Real-time quantification of microRNAs by stem–loop

Caifu Chen*,Dana A.Ridzon,Adam J.Broomer,Zhaohui Zhou,Danny H.Lee,

Julie T.Nguyen,Maura Barbisin,Nan Lan Xu,Vikram R.Mahuvakar,Mark R.Andersen,Kai Qin Lao,Kenneth J.Livak and Karl J.Guegler

Applied Biosystems,850Lincoln Centre Drive,Foster City,CA 94404,USA

Received May 24,2005;Revised July 8,2005;Accepted October 25,2005

ABSTRACT

A novel microRNA (miRNA)quantification method has been developed using stem–loop RT followed by TaqMan PCR analysis.Stem–loop RT primers are better than conventional ones in terms of RT effici-ency and specificity.TaqMan miRNA assays are spe-cific for mature miRNAs and discriminate among related miRNAs that differ by as little as one nucleot-ide.Furthermore,they are not affected by genomic DNA contamination.Precise quantification is achieved routinely with as little as 25pg of total RNA for most miRNAs.In fact,the high sensitivity,specificity and precision of this method allows for direct analysis of a single cell without nucleic acid purification.Like standard TaqMan gene expression assays,TaqMan miRNA assays exhibit a dynamic range of seven orders of magnitude.Quantification of five miRNAs in seven mouse tissues showed vari-ation from less than 10to more than 30000copies per cell.This method enables fast,accurate and sensitive miRNA expression profiling and can identify and monitor potential biomarkers specific to tissues or diseases.Stem–loop RT–PCR can be used for the quantification of other small RNA molecules such as short interfering RNAs (siRNAs).Furthermore,the concept of stem–loop RT primer design could be applied in small RNA cloning and multiplex assays for better specificity and efficiency.

INTRODUCTION

MicroRNAs (miRNAs)are naturally occurring,highly con-served families of transcripts (18–25nt in length)that are processed from larger hairpin precursors (1,2).miRNAs are

found in the genomes of animals (3–9)and plants (10–12).To date,there are $1000unique transcripts,including 326human miRNAs in the Sanger Center miRNA registry (13).

miRNAs regulate gene expression by catalyzing the cleav-age of messenger RNA (mRNA)(14–19)or repressing mRNA translation (19–21).They are believed to be critical in cell development,differentiation and communication (2).Specific roles include the regulation of cell proliferation and metabol-ism (22),developmental timing (23,24),cell death (25),haematopoiesis (26),neuron development (27),human tumorigenesis (28)and DNA methylation and chromatin modification (29).

Although miRNAs represent a relatively abundant class of transcripts,their expression levels vary greatly among species and tissues (30).Less abundant miRNAs routinely escape detection with technologies such as cloning,northern hybrid-ization (31)and microarray analysis (32,33).Here,we present a novel real-time quantification method for accurate and sens-itive detection of miRNAs and other small RNAs.This method expands the real-time PCR technology for detecting gene expression changes from macromolecules (e.g.mRNAs)to micro molecules (e.g.miRNAs).

MATERIALS AND METHODS

Targets,primers and probes (Supplementary Data)Seventeen miRNA genes were selected from the Sanger Center miRNA Registry at http://www.sanger.ac.uk/Software/Rfam/mirna/index.shtml.All TaqMan miRNA assays are available through Applied Biosystems (P/N:4365409).Stand-ard TaqMan Òassays for pri-miRNA precursors,pri-let-7a-3and pri-miR-26b and pre-miRNA precursor pre-miR-30a were designed using PrimerExpress Òsoftware (Applied Biosys-tems,Foster City,CA).All sequences are available in the section of the Supplementary Data.Synthetic miRNA oligo-nucleotides were purchased from Integrated DNA Technolo-gies (IDT,Coralville,IA).

*To whom correspondence should be addressed.Tel:+16506385245;Fax:+16506386343;Email:chencx@appliedbiosystems.com ÓThe Author 2005.Published by Oxford University Press.All rights reserved.

