Efficient Transfer of Synthetic Ribozymes into Cells Using Hemagglutinating Virus of Japan (HVJ)-Cationic Liposomes

We investigated the usefulness of ribozymes in inhibiting the expression of human T-cell leukemia virus type I (HTLV-I) gene. Two hammerhead ribozymes that were against HTLV-Irex (RR) and tax (TR) mRNA were synthesized. Both ribozymes were sequence-specific in the in vitro cleavage analysis of run-off transcripts fromtax/rex cDNA. Intracellular activities of the ribozymes were studied in HTLV-I tax cDNA-transfected rat embryonic fibroblasts (Rat/Tax cells), which expressed the Tax but not Rex. Ribozymes were delivered into cells using anionic or cationic liposomes fused with hemagglutinating virus of Japan (HVJ). Cellular uptake of ribozymes complexed with HVJ-cationic liposomes was 15–20 times higher cellular uptake than naked ribozymes, and 4–5 times higher than that of ribozymes complexed with HVJ-anionic liposomes. HVJ-cationic liposomes promoted accumulation of ribozymes in cytoplasm and accelerated transport to the nucleus. Tax protein levels were decreased about 95% and were five times lower when the same amount of TR was introduced into the cells using HVJ-cationic, rather than HVJ-anionic liposomes. Inactive ribozyme and tax antisense oligodeoxynucleotides reduced Tax expression by about 20%, whereas RR and tax sense oligodeoxynucleotides had no effect. These results suggest that the ribozymes’ effect against taxmRNA was sequence-specific, and HVJ-cationic liposomes can be useful for intracellular introduction of ribozymes.

The discovery of RNA molecules with sequence-specific RNAcleaving properties, called ribozymes, led to investigations of their potential use as specific inhibitors of gene expression (1)(2)(3). Several ribozyme configurations have been identified, of which the "hammerhead" (4) and "hairpin" (5) structures are the simplest, and most suitable, for biomedical applications (6 -8). Several factors appear to contribute to the intracellular efficacy of ribozymes and thus the success of ribozyme gene therapy. Most importantly, the ribozyme must co-localize with its molecular target in the appropriate cellular compartment and must be present in a sufficiently high concentration to promote hybridization.
Previous studies have generally assessed the catalytic activity of ribozymes in cell-free assay systems (6 -8). The present study describes an experimental system that allows one to assess the effects of ribozymes in living cells. The in vivo application of ribozymes will depend on the availability of efficient delivery methods. The methods of gene transfer are classified as viral or non-viral. Ribozymes generally have been introduced into cells via a viral infection or the transfection of expression vectors (6,9,10). Viral vectors can potentially lead to the incorporation of ribozyme-encoding genes into the cellular chromosomes, thereby allowing their permanent expression, and many studies now are aimed at developing suitable vectors that present a minimal associated risk to the host. Alternatively, ribozymes can be delivered to the cells by non-viral methods, such as lipophilic vesicles (liposomes) and cholesterol (6).
We recently developed a highly efficient method for gene transfer that involves the entrapment of DNA or RNA using hemagglutinating virus of Japan (HVJ, 1 Sendai virus) to enhance the fusion of anionic liposomes to cell membranes (11,12). However, especially in cultured cells, the level of transgene expression achieved with this method is somewhat lower than that obtained with some of the viral vectors. We then improved this gene delivery system using cationic lipids for the liposomes (13). In the present study, we compared the effects of ribozymes that had been transferred into living cells, using anionic or cationic liposomes.
