A simple and cost effective method to generate dsRNA for RNAi Studies in Invertebrates

doi: 10.2306/scienceasia1513-1874.2007.33.035A Simple and Cost Effective Method to Generate

dsRNA for RNAi Studies in Invertebrates

Chalermporn Ongvarrasoponea*, Yaowaluck Roshorma and Sakol Panyima,b

Institute of Molecular Biology and Genetics, Mahidol University, Salaya, Nakhon Pathom 73170,Thailand.b

Department of Biochemistry, Faculty of Science, Mahidol University, Rama VI Road, Bangkok 10400,Thailand.

a

* Corresponding author, E-mail: mbcov@mahidol.ac.th

Received 24 Feb 2006Accepted 21 Jul 2006

ABSTRACT: Using an E. coli strain deficient in ribonuclease III activity, we have developed a simple and costeffective in vivo bacterial expression system to generate large amounts of double-stranded RNA (dsRNA).This method, which involves E. coli culture, RNase A treatment of lysed cells and total RNA extraction, iseasier and less expensive than traditional in vitro transcription techniques. The system was validated byknocking down yellow head virus replication in shrimp OKA cell cultures. Results showed that the dsRNAsprepared in vitro and in vivo possessed similar potency in inhibiting viral replication. This methodology givesan alternative means to prepare large amounts of dsRNA at low cost.

KEYWORDS: double-stranded RNA, E. coli expression, in vitro transcription, yellow head virus, shrimp.

INTRODUCTION

RNA interference is a phenomenon wherebydouble-stranded RNA triggers a potent and specificinhibition of its homologous mRNA and was discoveredby Andrew Fire and colleagues in 1998 1. Once insidethe cell, double-stranded RNA (dsRNA) is cleaved by aribonuclease III homolog or Dicer into 21-23nucleotide-long small interfering RNA (siRNA) with a 2-nucleotide overhang at the 3’ end and a 5’ phosphate.siRNA is then incorporated into RNA induced silencingcomplexes (RISC), unwound, and the single-strandedantisense siRNA is targeted to a specific region of itscomplementary mRNA resulting in mRNA degradation.RNAi is widely used to study gene function, preventviral infection and undertake genome wide screeningfor potential candidates to treat cancer and infectiousdiseases 2,3 . In invertebrates, long dsRNA can beefficiently used to silence gene expression withoutactivation of dsRNA-activated protein kinase (PKR) orthe interferon response that has been shown to occurin mammalian cell systems. Double-stranded RNA canbe introduced into animals and cells by injection,electroporation and chemical mediated transfection.Therefore, a simple and cost effective approach toproduce long dsRNA will be useful for RNAi studies,especially in invertebrates.

Several approaches can be used to synthesize longdsRNA. In vitro transcription can be effectivelyemployed to synthesize a single-stranded RNA. The

gene of interest can be cloned into a plasmid vectorcontaining the T7, T3 or Sp6 RNA polymerasepromoters. The plasmid DNA template is linearizedand then gel purified. Single-stranded sense andantisense RNA are synthesized using the appropriateRNA polymerases and the plasmid DNA template isremoved by DNase I treatment. The single-strandedRNAs are isolated from protein contamination by phenolchloroform extraction and ethanol precipitation. Then,the two single-stranded RNAs can be annealed toproduce dsRNA 1,4,5 . In addition, DNA templatescontaining T7 promoter sequences on both ends canbe produced by a PCR technique using gene specificprimers containing T7 promoter sequences linked tothe 5’ end. After in vitro transcription, the dsRNA ispurified using solid phase adsorption to remove protein,nucleotides and oligonucleotides (http://www.ambion.com/techlib/prot/fm_1626.pdf). Theseapproaches give a high yield of dsRNA.

