One-step, zero-background ligation-independent cloning intron-containing hairpin RNA constructs for RNAi in plants

Authors


Author for correspondence:
Yule Liu
Tel: +86 10 62794013
Email: yuleliu@mail.tsinghua.edu.cn

Summary

  • The hairpin-based RNA interference (RNAi) technique plays an important role in exploring gene function in plants. Although there are several methods for making hairpin RNA (hpRNA) constructs, these methods usually need multiple relatively laborious, time-consuming or high-cost cloning steps. Here we describe a one-step, zero-background ligation-independent cloning (OZ-LIC) method for making intron-containing hpRNA (ihpRNA) constructs by our vector pRNAi-LIC.
  • To generate the ihpRNA constructs with zero-background, this method only requires treating two PCR products of target gene flanked with different LIC sequences and SmaI-linearized pRNAi-LIC vector by T4 DNA polymerase respectively, and then transforming these treated DNA mixture into Escherichia coli.
  • The ihpRNA constructs generated with our OZ-LIC RNAi vector can efficiently induce not only transient silencing of the exogenous marker genes and the endogenous resistance-related Nicotiana benthamiana SGT1 gene, but also stable transgenic suppression of Arabidopsis SGT1b gene.
  • Our new OZ-LIC method and RNAi vector will represent a powerful tool for gene knockdown in plants and may facilitate high-throughput determination of plant gene function.

Introduction

With the explosive release of gene sequences, large-scale approaches have been developed for converting this sequence information into functional information. T-DNA and transposon-based insertional mutant populations have provided the resources for the analysis of gene function (AzpirozLeehan & Feldmann, 1997; Martienssen, 1998; Kumar & Hirochika, 2001; Alonso et al., 2003). However, these approaches cannot disrupt or target all genes because of gene target bias, lack of phenotype or loss of insertions that cause lethality, even in model Arabidopsis plants. The function of most plant genes remains unknown.

After double-stranded RNA (dsRNA) was discovered as the trigger of RNA interference (RNAi) (Fire et al., 1998), dsRNA-based RNAi has been widely used to downregulate gene expression in various organisms and has become one of the most powerful tools for gene function analysis. More recently, artificial microRNA technology has also emerged to block gene expression in plants (Niu et al., 2006; Schwab et al., 2006; Ossowski et al., 2008). For plant RNAi, a popular approach is to use vectors to produce intron-containing hairpin RNA (ihpRNA) constructs, as the use of a functional intron in the hairpin sequence has the highest post-transcriptional gene silencing (PTGS) efficiency (Smith et al., 2000). The traditional ligase-based vectors were first used to generate ihpRNA constructs (Smith et al., 2000). The method based on these vectors usually requires several rounds of restriction enzyme digestion and ligation, and is therefore laborious and can only be used on a small scale. In addition to traditional ligase-based vectors, GATEWAY cloning system-based RNAi vectors play an important role in making ihpRNA constructs (Wesley et al., 2001). The GATEWAY ihpRNA vector-based cloning method requires cloning the target gene’s PCR products flanked by attB1 and attB2 sequences into the donor vector (e.g. pDONR207) by the GATEWAY BP reaction to generate an intermediate vector, and then cloning the target gene fragment into GATEWAY destination vectors by the GATEWAY LR reaction. Therefore, it usually needs two cloning steps to generate the final ihpRNA constructs. In addition, GATEWAY ihpRNA vector-based cloning often gives rise to unexpected recombination products with the loss-of-function intron in the antisense orientation with respect to the promoter (Helliwell et al., 2002; Helliwell & Waterhouse, 2003). However, the cost of enzymes used in the GATEWAY system is comparatively high, especially for researchers in developing countries. Recently, several PCR-based methods were developed to generate the ihpRNA cassettes. The DA-ihpRNA method was reported to produce ihpRNA cassettes using the own intron of target genes (Xiao et al., 2006). This method requires amplifying the ihpRNA cassette from the intron-containing genomic DNA which, although simple and rapid, has the limitation of the region of target fragments and still needs more traditional cloning steps to insert the ihpRNA cassettes into the final vectors. This method is unsuitable for making ihpRNA constructs of intronless genes or genes with only expressed sequences tag (EST) sequences available. Recently, a method of one-step hpRNA construction has been developed using the ZeBaTA vector (Chen et al., 2009), and the stem-loop hpRNA structure is obtained via overlapping of three PCR fragments, including two inverted repeat sequences of target genes and a loop fragment. However, when an intron is to be used as spacer, it will take more time to determine the orientation of the intron and pick up the correct clones in order to have the highest silencing efficiency. Mixed one-step overlap extension PCR method has been reported to generate hpRNA cassettes but it requires two more traditional cloning steps to generate the final constructs (Yan et al., 2009). More importantly, all these PCR-based methods are involved in PCR amplification of full-length hpRNA cassettes, including two inverted repeat sequences of target genes. This kind of overlapping PCR is often difficult because of the PCR suppression effect derived from the self-annealing of two inverted repeats (Broude et al., 2001). Therefore, all available ihpRNA construction methods have some limitations.