Ribozymes may become useful molecular therapies for several diseases of humans, including human T-cell leukemia virus type I (HTLV-I) infection, which is etiologically associated with adult T-cell leukemia (14,15). The HTLV-I Tax protein has oncogenic properties that may play a key role in tumorigenesis (16,17). We previously evaluated gene therapy for the treatment of HTLV-I-related diseases using tax antisense oligodeoxynucleotides (ODNs) (18 -20). In the present study, we describe an approach involving ribozyme-mediated cleavage of HTLV-I tax/rex mRNA. We investigated whether ribozymes * This work was funded by Grant 09470172 from the Ministry of Education, Science, Sports and Culture of Japan and grants from Funds for Comprehensive Research on Long Term Chronic Disease. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. § To whom correspondence and reprint requests should be addressed. Tel.: 81-992-75-5437; Fax: 81-992-75-2629. 1 The abbreviations used are: HVJ, hemagglutinating virus of Japan; FITC, fluorescein isothiocyanate; HTLV-I, human T-cell leukemia virus type I; mAb, monoclonal antibody; ODN, oligodeoxynucleotide; PBS, phosphate-buffered saline; PI, propidium iodide; Rat/Tax cell, tax-expressing rat embryonic fibroblast; RR, ribozyme against HTLV-I rex mRNA; RC, inactive control ribozyme derived from RR; TR, ribozyme against HTLV-I tax mRNA; TC, inactive control ribozyme derived from TR; TRITC, tetramethylrhodamine isothiocyanate; bp, base pair(s).
can cleave their target RNAs in the cells and whether HVJcationic liposome-mediated gene transfer allows efficient introduction of ribozymes into living cells. Targeted against HTLV-I tax/rex  mRNA-Ribozymes were chemically synthesized on a 1-mol scale  using a DNA synthesizer (model 8909 Expedite System, PerSeptive Biosystems, Framingham, MA). We generated two hammerhead ribozymes that targeted the tax/rex mRNA based on previously published sequence information (21). The ODNs were designed to bracket the tax and rex AUGs, which has been shown to inhibit HTLV-I Tax protein expression in experiments using antisense ODNs (18 -20).

Synthesis of Ribozymes
The ribozyme targeting the rex mRNA (RR) had the sequence 5Ј-G S G S G S CCUCCCUG*AUGAGGCCGAAAGGCCGAAACGGGU S C S -U S -3Ј. It cleaves the mRNA at position 5139 of the HTLV-I genome. The ribozyme targeting tax mRNA (TR), which cleaves the mRNA at position 7308 of the HTLV-I genome, had the sequence 5Ј-A S C S C S CUGGG-CUG*AUGAGGCCGAAAGGCCGAAAGUGGG S C S C S -3Ј (Fig. 1A). The HTLV-I-related sequences are underlined. As described previously by Ruffner et al. (22), we also designed inactive control ribozymes for RR (labeled RC) and TR (labeled TC), which inactivate the hammerhead motif by replacing G 5 with A (indicated by *). Each ribozyme contains phosphorothioate-modified nucleotides (indicated by "s"). The 5Ј end of each ribozyme was labeled with fluorescein isothiocyanate (FITC) using the FluoroPrime reagent (Pharmacia Biotech, Uppsala, Sweden).
Generation of an HTLV-I tax-expressing Plasmid and Run-off Transcription-Total cellular RNA was isolated from the HTLV-I-infected T-cell line, MT-2 cells (gift of Prof. I. Miyoshi, Kochi Medical School) by the acid guanidium thiocyanate-phenol-chloroform extraction method (23). HTLV-I tax cDNA from the MT-2 cells was obtained by polymerase chain reaction using HTLV-I tax/rex-specific primers as described previously (24). The plasmid pGEM-T was purchased from Promega Corp. (Madison, WI). The run-off transcription vector, pGEM-T-tax, was obtained by cloning the 220-bp HTLV-I tax cDNA (corresponding to HTLV-I bp 5095-7357, Fig. 1A) using the NcoI site in the multiple cloning region of pGEM-T. Run-off transcription was performed in the presence of 50 ng of pGEM-T-tax plasmid linearized with BamHI, pGEM-T-tax plasmid, 500 M of each dNTP, 40 mM Tris-HCl, pH 8.0, 10 mM MgCl 2 , 5 mM dithiothreitol, digoxigenin-labeled dUTP, and T7 RNA polymerase. After incubation at 37°C for 120 min, the reaction was terminated by adding stop solution (95% formamide and 10 mM EDTA, pH 8.0).
