell lines, we amplified the reverse-transcribed miRNAs through RT-PCR and visualized the PCR products on agarose gels. As shown in Fig. 4a, 2 mature miRNAs, 5p and 3p, derived from MER91C located on the human chromosome 17, were highly enriched by the 3 AGO proteins in the 3 human cell lines. However, mature miRNAs derived from the MER91C located on human chromosome 8 were weakly enriched by the AGO proteins; the 5ps were detected in all the 3 cell lines but the 3ps were not
detected in the HeLa cell line (Fig. 4b). The left arm (5p) of the mature miRNAs derived from MER96 located on human chromosome 3 was ubiquitously enriched by the 3 AGO proteins in the 3 human cell lines (Fig. 4c). The 5p of the predicted mature miRNA duplex was more enriched by AGO proteins than 3p of the miRNA. 4. Discussion The several consensus sequences of the MER transposons had almost perfect palindromic structure. We found that the paralogs of
Fig. 4. miRNAs that originated from MERs loaded onto cell line AGO proteins. (a) The left shows miRNA derived from MER91C located on chromosome 17 that was loaded onto AGO proteins of cell lines HeLa, HCT116, and HEK293. The right shows the secondary structures of miRNA derived from MER91C located on chromosome 17. miRNA derived from MER91C located on chromosome 8 (b) and miRNA derived from MER6 located on chromosome 3 (c) were loaded onto AGO proteins of cell lines HeLa, HCT116 and HEK293, respectively. The red line and blue lines indicate the predicted mature miRNA sequences in the secondary structures.
K. Ahn et al. / Gene 512 (2013) 422–428
the MER transposons also had palindromic structures and that their structures were similar to the hairpin structures of the miRNAs. In cases where the paralogs contained the core sequences of miRNAs such as seed regions that recognized their target nucleotide sequence (~ 7 bp), they could have a high impact on individual survival by producing miRNAs that cause translational repression or gene silencing. Therefore, we predicted which potential miRNAs could be derived from the MER paralogs by using the Repbase of Genetic Information Research Institute database. miRNAs that originate from transposable elements were recently identified in mammals and plants (Piriyapongsa and Jordan, 2007, 2008; Smalheiser and Torvik, 2005). A previous study demonstrated that some of the hsa-mir-548 genes originate from Made1 transposable elements that comprise nearly perfect palindromes (Piriyapongsa et al., 2007). In addition, it was reported that MER53 is the origin of the miR-1302 gene (Yuan et al., 2010). In that study, the authors examined 5839 MER53 elements and identified 36 potential paralogs of the miR-1302 gene among them. More than 100,000 copies of MER transposons exist in the haploid human genome (Jurka, 2006; Jurka et al., 1997). Interestingly, the MER families that were examined in this study had palindromic structures that were similar to the hairpin structures of miRNAs. In addition, they maintained a relatively high sequence identity among the members of each MER family. For example, the sequence identities between each individual copy and the consensus sequence of MER96 averaged 80%. Thus, we suggest that MER repeats that form palindromic structures could potentially produce miRNA. AGO proteins are involved in gene silencing mediated by RNAinduced silencing complex (RISC) by acting as the catalytic component of the RISC. AGO proteins could inhibit the expression of their targets either by using endonuclease activity or by attracting additional proteins that can affect translation, RNA stability, or chromatin structure. The majority of eukaryotic organisms possess more than one type of AGO protein and the functions of individual members of the family are often non-redundant. To identify miRNAs that originated from MERs in this study, we used human AGO1, AGO2, and AGO3 proteins, whose functions have been well studied. The major function of human AGO1 is related to heterochromatin silencing, whereas AGO2 is associated with RNAi, miRNA mediated gene silencing, and heterochromatin silencing (Hutvágner and Simard, 2008; Yang and Lai, 2010) and AGO3 is involved in transposon silencing in Drosophila (Hutvágner and Simard, 2008). We identified 3 potential miRNAs that were derived from the paralogs of the palindromic MER families and loaded onto AGO1, AGO2, and AGO3 proteins. These miRNAs originating from MERs could affect human gene expression or individual human survival through various functional pathways. Our results showed that 5p of the predicted mature miRNAs duplex was more enriched by the AGO proteins in the human cells than 3p. It was reported that the 3p miRNA strand is either degraded as merely a carrier strand or expressed abundantly as a potential functional guide miRNA. Also, well-conserved 3p miRNA* strands, particularly in the seed sequences of miRNAs, may afford potential opportunities for contributing to the regulation network (Guo and Lu, 2010). The miRBase data include substantial numbers of computationally predicted miRNA genes. In the case of the computationally predicted miRNAs were orthologous sequences, which was experimentally characterized miRNAs in another species. The repeat element-related miRNA families were found in the human, rhesus and mouse genomes with the lineage-specific expansion (Yuan et al., 2011). In addition, a number of current computational methods used for miRNA prediction do not consider TE-derived miRNAs (Bentwich et al., 2005; Li et al., 2006; Lindow and Krogh, 2005; Nam et al., 2005). Thus, considering computationally identified miRNAs, we suggest that the precise number of miRNA genes in the human genome is substantially higher than is currently believed. Although none of the MER paralogs was perfectly matched with established miRNA precursors, some of them showed relatively high sequence similarity to the
