|Home | About | Journals | Submit | Contact Us | Français|
Peptide-conjugated phosphorodiamidate morpholino oligomers (P-PMO) are single-stranded nucleic acid-like antisense agents that can reduce gene expression by sterically blocking complementary RNA sequence. P-PMO are water soluble and nuclease resistant, and they readily achieve uptake into cells in culture under standard conditions. Eight P-PMO, each 20 to 22 bases in length, were evaluated for their ability to inhibit influenza A virus (FLUAV) A/PR/8/34 (H1N1) replication in cell culture. The P-PMO were designed to base pair with FLUAV RNA sequences that are highly conserved across viral subtypes and considered critical to the FLUAV biological-cycle, such as gene segment termini and mRNA translation start site regions. Several P-PMO were highly efficacious, each reducing viral titer in a dose-responsive and sequence-specific manner in A/PR/8/34-infected cells. Two P-PMO, one designed to target the AUG translation start site region of PB1 mRNA and the other the 3′-terminal region of nucleoprotein viral genome RNA, also proved to be potent against several other FLUAV strains, including A/WSN/33 (H1N1), A/Memphis/8/88 (H3N2), A/Eq/Miami/63 (H3N8), A/Eq/Prague/56 (H7N7), and the highly pathogenic A/Thailand/1(KAN-1)/04 (H5N1). The P-PMO exhibited minimal cytotoxicity in cell viability assays. High efficacy by two of the P-PMO against multiple FLUAV subtypes suggests that these oligomers represent a broad-spectrum therapeutic approach against a high percentage of known FLUAV strains.
Influenza A virus (FLUAV) causes considerable morbidity and mortality worldwide each year and also poses a pandemic threat (6, 34). FLUAV is an enveloped negative-strand RNA virus, with eight genome segments that code for 11 known proteins (designated PB1, PB1-F2, PB2, PA, HA, NP, NA, M1, M2, NS1, and NS2) (4, 20), and it is classified into subtypes based on the antigenic nature of the surface glycoproteins hemagglutinin (HA) and neuraminidase (NA). Although 16 HA and 9 NA subtypes have been identified, only three FLUAV subtypes (H1N1, H2N2, and H3N2) have been associated with widespread disease in humans. In the past few years, however, several subtypes of avian influenza virus, notably H5N1, H7N7, and H9N2, have been reported to be capable of infecting and, in the case of the H5N1 viruses, causing severe pathology in humans (3, 16).
Although vaccines against matched FLUAV strains can reduce the duration and severity of illness in 60 to 80% of healthy adults, the rate of protection is lower in groups at higher risk of disease, such as the elderly and immunocompromised. Furthermore, vaccination may not provide protection against unexpected strains, such as the H5N1 strains that have caused human disease in Asia between 1997 and 2006. Currently available anti-influenza drugs are small-molecule compounds that act by interfering with essential viral protein functions (42). The usefulness of these drugs is variously limited, however, due to cost, uneven availability, and concerns over emergence of drug-resistant virus strains (14, 18, 23, 25) or side effects (2).
Several reports have described studies in which FLUAV was targeted with large molecules designed to inhibit viral amplification by interacting with viral RNA. DNAzymes (46), oligoaptamers (45), and short interfering RNA (10, 12) have all been documented to have antiviral activity in cell culture, and short interfering RNA has been documented to have antiviral activity in mice (11, 47), against FLUAV. Intravenous delivery in mice of a liposome-encapsulated antisense phosphorothioate oligonucleotide with sequence complementary to the translation start site region of PB2 mRNA reduced FLUAV titers in lung tissue and significantly increased overall survival rates (26, 27). To ultimately achieve clinical utility, any nucleic acid-based anti-FLUAV therapeutic will need to possess a number of favorable pharmacologic qualities, including in vivo stability, low toxicity, and the ability to reach viral RNA targets within relevant cell populations.
Phosphorodiamidate morpholino oligomers (PMO) are structurally similar to single-stranded DNA, in that each subunit includes a purine or pyrimidine base. Each base is joined to a novel backbone consisting of one morpholine ring and phosphorodiamidate linkage per subunit (43, 44). PMO are water soluble, nuclease resistant, and usually 20 to 25 subunits in length. PMO can interfere with gene expression by stably duplexing with complementary RNA through Watson-Crick base pairing, thus forming a steric block (13, 40). The entry of PMO into cells can be markedly increased by conjugation to arginine-rich peptide (ARP) (1, 8, 29). Recently, ARP-conjugated PMO (P-PMO) have been shown to produce antiviral activity against several RNA viruses in cell culture (1, 8, 17, 32, 48) and Ebola virus both in cell culture and in vivo (9). We describe here the evaluation in cell culture of several antisense P-PMO designed to target critical sequence regions in the FLUAV viral genome RNA (vRNA), cRNA, and/or mRNA. In this study, several P-PMO were found to have potent anti-FLUAV A/PR/8/34 (H1N1) activity. Two P-PMO with high efficacy, one targeting the PB1 translation start site region and the other the 3′-terminal region of NP vRNA, were then evaluated against A/WSN/33 (H1N1), A/Memphis/8/88 (H3N2), A/Eq/Miami/63 (H3N8), A/Thailand/1(KAN-1)/04 (H5N1), and A/Eq/Prague/56 (H7N7) and found to generate a greater than 85% reduction in titer of all viral strains in a sequence-specific manner.
