Engineered RNase P Ribozymes Are Efficient in Cleaving a Human Cytomegalovirus mRNA in Vitro and Are Effective in Inhibiting Viral Gene Expression and Growth in Human Cells*

By using an in vitro selection procedure, we have previously isolated RNase P ribozyme variants that efficiently cleave an mRNA sequence in vitro. In this study, a ribozyme variant was used to target the overlapping region of the mRNAs encoding human cytomegalovirus (HCMV) major transcription regulatory proteins IE1 and IE2. The variant is about 90 times more efficient in cleaving the IE1/IE2 mRNA sequence in vitro than the ribozyme derived from the wild type RNase P ribozyme. Our results provide the first direct evidence that a point mutation at nucleotide position 80 of RNase P catalytic RNA from Escherichia coli (U80→ C80) increases the rate of chemical cleavage, and another mutation at nucleotide position 188 (C188→ U188) enhances substrate binding of the ribozyme. Moreover, the variant is more effective in inhibiting viral IE1 and IE2 expression and growth in HCMV-infected cells than the wild type ribozyme. A reduction of about 99% in the expression level of IE1 and IE2 and a reduction of 10,000-fold in viral growth were observed in cells that expressed the variant. In contrast, a reduction of less than 10% in IE1/IE2 expression and viral growth was observed in cells that either did not express the ribozyme or produced a catalytically inactive ribozyme mutant. Thus, engineered RNase P ribozyme variants are highly effective in inhibiting HCMV gene expression and growth. These results also demonstrate the feasibility of engineering highly effective RNase P ribozymes for gene targeting applications, including anti-HCMV gene therapy.

By using an in vitro selection procedure, we have previously isolated RNase P ribozyme variants that efficiently cleave an mRNA sequence in vitro. In this study, a ribozyme variant was used to target the overlapping region of the mRNAs encoding human cytomegalovirus (HCMV) major transcription regulatory proteins IE1 and IE2. The variant is about 90 times more efficient in cleaving the IE1/IE2 mRNA sequence in vitro than the ribozyme derived from the wild type RNase P ribozyme. Our results provide the first direct evidence that a point mutation at nucleotide position 80 of RNase P catalytic RNA from Escherichia coli (U 80 3 C 80 ) increases the rate of chemical cleavage, and another mutation at nucleotide position 188 (C 188 3 U 188 ) enhances substrate binding of the ribozyme. Moreover, the variant is more effective in inhibiting viral IE1 and IE2 expression and growth in HCMV-infected cells than the wild type ribozyme. A reduction of about 99% in the expression level of IE1 and IE2 and a reduction of 10,000-fold in viral growth were observed in cells that expressed the variant. In contrast, a reduction of less than 10% in IE1/IE2 expression and viral growth was observed in cells that either did not express the ribozyme or produced a catalytically inactive ribozyme mutant. Thus, engineered RNase P ribozyme variants are highly effective in inhibiting HCMV gene expression and growth. These results also demonstrate the feasibility of engineering highly effective RNase P ribozymes for gene targeting applications, including anti-HCMV gene therapy.
Human cytomegalovirus (HCMV) 1 causes significant morbidity and mortality in immunocompromised or immunologically immature individuals, including neonates and AIDS patients (1,2). The emergence of drug-resistant strains of HCMV has posed a need for the development of new drugs and novel treatment strategies. RNA enzymes are being developed as promising gene-targeting agents to specifically cleave RNA sequences of choice (3,4). These ribozymes contain both a catalytic RNA domain that cleaves the target mRNA and a substrate-binding domain with a sequence antisense to the target mRNA sequence. Therefore, these gene-targeting ribozymes bind to the mRNA sequence through Watson-Crick interactions between the target sequence and the antisense sequence in the substrate-binding domain of the ribozyme. Compared with conventional antisense DNA and RNA, a ribozyme may have several unique features as it can cleave its target irreversibly, and multiple copies of its substrate can be cleaved by a single ribozyme molecule. Both hammerhead and hairpin ribozymes have been shown to cleave viral mRNA sequences and inhibit viral replication in cells infected with human viruses, whereas a ribozyme derived from a group I intron has been used to repair mutant mRNAs in cells (5)(6)(7)(8). Thus, ribozymes can be used as a tool in both basic research and clinical applications, such as in studies of developmental processes and in antiviral gene therapy (3,4).
RNase P is a ribonucleoprotein complex responsible for the 5Ј-end maturation of tRNAs (9,10). It catalyzes a hydrolysis reaction to remove a 5Ј leader sequence from tRNA precursors (pre-tRNA) and several small RNAs. In Escherichia coli, RNase P consists of a catalytic RNA subunit (M1 RNA) of 377 nucleotides and a protein subunit (C5 protein) of 119 amino acids (9,10). Under certain buffer conditions, such as 100 mM Mg 2ϩ , M1 RNA acts as a catalyst and cleaves pre-tRNAs in vitro in the absence of C5 protein (11). Extensive phylogenetic and biochemical analyses have provided significant insight into understanding the secondary structure (12,13) and the three-dimensional structure of M1 RNA (14 -17). These studies serve as a foundation for identifying the putative active site and substrate-binding site and for investigating the catalytic mechanism of this ribozyme.