Production of dsRNA using the commerciallyavailable kits based on the principle of in vitrotranscription on linearized DNA template or on PCRgenerated templates is widely used. However, it isincreasingly expensive when one needs to producelarge amounts of dsRNA for RNAi studies. In vivoproduction of dsRNA in the E. coli strain HT115, whichlacks ribonuclease III (RNase III) activity, with a protocolmodified from that of Timmon, et al., 2001 6 , can beused as an alternative approach to produce largeamounts of dsRNA at low cost. The E. coli strain HT115

36ScienceAsia 33 (2007)

was modified to express the T7 RNA polymerase froman IPTG inducible promoter 6 . In this study, theeffectiveness of viral derived dsRNA produced in vivoand in vitro to inhibit viral replication was compared inshrimp cells. A similar specific silencing effect to knockdown viral replication was demonstrated, suggestingthat dsRNA produced in vivo can be used as an alternativemethod to produce large amounts of dsRNA at lowcost.

MATERIALS AND METHODS

Plasmid Constructs

Recombinant plasmids expressing stem loop YHV-protease RNA were constructed in pGEM-T-easy vector(Promega, Madison, WI, USA) and pET3a vector(Novagen, Madison, WI, USA). A 400-bp cDNA fragmentin the coding region of the YHV-protease gene wasamplified using specific primers Pro-sense-Nde I (5’CAT ATG GGAATC GAC TAT CGT GAC TGC 3’) and Pro-anti-Pst I-Spe I (5’CTG CAG ACT AGT ATG CCG ACGATG TGA GCT CC 3’). The PCR fragment of the YHV-protease gene was cloned into the pGEM-T-easy vector(Promega) in the sense and antisense orientation. ThepGEM-T-easy vector containing sense-YHV-proteasewas digested with the restriction enzymes Spe I and PstI and ligated with a 200 bp PCR fragment of GFPsequences amplified with specific primers GFP-sense-Spe I (5’ AAG GCA CTA GTA TGG TGA GCA AGG GCGAGG 3’) and GFP-anti-Pst I; 5’ AAT TGC TGC AGC TGC

ACG CCG TAG GTC AG 3’) to construct a pGEM-T-easyvector containing sense-YHV-protease linked with GFP.This clone was digested with the restriction enzyme PstI and ligated with the Pst I - antisense-YHV-proteasefragment that was digested from the pGEM-T-easyvector containing antisense-YHV-protease. Therefore,the pGEM-T-easy vector containing the entire 1 kbinsert of the sense-YHV- protease, GFP and theantisense-YHV-protease was obtained and namedpGEM-T-easy-Pro (Fig 1A). This plasmid was used forin vitro production of stem loop YHV-protease dsRNAusing in vitro transcription7 . The 200-bp fragment ofGFP was the loop in the stem loop YHV- protease RNA.This region is cleaved after digestion with ribonucleaseA (RNase A) (Fig 1C). In addition, the entire 1 kb fragmentwas excised from pGEM-T-easy-Pro by digestion withNdeI and subcloned into the NdeI site of pET3a toconstruct pET3a-Pro (Fig 1B), which was used for invivo bacterial expression of the stem loop YHV-proteaseRNA.

Similarly, a recombinant plasmid encoding the GFPstem loop was constructed by amplifying a 400 bp GFPwith the primers GFP-sense-Nde I (5’ AAG GCA CTCATA TGG TGA GCA AGG GCG AGG 3’) and GFP-antisense-1-Xba I (5’ TGT TCT AGA ACT CCA GCT TGTGCC 3’), and a 600 bp GFP was amplified with theprimers GFP-sense-Nde I (5’ AAG GCA CTC ATA TGGTGA GCA AGG GCG AGG 3’) and GFP-antisense-2-XbaI (5’ TGT TCT AGA TTT GCT CAG GGC GGA CTG GGTGCT CAG 3’). The two PCR fragments were ligated at

Fig 1. Diagram of plasmid DNA constructs and double-stranded RNA product after RNase A digestion. A) pGEM-T-easy-Pro for in

vitro transcription. B) pET-3a-Pro for in vivo expression in HT115 bacterial host and C) dsRNA-Pro product from both methods.