In Vitro Ribozyme Reaction-The cleavage kinetics of the ribozymes were determined using tax mRNA obtained by run-off transcription from pGEM-T-tax. The cleavage reactions contained 2 pmol/ml tax mRNA in 50 mM Tris-HCl, pH 8.0, 0 -25 mM MgCl 2 , and 0.2-20 pmol/l ribozyme. After incubation for 1 h at 37°C, the reactions were stopped by adding 100 mM EDTA. The samples were heated for 5 min at 65°C, placed immediately on ice, loaded onto a 10% polyacrylamide, 8 M urea gel, and transferred to an Immobilon-S membrane (Millipore Corp., Bedford, MA). The cleavage products were detected using an antidigoxigenin antibody conjugated with alkaline phosphatase and visualized using an enzyme-linked color reaction (Boehringer Mannheim Corp., Mannheim, Germany).
Preparation of HVJ-Cationic Liposomes-HVJ-cationic liposomes were prepared as described elsewhere (12,13,25). In brief, a lipid mixture that contained 6 mg of phosphatidylcholine, 3 mg of cholesterol, and 0.75 mg of 3␤-[N-(NЈ,N-dimethylaminoethane)carbamoyl]cholesterol was dissolved in chloroform and evaporated using a rotary evaporator. The dried mixture was hydrated with 200 l of balanced salt solution (137 mM NaCl, 5.4 mM KCl, 10 mM Tris-HCl, pH 7.6) containing 100 g of ribozymes. Liposomes were prepared by vortex and extrusion. The liposomes were fused with HVJ that had been inactivated by ultraviolet light; the amount of virus used corresponded to 3 ϫ 10 4 hemagglutinating units. The HVJ-liposome complexes were separated from unfused HVJ by ultracentrifugation through a 30% (w/w) sucrose layer at 62,800 ϫ g for 90 min.
Generation of Tax-expressing Rat Embryonic Fibroblasts-Rat embryonic fibroblasts (Rat-1) transfected with the HTLV-I tax expression plasmid pH2R40M (Rat/Tax cells) have been described previously (17). The pH2R40M plasmid contains the SV40 promoter and polyadenylation signal, an R fragment of the HTLV-I long terminal repeat, a neomycin resistance gene, and a HindIII fragment derived from pX, which encodes intact Tax protein. The pH2R40M plasmid expresses the HTLV-I Tax protein, but not the Rex protein. The plasmid (5 g) was mixed with 10 g of Lipofectin (Life Technologies, Inc.) and added to the Rat-1 cells. Transfected cells were selected in medium containing with 800 g/ml G418 sulfate (Geneticin, Life Technologies, Inc.). The resulting tax-transfected Rat-1 cells (Rat/Tax cells) stably express the Tax protein. The Rat/Tax cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% heat-inactivated fetal calf serum, 4 mM L-glutamine, 50 units/ml penicillin, and 50 g/ml streptomycin at 37°C in a humidified 5% CO 2 atmosphere.
Analysis of Cellular Uptake of Ribozymes with Confocal Laser Scanning Microscopy and Flow Cytometry-To determine ribozyme uptake into the cells, 2 ϫ 10 5 Rat/Tax cells were incubated with 1 M naked FITC-labeled TR, 1 M FITC-labeled TR complexed with HVJ-cationic liposomes, or 1 M FITC-labeled TR complexed with HVJ-anionic liposomes for 24 h. FITC-labeled-rabbit IgG (10 g/ml) was used as a control. The cells were permeabilized with 70% ethanol for 30 min and incubated with phosphate-buffered saline (PBS) containing 250 g/ml propidium iodide (PI) for 15 min at room temperature in the dark. Cells from each group were mounted and analyzed by confocal laser scanning microscopy (Leica True Confocal Scanner 4D, Leica Lasertechnik GmbH, Heidelberg, Germany). The image consisted of 512 ϫ 512 pixels with the FITC detection system. Cells containing FITC-labeled ribozymes were counted by flow cytometry using an EPICS Profile counter (Coulter Electronics Inc., Hialeah, FL). The cut-off value for positive staining by PI was set at 10 on the y axis, and the cut-off value for cells positive for TR was set at 10 on the x axis in the histogram.