Studies on substrate recognition by M1 RNA and RNase P have led to the development of a general strategy in which M1 RNA and RNase P can be used as gene targeting tools to cleave any specific mRNA sequences (18). One of the unique features of RNase P holoenzyme and its catalytic RNA is their ability to recognize the structures, rather than the sequences of their substrates, which gives them the ability to hydrolyze different natural substrates in vivo or in vitro. Thus, M1 ribozyme can cleave an mRNA substrate as long as the target sequence hybridizes with its complementary sequence (designated as external guide sequence) to form a complex resembling the portion of a tRNA molecule that includes the acceptor stem, the T-stem, the 3Ј CCA sequence, and the 5Ј leader sequence (18,19) (Fig. 1A). A sequence-specific ribozyme, M1GS RNA, can be constructed by covalently linking a guide sequence to the 3Ј terminus of M1 RNA (20 -23). We have shown previously that M1GS ribozymes cleaved the mRNA sequence encoding the thymidine kinase(TK) of herpes simplex virus type 1 in vitro and inhibited TK expression in herpes simplex virus type 1-infected cells (20, 24 -27). More recently, an M1GS ribozyme (23) was constructed to target the mRNA coding for HCMV IE1 and IE2, the major transcription regulatory proteins required for viral gene expression and growth (1,28,29). A reduction of 80% in the expression of IE1 and IE2 and a reduction of 150-fold in viral growth were observed in HCMV-infected cells that expressed the ribozyme (23).
Targeted cleavage of mRNA by RNase P ribozyme provides a unique approach to inactivate any RNA of known sequence expressed in vivo. In order to develop this ribozyme for practical use both as a research tool and as a therapeutic agent for gene-targeting applications, further studies are needed to improve M1GS RNA catalytic efficiency in vitro and its efficacy in vivo. By using an in vitro selection procedure, we have recently isolated M1GS ribozyme variants that are more efficient in cleaving a specific mRNA sequence (i.e. TK mRNA) than that derived from the wild type M1 RNA (26).
In this study, we used one of these ribozyme variants to target the HCMV IE1 and IE2 mRNA sequence, and we investigated its activity in cleaving the target mRNA in vitro and its efficacy in inhibiting IE1 and IE2 gene expression and viral growth in cultured cells. Our results indicate that the ribozyme derived from the selected variant is at least 90 times more efficient in cleaving the target HCMV mRNA sequence in vitro than the ribozyme derived from the wild type M1 RNA sequence. Biochemical characterization of the variant indicates that a point mutation at nucleotide position 80 of M1 RNA (U 80 3 C 80 ) increases the rate of chemical cleavage, and another mutation at nucleotide position 188 (C 188 3 U 188 ) enhances substrate binding of the ribozyme. These results provide insights into the catalytic mechanism of how RNase P ribozyme efficiently cleaves an mRNA substrate and generate guidelines for construction of highly active M1GS ribozymes for genetargeting applications.
The constructed ribozyme variant is also more effective in inhibiting IE1 and IE2 expression and HCMV growth in cultured cells. A reduction of 99% in the expression of IE1 and IE2 and a reduction of 10,000-fold in HCMV growth were observed in cells that expressed the ribozyme variant. These results provide the direct evidence that engineered RNase P ribozyme variants can be highly effective in inhibiting HCMV gene expression and growth. More importantly, our study demonstrates the feasibility of developing effective RNase P ribozymes for gene-targeting applications, including the treatment and prevention of HCMV infections.

EXPERIMENTAL PROCEDURES
Antibodies, Viruses, and Cells-The monoclonal antibodies against human actin and HCMV gH were purchased from Sigma and Biodesign, Inc., respectively. The monoclonal antibodies c1202, c1203, and c1205, which react with HCMV proteins UL44, IE1/IE2, and UL83, respectively, were purchased from the Goodwin Institute for Cancer Research (Plantation, FL). Human foreskin fibroblasts and astrocytoma U373MG cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (Invitrogen) as described previously (23). HCMV (strain AD169) was obtained from American Type Culture Collection (ATCC) and was propagated in human foreskin fibroblasts and U373MG cells.
The DNA template for the in vitro transcription of RNA substrate ie37 was constructed by annealing the T7 promoter-containing oligonucleotide OliT7 with oligonucleotide sIE1 (5Ј-CGGGATCCTTTCTC-GGGGTTCTCGTTGCGATCCTCGGTCCTATAGTGAGTCGTATTA-3Ј). The DNA that codes for ie37-3 was generated by PCR to replace the entire 12-nucleotide-long 3Ј tail sequence of ie37 with a thymidine nucleotide. The DNA sequences that code for RNA substrates ie37-5 and ie37-3 were constructed by PCR using oligonucleotide sIE1 as the template, OliT7 as the 5Ј primer, and oligonucleotides oliie37-5 (5Ј-CG-GGATCCTTTCTCGGGGTTCTCGTTGCTATAGTGAGTCGTATTA-3Ј) and oliie37-3 (5Ј-AGGGTTCTCGT-3Ј) as the 3Ј primers, respectively.
Construction of Retroviral Plasmids and Ribozyme-expressing Cells-Retroviral constructs RvM1-IE, RvC-IE, RvC-V188-IE, RvC-R27-IE, RvV188-IE, and RvR27-IE were constructed by placing the DNA sequences that code for M1-IE, C-IE, C-V188-IE, C-R27-IE, V188-IE, and R27-IE under the control of the U6 promoter in the LXSN retroviral vector, respectively (20,31). To generate cell lines containing the M1GS-retroviral DNA sequence, the retroviral vector DNAs that contain the ribozyme sequence were transfected into amphotropic PA317 cells by using a mammalian transfection kit (Invitrogen). At 48 h post-transfection, culture supernatants that contained retroviral vector particles were collected and used to infect human U373MG cells. At 48 -72 h postinfection, neomycin (Invitrogen) was added to the culture medium at a final concentration of 600 g/ml. Cells were subsequently selected in the presence of neomycin for 2 weeks, and neomycin-resistant cells were cloned (20,31). The level of M1GS RNA expression in individual cell clones was determined by Northern analysis with a probe complementary to the M1 RNA sequence. Only those cell clones that expressed similar levels of ribozymes were used for subsequent experiments.