PMO were synthesized at AVI BioPharma Inc. (Corvallis, OR) by methods previously described (44). One of the arginine-rich peptides, NH2-R5F2R4C-CONH2 (abbreviated P4 in this report) or AC-NH-(RXR)4XB-COOH (X stands for 6-aminohexanoic acid) (abbreviated P7 in this report), was covalently conjugated to the 5′ end of each PMO used in this study (Fig. (Fig.1A).1A). The conjugation, purification, and analysis of P-PMO were by procedures previously described (29, 31). P-PMO of 20 to 22 subunits in length were designed to target, by complementary base pairing, eight sequence regions in the FLUAV vRNA, cRNA, and/or mRNA that have been identified as important in FLUAV RNA synthesis or translation (7, 19, 20, 35, 37, 49). The P-PMO sequences, name designations, and target locations are described in Table Table11 and Fig. Fig.1B.1B. A 20-mer PMO of random sequence having 50% G+C content (named “Dscr”) was prepared and conjugated to either P4 or P7 peptide for use as a control for non-sequence-specific activity of the two respective P-PMO chemistries. Antisense and negative control P-PMO sequences were screened with BLAST (http://www.ncbi.nlm.nih.gov/BLAST/) against primate and canine mRNA sequences. Additionally, the negative control was screened against all FLUAV sequences to preclude unintentional hybridization events. Prior to use, lyophilized P-PMO were resuspended with filter-sterilized distilled water to a concentration of 1 to 2 mM and were stored at 4°C.
Influenza virus A/PR/8/34 (PR/8) (H1N1) was kindly provided by Peter Palese (Mount Sinai School of Medicine, N.Y.). Influenza B/Lee/40 virus was purchased from American Type Culture Collection. These viruses were grown in the allantoic cavity of 10-day-old embryonated chicken eggs (Charles River Laboratories, Wilmington, MA) at 37°C. Allantoic fluid was harvested 48 h after virus inoculation, aliquoted, and stored at −80°C. FLUAV A/WSN/33 (WSN/33) (H1N1) and A/Memphis/8/88 (Mem/88) (H3N2) were generously provided by Yoshihiro Kawaoka (University of Wisconsin-Madison) and were grown in Madin-Darby canine kidney (MDCK) cells cultured in virus culture medium #A (Eagle's minimum essential medium [EMEM] supplemented with 0.3% bovine serum albumin [BSA], 1.0 μg/ml tosylsulfonyl phenylalanyl chloromethyl ketone [TPCK]-treated trypsin [Sigma], and 100 U/ml penicillin-streptomycin), and supernatant was stored at −80°C. A/Eq/Miami/63 (Miami/63) (H3N8) and A/Eq/Prague/56 (Prague/56) (H7N7) were obtained from Rocky Baker (Veterinary Diagnostic Laboratory, Oregon State University) and were grown in MDCK cells cultured in virus culture medium #B (Dulbecco's minimum essential medium [DMEM], 1.0 μg/ml TPCK-treated trypsin, and antibiotics), and supernatant was stored at −80°C. Influenza A/Thailand/1(KAN-1)/2004 (KAN-1) (H5N1) was obtained from a human patient with severe pneumonia, cultured, and stored as previously described (38).
