New HBV Therapy Research for siRNAs
"Potent and persistent in vivo anti-HBV activity of chemically modified siRNAs"
Nature Biotechnology 23, 1002 - 1007 (2005)
Published online: 24 July 2005
David V Morrissey1, 3, Jennifer A Lockridge1, 3, Lucinda Shaw1, Karin Blanchard1, Kristi Jensen1, Wendy Breen1, Kimberly Hartsough1, Lynn Machemer1, Susan Radka1, Vasant Jadhav1, Narendra Vaish1, Shawn Zinnen1, Chandra Vargeese1, Keith Bowman1, Chris S Shaffer1, Lloyd B Jeffs2, Adam Judge2, Ian MacLachlan2 & Barry Polisky1
1 Sirna Therapeutics, Inc., 2950 Wilderness Place, Boulder, Colorado 80301, USA.
2 Protiva Biotherapeutics, Inc., 100-3480 Gilmore Way, Burnaby, British Columbia V5G 4Y1, USA.
The efficacy of lipid-encapsulated, chemically modified short interfering RNA (siRNA) targeted to hepatitis B virus (HBV) was examined in an in vivo mouse model of HBV replication. Stabilized siRNA targeted to the HBV RNA was incorporated into a specialized liposome to form a stable nucleic-acid-lipid particle (SNALP) and administered by intravenous injection into mice carrying replicating HBV. The improved efficacy of siRNA-SNALP compared to unformulated siRNA correlates with a longer half-life in plasma and liver.
Three daily intravenous injections of 3 mg/kg/day reduced serum HBV DNA >1.0 log10. The reduction in HBV DNA was specific, dose-dependent and lasted for up to 7 d after dosing. Furthermore, reductions were seen in serum HBV DNA for up to 6 weeks with weekly dosing.
The advances demonstrated here, including persistence of in vivo activity, use of lower doses and reduced dosing frequency are important steps in making siRNA a clinically viable therapeutic approach.
Development of siRNAs as therapeutic agents will require improvements in both the stability of siRNAs1, 2, 3 and the efficiency and specificity of tissue-targeted delivery in vivo. We previously reported on a novel combination of chemical modifications that dramatically increase the in vivo stability of siRNA3. In an in vivo mouse model of HBV replication, stabilized siRNAs targeted to the HBV RNA (HBV263M) were dosed at 30 mg/kg three times per day resulting in a 90% reduction in serum HBV DNA. To find a more therapeutically viable dosing regimen, we have evaluated the use of lipid encapsulation as a means of improving the pharmacology of siRNA. The SNALP consists of a lipid bilayer containing a mixture of cationic and fusogenic lipids that enables the cellular uptake and endosomal release of the particle's nucleic acid payload. SNALPs are also coated with a diffusible polyethylene glycol-lipid (PEG-lipid) conjugate that provides a neutral, hydrophilic exterior and stabilizes the particle during formulation. The surface coating also shields the cationic bi-layer in vivo, preventing rapid systemic clearance. Upon administration, the PEG-lipid conjugate dissociates from the SNALP, transforming the particle into a transfection-competent entity4.
Two siRNA molecules, HBV263 and HBV1583, were selected and chemically stabilized for nuclease resistance using a previously reported method3 in which all 2'-OH residues on the RNA were substituted with 2'F, 2'O-Me or 2'H residues. We then analyzed in vitro potency and placed one to three ribonucleotides on the 5' ends of the antisense strands; these molecules were named HBV263M and HBV1583M. Subsequently, it was found that the placement of one, two or three ribonucleotides on the 5' end of the modified siRNA (named HBV263M and HBV1583M) improved the median inhibitory concentration (IC50) of the stabilized siRNAs by approximately fivefold (15 nM versus 3 nM; data not shown). The in vitro potency of the siRNAs was initially determined by using HepG2 cells with replicating HBV. The IC50s of the unmodified siRNAs (HBV263R and HBV1583R) and modified siRNAs (HBV263M and HBV1583M) were determined by standard cationic lipid transfection to be approx0.5 nM and 3 nM, respectively (data not shown).