The online version of this article has been published under an open access model.Users are entitled to use,reproduce,disseminate,or display the open access version of this article for non-commercial purposes provided that:the original authorship is properly and fully attributed;the Journal and Oxford University Press are attributed as the original place of publication with the correct citation details given;if an article is subsequently reproduced or disseminated not in its entirety but only in part or as a derivative work this must be clearly indicated.For commercial re-use,please contact journals.permissions@oxfordjournals.org

Nucleic Acids Research,2005,Vol.33,No.20e179

doi:10.1093/nar/gni178

Published online November 27, 2005

total RNA preparation

Mouse total RNA samples from brain,heart,liver,lung, thymus,ovary and embryo at day10–12were purchased from Ambion(P/N:7810,7812,7814,7816,7818,7824, 7826and7968).Ambion’s mouse total RNAs are derived from Swiss Webster mice.All RNA samples were normalized based on the TaqManÒGene Expression Assays for human or mouse glyceraldehyde-3-phosphate dehydrogenase(GAPDH) endogenous controls(P/N:4310884E and4352339E,Applied Biosystems).

Two cell lines,HepG2and OP9,were cultured using Gibco MEM(P/N:12492–021,Invitrogen,Carlsbad,CA) supplemented with10%fetal bovine serum(FBS)(P/N: SH30070.01,HyClone,Logan,UT).Trypsinized cells were counted with a hemocytometer.Approximately 2.8·106 suspended cells were pelleted by centrifugation(Allegra6, Beckman Coulter,Fullerton,CA)at1500r.p.m.for5min, washed with1ml Dulbecco’s phosphate-buffered saline(PBS) without MgCl2and CaCl2(P/N:14190078,Invitrogen,Carls-bad,CA).The cell pellets were re-suspended in140m l PBS and processed with three different sample preparation meth-ods.With thefirst method,a50m l sample(106cells)was mixed with an equal amount of Nucleic Acid Purification Lysis Solution(P/N:43055;Applied Biosystems)by pipet-ting up and down10times,and then spun briefly.The lysate was diluted1/10with1U/m l RNase inhibitor solution(P/N: N8080119;Applied Biosystems)before adding the solution to an RT reaction.In the second method,a50m l sample(106 cells)was used to purify total RNA using the mirVanaÔmiRNA Isolation Kit(P/N:1560,Ambion,Austin,TX) according to the manufacturer’s protocol.Purified total RNA was eluted in100m l of elution buffer.The third method involved diluting cells1/2with1·PBS,heating at95 C for5 min,and immediately chilling on ice before aliquotting dir-ectly into RT reactions.

miRNA detection using mirVanaÔmiRNA

detection kit

Solution hybridization-based miRNA analysis was carried out using the mirVanaÔmiRNA Detection Kit(Cat.#:1552, Ambion)according to the manufacturer’s protocol.RNA probes were synthesized by IDT.The radioisotope labeled RNA fragments were detected and quantitated with a Cyclone Storage Phosphor System(PerkinElmer,Boston,MA). Reverse transcriptase reactions

Reverse transcriptase reactions contained RNA samples including purified total RNA,cell lysate,or heat-treated cells,50nM stem–loop RT primer(P/N:4365386and 4365387,Applied Biosystems),1·RT buffer(P/N: 4319981,Applied Biosystems),0.25mM each of dNTPs, 3.33U/m l MultiScribe reverse transcriptase(P/N:4319983, Applied Biosystems)and0.25U/m l RNase inhibitor(P/N: N8080119;Applied Biosystems).The7.5m l reactions were incubated in an Applied Biosystems9700Thermocycler in a 96-or384-well plate for30min at16 C,30min at42 C,5min at85 C and then held at4 C.All Reverse transcriptase reac-tions,including no-template controls and RT minus controls, were run in duplicate.PCR