Vero cells in logarithmic phase were seeded in 24-well plates at 1 × 105 cells per well in DMEM containing 10% fetal calf serum, 2 mM l-glutamine, and antibiotics and were incubated at 37°C and 5% CO2. The next day, the cell growth medium was removed, the cells were rinsed once with serum-free DMEM, and 500 μl of serum-free DMEM containing either specified concentrations of P4-PMO or water (as mock treatment) was applied to culture wells and incubated at 37°C and 5% CO2 for 6 h. The treatment-containing media were then removed, and the cells were rinsed three times with media and infected with PR/8 or B/Lee/40 at a multiplicity of infection (MOI) of 0.05 in a volume of 200 μl media for 1 h. Following adsorption, the viral inoculum was removed and 1 ml of medium containing 4 μg/ml TPCK-trypsin (with or without P4-PMO, as specified) was added to each well, and the plates were incubated at 37°C and 5% CO2. At 24, 36, and 48 h postinfection (hpi), samples of virus culture supernatants were collected and titer measurements were performed by hemagglutination (HA) or plaque assays. HA assays were carried out using round-bottom 96-well plates. Serial twofold dilutions of virus samples were mixed with an equal volume of 0.5% chicken red blood cells (RBCs) (Charles River Laboratories) in phosphate-buffered saline and incubated on ice for 1 h. The titer was determined by noting the highest dilution of the virus which caused a hemagglutination reaction. Wells containing an adherent, homogeneous layer of erythrocytes were scored as positive. For plaque assays, serial 10-fold dilutions of virus-containing samples were added onto a monolayer of MDCK cells for 1 h, followed by an overlay of 1% semisolid agar. Two days after infection, plaques were visualized by staining with crystal violet and counted, and viral PFU/milliliter was calculated.
A quantitative colorimetric MTT cell proliferation assay kit (American Type Culture Collection) was used to quantify cell viability in response to treatment with P4-PMO used in the PR/8 experiments. Briefly, Vero cells were plated at a density of 1 × 104 cells per well in a flat-bottom 96-well culture plate and allowed to adhere overnight. The following day, 100 μl of serum-free DMEM containing either appropriate concentrations of P4-PMO or water was applied to culture wells in triplicate and incubated at 37°C and 5% CO2 for 24 h. After treatment, the MTT kit was used according to the supplier's instructions, and the absorbance of each well was determined on a microplate spectrophotometer (VERSAmax; Molecular Devices, Sunnyvale, CA) at a wavelength of 570 nm using the SOFTmax Pro program (Molecular Devices). Cytotoxicity was calculated by dividing the average optical density of treatment samples by the average of mock-treated samples.
MDCK cells in 12-well tissue culture plates at ~80% confluence were treated with specified concentrations of each P7-PMO or water as mock treatment in 1-ml/well virus diluent (virus culture medium #A without TPCK-trypsin) for 6 h. After removing the treatment-containing medium, the cells were infected with WSN/33 or Mem/88 at an MOI of 0.001 for 1 h at 37°C and 5% CO2, after which the inoculum was replaced with 2 ml/well virus culture medium #A without P7-PMO, and the cells were incubated at 37°C and 5% CO2. Aliquots of 200 μl were collected at 24 and 45 hpi and stored at −80°C. Viral replication was determined by standard plaque assay on confluent MDCK cells in the presence of 1.0 μg/ml TPCK-trypsin. The effect of P7-PMO on MDCK cell viability, under conditions identical to those of the above viral experiment but without virus, was determined by MTT assay in a manner similar to that described above for PR/8.
To evaluate the effect of preinfection treatment of cells with P7-PMO on subsequent determinations of viral titers, medium was removed and 500 μl of virus culture medium #B (but without TPCK-trypsin) containing either specified concentrations of P7-PMO, or water as mock-treatment, was added to cells and incubated at 37°C and 5% CO2 for 4 h. Treatment-containing media were then removed, and cells were infected with virus at an MOI of 0.0001 and allowed to adsorb for 1 h at room temperature on a shaker. Plates were then washed once, and 2 ml of virus culture medium #B, without P-PMO, was added to each well and incubated at 37°C and 5% CO2. Media supernatant samples were collected at 48 hpi, and virus titer was determined by HA assay. The effect of postinfection P7-PMO treatment on H3N8 virus replication was determined by adding P7-PMO directly to the virus culture medium at 1, 2, or 3 h after the viral infection period. Samples were collected at 48 hpi, and virus titer was determined by HA assay. HA assays were performed in 96-well round-bottom plates using twofold serial dilutions of 50 μl supernatant and the addition of 50 μl 0.4% chicken RBCs (PML Microbiologicals, Wilsonville, OR) to each well. Plates were incubated at room temperature for 2 h, and the titer was determined as for PR/8 above.
The effect of P7-PMO at concentrations of 10 to 400 μM on MDCK cell viability, under conditions similar to the above viral experiment but without trypsin or virus, was determined. As a positive control for cell death in this assay, the plant extract 6-prenyl naringenin (John Mata, Oregon State University) was used at 10 to 100 μg/ml. Briefly, triplicate samples of cells were exposed for 6 h to compound or mock treatment, the treatment was then removed, media were replenished, and the cells were incubated for 24 h. All samples were then assayed with the “Cell-Titer Blue” kit (Promega) according to manufacturer's instructions.