ScienceAsia 33 (2007) 37

the XbaI site and cloned into the NdeI site of pET3avector to construct pET3a-GFP for an in vivo expressionof dsRNA-GFP.

Bacterial Induction and dsRNA Purification

Plasmids, pET3a-Pro or pET3a-GFP weretransformed by the heat shock method into the HT115bacterial host, a RNase III deficient E. coli strain, withthe RNase III gene disrupted by a Tn10 transposonwhich contains a tetracycline-resistance marker. Thegenotype of HT115 is as follows: F-, mcrA, mcrB, IN(rrnD-rrnE)1, lambda-, rnc14::Tn10(DE3) lysogen:lacUV5 promoter-T7 polymerase. The strain wasmodified to express T7 RNA polymerase from anisopropyl-β-D-thiogalactopyranoside (IPTG) induciblepromoter 6 . Therefore, dsRNA can be produced in theHT115 bacterial host after induction with IPTG.A single colony of the HT115 bacteria containingpET3a-Pro or pET3a-GFP was grown in 3-5 ml LBmedia containing ampicillin (100 µg/ml) andtetracycline (12.5 µg/ml) overnight at 37 ºC. Thebacterial starter culture was diluted 100 fold with 2xYTmedia containing ampicillin (100 µg/ml) andtetracycline (12.5 µg/ml). The bacterial culture wasinoculated in 50 ml 2xYT medium and incubated at 37ºC until the OD600 reached 0.4. T7 RNA polymerase wasinduced to express dsRNA by the addition of 0.4 mMIPTG. The bacterial culture was further incubated at 37ºC for 4 hours. Normally, an OD600 per ml reached 1after 4 hours induction. The bacterial cells wereharvested by centrifugation at 6000xg for 5 min at 4 ºC.Every one OD-ml of cell pellet was resuspended in 50µThen, one microgram of RNase A in a total volume ofl of 0.1% SDS and boiled for 2 min to lyse the cells.65 µl buffer (300 mM sodium acetate, 10 mM Tris-ClpH 7.5, and 5 mM EDTA) was added and incubated at37 ºC for 5 min in order to remove the single-strandedRNA of the GFP loop (Fig 1C) and the total RNA of thebacterial host. The remaining dsRNA was purified with200 µl of TRIZOL reagent (Invitrogen, Carlsbad, CA,USA) following the manufacturers’ protocol. Theconcentration of dsRNA was estimated by UVspectrophotometry and adjusted to a finalconcentration of 1 µg/µl prior to storage at –80 oC untiluse.

dsRNA PrdsRNA Production by oduction by in vitrin vitroo T Transcription

ranscriptionThe pGEM-T-easy-Pro plasmid (harboring a stemloop construct of 400 bp of the protease gene of YHV)was purified with a QIAGEN midi prep column(QIAGEN, Hilden, Germany). In order to synthesizedsRNA in vitro, the plasmid DNA was linearized with SalI restriction enzyme and used as the template tosynthesize single-stranded RNA using a Ribomax invitro transcription kit (Promega) as described by the

manufacturer. After in vitro transcription, thesynthesized RNA was annealed to produce dsRNA byincubating the mixture at 70 ºC for 15 min and graduallyreducing the temperature to 22ºC8. The finalconcentration of dsRNA-Pro was determined by UVspectrophotometry at 260 nm.