To develop a simple and fast cloning vector for making RNAi constructs, we have adopted modified ligation-independent cloning (LIC) (Aslanidis & de Jong, 1990; Hsiao, 1993; Aslanidis et al., 1994; Palauqui et al., 1996; Dieckman et al., 2002; Dong et al., 2007) as an approach. The LIC strategy relies on the exonuclease activity of T4 DNA polymerase to generate fairly long (> 12 nt) sticky ends at the ends of the DNA fragments. These long sticky ends allow one to dovetail multiple DNA fragments in a directional manner. The resultant hybrids can be transformed into Escherichia coli without prior nick repair with DNA ligase. Ligation-independent cloning is much faster and more accurate than other cloning strategies because only a single transformation is required, and has been used to make Tobacco rattle virus-based virus induced gene silencing (VIGS) constructs (Dong et al., 2007). Here, we describe a rapid, one-step and zero background LIC (OZ-LIC) method for making ihpRNA constructs via our new plant RNAi vector pRNAi-LIC, with which we can readily construct ihpRNA for plant RNA silencing by one-step transformation, bypassing traditional restriction site limitation and ligase use. We have successfully applied this method to make 10 ihpRNA constructs with stems ranging from 300 bp to 700 bp in length, and achieved 100% positive cloning efficiency. In addition, the ihpRNA constructs generated with our vector can efficiently induce both transient silencing and stable transgenic suppression. These results suggest that our OZ-LIC method is reliable and reproducible for making ihpRNA constructs to facilitate the high-throughput gene functional study in plant research.

Materials and Methods

Plant and plasmid materials

Transgenic Nicotiana benthamiana plants containing the N gene and TMV-GFP (GFP tagged Tobacco mosaic virus) have been described previously and used for analysis of TMV resistance (Liu et al., 2002a). Arabidopsis thaliana Col-0 ecotype was used for plant transformation. Wild-type N. benthamiana was used for other functional assays. pYL41 is a pCAMBIA2300-based T-DNA vector which contains the duplicated CaMV 35S promoter and Nos terminator from pYL44 (Dong et al., 2007); pSAH18 is a -plasmid containing the ccdB gene in which the SmaI site in the original ccdB gene is removed without changing its amino acid sequence (A.H. Sha and Y.L. Liu, unpublished).

Plant RNAi vector pRNAi-LIC

pRNAi-LIC vector was generated using pYL41 as the skeleton. First, ccdB gene was amplified using pSAH18 as template and primers 5′-CC gagctc TAG AGC ACA CGA ccc ggg CTG TGT ATA AGG GAG C-3′ (5′-SacI-LIC4-SmaI-ccdB-3′) and 5′-CGG ggtacc GGG CCC TGA GGA GAA GAG ccc ggg AAT TCT CGA CTA AGT-3′ (5′-KpnI-LIC2-SmaI -ccdB-3′) and digested with SacI and KpnI and then cloned into vector pYL41 to make pYL41-ccdB. Second, the Pdk intron-containing chloramphenicol-resistance gene was PCR amplified using pHellsgate (Wesley et al., 2001) as template and primers 5′-CGG ggatcc gagctc GAC GAC AAG Acc cgg gTT CCG TGC TGG AAc caa CTG TAA TCA ATC CAA ATG (5′-BamHI–SacI–LIC1–SmaI–LIC3c-Pdk intron-3′) and 5′-CGG gggccc ttg GTA AGG AAA TAA TTA TTT TCT TTT TTC C-3′ (5′-ApaI–Pdk intron-3′) and then cloned into T-vector pEASY-T1 (Transgene, Beijing, China) to generate pPDK. There were two ApaI sites (one was introduced via primer while the other was in the vector pEASY-T1) in pPDK. Finally, the Pdk intron fragment of pPDK was ligated into ApaI site of pYL41-ccdB to generate pRNAi-LIC. pRNAi-LIC was used to make all ihpRNA constructs for silencing of the genes used in this manuscript, and was maintained in E. coli strain DB3.1, in which the ccdB gene is not lethal. The full-length sequence and detailed annotation of pRNAi-LIC was deposited in GeneBank (access number GQ870263; stock name in ABRC is pRNAi-LIC).

Ligation independent cloning procedure for making ihpRNA constructs

To make the ihpRNA construct for silencing of the gene of interest, the same target region is amplified twice by different oligo pairs. The first PCR product is amplified by the first oligo pair: 5′-CGA CGA CAA GAC CCT- gene specific forward primer-3′ and 5′-GAG GAG AAG AGC CCT- gene specific reverse primer-3′ (the underlined sequences are adaptors LIC1 and LIC2 respectively) and called ‘PCR Product 1’. The second PCR product is called ‘PCR Product 2’ and amplified by using purified PCR Product 1 as template and the universal oligo pair LIC3-TT-LIC2 (5′-CCA GCA CGG AAC CCT TGA GGA GAA GAG CCC T-3′) and LIC4-TT-LIC1 (5′-AGA GCA CAC GAC CCT TCG ACG ACA AGA CCC T-3′) (the underlined sequences are adaptors LIC3 and LIC4 respectively), which anneal to the end sequences of PCR Product 1 respectively. Alternately, to save time, PCR Product 2 of the same target region can be amplified by the third oligo pair: 5′-LIC3-TT-gene specific reverse primer-3′ and 5′-LIC4-TT-gene specific forward primer-3′. The PCR products were purified by polyethylene glycol (PEG)–MgCl2 precipitation as described previously (Dong et al., 2007) or by using a Gel purification kit in case there are nonspecific bands in addition to target fragments in the PCR products. Fifty nanogram samples of purified PCR Product 1 and 2 were treated with 0.5 Units T4 DNA polymerase (Fermentas, Beijing, China) in 1× reaction buffer and 5 mM dATP at 22°C for 30 min following 20 min of inactivation of T4 DNA polymerase at 75°C. The pRNAi-LIC vector was digested with SmaI and similarly treated with T4 DNA polymerase but with dTTP replacing dATP. Fifty nanogram samples of treated PCR Product 1 and 2 and treated pRNAi-LIC vector were mixed and incubated at 22°C for 15–30 min. Then the mixture was transformed into E. coli DH5α competent Cells (Tiangen, Beijing, China) and plated on Luria–Bertani (LB) medium containing kanamycin (25 mg l−1) and chloramphenicol (5 mg l−1) to screen the recombinants.