MDCK monolayers grown in EMEM with 10% fetal bovine serum and antibiotics, as described above, were pretreated for 4 h with the indicated concentrations of P7-PMO in media. The pretreatment media were then removed, and the cells were infected with KAN-1 (H5N1) virus at titers of 5 or 25 50% tissue culture infectious doses (TCID50) for 1 h. After removing the inoculum, media containing 2 μg/ml TPCK-trypsin and concentrations of P7-PMO identical to those in the pretreatment were added to the cells for 24 h. The amount of influenza virus present in the infected cell monolayer was then determined by the indirect enzyme-linked immunosorbent assay (ELISA) procedure for the measurement of the amount of viral nucleoprotein (NP), as previously described (38), except that the monoclonal antibody to FLUAV NP was purchased from Chemicon Inc. (Temecula, CA). The mean optical density at 450 nm of mock-treated infected samples was considered the 100% level of NP protein. The effect of P7-PMO on MDCK cell viability under cell culture and drug treatment conditions identical to those used in the KAN-1 (H5N1) viral experiment described above, but in the absence of virus, was assessed using the Cell Titer 96 Aqueous One Solution Cell Proliferation Assay (Promega) according to manufacturer's instructions.
Plasmid construction and cell-free assays. DNA corresponding to the coding sequence of firefly luciferase was subcloned into the multiple-cloning site of the T7 promoter-containing plasmid pCi-Neo (Promega) at the SalI and NotI sites. Subsequently, five pairs of complementary oligonucleotides, representing variations of the NP-AUG P4-PMO target sequence, with each having 0 to 4 mismatches in relation to the NP-AUG P4-PMO sequence, were duplexed and subcloned upstream of luciferase at the NheI and SalI sites. Each of the five plasmids was constructed such that the AUG translation initiation codon of luciferase is replaced with sequence corresponding to bases −7 to +13 relative to the A of the AUG translation initiation codon of the PR/8 NP gene (bases 39 to 58; GenBank accession no. NC_002019), which comprises the complete target site for the NP-AUG P4-PMO. A single AUG in the NP “leader” is in frame with the coding sequence of luciferase in the RNA produced from each of the five plasmids. After linearization of each plasmid [p(NP-0mp)luc, p(NP-1mp)luc, p(NP-2mp)luc, p(NP-3mp)luc, and p(NP-4mp)luc] with NotI, in vitro-transcribed RNA was produced using the T7-Megascript kit (Ambion) according to manufacturer's instructions. In vitro translations were carried out by programming rabbit reticulocyte lysate reactions with transcribed RNA at a final concentration of 1 nM, as previously described (33). The average light units produced by the set of reactions for each treatment were normalized to the means of all water-only control reactions and were expressed as a percentage of control reaction luciferase signal. Fifty percent effective concentrations were determined with GraphPad Prism graphing software (San Diego, CA).
Considerations in P-PMO sequence design for this study included our current understanding of both the function of various FLUAV genomic regions and P-PMO mechanisms of action. Previous reports indicate that the translation start site region of mRNA of viral nonstructural genes (1, 9, 41), or sequence involved in viral long-range RNA-RNA interaction (8, 15, 17), may include effectual P-PMO target sites. The P-PMO for this study were therefore designed to base pair with the AUG translation start site regions of the mRNAs involved in FLUAV RNA synthesis (PA, PB1, PB2, or NP) or one of the four terminal regions of the NP vRNA or cRNA (Fig. (Fig.1B;1B; Table Table11).
Eight P4-PMO were designed against the PR/8 (H1N1) sequence for the initial experiments in this study. The four AUG-targeted P4-PMO cover at least four nucleotides on either side of the AUG translation initiation codon of their respective mRNA target sequences and are also complementary to corresponding cRNAs.
The FLUAV RNA polymerase complex requires that the two termini of an RNA segment be partially duplexed in order to efficiently initiate RNA synthesis (37). Each segment of the FLUAV genome is believed to undergo a long-range RNA-RNA interactive event between the 5′ and 3′ ends during RNA synthesis, both in the production of cRNA or mRNA from vRNA and likewise from cRNA back to vRNA (49). NP has been reported to play critical roles in FLUAV genome replication, intracellular trafficking, packaging of the viral genome, and virus-host interactions (36). Therefore, four P4-PMO compounds were designed to base pair with one of the 5′- or 3′-terminal sequences of the NP vRNA or cRNA. In addition to the antisense P-PMO, a random-sequence P-PMO (designated “Dscr”) was synthesized to serve as a negative control compound.