Real-time PCR was performed using a standard TaqManÒPCR kit protocol on an Applied Biosystems7900HT Sequence Detection System(P/N:4329002,Applied Biosystems).The 10m l PCR included0.67m l RT product,1·TaqManÒUni-versal PCR Master Mix(P/N:4324018,Applied Biosystems), 0.2m M TaqManÒprobe,1.5m M forward primer and0.7m M reverse primer.The reactions were incubated in a384-well plate at95 C for10min,followed by40cycles of95 C for15s and60 C for1min.All reactions were run in triplicate.The threshold cycle(C T)is defined as the fractional cycle number at which thefluorescence passes thefixed threshold.TaqManÒC T values were converted into absolute copy numbers using a standard curve from synthetic lin-4miRNA.

The method for real-time quantification of pri-miRNA precursors,let-7a-3and miR-26b,and pre-miRNA precursor miR-30a was described elsewhere(34).

RESULTS

We proposed a new real-time RT–PCR scheme for miRNA quantification(Figure1).It included two steps:RT and real-time PCR.First,the stem–loop RT primer is hybridized to a miRNA molecule and then reverse transcribed with a Multi-Scribe reverse transcriptase.Next,the RT products are quan-tified using conventional TaqMan

PCR.

Figure1.Schematic description of TaqMan miRNA assays,TaqMan-based real-time quantification of miRNAs includes two steps,stem–loop RT and real-time PCR.Stem–loop RT primers bind to at the30portion of miRNA molecules and are reverse transcribed with reverse transcriptase.Then,the RT product is quantified using conventional TaqMan PCR that includes miRNA-specific forward primer,reverse primer and a dye-labeled TaqMan probes.The purpose of tailed forward primer at50is to increase its melting temperature(Tm) depending on the sequence composition of miRNA molecules.

e179Nucleic Acids Research,2005,Vol.33,No.20P AGE2OF9P AGE3OF9Nucleic Acids Research,2005,Vol.33,No.20e179

Figure2.Dynamic range and sensitivity of the TaqMan lin-4miRNA assay.(A)Amplification plot of synthetic lin-4miRNA over seven orders of magnitude. Synthetic RNA input ranged from1.3·10À3fM(equivalent to7copies per reaction)to1.3·104fM(7·107copies per reaction)in PCR;(B)Standard curve of the lin-4miRNA.The dynamic range and sensitivity of the miRNA quanti-fication scheme werefirst evaluated using a synthetic cel-lin-4 target.Synthetic RNA was quantified based on the A260value and diluted over seven orders of magnitude.The cel-lin-4 TaqMan miRNA assay showed excellent linearity between the log of target input and C T value,demonstrating that the assay has a dynamic range of at least7logs and is capable of detecting as few as seven copies in the PCR reaction(Figure2). Eight additional miRNA assays were also validated using mouse lung total RNA.The RNA input ranged from0.025to 250ng(Figure3).The C T values correlated to the RNA input (R2>0.994)over four orders of magnitude.A negative control assay,cel-miR-2,did not give a detectable signal,even in reactions with250ng mouse total RNA.

The expression profile offive miRNAs was determined in seven different mouse tissues to create a miRNA expression map.The copy number per cell was calculated based on the input total RNA(assuming15pg/cell)and the standard curve of synthetic lin-4target.Several interesting observations were made from this expression map.First,miRNAs are very abundant,averaging2390copies per cell in these tissues. The level of expression ranged from less than10to32090 copies per cell.Of the12miRNAs,miR-16and miR-323were the most and least abundant miRNAs,respectively,across all tissues.In addition,each tissue had a distinctive signature of miRNA expression.The overall level of miRNA expression was highest in mouse lung and lowest in embryos.Finally,the dynamic range of miRNA expression varied greatly from less than5-fold(let-7a)to more than2000-fold(miR-323)among these seven tissues(Table1).

To assess the need for RNA isolation,we added cell lysates directly to miRNA assays.The equivalent of2.5–2500cells were added directly to7.5m l RT reactions.When detected, the C T values correlated(R2>0.998)to the number of cells

in Figure3.Correlation of total RNA input to the threshold of cycle(C T)values for eight miRNA assays.Mouse lung total RNA input ranged from0.025to250ng per

RT reaction.A Caenorhabditis elegans miRNA(miR-2)was included as a negative control assay.