Quantitative Analysis of Tax Expression by Flow Cytometry-Rat/ Tax cells (2 ϫ 10 5 ) were incubated for 4 days without ribozymes, with 1 M naked TR, 1 M TR complexed with HVJ-anionic liposomes diluted by a factor of 1 ϫ 10 Ϫ2 , or 1 M TR complexed with HVJ-cationic liposomes diluted by a factor of 1 ϫ 10 Ϫ2 . The cells were permeabilized with 70% ethanol for 30 min. After incubation with saturating concentrations of anti-HTLV-I Tax mAb (1:500 dilution) for 30 min at 4°C, the cells were treated with FITC-labeled anti-mouse IgG (Cappel Research Products) as a second antibody. Control samples were treated with anti-FITC-labeled mouse IgG alone. The expression of Tax protein was quantified by flow cytometry (Coulter Electronics Inc).
Immunostaining for HTLV-I Tax Protein-Rat/Tax cells (3 ϫ 10 4 ) grown on glass coverslips for 4 days were treated with or without ribozymes as described above and fixed with 50% acetone, 50% methanol. Autofluorescence was quenched by treatment with 50 mM NH 4 Cl. The slides were rinsed in PBS containing 1% bovine serum albumin and incubated for 1 h in anti-HTLV-I Tax mAb (diluted 1:300). A tetramethylrhodamine isothiocyanate (TRITC)-conjugated secondary antibody, goat IgG fraction to mouse IgG (Cappel Research Products), was added at a 1:200 dilution and incubated for 30 min at room temperature. Negative control samples were similarly treated but were not exposed to the primary antibody. Confocal laser scanning microscopy (Leica True Confocal Scanner 4D) was used to determine the intracellular localization of both the HTLV-I Tax protein and ribozymes. Each image consisted of 1024 ϫ 1024 pixels with the two-color scanning analysis system.

RESULTS
In Vitro Cleavage of HTLV-I tax/rex mRNA by the Ribozymes-We synthesized two hammerhead ribozymes that were sequence-specific for a site downstream of the AUG start codons of the HTLV-I tax mRNA (TR) or rex mRNA (RR) (Fig.  1A). These sites had previously been targeted successfully by antisense ODNs (18,20) and thus were expected to be accessible for ribozyme-mediated cleavage. Incubation of the 220-base substrate HTLV-I tax/rex mRNA with RR resulted in the pre-dicted 97-base and 123-base cleavage products, whereas 148base and 72-base cleavage products were observed following incubation with TR (Fig. 1B). Both cleavage reactions were dose-dependent and were also dependent on the Mg 2ϩ concentrations (Fig. 1C). No cleavage products were observed when the mRNA was incubated with the inactive control ribozymes, RC and TC (Fig. 1B).
Ribozyme Uptake and Nuclear Transport after HVJ-Liposome-mediated Gene Transfer-We previously established a novel and highly efficient method for gene transfer using HVJliposomes (11,12). Gene delivery was even most efficient if the liposomes consisted of cationic rather than non-cationic lipids (13). In the present study, we have compared the gene transfer efficiency of ribozymes in the absence of HVJ-liposomes (naked ribozymes) and in the presence of HVJ-cationic or -anionic liposomes. To evaluate the intracellular distribution of the ribozymes, FITC-labeled ribozymes were visualized by confocal laser scanning microscopy. Cells of control samples containing FITC-conjugated rabbit IgG exhibited no staining ( Fig. 2A). Incubation of Rat/Tax cells with naked ribozymes for 24 h resulted in weak staining and spotty distribution of ribozymes in the cytoplasm of 1-3% of the cells (Fig. 2B). Examination of the cells at a higher magnification demonstrated that the naked ribozymes were distributed diffusely throughout the cytoplasm. TR complexed with HVJ-anionic liposomes showed a significantly improved cellular uptake compared with the naked ribozymes (Fig. 2C). Most of these ribozymes were retained in the cytoplasm and existed as vesicles in the endosomes or lysosomes (data not shown). Approximately 20% of cells also exhibited staining in their nuclei. In contrast, more than 90% of the Rat/Tax cells incubated with TR in the presence of HVJcationic liposomes exhibited cytoplasmic and nuclear fluorescence, which was more intense than in cells treated with anionic liposomes (Fig. 2D).