The degree of sequence conservation between six FLUAV subtypes and between individual strains within those subtypes was also considered during P-PMO design. FLUAV subtypes selected for sequence conservation analysis included the three human-infecting subtypes H1N1, H2N2, and H3N2, the three avian influenza virus subtypes that are considered to be the largest threat to humans, H5N1, H7N7 and H9N2, and the primarily equine subtype H3N8. Table Table22 summarizes the degree of sequence conservation at the target sites in FLUAV of the eight P-PMO antisense sequences used in this study. Near-perfect homology across all subtypes and strains was most pronounced for the target of the PB1-AUG P-PMO.
Eight antisense and the random-sequence Dscr P4-PMO (Table (Table1)1) were initially compared at a single concentration for their ability to inhibit PR/8 (H1N1) production over a period of 48 h in Vero cells. FLUAV titers were monitored by HA assay at 24, 36, and 48 hpi. In a trial comparing the four AUG region-targeting compounds, all but the PA-AUG-targeted P-PMO produced at least eightfold reductions in virus titer at 48 hpi (Fig. (Fig.2A).2A). The four P-PMO targeting the terminal ends of NP were compared in a separate trial under conditions identical to those of the AUG-targeting (Fig. (Fig.2B).2B). Again, all but one P4-PMO (NP-c5′) was capable of significantly reducing virus titers, some by as much as 16-fold.
Seven of the eight P4-PMO were selected for further dose-response studies by HA and plaque assays. PA-AUG was not tested further because of its relatively low anti-PR/8 activity in the single-concentration survey described above. The procedures and conditions used for the PR/8 dose-response studies were identical to those described for the previous single-dose experiment, except that titers were measured at 48 h only. An overall dose-response effect was observed for all seven antisense P4-PMO tested by HA assays (Fig. 3A and B). At a concentration of 10 μM, PB1-AUG, NP-AUG, NP-c3′, NP-v3′, and NP-v5′ all caused a greater than eightfold reduction in titer compared to controls. Plaque assay of the same P4-PMO yielded similar results (Fig. 3C and D). At 10 μM, NP-AUG, PB1-AUG, and NP-v3′ inhibited virus titer by over 3 log10 PFU/ml compared to a peak titer of approximately 6.5 log10 in the mock- or Dscr-treated controls (Fig. (Fig.33).
Based on their high activity against PR/8 (H1N1) and high conservation of target site sequence across and within FLUAV subtypes, notably H1N1 and H5N1 (see Table Table2),2), two of the P-PMO sequences, PB1-AUG and NP-v3′, were selected for evaluation against several other strains of FLUAV. The Dscr negative control and the two antisense sequences were prepared as P7-PMO and evaluated by plaque assays against WSN/33 (H1N1) and Mem/88 (H3N2), by HA assays against Miami/63 (H3N8) and Prague/56 (H7N7), and by ELISA against KAN-1 (H5N1). The recently developed P7 was selected as the conjugation peptide for these subsequent experiments, as it has been reported to transport PMO into cells with equal or greater efficiency than that provided by P4 peptide, yet it is more stable (31), less affected by serum (8), and appears to be less cytotoxic than P4 (Hong Moulton, AVI BioPharma, unpublished data; also see below).
Pretreatment of cells with PB1-AUG P7-PMO before infection with either WSN/33 or Mem/88 virus resulted in pronounced antiviral activity. Cells were incubated for 6 h with P7-PMO, infected, and then incubated for 24 h without P7-PMO. At 24 hpi, WSN/33 titers were reduced 12-fold at 10 μM and 82-fold at 20 μM, while Mem/88 titers were reduced 8-fold at 10 μM and 216-fold at 20 μM (Fig. (Fig.4).4). No inhibition of viral titer was observed with either virus at the lowest P7-PMO concentration (5 μM) tested. WSN/33, but not Mem/88, titers at 24 hpi were reduced about 20-fold by preinfection treatment of cells with 20 μM NP-v3′, although lower concentrations of this compound did not significantly affect replication titers. Replication of either virus was not affected by the presence of Dscr control P7-PMO at any concentration used throughout the assays. The titers of both viruses in the presence of any of the P7-PMO were similar to mock-treated infection controls by 45 hpi (data not shown), suggesting that there was insufficient intracellular presence of the oligomers to sustain their antiviral activity.
The effect of preinfection treatment with PB1-AUG and/or NP-v3′ P7-PMO on H3N8 and H7N7 viruses in MDCK cells was measured by HA assay. Cells were incubated for 4 h with P7-PMO, infected, and then incubated for 48 h without P7-PMO. As shown in Fig. Fig.5,5, NP-v3′ produced a dose-dependent response, with a greater than 85% reduction in virus titer at 15 μM, against both H3N8 and H7N7. However, the PB1-AUG P-PMO appears to have little or no effect on virus titer, and additional experiments will need to be conducted to further investigate this result. Although PB1-AUG appeared to have no effect by itself, when combined with NP-v3′ at a concentration of 15 μM each an additive effect was realized, producing a reduction in virus titers of more than 90%, which is an effect greater than NP-v3′ alone at 15 μM.