Viral Infection and Assays for Viral Gene Expression and Growth-5 ϫ 10 5 cells were either mock-infected or infected with HCMV at a multiplicity of infection (m.o.i.) of 0.05-1 as described previously (23). The infected cells were incubated for a certain period of time (as stated under the "Results") before they were harvested for viral mRNA or protein isolation. Total cellular RNA and protein samples were prepared from the cells as described previously (23,32). To measure the levels of viral immediate-early (IE) transcripts, some of the cells were also treated with 100 g/ml cycloheximide prior to and during infection.
To determine the mRNA expression using Northern analyses, the RNA fractions were separated in 1-2.5% agarose gels that contained formaldehyde, transferred to nitrocellulose membranes, hybridized with the 32 P-radiolabeled DNA probes containing the HCMV DNA sequence, and analyzed with a STORM840 PhosphorImager. The radiolabeled DNA probes used to detect M1GS RNAs, actin mRNA, HCMV immediate-early 5-kb RNA transcript, IE1 mRNA, IE2 mRNA, and US2 mRNA were synthesized from plasmids pFL117, p␤-actin RNA, pCig27, pIE1, pIE2, and pCig38, respectively, by using a random primed labeling kit (Roche Applied Bioscience) (23,33). The viral 5-kb RNA was transcribed from the region covering HCMV open reading frames between UL106 and UL110 (33). 2 To examine the protein expression using Western analyses, the denatured polypeptides from cell lysates were separated on SDS-polyacrylamide gels cross-linked with N,NЉ-methylenebisacrylamide, transferred electrically to nitrocellulose membranes, and reacted in an enzyme-linked immunoassay with antimouse IgG conjugated with alkaline phosphatase in addition to the antibodies against human actin and HCMV IE1/IE2, UL44, UL83, and gH. The membranes were subsequently stained with a chemiluminescent substrate with the aid of a Western chemiluminescent substrate kit (Amersham Biosciences) and quantitated with a STORM840 Phos-phorImager. Quantitation was performed in the linear range of RNA and protein detection.
To determine the efficacy of the ribozymes in inhibiting viral growth, 5 ϫ 10 5 cells were infected with HCMV at an m.o.i. of 0.5-2. The cells and medium were harvested at 1-day intervals through 7 days postinfection, and viral stocks were prepared by adding an equal volume of 10% skim milk, followed by sonication. The titers of the viral stocks were determined by infecting 1 ϫ 10 5 human foreskin fibroblasts and counting the number of plaques 10 -14 days postinfection. The values obtained were the average from triplicate experiments.
It has been shown that cloned U373MG cells exhibit clonal variability in their susceptibility to HCMV infection (1). To exclude the possibility that the observed inhibition of viral gene expression and growth is due to the variability of the particular cloned U373MG cells used in the experiments rather than the ribozyme-mediated inhibition, two additional sets of cloned cells that expressed different ribozymes were used to assay the efficacy of the ribozymes in inhibiting HCMV gene expression and growth. Moreover, three additional cloned cell lines that expressed each of the ribozymes were mixed and used as the oligoclonal cell populations. The antiviral efficacies of this set of oligoclonal cell populations were also determined.
Single Turnover Kinetic Analyses of in Vitro Cleavage Reactions-The cleavage reactions of substrate ie37 by different ribozymes were carried out in buffer A (50 mM Tris, pH 7.5, 100 mM NH 4 Cl, 100 mM MgCl 2 ), and single turnover kinetic analyses to determine the values of the observed cleavage rate (k obs ) were performed as described previously (26,34). Analyses were performed with a trace amount of radioactive substrate and an excess of ribozyme. The concentration of radioactive substrate was less than 0.01 nM, and the concentrations of ribozyme tested ranged from 0.05 to 5 nM. Variations in the amount of substrate did not affect the observed cleavage rate (k obs ) at a fixed excess ribozyme concentration, and the reaction followed pseudo firstorder kinetics. Pseudo first-order rate constants of cleavage (k obs ) were assayed at each ribozyme concentration by the slope of a plot Ϫln((F t Ϫ F e )/(1 Ϫ F e )) versus time using a Kaleidagraph program (Synergy Software, Reading, PA). F t and F e represent the fraction of the substrate at time t and the end point (Ͼ10 h) of the experiments, respectively. The values of the overall cleavage rate (k cat /K m ) s were calculated by the slope of a least squares linear regression (Kaleidagraph) of a plot of the values of k obs versus the concentrations of the ribozymes (26,34). The values obtained were the average of three experiments.
