We developed an intronic microRNA biogenesis reporter assay to be used in cells in vitro with four plasmids: one with intronic miRNA, one with the target, one to overexpress a regulatory protein, and one for Renilla luciferase. The miRNA was processed and could control luciferase expression by binding to the target sequence.
MicroRNAs (miRNAs) are short RNA molecules that are widespread in eukaryotes. Most miRNAs are transcribed from introns, and their maturation involves different RNA-binding proteins in the nucleus. Mature miRNAs frequently mediate gene silencing, and this has become an important tool for comprehending post-transcriptional events. Besides that, it can be explored as a promising methodology for gene therapies. However, there is currently a lack of direct methods for assessing miRNA expression in mammalian cell cultures. Here, we describe an efficient and simple method that aids in determining miRNA biogenesis and maturation through confirmation of its interaction with target sequences. Also, this system allows the separation of exogenous miRNA maturation from its endogenous activity using a doxycycline-inducible promoter capable of controlling primary miRNA (pri-miRNA) transcription with high efficiency and low cost. This tool also allows modulation with RNA-binding proteins in a separate plasmid. In addition to its use with a variety of different miRNAs and their respective targets, it can be adapted to different cell lines, provided these are amenable to transfection.
Precursor mRNA splicing is an important process for gene expression regulation in eukaryotes1. The removal of introns and the union of exons in mature RNA is catalyzed by the spliceosome, a 2 megadalton ribonucleoprotein complex composed of 5 snRNAs (U1, U2, U4, U5, and U6) along with more than 100 proteins2,3. The splicing reaction occurs co-transcriptionally, and the spliceosome is assembled at each new intron guided by the recognition of conserved splice sites at exon-intron boundaries and within the intron4. Different introns might have different splicing rates despite the remarkable conservation of the spliceosome complex and its components. In addition to the differences in splice site conservation, regulatory sequences distributed on introns and exons can guide RNA-binding proteins (RBP) and stimulate or repress splicing5,6. HuR is a ubiquitously expressed RBP and is an important factor to control mRNA stability7. Previous results from our group showed that HuR can bind to introns containing miRNAs, indicating this protein might be an important factor to facilitate miRNA processing and maturation, also leading to the generation of alternative splicing isoforms6,8,9.
Many microRNAs (miRNAs) are coded from intronic sequences. Whereas some are part of the intron, others are known as “mirtrons” and are formed by the entire intron10,11. miRNAs are short non-coding RNAs, ranging from 18 to 24 nucleotides in length12. Their mature sequence shows partial or total complementarity with target sequences in mRNAs, therefore affecting translation and/or mRNA decay rates. The combinations of miRNAs and targets drive the cell to different outcomes. Several miRNAs can drive cells to pro- or anti-tumoral phenotypes13. Oncogenic miRNAs usually target mRNAs that trigger a suppressive characteristic, leading to increased cellular proliferation, migration, and invasion14. On the other hand, tumor-suppressive miRNAs might target oncogenic mRNAs or mRNAs related to increased cell proliferation.
The processing and maturation of miRNAs are also dependent on their origin. Most intronic miRNAs are processed with the participation of the microprocessor, formed by the ribonuclease Drosha and protein co-factors12. Mirtrons are processed with the activity of the spliceosome independently of Drosha15. Considering the high frequency of miRNAs found within introns, we hypothesized that RNA-binding proteins involved with splicing could also facilitate the processing and maturation of these miRNAs. Notably, the RBP hnRNP A2/B1 has already been associated with the microprocessor and miRNA biogenesis16.
