Gene Therapy (2001) 8, 1863–1871  2001 Nature Publishing Group All rights reserved 0969-7128/01 $15.00 www.nature.com/gt RESEARCH ARTICLE Long-term RNase P-mediated inhibition of HIV-1 replication and pathogenesis

HJ Hnatyszyn, G Spruill, AK Young, R Seivright and G Kraus Department of Microbiology and Immunology, University of Miami, Miami, FL, USA

Advances in genetic analysis and a greater understanding structs and transfected into a CD4+ T cell line that was there- of human immunodeficiency virus (HIV) molecular pathogen- after infected with HIV-1 MN. CD4+ T cells treated with the esis have identified critical viral targets for gene interference active U5 EGS (560) were observed to maintain CD4+ strategies. RNase P molecules have been proposed as a expression and did not produce HIV p24 gag antigen, form novel approach for gene targeting based upon their potent syncytia or undergo apoptosis up to 30 days after infection. catalytic activity, as well as versatile external guide Identical cells expressing the inactivated form of the U5 sequence (EGS) which can be modified to specifically recog- RNase P EGS completely down-regulated CD4 expression, nize almost any target mRNA. We designed a truncated produced elevated levels of HIV-1, formed large syncytia EGS to specifically recognize the highly conserved U5 and underwent apoptosis similar to untreated cells. HIV-1 region of HIV-1 mRNA and mediate subsequent cleavage of replication and related cytopathology can be effectively hybridized mRNA by the RNase P enzyme component. The inhibited in CD4+ T cells expressing a protective U5 EGS active U5-EGS (560), as well as a disabled U5 EGS (560D) (560). Gene Therapy (2001) 8, 1863–1871. control, were cloned into plasmids containing proviral con-