PrimarPrimary Cultury Cultury Culture of Le of Le of Lymphoid Cells (Oka Cells)ymphoid Cells (Oka Cells)Primary Oka cell culture of P. monodon was preparedas described by Assavalapsakul, et al., 20039. Briefly,lymphoid tissues were isolated from 2 kg of juvenileshrimp (10-15 g each) and washed in washing medium(2x Leibovitz’s L15 medium containing 100 IU/mlpenicillin, 100 mg/ml streptomycin, 15% fetal bovineserum (FBS) and 5% lactalbumin). The tissues wereminced into small pieces and washed in completemedium (washing medium with 15% shrimp meatextract). The minced tissues were seeded onto a 24 wellplate and incubated at 26 ºC until a monolayer formed.Transfection of dsRNA into Primarransfection of dsRNA into Primary Cultury Cultury Culture ofe ofOka Cells

Primary cultures of Oka cells at 70% confluence in24-well plates were transfected with 1 µg of dsRNAusing the transmessenger RNA transfection kit(QIAGEN). Forty hours post-transfection, cells wereinfected with YHV at different dilutions from 3x104 to3x101 particles. After YHV infection for 1.5 hours, theexcess virus was removed. The fresh complete mediumwas added and was incubated at 26 ºC. The cells andculture medium were collected at 72 hours after YHVinfection.

µl per well) corresponding to 3x10In this study, the amount of the purified YHV (1504 to 3x101 particleswas calculated from 10-4 to 10-7 dilutions of the stockYHV containing 2x109 TCID50/ml. Tissue cultureinfectious dose 50 (TCIDvirus causing cytopathic effect in 50% of the inoculated50), which is the dilution ofcell culture, was determined by incubating the primarylymphoid cell culture with 10-fold serial dilutions ofYHV. The number of cytopathic foci that representedYHV infection was counted after staining with crystalviolet 9 .

RT-PCR of the Helicase Gene of YHV and ActinTotal RNA from YHV infected Oka cells wasextracted using TRI Reagent@- LS (Molecular ResearchCenter, Inc., Cincinnati, Ohio, USA). The first strandcDNA was synthesized using ImProm-IITM reversetranscriptase (Promega) and Oligo dT primer accordingto the manufacturer’s instructions. Gene specificprimers for the helicase gene (800 bp) of YHV (5’ CAAGGA CCA CCT GGT ACC GGT AAG AC 3’ and 5’ GCGGAA ACG ACT GAC GGC TAC ATT CAC 3’) and actinprimers (5’ GAC TCG TAC GTG GGC GAC GAG G 3’ and

385’ AGC AGC GGT GGT CAT CTC CTG CTC 3’) were usedto simultaneously amplify YHV and actin mRNAs7 . ThePCR products that used to detect the expression ofYHV and actin were analysed by agarose gelelectrophoresis.

Westerestern blot Analysis

n blot AnalysisOka cell culture medium was collected to determinethe YHV replication. The medium (200 µl) was mixedwith an equal volume of 4xSDS sample buffer andboiled for 10 min. Equal amounts of proteins wereelectrophoresed in 10% SDS-polyacrylamide gels andstained by Coomassie brilliant blue as a loading control.In addition, the proteins were transferred onto a PVDFmembrane (Bio-Rad Laboratories, Inc., Hercules, CA,USA) with a SemiDry electrophoresis transfer apparatus(Bio-Rad Laboratories, Inc.). The membrane wasblocked overnight at 4 ºC in phosphate buffered saline(PBS) containing 5% skimmed milk. The YHV-structuralprotein was detected by incubating the membrane withmouse anti-gp116 antiserum in 5% skimmed milk inPBS containing 0.2% Tween-20 (PBST) (dilution 1:2000)for 1 hour at room temperature10 . After washing withPBST, the membrane was incubated with horseradishperoxidase conjugated goat anti-mouse polyclonalantibodies (Sigma Chemical, St. Louis, MO, USA)(dilution 1:8000). The signal was detected by ECL PlusWestern Blotting Detection Reagent (AmershamBiosciences, Buckinghamshire, UK).RESULTS AND DISCUSSION

