Inhibition of HCV NS3 protease by RNA aptamers in cells

Abstract

Non-structural protein 3 (NS3) of hepatitis C virus (HCV) has two distinct activities, protease and helicase, which are essential for HCV proliferation. In previous work, we obtained RNA aptamers (G9-I, II and III) which specifically bound the NS3 protease domain (ΔNS3), efficiently inhibiting protease activity in vitro. To utilize these aptamers in vivo, we constructed a G9 aptamer expression system in cultured cells, using the cytomegarovirus enhancer + chicken β-actin globin (CAG) promoter. By conjugating the cis-acting genomic human hepatitis delta virus (HDV) ribozyme and G9-II aptamer, a chimeric HDV ribozyme-G9-II aptamer (HA) was constructed, which was used to produce stable RNA in vivo and to create tandem repeats of the functional unit. To target the transcribed RNA aptamers to the cytoplasm, the minimal mutant of constitutive transport element (CTE), derived from type D retroviruses, was conjugated at the 3′ end of HA (HAC). Transcript RNAs from (HA)_n and (HAC)_n were processed into the G9-II aptamer unit by the cis-acting HDV ribozyme, both in vitro and in vivo. Efficient protease inhibition activity of HDV ribozyme-G9-II aptamer expression plasmid was demonstrated in HeLa cells. Protease inhibition activity level of tandem chimeric aptamers, (HA)_n and (HAC)_n, rose with the increase of n from 1 to 4.

INTRODUCTION

Hepatitis C virus (HCV) is the major etiological agent of post-transfusion non-A, non-B hepatitis. Chronic HCV infection is a serious disease throughout the world and can also develop into chronic hepatitis, liver cirrhosis or hepatocellular carcinoma. Although the number of HCV carriers has increased to ∼300 million worldwide, the principle drugs against HCV are all based on interferon. Currently, although combination treatment with interferon and the general antiviral nucleoside mimic, ribavirin, affords a more favorable response than interferon alone, patients not responding to either therapy have an increased risk of further liver disease, notably cirrhosis and liver cancer. To improve therapeutic options, the development of a greater range of drugs and more effective methods to combat HCV is both desirable and necessary (reviewed in 1).

HCV has a single positive-stranded RNA genome of ∼9.6 kb in length that encodes a large polyprotein consisting of ∼3010 amino acids. This precursor polyprotein is processed into a range of structural (core protein, C; envelope glycoproteins, E1 and E2) and non-structural (NS2, NS3, NS4A, NS4B, NS5A and NS5B) proteins by host signal peptidases and two viral proteases, NS2-3 and NS3 (reviewed in 2). The NS3 protein contains a trypsin-like serine protease domain at its N-terminus, while its C-terminal domain has helicase activity. It has been clearly demonstrated that the NS3 protease requires the cofactor NS4A to efficiently cleave the rest of the non-structural proteins (NS3, NS4A, NS4B, NS5A and NS5B). Since NS3 is necessary for subsequent viral replication, it is thought that development of a specific inhibitor of NS3 protease activity would be an attractive target for new anti-HCV drugs (reviewed in 3,4). The target NS3 or NS3-4A complexes are localized almost exclusively in the cytoplasm.

In previous work, we obtained RNA aptamers specific to the protease domain of NS3 by in vitro selection (5). The selected RNA aptamers (G9-I, -II and -III) bound specifically to the NS3 protease domain and strongly inhibited NS3 protease activity in vitro. Furthermore, alanine-scanning mutagenesis of the NS3 protease domain revealed an aptamer binding site (6), which corresponds to an exosite of the NS3 protease (7,8). To demonstrate the efficiency of these aptamers in vivo, their activity was tested by direct transfection of G-9 aptamers into HeLa cells in a transient enzyme (NS3-4A) and substrate expression system, where efficient inhibition was observed, as seen in vitro (unpublished results).

In this paper, we describe the design of chimeric cis-acting ribozyme and aptamer units to express G9 RNA aptamers in vivo with the correct folding and in large amounts. A cytoplasmic transport signal sequence was also attached to ensure proper cellular localization. This ‘ribozyme-aptamer signal’ unit can be repeated in tandem, increasing the aptamer dosage. In addition, a strong RNA polymerase II promoter was used to induce high expression of these units in cells.