To examine the effect of PB1-AUG and NP-v3′ P7-PMO on the replication of an H5N1 isolate, MDCK cells were incubated with P7-PMO for 4 h, infected with two different levels of KAN-1 (H5N1) virus, incubated again with P7-PMO, and assayed for activity by ELISA at 24 hpi. Both antisense P7-PMO generated dose-dependent inhibition at either virus infection level, with NP-v3′ consistently generating moderately greater antiviral effect at the lower doses than PB1-AUG. At the lower virus infection level (5 TCID50), NP-v3′ at 10 μM produced 88% and PB1-AUG 57% inhibition of viral NP protein levels (Fig. (Fig.6A).6A). In contrast, the Dscr control P7-PMO had less than 8% activity at this dose. At the higher virus infection level (25 TCID50), both antisense P7-PMO produced a greater than 85% reduction in viral NP protein at 20 μM, a dose at which the Dscr control P7-PMO had no activity (Fig. (Fig.6B6B).
In order to assess the sequence specificity and cytotoxicity of the P4-PMO used in this study, each was tested at 20 μM for inhibition activity against influenza B virus (IBV). As with the PR/8 (H1N1) inhibition assay, IBV was inoculated at 0.05 MOI in Vero cells, and titers were determined at 48 hpi by HA assay. IBV grew to nearly as high a titer at 48 h as PR/8 had. No difference in IBV titer between cells treated with any of the P4-PMO and cells receiving mock treatment was observed (data not shown). This was not surprising, as the degree of sequence conservation between IBV and FLUAV at any of the P4-PMO target sites is below 82% (data not shown). We interpret this result as evidence that the P4-PMO had minimal cytopathic or generic antiviral activity, as IBV titer would be expected to diminish compared to that for mock-treated samples if any of the P4-PMO had caused such nonspecific effects.
All P4- and P7-PMO were tested for cytotoxicity with standard cell viability assays in the absence of virus, under the same culture conditions, and with the same, or wider, range of P-PMO concentrations as those in the various antiviral experiments. The P4-PMO exhibited a somewhat higher impact on cell viability than P7-PMO did, with concentrations of some P4-PMO in the 10 to 20 μM range, resulting in an approximately 20% loss in cell viability after 24 h of incubation. In a variety of trials with MDCK cells, P7-PMO generated less than 10% reduction in cell viability at all concentrations up to and including 20 μM for all treatment durations (data not shown). However, at 40 μM, with a 24-h treatment under the conditions of the H5N1 antiviral assay, each P7-PMO reduced MDCK cell viability from 15 to 30%. In an attempt to derive a “Selectivity Index” (SI) (the ratio of the concentration of drug causing 50% cytotoxicity [CC50] divided by the concentration of drug causing a 50% inhibition of viral production) relevant to the various conditions of this study, we sought to determine the CC50 for the PB1-AUG and NP-v3′ P7-PMO. MDCK cells were treated for 6 h with concentrations of each P7-PMO from 10 to 400 μM, using culture conditions identical to those under which the H3N8 and H7N7 antiviral experiments were conducted but in the absence of trypsin or virus. Cell viability was determined at 24 h after the treatment period. Surprisingly, we were unable to cause a 50% loss in cell viability under these conditions. The experiment was repeated three times, with similar results. A concentration of 400 μM of either oligomer resulted in a 15 to 30% loss in cell viability (data not shown). An unrelated natural plant extract, 6-prenyl naringenin, generated a dose-responsive pattern of cytotoxicity, with a CC50 of approximately 20 μg/ml. Based on a 50% inhibition of viral production of approximately 10 μM for both PB1-AUG (Fig. (Fig.4)4) and NPv3′ (Fig. (Fig.5),5), we conclude that the SI for either antisense P7-PMO is over 40 under these conditions.
Together, these results indicate that under the various experimental conditions used in this study P-PMO did not have a significant impact on cell viability, and the observed antiviral activity was sequence specific.
Having established the potent anti-FLUAV activity of P-PMO in settings in which cells were pretreated with compound for 4 or 6 h before viral infection, we investigated the effect of treating cells only after viral infection. At 1, 2, or 3 h after the completion of a 1-h infection period with H3N8, 15 μM of Dscr, NP-v3′, and/or PB1-AUG P7-PMO was added and allowed to remain in the culture medium (which contained trypsin). The results from these experiments showed that, at 48 hpi, NP-v3′ provided a 70% reduction in virus titer if treatment was begun 1 h after the completion of the infection period (Fig. (Fig.7).7). Reduction in titer decreased to 40% or 20% if treatment was begun 2 or 3 h, respectively, after the infection period. As with the preinfection protocol, PB1-AUG and Dscr had no observable effect on H3N8.