Gene silencing vectors

The ihpRNA constructs for silencing GFP, GUS, NbSGT1 and AtSGT1b genes were generated as described earlier and named pRNAi-GFP, pRNAi-GUS, pRNAi-NbSGT1 and pRNAi-AtSGT1b, respectively. The PCR primers used to generate PCR Product 1 for making these ihpRNA constructs are OX1 and OX2 (for GFP; AF007834), OX3 and OX4 (for GUS; AY292368), OX5 and OX6 (for NbSGT1; AY899199), and OX7 and OX8 (for AtSGT1b; At4g11260). The vectors for silencing the other six genes were generated similarly and these six genes include Arabidopsis TIR-NBS-LRR protein-like gene (AtTNL; At4g19520), tobacco ERD2-like genes NbERD2a (GU388433) and NbERD2b (GU388432), cotton MSH1-like gene (GhMSH1; TC16563) and coat protein genes of ORSV (ORSV-CP; EU683879) and CyMV (CyMV-CP; FJ858205). These constructs were confirmed by PCR, restriction enzyme digestion and partial DNA sequencing. All primers used in this study are listed in Supporting Information Table S1.

Agroinfiltration

Nicotiana benthamiana plants were grown in pots at 25°C in a growth chamber under 16 h light : 8 h dark cycle. All ihpRNA constructs were transformed into Agrobacterium strain GV3101 and screened on LB plates containing kanamycin (50 μg ml−1). Agrobacterium was grown in LB at 28°C for 2 d. The cultures were spun down at 2500 g for 5 min and cells were resuspended in infiltration medium (10 mM 2-(N-morpholino) ethanesulfonic acid (MES), 10 mM MgCl2, 200 μM acetosyringone), adjusted to OD600 = 1.0 and incubated at room temperature for 4 h before agroinfiltration. Agroinfiltration of N. benthamiana plants were performed with a 1-ml needleless syringe as described previously (Liu et al., 2002a). For TMV resistance analysis, Agrobacterium culture containing pRNAi-NbSGT1 or pRNAi-LIC vector alone were mixed with Agrobacterium culture containing TMV-GFP at the ratio of 100 : 1 or 1 : 1, and then infiltrated into the leaves of N-containing N. benthamiana plants. The existence of TMV-GFP was observed under UV light. All the experiments were repeated three times and three plants were used for each constructs per time.

GFP imaging

Green fluorescent protein fluorescence was detected using a long-wavelength UV lamp (Black Ray model B 100A; UV products; Upland, CA, USA) and photographs were taken using CannonA630 digital camera.

GUS staining

Glucuronidase (GUS) staining was performed as reported elsewhere with a slight modification (Jefferson et al., 1987). The infiltrated leaves were detached 4 d post-agroinfiltration and put into GUS staining buffer (50 mM phosphate buffer (pH7.0), 10 mM Na2EDTA, 5 mM K3Fe(CN)6, 5 mM K4Fe(CN)6, 0.1% X-Gluc) and infiltrated under vacuum at 0.1 Mpa for 5 min. The leaves were then transferred to the 37°C incubator in the dark. When the blue color appeared in the leaves, the leaves were washed in 50 mM phosphate buffer (pH7.0), 50%, 75% and 100% ethanol sequentially for 5 min each and then placed into 75% ethanol to decolorize to the ideal state.

Plant transformation and analysis of Arabidopsis response to auxin

Arabidopsis ecotype Col-0 was transformed with pRNAi-AtSGT1b using the floral dip method descried by Zhang et al. (Zhang et al., 2006). Transformants were screened on selective medium containing 50 mg l−1 kanamycin to obtaine primary transgenic plants, referred to as T1 plants. The presence of AtSGT1b ihpRNA expression cassette in transgenic plants was confirmed by PCR using primers OX7 and OX13. For analysis of Arabidopsis response to auxin, the seeds of Arabidopsis T1 plants were germinated on Murashige and Skoog (MS) plates containing 0.06 μM 2, 4-Dichlorophenoxyacetic acid (2, 4-D).

RNA isolation and RT-PCR analysis

Total RNA was extracted at different time-points from leaves of N. benthamiana or Arabidopsis plants using TRNzol solution (Tiangen) and treated by RNase-free DNase I (Fermentas) to remove the potential DNA contamination. First-strand cDNA was synthesized using 1 μg of total RNA, Oligo(dT)15 or gene-specific primers and M-MuLV Reverse Transcriptase (Fermentas) according to the manufacturer’s protocol. Semiquantitative reverse-transcription polymerase chain reaction (RT-PCR) was performed as described in Liu et al. (2002a). For RT-PCR, primers that anneal outside the target region for silencing were used to ensure that the endogenous genes were tested without interference from the T-DNA insertion containing the target genes. The RT-PCR primers were OX9 and OX10 (for GFP), OX11 and OX4 (for GUS), OX5 and OX12 (for NbSGT1), OX15 and OX16 (for NbActin, used as a control reference of N. benthamiana RNA), TMV-1 and TMV-2 (for TMV detection), OX7 and OX17 (for AtSGT1b), and OX18 and OX19 (for AtTubulin, used as a control reference for Arabidopsis RNA). Sequences of these primers are listed in Table S1. The leaves were collected for RNA extraction from N. benthamiana 3 d post-agroinfiltrated with pRNAi-GFP and pRNAi-GUS, and 7 d post-agroinfiltrated with pRNAi-NbSGT1. Whole Arabidopsis plants were used for RNA isolation. Each experiment was repeated three times.