The HBV-targeted siRNAs were encapsulated in lipid nanoparticles to form SNALPs, with the final nucleic acid/lipid ratio and mean particle size being similar for all SNALPs regardless of their nucleic acid payload. The mean particle size of 14 SNALP preparations was 140 plusminus 12 nm and polydispersity was 0.11 plusminus 0.02. Mean encapsulation efficiency was 93% plusminus 3% and the final nucleic acid/lipid ratio was 48 plusminus 8 mug/mumol. Based upon four replicate HBV263M-SNALP preparations used in these studies, mean particle size was 141 plusminus 14 nm, polydispersity was 0.12 plusminus 1, mean encapsulation efficiency was 94% plusminus 4% and nucleic acid/lipid ratio was 42 plusminus 4 mug/mumol.
The in vitro potency of HBV263M-SNALP and HBV1583M-SNALP was determined in HBV-replicating HepG2 cells. The siRNAs-SNALP reduced hepatitis B surface antigen (HBsAg) in HepG2 cells compared to untreated cells in a dose-dependent manner, whereas control siRNA had no substantial effect on HBsAg. Up to 98% reduction in HBsAg was observed in HepG2 cells treated with 25 nM siRNA-SNALP and the IC50 of HBV263M and HBV1583M was approx1 nM. These results demonstrate that siRNA targeted to HBV and formulated as a SNALP has high potency and activity in cell culture, indicating efficient cell entry and interaction with the RNA-induced silencing complex.
To measure the biodistribution of SNALPs, HBV263M-SNALP was prepared containing the nonexchangeable lipid label 3H-CHE, as described previously6, 7. Substantial quantities of SNALP accumulated in liver (28% plusminus 1.7%) and spleen (8.2% plusminus 2.8%) by 24 h after injection. Remarkably little SNALP accumulated in lung (0.3% plusminus 0.1%). The biodistribution of siRNA-SNALP in mouse liver was examined in greater detail using cyanine 3 (Cy3)-end labeled HBV263M. The Cy3-labeled HBV263M-SNALP was injected into female CD1 mice at a dose of 3 mg/kg. Intense staining was observed 2 h after injection (Supplementary Fig. 3 online). Regions of fluorescence were seen in hepatocytes at 2, 8 and 24 h after dosing, although staining intensity decreased slightly at 8 h, and was primarily punctate staining after 24 h. Diffuse background Cy3 staining increased between 8 and 24 h, consistent with release of Cy3-labeled siRNA from SNALPs through fusion with the endosomal membrane. An intense signal was observed in some Kupffer cells and vascular endothelial cells surrounding hepatocytes and in blood vessels up to 72 h after dosing, but the intensity of staining decreased in all liver cell types by day 7 after dosing. In the mouse, unstabilized, unformulated siRNA was rapidly eliminated from the plasma compartment with an elimination half-life of approx2 min. Administration of an siRNA that had been chemically stabilized improved the elimination half-life (T1/2 = 0.8 h or 49 min) compared to the unstabilized compound. However, the most dramatic improvement to the pharmacokinetics resulted from formulation of the siRNA with SNALP (T1/2 = 6.5 h).
The half-life of HBV263M-SNALP in mouse plasma and liver was also evaluated in mice treated with the same chronic regimen used in efficacy studies. The half-life of HBV263M-SNALP was estimated to be 12.4 or 6.1 h in plasma (Fig. 2b) and 11.2 or 15.1 h in liver (Fig. 2c) after the first daily or fourth weekly dose, respectively. The area under the curve (AUCto6h) was similar after the first daily and fourth weekly doses (40.7 and 48.3 h*ng/mg, respectively), suggesting that HBV263M-SNALP had similar pharmacokinetic properties throughout the 5-week experimental protocol. The maximum plasma concentration followed the same trend as AUCto6h (data not shown).