In an attempt to characterize the loss of P-PMO activity that may result from a variable number of mispairs between the base sequence of a P-PMO and its RNA target sequence, the NP-AUG P4-PMO was tested in cell-free in vitro translation inhibition assays against a panel of in vitro-transcribed RNAs. Five RNAs were produced from a series of plasmids that were constructed such that the target sequence for NP-AUG P4-PMO was fused in frame directly upstream of luciferase and downstream from a T7 promoter. The plasmids differ incrementally in the number of mismatches, from 0 to 4, present in the NP-AUG P4-PMO target “leader-sequence” in relation to the NP-AUG P4-PMO sequence. Table Table33 lists the NP AUG region leader sequences of the five plasmids. The mismatched RNA target sequences were designed to reflect the most likely natural sequence variations in the NP-AUG region, as derived from the influenza sequence databases. In vitro translations with 1 nM RNA in rabbit reticulocyte lysate were carried out with the five different RNAs against a 6-point dose-response of NP-AUG and Dscr P4-PMO. The results showed that a single mismatch between P4-PMO and target RNA caused little reduction of P4-PMO activity compared to no mismatches (Fig. (Fig.8).8). Two or more mismatches, however, generated a considerable loss of inhibitory activity. The 50% effective concentrations of the NP AUG P4-PMO against the RNAs with which it had zero, one, two, three, or four mismatches were approximately 30, 50, 200, 400 and 550 nM, respectively. All concentrations of Dscr generated little or no inhibition of translation of any of the RNAs used in the in vitro translation assays.
Currently available vaccines and therapeutics are not sufficient to effectively prepare for and respond to the public health threat that FLUAV represents (6, 21, 22, 28). This report documents an initial investigation into the viability of using P-PMO as an anti-FLUAV agent. Eight P-PMO compounds were designed to duplex with regions of FLUAV RNA believed to be critical to one or more events of the virus biological cycle. Four of the P-PMO each target one of the AUG translation start site regions of viral mRNA encoding PB1, PB2, PA, and NP. The other four P-PMO each target one of the 5′- or 3′-terminal regions of the NP vRNA or cRNA. Several FLUAV strains were subject to P-PMO treatment in cell cultures. Two P-PMO, one targeting the PB1 translation initiation region and the other the 3′-terminal region of the NP vRNA, were shown to produce high antiviral efficacy against multiple subtypes of FLUAV.
Eight antisense P4-PMO compounds were initially compared for efficacy and specificity against PR/8 (H1N1) (Fig. (Fig.2).2). Each of the eight generated some degree of antiviral activity. Of the four AUG region-targeting compounds, the one designed against the PA segment was the least effective, while the other three, specific for PB1, PB2, and NP segments, produced marked virus titer reduction (Fig. (Fig.2A).2A). All AUG region-targeting P4-PMO have sequence complementary to both their respective mRNA and cRNA. Based on current understanding, we assume that these P-PMO likely inhibited virus production by steric interference with mRNA translation; however, the possibility exists that these compounds disrupted the synthesis of vRNA from cRNA. The lack of inhibitory effect by the PA-targeting P-PMO may have been due to inefficient duplex formation between P-PMO and RNA target or to duplex formation that was not highly disruptive to any critical molecular event. Alternatively, lowered levels of PA gene product may simply have had little impact on viral amplification in this setting.
The high efficacy of P-PMO designed to base pair to one or the other end of NP vRNA (NP-v3′ and NP-v5′) represents further confirmation that access to both vRNA termini is important in NP RNA synthesis. We found the low level activity of NP-c5′ puzzling, as it was designed to anneal to sequence of consequence in two separate species of NP RNA: the 5′-terminal region of the NP cRNA and in the 5′-untranslated region of NP mRNA.
The data presented in this report document high efficacy of PB1-AUG and/or NP-v3′ P-PMO against a total of six FLUAV strains from five different FLUAV subtypes. Due to the high sequence conservation of the two P-PMO target sites in FLUAV (see Table Table2),2), these two oligomers could be expected to similarly inhibit other FLUAV strains and subtypes. The lack of efficacy by NP-v3′ P7-PMO against Mem/88 (H3N2) (Fig. (Fig.4B)4B) was likely due to the presence of two mismatches between the sequence of the oligomer and its target RNA. The lack of activity of PB1-AUG P7-PMO against the H3N8 and H7N7 strains (Fig. (Fig.5),5), in light of its high efficacy against the other four FLUAV strains tested, is difficult to reconcile. However, the PB1 gene sequence for both the H3N8 and H7N7 strains used in this study are not available in the influenza sequence databases, and it is possible that sequence mismatches exist that would reduce duplex formation.