Results

Development of a LIC strategy for making ihpRNA constructs

To adopt a LIC strategy for the simple and cost-effective cloning of the intron-containing inverted repeat inserts into the vectors, we developed a new plant T-DNA RNAi vector, pRNAi-LIC. In pRNAi-LIC, there are four different adaptors (LIC1–LIC4), Pdk intron and ccdB gene between the duplicated CaMV 35S promoter and Nos terminator. The Pdk intron-containing chloramphenicol resistance gene is placed between LIC3 and LIC2, and ccdB gene resides between LIC2 and LIC4; there are three SmaI sites between LIC1 and LIC3, between LIC2 and ccdB , and between ccdB and LIC4 (Fig. 1a). Four 14 bp adaptors (LIC1–LIC4) are exposed after digestion of pRNAi-LIC with SmaI and used to generate the sticky ends at the ends of both the linearized vector and Pdk intron fragments by T4 DNA polymerase. The chloramphenicol resistance gene inside the Pdk intron only allows the growth of the recombinant colonies containing the Pdk intron, and the ccdB gene allows us to screen putative recombinant colonies rapidly and accurately, a strategy employed by the widely-used GATEWAY technology. To make the ihpRNA construct for gene silencing, the target region of the gene of interest is PCR amplified twice to obtain two PCR products with different ends. PCR Product 1 is obtained by using gene-specific primers and any target gene-containing DNA as the template. As the LIC1 sequence is included in the forward primer and LIC2 sequence is included in the reverse primer, PCR Product 1 has terminal LIC1 and LIC2 sequences (Fig. 1b). PCR Product 2 is obtained by using PCR Product 1 as the template and the universal PCR primer pair LIC4-TT-LIC1 and LIC3-TT-LIC2. Thus, PCR Product 2 has terminal LIC3 and LIC4 sequences (Fig. 1b). Digestion of pRNAi-LIC with SmaI generates the Pdk intron fragments flanked by adaptors LIC2 and LIC3 and vector backbone fragments flanked by adaptors LIC1 and LIC4, and subsequent treatment with T4 DNA polymerase generates sticky ends. Both PCR Product 1 and PCR Product 2 carrying the LIC adaptor sequences are similarly treated with T4 DNA polymerase, resulting in sticky ends complementary to the according sticky LIC adaptor sequences on the treated vector and Pdk intron. Incubation of treated PCR products and treated linearized vector and the Pdk intron can form strong bonds via hydrogen bonds (Fig. 1b; see the Materials and Methods section for details) and the mixture can be transformed directly into E. coli without the ligation step (see the Materials and Methods section for details). Gaps can be ligated automatically in E. coli. Our new plant ihpRNA vector pRNAi-LIC can thus be used for fast cloning because any number of DNA sequences carrying compatible ends can be combined with pRNAi-LIC.

Figure 1.

 Schematic diagram of intron-containing hairpin RNA (ihpRNA) construction. (a) Plant ihpRNA vector pRNAi-ligation-independent cloning (LIC). The chloramphenicol resistance gene (Cmr) containing Pdk intron, a ccdB expression cassette, four adaptors (LIC1–LIC4) and three SmaI sites are cloned between the duplicated 35S CaMV promoter and Nos terminator of pYL41, a pCAMBIA2300-based T-DNA vector. (b) LIC of ihpRNA constructs. The target fragment of the gene of interest is PCR amplified using gene-specific primers carrying the adaptors LIC1 and LIC2 to obtain PCR Product 1, which has terminal sequences LIC1 and LIC2. PCR Product 2 is amplified by using PCR Product 1 as template and universal primers (LIC4-TT-LIC1 and LIC3-TT-LIC2) and thus has terminal sequences LIC3 and LIC4. PCR Products 1 and 2 are further treated with T4 DNA polymerase in the presence of dATP. pRNAi-LIC vector is digested with SmaI and treated with T4 DNA polymerase at the presence of dTTP. The treated vector and two PCR products with sticky ends are mixed and then transformed into Escherichia coli cells, such as DH5α. Primers OX13 and OX14 are marked and used for partial DNA sequencing. SacI and BamHI are indicated and used for the identification by restriction enzyme digestion.