The immunostimulatory properties of the siRNA-SNALP formulation were evaluated because recent studies have demonstrated that unmodified, synthetic siRNA can induce a high level of type I interferon and inflammatory cytokines in mammalian cells8, 9. The potential for siRNAs-SNALP to induce a cytokine response in vivo, and the impact of siRNA chemical modifications on such a response, was examined by intravenous administration of HBV263M-SNALP and HBV263invM-SNALP as well as versions of both, designated HBV263R-SNALP and HBV263invR-SNALP, in which the duplex was composed of unmodified ribonucleotides except for the addition of inverted deoxy abasic end caps. The addition of the end caps provides some level of stability since completely unmodified siRNA has a half-life in serum of approximately 1.5 min.
Strikingly, HBV263R-SNALP and HBV263invR-SNALP strongly induced interferon alpha (IFN-alpha) and inflammatory cytokines in the serum of injected mice (Fig. 3a). In contrast, administration of HBV263M-SNALP or HBV263invM-SNALP induced no detectable increase in serum interferon-alpha (IFN-alpha) or inflammatory cytokines (IL-6, TNF-alpha). In addition, mice treated with unmodified siRNA-SNALP, either HBV263R-SNALP or HBV263invR-SNALP, had significant levels of interferon in the liver, whereas interferon was not detected in samples from mice treated with the modified HBV263M-SNALP or HBV263invM-SNALP. These data indicate that the siRNA modification protocol described above not only increases siRNA in vivo stability but also abrogates undesirable immunostimulatory properties of unmodified siRNA. The toxicity of HBV263M-SNALP and HBV263R-SNALP was examined by monitoring serum transaminase concentrations. In mice treated with 3 mg/kg of the unmodified HBV263R-SNALP, the serum aspartate aminotransferase (AST) level was threefold higher than that of control and the serum alanine aminotransferase (ALT) level was twofold higher than that of control, whereas levels of AST and ALT were not significantly altered in mice treated with 3 mg/kg of the chemically modified HBV263M-SNALP. The elevated AST and ALT levels indicated liver inflammation, which correlated with the observed cytokine induction from the unmodified siRNA-SNALP. Animals treated with HBV263R-SNALP showed other symptoms of systemic toxicity including transient lymphopenia and thrombocytopenia, decrease in body weight and piloerection (data not shown). These adverse effects were not evident in animals treated with the chemically modified HBV263M-SNALP.
A mouse model of HBV replication was used to evaluate the in vivo efficacy of HBV263M-SNALP and HBV1583M-SNALP. In this model, NOD.CB17-Prkdcscid/J mice were given 0.3 mug HBV vector DNA by hydrodynamic injection10, resulting in HBV replication in the liver and persistent viral titers in the serum3, 11. After hydrodynamic injection of the HBV vector, the mice were allowed to recover for 6 d before administration of the siRNA-SNALP by standard intravenous injection. This recovery period allowed liver ALT/AST and liver histopathology to return to normal (data not shown10) before siRNA-SNALP treatment.
Animals were treated with HBV263M-SNALP, HBV263invM-SNALP, HBV263R-SNALP or HBV263invR-SNALP at 0.3 or 3 mg/kg/day for 3 consecutive days and killed 3, 7, or 14 d after the last treatment. Reductions in HBV serum titers were observed with both modified and unmodified siRNA on days 3 and 7, but were lost by day 14. At the 3-d time point (Fig. 4a), very similar decreases in serum HBV DNA were observed in both the HBV263M-SNALP and HBV263R-SNALP groups treated with 3 mg/kg/day. A 1.26 log10 (P < 0.0001, 0.99-1.54, 95% confidence interval (CI)) reduction was seen in the HBV263M-SNALP treated group as compared to the HBV263invM-SNALP-treated group, and a 1.14 log10 (P < 0.0001, 0.88-1.41, 95% CI) reduction was observed compared to the saline group (control). Mice treated with 3.0 mg/kg/day of HBV263R-SNALP exhibited an HBV serum titer that was 1.05 log10 (P < 0.0001, 0.71-1.38, 95% CI) less than that of the HBV263invR-SNALP treated group, and 1.13 log10 (P < 0.0001, 0.80-1.45, 95% CI) less than the saline group. In the 0.3 mg/kg/day groups, modest decreases (approx0.3 log10) in HBV serum titer were observed with either the stabilized or unmodified siRNA-SNALPs, as compared to the matched controls. Similar results were also observed with serum HBsAg levels at day 3, with 90% and 95% decreases in the 3 mg/kg/day HBV263M-SNALP- and HBV263R-SNALP treated groups respectively, as compared to the matched inverted control groups. In the 0.3 mg/kg/day active groups (HBV263M-SNALP and HBV263R-SNALP), the activity was more pronounced than with the serum HBV DNA endpoint, with approx50% reduction observed with either siRNA chemistry.