We have previously reported that several RNA-binding proteins, such as hnRNPs and HuR, are associated with intronic miRNAs by mass spectrometry17. HuR’s (ELAVL1) association with miRNAs from the miR-17-92 intronic cluster was confirmed using immunoprecipitation and in silico analysis9. miR-17-92 is an intronic miRNA cluster composed of six miRNAs with increased expression in different cancers18,19. This cluster is also known as “oncomiR-1” and is composed of miR-17, miR-18a, miR-19a, miR-20, miR-19b, and miR-92a. miR-19a and miR-19b are among the most oncogenic miRNAs of this cluster19. The increased expression of HuR stimulates miR-19a and miR-19b synthesis9. Since intronic regions flanking this cluster are associated with HuR, we developed a method to investigate if this protein could regulate miR-19a and miR-19b expression and maturation. One important prediction of our hypothesis was that, as a regulatory protein, HuR could facilitate miRNA biogenesis, leading to phenotypic alterations. One possibility was that miRNAs were processed by the stimulation of HuR but would not be mature and functional and, therefore, the effects of the protein would not directly impact the phenotype. Therefore, we developed a splicing reporter assay to investigate whether an RBP such as HuR could affect the biogenesis and maturation of an intronic miRNA. By confirming miRNA processing and maturation, our assay shows the interaction with the target sequence and the generation of a mature and functional miRNA. In our assay, we couple the expression of an intronic miRNA cluster with a luciferase plasmid to check for miRNA target-binding in cultured cells.
An overview of the protocol described here is depicted in Figure 1.
1. Plasmid construction
2. Cell culture
NOTE: The HeLa-Cre cell line was a gift from Dr. E. Makeyev21, and the papillary thyroid cancer cell (BCPAP) was kindly provided by Dr. Massimo Santoro (University "Federico II", Naples, Italy). HeLa cell, papillary thyroid cancer cell (BCPAP), and HEK-293T were used to overexpress HuR.
3. HuR overexpression
4. Total RNA isolation
5. Determination of total RNA concentration and quality
6. Reverse transcription
7. Quantitative PCR
8. The reporter assay
Our initial hypothesis was that HuR could facilitate intronic miRNA biogenesis by binding to its pre-miRNA sequence. Thus, the connection of HuR expression and miR-17-92 cluster biogenesis could point to a new mechanism governing the maturation of these miRNAs. Overexpression of HuR upon transfection of pFLAG-HuR was confirmed in three different cell lines: HeLa, BCPAP, and HEK-293T (Figure 2). As controls, untransfected cells and cells transfected with empty pFLAG vectors were used. Importantly, we observed that HuR overexpression in these cells stimulates the expression of miR-19a (Supplementary Table 1 results reported in Gatti da Silva et al.9).
To confirm that the induced miRNAs were genuinely functional, an in vitro reporter system was designed, as described here. The artificial intron in pRD-RIPE separates two coding regions of eGFP. This cassette is under the control of a tetracycline responsive element (TRE). The expression of eGFP is, thus, controlled by the presence of the antibiotic doxycycline (DOX) in the cell medium (Figure 3).
An in silico search using miRbase and TargetScan revealed possible mRNA targets for miR-19a and miR-19b (components of the miR-17-92 cluster). As a potential target for both miRNAs, the sequence 5' TTTGCACA 3', found on the RAP-IB 3' UTR region, was chosen (Supplementary Table 2). RAP-IB is a GTPase member of the Ras-associated protein family (RAS). The induced miRNAs were correctly processed and functional in our assay as it hybridized to the target sequence cloned next to luciferase (see Figure 3, pmiR-GLO target). Therefore, it blocked luciferase translation, which was reflected in reduced luciferase activity (Figure 3). As a control for this assay, we scrambled the target sequence, replacing it for 5' GGGTAAA 3' in Luc-scrambled plasmid (target and scrambled sequences are underlined in the oligos sequences in the Table of Materials).
In regular conditions and without induction of the pRD-miR-17-92 plasmid, endogenous miR-19a and/ or miR-19b found the target on pmiR-GLO, which led to reduced luciferase activity (data not shown). After DOX supplementation, a slight and not statistically significant reduction in luciferase activity was observed in cells with a scrambled target sequence. This might be due to the presence of target sequences for other miRNAs in this sequence. A strong reduction in luciferase activity was observed with RAP-IB 3'UTR upon increased miR-19a and miR-19b expression from pRD-miR-17-92 in comparison to HeLa-Cre cells (see Figure 4, compare orange and black bars). The transfection of pFLAG-HuR in these cells by itself did not change the luciferase activity (see Figure 4, compare gray and black bars). However, overexpression of HuR coupled with doxycycline supplementation, stimulating expression of the miR-17-92 cluster, further reduced luciferase activity (see Figure 4, compare gray and red bars). This indicates a positive regulation of HuR over miR-19a and miR-19b, coupled with correct processing and maturation of these miRNAs, which were able to successfully find their targets (HeLa-Cre miR-17-92-HuR-luc) (see Figure 4).