MATERIALS AND METHODS

Construction of plasmids

EX taq (Takara) was used for PCR. The ΔNS3 (9) and NS3-4A expression plasmids (10), G9-II aptamer coding plasmids (5) and genomic HDV ribozyme, CdS4, coding plasmids (11) have been described previously. In the case of PCR using the pUC vector as the template, M13 universal forward and reverse primers were used.

The HDV-G9-II aptamer was constructed as follows. The G9-II aptamer was inserted into the stem IV region of the genomic HDV ribozyme. The double-stranded DNA containing the G9-II aptamer sequence was amplified by first PCR mutagenesis with the corresponding (+) and (–) primers containing regions of 20 bases of genomic HDV ribozyme attached at the 5′ end using the G9-II plasmid as the template. Using this double-stranded DNA as one of the primers for a second PCR, the cDNA of the HDV-G9-II aptamer was amplified with an additional (+) primer containing the T7 promoter, using plasmid cis-MTS-U (11) as the template. The genomic HDV ribozyme and G9-II-aptamer conjugated PCR products were digested with XhoI and BamHI, and inserted into pUCT7 (12) to produce pHDV-G9-II (pHA).

The constitutive transport element (CTE) M45 coding plasmid was constructed by cloning of the relevant PCR fragment, to which BamHI and SalI linkers had been added at each end using chemically synthesized oligonucleotides, into pGEM Teasy (Promega). The CTE M45-ribozyme-aptamer plasmid (pHAC) was constructed by insertion of the PCR product of CTE M45 double digested with BamHI and SalI into pHA digested with the same enzymes. The SalI site of pHAC was mutated to KpnI using the Quik Change™ Site-Directed Mutagenesis Kit (Stratagene) to generate pHAC(KpnI) for the construction of expression plasmids.

The plasmids containing tandem HDV-G9-II aptamers, p(HA)_n were constructed as follows. To generate p(HA)₂, the PCR fragment of HA (pHA used as the template) double digested with XhoI and HindIII was subcloned into pHA double digested with SalI and HindIII, thereby fusing the SalI and XhoI sites. To increase the value of n, further fragments were inserted. In the case of p(HAC)_n, concatenation of HAC was achieved using the same method used for the construction of p(HA)_n. For the construction of the plasmid carrying the inactive HDV-G9-II aptamer (pHmA), the 763C to U substitution was introduced by mutagenic PCR. Finally, tandem inactive HDV-G9-II aptamers, p(HmA)_n and p(HmAC)_n, were constructed by the same method as p(HA)_n and p(HAC)_n.

To introduce the 3′ processing HDV ribozyme into the XhoI–BglII site of the pCAGGS expression vector (13), a KpnI site (next to XhoI) in the upstream region of HDV86 (12) and a BglII site in the downstream region were introduced into HDV86 by mutagenic PCR using substituted primers. The XhoI and BglII digested PCR fragment of HDV86 was inserted into pCAGGS digested with the same enzymes to generate the cytomegalovirus enhancer + chicken β-actin globin (CAG) promoter-driven vector h/pCAGGS. To construct the G9-II-expressing plasmid Ah/pCAGGS, the XhoI and KpnI digested PCR fragment of G9-II (where the restriction enzyme sites had been introduced by substituted PCR primers) was inserted into h/pCAGGS. Finally, (HA)_nh/pCAGGS, (HAC)_n h/pCAGGS and inactive HDV derivatives (HmA)_nh/pCAGGS and (HmAC)_nh/pCAGGS were constructed using the same method.

The plasmids constructed were sequenced using a BigDye Terminator Cycle Sequencing Kit on a 377 automatic sequencer (Applied Biosystems). For the protease inhibition assay in HeLa cells, the construction of the substrate expression plasmid (pC5abY), and the enzyme expression plasmids (pCMV/34s-2-FLAG and pCMV/34s-2M-FLAG) have been described elsewhere (N.Kakiuchi, unpublished results).

Preparation of RNA aptamers in vitro

The double-stranded DNA generated by PCR (EX Taq, Takara) was used as a template for in vitro transcription by T7 RNA polymerase (T7 Ampliscribe Kit, Epicentre Technologies). Reaction conditions were as described previously (5,14). To prepare the ³²P-labeled RNAs for analysis of the self-cleavage reaction, plasmid DNAs containing the HDV-G9-II aptamer and CdS4, linearized with BamHI, were used for in vitro transcription (14).