Fast and zero-background cloning of plant ihpRNA constructs

In order to demonstrate that our new plant RNAi vector pRNAi-LIC can be used to make ihpRNA constructs easily with the OZ-LIC method, we used PCR to obtain the corresponding PCR Product 1 and PCR Product 2 of 10 genes including N. benthamiana SGT1 (NbSGT1), Arabidopsis SGT1 (AtSGT1b) and two marker genes (GFP and GUS) using gene-specific and universal primer sets respectively (see the Materials and Methods section for details). The gene-specific primer set was used to obtain PCR Product 1 and the universal primer set (LIC4-TT-LIC1 and LIC3-TT-LIC2) was used to obtain PCR Product 2 by using PCR Product 1 as a template. PCR Product 1 (with terminal sequences LIC1 and LIC2) and PCR Product 2 (with terminal sequences LIC3 and LIC4) were treated with T4 DNA polymerase and further mixed with SmaI-linearized pRNAi-LIC vector (containing Pdk intron) similarly treated with T4 DNA polymerase. After incubation at 22°C for 15 min, this mixture was transformed into E. coli DH5α cells and transformants were selected on LB plates containing both kanamycin and chloramphenicol. We generated 10 ihpRNA constructs for silencing, with target regions ranging in size from 300 bp to 700 bp. The ihpRNA constructs for silencing GFP, GUS, NbSGT1 and AtSGT1b genes were named pRNAi-GFP, pRNAi-GUS, pRNAi-NbSGT1 and pRNAi-AtSGT1b, respectively. These constructs were verified by double digestion of SacI and BamHI (Fig. 2a) and by sequencing the vector–insert junctions and target gene fragment–Pdk intron junctions using primer OX13 and primer OX14. Double digestion of the right recombinant plasmids with SacI and BamHI typically results in a 9.7 Kb vector fragment and two more DNA fragments – one c. 622 bp plus size of target fragment and the other c. 1053 bp plus size of target fragment – however, if there are BamHI or SacI sites in the target gene fragment, more bands will appear (Fig. 2a, lanes 1 and 5). In our hands, this OZ-LIC protocol for making ihpRNA constructs had a 100% success rate. We tested 10 transformants for each ihpRNA construct by PCR and restriction digestion, and all transformants gave the predicted pattern. For example, all 10 clones of pRNAi-AtSGT1b gave the predicted PCR bands (Fig. 2b) and the predicted digestion pattern (Fig. 2c). These results suggest that our new plant ihpRNA vector pRNAi-LIC can be used to make ihpRNA constructs with zero background by the OZ-LIC method. It only takes c. 1 h to start E. coli transformation step to get the ihpRNA constructs once PCR products are purified. This OZ-LIC method is thus simple and fast, and may facilitate high-throughput cloning of ihpRNA constructs for large scale gene functional analysis.

Figure 2.

 Zero-background cloning of intron-containing hairpin RNA (ihpRNA) constructs. (a) The pattern of double digestion by BamHI and SacI of 10 ihpRNA constructs for silencing GFP, GUS, ORSV-CP, CyMV-CP, GhMSH1, NbSGT1, NbERD2a, NbERD2b, AtSGT1b and AtTNL (Lanes 1–0 in order). All constructs gave the predicted bands. Lanes 1 and 5 have one more overlapped band because of the existence of one BamHI site in GFP and GhMSH1 target region. The cloning efficiency into pRNAi-ligation-independent cloning (LIC) is zero-background, as shown by colony PCR (b) and by double digestion with BamHI and SacI on colonies obtained from transformants of pRNAi-AtSGT1b (c). All the colonies tested here contain the correct inserts. Lanes 1–10 in (b) represent 10 independent colonies, Lanes 0 and 11 are negative and positive control, respectively. M is a DNA size marker.

Silencing of two marker genes by transient expression of ihpRNA constructs

Coinfiltration is a transient system and has been widely used to determine the functionality of RNAi constructs for silencing of marker genes (Johansen & Carrington, 2001; Llave et al., 2002; Wang et al., 2005; Carvalho et al., 2008). In order to demonstrate that the ihpRNA constructs described above can be used for gene silencing, we first tested the ability of pRNAi-GFP for silencing GFP gene by Agrobacterium-mediated transient expression. Agrobacterium cultures containing pCAMBIA1302 were mixed with Agro-bacterium containing the pRNAi-LIC vector or pRNAi-GFP, respectively, and infiltrated into different parts of the same leaves of N. benthamiana plants. When the control vector pRNAi-LIC and pCAMBIA1302, a T-DNA vector that contains GFP gene driven by the constitutive CaMV 35S promoter, were used together, the infiltration areas gave strong green fluorescence under UV light (Fig. 3a; left). When pRNAi-GFP and pCAMBIA1302 were used together, however, the infiltration areas gave much weaker GFP fluorescence (Fig. 3a; right). To confirm the knockdown of the GFP gene at the molecular level, we performed semiquantitative RT-PCR. In the pRNAi-GFP infiltrated leaf areas, GFP mRNA level was reduced compared with the control infiltrated with pRNAi-LIC vector alone (Fig. 3c). In both RNA samples, RNA levels of endogenous actin were similar (Fig. 3c) and served as an internal control. We also tested the ability of pRNAi-GUS for silencing GUS gene similarly. Agrobacterium cultures containing pCAMBIA1301 were mixed with Agrobacterium containing pRNAi-LIC vector or pRNAi-GUS, respectively, and infiltrated into different parts of the same leaves of N. benthamiana plants. The infiltrated leaves were detached for GUS staining 3 d post infiltration. When control vector pRNAi-LIC and pCAMBIA1301, a T-DNA vector that expresses GUS gene driven by CaMV 35S promoter were used together, the infiltration areas gave a strong GUS-staining blue signal (Fig. 3b; left); by contrast, when pRNAi-GUS and pCAMBIA1301 were used together, the infiltration areas gave a much weaker blue signal (Fig. 3b; right). To confirm the knockdown of GUS gene at the molecular level, semiquantitative RT-PCR was performed. In the pRNAi-GUS infiltrated leaf areas, GUS mRNA level was reduced compared with the control infiltrated with pRNAi-LIC vector alone (Fig. 3d). In both RNA samples, actin RNA levels were similar (Fig. 3d) and served as an internal control. These results suggest that the ihpRNA constructs generated with our pRNAi-LIC vector can efficiently knockdown the expression of genes, at least marker genes.

Figure 3.