Determination of Binding Dissociation Constant (K d ) and Apparent Reaction Rate Constant k app -The procedures to measure the equilibrium dissociation constants (K d ) of complexes of the ribozymes and the substrates were modified from Pyle et al. (35). In brief, various concentrations of M1GS RNAs were preincubated in buffer E (50 mM Tris, pH 7.5, 100 mM NH 4 Cl, 100 mM CaCl 2 , 3% glycerol, 0.1% xylene cyanol, 0.1% bromphenol blue) for 10 min before mixing with an equal volume of different concentrations of substrate RNA preheated under identical conditions. The samples were incubated for 10 -120 min to allow binding, then loaded on a 5% polyacrylamide gel, and run at 10 watts. The electrophoresis running buffer contained 100 mM Tris-Hepes, pH 7.5, and 10 mM MgCl 2 (35). The K d value was then extrapolated from a graph plotting percent of product bound versus M1GS RNA concentration. The values were the average of three experiments.
The experimental procedures to determine the apparent reaction rate constant k app were carried out as described previously (26). In brief, equimolar amounts of substrates and ribozymes were incubated in buffer B (16.6 mM PIPES, 40 mM Tris-HCl, pH 6.0, 100 mM NaCl, 100 mM CaCl 2 ) at 37°C for 10 min to allow binding. After binding, the bound ribozyme-substrate complexes were separated from the unbound substrates using G-50 Sephadex gel filtration columns (Roche Applied Science). The eluted samples were examined in 5% polyacrylamide nondenaturing gels for the presence of labeled ribozyme-substrate complexes, and the concentrations of the complexes were quantitated by a STORM840 PhosphorImager. Ribozyme-substrate complexes were then diluted in different concentrations (2-70 nM) and incubated in buffer C (16.6 mM PIPES, 40 mM Tris-HCl, pH 6.0, 100 mM NaCl, 100 mM MgCl 2 ) at 37°C. Aliquots were withdrawn from reaction mixtures at regular intervals (from 0 to 160 min). The cleavage products were separated on 15% denaturing gels, autoradiographed, and quantitated with a STORM840 PhosphorImager. The apparent rate constant, k app , was calculated by the slope of the plot of ln(S 0 /S t ) versus time, where S 0 equals the initial substrate concentration and S t equals the substrate concentration at a given time point. The values obtained were the average of three experiments. Cleaving the  IE1/IE2 mRNA Sequence-IE1 and IE2, which are the viral immediate early (IE or ␣) gene products and share 85 aminoterminal amino acids due to alternative splicing and polyadenylation of transcripts initiating at a strong promoter-enhancer (1,28,29), function as the major transcriptional regulators and are required for the expression of viral early (␤) and late (␥) genes (36,37). Therefore, targeting the overlapping region (e.g. exons 1-3) of the mRNAs coding for IE1 and IE2 should simultaneously shut down the expression of both proteins and may yield more effective inhibition of viral replication. Choosing a target region that is accessible to binding of M1GS ribozymes is important in order to achieve efficient targeting. This is because most mRNA species inside cells are associated with proteins and are present in folded conformations. We have used an in vivo mapping approach with dimethyl sulfate (20,38,39) to determine the accessible regions of IE1 and IE2 mRNAs. A position 39 nucleotides upstream from the 3Ј terminus of exon 3 was chosen as the cleavage site for the M1GS ribozyme. The targeted region (designated as IE mRNA), which is in the overlapping region of IE1 and IE2 mRNAs (40), is one of the sequences most accessible to dimethyl sulfate modification and is likely accessible to ribozyme binding (data not shown).

Engineered RNase P Ribozymes Are Efficient in
We have used previously (26) an in vitro selection procedure to isolate M1GS RNA variants that are more efficient in cleaving the TK mRNA sequence than the ribozyme derived from the wild type M1 RNA. The objective of the study was to generate highly active RNase P ribozyme variants that can be used for targeting TK mRNA as well as any other mRNAs and to study the variants to understand the catalytic mechanism of RNase P ribozymes in cleaving mRNA substrates. However, little is currently known about the mechanism of how some of these active variants increase their cleavage activity in vitro. Whether these variants can be used for the construction of highly active ribozymes for inhibition of HCMV gene expression and growth has not been extensively studied. In this study, we addressed these issues by studying one of the variants, R27. This variant, as well as its derivative ribozyme (i.e. R27-IE), are among the most active M1GS RNAs in cleaving the TK as well as the IE1/IE2 mRNA sequences in vitro (see below and Table I). Thus, we chose to characterize the activity of R27 in cleaving IE1/IE2 sequence in vitro and its efficacy in inhibiting IE1/IE2 expression in HCMV-infected cells, and we investigate the mechanism of how the mutations increase its catalytic activity. We hope that these results will provide insight into how RNase P ribozyme efficiently cleaves an mRNA substrate and, furthermore, will generate general guidelines for construction of highly active ribozymes for gene-targeting applications.
