Total RNA extraction from tissues for microRNA and target gene expression analysis: not all kits are created equal
© The Author(s). 2018
Received: 11 July 2017
Accepted: 28 February 2018
Published: 16 March 2018
microRNAs (miRNAs) are short non-coding RNAs that fine-tune gene expression. The aberrant expression of miRNAs is associated with many diseases and they have both therapeutic and biomarker potential. However, our understanding of their usefulness is dependent on the tools we have to study them. Previous studies have identified the need to optimise and standardise RNA extraction methods in order to avoid biased results. Herein, we extracted RNA from murine lung, liver and brain tissues using five commercially available total RNA extraction methods. These included either: phenol: chloroform extraction followed by alcohol precipitation (TRIzol), phenol:chloroform followed by solid-phase extraction (column-based; miRVana and miRNeasy) and solid-phase separation with/without affinity resin (Norgen total and Isolate II). We then evaluated each extraction method for the quality and quantity of RNA recovered, and the expression of miRNAs and target genes.
We identified differences between each of the RNA extraction methods in the quantity and quality of RNA samples, and in the analysis of miRNA and target gene expression. For the purposes of consistency in quantity, quality and high recovery of miRNAs from tissues, we identified that Phenol:chloroform phase separation combined with silica column-based solid extraction method was preferable (miRVana microRNA isolation). We also identified a method that is not appropriate for miRNA analysis from tissue samples (Bioline Isolate II). For target gene expression any of the kits could be used to analyse mRNA, but if interested in analysing mRNA and miRNA from the same RNA samples some methods should be avoided.
Different methods used to isolate miRNAs will yield different results and therefore a robust RNA isolation method is required for reproducibility. Researchers should optimise these methods for their specific application and keep in mind that “total RNA” extraction methods do not isolate all types of RNA equally.
microRNAs (miRNAs) are a family of short (~ 22 nucleotide) non-coding RNAs that are essential for controlling the regulation of gene expression. They act via binding to specific sequences in the 3′-untranslated region (3′-UTR) of mRNAs, leading to either translational repression or message decay. miRNAs have been linked to distinct biological processes in normal development and their altered expression is implicated in many diseases including neurological disorders, immune diseases and cancers . miRNA profiling studies have helped to elucidate their critical roles in developmental biology such as regulation of cell cycle , apoptosis  and differentiation . Furthermore, their altered expression in disease makes them attractive as diagnostic and prognostic biomarkers, and as targets for manipulation as therapeutics.
With the increasing interest in miRNAs, particularly as diagnostic and prognostic biomarkers, a number of platforms have been developed to analyse miRNA expression. Current detection methods of miRNAs involve cloning, in situ hybridisation, microarray, bead-based flow cytometry, northern blot, next generation sequencing and quantitative real time PCR (qRT-PCR) [5, 6]. These methods each have their advantages and disadvantages. However, all rely on the quality of the in-put material. Detection methods such as qRT-PCR and microarray are dependent on the assumption that the RNA extraction method isolates all miRNAs equally [7–9]. It is becoming evident that here is a crucial need to optimise and standardise the isolation of miRNA, given the discrepancies among many studies, whereby the recovery of miRNAs is dependent on the extraction method used [10–16]. These initial studies have focussed on miRNA recovery from plasma, urine and other bodily fluids, however to date few papers have compared differences in miRNA extraction efficiency from solid tissues.
The development of a miRNA for therapeutic use also requires knowledge of its specific gene targets, a field of research which is in its infancy. Therefore, it may be of specific importance for researchers to analyse both miRNA and target gene expression in the same RNA sample. However, different extraction methods may bias towards isolation of either long or short RNAs, potentially impacting the findings of the study. This research used five commonly used RNA isolation methods to extract total RNA including miRNAs from solid tissues, and compared the extraction efficiency, quality and yield of resultant RNAs. We analysed miRNA and target gene expression using qRT-PCR from brain, liver and lung tissues from mice. The data presented herein could serve as a useful reference for researchers interested in analysing both miRNA and target gene expression in animal tissues for diagnostic and prognostic use or miRNA validation.
Brain, liver and lung tissue samples were from 8 to 10 week old male C57Bl/6 mice (n = 3). Mice were euthanised by cervical dislocation, then tissue samples were harvested and immediately snap frozen using 100% ethanol and dry ice, and stored at − 80 °C until used. The use of mice was in accordance with local institutional ethical guidelines.