Table1.Expression profiles of five miRNAs across seven mouse tissues

miRNA ID Copy number per cell cFold-change Brain Heart Liver Lung Thymus Ovary Embryo Average

let-7a20101420700239014203120105017305

miR-161024013520302208032090111005210140206

miR-2070300130580199042062059028

miR-2167025404450797035505310390355020

miR-2229010203105901305604042026 Average224020401140443035402600750239017

Mouse or human total RNA samples from brain,heart,liver,lung,thymus,ovary and embryo(at day10–12)were purchased from Ambion.Copy number per cell is estimated based on standard curve of lin-4synthetic miRNA.A total of150ng RNA(or equivalent to approximately10000cells assuming15pg of total RNA per cell) was added to each RT reaction.RNA input was normalized based on TaqMan GAPDH endogenous control(P/N:4352339E).

e179Nucleic Acids Research,2005,Vol.33,No.20P AGE4OF9the RT reactions over at least three orders of magnitude (Figure4).

The effect of non-specific genomic DNA on TaqMan miRNA assays was also tested for12assays.Results showed no difference in C T values in the presence or absence of5ng of human genomic DNA added to the RT reactions,suggesting that the assays are highly specific for RNA targets(data not shown).Based on this observation,we added heat-treated cells directly to miRNA quantification assays.Figure5illustrates the comparison of miRNA quantification using purified

total Figure4.Dynamic range of eight TaqMan miRNA assays using OP9cell lysates.The number of cell input ranged from3to2500cells per RT.A Caenorhabditis elegans miRNA(miR-2)was used as a negative

control.

Figure5.Comparison of heat-treated cells,cell lysate and total RNA for real-time quantitation of10miRNAs.The level of miRNA expression is measured in the threshold cycles(C T).Approximately400HepG2cells were analyzed per PCR.

P AGE5OF9Nucleic Acids Research,2005,Vol.33,No.20e179

RNA,cell lysates and heat-treated cells derived from an equal number of HepG2cells.Adding heat-treated cells directly to the miRNA assays produced the lowest C T values,and good concordance was observed among all three different sample preparation methods.

The reproducibility of TaqMan miRNA assays was examined by performing12miRNA assays with 16replicates performed by two independent operators (data not shown).The standard deviation of the C T s averaged 0.1,demonstrating the high precision of the assays.

Solution hybridization-based miRNA northern analysis was used as an independent technology to compare with TaqMan miRNA assays (Figure 6).We observed that hybridization-based miRNA analyses were less reproducible and that con-cordance with TaqMan assays varied from target to target.There was a general concordance between the two methods (R 2¼0.916)for miR-16across five mouse tissue samples.However,correlations were relatively low for less abundant miRNAs,such as miR-30(R 2¼0.751).

Hybridization methods can lack specificity for the mature miRNAs.We investigated the ability of the TaqMan miRNA assays to differentiate between the mature miRNAs and their longer precursors,using synthetic targets for pri-miRNA pre-cursors,pri-miR-26b and pri-let-7a and pre-miRNA precursor pre-miR-30a (Table 2).TaqMan assays designed to detect either precursors or mature miRNAs were tested with synthetic targets averaging 1.5·108copies per RT reaction (1.3·107copies per PCR).TaqMan miRNA analyses with only pri-miRNA precursor molecules produced C T values at least 11cycles higher than analyses with mature miRNA ones.This result implies that if mature miRNA and precursor

were at an equal concentration,the latter would contribute <0.05%background signal to the assay of mature target.For pre-miR-30a where the mature miRNA miR-30a-3p is located at the 30end of the pre-miR-30a sequence,a

difference

Figure 6.Comparison of TaqMan miRNA miR-16assay to solution-based northern hybridization analysis.Total RNAs from mouse kidney,liver,lung,spleen and testicle tissues were used.