Keywords: RNase P; external guide sequence; HIV; gene therapy

Introduction are thought to be mediated by RNase H which cleaves the target mRNA strand of an RNA-DNA hybrid.6,7 The limitations of combination antiviral drug therapies Cleavage at non-targeted sites is a problem associated for human immunodeficiency virus (HIV-1) have lead to with antisense technologies as RNase H does not require the design and development of alternative approaches to a 100% complementary duplex to degrade hybridized 1 therapeutic intervention. Indeed, the high mutation rate mRNA.8–10 Furthermore, antisense molecules act in a 1:1 of HIV generates new variants of the virus which are stoichiometric relationship with target mRNA and pos- resistant to conventional therapies. In addition to viral sess no inherent ‘catalytic turnover’, thereby requiring resistance, current treatments for HIV infection are hin- continuous administration of these molecules to elicit a dered by their cost, availability, patient noncompliance sustained therapeutic effect in treated cells.5 ‘Hairpin’ or 2 and toxicity. As accessible and affordable alternatives, ‘hammerhead’ ribozymes may be able to overcome the strategies involving gene therapy have been proposed to deficiencies encountered with antisense technologies, as restore immune function in infected patients using a var- these RNA enzymes possess autocatalytic activity that iety of effector molecules to inhibit HIV replication in + permits the repeated recognition and cleavage of mul- CD4 cells, reduce the cytopathic effects of viral infection tiple target mRNA transcripts within an infected cell.11 3 and/or induce death of infected cells. Exploration of As with antisense technologies, these catalytic RNA mol- mechanisms for regulation of HIV replication have led to ecules can be designed to specifically recognize target the design and evaluation of a large number of genetic mRNA sequences from foreign or aberrant genes, but are antiviral strategies including the expression of limited by the requirement for the presence of specific dominant/negative mutant proteins, RNA decoys, intra- nucleotide sequences (-GUX-) in the target mRNA for 1 cellular antibodies and gene interference therapies. cleavage to occur.12 Furthermore, a single point mutation The most common approaches to gene interference for in the required GUX sequence could render ribozymes the treatment of HIV infection have mainly involved anti- ineffective for target mRNA cleavage.5 With technologi- 3 sense technologies or ribozymes. However, therapeutic cal advances in genetic analysis and a greater under- approaches based on these technologies have encoun- standing of HIV molecular pathogenesis that have per- 4,5 tered limitations. Inhibitory effects of antisense mol- mitted the identification of specific gene targets, it is ecules, either oligodeoxynucleotides or antisense RNA, imperative to develop gene interference strategies that combine the attributes of antisense and ribozyme technologies and overcome their limitations. Correspondence: HJ Hnatyszyn, Department of Microbiology and Immu- nology, University of Miami (FL), 1550 NW 10th Avenue (R-138), One promising gene interference technology has been Miami, Florida 33136, USA investigated and characterized for over 20 years follow- Received 19 April 2001; accepted 17 September 2001 ing its discovery by Sidney Altman in the late 1970s.13 RNase P-mediated inhibition of HIV-1 pathogenesis HJ Hnatyszyn et al 1864 RNase P, a ribonucleoprotein that consists of enzyme and tective 5’cap. A disabled EGS devoid of the T-stem and RNA subunits, catalyzes the hydrolysis reaction that loop required for RNase P enzyme recognition and cleav- removes the 5’ leader sequence from tRNA precursors to age was constructed to control for antisense inhibition form mature tRNA molecules.14,15 This essential enzy- (Figure 1b). For our experiments, the EGS is complemen- matic activity has been found in all cell types, including tary to 10 base pairs from the HIV-1 U5 region. This prokaryotes and eukaryotes.16 A general strategy for gene length of EGS should be suitable for target specificity targeting has been developed based upon the fact that the without creating aberrant secondary structure that may 3’ proximal sequence of the RNA component of RNase P, interfere with target mRNA hybridization or RNase P- called the external guide sequence (EGS), can be modified mediated cleavage. The genes encoding for the active and to hybridize with complementary target mRNA.4,17,18 The disabled U5 EGS was cloned into Moloney-based retrovi- EGS is designed to leave a 5’-ACCAC-3’ unpaired stretch ral vectors for delivery to target cells. The following required for cleavage by the RNase P protein component investigation describes the potent inhibition of HIV-1 following hybridization. In principle, any mRNA of replication and viral pathogenesis in CD4+ T cells known sequence can be targeted for precise cleavage by expressing this novel RNase P molecule. RNase P through use of custom-designed EGS RNA without the restriction of a specific nucleotide sequence in the target mRNA.18,19 Subsequent studies of substrate Results recognition and RNase P cleavage requirements have provided insight into structural modifications of the RNA Retroviral gene expression component of RNase P that may be implemented to The CD4+ T cell line, M4C8, was transfected with retrovi- enhance target mRNA cleavage.20 As such, RNase P mol- ral expression plasmids containing single copies of the ecules combine the attributes of antisense technologies, genes for either the U5 EGS (560) or the disabled U5 EGS as well as ribozymes and transcend several of their (560D) driven by the tRNAVAL promoter. The disabled limitations. U5 EGS (560D) lacked the T-stem loop required for recog- We have designed a truncated EGS that specifically nition and cleavage of target mRNA by the RNase P recognizes and hybridizes to the U5 region of the 5’ enzyme, but shared the identical EGS with the active U5 leader sequence of HIV-1 (Figure 1a). This region is EGS (560) to control for antisense effect. Following selec- highly conserved across HIV-1 clades and is present on tion of transfected cells for neomycin resistance, all forms of HIV RNA including incoming RNA strands untreated M4C8 cells and cells expressing either the and mRNA transcripts expressed following infection.21–24 active or disabled U5 EGS molecule were evaluated for Successful cleavage at this region of the viral RNA results the presence and expression of the neomycin resistance in degradation of the genome due to removal of the pro- (NeoR) gene, as well as each U5 EGS (560 and 560D) using real-time PCR (Table 1). These assays were used to con- firm transfection of the retroviral plasmids, as well as expression of the genes carried by these constructs and eliminate target cell clones that generated spontaneous resistance to G418. Theoretically, Northern blot analysis could be used to detect EGS expression from the tRNA promoter. However, the size of the EGS (>35 base pairs) makes detection difficult and quantitation impractical. All cells were removed from G418 selection 7 days before RNA isolation as subsequent HIV infection and analysis would be completed on cells no longer cultured in the presence of G418. cDNA was generated from isolated RNA samples and evaluated for NeoR and U5 EGS (560 and 560D) expression, as well as expression of a common ‘housekeeping’ gene, Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) using real-time RT-PCR. The crossing points, defined as the cycle where fluorescent signal from amplified products enters the linear phase, were used in a ratio as a comparative measure of NeoR expression standardized by the level of GAPDH for each experimental and untreated cell population. Expression of the NeoR gene was confirmed for all cultures of experi- mental and untreated M4C8 cells and there was no sig- nificant difference in levels of gene expression (Table 1). Furthermore, the NeoR mRNA transcript could be detected in these cell populations at greater than 60 days after selection and removal of G418. Expression of each Figure 1 Diagrams of the truncated U5 EGS (560) and (560D). (a) The RNase P EGS was confirmed for each corresponding active form of the U5 EGS (560) including the T-stem and loop, as well M4C8 population following selection using real-time as the EGS of the acceptor stem that is complementary to the U5 region PCR. Each of the U5 EGS were expressed at similar levels of HIV-1 mRNA. (b) The disabled form U5 EGS (560D). The T-stem and loop required for RNase P enzyme recognition and binding have been when compared with levels of the ‘housekeeping’ control R deleted. The EGS of the acceptor stem is identical to the active form to gene, GAPDH (Table 1). Thus, expression of the Neo control for antisense inhibition of HIV-1 mRNA. gene and each U5 EGS expressed by the retroviral plas-

Gene Therapy RNase P-mediated inhibition of HIV-1 pathogenesis HJ Hnatyszyn et al 1865 Table 1 Detection of vector gene expression for experimental and control M4C8 cells

Cell line GAPDH (cycle) NeoR (cycle) Ratio U5 EGS 560 Ratio U5 EGS 560D Ratio (cycle) (cycle)

M4C8 16.94 0 0 0 0 0 0 M4C8 + U5 EGS 560 15.93 19.66 1.234 18.30 1.148 0 0 M4C8 + U5 EGS 560D 15.94 21.90 1.374 0 0 20.62 1.293