RNA quantity and quality assessment
RNA samples were quantitated using the Qubit HS RNA kit (Thermo Fisher Scientific) with a Qubit 3.0 Fluorometer (Thermo Fisher Scientific), following manufacturer’s instructions. The 260/280 and 260/230 ratios of absorbance values were used to assess the purity of RNA using a Nanodrop ND-1000 spectrophotometer (Thermo Fisher Scientific). A 260/280 ratio of ~ 2.0 and 260/230 ratio in the range of 2.0–2.2 was accepted as “pure” for RNA. Lower ratios may indicate the presence of protein, phenol, EDTA, carbohydrates or other contaminants that absorb at or near 260, 230 or 280 nm. The miRNA content was quantitated using a Qubit microRNA assay kit (Thermo Fisher Scientific) with a Qubit 3.0 Fluorometer, according to manufacturer’s recommendations. The integrity of the RNA was determined for all samples using an Agilent Bioanalyzer 2100 (Agilent Technologies, Diegem, Belgium) with an Agilent RNA 6000 kit (Agilent Technologies). RNA Integrity Numbers (RINs) were used to evaluate the integrity of the RNA samples with > 7.0 considered intact and < 7.0 considered degraded. The quality assessment of RNA samples is provided in Additional file 1: Figure S2, Additional file 2: Figure S3, Additional file 3: Figure S4.
Reverse transcription and quantitative real time PCR (qRT-PCR)
For miRNA detection, 10 ng of the RNA samples were reverse transcribed in technical duplicates using a TaqMan microRNA reverse transcription kit (Thermo Fisher Scientific) with Taqman microRNA primers specific for the miRNAs of interest, following manufacturer’s instructions (see Additional file 4: Table S1). Reverse transcription products were diluted to 50 μl and 5 μl of diluted sample used in single qPCR reactions, with a total volume of 20 μl. qRT-PCR was performed using Taqman microRNA assays (Thermo Fisher Scientific) and TaqMan universal PCR master mix II, no UNG (Thermo Fisher Scientific) on a ViiA7 Real-Time PCR system (Applied Biosystems/Thermo Fisher Scientific), using recommended PCR cycling conditions. For analysis of target gene mRNA expression 250 ng of RNA was reverse transcribed into cDNA using a Quantitect reverse transcription kit (Qiagen) in duplicate reactions. For qRT-PCR 2 μl of cDNA was added to SensiMix SYBR Hi-ROX (Bioline), Quantitect Primer Assays (Qiagen; see Additional file 4: Table S1) and made up to a final volume of 20 μl with RNase free water. qRT-PCR was run on a ViiA7 Real-time PCR system, using recommended cycling conditions. Reverse transcription and set-up of qPCR reactions was performed with a QIAgility automated liquid handling system (Qiagen) to decrease variation due to manual pipetting error.
qRT-PCR data was analysed using GenEx qPCR analysis software (Exiqon, Ver 126.96.36.1998), taking PCR efficiency into account. GeNorm and Normfinder were used to select appropriate reference genes for normalisation. One-way ANOVA was used to compare the different extraction methods for each tissue and Student’s t-tests (2 tailed, unpaired) used to detect differences in normalised gene expression with p-value < 0.05 considered statistically significant.
Comparison of yield, purity and integrity of RNA extracted from tissues
Analysis of RNA extracted from mouse tissues
microRNA expression in mouse tissues extracted with different kits
miRNA (miR-7 and miR-34a) target gene (EGFR and AXL) expression in tissues
In this study we hypothesised that different methods of RNA isolation will impact results when analysing gene expression in tissues. RNA extraction methods can be broadly characterised into either phenol:chloroform extraction followed by alcohol precipitation (TRIzol), phenol:chloroform followed by solid-phase extraction (column-based; miRVana and miRNeasy) and solid-phase separation with/without affinity resin (Norgen total and Isolate II). These methodologies were primarily developed for the extraction of long mRNAs and have been based on the assumption that all RNAs are purified equally. Moreover, there are a number of considerations for choosing a particular RNA extraction method such as the quality, quantity, price, and ease of use (time to extraction). Herein, we present data that shows the method used to extract the RNA will also produce variable results compared to different methods in downstream applications and therefore is another important consideration.
