Quantitation of Physiological and Biochemical Barriers to siRNA

(12) Following systemic administration, siRNA has to transit through the circulation, ..... AURC, area under the mRNA KD response curve. ..... D.S.: D...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/molecularpharmaceutics

Quantitation of Physiological and Biochemical Barriers to siRNA Liver Delivery via Lipid Nanoparticle Platform Yan Xu,*,† Mei Ou,† Ed Keough,‡ Jeff Roberts,†,§ Ken Koeplinger,† Mike Lyman,† Scott Fauty,∥ Ed Carlini,† Melissa Stern,† Rena Zhang,† Suzie Yeh,† Elizabeth Mahan,† Yi Wang,‡ Don Slaughter,†,⊥ Marian Gindy,# Conrad Raab,† Charles Thompson,† and Jerome Hochman† †

Pharmacokinetics, Pharmacodynamics, and Drug Metabolism, ‡RNA Discovery Biology, ∥Laboratory Animal Resources, and Pharmaceutical Sciences, Merck Research Laboratories, Merck & Co., Inc., 770 Sumneytown Pike, West Point, Pennsylvania 19446, United States #

S Supporting Information *

ABSTRACT: Effective delivery of small interfering RNA (siRNA) requires efficient cellular uptake and release into cytosol where it forms an active complex with RNAi induced silencing complex (RISC). Despite rapid developments in RNAi therapeutics, improvements in delivery efficiency of siRNA are needed to realize the full potential of this modality in broad therapeutic applications. We evaluated potential physiological and biochemical barrier(s) to the effective liver delivery of siRNA formulated in lipid nanoparticle (LNP) delivery vehicles. The comparative siRNA delivery performance of three LNPs was investigated in rats. They were assembled with either C14- or C18-anchored PEG-lipid(s), cationic lipid(s), and various helper lipid(s) and contained the same siRNA duplex. These LNPs demonstrated differentiated potency with ED50’s ranging from 0.02 to 0.25 mg/kg. The two C14PEG-LNPs had comparable siRNA exposure in plasma and liver, while the C18-PEG-LNP demonstrated a higher plasma siRNA exposure and a slower but sustained liver uptake. RISC bound siRNA within the liver, a more proximal measure of the pharmacologically active siRNA species, displayed loading kinetics that paralleled the target mRNA knockdown profile, with greater RISC loading associated with more potent LNPs. Liver perfusion and hepatocyte isolation experiments were performed following treatment of rats with LNPs containing VivoTag-fluorescently labeled siRNA. One hour after dosing a majority of the siRNA within the liver was associated with hepatocytes and was internalized (within small subcellular vesicles) with no significant cell surface association, indicating good liver tissue penetration, hepatocellular distribution, and internalization. Comparison of siRNA amounts in hepatocytes and subcellular fractions of the three LNPs suggests that endosomal escape is a significant barrier to siRNA delivery where cationic lipid seems to have a great impact. Quantitation of Ago-2 associated siRNA revealed that after endosomal escape further loss of siRNA occurs prior to RISC loading. This quantitative assessment of LNP-mediated siRNA delivery has highlighted potential barriers with respect to endosomal escape and incomplete RISC loading for delivery optimization efforts. KEYWORDS: siRNA, lipid nanoparticle, delivery efficiency, liver, subcellular fractionation



INTRODUCTION

Additionally, duration of the mRNA KD could be days to weeks in animal studies.4,5 Since the chemical properties of siRNAs against different gene targets are relatively consistent, discovery of siRNA against novel targets could be less complex when compared to small molecules. Although siRNA offers long-term promise as a therapeutic modality, several challenges have to be addressed for it to reach its full potential in the clinic, most notably, achieving efficient and safe delivery to the target tissues and cells.6−11 In contrast