M4C8, Molt 4 Clone 8 CD4+ T cell line; U5 EGS (560), active RNase P; M4C8+ U5 EGS (560D), disabled RNase P; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; NeoR, neomycin resistance gene. Ratios are calculated by dividing the NeoR, U5 EGS (560) or U5 EGS (560D) crossing points (cycle) with the corresponding GAPDH crossing point determined using real-time RT-PCR. mids was confirmed for M4C8 cells transfected with no detectable toxicity associated with the active or either the disabled or active U5 EGS used for this disabled U5 EGS molecules. investigation. Inhibition of HIV-1 production Target cell proliferation following treatment Triplicate cultures of experimental and untreated M4C8 Experimental (active or disabled U5 EGS) and untreated cells were exposed to HIV-1 MN and evaluated for p24 M4C8 cells were subsequently evaluated for cell prolifer- gag antigen production, HIV-1 proviral DNA, levels of ation using a commercial colorimetric assay based upon CD4 expression and the percentage of apoptotic cells the conversion of tetrazolium salts to formazan dye by over time. Cell and supernatant samples were harvested viable, metabolically active cells. Cells were removed in triplicate from each culture at regular intervals for a from G418-supplemented media for a period of 7 days period of 32 days. As an indication of productive HIV before evaluated of cell viability and proliferation. infection, levels of p24 antigen were detectable in the Experimental and control cells were seeded in quadrupli- supernatants of untreated M4C8 and experimental cells cate at 103 or 104 cells per well in a 96-well plate and expressing the disabled U5 EGS (560D) molecule at 7 evaluated for formazan dye conversion after 5 days days after infection (Figure 3a). The level of p24 antigen (Figure 2). Levels of cell proliferation for non-transfected in the supernatant of these cultures continued to increase and M4C8 cells expressing the U5 EGS (560) were not until reaching the maximum threshold of the assay significantly different from one another. M4C8 cells between 10 to 14 days after infection. Levels of p24 anti- expressing the disabled U5 EGS (560D) were slightly gen remained elevated in these control cultures to 32 more active than cells expressing the U5 EGS (560) or days. In complete contrast, no p24 antigen could be untreated controls. This small difference in cell growth is detected in the supernatants from cultures of M4C8 cells consistent for each M4C8 culture expressing the inactive expressing the active U5 EGS (560) and exposed to HIV- U5 EGS (560D) and is not attributed to transfection or 1 MN over the same period of time. Furthermore, components of the retroviral plasmid. Nevertheless, all untreated M4C8 and experimental M4C8 expressing M4C8 cells utilized for this investigation had similar lev- either the active or disabled U5 EGS were permissive to els of viability, metabolism and proliferation indicating HIV-2 KR infection as indicated by elevated p26 antigen levels in culture supernatants and the rapid formation of syncytia in infected cultures (Figure 3b). This observation would indicate that the selected cells are permissive to HIV-2 infection but the active U5 EGS (560) specifically inhibits HIV-1 infection and replication. Collectively, these data suggest that M4C8 cells expressing the active U5 EGS (560) were either resistant to HIV-1 infection due to cleavage of incoming virus or did not produce HIV-1 particles as viral mRNA is cleaved following transcription.

Inhibition of proviral DNA integration Genomic DNA was isolated from cell samples and evalu- ated for the presence of HIV-1 proviral DNA using PCR (data not shown). A DNA fragment corresponding to the conserved region of the HIV-1 gag protein was readily amplified in untreated M4C8 and M4C8 expressing the disabled U5 EGS (560D). However, no proviral DNA

+ could be amplified from genomic DNA isolated from Figure 2 Cell proliferation assays for experimental and untreated CD4 M4C8 expressing the active U5 EGS (560) throughout the M4C8 cells as an indication of toxicity and changes in cellular activity. 32 day period after exposure to HIV-1. These results cor- Two concentrations of cells (103 and 104) were evaluated for both untreated and cells expressing either the active or disabled U5 EGS. The responded with the contrasting levels of p24 antigen levels of formazan dye converted from tetrazolium salts by viable cells measured for each culture supernatant and indicated that were measured at a wavelength of 560 nm for each sample. the active U5 EGS (560) may interfere with the HIV cycle

Gene Therapy RNase P-mediated inhibition of HIV-1 pathogenesis HJ Hnatyszyn et al 1866

Figure 4 Percentage of CD4+ cells in HIV-1 MN infected cultures of untreated and experimental M4C8 cells. Cell samples were harvested over a period of 32 days and analyzed for CD4 expression using a phyco- erythrin labeled anti-human CD4 monoclonal and flow cytometry.

M4C8 and cells expressing the disabled U5 EGS (560D) had very low numbers of cells expressing CD4 (2–10%) which was observed to 32 days after infection. However, M4C8 cells treated with the active U5 EGS (560) had no decline in population of cells expressing CD4 (92–98%) over the same period. This observation suggests that M4C8 cells expressing the active U5 EGS (560) did not undergo loss of CD4 expression observed in control cells infected with HIV-1 MN.

Inhibition of apoptosis by active U5 EGS (560) Cell cultures were also evaluated for the number of cells undergoing apoptosis using Annexin V stain and flow Figure 3 Levels of HIV-1 p24 gag antigen and HIV-2 p26 gag antigen cytometry over a period of 30 days (Figure 5). Before in experimental and untreated M4C8 cell cultures as a measure of viral infection with HIV-1 MN, cultures of non-transfected, infection. 105 cells for each cell population were incubated with either experimental and control M4C8 had low levels of cells HIV-1 MN or HIV-2 KR at an MOI of 0.01. Supernatants were evaluated undergoing apoptosis (2.66–5.12%). At 7 days after infec- for levels of gag antigen as a measure of HIV infection and viral pro- duction. (a) Levels of HIV-1 MN p24 gag antigen. ᮀ, M4C8 #1; ᭿, M4C8 #2; ̅, M4C8 + U5 EGS (560)#1; ̆, M4C8 + U5 EGS (560)#2; ᭺, M4C8 + U5 EGS (560D)#1; ¼, M4C8 + U5 EGS (560D)#2. (b) Levels of HIV-2 KR p26 gag antigen. HIV-2 KR has a rapid pathogenesis in M4C8 cells. By 14 days after infection with HIV-2 KR, most cells in infected cultures were dead. ᮀ, M4C8 #1 non-infected; ᭿, M4C8 #1; ̆, M4C8 + U5 EGS (560)#1; ¼, M4C8 + U5 EGS (560D)#1.

before integration of proviral DNA into the host cell genome.