The use of this reporter system confirmed that HuR induces expression and proper maturation of miR-19a and miR-19b in HeLa cells.
Figure 1: A conceptual overview of the protocol described here. Plasmids containing the miRNA, target sequence, regulatory protein, and pTK-Renilla were transfected in cultured cells. After antibiotic selection, these cells were used for inducing miRNA expression (with doxycycline), with subsequent quantification of luciferase activity, and for RNA analysis to confirm HuR overexpression. Please click here to view a larger version of this figure.
Figure 2: Confirmation of FLAG-HuR over-expression in (A) HeLa, (B) BCPAP, and (C) HEK293T cells by qPCR. Transfection with empty pFLAG was used as a control (white bars). β-actin was used to normalize Ct values. The y-axis represents the fold change of expression calculated after normalization. The error bars represent standard errors calculated from three independent measurements. **P <0.005, ****P < 0.0005 (figure adapted from Gatti da Silva et al.9). Please click here to view a larger version of this figure.
Figure 3: Intronic miRNA reporter system. (A) Schematic representation of the pRD-RIPE plasmid that generated pRD-miR-17-92. pre-miR-17-92 was inserted inside the intron and was under the control of a dox-inducible promoter; on the right, the pmiR-GLO plasmid with RAP-IB 3'UTR target sequence (pmiRGLO-RAP-IB); the control had the scrambled target sequence (pmiRGLO-scrambled). (B) Detail on the sequence of pmiRGLO-RAP-IB, focusing on the target sequence complementary to miR-19a and miR-19b. Luciferase activity is reduced upon the interaction of miRNA and the target (figure adapted from Gatti da Silva et al.9). Please click here to view a larger version of this figure.
Figure 4: HeLa-Cre cells were co-transfected with reporter plasmids pTK-Renilla, pmiRGLO-RAP-IB, pmiRGLO-scrambled, pRD-miR-17-92, and pFLAG-HuR. Luciferase activity on untransfected HeLa-Cre cells (black), HeLa-Cre miR-17-92-scrambled (white), HeLa-Cre-HuR (gray), HeLa-Cre miR-17-92-luc (orange), and HeLa-Cre miR-17-92-HuR-luc (red). Luciferase activity was quantified as described in the protocol as the relative activity of firefly luciferase to Renilla luciferase (observed with pTK-Renilla) and normalized after doxycycline addition. It is shown as "relative light units" (RLU) on the y-axis. **P < 0.005; ****P < 0.0005 (figure adapted from Gatti da Silva et al.9). Please click here to view a larger version of this figure.
Supplementary Table 1: Raw Ct values observed for qPCR to analyze miR-19 expression. Please click here to download this Table.
Supplementary Table 2: RAP-IB 3'UTR sequence and miR-19 target sequence. Please click here to download this Table.
Pre-mRNA splicing is an important process for gene expression regulation, and its control can trigger strong effects on cell phenotypic modifications22,23. More than 70% of miRNAs are transcribed from introns in humans, and we hypothesized that their processing and maturation could be facilitated by splicing regulatory proteins24,25. We developed a method to analyze intronic miRNA processing and function in cultured cells. Our assay uses four different plasmids and allows us to test the maturation of endogenous and exogenous or induced miRNAs. The method described here can be performed in 2-3 days and does not require expensive reagents.
Considering HuR is an RNA binding protein important for mRNA stability and is immune-precipitated with miR-18 and miR-19a miRNAs17, this assay tested if it could drive intronic miRNA biogenesis and maturation. HuR is not found in the core of spliceosomes but interacts with proteins involved with mRNA stabilization26. Consistently, overexpression of HuR has been associated with the development of ovary and bladder cancers27,28. In papillary thyroid cancer cells, we observed that HuR increases cell proliferation, migration, and invasion9. HuR also affects the mRNA stability of cell cycle regulators such as PTEN, cyclin A2, and cyclin D2, increasing cell proliferation and migration and leading to tumorigenic characteristics29,30. However, a significant part of these mRNAs also have target sequences for miR-19a and miR-19b. Notably, HuR can bind to the intronic region where these miRNAs are transcribed, leading to the notion that it can control the splicing and maturation of these miRNAs and, consequently, enhance tumorigenic features.