This study demonstrates that RNA isolation methods vary in terms of the quantity and quality of RNA samples, and in the analysis of miRNA and target gene expression. We chose organs that can be difficult to isolate RNA from for different reasons. For example, RNA isolation from brain is often noted to result in low yields due to its high lipid content. This was particularly evident with the Bioline Isolate II kit. The manufacturers’ protocol suggests that up to 5 μg of RNA should be purified from 10 mg of rat/mouse brain. However, the miRNeasy handbook suggests that 5–20 μg of RNA should be achievable from brain with the same amount of input material. Troubleshooting on the manufacturer’s website describes one problem with lipid rich tissues is clogging of the column and suggests decreasing the sample amount or increasing the lysis buffer volume. Our extractions were performed within the suggested limits of the protocol. Subsequently, they suggest performing pre-extraction using TRIsure reagent (Bioline’s equivalent of TRIzol) then column-based separation with the kit to clean-up the aqueous phase. This method would then be comparable to miRVana and miRNeasy RNA extraction methods and may result in better yields for this tissue. Additionally, the RNA quality correlates to tissue-specific responses to physiological stress both prior to and following tissue death. Tissues such as lung and liver characteristically have low RINs and a small 28S peak as these tissues are prone to faster degradation of RNA by high levels of nucleases. To inactivate nucleases the tissue is rapidly frozen and thawing prevented until the weighed material is added to the extraction buffer. Although, thawing of tissue was prevented, we cannot rule out the time from euthanasia of the animal and excision of the organs to rapid-freezing as a contributing factor to the overall low RIN values and some degradation products observed with all lung samples, irrespective of isolation kit used. An alternative method to inactivate nucleases is using a RNA stabilising solution such as RNAlater (ThermoFisher Scientific), which allows unfrozen tissue samples to be processed later. It may help ensure the integrity of the RNA, but it is not compatible with all RNA isolation procedures and users should check their specific method prior to performing extractions. The addition of a peak after 60 s in some samples in the Bioanalyser analysis suggests gDNA contamination. An additional DNase treatment can be performed on columns with some kits, however a gDNA elimination step is highly recommended during the reverse transcription step. The inclusion of this step in our methods may explain why there was no interference when analysing mRNA expression of the Norgen lung samples, but it did impact miRNA expression analysis as the Taqman miRNA assay protocol does not include gDNA removal. Furthermore, the purity of RNA samples is a critical factor as even though samples with RIN > 7 should perform okay in most downstream applications, those that involve enzymatic reactions such as qPCR can be inhibited by nucleases, metal ions or organic contaminants. In general, the Bioline Isolate II RNA samples had lower 260/230 ratios and contaminants could be contributing to its poor performance. Finally, differences were observed in the enrichment for miRNAs in the total RNA preparation. As miRNAs represent a small fraction of the overall RNA repertoire, it can be difficult to determine their contribution from the integrity analysis. The Qubit microRNA assay offers highly selective detection of low amounts of small RNAs even in the presence of common contaminants. High percentages of miRNAs were commonly detected with miRVana and Norgen extraction methods, however given the integrity analysis for Norgen total RNA preparations, it is possible that the miRNA enrichment may be overestimated as smaller RNA fragments are being detected following degradation or oxidation of larger RNAs (mRNAs, rRNA or tRNAs) or in the case of the lung samples DNA contamination perturbing the Qubit results. Hence, the quantity, quality and purity of RNA samples are highly important factors for choosing a particular RNA extraction method and further research for how to optimise these methods for your tissue of interest can help improve these.
Profiling of miRNAs can give insight as biomarkers for diagnostic or prognostic applications. For example, panels of miRNAs can be used to classify different cancer phenotypes, predict recurrence or response to therapies [1, 12, 29–31]. However, for discovery and clinical application of miRNA-based biomarkers there needs to be optimised and standardised practices for how the RNA is extracted to prevent conflicting results. For example a meta-analysis of 63 published studies by Zhou et al. (2014) found inconsistent and even contrasting results when assessing the prognostic value of the oncomiR, miR-21 . The authors attribute the heterogeneity to differences in the source of the samples, the detection methods and normalization methods but did not allude to the methods for extracting RNA, which we show also contributes.
The biological functions and targets of very few miRNAs have been experimentally validated, but this is critical to realising the potential of miRNA-based therapy. Furthermore, uncovering tissue-specific biological functions will help identify their therapeutic action and potential off-target effects in normal tissues. Validation of miRNA-mRNA interactions is primarily performed in cell culture assays, which involve artificial manipulation of endogenous miRNAs. However, the levels achieved through cell culture manipulation (e.g. transfection of mimics) are not usually at physiological levels observed in vivo therefore, it is important to recapitulate the results in appropriate animal models. Currently, no reports have directly compared miRNA and target gene detection from total RNA samples using different extraction methods. We show that RNA extraction methods differ in their efficiency in isolating both short and long RNAs and therefore careful consideration must go into choosing an appropriate method if you wish to detect both from the same sample.