Active U5 EGS (560) preserves CD4 expression The levels of p24 antigen in the culture supernatant also correlated well with the loss of CD4 from the surface of untreated M4C8 and cells expressing the disabled U5 EGS (560D) (Figure 4). Before exposure to HIV-1 MN, greater than 95% of all cells, including experimental and control cultures, were observed to express CD4 on their surface as detected by flow cytometry. By 11 days after Figure 5 Levels of apoptotic cells in cultures of untreated and experi- infection, there was little or no decrease in CD4 levels for mental M4C8 cells exposed to HIV-1 MN. Samples were harvested at untreated M4C8 cells or those expressing the active U5 regular intervals and evaluated for apoptosis using Annexin V staining for phosphatidyl serine and flow cytometry. Samples harvested past day EGS (560). However, only 55–60% of the M4C8 treated 14 indicated the generation of CD4Ϫ, HIV-1 producing M4C8 clones in with the disabled U5 EGS (560D) expressed CD4 on their untreated M4C8 cells and cells treated with the disabled U5 EGS (560) surface. At 21 days after infection, both the untreated that were resistant to virus-mediated death.

Gene Therapy RNase P-mediated inhibition of HIV-1 pathogenesis HJ Hnatyszyn et al 1867 tion with HIV-1, the percentage of cells undergoing or CD4+ cells expressing the active U5 EGS (560) have a sur- completing apoptosis was elevated in untreated M4C8 vival advantage over untreated M4C8 or cells expressing cells (17%) and more so in M4C8 cells expressing the dis- a disabled U5 RNase P EGS following exposure to abled U5 EGS (560D) (27.63%). However, the percentage HIV-1 MN. of apoptotic cells remained low (5.12%) for M4C8 treated with the active U5 EGS (560). At 11 and 14 days after Discussion infection with HIV-1 MN, levels of apoptotic cells were elevated in untreated M4C8 (7.63–12.9%) and cells The RNase P complex may offer an excellent alternative expressing the disabled U5 EGS (560D) (23–29.22%). Over to conventional gene interference therapies for the treat- the same time period, the levels of apoptotic cells ment of infectious diseases and human malignancies. The exposed to HIV-1 MN and expressing the active U5 EGS RNA component contains the external guide sequence (560) remained similar to background levels for unin- (EGS) which can be designed to recognize and hybridize fected control cultures (2.3–4.8%). Cell samples harvested to almost any target mRNA. The enzyme component of after day 14 indicated the generation of CD4Ϫ, HIV-1-pro- RNase P required for cleavage of the RNA hybrid is ducing M4C8 clones in untreated M4C8 cells and cells ubiquitously expressed in all cells that express tRNA and treated with the disabled U5 EGS (560) that were resistant translate proteins. The result is an effective and versatile to virus-mediated death. gene interference system that is independent of the target These observations were corroborated with light mRNA sequence and can be applied in every cell type. microscopic evaluation of the control and HIV infected In this investigation, we designed a truncated RNase cultures (Figure 6). M4C8 expressing the disabled U5 P EGS that targeted the highly conserved U5 region of EGS (560D) and exposed to HIV-1 MN developed large the HIV-1 genome present in incoming viral genomes, and numerous syncytia by day 7. Syncytia remained as well as viral mRNA transcripts produced following present to day 14 after which viable HIV-infected, CD4Ϫ infection. A CD4+ T cell line expressing the U5 EGS (560) clones began to develop. Similar, but less numerous, syn- was protected from HIV-1 infection with no loss of CD4 cytia were observed in untreated M4C8 cultures infected expression, no detectable integrated proviral DNA, no with HIV-1 at 7 days after infection. Once again these measurable viral replication and no observable cytopath- giant multi-nucleated cells were observed to day 14 when ology. In contrast, untreated M4C8 cells and cells CD4Ϫ, HIV-infected isolates gained prominence. Con- expressing a disabled U5 EGS (560D) antisense control versely, cultures of M4C8 cells expressing the active U5 molecule were permissive to infection with HIV-1 MN EGS (560) as well as non-infected controls for each cell and presented with the key characteristics of HIV cyto- type, remained free of syncytia formation and CD4+ to pathology. These data indicate that inhibition of HIV-1 30 days after infection. These observations suggest that infection or replication was not due to an antisense effect as cells expressing the disabled U5 EGS (560D), with the identical hybridization region but lacking the T-stem for enzyme recognition, elicited no such results. Further- more, this inhibition of viral infection observed with CD4+ T cells treated with the active U5 EGS (560) was specific for HIV-1, as these experimental cells as well as controls were readily infected with HIV-2. Since the tar- get sequence for U5 EGS (560) is not present in HIV-2, this specificity is not surprising. However, this obser- vation further suggests that the inhibition observed in cells expressing the active RNase P EGS is not due to the cells being rendered non-permissive to HIV, but rather specifically protected from HIV-1 strains with genomes containing the target sequence. Theoretically, the inhi- bition of HIV-1 infection and/or replication should extend to most strains of HIV-1 as the target sequence is highly conserved across viral clades. The point in the viral replication cycle where RNase P- mediated inhibition of HIV-1 infection occurs is unclear. One possible scenario would suggest that the U5 EGS (560) could inhibit viral infection following viral entry into the cytoplasm and before generation of proviral DNA. In the current investigation, the retroviral plasmids transfected into the M4C8 cells could potentially produce Figure 6 Morphology of untreated and experimental M4C8 cells follow- two different transcripts of the U5 EGS. One would be R ing exposure to HIV-1 MN. Cells were assessed using Varicel phase con- fused with the neomycin resistance (Neo ) mRNA tran- trast microscopy for detailed imaging of cell surfaces. Each scale bar rep- script driven by the Moloney LTR promoter and would resents 50 ␮m. (a) Untreated, uninfected M4C8 cells for comparison; (b) subsequently be exported to the cytoplasm for translation untreated M4C8 exposed to HIV-1 MN at 14 days after infection; (c) of the selectable marker. This fusion NeoR-U5 EGS (560) M4C8 expressing the active U5 EGS (560) and exposed to HIV-1 MN could bind to the incoming viral RNA and inhibit reverse at 30 days after infection; (d) M4C8 expressing the disabled U5 EGS (560D) and exposed to HIV-1 MN at 14 days after infection. Arrows transcription and proviral integration. In this instance, indicate large cell syncytia formed by the fusion of HIV-1-infected cells inhibition would not be mediated by RNase P cleavage, with adjacent CD4+ cells. as the enzyme component of the RNase P complex is