To quantitate the cellular uptake of TR complexed with HVJ-cationic liposomes, cells containing FITC-labeled ribozymes were counted by flow cytometry. Positive staining was indicated by a signal shift to the right on the x axis, compared with that of a negative control sample stained with mouse IgG (Fig. 3, left panels). Rat/Tax cells positive for nuclear staining by PI were located above the cut-off value of 10 on the y axis (Fig. 3, right panels, regions a and b). Cells containing FITClabeled ribozymes were located to the right of the cut-off value of 10 on the x axis (regions b and d). Region b of the histogram represented double-positive cells, whose proportion we estimated. These analyses indicated that naked TR was present in 5% of the 10 4 cells after a 24-h incubation (Fig. 3A), whereas ribozymes complexed with HVJ-anionic liposomes had been taken up into 22.8% of the cells (Fig. 3B). In contrast, ribozymes complexed with HVJ-cationic liposomes demonstrated a significantly increased uptake and were delivered into 96.1% of the cells (Fig. 3C). Thus, the uptake of ribozymes complexed with HVJ-cationic liposomes was 15-20 times increased compared with naked ribozymes and 4 -5 times increased compared with ribozymes complexed with HVJ-anionic liposomes.
Sequence-specific Ribozyme-mediated Cleavage of tax mRNA in Rat/Tax Cells-Rat/Tax cells express high levels of tax mRNA, but no rex mRNA, because the tax cDNA used transfected these cells contained naked 5Ј-rex sequences (17) (Fig.  4A). Thus, the tax mRNA from Rat/Tax cells could be cleaved by TR but not by RR. We used Western blot analysis to investigate whether the ribozymes introduced into the cells by HVJ-cationic liposome-mediated gene transfer possessed sequence-specific catalytic activity and therefore down-regulated Tax expression. The Tax protein signals were quantified by densitometric scanning. The 40-kDa Tax protein was highly detectable in Rat/Tax cells treated with 1 M naked TR (Fig.  4B, lane 1). Treatment with 1 M TR complexed with HVJcationic liposomes significantly reduced Tax expression by approximately 95%, compared with cells treated with 1 M naked TR (Fig. 4B, lane 2). In contrast, Tax protein expression was not affected by treatment with 1 M RR complexed with HVJcationic liposomes (Fig. 4B, lane 3). A reduction of nearly 20% in Tax protein expression was observed in cells treated with 1 M TC complexed with HVJ-cationic liposomes (Fig. 4B, lane  4). However, this effect of TC may be due to an antisense effect of the flanking TR sequences. Consistent with this hypothesis, a similar reduction in Tax expression of about 20% was also found in cells treated with 1 M tax antisense ODNs (18,20) complexed with HVJ-cationic liposomes (Fig. 4B, lane 5). In contrast, Tax expression was not reduced after treatment with 1 M tax sense ODNs, complexed with HVJ-cationic liposomes (Fig. 4B, lane 6).
Inhibition of Tax Protein Synthesis in TR-treated Cells-We compared the degree of down-regulation of Tax expression obtained when cationic or anionic liposomes that carried equal amounts of TR were used. Under these conditions, TR complexed with HVJ-cationic or HVJ-anionic liposomes produced a similar level of Tax down-regulation (data not shown). To evaluate the ribozyme achieved with each type of HVJ-liposome more precisely, we also compared Tax suppression after ribozyme transfer using diluted HVJ-liposomes. When the Rat/ Tax cells treated with 1 M TR complexed with HVJ-cationic liposomes diluted by a factor of 1 ϫ 10 Ϫ2 , a reduction of Tax expression occurred (Fig. 5, lanes 2-4). In contrast, 1 M TR complexed with HVJ-anionic liposomes diluted by a factor of 1 ϫ 10 Ϫ1 showed a decrease in Tax protein expression (Fig. 5,  lanes 5-7). Densitometric scanning analysis of a Tax immunoblot demonstrated that Tax signals were approximately 5 times lower in cells incubated with 1 M TR complexed with 1 ϫ 10 Ϫ2 diluted HVJ-cationic liposomes (Fig. 5, lane 3), compared with equally diluted HVJ-anionic liposomes (Fig. 5, lane 6).
