High doses of siRNAs induce eri-1 and adar-1 gene expression and reduce the efficiency of RNA interference in the mouse

Construction of dsRNA-expressing plasmids

To express dsRNA in E. coli, a pET2P plasmid was constructed by positioning a tac promoter into pET-22b [20] in an opposite direction to the T7 promoter separated by multiple cloning sites. The DNA fragment encoding the second half of the HBVP (type B hepatitis virus, subtype adr, polymerase) gene was PCR-amplified with the following primers: HBVP-sense 5′-GGAATTCGTCTTGGGTATACATTTGACC-3′ (the introduced EcoRI site is underlined) and HBVP-antisense 5′-GGGGTACCAGAGGACAACAGAGTTG-3′ (the introduced KpnI site is underlined). The DNA fragment containing exon 2 and exon 3 of seven mouse eri-1 exons (MERI-1) (GenBank® accession number NM_026067) was amplified from mouse cDNA with following primers: eri-1-sense 5′-CGGAATTCGCAGACTTGAT-3′ (the introduced EcoRI site is underlined) and eri-1-antisense 5′-CC-GGTACCTGGCCTCACATA-3′ (the introduced KpnI site is underlined). The DNA fragment encoding part of the avian influenza virus NP (nucleoprotein) was amplified from cDNA with following primers: np-sense 5′-GCGAATTCTCTGCACTCATCCTGAGAGG-3′ (the introduced EcoRI site is underlined) and np-antisense 5′-CGGGTACCTACTCCTCTGCATTGTCTCC-3′ (the introduced KpnI site is underlined). The DNA fragments obtained above were first cloned into pBluescript II KS (Stratagene) with the introduced restriction enzyme recognition sites and sequence-verified, then subcloned into pET-2P between EcoRI and KpnI sites to obtain pET2P-HBVP, pET2P-MERI-1 and pET2P-NP respectively.

Reporter plasmid pCMV-iHBS was constructed as described in [21] containing an HBsAg (type B hepatitis virus surface antigen)-coding sequence placed downstream of mouse Igκ-chain leader sequence, which enables the expressed protein to secrete to the outside the cell. The secretory plasmid was used for both cell culture assay and animal testing.

Expression of dsRNA in E. coli and esiRNA preparation

The dsRNA expression vectors obtained above were transformed into E. coli strain BL21(DE3) (Stratagene). Single colonies were inoculated into 2 ml of LB (Luria–Bertani) medium supplemented with 100 μg/ml ampicillin and cultured with shaking (250 rev./min) at 37 °C overnight. The seed culture was inoculated into 200 ml of fresh medium and continued growing until the D₅₅₀ reached 0.6, and IPTG (isopropyl β-D-thiogalactoside) was added to final concentration of 0.25 mmol/l to induce for 3 h before cells were harvested by centrifugation at 3800 g for 15 min. The dsRNA purification and esiRNA preparation were manipulated with the method described previously [18]. Briefly, RNA-containing cell lysate in 20% ethanol was passed through a Whatman® fibrous cellulose CF-11 column equilibrated with 20% ethanol containing 1× STE (10 mM Tris/HCl, 100 mM NaCl and 1 mM EDTA, pH 8.0). After washing with 5 column vol. of 1× STE containing 18% ethanol, dsRNA was then eluted out of the column with 20 ml of 1× STE, ethanol-precipitated and resuspended in 500 μl of RNase-free water. The concentration and purity of the dsRNA preparation was determined by measuring the absorbance at 260 nm and 280 nm.

To prepare siRNAs, 100 μg of purified long dsRNA was digested with 100 μg of recombinant RNase III (Ambion) in a 400 μl reaction mixture containing 50 mM Tris/HCl (pH 7.5), 50 mM NaCl, 10 mM MnCl₂ and 1 mM DTT (dithiothreitol) at 37 °C for 1 h. The digestion mixture was separated on a 15% non-denaturing polyacrylamide gel, and the esiRNAs of approx. 21–23 bp in length were recovered from gel with the method described by Yang et al. [22].

Cell culture and transfection

CHO (Chinese-hamster ovary) cells were grown at 37 °C in an atmosphere of 5% CO₂ in Dulbecco's modified Eagle's medium supplemented with 10% foetal calf serum (Biological Industries, Kibutz Beit Haemek, Israel), streptomycin (100 μg/ml) and penicillin (100 units/ml). To establish a cell line that constitutively expresses HBsAg, we transfected 600 ng/ml pCMV-iHBS plasmid DNA into CHO cells in a 24-well plate (70% confluence) using Lipofectamine™ 2000 (Invitrogen), according to the manufacturer's instructions. The level of HBsAg in the medium was measured 72 h after transfection, and G418 was added to cells to a concentration of 800 μg/ml. G418-resistant cells were then serially diluted to make constitutively expressive clonal HBsAg strains named CHO-iHBS cell strains.