Ribozyme R27-IE was constructed by covalently linking the 3Ј terminus of R27 RNA to a guide sequence of 13 nucleotides that is complementary to the targeted mRNA sequence. R27 contains point mutations U 80 3 C 80 and C 188 3 U 188 (Fig. 1C) (26). Little is known about the functional role of U 80 in M1GS RNA cleavage, whereas our recent nuclease footprint analyses and UV cross-linking studies suggest that C 188 is in close proximity to the 3Ј tail sequence of an mRNA substrate and is probably involved in binding the mRNA substrate (24,25,27). In order to study the function of these two mutations in the cleavage of the IE mRNA sequence, we constructed two additional ribozyme variants, V80-IE and V188-IE, which were derived from M1 RNA and contained the mutations U 80 3 C 80 and C 188 3 U 188 , respectively. Two other M1GS ribozymes, M1-IE and C-IE, were also constructed by linking the guide sequence to the 3Ј termini of the wild type M1 RNA sequence and mutant C102 RNA sequence, respectively. C102 RNA is derived from M1 RNA and contains several point mutations at the P4 catalytic domain and is at least 1 ϫ 10 5 -fold less active than M1 RNA (30). Cleavage of RNA substrate ie37, which contains the targeted HCMV mRNA sequence of 37 nucleotides (Fig. 1B), was observed in the presence of R27-IE, V80-IE, V188-IE, and M1-IE (Fig. 2, lanes 3 and 4, and Table I). In contrast, cleavage of the same substrate by C-IE was barely detected due to the point mutations at the catalytic center region (Fig. 2, lane 5). To determine whether the variant is more active in cleaving the IE mRNA sequence, cleavage of ie37 by the ribozymes was assayed, and the overall cleavage rate (k cat /K m ) s of the reactions was determined using kinetic analyses under single turnover conditions. Ribozyme R27-IE, V80-IE, and V188-IE were about 90, 8, and 16 times more active than M1-IE that was derived from the wild type M1 RNA sequence (Table I). These observations indicate that ribozyme variants selected for their efficient activities in cleaving TK mRNA are also very active in cleaving IE mRNA in vitro, suggesting that the selected R27 domain can be used for the construction of ribozymes that are highly efficient in cleaving other mRNAs. Ribozyme R27-IE still exhibited at least 80-fold higher cleavage activity than M1-IE when a 450-nucleotidelong substrate of IE2 mRNA sequence was used (data not shown). This substrate includes the ie37 sequence and, in addition, contains a 150-nucleotide-long 5Ј leader and a 287nucleotide-long 3Ј tail sequence. These results suggest that R27-IE may also cleave the full-length IE1 and IE2 mRNAs more efficiently than M1-IE.
Ribozyme Variant Increases Its Activity by Enhancing the Rate of Cleavage-The values of (k cat /K m ) s obtained under single turnover conditions reflect the rates of substrate binding and chemical cleavage of the phosphodiester linkage. Accordingly, experiments were carried out with the variant to determine whether a change in the rates of these steps contributed to the increased (k cat /K m ) s values. To determine whether the mutations affect the rate of chemical cleavage, ie37 was allowed to form active complexes with the ribozymes in the presence of divalent ions, and the apparent reaction rate constant, k app , for the complexes was measured. CaCl 2 was used instead of MgCl 2 as the source of divalent ions, in order to reduce the rate of cleavage while allowing proper folding of the ribozymes and substrates and preserving the interactions between the ribozyme and the substrate in an active ribozymesubstrate complex. In the presence of CaCl 2 , M1 RNA cleaves a pre-tRNA substrate at least 50 times slower than it does in the presence of MgCl 2 (26,27,41,42). Moreover, the rate of cleavage of ie37 was at least 200-fold slower in the presence of CaCl 2 than in the presence of MgCl 2 (data not shown). Meanwhile, nuclease footprint analyses and UV cross-linking studies indicate that the regions of the ribozymes that potentially interact with the substrates in the presence of Ca 2ϩ ions are similar to those found in the presence of Mg 2ϩ ions, suggesting that the interactions between the substrates and ribozymes are similar in the presence of these different divalent ions (26,27,41,42). To assay the values of k app , equimolar amounts of ribozymes and ie37 were mixed in the presence of CaCl 2 , and the ribozyme-substrate complexes were separated from the unbound substrates using G-50 Sephadex columns. The ribozyme-ie37 complexes were then diluted in different concentrations and further incubated in the presence of 100 mM MgCl 2 to allow cleavage. The k app values were independent of the concentrations of the complexes within the tested range of 2-70 nM (Table II). These results suggest that substrate cleavage takes place predominantly in the bound complexes (cis-cleavage) as observed previously (25,26). The values of k app for the selected variants were at least 1 ϫ 10 2 -fold higher than those of C-IE (Table II). Moreover, the values of k app for V188-IE were similar to those of M1-IE, suggesting that this variant exhibits a similar rate of chemical cleavage as M1-IE. Meanwhile, the values of k app for R27-IE and V80-IE were about 10 times higher than those of M1-IE (Table II). V80-IE only contains the mutation U 80 3 C 80 , which is also found in R27-IE. These results suggest that this mutation is responsible for the increased cleavage activity of V80-IE. A "rescued" ribozyme, rV80-IE, was derived from V80-IE, in which C 80 was changed back to U 80 and the wild type sequence was restored. Ribozyme rV80-IE exhibited similar activity as M1-IE (data not shown), suggesting that the mutated C 80 functions to enhance the rate of cleavage (k app ) and is indeed responsible for the increased activity of V80-IE.
Ribozyme Variant Increases the Substrate Binding by Enhancing the Interactions to the 3Ј Tail Sequence of the mRNA Substrate-An increase in the overall cleavage rate ((k cat /K m ) s ) of the variants may also be due to additional tertiary interactions between the ribozyme and the substrate. These interactions may result in better binding and docking of the substrate to the active site of the ribozyme. To determine whether this is the case, the binding affinities of the ribozymes to the IE mRNA sequence were determined by a gel-shift binding assay, in which ribozyme-substrate complexes were separated in polyacrylamide gels under non-denaturing conditions. To determine the binding affinity of the ribozymes to the substrate rather than to the product, the binding assays were carried out under the buffer conditions that contain Ca 2ϩ but not Mg 2ϩ . During the binding assays in the presence of CaCl 2 , the cleavage products were barely detected, although substantial amounts of products were found under the optimal cleavage conditions (e.g. 100 mM MgCl 2 ) (data not shown) (26,27).