To detect the self-cleavage activity of tandem HDV aptamers, (HA)_n and (HAC)_n (n = 2–4), the relevant plasmids, p(HA)_n(KpnI) and p(HAC)_n, were linearized with KpnI, and used as a template for transcription as described above. At 3, 6 and 22 h, aliquots of the reaction mixture were fractionated by electrophoresis on 8% PAGE containing 7 M urea and detected using a bioimaging analyzer (BAS2000; Fuji Film).

NS3 protease cleavage activity in vitro using a synthetic substrate

The protease inhibition assay was essentially as described previously (5,15). A dansyl-labeled peptide substrate 17mer (86 µM) was added to the premixture of the protease (0.8 µM) and variants of G9 aptamers, in the presence of the NS4A peptide (22.5 µM). The reaction mixture (25 µl) was incubated for 40 min at 25°C, and then products were analyzed on reverse phase HPLC.

Transfection of RNA aptamers or plasmids into cells

Cells (HeLa and HepG2) were maintained in Dulbecco’s modified Eagle’s medium (DMEM; Life Technology) containing 10% heat-inactivated fetal bovine serum (Life Technology) and 100 mg/l kanamycin, at 37°C in a humidified atmosphere of 5% CO₂. Cells were plated at an appropriate density so as to be 50–80% confluent. After 1 day, nearly confluent monolayers of HeLa cells and HepG2 cells were transfected using FuGENE 6 (Roche, IN, USA) according to the manufacturer’s protocol.

The enzyme expression vector that is modified of pCMV/N1027–1215 (10) (pCMV/34s-2-FLAG; HCV 1027–1908 coding NS3-NS4A, 0.5 µg/well) and substrate expression vector (pC5abY; CFP-GDDIVCC*SMST-YFP, *; cleavage site between NS5A and NS5B, 0.12 µg/well) were transfected into HeLa cells (seeding 0.4 × 10⁵ in 1 ml per well). After 24 h of incubation, RNA aptamer (∼2.5 µg) was transfected using DMRIE (Life Technology), incubated for 1 day, harvested and subjected to western blot analysis.

For endogenous delivery of RNA aptamers, HDV-G9-II aptamer expression plasmids (0.5 µg/well) were cotransfected into HeLa cells with the substrate expression plasmid (C5abY; 0.12 µg/well) and enzyme or mutant enzyme expression plasmid: pCMV/34s-2-Flag or pCMV/34s-2M-Flag (NS3-4A; 0.5 µg/well) using FuGENE 6. The empty vector (pCAGGS) was used as control. After 3 days, cells were lysed and subjected to western blot analysis.

Detection of protease inhibition activity in cells

The concentration of proteins in cell lysates was quantified using the Bio-Rad Protein Assay (Bio-Rad). Cell lysates containing 5–6 µg of proteins were separated on a 15% SDS–PAGE gel, then transferred to PVDF membranes (Immobilon P; Millipore). The membrane was blocked with 10% skim milk in Tris-buffered saline (TBS) with 0.1% Tween-20 for 1 h and stained with antiserum against GFP peptides (1:1000 dilution; Clontech) in TBS with 0.1% Tween-20. The blot was probed with biotin labeled anti-rabbit IgG (Wako) as the secondary antibody. The signal was amplified by incubation with ABC reagent (Vector Laboratories), probed with ECL™ reagent (Amersham Pharmacia Biotech) and detected by X-ray film exposure (Fuji Film RX-U). Signals were quantified with an ARCUS II image scanner (AGFA) using the Quantity One quantification software package (Protein Databases Inc.). Protease inhibition activities caused by delivery of RNAs into cells were determined by the quantification of the fraction of substrate cleaved by NS3.