At the 7-d time point, a significant difference was observed in the specificity of anti-HBV activity between the stabilized and unmodified siRNAs-SNALP. A decrease in serum HBV DNA levels of 1.04 log10 (P < 0.0001, 0.73-1.34, 95% CI) was seen in the 3 mg/kg/day HBV263M-SNALP treated group as compared to .89 log10 (P < 0.0001, 0.63-1.15, 95% CI) in the HBV263invM-SNALP treated group. No activity was seen at the 0.3 mg/kg dose (Fig. 4c). Decreases of 86% and 32% in serum HBsAg levels were observed in the groups treated with 3.0 and 0.3 mg/kg/day HBV263M-SNALP, respectively, as compared to the HBV263invM-SNALP-treated or saline groups. The HBV263R-SNALP-treated group displayed anti-HBV activity similar to that of the stabilized siRNA-SNALP at either dose level with both the HBV DNA or HBsAg endpoints on day 7. However, in contrast to the stabilized inverted control siRNA, HBV263invR-SNALP showed significant (P < 0.0001) nonspecific activity in reducing both HBV DNA and HBsAg levels. In fact, an equivalent decrease (approx0.8 log10) in HBV DNA levels was observed in both the high dose HBV263R-SNALP and HBV263invR-SNALP groups as compared to the saline-treated group. This nonspecific activity observed with the unmodified siRNA is likely the result of the cytokine response observed with this siRNA chemistry (Supplementary Fig. 6 online).
To examine the dose-dependent activity of HBV263M-SNALP, we treated mice with 1, 3 or 5 mg/kg/day HBV263M for 3 d; serum HBV DNA was measured 3 d after the last treatment. A dose-dependent reduction in the serum HBV DNA titer was observed, with a maximum decrease of 1.63 log10 (P < 0.0001, 1.22-2.03, 95% CI) at the 5 mg/kg dose relative to saline (Fig. 4e). The in vivo efficacy of SNALP-formulated siRNA as an anti-HBV treatment was also examined using a chronic dosing regimen. Mice were treated with HBV1583M-SNALP at 3 mg/kg/day for 3 d and then weekly for 5 weeks. The use of the HBV site 1583 siRNA provided an opportunity to demonstrate anti-HBV activity with a second independent siRNA targeting a separate site on the HBV RNA. The results show that HBV1583M-SNALP reduced serum HBV DNA titers by an average of 0.85 log10 relative to controls after 2, 4 or 6 weeks of treatment (Fig. 4f). Thus, the reduction of serum HBV DNA was statistically equivalent over the 6-week course of the experiment. This result also provides further support for a specific siRNA-mediated targeting of HBV RNA, rather than nonspecific off-target effects.
A notable result of these studies was the distinct difference between the in vivo behavior of the chemically stabilized and unmodified siRNAs observed with respect to the specificity of anti-HBV activity and the acute inflammation associated with HBV-SNALP treatment. A likely explanation for both observations are the significant levels of IFN-alpha and inflammatory cytokines induced by the unmodified HBV siRNA8. Interferons have well-characterized antiviral activity12, and their induction by siRNA in vitro has been implicated in mediating off-target gene effects13. The anti-HBV activity exhibited by the unmodified HBV263invR-SNALP may reflect such a nonspecific response. Similarly, the anti-HBV effects of the unmodified active HBV263R-SNALP may not be completely attributable to an RNAi-based mechanism. In contrast, the stabilized inverted siRNA, shown in Figure 3 to be nonimmunostimulatory, had minimal effects on HBV replication. Recent reports have highlighted the sequence-specific activation of innate immunity by siRNA, apparently through engagement of Toll-like receptor 7 (refs. 8,9). Our findings indicate that chemically modified siRNAs can abrogate recognition by the innate immune system, thereby reducing toxicity.