 Silencing of two marker genes GFP and GUS by transient expression of the intron-containing hairpin RNA (ihpRNA) constructs. (a) Silencing effect of green fluorescent protein (GFP) ihpRNA on transient expression of GFP gene. Agrobacterium cultures containing pCAMBIA1302 (35S-GFP) were mixed with Agrobacterium containing pRNAi-ligation-independent cloning (LIC) vector (left) or pRNAi-GFP (right), respectively and infiltrated into different parts of the same leaves of Nicotiana benthamiana plants. The pictures were taken 3 d post agroinfiltration under UV light. Agroinfiltration with pRNAi-GFP resulted in weaker GFP fluorescence compared with that with vector alone. (b) Silencing effect of GUS ihpRNA on transient expression of GUS gene. Agrobacterium cultures containing pCAMBIA1301 (35S-GUS) were mixed with Agrobacterium containing pRNAi-LIC vector (left) or pRNAi-GUS (right), respectively, and infiltrated into different parts of the same leaves of N. benthamiana plants. The GUS staining was performed 3 d post agroinfiltration. Agroinfiltration with pRNAi-GUS resulted in much weaker blue signal compared with that with vector alone. (c,d) Reverse-transcription polymerase chain reaction (RT-PCR) analysis was performed using the leaf tissues agroinfiltrated by GFP and GUS ihpRNA constructs to confirm the silencing of GFP (c) and GUS (d) genes. The first-strand cDNA was generated from total RNA isolated from the leaf tissues agroinfiltrated with pRNAi-LIC vector alone or ihpRNA constructs. The amount of GFP and GUS RNA was reduced in the leaf tissue infiltrated with pRNAi-GFP (c, bottom lane 2) and pRNAi-GUS (d, bottom, lane 2) compared with that infiltrated with pRNAi-LIC vector (lane 1). The amount of actin RNAwas similar and used was as an internal control (c and d, upper). Lane 1, the leaf area agroinfiltrated with pRNAi-LIC vector alone; lane 2, the leaf area agroinfiltrated with pRNAi-GFP (c, lane 2) and pRNAi-GUS (d, lane 2); M is a DNA size marker.

Systemic silencing of NbSGT1 by agroinfiltration

In order to demonstrate that our ihpRNA constructs can be used for silencing corresponding endogenous genes, we tested whether pRNAi-NbSGT1 can inhibit the expression of the NbSGT1 gene in N-containing N. benthamiana plants. SGT1 (suppressor of G(2) allele of skp1) has been reported to participate in diverse physiological processes, including cell cycle progression in yeast, plant defense against pathogens and plant hormone signaling (Stuttmann et al., 2008). Virus induced gene silencing of NbSGT1 in N-containing transgenic plants compromises N-mediated resistance to TMV (Liu et al., 2002b; Peart et al., 2002). Agrobacterium cultures containing TMV-GFP were mixed with Agrobacterium containing either control pRNAi-LIC vector or pRNAi-NbSGT1 at 1 : 100 ratio (OD600) respectively and infiltrated into different parts of the same leaves of N-containing N. benthamiana plants. In the pRNAi-NbSGT1-infiltrated leaf area (Fig. 4a; right), there was stronger green fluorescence of GFP compared with that in the pRNAi-LIC vector-infiltrated leaf area (Fig. 4a; left), indicating there was more TMV-GFP in the pRNAi-NbSGT1-infiltrated leaf area. Surprisingly, TMV-GFP did spread to the upper uninfiltrated leaves 13 d postinfiltration (data not shown), suggesting that the agroinfiltration by pRNAi-NbSGT1 may result in systemic silencing of NbSGT1 gene. To check this hypothesis, Agrobacterium cultures containing TMV-GFP were mixed with Agrobacterium containing either pRNAi-LIC vector or pRNAi-NbSGT1 at 1 : 1 ratio (OD600), respectively, and infiltrated into the leaves of different N-containing N. benthamiana plants. In pRNAi-LIC vector infiltrated plants, TMV-GFP did not spread to the upper leaves (Fig. 4b; left). By contrast, TMV-GFP overcame N-mediated resistance and did spread to the upper leaves in the pRNAi-NbSGT1 infiltrated plants (Fig. 4b; right). The existence of TMV-GFP in the upper leaves of pRNAi-NbSGT1 infiltrated plants was confirmed by RT-PCR (Fig. 4c). To check whether the loss of TMV resistance in pRNAi-NbSGT1 infiltrated plants was caused by the silencing of NbSGT1, we performed RT-PCR. In pRNAi-NbSGT1 infiltrated plants, NbSGT1 RNA level was reduced greatly compared with that in pRNAi-LIC vector infiltrated plants (Fig. 4d). In both leaf tissue RNA samples, actin RNA levels were similar (Fig. 4d) and served as an internal control. These results indicated that more TMV accumulation and the loss of TMV resistance in pRNAi-NbSGT1 infiltrated area or plants were indeed caused by the knocking down of NbSGT1. These results also suggest that the ihpRNA constructs generated with our pRNAi-LIC vector can efficiently silence the corresponding endogenous genes, at least at transiently.

Figure 4.