Gene Therapy RNase P-mediated inhibition of HIV-1 pathogenesis HJ Hnatyszyn et al 1868 localized in the nucleolus and nucleoplasm of the thereby preventing the down-regulation of CD4 from the nucleus.25–27 Rather the inhibition would be attributed to surface of cells treated with the active U5 EGS (560). Fur- antisense activity and RNase degradation of RNA thermore, RNase P-mediated inhibition of viral transcrip- hybrids. However, this mechanism is unlikely as similar tion would preclude translation of structural proteins, inhibition of viral infection would have been expected preventing detection of the p24 gag antigen and with the disabled U5 EGS (560D), which shared the same expression of HIV gp160 involved in the formation of cell EGS as the active form but no site for binding the RNase syncytia. Finally, inhibition of viral replication at this P enzyme. Thus, it is unlikely that the inhibition of viral stage of the viral cycle would prevent expression of viral replication is due to the antisense effect of the U5 EGS gene products that promote apoptosis of infected cells. fused to the NeoR transcript. These last two scenarios would explain the observations Inhibition of incoming virus may still be a viable possi- and data obtained from CD4+ T cells expressing the active bility when the pre-integration complex (PIC) of the lenti- U5 EGS (560) and lack thereof for cells expressing the virus enters the nucleus of the infected cell, where both inactive U5 RNase P EGS. However, the inability to the RNase P enzyme and the second form of the active detect HIV-1 proviral DNA in cells expressing the active U5 EGS (560) transcript coexist.28,29 Replication of HIV-1 U5 EGS (560) lends support to the inhibition of incoming in non-dividing and slowly proliferating cell populations viral RNA in the PIC rather than post-reverse transcrip- is dependent upon the active import of the viral PIC into tion and integration. Further investigation involving PCR the nucleus.30 Formation of the PIC is required for reverse techniques for proviral DNA and mRNA transcripts, as transcription of viral genomic RNA to proviral DNA. well as immunohistochemistry for viral protein pro- When the PIC initially arrives at the perinuclear region duction may provide insight into the precise point where and the nucleus, the viral genomic RNA is present along U5 EGS (560) interrupts the HIV-1 cycle of infection. with incomplete proviral DNA molecules. At this point, The use of targeted RNase P EGS molecules as specific the viral RNA is susceptible to EGS binding and RNase therapeutic and research tools is a new approach to gene P cleavage. Expression of the second U5 EGS (560) is interference strategies for infectious diseases and human driven by the tRNA promoter and produces a single U5 malignancies. Carefully designed RNase P EGS mol- EGS (560) molecule. This molecule should remain ecules offer the advantages attributed to antisense and localized in the nuclear region of cells expressing the EGS ribozyme technologies without the disadvantages of where the protein component of RNase P is in abun- inappropriate cleavage or the requirement for specific dance. When the PIC including the HIV genomic RNA is sequences within the target mRNA. As a potential ther- localized in this region, it is possible that the U5 EGS apy for HIV infection, these RNase P molecules could be (560) could hybridize with its target sequence followed used to treat non-infected, autologous stem cell grafts, by cleavage by the RNase P enzyme. Such a mechanism thereby protecting these altered cells from infection fol- would prevent production of proviral DNA and sub- lowing transplantation to reconstitute the immune sys- sequent integration into the host cell genome thereby tem of HIV patients.31 Future directions could include the interrupting the viral cycle and preventing infection. This production of synthetic RNase P EGS molecules that mechanism could account for the results obtained using could be administered much like pharmaceuticals or in real-time PCR to detect HIV-1 proviral DNA in cellular combination with targeted drug delivery systems,32 as an genomic DNA isolated from experimental and control effective intervention for HIV infection. cells exposed to HIV-1. HIV-1 proviral DNA was readily amplified and detected in genomic DNA from untreated M4C8 and cells expressing the inactive U5 EGS (560D) Materials and methods infected with HIV-1 MN throughout the investigation. However, no proviral DNA could be detected in M4C8 Cell culture and viral stocks + cells expressing the active U5 EGS (560) over the same The MOLT-4 clone 8 (M4C8) cell line is a CD4 T cell line time periods, suggesting reverse transcription of viral permissive to both HIV-1 and HIV-2.33 M4C8 were the RNA was not completed and integration was not achi- kind gift of M Hayami (Institute for Virus Research, eved. These observations, along with the other measures Kyoto University, Kyoto, Japan). HIV-1 MN was pro- of viral infection and production in cells treated with duced in our laboratory using a constitutively expressing 4 active U5 EGS (560) and antisense control molecules, lend T cell line and titered at 10 TCID50. HIV-2 KR used for support to this scenario of RNase P-mediated inhibition RNase P EGS specificity and permissive control experi- of viral RNA as the PIC arrives at the nucleus. This stage ments was also produced in our laboratory using of viral infection would serve as the most effective time M4C8 cells. for an active EGS to mediate RNase P inhibition, as the number of viral RNA targets to be eliminated and pre- RNase P EGS design and vector construction vent infection is limited to incoming viral RNA genomes. The Moloney murine leukemia virus (MoMLV) vectors One final scenario could suggest that proviral inte- pLNL-6-560 and pLNL-6-560D were engineered by gration is achieved in cells expressing the active U5 EGS inserting a t-RNAval promoter RNase P cassette into the (560), but is not detected using PCR and cellular DNA unique NheI site located in the U3 region of the MoMLV samples. This would imply that RNase P-mediated inhi- 3’LTR. The 560 RNase P molecule was created by bition would have to occur at the level of viral genome annealing of the upper-strand primer 560US (5’CGGGAT transcription. This scenario is attractive, as all of the CCACATCAAGGTTCGTGCCCTTCCAGACGGGCAAC components for recognition, hybridization and sub- CATACGCGTCG 3’) and lower-strand primer 560LS (5’C sequent cleavage of target mRNA would be localized in GACGCGTATGGTTGCCCGTCTGGAAGGGCACGAA 3’). the nuclear region of infected cells. Transcription of early The annealed product was subject to PCR amplification genes, such as nef, tat and rev, would not be completed (94°,48° and 72°C for 10 s at each temperature). The frag-