For esiRNA dose–response experiments, CHO-iHBS cells from six-well plates (70% confluence, approx. 5×10⁶ cells) were transfected with 4–10 μg of esiHBVP using Gene Pulser Xcell™ system (Bio-Rad) according to the manufacturer's instructions. Cells were immediately seeded into new six-well plates with fresh medium. Every 24 h, medium was removed for analysis, and the cells were replenished with fresh medium. Secretory HBsAg in the medium was analysed using an ELISA.

Animal care and treatment

Male ICR mice (6–8 weeks old, 18–20 g; Shanghai Laboratory Animal Center, Shanghai, China) were used for hydrodynamic injection (high-volume intravenous injection). For dose–response experiments, appropriate amounts of esiRNAs and reporter plasmid pCMV-iHBS in 2 ml of Ringer's buffer were injected into mice within 5 s. Control mice were injected with 2 ml of Ringer's buffer. The procedure was performed in accordance with the requirements of the Shanghai Laboratory Animal Center of Shanghai, which proved the procedure to be safe for animals. Mouse blood was obtained on the indicated days after injection, and the titre of HBsAg in sera was measured using the ELISA.

RT (reverse transcription)–PCR analysis

Total RNA was isolated from freshly harvested livers of mice injected with pCMV-iHBS plasmid DNA and siRNAs using a Qiagen RNA isolation kit. RT was performed from total RNA using RNase-free MMLV (Moloney murine leukaemia virus) reverse transcriptase (Takara, Osaka, Japan). To correct the amplification process for tube-to-tube variability in amplification efficiency, β-actin mRNA was used as an internal standard for the semiquantification of the RT–PCR. The primers for β-actin were 5′-TGATGGACTCCGGTGACGG-3′ (forward) and 5′-TGTCACGCACGATTTCCCGC-3′ (reverse). After normalization with β-actin amplicon (179 bp), the same amount of cDNA was used as a template to amplify meri-1 and adar-1 genes using the following primers: meri-1-sense primer 5′-CGGAATTCGCAGACTTGAT-3′ and meri-1-antisense primer 5′-CCGGTACCTGGCCTCACATA-3′; adar-1-sense primer 5′-GCTCTAGAGTTCCAGTACTGTGTAGCAGT-3′ and adar-1-antisense primer 5′-ATGCGAATTCGGATCCTTGGGTTCGTGAGGAGGTCC-3′. The PCR program was set up as follows: denaturing at 94 °C for 1 min, annealing at 52 °C for 0.5 min and extension at 72 °C for 1 min. The number of amplification cycles was 30 for meri-1 and adar-1 genes and 25 for β-actin. Then 15, 20, 25, 30, 35, 37 and 40 cycles of each kind of RT–PCR were performed to verify that under the described conditions the PCR-amplification of each fragment was still in the linear range. Samples were analysed on a 2% agarose gel stained with ethidium bromide. The density of bands was quantified by using a Molecular Imager FX Pro Fluorescent Imager (Bio-Rad).

Higher doses of esiRNA induced stronger rebound of HBsAg expression after a period of suppression in CHO-iHBS cells

It is clear that inhibition of gene expression is sequence-specific and dose-dependent, and that the RNAi effect is transient and usually lasts 3–4 days. Previous work in our laboratory suggested that the expression of homologous gene rebounds after 3–4 days of suppression by esiRNAs and that the rebound effect was stronger in the cells or in animals challenged with higher doses of siRNAs than those challenged by lower doses of siRNAs (Z. Qian, B. Xuan and J. Hong, unpublished work).

To examine inhibitory effects of esiRNA of HBVP, we first linearized HBsAg-expressing plasmid pCMV-iHBS and integrated it into the genome of CHO cells to establish a cell line in which HBsAg is constitutively expressed (see the Materials and methods section). We then transfected these CHO-iHBS cells with 4 μg or 10 μg of esiHBVP dissolved in PBS. Approx. 5×10⁶ cells/well were used for transfection and the same volume of PBS without any DNA was used as a negative control. We measured the concentration of HBsAg secreted into the medium at various time points after transfection and normalized the expression of secretory HBsAg relative to the negative control. The results showed a continuous increase of inhibition of HBsAg expression in cells transfected with 4 μg of esiHBVP from 24 h to 72 h before a slight rebound at 96 h post-transfection, while cells given 10 μg of esiHBVP elicited a better inhibitory effect at an earlier stage and began to rebound at 72 h post-transfection (Figure 1). It seems that the inhibitory effect of RNAi began to be lost at later time points and the overall expression level of the gene in the cells began to rise. Interestingly, the cells given higher doses of siRNA showed a much higher rebound at 96 h after transfection. To explain this phenomenon, it might be possible that some sort of repelling mechanism is triggered in the cell when large amounts of siRNA were introduced into cells to protect cells from RNA viral infection.