We quantified the inhibitory effects of TR on Tax expression by flow cytometry using an anti-Tax mAb as reported previously (20). Rat/Tax cells expressed Tax protein at a high level (Fig. 6A). In cells treated with 1 M naked TR, the level of Tax expression was almost unaffected (Fig. 6B). Incubation of the cells with 1 M TR complexed with HVJ-anionic liposomes diluted by a factor of 1 ϫ 10 Ϫ2 (Fig. 6C), however, considerably down-regulated Tax expression, although to a lesser extent than did 1 M TR complexed with an equal volume of HVJcationic liposomes (Fig. 6D).
Tax Protein Degradation Induced by TR Accumulation in Rat/Tax Cells-The intracellular distribution and kinetics of ribozyme activity in the Rat/Tax cells was visualized using FITC-labeled ribozymes. The relationship between the ribozymes and Tax protein was investigated by a two-color confocal laser-fluorescence-microscopy scanning system using FITC-labeled ribozyme, which produces a green signal, and TRITC-labeled Tax protein, visible as an orange signal. A yellow signal indicated the co-localization of ribozyme and Tax protein. All untreated Rat/Tax cells expressed Tax protein both in the cytoplasm and in the nucleus, and the expression level did not differ among individual cells (data not shown). When the cells were incubated with TR complexed with HVJ-cationic liposomes for 4 days, Tax protein expression decreased markedly, resulting in the degradation of Tax protein, as indicated by the loss of orange signals (Fig. 7A). Approximately 5% of the cells did not incorporate TR and still expressed Tax protein (Fig. 7A, arrow). In contrast, following the introduction of RR mediated by HVJ-cationic liposomes, yellow signals were clearly evident in the cytoplasm and nuclei of the Rat/Tax cells (Fig. 7B). This observation suggests that Tax protein was still highly expressed following the nuclear deposition of RR. DISCUSSION The present study evaluated the efficiency of a method of gene transfer using anionic or cationic liposomes complexed with HVJ to introduce hammerhead ribozymes targeted against HTLV-I tax/rex mRNA into living cells. We also investigated the sequence specificity and activity of these ribozymes in the cells. We found that transfer efficiency was higher for ribozymes complexed with HVJ-cationic liposomes rather than with HVJ-anionic liposomes. The ribozymes specifically cleaved the HTLV-I tax/rex target mRNA, and TR suppressed Tax protein expression in the cells.
The ultimate goal of gene therapy for HTLV-I infections is the inactivation of viral genes in the infected cells. Tax protein has been suggested to play a role in tumorigenesis by modulating the expression of cellular genes (26). We previously reported that Rat/Tax cells were transformed and exhibited a marked increase in a cell refraction cell density (17,27). Treatment of Rat/Tax cells and tax-expressing murine fibroma cells derived from an HTLV-I tax transgenic mouse with TR that was complexed with HVJ-cationic liposomes reduced the density to the cells and inhibited the transformation, suggesting that TR can induce the ablation of tumorigenesis of HTLV-I tax-expressing cells. 2 TR that was complexed with HVJ-cationic liposomes successfully cleaved tax mRNA in the HTLV-I infected synoviocytes obtained from patients with chronic arthritis and induced apoptosis in these tax-expressing synoviocytes (28). These are several reasons for choosing ribozyme-mediated, rather than antisense-based, strategies for gene therapy in the treatment of disorders such as HTLV-I infection. We reported previously that antisense ODNs against tax successfully reduced Tax protein expression by up to 80% in HTLV-I-tax transformed cells in vitro and in vivo (18). However, this approach required large doses (10 -20 M) of tax antisense ODNs to obtain a significant inhibition of Tax expression. Attempts to inhibit gene expression using antisense ODNs have often been complicated by non-antisense effects, such as direct interactions with proteins (hybridization-independent effects) or hy- bridization to other mRNAs (hybridization-dependent effects) (29). In contrast, ribozymes, which contain self-cleaving RNA structures, allow the development of more sequence-specific strategies to inhibiting gene expression (30,31). After cleavage of the target mRNAs and dissociation from the cleaved molecules, ribozymes can be recycled to catalyze additional reactions (6,7). Accordingly, ribozyme-based approaches likely allow more sequence-specific and efficient gene inhibition, while requiring lower concentrations than antisense ODN-based approaches.