Double stranded RNA triggers a potent and specificinhibition of its cognate mRNA in the process of RNAinterference. dsRNA can be produced using both invitro transcription and in vivo expression in bacteria. Alarge amount of dsRNA is required for use in genesilencing experiments to study gene function and viraldefense mechanisms. In this study, we have developeda simple approach with relatively low cost to producelarge amounts of dsRNA in E. coli strain HT115, whichlacks ribonuclease III, an enzyme that normallydegrades dsRNA. Induction of dsRNA production canbe caused by adding IPTG to induce expression of theT7 RNA polymerase in E.coli. Comparison of dsRNA ofthe protease gene of YHV (dsRNA-Pro) produced by invitro transcription and in vivo expression is shown inFig 2. A major band of 400 bp of dsRNA-Pro wasobserved using both methods. The faint smearbackground of the dsRNA-Pro purified from thebacterial cells was due to non-specific contaminationof the host RNA. However, in vivo dsRNA-Pro can beused as effectively as the in vitro dsRNA-Pro. In thisexperiment, a total of 1.5 mg was obtained from a 50ml bacterial culture after digestion with RNase A.

ScienceAsia 33 (2007)

Therefore, one OD600-ml of the bacterial cell pelletyielded 30 µg dsRNA-Pro. In this method, the productionof dsRNA can be scaled up by simply growing a largeramount of the bacterial culture. On the other hand, anin vitro transcribed hairpin RNA requires a linearizedDNA template of at least 5-10 µg in order to producemilligram amounts of dsRNA. It is time consuming,labor intensive and expensive to produce and purify alarge amount of linearized DNA template. In addition,the reagents used for an in vitro transcription arerelatively expensive.

The effectiveness of dsRNA-Pro produced by invitro transcription and in vivo bacterial expression wascompared. An unrelated dsRNA of green fluorescentprotein (dsRNA-GFP) was used as a control for thespecificity of dsRNA-Pro on viral replication. The resultshowed that the YHV genome (Fig 3) and YHV-structuralprotein, gp-116 (Fig 4) were detected at 72 hours postinfection in the cells that were not transfected withdsRNA-Pro (mock) and in those transfected withdsRNA-GFP. The level of the YHV genome and gp-116protein were viral dose dependent. Transfection of anin vitro transcribed dsRNA-Pro specifically knockeddown YHV genome (Fig 3) and protein expression (Fig4) in those cells infected with 30 to 3x104 YHV particles.The dsRNA-Pro produced in vivo in E. coli showed asimilar potency to the in vitro transcribed dsRNA-Proin knocking down YHV replication (Fig 3 and 4). Inaddition, dsRNA-Pro produced in vivo showed no

Fig 2. Double-stranded RNA of the protease gene of YHV

(dsRNA-Pro) produced using in vitro transcription andin vivo expression in bacteria. dsRNA-Pro was producedas an in vitro transcribed hairpin RNA from a linearizedpGEM-T-easy-Pro template and treated with RNase A(lane 1). The bacteria E. coli strain HT115 harboringpET3a-Pro was induced with 0.4 mM IPTG to expressdsRNA-Pro in vivo. Total RNA was isolated and treatedwith RNase A (lane 2). DsRNAs were electrophoresedthrough a 1.5% agarose gel and visualized by stainingwith ethidium bromide. Lane M represents 1 kb PlusDNA Ladder (Invitrogen).

ScienceAsia 33 (2007) 39

Fig 3. DsRNA-Pro produced in vitro and in vivo showed

similar specific inhibition of YHV replication in Okacells. The amount of YHV genome was determined byRT-PCR. Mock represents cells without dsRNA trans-fection. Control cells (C) were not infected with YHV.The numbers 4, 3, 2, and 1 represent the amounts ofYHV infected cells at 3x104, 3x103, 3x102 and 30, re-spectively. Actin was used as an internal control for equalloading. Lane M represents 1 kb Plus DNA Ladder(Invitrogen).