Detection of processing and intracellular localization of HDV-G9-II aptamers in cells by northern blot analysis

HDV-G9-II aptamer expression plasmids were transfected into HepG2 cells at 24 h after passage (3 × 10⁶). For fractionation of nuclear and cytoplasmic RNAs, cells were collected at 2–4 days post-transfection and separated by the modified protocol of NE-PER™ Nuclear and Cytoplasmic Extraction Reagents (Pierce). After isolating the cytoplasmic fractions, 100 µl of NP-40 buffer [20 mM Tris–HCl (pH 7.5), 50 mM KCl, 10 mM NaCl, 1 mM EDTA, 0.5% NP-40] was added to the pellets and incubated on ice for 5 min. Next, the solution was centrifuged and the pellets were resuspended in 100 µl of NP-40 buffer to use as nuclear fractions. The cytoplasmic RNAs were extracted with ISOGEN (Nippon Gene) according to the manufacturer’s protocol. The RNA content was determined photometrically.

For northern blot analysis, samples from each fraction containing 15–30 µg RNA were denatured (65°C, 15 min) with formaldehyde before electrophoresis. RNA samples were fractionated by electrophoresis through 3% NuSieve (3:1) agarose gel (FMC Inc., Rockland, ME, USA) containing 2.1 M formaldehyde and transferred onto an Immobilon Nylon-N+ membrane (Millipore). Subsequently, RNAs were heat fixed for 2 h at 80°C and UV crosslinked. The blot was prehybridized for 2 h at 42°C in 50% formamide, 6× SSC, 5× Denhard’s, 0.5% SDS, containing 100 µg/ml denatured salmon sperm DNA, and hybridized overnight at 42°C with the specific probes. Probe DNA complementary to the G9-II aptamer was labeled with [α-³²P]dCTP by PCR using EX taq (Takara). Filters were washed and analyzed using a bioimaging analyzer (BAS2000; Fuji Film). For validation of the RNA fractionation, the filters were re-hybridized with a probe specific to the nuclear U6 snRNA (16).

RESULTS

The HDV-G9-II aptamer is effective in vitro and in HeLa cells

When G9 RNA aptamers (∼2.5 µg/ml) were directly introduced into HeLa cells transiently expressing the NS3 protein, NS4A peptide and a substrate that has the NS5AB junction sequence, the aptamers showed a dose-dependent protease inhibition activity (∼50% inhibition; unpublished results). The G9-II aptamer proved to be the most efficient inhibitor. As the next step, to express RNA aptamers in cells, we constructed a G9 aptamer expression system using the CAG promoter (13). To neglect the probability of incorrect aptamer folding and the instability of the aptamer expressed by itself, a ribozyme– aptamer chimera that produced small RNA units from the relevant RNA transcript in cells using the cis-acting ribozyme was designed. For this purpose, the HDV ribozyme was used (Fig. 1A) because it cleaves efficiently at a precise position under physiological conditions. In the tertiary structure, the ribozyme’s stem IV region is protruding (17) and it is possible to change its size without any loss of ribozyme activity (11). This means the aptamer sequence can be inserted into the stem IV region and the chimera should retain both aptamer and self-cleavage activities. Furthermore, the HDV ribozyme is stable in vivo (18), so the chimeric construct is likely to remain stable and functional in cells. The G9-II aptamer was inserted into stem IV of the genomic HDV ribozyme, CdS4 (11), to generate a chimeric HDV-G9-II aptamer (Fig. 1A). During run-off in vitro transcription, self-cleavage of the HDV-G9-II aptamer was confirmed to occur at a precise position (data not shown) and cleavage efficiency was determined using isolated precursor HDV-G9-II aptamers. Compared with CdS4 (12 ± 2/min), cleavage efficiency of the chimera was reduced by ∼1 order of magnitude (HDV-G9-II: 1.3 ± 0.3/min), but it was self-cleaved completely and released 5′ cleaved fragment (14 nt).

Figure 1.