The potential to selectively inhibit gene expression with small interfering RNA (siRNA) has led to an explosion of basic research applications as well as enthusiastic pursuit of siRNA as a novel therapeutic modality. Over the past decade at least 21 siRNA drug candidates have been developed for more than a dozen diseases, including viral infection, oncology, metabolic disease, and genetic disorders.1 Interest in siRNA as a new therapeutic approach stems from its ability to catalytically and potently downregulate expression of virtually any protein in mammalian cells based on sequence homology to mRNA transcript. With direct transfection into cells in vitro, the RNAi mechanism is exquisitely potent in knockdown (KD) of targeted mRNA with intrinsic IC50 within the pM range.2,3 © 2014 American Chemical Society

Received: Revised: Accepted: Published: 1424

October 2, 2013 February 24, 2014 March 3, 2014 March 3, 2014 dx.doi.org/10.1021/mp400584h | Mol. Pharmaceutics 2014, 11, 1424−1434

Molecular Pharmaceutics

Article

Table 1. Physical and Compositional Properties of Lipid Nanoparticles (LNPs) lipid components (molar ratio) LNP identifier LNP1 LNP2 LNP3

DLinDMAb (58%) DLinDMAb (58%) lipid Af (58%)

cholesterol (30%) cholesterol (30%) cholesterol (30%)

DSPCc (10%) DSPCc (10%) DSPCc (10%)

C18-PEG-c-DSAd (2%) C14-PEG-DMGe (2%) C14-PEG-DMGe (2%)

siRNA

zeta potentiala (mV)

particle sizea (nm)

ApoB ApoB ApoB

3.2 (0.5) 5.9 (0.5) 6.8 (0.9)

72.2 (0.5) 82.1 (0.7) 77.3 (0.3)

Mean (SE), n ≥ 3. bDLinDMA, 1,2-dilinoleyloxy-N,N-dimethyl-3-aminopropane. cDSPC, 1,2-distearoyl-sn-glycero-3-phosphocholine. dC18-PEGc-DSA, poly(ethylene glycol)2000-distearyloxypropylamine. eC14-PEG-DMG, poly(ethylene glycol)2000-dimyristoylglycerol. flipid A, structure not disclosed. a

omeU;omeU;omeU;fluA;fluA;omeC;fluA;fluA;omeU;omeU;omeC;omeC;omeU;fluG (VivoTag680);fluA;fluA;fluA;omeU;dTsdT-iB. Nucleotide modifications are represented as deoxy (d), 2′-fluoro (flu), 2′-O-methyl (ome), and ribo (r), with abbreviations immediately preceding the modified base. Phosphorothioate linkage replacing phosphodiester bond is represented by a subscript of ‘s’ between the two linked nucleotides. The passenger strand is blocked with an inverted abasic nucleotide (iB) on both the 5′ and the 3′ ends. A fluorochrome VivoTag680 (PerkinElmer, Waltham, MA) is coupled to the nucleotide at position 15 in the guide strand (GS) to facilitate siRNA visualization by fluorescence imaging. Table 1 summarized the composition and physical properties of the three selected LNPs. They were assembled in a similar manner as described in ref 13. Cholesterol and 1, 2-distearoylsn-glycero-3-phosphocholine (DSPC) were purchased from Northern Lipids Inc. (Burnaby, BC, Canada). Poly(ethylene glycol)2000-dimyristoylglycerol (PEG-DMG) and poly(ethylene glycol)2000-distearyloxypropylamine (PEG-c-DSA) was manufactured by NOF Corporation (Kanagawa, Japan). Cationic lipids of 1,2-dilinoleyloxy-N,N-dimethyl-3-aminopropane (DLinDMA)14 and lipid A were synthesized at Merck (West Point, PA). In Vivo Studies. Female Sprague−Dawley rats, obtained from Charles River Laboratories, were approximately 7−8 weeks old at the time of study (body weight of ∼200 g). All studies were performed in Merck Research Laboratories’ AAALAC-accredited West Point (PA) animal facility using protocols approved by the Institutional Animal Care and Use Committee. siRNA-LNPs were tail-vein injected to rats at a volume of 1.0 to 1.5 mL with an infusion rate of approximately 3 mL/min. Following LNP administration rats were placed in cages with normal diet and water ab libitum. At terminal necropsies (48 h post dose for the dose−response study, and 5 min, 30 min, 1 h, 4 h, 24 h, 72 h, and 168 h post dose for the kinetic study), rats were euthanized by CO2 asphyxiation followed by exsanguination. Biopsy punches were collected from the right medial lobe of the liver, placed in RNALater (Ambion, Grand Island, NY), and stored at 4 °C for mRNA quantification, or flash-frozen and stored at −80 °C for siRNA quantification. In a separate group of jugular vein cannulated rats, blood was sequentially collected from the jugular vein into EDTA tubes at 2 min, 5 min, 15 min, 30 min, 1 h, 2 h, 4 h, 6 h, and 24 h post dose. Plasma was obtained from the supernatant post 3000g centrifugation of the blood samples for 10 min at 4 °C (Eppendorf 5810R) and stored at −80 °C for siRNA measurement. Rat Liver Perfusion and Hepatocyte Isolation. Liver perfusion and hepatocyte isolation were conducted as described by Quistorff et al.15 with some modifications. Briefly, rats were anesthetized with ketamine at 1 or 24 h following a single iv administration of siRNA-LNP through the tail vein. A small