In order to test that, this assay was developed with four different plasmids. The first contains the miRNA cluster miR-17-92 inserted inside an intron, and its expression is controlled upon doxycycline addition. The second plasmid is the commercial pmiR-GLO carrying the target sequence next to the 3' of the luc2 gene. Binding of the miRNA to this target sequence affects luc2 expression, which can be measured upon a luminescence assay. The third plasmid is pFLAG-HuR, which induces the overexpression of this regulatory protein. Finally, pTK-Renilla is also used to include the Renilla luciferase fluorescence. Our results indicated that HuR stimulates miR-17-92 expression and increases the association of its components with the target sequences, therefore resulting in mature and functional miRNA molecules. We describe the association of HuR with miRNA in another paper9. The target sequences used in this study were specific to miR-19a and miR-19b miRNAs. However, this method can also be adapted to the use of shorter or longer target sequences, including multiple target sites. In this case, unspecific binding of endogenous miRNAs to the target sequence or sequences nearby must be considered in the analysis. The sequences used might have other possible target sites, which might affect luciferase activity. The method described here could also be used to validate different targets of a miRNA or a group of miRNAs transcribed together, enabling functional studies of miRNAs and their specific phenotypic alterations. It should also be noted that it allows the results to be compared among different cell lines and genetic backgrounds.
The most critical steps for the broader use of this protocol in mammalian cells are the efficiency of the transfection and induction steps. Since different plasmids are transfected into the cells, adequate controls are also critical. These parameters change with the size of the plasmids and are affected by the extensive time in cell culture. Therefore, an important point to be considered before beginning the assay is the ease of transfection of the chosen cells. Additionally, the extensive use of different antibiotics to select for the different plasmids (gentamicin, puromycin, and doxycycline) and activate the expression can be harmful to cells and reduce the efficiency of the following experiments. It is important that this assay is first optimized with cells with known transfection efficiency.
The reporter successfully showed that increased HuR expression could induce miRNA biogenesis and confirmed its functional maturation. Importantly, overexpression of HuR did not affect the amount of intron produced since we did not observe reduced luciferase activity in this condition (Figure 4). Further experiments can be performed to characterize RBPs and miRNAs using this reporter system. For example, it would be important to create single mutations in miRNA sequences or in their flanking regions and analyze the impact of regulatory proteins in miRNA biogenesis. This would allow the specific characterization of the biogenesis of individual miRNAs and could also improve the knowledge on conserved sequences for miRNA processing and maturation.
We consider this is an important methodological contribution toward the study of miRNA biogenesis and maturation and the role of splicing regulatory proteins in these processes. It could also be applied to studies on the regulation and control of mRNAs and miRNAs in different types of cells.
The authors have nothing to disclose.
The authors are grateful to E. Makeyev (Nanyang Technological University, Singapore) for the HeLa-Cre cells and pRD-RIPE and pCAGGS-Cre plasmids. We thank Edna Kimura, Carolina Purcell Goes, Gisela Ramos, Lucia Rossetti Lopes, and Anselmo Moriscot for their support.