Previous research has commonly assumed that all kinds of RNA are purified equally. However, several reports have emerged indicating differences in extraction efficiency depending on the method used for RNA isolation. Kim et al. (2011) retracted their paper in Molecular Cell as they found that their conclusions about differences in miRNA expression from cells grown at different confluencies (high density versus low density) and when they were detached from the culture dish (adherent versus suspension) were actually accounted for by differences in RNA extraction efficiency using the TRIzol method . They found that miRNAs with low GC content and stable secondary structure were lost during extraction, rather than being degraded in the cells as they had initially published . Previous results from El-Khoury et al. (2016) found differences in miRNA recovery from cells, plasma and urine/plasma-derived exosomes when comparing TRIzol LS, miRNeasy serum/plasma and miRCURY biofluid extraction kits . The authors found that the miRCURY kit isolated highly pure RNA but poorly recovered miRNAs, TRIzol yielded low purity RNA which impacted PCR efficiency, whilst miRNeasy yielded poor quality RNA but performed the best for miRNA detection. Likewise, McAlexander et al. (2013) found differences in miRNA extraction from plasma and cerebrospinal fluid . They compared miRVana, miRCURY Cell and Plant kit and TRIzol LS extractions with and without glycogen as a carrier. miRVana was similar with or without glycogen, while miRCURY without glycogen had slightly lower miRNA recovery than miRVana. However, glycogen greatly improved miRNA recovery with the miRCURY kit but exacerbated the low yield and variability with TRIzol extraction. They then went on to compare miRCURY Cell and Plant kit to miRCURY Biofluids and miRNeasy serum/plasma extraction methods with glycogen as a carrier and found that miRCURY Biofluids had the highest relative abundance of spiked-in exogenous miRNA and concluded this was the superior method for their particular application. Collectively, these studies highlight the differences in RNA recovery using different extraction methods and suggest the need to optimise for your cell, tissue, and/or fluid of interest.
There are several steps during the work-flow that can introduce experimental variation. Beginning with the method for fixing or freezing the tissue, storage of the material, RNA extraction method, method used for reverse transcription and qPCR amplification or other downstream platforms. In this study we have attempted to control for some of these variables by flash-freezing the tissues, storage at − 80 °C, preventing thawing before extraction and using the most recommended practices for miRNA expression analysis. For instance, the use of sequence specific RT primers as opposed to a universal RT primer has been shown to be superior for specific product amplification . qPCR is considered the gold standard for analysing miRNA and target expression due to its sensitivity and specificity over global profiling platforms. Importantly, the quality of the results obtained is more important than the sheer number of miRNAs profiled from large sequencing based platforms. Furthermore, although we did not compare the effect of miRNA enrichment from our RNA samples previous reports have advised against it, particularly for global miRNA profiling platforms. For instance, Redshaw et al. (2013) found that the short RNA enrichment procedure results in significantly less relative miRNA levels when comparing total RNA and enriched material, specifically the enrichment procedure reduced the copy number of miRNAs by up to 25% of that present from pre-enriched samples and that this loss varied for different miRNA sequences. Therefore, miRNA data from total and short RNA preparations may not be directly comparable. Furthermore, it has been suggested that larger RNA can act as a carrier for small RNAs . Whether miRNA enriched methods for all companies would result in better recovery from tissues remains to be determined. Although Bioline suggest a separate kit for miRNA isolation, we used the Isolate II total RNA kit to directly compare with other extraction methods as the miRNA specific kit does not allow isolation of both short and long RNAs from the same elution. Rather the miRNA species are enriched for and isolated first, and then long RNAs can be extracted after, resulting in two separate fractions. Although the Isolate II is not optimised to isolate small RNAs such as the miRVana or miRNeasy kits, it also wasn’t superior for the target gene expression. Given the large discrepancy between certain kits researchers should opt for a more robust extraction method.
Comparative analysis of five commonly used RNA extraction methods revealed the miRVana kit was the best performer for producing consistently high yield and quality RNA without compromising miRNA and target gene detection levels. Although this kit was the most expensive, its high quality output and ease of use makes it a preferable method for RNA extraction, particulary if researchers wish to use the same RNA preparation for miRNA and gene expression analyses, as the other methods had lower detection levels and more variability. Empirically, this needs to be optimised for a specific application. A final caution to researchers is that just because a kit is described as isolating total RNA and is appropriate for mammalian tissues, cultured cells or biological fluids; it does not mean that it yields all types of RNAs equally.