 Agroinfiltration with pRNAi-NbSGT1 abolished N-mediated resistance to Tobacco mosaic virus (TMV). (a) Silencing effect of NbSGT1 intron-containing hairpin RNA (ihpRNA) on N-mediated resistance to TMV at the agroinfiltrated leaves. Agrobacterium cultures containing TMV-green fluorescent protein (GFP) were mixed with Agrobacterium containing pRNAi-ligation-independent cloning (LIC) vector (left) or pRNAi-NbSGT1 (right) at the ratio of 1 : 100 (OD600), respectively, and infiltrated into different parts of the same leaves of Nicotiana benthamina (NN) plants containing the N. The pictures were taken 7 d post agroinfiltration under UV light. Agroinfiltration with pRNAi-NbSGT1 resulted in more TMV-GFP accumulation compared to that with the vector alone. (b) Agroinfiltration with pRNAi-NbSGT1 abolished N-mediated resistance to TMV. Agrobacterium cultures containing TMV-GFP were mixed with Agrobacterium containing pRNAi-LIC vector (left) or pRNAi-NbSGT1 (right) at the ratio of 1 : 1, respectively, and infiltrated into different NN plants. Pictures were taken from upper leaves under UV illumination 15 d post agroinfiltration. Movement of TMV-GFP from the inoculated leaf into the upper leaves indicates loss of resistance to TMV. (c) Reverse-transcription polymerase chain reaction (RT-PCR) to conform the existence of TMV in the upper leaves. RT-PCR was performed using TMV MP gene-specific primers. Lane 1, pRNAi-LIC vector agroinfiltrated NN plant; lane 2, pRNAi-NbSGT1 agroinfiltrated NN plant. M is a DNA marker. (d) RT-PCR to confirm the silencing of NbSGT1 in the upper leaves of the agroinfiltrated plants. Total RNA was used to generate first-strand cDNA with oligo(dT)15. This cDNA was used in a PCR to amplify NbSGT1 fragment outside of the target region. Lane 1, pRNAi-LIC vector agroinfiltrated NN plant; lane 2, pRNAi-NbSGT1 agroinfiltrated NN plant. M is a DNA marker.

Silencing of AtSGT1b by expressing ihpRNA construct in transgenic plants

Arabidopsis with AtSGT1b mutation is more insensitive to auxin (Gray et al., 2003). In order to demonstrate that our ihpRNA constructs can be used for silencing corresponding endogenous genes in stable transgenic plants, we transformed Arabidopsis plants with pRNAi-AtSGT1b to obtain AtSGT1b ihpRNA plants. The seeds of Arabidopsis T1 plants were germinated in MS plates containing 0.06 μM 2,4-D, synthetic auxin. Silencing of AtSGT1b conferred auxin-resistant root growth for eight of nine independent transgenic lines tested. As shown in Fig. 5a, the progenies of the transgenic line 4 showed segregation for normal short and long root when grown on 2,4-D medium, whereas all wild-type plants showed a short root on this medium. We performed PCR to test whether the short root phenotype in the progenies of transgenic line 4 results from the absence of transgene derived from genetic segregation. The PCR results showed that there is no AtSGT1b ihpRNA transgene in the genomes of the two plants with the shorter root phenotype (Fig. 5b, lanes 1 and 6). We also performed RT-PCR to check the amount of AtSGT1b mRNA. In all plants, the amounts of actin RNA were similar (Fig. 5c) and served as an internal control. Among wild-type plants and two progenies of transgenic line 4 with shorter roots, the amounts of AtSGT1b mRNA were similar (Fig. 5c, lanes WT, 1 and 6). In the AtSGT1b ihpRNA transgenic plants with the longer root phenotype, the amounts of AtSGT1b mRNA were reduced greatly compared with those in the control wild-type plant (Fig. 5c, lanes 2–5). These results suggest that the ihpRNA constructs generated with our pRNAi-LIC vector can efficiently silence the corresponding endogenous genes in stable transgenic plants.

Figure 5.

 Phenotype of AtSGT1b intron-containing hairpin RNA (ihpRNA) transgenic plants response to 2,4-D. (a) All wild-type (WT) Arabidopsis control plants are sensitive to 2,4-D and have shorter roots (left). However, some progenies of AtSGT1b ihpRNA transgenic T1 plant line 4 are insensitive to 2,4-D and have longer roots (plants 2–5), and some are sensitive (plants 1 and 6) on the Murashige and Skoog (MS) plates containing 0.06 μM 2,4-D because of normal segregation (right). (b) A PCR to check the existence of transgene in the progenies of the AtSGT1b ihpRNA transgenic line 4. The segregation at DNA level is shown by PCR using total DNA as templates. (c) Reverse-transcription polymerase chain reaction (RT-PCR) to confirm the silencing of AtSGT1b in the AtSGT1b ihpRNA transgenic plants insensitive to 2,4-D. AtSGT1b mRNA levels of the insensitive plants (lanes 2–5) were greatly reduced compared with that of WT plant and plants (lanes 1, 6) sensitive to 2,4-D. M is a DNA marker.

Discussion

In this work, we developed a new plant RNAi vector to make ihpRNA constructs with zero background, only needing one ligation independent cloning step. We have demonstrated that the ihpRNA constructs generated with this RNAi vector based on OZ-LIC method can efficiently induce both transient silencing and stable transgenic suppression of target genes. The OZ-LIC method and our plant RNAi vector may facilitate high-throughput determination of plant gene function.