Fig 4. Bacterially expressed dsRNA-Pro inhibits YHV

replication at the protein level similar to in vitrotranscribed dsRNA-Pro. Western blot analysis wasperformed on the culture medium collected at 72 hrpost YHV infection for analysis of YHV structuralprotein with anti-gp116 antibody. Mock represents cellswithout dsRNA transfection. Control cells (C) were notinfected with YHV. The numbers 4, 3, 2, and 1 representthe amount of YHV with which the cells were infected:3x104, 3x103, 3x102 and 30 viral particles, respectively.

obvious toxic effect on the transfected cells, and couldbe injected into shrimp (data not shown). Similarly, acrude extract of the bacterially expressed virus deriveddsRNA has been used to protect plants against viralinfections 11 .

Taken together, in vivo production of dsRNA usingan E.coli bacterial system can be effectively employedto produce a large amount of dsRNA with relatively lowcost for RNAi studies. Production of a large amount ofdsRNA is required for silencing the expression of a genein vivo. DsRNA-Pro at the dose of 25 µg per shrimp wasinjected into shrimp in order to knock down YHVreplication and reduce mortality 12 . It should be notedthat the cost of production of 30 mg dsRNA by the invivo system was approximately one-third of the in vitromethodology. The method to generate dsRNA in vivo isrelatively simple, involving E. coli culturing, RNase Atreatment of the lysed cells and total RNA extraction.

It thus gives an alternative means to produce a largeamount of dsRNA at low cost.

ACKNOWLEDGEMENTS

We thank Wanlop Chinnirunvong for technicalassistance and Prof. Duncan R. Smith for critical readingof the manuscript. This work was supported by aThailand Research Fund (TRF) Senior Research ScholarAward to S.P., and a Commission on Higher Educationand TRF grant to C.O.

REFERENCES

1.Fire A, Xu S, Montgomery MK, Kostas SA, Driver SE and

Mello CC (1998) Potent and specific genetic interference bydouble-stranded RNA in Caenorhabditis elegans. Nature391391:806-11.

2.Meister G and Tuschl T (2004) Mechanisms of gene silencing

by double-stranded RNA. Nature 431431:343-9.

3.Novina CD and Sharp PA (2004) The RNAi revolution. Nature

430430:161-4.

4.Timmons L and Fire A (1998) Specific interference by

ingested dsRNA. Nature 395395:8.

5.Betz N and Worzella T (2003) Take the RiboMaxTM Express

for RNAi. Promega Note 8484:7-11.

6.Timmons L, Court DL and Fire A (2001) Ingestion of

bacterially expressed dsRNAs can produce specific and potentgenetic interference in Caenorhabditis elegans. Gene 263263:103-12.

7.Tirasophon W, Roshorm Y and Panyim S (2005) Silencing of

yellow head virus replication in penaeid shrimp cells bydsRNA. Biochem Biophys Res Commun 334334:102-7.

8.Worby CA, Simonson-Leff N and Dixon JE (2001) RNA

interference of gene expression (RNAi) in cultured Drosophilacells. Sci STKE 2001:PL1.

9.Assavalapsakul W, Smith DR and Panyim S (2003)

Propagation of infectious yellow head virus particles priorto cytopathic effect in primary lymphoid cell cultures ofPenaeus monodon. Dis Aquat Organ 5555:253-8.

10.Assavalapsakul W, Tirasophon W, and Panyim S (2005)

Antiserum to the gp116 glycoprotein of yellow head virusneutralizes infectivity in primary lymphoid organ cells ofPenaeus monodon. Dis Aquat Organ 6363:85-8.

11.Tenllado F, Martinez-Garcia B, Vargas M and Diaz-Ruiz JR

(2003) Crude extracts of bacterially expressed dsRNA canbe used to protect plants against virus infections. BMCBiotechnol 3:3.

12.Yodmaung S, Tirasophon W, Roshorm Y, Chinnirunwong W

and Panyim S (2006) YHV-protease dsRNA inhibits YHVreplication in Penaeus monodon and prevents mortality. BiochemBiophys Res Commun 341341:351-6.