Both TR and RR successfully cleaved a target tax/rex mRNA in a cell-free system. The specificity of the TR-or RR-mediated cleavage was supported by our findings that inactive ribozymes that contained mutations in their catalytic domains (TC and RC) showed no cleavage activity. However, only TR, but not RR, effectively cleaved tax mRNA in Rat/Tax cells. These results confirmed the sequence-specific activities of our synthetic ribozymes in vitro and in cultured cells. In the cultured cells, however, we also observed a nearly 20% reduction in Tax expression after treatment with TC. It is likely that this effect is due to antisense effects of the flanking sequences of TR, because a similar down-regulation of Tax expression occurred in the presence of tax antisense ODNs. Scherr et al. (32) recently reported a noticeable antisense effect of ribozymes, because the incubation of HeLa cells with inactive variants of hammerhead ribozymes targeted against N-ras also resulted in a 20% reduction of N-ras levels. Thus, whereas the effects of ribozymes in cultured cells mainly result from the cleavage activities of the ribozymes, other factors (e.g. antisense effects) also may contribute to the gene inhibition. Although these effects may only be weak, they should not be ignored.
Only a few studies have previously investigated the activity of ribozymes in the living cell, which is difficult for ribozymes to penetrate effectively (7,9,10). Several factors appear to contribute to the efficiency of ribozyme uptake and activity in living cells. For example, unmodified RNA is subject to rapid degradation by nucleases upon its delivery to the cells. The half-life of a hammerhead ribozyme in serum is less than 0.1 min (33). However, a combination of modifications, including the introduction of several phosphorothioate linkages at the 5Ј-end of the ribozyme, can substantially increase ribozyme stability (34,35). We therefore used hammerhead ribozymes containing extensive phosphorothioate modifications at the 5Јand 3Ј-ends, which have been have been shown to confer higher stability in serum than the 5Ј-end modification alone, without reducing the catalytic efficiency of the ribozyme (36).
The efficacy of ribozymes in inhibiting gene expression also depends on their cellular uptake. Ribozymes generally have been introduced into cells by viral infection or by the transfection of expression vectors (6,9,10). Viral vectors can potentially lead to the incorporation of ribozyme-encoding genes into cellular chromosomes, thereby allowing their permanent expression. Consequently, many ongoing studies are aimed at developing suitable vectors that present a minimal risk to the host. However, in addition to the possibility that viral expression vectors may exert their own biological effects, they generally require long periods of incubation and usually have a low transfection efficiency. The present study used a ribozyme delivery system that employs cationic liposomes fused to the viral coat of HVJ. Ribozymes complexed with these HVJ-cationic liposomes showed a 15-20 times higher cellular uptake than naked ribozymes, and a 4 -5 times higher uptake than ribozymes complexed with HVJ-anionic liposomes. We also assessed the effect of the HVJ-cationic liposomes on cellular viability in Rat/Tax cells and mouse fibroblast, Balb/3T3 cells. There was no indication of significant cytotoxicity, even when excess amounts of HVJ-liposome-complexed ribozymes were used (data not shown).