Construction of several ribozyme–aptamer conjugates and their activities in vitro and in vivo. (A) To construct the HDV-G9-II (HA) aptamer– ribozyme conjugate, stem IV (dotted box) of CdS4 was replaced with the G9-II aptamer. The arrow indicates the self-cleavage site of the cis-acting genomic HDV ribozyme, CdS4 (11). Numbering of the HDV ribozyme is based on that of Makino et al. (27). In the next step CTE M45 was conjugated. The inactive HDV aptamer variants, HmA and HmAC, were constructed by substitution of C763 to U. (B) Inhibition of proteolytic activity of ΔNS3 in vitro by HDV-G9-II and derivatives of the genomic HDV ribozyme. Inhibition activity was measured with ΔNS3 and a 17mer substrate by HPLC (15). (C) Inhibition of proteolytic activity in HeLa cells with HDV-G9-II. RNAs were transfected into cells along with transiently expressed NS3 protease and its substrate. Protease inhibition activity was detected by western blot analysis. The error bars indicate the standard deviation of three different experiments. To normalize the data, proteolytic activity (cleavage of substrate) in the absence of RNA was set at 100%. (D) Inhibition of proteolytic activity in vitro by monomer units (HA, HAC and HmAC) and a tandem derivative, (HmAC)₄. The error bars indicate the standard deviation of three different experiments. To normalize the data, proteolytic activity (cleavage of substrate) in the absence of RNA was set at 100%.

The HDV-G9-II aptamer showed similar protease inhibition activity in vitro (∼90% inhibition) against ΔNS3 as observed for G9 aptamers. Interestingly, some inhibition activity (∼30%) was found for CdS4 alone. In fact, the conserved sequence (5′-GAAUGGGAC-3′) of G9 aptamers exists in J4/2 and stem II of CdS4 (764–771). To investigate whether this weak activity was caused by this common sequence, a protease inhibition assay was carried out using mutated HDV ribozyme, A765G and A765C variants (19). Both of these variants lost their inhibition activity almost completely; therefore, this region of the HDV ribozyme interacts moderately with the NS3 protease (Fig. 1B). Next, protease inhibition activity in HeLa cells was tested by directly introducing HDV-G9-II aptamers (Fig. 1C). The HDV-G9-II aptamer showed more effective protease inhibition (50%) than G9-II alone in HeLa cells.

Attachment of a cytoplasmic transportation signal sequence to HDV-G9-II aptamers and tandem of repeating this unit

Since the NS3 protein is complexed with NS4A and localizes to the endoplasmic reticulum in cells (20), transportation of HDV-G9-II aptamers from the nucleus to the cytoplasm is important. It has been reported that the CTE of type D retroviruses promotes the nuclear export of unspliced viral transcripts in interaction with TAP (21). CTE M45 is the minimal mutant of CTE (22). To overcome the problem of aptamer localization, CTE M45 was attached to the 3′ end of HDV-G9-II (HA) creating HAC (Fig. 1A). To increase the copy number of aptamers in cells, further copies of the HDV-G9-II aptamer were ligated in tandem, giving (HA)_n and (HAC)_n (n = 1–4), which produce the small unit (HA or HAC) containing G9-II, through processing by the cis-acting HDV ribozyme. As a control, inactive HDV mutants (HmA and HmAC), which contained the 763 C to U substitution (Fig. 1A), resulting in no self-cleavage, were constructed and showed protease inhibition activity at a similar level to HA and HAC (Fig. 1D).

The activity of the monomer unit (HA, HAC and HmAC) and the inactive HDV tandem aptamer (HmAC)₄ were compared in in vitro proteolytic assay. The amount of (HmAC)₄ used was one-quarter that of HA, HAC and HmAC, thus an equal number of monomer unit aptamers was present in each RNA sample. The inactive HDV tandem variant (HmAC)₄ showed decreased efficiency (∼20% inhibition in Fig. 1D). This strongly suggests that the use of active HDV aptamers in a tandem system, which is able to produce small G9-II units, is effective.

For all tandem HDV aptamers (HA)_n and (HAC)_n (n = 1–4), self-cleavage activities were confirmed by in vitro run-off transcription of relevant digested plasmids and the expected cleaved ladder was detected on the gel image (Fig. 2A and B). The attached CTE portion forming the stem–loop at the 3′ end of HA acts favorably to support proper folding of the long tandem HDV derivatives and to avoid the misfolding of the HDV ribozyme structure. In the inactive tandem variants, (HmA)_n and (HmAC)_n, no self-cleavage was detected after 1 day (data not shown).

Figure 2.

Self-cleavage reaction of tandem repeated HDV-G9-II aptamers in vitro. (A) Self-cleavage pattern of run-off in vitro transcription of (HA)_n (n = 1–4) on denaturing PAGE. (B) Self-cleavage pattern of run-off in vitro transcription of (HAC)_n (n = 1–4). The cleavage pattern is schematically drawn at the bottom. White triangles indicate the self-cleavage sites. Asterisks indicate the small unit produced from the tandem derivatives due to self-cleavage.