The authors wish to thank Dianne Beveridge for carefully reading and reviewing the manuscript and Lisa Stuart for assistance in animal ethics application and approval.
Nil from independent sources.
Availability of data and materials
All data generated or analysed during this study are included in this published article [and its Additional files].
RB helped design the experiments, performed experimental work, analysed data and drafted the manuscript. ME, JH, TK and KR performed experimental work and analysed data. PL helped to design the experiments and to analyse and interpret data. All authors read and approved the final manuscript.
Animals were sourced, housed and treated in accordance with institutional guidelines approved by the Harry Perkins Institute of Medical Research Ethics Committee under application #AEO35.
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Wang J, Chen J, Sen S. MicroRNA as biomarkers and diagnostics. J Cell Physiol. 2016;231(1):25–30.View ArticleGoogle Scholar
- Bueno MJ, Malumbres M. MicroRNAs and the cell cycle. BBA Mol Basis Dis. 2011;1812(5):592–601.View ArticleGoogle Scholar
- Lynam-Lennon N, Maher SG, Reynolds JV. The roles of microRNA in cancer and apoptosis. Biol Rev. 2009;84(1):55–71.View ArticleGoogle Scholar
- Ivey KN, Srivastava D. MicroRNAs as regulators of differentiation and cell fate decisions. Cell Stem Cell. 2010;7(1):36–41.View ArticleGoogle Scholar
- Koshiol J, Wang E, Zhao Y, Marincola F, Landi MT. Strengths and limitations of laboratory procedures for microRNA detection. Cancer Epidemiol Biomark Prev. 2010;19(4):907–11.View ArticleGoogle Scholar
- Redshaw N, Wilkes T, Whale A, Cowen S, Huggett J, Foy CA. A comparison of miRNA isolation and RT-qPCR technologies and their effects on quantification accuracy and repeatability. BioTechniques. 2013;54(3):155–64.View ArticleGoogle Scholar
- Chugh P, Dittmer DP. Potential pitfalls in microRNA profiling. Wiley Interdiscip Rev: RNA. 2012;3(5):601–16.View ArticleGoogle Scholar
- Podolska A, Kaczkowski B, Litman T, Fredholm M, Cirera S. How the RNA isolation method can affect microRNA microarray results. Acta Biochim Pol. 2011;58(4):535–40.Google Scholar
- Ach RA, Wang H, Curry B. Measuring microRNAs: comparisons of microarray and quantitative PCR measurements, and of different total RNA prep methods. BMC Biotechnol. 2008;8(1):69.View ArticleGoogle Scholar
- Moldovan L, Batte KE, Trgovcich J, Wisler J, Marsh CB, Piper M. Methodological challenges in utilizing miRNAs as circulating biomarkers. J Cell Mol Med. 2014;18(3):371–90.View ArticleGoogle Scholar
- Li Y, Kowdley KV. Method for microRNA isolation from clinical serum samples. Anal Biochem. 2012;431(1):69–75.View ArticleGoogle Scholar
- Saplacan RMM, Mircea PA, Balacescu L, Balacescu O. MicroRNAs as non-invasive screening biomarkers of colorectal cancer. Clujul Med. 2015;88(4):453–6.View ArticleGoogle Scholar
- El-Khoury V, Pierson S, Kaoma T, Bernardin F, Berchem G. Assessing cellular and circulating miRNA recovery: the impact of the RNA isolation method and the quantity of input material. Sci Rep. 2016;6:19529.View ArticleGoogle Scholar
- Channavajjhala SK, Rossato M, Morandini F, Castagna A, Pizzolo F, Bazzoni F, Olivieri O. Optimizing the purification and analysis of miRNAs from urinary exosomes. Clin Chem Lab Med. 2014;52(3):345–54.View ArticleGoogle Scholar
- McAlexander MA, Phillips MJ, Witwer KW. Comparison of methods for miRNA extraction from plasma and quantitative recovery of RNA from cerebrospinal fluid. Front Genet. 2013;4:83.View ArticleGoogle Scholar