Table 2.Discrimination between mature miRNAs and their pri-or pre-miRNA precursors ID

Synthetic miRNA (No.of copies)Synthetic precursor (No.of copies)Total RNA (ng)C T

miRNA

Precursor

miR-26b

1.5·108

0016.5ND 0 1.5·10

8

027.418.7007.521.928.7000ND ND let-7a

1.5·108

0016.5ND 0 1.5·108

029.519.4007.519.934.7000ND ND miR-30a-3p

1.5·108

0015.8ND 0 1.5·108

024.217.2007.525.430.3000ND ND miR-30a-5p

1.5·108

0016.3ND 0 1.5·108

029.118.9007.522.130.30

39.6ND

Note:ND represents no detectable PCR products after 40cycles.The copy number of synthetic miRNAs in RT was estimated based on the A 260values.Only 9%of RT product was added to PCR.Total RNA from human lung was used.Pre-miRNA precursors,pri-let-7a-3and pri-miR-26b,and pre-miRNA precursor pre-miR-30a were examined.

e179Nucleic Acids Research,2005,Vol.33,No.20P AGE 6

OF

9

of 8.4C T was observed.The results showed that TaqMan miRNA assays are specific to mature miRNAs.However,the assay specificity is better if the miRNA is located at the 50strand of the pre-miRNA precursor.Experiments analyzing total RNA instead of synthetic targets indicated that the pre-cursors are at least two orders of magnitude less abundant than mature miRNAs,based on C T differences of 7or more for miR-26b-1and let-7a-2precursors.Considered together,these results suggest that the TaqMan miRNA assays are highly specific for the mature miRNAs.

The ability of the TaqMan miRNA assays to discriminate miRNAs that differ by as little as a single nucleotide was tested with the five synthetic miRNAs of let-7a,let-7b,let-7c,let-7d and let-7e (Figure 7).Each miRNA assay was examined against each miRNA.Relative detection efficiency was cal-culated from C T differences between perfectly matched and mismatched targets,assuming 100%efficiency for the perfect match.Very low levels of non-specific signal were observed,ranging from zero to 0.3%for miRNAs with 2–3mismatched bases and only 0.1–3.7%for the miRNAs that differed by a single nucleotide.Most cross-reactions resulted from G–T mismatches during the RT reaction (let-7a assay versus let-7c target etc.).Only the targeted miRNA was detected if more than three mismatched bases between any two miRNAs were present.

We compared the discrimination ability of the TaqMan miRNA assays to that of solution-based hybridization analysis (Figure 8).In our hands,the hybridization method discrimin-ated well between let-7a and let-7b.However,poor or no discrimination was observed among let-7a,let-7c and let-7d,which differ by 1–3nt.

We speculated that stem–loop primers might provide better RT efficiency and specificity than linear ones.Base stacking of the stem might enhance the thermal stability of the RNA–DNA heteroduplex.Furthermore,spatial constraint of the stem–loop would likely improve the assay specificity in comparison to conventional linear RT primers.We compared the sensitivity

and specificity of the stem–loop and linear RT primers using synthetic miRNAs for let-7a (Figure 9).We observed several advantages for the stem–loop RT.First,in the presence of the synthetic let-7a target,the C T values between linear and stem–loop RT methods differed by 7,indicating that the efficiency of stem–loop RT was at least 100times higher.Secondly,stem–loop RT discriminated better between miRNAs that differ by two bases based on D C T values.Finally,the stem–loop RT was at least 100times better able to discriminate between the mature miRNA and its precursor,based on the D C T (precursor versus mature)of 7.

DISCUSSION

Since the discovery of miRNAs,remarkable advances in the characterization of these gene families have delineated the mechanism for their functions in gene regulation (35).As a result,extensive surveys have begun to identify miRNA bio-markers specific for tissue types or disease status.These stud-ies will benefit from methods that allow for both accurate identification and quantification of miRNAs.