427
miRNA precursors. It could explain that after transposition of MER elements on the other genomic regions, they have influenced by various mutation events during evolutionary radiation. Therefore, they could not be perfectly matched with established miRNAs. It is possible that MER paralogs could produce novel miRNA. As shown in Fig. 4, through the immunoprecipitation of AGO proteins, we identified 3 potential miRNAs that originated from MER that were loaded on the AGO proteins. Therefore, our study contributes to an improved understanding of the relationship between miRNAs and MER transposons. Supplementary data to this article can be found online at http:// dx.doi.org/10.1016/j.gene.2012.08.028. Author contributions KA conceived and designed the experiments, and KA and J-AG performed the experiments. KA performed the computational analysis, and KA and H-SH analyzed the data. H-SK contributed the reagents, materials, and analysis tools. Finally, KA, KH, and H-SK wrote the paper. Competing interests The authors declare that no competing interests exist. Acknowledgments We thank Dr. Jungnam Lee for her useful comments during the preparation of this manuscript. This work was supported by a grant from the Next generation BioGreen 21 Program (No. PJ0081062011), Rural Development Administration, Republic of Korea. References Althaus, C.F., et al., 2012. Tailored enrichment strategy detects low abundant small noncoding RNAs in HIV-1 infected cells. Retrovirology 9, 27. Bentwich, I., et al., 2005. Identification of hundreds of conserved and nonconserved human microRNAs. Nat. Genet. 37, 766–770. Denli, A.M., Tops, B.B.J., Plasterk, R.H.A., Ketting, R.F., Hannon, G.J., 2004. Processing of primary microRNAs by the microprocessor complex. Nature 432, 231–235. Guo, L., Lu, Z., 2010. The fate of miRNA* strand through evolutionary analysis: implication for degradation as merely carrier strand or potential regulatory molecule? PLoS One 5, e11387. Hutvágner, G., Simard, M.J., 2008. Argonaute proteins: key players in RNA silencing. Nat. Rev. Mol. Cell Biol. 9, 22–32. Jurka, J., 2006. MER135: conserved mammalian repeat, probably derived from a nonautonomous DNA transposon. Repbase Rep. 6, 388. Jurka, J., Kapitonov, V.V., Klonowski, P., Walichiewicz, J., Smit, A.F., 1997. Identification of new medium reiteration frequency repeats in the genomes of Primates, Rodentia and Lagomorpha. Genetica 98, 235–247. Kincaid, R.P., Burke, J.M., Sullivan, C.S., 2012. RNA virus microRNA that mimics a B-cell oncomiR. Proc. Natl. Acad. Sci. U. S. A. 109, 3077–3082 (Epub 2012, 30). Klase, Z., et al., 2007. HIV-1 TAR element is processed by Dicer to yield a viral microRNA involved in chromatin remodeling of the viral LTR. BMC Mol. Biol. 8, 63. Krichevsky, A.M., King, K.S., Donahue, C.P., Khrapko, K., Kosik, K.S., 2003. A microRNA array reveals extensive regulation of microRNAs during brain development. RNA 9, 1274–1281. Lee, Y.S., et al., 2004. Distinct roles for Drosophila Dicer-1 and Dicer-2 in the siRNA/miRNA silencing pathways. Cell 117, 69–81. Li, S.-C., Pan, C.-Y., Lin, W.-c., 2006. Bioinformatic discovery of microRNA precursors from human ESTs and introns. BMC Genomics 7, 164. Lindow, M., Krogh, A., 2005. Computational evidence for hundreds of non-conserved plant microRNAs. BMC Genomics 6, 119. Nam, J.-W., et al., 2005. Human microRNA prediction through a probabilistic co-learning model of sequence and structure. Nucleic Acids Res. 33, 3570–3581. Okamura, K., Ishizuka, A., Siomi, H., Siomi, M.C., 2004. Distinct roles for Argonaute proteins in small RNA-directed RNA cleavage pathways. Genes Dev. 18. Okamura, K., Liu, N., Lai, E.C., 2009. Distinct mechanisms for microRNA strand selection by Drosophila Argonautes. Cell 36, 431–444. Piriyapongsa, J., Jordan, I.K., 2007. A family of human microRNA genes from miniature inverted-repeat transposable elements. PLoS One 2, e203. Piriyapongsa, J., Jordan, I.K., 2008. Dual coding of siRNAs and miRNAs by plant transposable elements. RNA 14, 814–821. Piriyapongsa, J., Mariño-Ramírez, L., Jordan, I.K., 2007. Origin and evolution of human microRNAs from transposable elements. Genetics 176, 1323–1337. Shalgi, R., Pilpel, Y., Oren, M., 2010. Repression of transposable-elements—a microRNA anti-cancer defense mechanism? Trends Genet. 26, 253–259.
428
K. Ahn et al. / Gene 512 (2013) 422–428
Smalheiser, N.R., Torvik, V.I., 2005. Mammalian microRNAs derived from genomic repeats. Trends Genet. 21, 322–326. Xu, Q., et al., 2010. Discovery and comparative profiling of microRNAs in a sweet orange red-flesh mutant and its wild type. BMC Genomics 11, 246. Yang, J.-S., Lai, E.C., 2010. Dicer-independent, Ago2-mediated microRNA biogenesis in vertebrates. Cell Cycle 9, 4455–4460.
Yuan, Z., et al., 2010. Origin and evolution of a placental-specific microRNA family in the human genome. BMC Evol. Biol. 10, 346. Yuan, Z., Sun, X., Liu, H., Xie, J., 2011. MicroRNA genes derived from repetitive elements and expanded by segmental duplication events in mammalian genomes. PLoS One 6, e17666. Zhang, R., Peng, Y., Wang, W., Su, B., 2007. Rapid evolution of an X-linked microRNA cluster in primates. Genome Res. 17, 612–617.