Gene Therapy RNase P-mediated inhibition of HIV-1 pathogenesis HJ Hnatyszyn et al 1869 ment was digested by BamHI and MluI and ligated into 3GAP primers with the DNA amplification SYBR Green the modified pLNL-6 t-RNA containing vector. The same kit (Roche Diagnostics) and real-time PCR was carried procedure was used to construct the pLNL-6-560D vector out (30 cycles at 95°C 1 s, 60°C 5 s, 72°C for 10 s). Melting using the following upper-strand primer 560D US (5’AG point analysis using the LightCycler confirmed the gener- GATCCGGCCAAACATCCCAGACGGGCACGCGTCG 3’) ation of a single product corresponding to the GAPDH and lower-strand primer 560DLS (5’CGACGCGTGCCCG fragment. TCTGGGATGTTTGGCCGGATCCT 3’). The integrity of To confirm selected cell populations were transfected the new constructs were confirmed by DNA sequencing. with each retroviral expression plasmid, real-time PCR analysis was completed using primers specific for the Transfection and selection of M4C8 cells neomycin resistance (NeoR) selectable marker and NeoR M4C8 cells were seeded at 105 cells per ml of media into plasmid DNA standards. Again 2 ␮l of cDNA was ana- uncoated T25 suspension flasks. Cells were transfected lyzed with the DNA amplification SYBR Green kit using with the U5 EGS (560) or (560D) control plasmid (2 ␮g) the 5neo (5’ GATCAAGAGACAG GATGAGG 3’) and using Fugene (Roche Biochemicals, Indianapolis, IN, 3neo (5’ CAGCCATGATGGATACTTTC 3’) primers. The USA) according to the manufacturer’s specifications. real-time PCR conditions were one cycle at 95°C for 30 Transfected cells were selected for neomycin resistance s, followed by 30 cycles at 95°C 1 s, 60°C 5 s, 72°C for 20 by culturing cells in the presence of G418 (500 ␮g/ml of s. Melting point analysis using the LightCycler confirmed media). Selected cells were maintained in RPMI sup- the generation of a single product corresponding to the plemented with 10% FCS, 1% penicillin-streptomycin sol- NeoR fragment (melting point = 93.49°C). ution and G418. Seven days before infection with HIV-1 In addition, DNA isolated from transfected and non- or HIV-2, the selection medium was removed from the transfected cell populations were analyzed for the pres- transfected cells and replaced with normal culture ence of the NeoR gene to provide insight into the number medium without G418. of plasmid copies per cell. DNA samples from 100 000 cells were initially analyzed using real-time PCR and pri- Cell proliferation assays mers specific for the B-7 gene to ensure equal amounts Colorimetric assays for the non-radioactive quantification of DNA were used for each reaction. Using real-time PCR of cellular proliferation, viability and cytotoxicity were and serial dilutions of a NeoR containing plasmid (2.6 × completed on selected M4C8 cells (104/well) in the pres- 105 to 2.6 × 109 copies), DNA samples were analyzed for ence and absence of G418 to ensure all cells used for this the presence of the NeoR gene using PCR conditions investigation had similar proliferation rates. The Cell Pro- described above. The average number of plasmid copies liferation Kit I (Roche Biochemicals) was used according per cell was calculated based upon DNA extraction, cell to the manufacturer’s specifications. Levels of the forma- number and quantitative PCR. The average amount of zan dye generated by cells in this assay were measured DNA extracted from 100 000 M4C8 cells (total extractions (570–630 nm) using a scanning multiwell spectrophoto- = 10) was calculated to be 0.6 pg/cell. Since the amount meter (PerkinElmer Wallac, Gaithersburg, MD, USA) and of DNA for each PCR reaction was known, we could directly correlate to the proliferation of viable cells. approximate the number of cells per PCR reaction and subsequently calculate the number of plasmid copies per Confirmation of vector expression using real-time RT- cell based upon the quantitative PCR results and the plas- PCR mid standards. The number of copies of retroviral Total RNA was isolated from M4C8, M4C8 + U5 EGS 560 plasmid per cell was calculated to be approximately 6.85 and M4C8 + U5 EGS 560D cells using a High Pure RNA copies per cell for M4C8 + U5 EGS 560 and 11.85 copies Isolation kit (Roche Diagnostics). 300 ng of each RNA per cell for M4C8 + U5 EGS 560D. sample, as well as serial dilutions of in vitro transcribed Finally, levels of expression for each RNase P EGS (U5 GAPDH RNA were analyzed using real-time RT-PCR EGS 560 and 560D) was measured using real-time PCR and the LightCycler. Following the instructions of the analysis with control plasmid DNA. Again 2 ␮l of cDNA RNA amplification SYBR Green kit (Roche Diagnostics) was analyzed with the DNA amplification SYBR Green and adding the 5GAP (5’GAAGGTGAAGGTCGGAGTC kit using the LNL6-2876 (5’AGGTCACTGCGTGGATGG 3’) and 3GAP (5’GAAGATGGTGATGGGATTTC 3’) pri- 3’) and either MN560LS or primers corresponding to each mers with 300 ng of total RNA, real-time PCR was perfor- EGS. The real-time PCR conditions were one cycle at med (one cycle at 52°C, for 10 min; 30 cycles at 95°C1s, 95°C for 30 s, followed by 30 cycles at 95°C 2 s, 51°C5 60°C 5 s, 72°C for 10 s). This reaction provided a quanti- s, 72°C for 45 s. Melting point analysis using the Light- tative measurement of RNA isolated from each cell sam- Cycler confirmed the generation of a single product cor- ple to compare with spectrophotometric quantitation. responding to each U5 EGS fragment (U5 EGS 560 = Levels of GAPDH for each sample as a measure of RNA 93.0°C and U5 EGS 560D = 93.5°C). isolation were similar for each transfected and control cell sample (data not shown). Infection of M4C8 with HIV-1 MN and HIV-2 KR cDNA was generated using 300 ng of each RNA sam- Experimental (expressing the active or disabled U5 EGS) ple with random hexamers according to the manufac- and control M4C8 cells (105 cells in 500 ␮l of media) were turer’s specifications using the Gene Amp RNA PCR placed in 1.5 ml microcentrifuge tubes. HIV-1 MN (MOI Core Kit (Applied Biosystems, Foster City, CA, USA). = 0.01) was added to each tube and incubated with the Two ␮l of cDNA was analyzed using quantitative PCR cells for 6 h. Following incubation, cells were washed five and the LightCycler for levels of GAPDH, neomycin times with culture media via centrifugation to remove resistance and each RNase P EGS. Levels of GAPDH remaining virus. Cells were resuspended in 7 ml of fresh were analyzed to determine that similar levels of cDNA culture media and primary (time = zero) cell and super- were generated from each RNA sample using 5GAP and natant samples were harvested for analysis. Supernatants