The advances demonstrated in this work, including persistence of activity, use of lower doses and reduced dosing frequency, are important steps towards making siRNA therapeutics a reality. As shown in Figure 7, a significant reduction in HBV serum titers, as assayed by either HBV DNA or HBsAg, is observed at a time point 7 d after the last of the three daily injections. This persistent in vivo activity is a major advance relative to previous reports showing silencing activity for only 18 h (ref. 3) or 24 h (ref. 14). In addition, the doses demonstrating efficacy in these prior reports were significantly higher than those shown in this study. The daily loading doses demonstrating efficacy in the current study are 30-fold lower than those reported previously (1 versus 30 mg/kg). Moreover, the single maintenance dose given each week is approximately two logs lower than doses used previously when the average daily dose is calculated (0.43 versus 30 mg/kg). If scaled directly for a 70-kg human, it is obvious that doses of 30 mg/kg/d (>2,100 mg) would be untenable. In vivo siRNA activity reported3, 14 relied on daily IV dosing, which is also clinically untenable. In contrast, we have demonstrated sustained long-term anti-HBV activity with once weekly siRNA maintenance doses. This dosing regimen is also likely to minimize the potential for oligonucleotide-related toxicities, which have generally been dose- and concentration-dependent.
This report demonstrates that SNALP formulation of stabilized siRNA is an efficient method for delivering siRNA to mouse liver and reducing HBV DNA titer in vivo. The improved efficacy of SNALP-formulated siRNA correlates with a longer half-life in plasma and liver, and reduced toxic and immunostimulatory side effects. Compared to previous reports on systemic dosing of siRNA to the liver3, 14, this work has demonstrated more robust and persistent biological activity, use of lower doses and reduced dosing frequency. These are all important steps towards making siRNA therapeutics a reality.
Oligonucleotide synthesis and characterization.
All RNAs were synthesized at Sirna Therapeutics by standard procedures15. Complementary strands were annealed in PBS, desalted and lyophilized. The sequences of the active HBV siRNAs are shown in Figure 1. The modified siRNAs used in vivo are termed HBV263M and HBV1583M, while versions containing unmodified ribonucleotides and inverted abasic terminal caps are called HBV263R and HBV1583R. Some pharmacokinetic studies were done with siRNA targeting two other sites, HBV1580M and HBV1580R.
The siRNA sequences for HCV irrelevant control are:
sense strand: 5'-CUGAUAGGGUGCUUGCGAGTT-3'
antisense strand: 5'-CUCGCAAGCACCCUAUCAGTT-3'
The inverted control sequences are inverted from 5' to 3'.
Except where otherwise stated, all animal studies were conducted at Sirna Therapeutics with approval by Sirna's Institutional Animal Care and Use Committee. Sirna's animal facility is AAALAC accredited. Animal studies overseen by Protiva were completed with local Animal Care and Use Committee oversight and in accordance with the Canadian Council on Animal Care guidelines.
Encapsulation of siRNA.
siRNAs were encapsulated into liposomes with a lipid composition of DSPC:Chol:PEG-C-DMA:DLinDMA (20:48:2:30 molar percent) by an adaptation of a previously published method16 whereby detergents are replaced by ethanol for the solubilization of lipid components17. Synthetic cholesterol was from Sigma whereas the phospholipid DSPC (1,2-distearoyl-sn-glycero-3-phosphocholine) was from Avanti Polar Lipids. DLinDMA and PEG-C-DMA were synthesized by Protiva. Nucleic acid encapsulation was determined using a RiboGreen assay (Molecular Probes), comparing fluorescence in the presence and absence of Triton X-100. Particle size was determined using a Malvern Instruments Zetasizer 3000HSA.