Inhibition of WSN/33 and H3N2 but not H3N8 and H7N7 by PB1-AUG P7-PMO, and conversely the inhibition of H3N8 and H7N7 but not WSN/33 by NP v3′ P7-PMO, indicates that both P7-PMO had low cytopathic effect. Further confirmation that the toxicity of the P7-PMO was low under the conditions of this study was provided by the negligible effect of the Dscr P7-PMO on viral titers and by cell viability assays, including the approximation of an SI of more than 40 for both PB1-AUG and NP-v3′.
Simultaneous use of both PB1-AUG and NP-v3′ P7-PMO may be able to provide greater protection against a spectrum of FLUAV strains, as well as greater efficacy against certain individual strains, than either oligomer alone. Furthermore, due to the dissimilar target locations within the FLUAV genome of the two oligomers, we expect that their simultaneous use would reduce the probability of emergence of resistant FLUAV through genetic drift or shift. The PB1-AUG P-PMO target region appears to be especially invariant across FLUAV strains (Table (Table2),2), suggesting that mutations in this particular region may result in nonviable virus.
In experiments with WSN/33 and H3N2, where treatment was applied before infection only, antiviral effect was observed at 24 hpi (Fig. (Fig.4).4). At 45 hpi, however, no antiviral activity was observed. The persistence of antiviral activity in cells for a minimum of 24 h after a 6-h treatment period suggests that P-PMO remains active and stable for many hours after its initial entry into cells. We assume that between 24 and 45 h a number of events, including cell division and rapid replication of the virus strains used, combined to overcome the P-PMO antiviral effect. The possibility of escape mutant amplification in the presence of the P-PMO was not investigated in this study but will likely be the subject of future study.
In an experiment where P7-PMO treatment was applied to cells at different time points exclusively after infection, considerable antiviral efficacy (70% reduction) was observed if the treatment began 1 h after the infection period ended. However, if P-PMO treatment was delayed until 3 h after the infection period, little (20%) reduction of viral titer occurred (Fig. (Fig.7).7). These data suggest that the additional 1 to 2 h before the commencement of treatment may have allowed viral replication enough of a “head-start” that the P7-PMO was unable to intervene in consequential events of the virus life cycle in a timely and effective manner. However, we note that, as in the H5N1 experiment (Fig. (Fig.6),6), the P7-PMO was added directly to trypsin-containing growth medium. It is probable that much of the P7 peptide, which contains arginine:arginine repeats, was degraded before transporting PMO into cells (31). The rate of degradation of the conjugate peptide in vivo may not be a major issue for drug delivery, however, as ARPs have been shown to effectively deliver macromolecular cargo in vivo in several studies (5, 9, 30, 39).
NP-AUG P4-PMO, which targets the NP mRNA translation start site region, was used to investigate the effect of mispairing between the base sequence of a P4-PMO and its target RNA sequence. Reporter constructs containing cloned target sequence variants were used to determine translation inhibition via in vitro transcription/translation reactions. The data indicate that a single mismatch between RNA target and P4-PMO results in a negligible loss of activity, while more than one mispair results in a substantial loss of inhibition (Fig. (Fig.8).8). This degree of mismatch tolerance by a P-PMO suggests a favorable characteristic for a potential therapeutic agent, in that single viral mutations at the P-PMO target site may be insufficient to produce drug resistance, yet the likelihood of undesirable off-target reactivity with host RNA is minimized.
In summary, this report documents several locations in the FLUAV genome as productive targets for P-PMO antiviral intervention, including the translation-initiator regions of several viral mRNAs and the NP segment vRNA terminal regions. Furthermore, the potent inhibition by two P7-PMO of several FLUAV subtypes, including a highly pathogenic H5N1 isolate, raises the possibility that the two oligomers could be combined to produce a broad-spectrum therapeutic appropriate against a high percentage of FLUAV strains. The level and duration of suppression of multiple FLUAV subtypes by P7-PMO suggests that these compounds should be evaluated in vivo.
We thank the Chemistry Group at AVI BioPharma for expert production of P-PMO and Andrew Kroeker and Robert Blouch for superb technical assistance.
The work was supported partly by NIH grants AI56267 (to J. Chen) and P50-CA112967 (to R. Hynes).
Published ahead of print on 11 September 2006.