Recombinant DNA | |||
pCAGGS-Cre (Cre- encoding plasmid) | A kind gift from E. Makeyev from Khandelia et al., 2011 | ||
pFLAG-HuR | Generated during this work | ||
pmiRGLO-RAP-IB | Generated during this work | ||
pmiRGLO-scrambled | Generated during this work | ||
pRD-miR-17-92 | Generated during this work | ||
pRD-RIPE-donor | A kind gift from E. Makeyev from Khandelia et al., 2011 | ||
pTK-Renilla | Promega | E2241 | |
Antibodies | |||
anti-B-actin | Sigma Aldrich | A5316 | |
anti-HuR | Cell Signaling | mAb 12582 | |
IRDye 680CW Goat anti-mouse IgG | Li-Cor Biosciences | 926-68070 | |
IRDye 800CW Goat anti-rabbit IgG | Li-Cor Biosciences | 929-70020 | |
Experimental Models: Cell Lines | |||
HeLa-Cre | A kind gift from E. Makeyev from Khandelia et al., 2011 | ||
HeLa-Cre miR17-92 | Generated during this work | ||
HeLa-Cre miR17-92-HuR | Generated during this work | ||
HeLa-Cre miR17-92-HuR-luc | Generated during this work | ||
HeLa-Cre miR17-92-luc | Generated during this work | ||
HeLa-Cre miR17-92-scrambled | Generated during this work | ||
Chemicals and Peptides | |||
DMEM/high-glucose | Thermo Fisher Scientific | 12800-017 | |
Doxycycline | BioBasic | MB719150 | |
Dual-Glo Luciferase Assay System | Promega | E2940 | |
EcoRI | Thermo Fisher Scientific | ER0271 | |
EcoRV | Thermo Fisher Scientific | ER0301 | |
Geneticin | Thermo Fisher Scientific | E859-EG | |
L-glutamine | Life Technologies | ||
Opti-MEM I | Life Technologies | 31985-070 | |
pFLAG-CMV™-3 Expression Vector | Sigma Aldrich | E6783 | |
pGEM-T | Promega | A3600 | |
Platinum Taq DNA polymerase | Thermo Fisher Scientific | 10966-030 | |
pmiR-GLO | Promega | E1330 | |
Puromycin | Sigma Aldrich | P8833 | |
RNAse OUT | Thermo Fisher Scientific | 752899 | |
SuperScript IV kit | Thermo Fisher Scientific | 18091050 | |
Trizol-LS reagent | Thermo Fisher | 10296-028 | |
trypsin/EDTA 10X | Life Technologies | 15400-054 | |
XbaI | Thermo Fisher Scientific | 10131035 | |
XhoI | Promega | R616A | |
Oligonucleotides | |||
forward RAP-1B pmiRGLO | Exxtend | TCGAGTAGCGGCCGCTAGTAAG CTACTATATCAGTTTGCACAT |
|
reverse RAP-1B pmiRGLO | Exxtend | CTAGATGTGCAAACTGATATAGT AGCTTACTAGCGGCCGCTAC |
|
forward scrambled pmiRGLO | Exxtend | TCGAGTAGCGGCCGCTAGTAA GCTACTATATCAGGGGTAAAAT |
|
reverse scrambled pmiRGLO | Exxtend | CTAGATTTTACCCCTGATATAGT AGCTTACTAGCGGCCGCTAC |
|
forward HuR pFLAG | Exxtend | GCCGCGAATTCAATGTCTAAT GGTTATGAAGAC |
|
reverse HuR pFLAG | Exxtend | GCGCTGATATCGTTATTTGTG GGACTTGTTGG |
|
forward pre-miR-1792 pRD-RIPE | Exxtend | ATCCTCGAGAATTCCCATTAG GGATTATGCTGAG |
|
reverse pre-miR-1792 pRD-RIPE | Exxtend | ACTAAGCTTGATATCATCTTG TACATTTAACAGTG |
|
forward snRNA U6 (RNU6B) | Exxtend | CTCGCTTCGGCAGCACATATAC | |
reverse snRNA U6 (RNU6B) | Exxtend | GGAACGCTTCACGAATTTGCGTG | |
forward B-Actin qPCR | Exxtend | ACCTTCTACAATGAGCTGCG | |
reverse B-Actin qPCR | Exxtend | CCTGGATAGCAACGTACATGG | |
forward HuR qPCR | Exxtend | ATCCTCTGGCAGATGTTTGG | |
reverse HuR qPCR | Exxtend | CATCGCGGCTTCTTCATAGT | |
forward pre-miR-1792 qPCR | Exxtend | GTGCTCGAGACGAATTCGTCA GAATAATGTCAAAGTG |
|
reverse pre-miR-1792 qPCR | Exxtend | TCCAAGCTTAAGATATCCCAAAC TCAACAGGCCG |
|
Software and Algorithms | |||
Prism 8 for Mac OS X | Graphpad | https://www.graphpad.com | |
ImageJ | National Institutes of Health (NIH) | http://imagej.nih.gov/ij |