The binding affinities of the ribozymes to the substrates, measured as the dissociation constant (K d ), were determined and are shown in Table I. V188-IE and R27-IE exhibited about 65-fold higher binding affinity to ie37 than M1-IE. Since V188-IE only contains mutation C 188 3 U 188 , which is also found in R27-IE, it is likely that this mutation is responsible for

) s ) and binding affinity (K d ) in reactions of substrates ie37 and ie37-3 with
RNase P mutants and the ribozyme (M1-IE) derived from the wild type M1 RNA sequence Single turnover kinetic analyses to determine the values of (k cat /K m ) s were carried out in buffer A (50 mM Tris, pH 7.5. 100 mM NH 4 Cl, 100 mM MgCl 2 ) as described previously (26,30). Binding assays were carried out in buffer E (50 mM Tris, pH 7.5, 100 mM NH 4 Cl, 100 mM CaCl 2 , 3% glycerol, 0.1% xylene cyanol, 0.1% bromphenol blue), using a protocol modified from Pyle et al. (35).  (24,25,27) that different regions of the ribozymes potentially interact with specific parts of an mRNA substrate. To determine further which part of the substrate is important for the increased binding affinity between the substrates and the ribozymes, binding assays were carried out using substrates ie37-5 and ie37-3, which were derived from ie37 by deleting the 5Ј leader and 3Ј tail sequence, respectively (Fig. 1B). The binding affinity of M1-IE to ie37-3 was three times weaker than its affinity to ie37 (Table I). The binding affinities of V188-IE and R27-IE to ie37-3, however, were about 190 times weaker than those to ie37. Indeed, the K d values of V188-IE and R27-IE to ie37-3 were similar to those of M1-IE. Meanwhile, the 5Ј leader sequence appears to be not important for the increased binding affinity of the variants since R27-IE and V188-IE still exhibit 60 times higher binding affinity to ie37-5 than M1-IE (data not shown). These results suggest that the presence of the 3Ј tail sequence contributes to the increased binding affinity, and deleting this sequence results in the loss of the increased affinity.
The results from kinetic analyses of the cleavage of ie37-3 by the ribozymes further support the notion that the variants strengthen their substrate binding by enhancing their interactions with the 3Ј tail sequence of the substrate. V188-IE did not exhibit higher activity ((k cat /K m ) s ) than M1-IE in cleaving ie37-3, whereas this variant was about 15 times more active in cleaving ie37 than M1-IE (Table I). Cleavage of this substrate by V188-IE and R27-IE was about 26 times slower than their FIG. 1. A, schematic representation of a natural substrate (precursor tRNA (ptRNA)), a small model substrate (external guide sequence (EGS):mRNA) for ribonuclease P and M1 RNA from E. coli, and a complex formed between a M1GS RNA and its mRNA substrate. A part of the external guide sequence-mRNA and M1GS-mRNA complexes resembles the acceptor stem and T-stem domains of a (precursor tRNA) (shown in boldface type). The site of cleavage by RNase P or M1 RNA is marked with a solid arrowhead. B, schematic representation of the substrate used in the study. The targeted sequences that bind to the guide sequences of the ribozymes are boxed. The regions upstream and downstream from the targeted sequence represent the 5Ј leader sequence and the 3Ј tail sequence, respectively. C, the mutated positions (U 80 and C 188 ) (boxed) of ribozyme variant R27 at the secondary structure of the RNA subunit (M1 RNA) of RNase P from E. coli (12,13). The mutated nucleotides (A 347 C 348 3 C 347 U 348 , C 353 C 354 C 355 G 356 3 G 353 G 354 A 355 U 356 ) in C102 (30), which was used to construct C-IE, C-V188-IE, and C-R27-IE, are also shown (circled). cleavage of ie37, respectively. In contrast, the rate of cleavage of ie37-3 by M1-IE was only about 2-fold slower than that of ie37. Thus, the presence of the 3Ј tail sequence appears to be important for the increased cleavage efficiency of V188-IE and R27-IE.
Retroviral Vector-mediated Expression of Ribozyme in Human Cells-The highly active R27-IE and V188-IE as well as M1-IE and C-IE were cloned into the retroviral LXSN vector and placed under the control of the small nuclear U6 RNA promoter. This promoter, from which transcripts are highly expressed and primarily localized in the nucleus, is transcribed by RNA polymerase III and has been shown previously to express M1GS RNA and other RNAs steadily in human cells (31,(43)(44)(45)(46). Two additional ribozymes, C-V188-IE and C-R27-IE, were also cloned into the retroviral vector. C-V188-IE and C-R27-IE contain the mutations found in C102 and are catalytically inactive but contain the same guide sequence as V188-IE and R27-IE (Fig. 2, lanes 1 and 2). Indeed, C-V188-IE and C-R27-IE exhibited similar binding affinities to ie37 as V188-IE and R27-IE (Table I). Therefore, these ribozymes can be used as the control for the antisense effect of the guide sequence.