We recently evaluated in vivo immunogenic reactions by the repeated injection of HVJ-liposomes into lung of rats. 3 Although anti-HVJ antibody titers were markedly elevated after the second administration of the HVJ-liposomes, the reporter gene was efficiently expressed, and only minimal inflammatory changes were detected after repeated administration. HVJliposomes adhere to and fuse with their target cells rapidly (i.e. within 1-2 min) (37), likely before they can be neutralized by antibody. No cytotoxic T-lymphocytes against HVJ were generated even after the repeated injection of HVJ-liposomes into the portal vein of rats. 4 Thus, we conclude that HVJ-liposomemediated method of gene transfer is safe, only weakly immunogenic, and highly efficient approach in living cells.
The efficiency of ribozymes also depends on the kinetics and location of their accumulation in the cells. The HVJ-cationic liposome-based gene transfer resulted in an accelerated transport of the ribozyme to the nucleus, with the nuclear localization of the ribozyme occurring in approximately 93% of the cells within 24 h. In contrast, HVJ-anionic liposomes were retained in the lysosomes, and ribozyme translocation into the nucleus had occurred in only 23% of the cells after 24 h of treatment. Finally, the naked ribozymes appeared to be localized in endosome vesicles and/or lysosomes for up to 24 h of treatment. This is consistent with the reported distribution of naked ODNs in cells (38). Naked ODNs and ODNs that are complexed with cationic lipids exhibit different intracellular behaviors as a result of their size difference (39). Thus, naked ODNs are taken up by pinocytosis, whereas ODNs associated with cationic lipids are taken up by phagocytosis and exhibit a delayed transfer to the lysosomes. Cationic liposomes can fuse with the cell membranes, allowing their content to be directly transferred into the cytoplasm, thus, avoiding their uptake by lysosomes (40). As a result, ribozymes complexed with treated with HVJcationic liposomes can escape lysosomal degradation by passing through the cell membrane, thereby leading to the accumulation of catalytically active ribozymes in the cytoplasm and nucleus.
Several factors may help to explain the difference in transfer efficiency and activity between ribozymes complexed with HVJ-cationic liposomes and HVJ-anionic liposomes. First, the trapping efficiency of ribozymes into the liposomes was about 6 times higher for HVJ-cationic liposomes than for HVJ-anionic liposomes. We previously analyzed the efficiency with which DNAs were trapped into liposomes (12,41). We found that, whereas the trapping efficiency of HVJ-cationic liposomes was about 60%, the efficiency of HVJ-anionic liposomes was only about 10%. Thus, the use of cationic lipids facilitates the entrapment of large amounts of negatively charged molecules such as plasmids, proteins, ODNs, and ribozymes in the liposomes. Second, the fusion efficiency of cationic liposomes with HVJ appears to be about 5 times higher than that of anionic liposomes. Third, the strong negative net charge of anionic liposomes seems to reduce the possibility that these liposomes associate and fuse with the plasma membranes of cultured cells. In contrast, HVJ-cationic liposomes, which have a neutral net charge, fuse efficiently with the cells (13).
To evaluate the cellular effects of the ribozymes in more detail, we compared the effects of ribozymes complexed with diluted HVJ-cationic or HVJ-anionic liposomes. The densitometric analysis of Tax Immunoblot demonstrated that Tax levels were approximately 5 times lower in cells incubated with 3 M. Yoshida and Y. Kaneda, submitted for publication. 4 M. Hirano and Y. Kaneda, submitted for publication. TR complexed with HVJ-cationic liposomes diluted by a factor of 1 ϫ 10 Ϫ2 , compared with TR complexed with equally diluted HVJ-anionic liposomes. Our findings suggest that HVJ-cationic liposomes are more effective in introducing ribozymes into cultured cells than HVJ-anionic liposomes, because the HVJcationic liposome-mediated delivery resulted in a more efficient ribozyme entrapment, a more effective cellular delivery, and an accelerated transport to the nucleus.
In summary, our results show that HVJ-cationic liposomes can be useful in delivering macromolecules such as ODNs and ribozymes into living cells. We believe that HVJ-cationic-liposome-mediated gene transfer method can be useful in molecular therapies for various diseases, including HTLV-I infection.