With the rapid release of various organisms’ genomic and EST sequences, there is an urgent need to develop an easy, cost-effective and high-throughput reverse genetics method for functional genomics study. One of the most powerful and popular tools in reverse genetics is ihpRNA-induced RNA silencing. In the field of plant science, several vectors are available to generate ihpRNA constructs using different strategies. For some vectors such as pHANNIBAL (Helliwell & Waterhouse, 2003), the ihpRNA cloning procedure requires several rounds of restriction enzyme digestion and ligation and is quite laborious and time-consuming. GATEWAY-based ihpRNA cloning procedure is relatively easy and has been widely used (Wesley et al., 2001). Compared with GATEWAY-based ihpRNA vectors, our OZ-LIC based ihpRNA vector has many obvious advantages. First, it only requires a single transformation step to generate final ihpRNA constructs, while the GATEWAY-based ihpRNA cloning method usually needs two cloning steps. Therefore, the OZ-LIC method will be faster and more cost effective. Second, the enzymes such as BP and LR Clonases used in the GATEWAY-based method are comparatively expensive. However, T4 DNA polymerase used in the LIC method is relatively cheap. Third, with our OZ-LIC based ihpRNA vector, we can achieve almost zero-background cloning efficiency. However, with GATEWAY-based ihpRNA vectors such as pHellsgate 8, two major categories of ihpRNA constructs are usually produced and only some transformants are desired, which means more colonies have to be further screened. Fourth, OZ-LIC method uses gene-specific primers with shorter adaptors (LIC adaptors LIC1/LIC2 are 14 nt vs GATEWAY adaptors attB1/attB2 are 29 nt), and one universal primer set can be used to amplify PCR Product 2 of any target genes. This means that the OZ-LIC method has a lower primer cost, especially for making a number of ihpRNA constructs. Recently, some methods were reported to generate the hpRNA cassettes via overlapping PCR (Xiao et al., 2006; Chen et al., 2009; Yan et al., 2009), but PCR failure may occur during amplification of the full-length hpRNA cassettes because of self-annealing of two inverted repeat fragments. In addition, the hpRNA cassettes still need to be cloned into the final vectors by one or two traditional cloning steps in these methods. By contrast, our OZ-LIC method requires simple PCR amplification and only needs a single-step transformation without prior ligation. Therefore, our OZ-LIC method overcomes almost all the shortcomings associated with the available vectors, and is a simpler, faster, less laborious and more cost-effective way to make ihpRNA constructs than the existing methods.

The introduction of an intron as the loop for ihpRNA constructs not only overcomes the cloning difficulty that most bacteria cannot tolerate plasmids with an inverted repeat, but can also induce PTGS with almost 100% efficiency (Smith et al., 2000). We found that including an antibiotic gene inside the intron can dramatically increase the cloning efficiency. When using the Pdk intron without the antibiotic gene as the loop region in the vector, we always got a high percentage of false-positive transformants, as reported previously (Conley et al., 1986; Wesley et al., 2001). To overcome this problem, we used the chloramphenicol-resistance gene containing the Pdk intron as the loop region in the vector, as described previously (Wesley et al., 2001). We found that the use of the ethanol precipitation-purified SmaI-linearized vector slightly increased the transformation efficiency. To date, we have successfully generated ihpRNA constructs for 10 target genes with a 100% success rate. This method can be easily manipulated even by novices to molecular cloning. This vector has been deposited into ABRC and the detailed protocol is described in the Supporting Information, Notes S1.

We verified the silencing effect of the ihpRNA constructs of two marker genes (GFP and GUS) and an endogenous defense-related gene, N. benthamiana SGT1 by Agrobacterium-mediated transient expression. Agroinfiltra-tion with all ihpRNA constructs induced gene silencing locally and the corresponding RNA level was reduced. Surprisingly, agroinfiltration with pRNAi-NbSGT1 resulted in systemic silencing of NbSGT1 and the loss of N-mediated resistance to TMV (Fig. 4c), confirming that the effect of siRNA generated by dsRNA is noncell autonomous (Palauqui & Vaucheret, 1995; Palauqui et al., 1996; Voinnet, 2005). Furthermore, we successfully silenced AtSGT1b in transgenic Arabidopsis plants transformed with pRNAi-AtSGT1b. As expected, the silencing of AtSGT1b conferred auxin-resistant root growth, resulting in longer roots in the MS plates containing 0.06 μM 2,4-D. Eight of nine tested AtSGT1b ihpRNA transgenic lines showed the expected phenotype, suggesting that PTGS efficiency induced by our RNAi vector (c. 89%) is similar to that induced by the well-documented vectors such as pHANNIBAL and pHellsgate 8 or 12 (Wesley et al., 2001). These results indicated that the ihpRNA constructs generated with our new RNAi vector are effective for gene silencing.

Using our OZ-LIC ihpRNA method, we can readily make ihpRNA constructs in one ligation-independent cloning step. This protocol only needs a simple PCR reaction, treatment by T4 DNA polymerase and a single transformation step to make ihpRNA constructs with almost zero-background cloning efficiency. It also bypasses either restriction enzyme limitation or ligation requirement. All the advantages may enable it to be used for making ihpRNA constructs of plant genes in a high-throughput manner to silence any genes, especially genes that are only partially sequenced or where there are no available T-DNA mutants (in the case of Arabidopsis). The adoption of the OZ-LIC method for chemically induced expression of ihpRNA may be useful to silence the genes whose knockout causes lethality in plants. In addition, LIC strategy and our pRNAi-LIC vector may also be useful for making constructs for expression of genes, including artificial microRNAs, by adding adaptors LIC 1 and LIC 4 at their ends by PCR. In mammalian and Drosophila systems, ihpRNA construct-mediated RNA silencing is also one of the main approaches for gene silencing and similar LIC strategy may be adopted to make ihpRNA constructs for silencing in those systems.

Acknowledgements

The authors thank Amy Lelyveld for thoughtful comments and critical reading of the manuscript. We are grateful to CAMBIA (Canberra, ACT, Australia) for T-DNA binary vectors pCAMBIA1301 and pCAMBIA1302, to Dr Peter Waterhouse for pHellsgate, to Dr S. P. Dinesh-Kumar for the seeds of N-containing N. benthamiana plants and TMV-GFP construct. This work is supported by the National Natural Science Foundation of China (Grant no. 30725002, 30930060); the National Transgenic Program (Grant no. 2008ZX08009-003, 2008ZX08005-001); and the National High-Tech (863) Program of China (Grant no. 2007AA10Z412).

Ancillary