Cell culture studies.
Hep G2 cells were grown in EMEM (Cellgro) with nonessential amino acids, sodium pyruvate (90%), and 10% fetal bovine serum (HyClone). Replication-competent cDNA was generated by the excision and re-ligation of the HBV genomic sequences from the psHBV-1 vector. Hep G2 cells were plated (3 times 104 cells/well) in 96-well microtiter plates and incubated overnight. A cationic lipid/DNA complex was formed containing (at final concentrations) cationic lipid (11-15 mug/ml), and re-ligated psHBV-1 (4.5 mug/ml) in growth medium. After a 15-min incubation at 37 íC, 20 mul of the complex was added to the plated Hep G2 cells in 80 mul of growth medium minus antibiotics. After 7.5 h at 37 íC, the medium was then removed, the cells rinsed once with medium, and 100 mul of fresh medium was added to each well. We added 50 mul of the siRNA-SNALP formulation (diluted into media at a 3times concentration) per well, with three replicate wells per concentration. The cells were incubated for 4 d, the medium was then removed, and assayed for HBsAg levels.
HBsAg enzyme-linked immunosorbent assay (ELISA).
Levels of HBsAg were determined using the Genetic Systems/Bio-Rad HBsAg ELISA kit, as per the manufacturer's instructions. The absorbance of cells not transfected with the HBV vector was used as background for the assay, and thus subtracted from the experimental sample values.
Tissue distribution studies and pharmacokinetic analysis.
A tissue distribution study was conducted by Protiva. Radiolabeled SNALP were prepared for short-term distribution studies by incorporation of 2 muCi/mg lipid of the nonexchangeable lipid label 3H-CHE (ref. 5). SNALP were administered by lateral tail vein injection in 8-week-old male A/J mice. At 24 h, mice were euthanized by CO2 inhalation, tissues were collected and homogenized in lysing matrix tubes containing 500 mul of distilled water. Liver lysate (100 mul) and 200 mul of all other tissue lysates were assayed for radioactivity by liquid scintillation counting with Picofluor 40 (Perkin Elmer). Values were corrected for plasma levels at 24 h (ref. 5).
Plasma and liver pharmacokinetics after single or chronic dosing.
Single dose. Female CD-1 mice were obtained from Harlan Sprague Dawley and weighed approximately 25-30 g at the time of the study. Animals were administered a 30-mg/kg dose of the unformulated compounds (HBV1580R and HBV1580M) or a 3-mg/kg dose of the formulated HBV263M. Intact siRNA was determined using a hybridization method described below (HBV1580R and formulated HBV263M) or by determining total radioactivity followed by phosphor image analysis of samples separated by PAGE18.
Chronic dosing. Female CD-1 mice were obtained from Harlan Sprague Dawley and weighed approximately 25-30 g at the time of the study. Animals received either a single dose, or multiple doses using the same dosing regimen as that for long-term efficacy studies described below. For the chronic dosing, siRNA was administered as a standard intravenous bolus (100-120 mul) at a dose of approx3 mg/kg into a lateral tail vein on days 1, 2, 3 of week 1 (loading doses), and then day 1 of weeks 2, 3, 4, and 5 (maintenance doses 1-4). Blood was sampled 10 min, 1 and 6 h after each of the three loading doses (week 1) and 10 min, 1 and 6 h, and 3 d after the first and last loading doses (weeks 2 and 5). Animals were euthanized at the appropriate time points by CO2 inhalation followed immediately by exsanguination. Blood was obtained via cardiac puncture and EDTA plasma collected. After exsanguination, animals were perfused with sterile saline. A sample of liver (approx100 mg) was placed in a preweighed homogenization tube and frozen on dry ice.