Current methods for detection and quantification of miR-NAs are largely based on cloning,northern blotting (5),or primer extension (36).Although microarrays could improve the throughput of miRNA profiling,the method is relatively limited in terms of sensitivity and specificity (32,33).Low sensitivity becomes a problem for miRNA quantification because it is difficult to amplify these short RNA targets.Furthermore,low specificity may lead to false positive

signal

Figure 7.Discrimination power of let-7miRNA assays.Relative detection (%)calculated based on C T difference between perfectly matched and mismatched targets.A total of 1.5·108copies of synthetic RNA was added to RT reaction.The concentration was estimated based on the A 260

values.

Figure 8.Poor discrimination of miRNAs with solution hybridization-based (northern)

analysis.

Figure 9.Specificity of TaqMan miRNA assays between stem–loop and linear RT primers.Mature let-7a-specific assay was tested against let-7a,let-7e and pri-miR precursor let-7a-3.D C T represents the C T difference between two targets or methods.A total of 1.5·108copies of synthetic targets were added to each RT reaction.

P AGE 7

OF

9Nucleic Acids Research,2005,Vol.33,No.20e179

Real-time PCR is the gold standard for gene expression quantification(38,39).It has been a long challenge for scient-ists to design a conventional PCR assay from miRNAs aver-aging$22nt in length.We developed a novel scheme to design TaqMan PCR assays that specifically quantify miRNA expression levels with superior performance over existing conventional detection methods.We have designed and validated assays for222human miRNAs(Chen et al., unpublished data).These assays combine the power of PCR for exquisite sensitivity,real-time monitoring for a large dynamic range and TaqMan assay reporters to increase the specificity.In our hands,miRNA precursors were at least2000 times less effective targets than mature miRNAs(Table2). Because these assays are insensitive to precursors or genomic DNA,we were able to add heat-treated cells directly to the assays,eliminating the need for sample preparation.For applications where both mature miRNAs and their precursors need to be assayed,conventional TaqMan assays can be used in parallel to specifically detected precursors.

We observed the better specificity and sensitivity of stem–loop RT primers than conventional linear ones likely due to the base stacking and spatial constraint of the stem–loop structure (Figure9).The base stacking could improve the thermal sta-bility and extend the effective footprint of RT primer/RNA duplex that may be required for effective RT from relatively shorter RT primers.The spatial constraint of the stem–loop structure may prevent it from binding double-strand genomic DNA molecules and,therefore,eliminate the need of TaqMan miRNA assays for RNA sample preparation.Stem–loop RT primers can potentially be used for multiplex RT reactions and small RNA cloning for possibly better efficiency and specificity.

There is an increasing need for sensitive and specific whole miRNA profiling.The ability to effectively profile miRNAs could lead to the discoveries of disease-or tissue-specific miRNA biomarkers,as well as contribute to the understanding of how miRNAs regulate stem cell differentiation.Our stem–loop RT–PCR method should provide a practical solution for these studies.We are currently developing multiplex approaches that should further increase the utility of this method.

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online.

ACKNOWLEDGEMENTS

We greatly thank Kelly McDonald,Fenton Williams,Will Bloch,Neil Straus,and Victor Ambros for critical reading of the manuscript.Funding to pay the Open Access publication charges for this article was provided by Applied Biosystems. Conflict of interest statement.None declared.REFERENCES

1.Ambros,V.(2004)The functions of animal microRNAs.Nature,431,

350–355.

2.Bartel,D.P.(2004)MicroRNAs:genomics,biogenesis,mechanism,and

function.Cell,116,281–297.

3.Mourelatos,Z.,Dostie,J.,Paushkin,S.,Sharma,A.,Charroux,B.,Abel,L.,

Rappsilber,J.,Mann,M.and Dreyfuss,G.(2002)miRNPs:a novel class of ribonucleoproteins containing numerous microRNAs.Genes Dev.,16, 720–728.