Construction of a HDV-G9-II aptamer expression plasmid under the control of the CAG promoter

The efficiency of the G9 aptamer and chimeric HDV-G9 aptamers could be detected by direct transfection into HeLa cells. As a next step, HDV-G9-II aptamers were put into an expression vector system in mammalian cells (Fig. 3A). To construct expression vectors for (HA)_n and (HAC)_n, we therefore chose to use the strong, RNA polymerase II-dependent CAG promoter (13). The CAG promoter consists of the chicken β-actin promoter with the CMV immediate-early enhancer upstream, and a splicing acceptor site from the β-globin gene downstream. This composite promoter produces high expression levels of foreign genes, which are expressed as novel eukaryotic pre-mRNAs possessing a 3′ poly(A) tail and containing an intron (∼910 nt). After splicing, the expressed RNA contains 5′ cap flanking regions (∼100 nt) up to the first cleavage site of HA or HAC. To release the 3′ dispensable fragment [poly(A) tail] from the transcript RNA aptamers, the HDV ribozyme was used as the 3′ processing ribozyme (h; Fig. 3A). It is located upstream of the poly(A) recognition site and 3′ end of the tandem HA [designated (HA)_nh/pCAGGS] or HAC [designated (HAC)_nh/pCAGGS] units. An expression vector carrying the inactive HDV ribozyme–aptamer conjugate was also constructed in the same way [designated (HmA)_nh/pCAGGS and (HmAC)_nh/pCAGGS].

Figure 3.

Detection of the self-cleavage reaction and intracellular localization of tandem HDV-G9-II aptamers in HepG2 cells by northern blot analysis. (A) Schematic diagram of an aptamer expression plasmid and processing of the transcript RNA. The HDV-G9-II aptamer (HA)_n (indicated by a black box) is designed to be transcribed under the control of the CAG promoter (chicken β-actin promoter + cytomegalovirus enhancer, indicated by CAG), whose parental plasmid vector is pCAGGS (13). ‘h’ in a gray box indicates the 3′ processing HDV ribozyme. RβGpA indicates rabbit β-globin poly(A) cDNA. Black triangles indicate the self-cleavage sites. (B) Stability of HDV-G9-II aptamers (HA)₄h and (HAC)₄h expressed in HepG2 cells. N, nuclear fraction; C, cytoplasmic fraction. To confirm fractionation, U6 RNA is shown in each fraction. Approximately equal levels of monomer unit (HA and HAC) were observed in cytoplasmic fractions 2–4 days after transfection. (C) Comparison of the amount of monomer unit (HA and HAC) produced by the self-cleavage of (HA)_nh and (HAC)_nh (n = 1, 4). (D) Comparison of RNAs expressed from (HAC)₄h and (HmAC)₄h. Cells were harvested 2 days after transfection (C and D).

Detection of processing and localization of expressed tandem HDV aptamers in vivo

To test the self-cleavage activity and localization of expressed tandem HDV aptamers, (HA)₄h/pCAGGS and (HAC)₄h/pCAGGS were transfected into HepG2 cells. Between 2 and 4 days after transfection, cell lysates were fractionated into nuclear and cytosolic fractions. RNA was isolated from each fraction and the HDV-G9-II aptamer was detected by northern blotting using the G9-II probe. As observed clearly in the northern blot analysis (Fig. 3B), tandem HDV aptamers, (HA)₄h and (HAC)₄h, were expressed efficiently and processed to their monomer units in cells as well as in vitro (Fig. 2). Equal molecular amounts of RNA were loaded into each lane of the gel, however, the total amount of RNA in the cytosol is two to three times more than that in the nucleus. Considering the amount of total fractionated RNA, more than half of the expressed HDV-G9-II aptamer is localized in the cytoplasmic fraction. Furthermore, a comparison of samples taken 3 and 4 days after transfection showed no difference in gel image, demonstrating the continued stability of the HDV aptamer 4 days after transfection.