To generate the M1GS ribozyme-expressing cell lines, amphotropic packaging cells (PA317) (31) were transfected with LXSN-M1GS DNAs to produce retroviral vectors that contained the genes for M1GS RNA. Human U373MG cells were then infected with these retroviral vectors, and cells that expressed the retroviral vectors and ribozymes were cloned. The constructed cell lines and a control cell line in which cells were transfected with LXSN vector DNA were indistinguishable in terms of cell growth and viability for up to 3 months (data not shown), suggesting that the expression of the ribozymes did not exhibit significant cytotoxicity. Northern analyses with a DNA probe that is complementary to M1 RNA were carried out to determine the level of ribozyme expression in individual cell clones (Fig. 3B), using the expression of actin mRNA as the internal control (Fig. 3A). The M1GS RNAs were primarily expressed in the nuclei as they were only detected from the nuclear but not cytoplasmic RNA fractions (data not shown). Only the cell lines that expressed similar levels of these ribozymes were used for further studies in tissue culture.
Ribozyme Variants Are More Effective in Inhibiting IE1/IE2 Expression-To determine the efficacy of the ribozymes in reducing IE1/IE2 expression, cells were infected with HCMV at an m.o.i. of 0.05-1. Total RNAs were isolated from cells that were pretreated with 100 g/ml cycloheximide and then were either mock-infected or infected with HCMV. Under this condition, only viral IE mRNAs were synthesized (1). The levels of the IE1 and IE2 mRNAs (Fig. 4B) in the infected cells were determined by Northern analyses, using the level of the 5-kblong viral immediate-early transcript as the internal control (Fig. 4A). The viral 5-kb RNA is transcribed from the region covering HCMV open reading frames between UL106 and UL110 (33). 2 The expression of this RNA is not regulated by IE1/IE2 under the assay conditions (33). A reduction of about 99, 93, and 80% (average of three experiments) in the levels of IE1 and IE2 mRNA expression was observed in cells that expressed R27-IE, V188-IE, and M1-IE, respectively (Table  III). In contrast, cells that expressed C-IE, C-V188-IE, and C-R27-IE only exhibited a reduction of less than 10% (Table  III). Since these control ribozymes contain the same guide sequence as M1-IE, V188-IE, and R27-IE but do not exhibit catalytic activity, the low level of inhibition found in cells that expressed C-IE, C-V188-IE, and C-R27-IE RNAs was probably due to an antisense effect. Thus, these results suggest that the catalytic cleavage of the target mRNAs by M1-IE, V188-IE, and R27-IE is responsible for the significant reduction of IE1 and IE2 mRNA expression in cells that expressed these ribozymes.
The levels of IE1 and IE2 proteins in M1GS-expressing cells are expected to be lowered due to reduced levels of IE1/IE2 mRNAs. Proteins were isolated from cells at 24 -48 h postinfection, separated in SDS-polyacrylamide gels, and transferred to membranes. One membrane was stained with an anti-IE1/ IE2 antibody (anti-IE1/IE2) (Fig. 5B) and another membrane with a monoclonal antibody against human actin (anti-actin) (Fig. 5A). The latter serves as an internal control for the quantitation of IE1/IE2 protein expression. The results of three independent experiments are summarized in Table III; a reduction of about 98, 93, and 80% in the level of IE1 and IE2 proteins was observed in cells that expressed R27-IE, V188-IE, and M1-IE RNA, respectively. In contrast, a reduction of less than 10% was found in cells that expressed C-IE, C-V188-IE, and C-R27-IE RNAs.

TABLE II
The apparent rate constants (k app ) of the ribozyme-substrate complexes The ribozyme-ie37 complexes were separated from the unbound substrates by Sephadex G-50 gel filtration columns and diluted in different concentrations before the rate of cleavage was assayed. The cleavage products were separated on 15% polyacrylamide gels and quantitated with a STORM840 PhosphorImager.  1 and 6) or from cell lines that expressed R27-IE (lanes 2 and 7), V188-IE (lanes 3 and 8), M1-IE (lanes 4 and 9), and C-IE (lanes 5 and 10). Equal amounts of each RNA sample (30 g) were separated on 1 (A) or 2.5% (B) agarose gels that contained formaldehyde, transferred to a nitrocellulose membrane, and hybridized to a 32 P-radiolabeled probe that contained the DNA sequence coding for actin (A) or M1 RNA (B). The actin mRNA expression (A) was used as the internal control.

Ribozyme Variants Are More Effective in Inhibiting Viral
Gene Expression and Growth-Inhibition of IE1 and IE2 expression is expected to result in a reduction of the expression of both viral early (␤) and late (␥) genes (1,36,37). To determine whether this is the case, cells were infected with HCMV at an m.o.i. of 0.1-1 for 48 -72 h. The level of the US2 mRNA (an early mRNA) (Fig. 4D) as well as the protein levels of UL44 (an early and late protein), UL83 (a late protein), and gH (a late protein) (Fig. 5, C and D, data not shown) were determined. The level of the HCMV 5-kb transcript and the protein level of human actin were used as the internal controls. A reduction of about 97-98, 91-92, and 74 -76% in the expression levels of these genes was observed in cells that expressed R27-IE, V188-IE, and M1-IE RNA, respectively. No significant reduction was detected in cells that expressed C-IE, C-V188-IE, and C-R27-IE (Figs. 4 and 5 and Table III). These results suggest an overall inhibition of viral early and late gene expression in the cells that expressed M1-IE, V188-IE, and R27-IE.
To determine whether viral growth was blocked in the ribozyme-expressing cells, cells were infected with HCMV at an m.o.i. of 0.5-2. Virus stocks were prepared from the infected cultures at 1-day intervals through 7 days postinfection, and the count of plaque-forming unit was determined by measuring the viral titer in human fibroblasts. After 5 days of postinfec-tion, a reduction of about 10,000-, 1000-, and 150-fold in viral yield was observed in cells that expressed R27-IE, V188-IE, and M1-IE, respectively (Fig. 6). No significant reduction was found in those cells that expressed the control ribozymes C-IE, C-V188-IE, or C-R27-IE (Fig. 6, data not shown).