Gene Therapy RNase P-mediated inhibition of HIV-1 pathogenesis HJ Hnatyszyn et al 1870 and cell samples were harvested from each culture every References 7 days for analysis for HIV infection, induction of apoptosis and levels of CD4 expression. In order to assess 1 Wong-Staal F, Buchschacher Jr GL. Gene therapy and the two specificity of the RNase P EGS, duplicate cell samples faces of HIV. Gene Therapy 2000; 7: 1351–1352. were infected with HIV-2 KR (MOI = 0.01) and processed 2 Dornburg R, Pomerantz RJ. HIV-1 gene therapy: promise for the using an identical protocol. future. Adv Pharmacol 2000; 49: 229–261. 3 Lamothe B, Joshi S. Current developments and future prospects for HIV gene therapy using interfering RNA-based strategies. Detection of HIV-1 proviral DNA using real-time PCR Front Biosci 2000; 5: D527–D555. Genomic DNA was isolated from cell samples harvested 4MaMet al. Intracellular mRNA cleavage induced through acti- every 7 days after infection using a commercial kit vation of RNase P by nuclease-resistant external guide (Perfect Blood gDNA Mini, Eppendorf Scientific, sequences. Nat Biotechnol 2000; 18:58–61. Westbury, NY, USA). DNA was quantitated using spec- 5 Cobaleda C, Sanchez-Garcia I. In vivo inhibition by a site-specific trophotometry and analyzed for the presence of HIV-1 catalytic RNA subunit of RNase P designed against the BCR- proviral DNA as an indication of infection using real- ABL oncogenic products: a novel approach for cancer treatment. Blood 2000; 95: 731–737. time PCR and the LightCycler. PCR mixtures were pre- 6 Minshull J, Hunt T. The use of single-stranded DNA and RNase pared according to the DNA Amplification SYBR Green H to promote quantitative ‘hybrid arrest of translation’ of I Kit (Roche Biochemicals) using the SK38 (5’ mRNA/DNA hybrids in reticulocyte lysate cell-free trans- ATAATCCACCTATCCCAGTAGGAGAAAT 3’) and lations. Nucleic Acids Res 1986; 14: 6433–6451. SK39 (5’ TTTGGTCCTTGTCTTATGTCCAGAATGC 3’) 7 Walder RY, Walder JA. Role of RNase H in hybrid-arrested oligonucleotides specific for the viral sequence encoding translation by antisense oligonucleotides. Proc Natl Acad Sci USA a region of HIV-1 gag protein. The LightCycler was util- 1988; 85: 5011–5015. ized for PCR amplification and detection of HIV-1 pro- 8 Giles RV, Tidd DM. Increased specificity for antisense oligo- viral DNA as a measure of viral infection of experimental deoxynucleotide targeting of RNA cleavage by RNase H using and untreated CD4+ T cells over time. The real-time PCR chimeric methylphosphonodiester/phosphodiester structures. Nucleic Acids Res 1992; 20: 763–770. conditions were 30 cycles at 95°C for 1 s, 60°C for 5 s, ° 9 Giles RV, Spiller DG, Tidd DM. Chimeric oligodeoxynucleotide 72 C for 20 s. Melting curve profiles were completed for analogues: enhanced cell uptake of structures which direct ribo- each amplified product over a temperature range of 65 nuclease H with high specificity. Anticancer Drug Des 1993; 8: to 88°C. 33–51. 10 Giles RV et al. Single base discrimination for ribonuclease H- P24/P26 antigen capture assays dependent antisense effects within intact human leukaemia cells. Nucleic Acids Res 1995; 23: 954–961. HIV-1 MN replication and production was monitored by 11 Kruger K et al. Self-splicing RNA: autoexcision and autocycliz- determination of the p24 gag antigen concentration ation of the ribosomal RNA intervening sequence of tetra- (Coulter p24 Antigen Assay, Beckman-Coulter, Miami, hymena. Cell 1982; 31: 147–157. FL, USA) in the culture supernatants harvested at various 12 Cech TR, Bass BL. Biological catalysis by RNA. Annu Rev time-points. Productive HIV-2 KR infection was moni- Biochem 1986; 55: 599–629. tored by determination of the p26 gag antigen concen- 13 Altman S. The road to RNase P. Nat Struct Biol 2000; 7: 827–828. tration (Coulter p26 SIV Antigen Assay) in the culture 14 Altman S, Kirsebom L, Talbot S. Recent studies of ribonuclease supernatants. P. FASEB J 1993; 7:7–14. 15 Altman S et al. Catalysis by the RNA subunit of RNase P – a minireview. Gene 1989; 82:63–64. CD4 and Annexin V detection using flow cytometry 16 Altman S, Wesolowski D, Puranam RS. Nucleotide sequences Non-infected and infected cell samples were evaluated in of the RNA subunit of RNase P from several mammals. Geno- triplicate for CD4 expression and levels of phosphatidyl- mics 1993; 18: 418–422. serine as a measure of apoptosis using flow cytometry. 17 Altman S. RNA enzyme-directed gene therapy. Proc Natl Acad Cells were fixed with 10% buffered formalin before stain- Sci USA 1993; 90: 10898–10900. ing and analysis. Anti-human CD4 monoclonal antibody 18 Yuan Y, Hwang ES, Altman S. Targeted cleavage of mRNA by human RNase P. Proc Natl Acad Sci USA 1992; 89: 8006–8010. (OKT4; Beckman-Coulter) labeled with phycoerythrin 19 Yuan Y, Altman S. Substrate recognition by human RNase P: (PE) was used to detect CD4 expression by non-infected identification of small, model substrates for the enzyme. EMBO and HIV infected M4C8 cells and controls. An appropri- J 1995; 14: 159–168. ate nonspecific murine immunoglobulin control (Mouse 20 Ma MY et al. Nuclease-resistant external guide sequence- IgG2b; Myelin BP, BD Biosciences, San Diego, CA, USA) induced cleavage of target RNA by human ribonuclease P. Anti- was included with each M4C8 cell sample to monitor sense Nucleic Acid Drug Dev 1998; 8: 415–426. nonspecific staining via the Fc receptor of normal lym- 21 Ojwang JO et al. Inhibition of human immunodeficiency virus phocytes. Nonspecific binding values were less than 2% type 1 expression by a hairpin ribozyme. Proc Natl Acad Sci USA for all samples analyzed. An apoptosis detection kit 1992; 89: 10802–10806. (Beckman-Coulter) utilizing FITC-labeled Annexin V, a 22 Yamada O et al. Activity and cleavage site specificity of an anti- calcium and phospholipid-specific protein, to bind to HIV-1 hairpin ribozyme in human T cells. Virology 1994; 205: 121–126. phosphatidyl serine was used to detect apoptosis 23 Yamada O et al. Intracellular immunization of human T cells following infection with HIV. with a hairpin ribozyme against human immunodeficiency virus type 1. Gene Therapy 1994; 1:38–45. 24 Leavitt MC et al. Transfer of an anti-HIV-1 ribozyme gene into Acknowledgements primary human lymphocytes. Hum Gene Ther 1994; 5: 1115–1120. 25 Jacobson MR et al. Nuclear domains of the RNA subunit of We would like to thank Dr Rebecca Geffin for titering the RNase P. J Cell Sci 1997; 110: 829–837. HIV-1MN stocks produced in our laboratory. 26 Bartkiewicz M, Gold H, Altman S. Identification and charac-