To evaluate the efficiency of repeating the aptamer unit, (HA)_nh/pCAGGS and (HAC)_nh/pCAGGS (n = 1 and 4) were transfected into HepG2 cells (Fig. 3C). The expressed aptamers, HAh and HACh, from the monomer (n = 1) unit plasmids self-cleaved and produced the small unit, but the band intensity of the cleaved unit was weak. In contrast, the tandem repeat derivatives (n = 4) showed strong intensity of the monomer bands, indicating an increase in copy number of the aptamer unit due to self-cleavage. To compare self-cleavage of (HAC)₄ with (HmAC)₄, (HAC)₄h/pCAGGS and its inactive HDV variant (HmAC)₄h/pCAGGS were transfected into HepG2 cells and cleavage ability was compared after 2 days (Fig. 3D). Tandem (HAC)₄ showed self-cleavage producing the HAC unit, but the inactive HDV mutant did not show self-cleavage. These results are consistent with the events observed in vitro.

Expressed tandem HDV aptamers inhibit protease activity in vivo

Next, the effectiveness of HDV aptamers in an aptamer expression vector system was tested. The HDV aptamer expression vector was cotransfected into HeLa cells with a substrate (C5abY) expression plasmid and an enzyme (NS3-4A) expression plasmid. Three days after transfection, cell lysates were treated and protease inhibition activity was detected by western blot analysis and quantified. As shown in Figure 4A, by increasing the n value of (HAC)_nh/pCAGGS from 1 to 4, the amount of the cleaved substrate (C5abY) products was reduced. The same result was also detected in cells transfected with (HA)_nh/pCAGGS (data not shown). This evidence indicates that the tandem-repeating system acts efficiently for increasing the cellular concentration of the aptamers in HeLa cells. To confirm that the substrate cleavage was caused by NS3-4A, a mutant NS3-4A(Ser1165→Ala) expression plasmid, which expresses mutant NS3 without protease activity, was also transfected and no substrate cleavage due to the expressed tandem HDV aptamer was seen (data not shown).

Figure 4.

Detection of the protease inhibition activity of HDV NS3 aptamers in HeLa cells by western blot analysis. (A) Comparison of protease inhibition activity of (HAC)_nh/pCAGGS (n = 1–4) in HeLa cells. ‘cont.’ indicates treatment with pCAGGS instead of (HAC)_nh/pCAGGS. (B) Com parison of the protease inhibition activity of monomer and tandem variants in HeLa cells. (C) Comparison of the protease inhibition activity of tandem variants with/without the 3′ processing HDV ribozyme: (HA)₄h/pCAGGS, (HA)₄/pCAGGS, (HAC)₄h/pCAGGS, (HAC)₄/pCAGGS and (HmAC)₄h/pCAGGS. Substrate cleavage (%) was calculated by western blot analysis. The values show the average from three different experiments.

It is clearly shown in Figure 4B that A/pCAGGS without 5′ and 3′ processing HDV ribozyme activity showed the aptamer to have almost no effect, giving the same level as pCAGGS (the empty vector). There was no big difference in activity of the monomer units, but the HDV-G9-II aptamer is very effective in the tandem system. Tandem repeats of the inactive HDV ribozyme, (HmAC)₄, showed the same lower inhibition activity observed in vitro (Fig. 1D). For the pCAGGS control, (HA)₄h showed better inhibition efficiency (∼41% inhibition in Fig. 4C) than its 3′ processing HDV ribozyme deleted derivative (HA)₄ (∼19% inhibition). In all cases, aptamers with the 3′ processing HDV ribozyme (h) showed better inhibition efficiency than aptamers without 3′ processing HDV ribozyme. In pol II transcription systems, the 5′ cap and 3′ poly(A) are very important for mRNA transport to the cytosol. This result indicates that 3′ processing by the HDV ribozyme occurs slowly and does not interfere with the function of poly(A) for nuclear export of expressed HDV-G9-II aptamers. Most of the 5′ capped or 5′ cleaved processed HDV aptamer with poly(A) tails were transported immediately to the cytosol, then self-cleavage occurred at each position and the 3′ fragment was released by the 3′ processing HDV ribozyme.