It has been shown that cloned U373MG cells exhibit clonal variability in their susceptibility to HCMV infection (1). To exclude the possibility that the observed inhibition of viral gene expression and growth is due to the variability of the particular cloned U373MG cells used in the experiments rather than the ribozyme-mediated inhibition, two additional sets of cloned cells that expressed different ribozymes were used to assay the efficacy of the ribozymes in inhibiting HCMV gene expression and growth. Moreover, three additional cloned cell lines that expressed each of the ribozymes were mixed and used as the oligoclonal cell populations. The antiviral efficacies of this set of oligoclonal cell populations were also determined. The levels of inhibition of viral gene expression and growth in these cell lines are similar to the levels observed in the set of cloned cell lines described in Table III and Fig. 6. These results indicate that the observed inhibition of viral gene expression and growth is due to the ribozyme-mediated cleavage but not the clonal variability of the susceptibility of the U373MG cells to HCMV infection.  8 -12, 14 -18, and 20 -24) and were harvested at either 10 (A-C) or 24 h (D) postinfection. Viral infection was carried out either in the absence (D) or presence (A-C) of 100 g/ml cycloheximide. RNA samples were isolated from parental U373MG cells (Ϫ, lanes 1, 6, 7, 12, 13, 18, 19, and 24) and cell lines that expressed R27-IE (lanes 2, 8, 14, and 20), V188-IE (lanes 3, 10, 15, and 21), C-IE (lanes 4, 10, 16, and 22), and M1-IE (lanes 5, 11, 17, and 23). These samples (30 g) were then separated on agarose gels that contained formaldehyde, transferred to a nitrocellulose membrane, and hybridized to 32 P-radiolabeled probes that contained the cDNA sequence of the HCMV 5-kb transcript (lanes 1-6), IE1 mRNA (lanes 7-12), IE2 mRNA (lanes [13][14][15][16][17][18], and US2 mRNA (lanes 19 -24). The hybridized products corresponding to the 5-kb RNA, IE1, IE2, and US2 mRNAs were about 5, 1.9, 2.2, and 1.5 kb, respectively (1,28,33). RNA probes were used in great excess of the detected RNA species.

DISCUSSION
Ribozymes have been shown to be promising antiviral agents for inhibition of viral gene expression and replication (3,4). The M1GS-based technology represents an attractive approach for gene inactivation since it generates catalytic and irreversible cleavage of the target RNA by using M1 RNA, a highly active RNA enzyme found in nature (9,10). M1 RNA may further increase its activity in cultured cells by interacting with the cellular proteins (9,20). These properties, as well as the simple design of the guide sequence, make M1GS an attractive and unique gene-targeting agent that can be used for antiviral as well as other in vivo applications.
Our previous studies indicate that expression of an M1GS RNA derived from the wild type M1 RNA sequence inhibits HCMV gene expression by 75-80% and reduces viral growth by 150-fold in cultured cells (23). Understanding the mechanism of ribozyme cleavage and improving its efficacy is essential in developing the M1GS ribozyme for practical anti-HCMV appli-cations. However, little is currently known about the ratelimiting step of the M1GS RNA cleavage reaction in cells. Equally unclear is whether the efficacy of the ribozymes can be improved and, if so, how it can be improved. In the present study, we showed that an RNase P variant, R27-IE, exhibited 90 times higher rate of cleavage in vitro in cleaving the IE1/IE2 mRNA sequence than the ribozyme (i.e. M1-IE) derived from the wild type M1 RNA sequence. Moreover, R27-IE inhibited the expression of IE1 and IE2 in cultured cells by more than 98% and was more effective in cultured cells than M1-IE, which reduced IE1 and IE2 expression by about 80% (Table III). A reduction of about 10,000-fold in viral growth was observed in the R27-IE-expressing cells, whereas a reduction of about 150fold was observed in M1-IE-expressing cells (Fig. 6). In contrast, a reduction of less than 10% in the levels of IE1/IE2 expression and viral growth was observed in cells that expressed C-IE, C-V188-IE, or C-R27-IE. Ribozymes C-IE, C-V188-IE, and C-R27-IE had the identical guide sequence and  1, 7, 13, and 19) or infected with HCMV (m.o.i. ϭ 0.5) (lanes 2-6, 8 -12, 14 -18, and 20 -24) and were harvested at 24 -72 h postinfection. Protein samples were isolated from parental U373MG cells (Ϫ, lanes 1, 5, 7, 11, 13, 17, 19, and 23) or from cell lines that expressed R27-IE (lanes 2, 8, 14, and 20), V188-IE (lanes 3, 9, 15, and 21), C-IE (lanes 4, 10, 16, and 22), and M1-IE (lanes 6, 12, 18, and 24). These samples (40 g) were then separated in either SDS-9% polyacrylamide gels (A) or SDS-7.5% polyacrylamide gels (B-D) and transferred to membranes. One membrane was allowed to react with a monoclonal antibody (anti-actin) against human actin (lanes 1-6) (A) whereas other membranes were stained with the monoclonal antibodies against viral IE1/IE2 (lanes 7-12) (B), UL44 (lanes 13-18) (C), and UL83 proteins (lanes 19 -24) (D). All antibodies were used in great excess of the detected antigens.