4.Lau,N.C.,Lim,L.P.,Weinstein,E.G.and Bartel,D.P.(2001)An abundant

class of tiny RNAs with probable regulatory roles in Caenorhabditis elegans.Science,294,858–862.

5.Lagos-Quintana,M.,Rauhut,R.,Lendeckel,W.and Tuschl,T.(2001)

Identification of novel genes coding for small expressed RNAs.Science, 294,853–858.

6.Lagos-Quintana,M.,Rauhut,R.,Yalcin,A.,Meyer,J.,Lendeckel,W.and

Tuschl,T.(2002)Identification of tissue-specific microRNAs from

mouse.Curr.Biol.,12,735–739.

7.Lagos-Quintana,M.,Rauhut,R.,Meyer,J.,Borkhardt,A.and Tuschl,T.

(2003)New microRNAs from mouse and human.RNA,9,175–179. 8.Lai,E.C.(2002)Micro RNAs are complementary to30-UTR sequence

motifs that mediate negative post-transcriptional regulation.Nature

Genet.,30,363–3.

9.Lee,R.C.and Ambros,V.(2001)An extensive class of small RNAs in

Caenorhabditis elegans.Science,294,862–8.

10.Llave,C.,Xie,Z.,Kasschau,K.D.and Carrington,J.C.(2002)Cleavage of

Scarecrow-like mRNA targets directed by a class of Arabidopsis miRNA.

Science,297,2053–2056.

11.Park,W.,Li,J.,Song,R.,Messing,J.and Chen,X.(2002)CARPEL

FACTORY,a Dicer homolog,and HEN1,a novel protein,act in

microRNA metabolism in Arabidopsis thaliana.Curr.Biol.,12,

1484–1495.

12.Reinhart,B.J.,Weinstein,E.G.,Rhoades,M.W.,Bartel,B.and Bartel,D.P.

(2002)MicroRNAs in plants[Erratum(2002)Genes Dev.,16,2313.].

Genes Dev.,16,1616–1626.

13.Griffiths-Jones,S.(2004)The microRNA Registry.Nucleic Acids Res.,

32,D109–D111.

14.Hutvagner,G.and Zamore,P.D.(2002)A microRNA in a multiple-

turnover RNAi enzyme complex.Science,297,2056–2060.

15.Zhang,H.,Kolb,F.A.,Brondani,V.,Billy,E.and Filipowicz,W.(2002)

Human Dicer preferentially cleaves dsRNAs at their termini without a requirement for ATP.EMBO J.,21,5875–5885.

16.Zhang,H.,Kolb,F.A.,Jaskiewicz,L.,Westhof,E.and Filipowicz,W.

(2004)Single processing center models for human Dicer and bacterial RNase III.Cell,118,57–68.

17.Zhang,S.and Semino,C.E.(2003)Design peptide scaffolds for

regenerative medicine.Adv.Exp.Med.Biol.,534,147–163.

18.Doench,J.G.,Petersen,C.P.and Sharp,P.A.(2003)siRNAs can function as

miRNAs.Genes Dev.,17,438–442.

19.Doench,J.G.and Sharp,P.A.(2004)Specificity of microRNA target

selection in translational repression.Genes Dev.,18,504–511.

20.Wightman,B.,Ha,I.and Ruvkun,G.(1993)Posttranscriptional regulation

of the heterochronic gene lin-14by lin-4mediates temporal pattern

formation in C.elegans.Cell,75,855–862.

21.Olsen,P.H.and Ambros,V.(1999)The lin-4regulatory RNA controls

developmental timing in Caenorhabditis elegans by blocking LIN-14 protein synthesis after the initiation of translation.Dev.Biol.,216,

671–680.

22.Brennecke,J.,Hipfner,D.R.,Stark,A.,Russell,R.B.and Cohen,S.M.

(2003)Bantam encodes a developmentally regulated microRNA that controls cell proliferation and regulates the proapoptotic gene hid in Drosophila.Cell,113,25–36.