- Eldh M, Lötvall J, Malmhäll C, Ekström K. Importance of RNA isolation methods for analysis of exosomal RNA: evaluation of different methods. Mol Immunol. 2012;50(4):278–86.View ArticleGoogle Scholar
- Kalinowski FC, Brown RA, Ganda C, Giles KM, Epis MR, Horsham J, Leedman PJ. microRNA-7: a tumor suppressor miRNA with therapeutic potential. Int J Biochem Cell Biol. 2014;54:312–7.View ArticleGoogle Scholar
- Giles KM, Brown RA, Ganda C, Podgorny MJ, Candy PA, Wintle LC, Richardson KL, Kalinowski FC, Stuart LM, Epis MR, et al. microRNA-7-5p inhibits melanoma cell proliferation and metastasis by suppressing RelA/NF-kappaB. Oncotarget. 2016;7(22):31663–80.View ArticleGoogle Scholar
- Horsham JL, Ganda C, Kalinowski FC, Brown RA, Epis MR, Leedman PJ. microRNA-7: a miRNA with expanding roles in development and disease. Int J Biochem Cell Biol. 2015;69:215–24.View ArticleGoogle Scholar
- Horsham JL, Kalinowski FC, Epis MR, Ganda C, Brown RA, Leedman PJ. Clinical potential of microRNA-7 in cancer. J Clin Med. 2015;4(9):1668–87.View ArticleGoogle Scholar
- Saito Y, Nakaoka T, Saito H. microRNA-34a as a therapeutic agent against human cancer. J Clin Med. 2015;4(11):1951–9.View ArticleGoogle Scholar
- Pfeffer SR, Yang CH, Pfeffer LM. The role of miR-21 in cancer. Drug Develop Res. 2015;76(6):270–7.View ArticleGoogle Scholar
- Korpal M, Kang Y. The emerging role of miR-200 family of microRNAs in epithelial-mesenchymal transition and cancer metastasis. RNA Biol. 2008;5(3):115–9.View ArticleGoogle Scholar
- Kim Y, Yeo J, Kim B, Ha M, Kim V. Short structured RNAs with low GC content are selectively lost during extraction from a small number of cells. Mol Cell. 2012;46(6):893–5.View ArticleGoogle Scholar
- Bhatnagar S, Weber GJ, Vikranth S, Viktoriya S, Eugenia W. Compatible changes of lead microRNAs in circulating plasma and brain in senescence-accelerated aging and alzheimer's disease mouse models. Aging Sci. 2014;2(125) https://doi.org/10.4172/2329-8847.1000125.
- Kredo-Russo S, Ness A, Mandelbaum AD, Walker MD, Hornstein E. Regulation of pancreatic microRNA-7 expression. Experimental Diabetes Res. 2012;2012:7.View ArticleGoogle Scholar
- Choi J-W, Kang S-M, Lee Y, Hong S-H, Sanek NA, Young WS, Lee H-J. MicroRNA profiling in the mouse hypothalamus reveals oxytocin-regulating microRNA. J Neurochem. 2013;126(3):331–7.View ArticleGoogle Scholar
- Wong L, Lee K, Russell I, Chen C. Endogenous controls for real-time quantitation of miRNA using TaqMan® microRNA assays. In: Applied Biosystems application note publication 127AP11–01 available at wwwappliedbiosystemscom; 2007.Google Scholar
- Skinner HD, Lee JH, Bhutani MS, Weston B, Hofstetter W, Komaki R, Shiozaki H, Wadhwa R, Sudo K, Elimova E, et al. A validated miRNA profile predicts response to therapy in esophageal adenocarcinoma. Cancer. 2014;120(23):3635–41.View ArticleGoogle Scholar
- Tang S, Wu WKK, Li X, Wong SH, Wong N, Chan MTV, Sung JJY, Yu J. Stratification of digestive cancers with different pathological features and survival outcomes by microRNA expression. Sci Rep. 2016;6:24466.View ArticleGoogle Scholar
- Kleivi Sahlberg K, Bottai G, Naume B, Burwinkel B, Calin GA, Borresen-Dale AL, Santarpia L. A serum microRNA signature predicts tumor relapse and survival in triple-negative breast cancer patients. Clin Cancer Res. 2015;21(5):1207–14.View ArticleGoogle Scholar
- Zhou X, Wang X, Huang Z, Wang J, Zhu W, Shu Y, Liu P. Prognostic value of miR-21 in various cancers: an updating meta-analysis. PLoS One. 2014;9(7):e102413.View ArticleGoogle Scholar
- Kim Y-K, Yeo J, Ha M, Kim B, Kim VN. Cell adhesion-dependent control of microRNA decay (RETRACTED). Mol Cell. 2011;43(6):1005–14.View ArticleGoogle Scholar