Further tandem HDV aptamer expression plasmids were constructed, to generating (HA)_nh/pCAGGS (n = 4, 6, 8, 10), and (HAC)_nh/pCAGGS (n = 4, 8, 10, 12). In each experiment, the same amount of plasmid was cotransfected and the efficiency of substrate cleavage was quantified. Most of the HA and HAC tandem HDV aptamer derivatives showed similar levels of protease inhibition activity. The efficiency of this multiple independent HDV aptamer expression system reached a plateau value at n of ∼4 (data not shown).

DISCUSSION

The characteristic function of G9 aptamers targeted to HCV NS3 protease in cultured cells requires proper folding into the active conformation for binding to NS3, and stability from other cellular factors. For these purposes, it is important to eliminate the 5′ and 3′ extra flanking regions from the aptamer in the expressed RNA. In the example of 5′ and 3′ trimming using hammerhead ribozymes, tandem usage of its cassette system has been reported (23). We used cis-acting HDV ribozyme. HDV has been discovered in humans and its catalytic domain (HDV ribozyme; reviewed in 24) functions for efficient cleavage activity under physiological conditions. Aptamers can be inserted into the non-functional stem IV region of the HDV ribozyme. The HDV ribozyme (3′ cleaved fragment) forms a compact, packed tertiary structure via extensive hydrogen bonding (17) and is stable in vivo (18). The HDV ribozyme–aptamer conjugate appears to be protected from exonuclease activity at both 5′ and 3′ ends in the tightly packed structure of the HDV ribozyme, resulting in continued stability up to 4 days after transfection. The most useful benefit of the constructed aptamer expression vector is the tandem system, using the advantages of the HDV ribozyme, which operates effectively to produce small stable active units in the cell.

Other examples of short, connected aptamers tested in vivo have been reported. Hoechst dye 33258 aptamers were shown to act as a translational switch for controlling gene expression in CHO cells (25), and B52 specific aptamers are able to inhibit Drosophila B52-stimulated pre-mRNA splicing in vivo (26). In the case of HDV-G9 aptamers (HA and HAC), the tandem aptamers with inactive ribozyme, (HmA)_n and (HmAC)_n, do not show any enhanced activity. In contrast, the other tandem vectors, (HA)_nh and (HAC)_nh, operated effectively in vitro and in vivo. This system could be applied to introduce other catalytic RNAs into the cell. Moreover, it might be possible to express multiple functional RNAs simultaneously through transfection of a single plasmid by inserting different catalytic RNAs into stem IV of each tandem HDV ribozyme.

To stimulate nuclear export of the expressed aptamer RNAs to the cytosol, CTE-M45 was conjugated to the 3′ end of HA to construct HAC. However, in the expression vector system, both (HA)_nh and (HAC)_nh showed similar levels of efficiency. There are two possible reasons for this result. First, since the samples were taken 3 days after transfection, the reaction is in a steady state, so differences between them may no longer be visible. Secondly, CTE-M45 is the minimal mutant of CTE function, so its effect is rather weak in this system. However, its addition to HA does not interfere with the ribozyme or aptamer functions in cultured cells. It can be concluded that by using this kind of construction system it might be possible to include two functions in the cis-acting HDV ribozyme, with one functional RNA in stem IV and the other linked at the 3′ end, where the 3′ attachment of a catalytic functional RNA does not interfere with the formation of the active structure of the HDV ribozyme and formation of the rigid structure.

The NS3 aptamer expression system using the HDV ribozyme was used in a pol II system. As shown in this study, general eukaryotic mRNA nuclear export factors (5′ cap and polyadenylation) were not affected by the insertion of multiple HDV ribozymes and functioned smoothly to export the expressed aptamer RNA to the cytosol. There, most of the chimeric HDV ribozyme in expressed RNAs acted to release the dispensable 5′ and 3′ fragments thus producing small active units. There are many reports of aptamers that can bind to nucleic acids or non-nucleic acid targets such as small organic molecules and proteins. They might be broadly useful for biological study, or as biosensors, diagnosis tools or therapeutic reagents. However, there are limited examples of their applications in vivo. This tandem system using the HDV ribozyme will be useful for the expression of functional RNAs in cultured cells. The effect of the constructed tandem HDV aptamer expression vector in the virus genome expression system is under investigation. Continuation of these research projects has the potential to lead to the application of protease inhibitors as antiviral compounds against HCV.