23.Reinhart,B.J.,Slack,F.J.,Basson,M.,Pasqulnelli,A.E.,Bettinger,J.C.,

Rougvie,A.E.,Horvitz,H.R.and Ruvkun,G.(2000)The21-nucleotide let-7RNA regulates developmental timing in Caenorhabditis elegans.

Nature,403,901–906.

24.Ambros,V.(2003)MicroRNA pathways in flies and worms:growth,

death,fat,stress,and timing[Erratum(2003)Cell,114,269.].Cell,

113,673–676.

25.Baehrecke,E.H.(2003)miRNAs:micro managers of programmed cell

death.Curr.Biol.,13,R473–R475.

26.Chen,C.-Z.,Li,L.,Lodish,H.F.and Bartel,D.P.(2004)MicroRNAs

modulate hematopoietic lineage differentiation.Science,303,83–86.

e179Nucleic Acids Research,2005,Vol.33,No.20P AGE8OF927.Johnston,R.J.and Hobert,O.(2003)A microRNA controlling left/right

neuronal asymmetry in Caenorhabditis elegans.Nature,426,

845–849.

28.Michael,M.Z.,O’Connor,S.M.,van Holst Pellekaan,N.G.,Young,G.P.

and James,R.J.(2003)Reduced accumulation of specific microRNAs in colorectal neoplasia.Mol.Cancer Res.,1,882–1.

29.Bao,N.,Lye,K.-W.and Barton,M.K.(2004)MicroRNA binding sites

in Arabidopsis class III HD-ZIP mRNAs are required for methylation of the template chromosome.Dev.Cell,7,653–662.

30.Kim,J.,Krichevsky,A.,Grad,Y.,Hayes,G.D.,Kosik,K.S.,Church,G.M.

and Ruvkun,G.(2004)Identification of many microRNAs that copurify with polyribosomes in mammalian neurons.Proc.Natl Acad.Sci.USA, 101,360–365.

31.Lim,L.P.,Glasner,M.E.,Yekta,S.,Burge,C.B.and Bartel,D.P.(2003)

Vertebrate microRNA genes.Science,299,1540.

32.Krichevsky,A.M.,King,K.S.,Donahue,C.P.,Khrapko,K.and Kosik,K.S.

(2003)A microRNA array reveals extensive regulation of microRNAs during brain development[Erratum(2004)RNA,10,551.].RNA,9,

1274–1281.33.Liu,C.-G.,Calin,G.A.,Meloon,B.,Gamliel,N.,Sevignani,C.,

Ferracin,M.,Dumitru,C.D.,Shimizu,M.,Zupo,S.,Dono,M.et al.(2004) An oligonucleotide microchip for genome-wide microRNA profiling in human and mouse tissues.Proc.Natl Acad.Sci.USA,101,9740–9744.

34.Schmittgen,T.D.,Jiang,J.,Liu,Q.and Yang,L.(2004)A high-throughput

method to monitor the expression of microRNA precursors.Nucleic Acids Res.,32,e43.

35.He,L.and Hannon,G.J.(2004)MicroRNAs:small RNAs with a big role

in gene regulation.Nature Rev.Genet.,5,522–531.

36.Zeng,Y.and Cullen,B.R.(2003)Sequence requirements for micro RNA

processing and function in human cells.RNA,9,112–123.

37.Allawi,H.T.,Dahlberg,J.E.,Olson,S.,Lund,E.,Olson,M.,Ma,W.-P.,

Takova,T.,Neri,B.P.and Lyamichev,V.I.(2004)Quantitation of

microRNAs using a modified Invader assay.RNA,10,1153–1161. 38.Livak,K.J.and Schmittgen,T.D.(2001)Analysis of relative gene

expression data using real-time quantitative PCR and the2[-Delta Delta C(T)]Method.Methods,25,402–408.

39.Heid,C.A.,Stevens,J.,Livak,K.J.and Williams,P.M.(1996)Real time

quantitative PCR.Genome Res.,6,986–994.

P AGE9OF9Nucleic Acids Research,2005,Vol.33,No.20e179