Quantification of siRNA in plasma and liver samples was done using a sandwich hybridization assay with a working concentration range of 0.027-7.013 ng/ml for the sense strand and 0.026-6.663 ng/ml for the antisense strand. Liver samples were prepared at a concentration of 100 mg/ml in tissue homogenization buffer (3 M guanidine isothiocyanate, 0.5 M NaCl, 0.1 M Tris pH 7.5, 10 mM EDTA). This mixture was homogenized once in a Bio-101 Homogenizer (Savant) with a speed setting of 6.0 and a run time of 10 s. The homogenized liver solutions were diluted to 10 mg/ml in 1 M GITC Buffer (1 M guanidine isothiocyanate, 0.5 M NaCl, 0.1 M Tris pH 7.5, 10 mM EDTA), then used in the assay at further dilution (1:2-1:10). The plasma samples were diluted more than 25-fold in 1 M GITC buffer. Duplex was estimated by doubling the concentration of the sense strand.
WinNonLin Professional (ver 3.3) was used to conduct noncompartmental pharmacokinetic analysis.
In vivo localization of Cy3-labeled siRNA in mouse livers.
Female CD1 mice, 12 weeks of age, were injected intravenously with Cy3-labeled HBV263M, at a dose of 3 mg/kg per mouse. At a series of time points after injection (2 h, 8 h, 1 d, 2 d, 3 d, 6 d and 7 d), livers were harvested from saline-perfused mice. Portions were placed in a 4% paraformaldehyde, 0.5% glutaraldehyde solution for a minimum of 4 h. Before embedding and sectioning, liver pieces were incubated overnight at 4 íC in 30% sucrose. Tissues were embedded in Cryogel (Instrumedics), and cryogenically sectioned. Slides of sectioned livers were mounted with Prolong Antifade Reagent (Molecular Probes), and images were acquired with an Olympus DSU Spinning Disk confocal microscope on a 1 times 81 platform. The acquisition and analysis program used was Slidebook 4.0 v.10.
Analysis of liver toxicity and immune stimulation.
Liver toxicity and immune stimulation studies were conducted by Protiva. We obtained 6- to 8-week-old CD1 ICR mice from Harlan and subjected them to a 3-week quarantine and acclimation period before the experiment. siRNA-SNALPs were administered as an intravenous injection in the lateral tail vein in 0.2 ml PBS for 3-5 s. Blood was collected by cardiac puncture and processed as plasma for cytokine and CBC differential analysis or serum-for-serum chemistries. Blood cell counts and clinical analysis were performed at the Central Laboratory for Veterinarians.
All cytokines were quantified using sandwich ELISA kits. These were mouse interferon-alpha (PBL Biomedical), and mouse IL-6 and TNF-alpha (BD Biosciences).
HBV vector-based mouse model.
To assess the activity of chemically stabilized siRNAs against HBV, we carried out systemic dosing of the siRNA following hydrodynamic injection (HDI)3, 10, 11 of the HBV vector in mouse strain NOD.CB17-Prkdcscid/J (Jackson Laboratory). Female mice were 5-6 weeks of age and approx20 grams at the time of the study. The HBV vector used, pWTD, is a head-to-tail dimer of the complete HBV genome19. For a 20-gram mouse, a total injection of 1.6 ml containing pWTD in saline, was injected into the tail vein within 5 s. A total of 0.3 mug of the HBV vector was injected per mouse. Standard systemic dosing of siRNAs was at 1, 3 or 5 mg/kg/day. To allow recovery of the liver from the disruption caused by HDI, systemic dosing was started 6 d after HDI.
HBV DNA analysis.
Viral DNA was extracted from 50 mul mouse serum using QIAmp 96 DNA Blood kit (Qiagen), according to manufacturer's instructions. HBV DNA levels were analyzed using an ABI Prism 7000 sequence detector (Applied Biosystems). Quantitative real-time PCR was carried out using the following primer and probe sequences: forward primer 5'-CCTGTATTCCCATCCCATCGT-3' (HBV nucleotide 2006-2026), reverse primer 5'-TGAGCCAAGAGAAACGGACTG-3' (HBV nucleotide 2063-2083) and probe FAM 5'-TTCGCA AAATACCTATGGGAGTGGGCC-3' (HBV nucleotide 2035-2062). The psHBV-1 vector, containing the full length HBV genome, was used as a standard curve to calculate HBV copies per ml of serum.