Gene Therapy RNase P-mediated inhibition of HIV-1 pathogenesis HJ Hnatyszyn et al 1871 terization of an RNA molecule that copurifies with RNase P matrix protein regulate nuclear import of the HIV-1 pre-inte- activity from HeLa cells. Genes Dev 1989; 3: 488–499. gration complex. J Mol Biol 2000; 299: 359–368. 27 Kato Y et al. Relationships between the activities in vitro and in 31 Gervaix A et al. Gene therapy targeting peripheral blood CD34+ vivo of various kinds of ribozyme and their intracellular localiz- hematopoietic stem cells of HIV-infected individuals. Hum Gene ation in mammalian cells. J Biol Chem 2001; 276: 15378–15385. Ther 1997; 8: 2229–2238. 28 Popov S et al. Viral protein R regulates nuclear import of the 32 Hnatyszyn HJ, Kossovsky N, Gelman A, Sponsler E. Drug deliv- HIV-1 pre-integration complex. EMBO J 1998; 17: 909–917. ery systems for the future. PDA J Pharm Sci Technol 1994; 48: 29 Bukrinsky MI et al. Association of integrase, matrix, and reverse 247–254. transcriptase antigens of human immunodeficiency virus type 33 Kikukawa R et al. Differential susceptibility to the acquired 1 with viral nucleic acids following acute infection. Proc Natl immunodeficiency syndrome retrovirus in cloned cells of Acad Sci USA 1993; 90: 6125–6129. human leukemic T cell line Molt-4. J Virol 1986; 57: 1159–1162. 30 Haffar OK et al. Two nuclear localization signals in the HIV-1

Gene Therapy