Higher doses of esiRNA reduced the RNAi effect faster than lower doses in mice

We then tested the inhibitory effect of esiHBVP in mice by hydrodynamic injection (high-volume injection in tail vein), an effective method to transfer nucleic acid to up to 40% of mouse liver cells [23,24]. We treated separate groups of mice with 10 μg of reporter plasmid pCMV-iHBS, together with indicated amount of siRNAs. A group of mice was given additional esiNP as non-specific control. At days 1, 4 and 7 after injection, we measured the amount of HBsAg in serum. Similar to the findings obtained in cell experiments, all challenged mice revealed a reduced level of expression by siRNAs (Figure 2). However, on day 4 after injection the mice given 10 μg of siRNA showed 58% of the HBsAg level in the serum of uninjected mice, compared with approx. 22% in mice given 1 μg of siRNA. On day 7 after injection, the HBsAg level of the group given 10 μg of siRNA came back to 70% of the control, and those with the lower dose of siRNA also showed a slight increase of secretory level compared with 3 days before. Remarkably, mice injected with 1 μg of esiHBVP and 9 μg of unrelated esiNP displayed a much stronger rebound of secretion level, a quite similar pattern to that of mice given 10 μg of esiRNA. These results suggest that the RNAi machinery may be down-regulated at some point after large amounts of siRNAs are introduced into the cells.

Silencing of meri-1 gene could make RNAi more effective in mouse liver

Kennedy et al. [19] identified an exonuclease activity encoded by gene eri-1 in C. elegans, whose normal function suppresses RNAi in most tissues. The eri-1 gene encodes an evolutionarily conserved protein with domains homologous with nucleic-acid-binding and exonuclease proteins. Mouse genome database searches revealed a mouse orthologue of eri-1 (named meri-1) with seven exons. We amplified a fragment overlapping and including the whole of exon 2 and the N-terminal part of exon 3 from mouse cDNA and made its esiRNA as described in the Materials and methods section. We treated mice with reporter plasmid pCMV-iHBS and different amounts of esiHBVP together with or without esiMERI-1 using hydrodynamic injection. The amount of HBsAg in serum after injection was assayed to monitor RNAi efficacy. The mice injected with 1 μg or 10 μg of esiHBVP accompanied by 1 μg of esiMERI-1 displayed a much lower secretion level of HBsAg, compared with those not simultaneously given 1 μg of esiMERI-1 (Figure 3), suggesting that the decline (if not the loss) of the function of exonuclease MERI-1 may increase the efficacy of the RNAi machinery in mouse liver; this enhancement started the first day of injection and lasted for at least 7 days.

RT–PCR analysis of the expression level of ERI-1 in mice

In C. elegans, the eri-1 gene was identified to be an exonuclease that negatively regulates RNAi machinery. We then examined whether or not the expression level of meri-1 in the liver changed when siRNA was introduced into the body. Various amounts of esiHBVP or non-related control esiNP were injected into mice by using hydrodynamic injection method. At 4 days after administration, total RNA was extracted from livers and RT–PCR was performed using meri-1 and adar-1 gene-specific primers. All reactions were normalized with β-actin. As shown in Figure 4, the mRNA levels of meri-1 and adar-1 genes were increased markedly by the introduction of exogenous siRNAs. The group injected with 10 μg of esiHBVP showed a near 3-fold increase of mRNA level with the meri-1 gene and over 4-fold increase with the adar-1 gene than the uninjected group. The increase was also observed in the group injected with 1 μg esiHBVP plus 9 μg of non-specific esiNP. However when 1 μg of esiERI-1 was injected into mice together with 10 μg of esiHBVP, the mRNA levels of both meri-1 and adar-1 were reduced. In particular, the meri-1 mRNA showed a level close to that of mice injected with only 1 μg of esiHBVP. It seems that the administration of high doses of exogenous siRNAs, either 10 μg of esiHBVP or 1 μg of esiHBVP plus 9 μg of esiNP, induced the expression of meri-1 and adar-1 genes, and the addition of 1 μg of esiERI-1 offset, to some extent, the increase of meri-1 mRNA.