to most freely diffusible small molecules for which rapid equilibrium between the systemic circulation and the drug target enables pharmacological engagement, siRNA requires a delivery vehicle and/or targeting ligand to facilitate its effective delivery to the site of action.12 Following systemic administration, siRNA has to transit through the circulation, distribute to the appropriate organ/tissues, cross the vascular endothelium, and permeate the extracellular matrix to be taken up by the target cell population in a manner that enables release of siRNA into the cytosol in sufficient quantities to allow the formation of the active RNA induced silencing complex (RISC). Since siRNA is a highly charged hydrophilic macromolecule (∼14 kDa polyanion), it is not capable of passively permeating cell membranes. Consequently siRNA delivery entails internalization by the target cells followed by endosomal−lysosomal trafficking eventually leading to escape to the cytosol or the degradation of siRNA (and/or vehicle) in the acidic lysosomal compartment. Facilitating siRNA escape from endosomes and lysosomes using delivery vehicles prior to degradation represents one key strategy to enhance its effective cytosolic delivery upon internalization. One potential hurdle to the successful improvement in siRNA delivery efficiency is the lack of appropriate techniques with which to quantitatively study the distribution and subcellular trafficking of siRNA to elucidate barriers that may limit its effective delivery. To address this, we investigated the siRNA liver delivery following single intravenous (iv) bolus administrations of three lipid nanoparticle (LNP) vehicles, which had differentiated potency, in rats. To quantify the physiological and biochemical delivery barrier(s), we measured the siRNA pharmacokinetics (PK) in plasma and liver, evaluated siRNA uptake into the target cells by isolating hepatocytes following liver perfusion, and assessed the subcellular trafficking of siRNA in hepatocytes by confocal fluorescence microscopy and subcellular fractionation (i.e., differential centrifugation followed by a density gradient separation). Characterization of the siRNA delivery barriers and elucidation of the underlying mechanisms would facilitate a better design of siRNA sequences and delivery systems and contribute to the discovery and development of improved RNAi-based therapy.



EXPERIMENTAL SECTION Preparation of siRNA-Lipid Nanoparticles (LNPs). Chemically modified Apo-lipoprotein B (ApoB) siRNAs, targeting the ApoB mRNA transcript (Accession # NM 019287), were synthesized at Merck (Rahway, NJ). The primary sequence (presented in the 5′ to 3′ direction) and the chemical modification pattern of the ApoB siRNA duplex are shown as follows: Guide Strand, rA;srU;srU;someU;omeC;fluA;fluG;fluG;fluA;fluA;omeU;omeU;fluG;fluU;omeU;fluA;fluA;fluA;fluG;omeU;someU; Passenger Strand, iB;omeC;1425

dx.doi.org/10.1021/mp400584h | Mol. Pharmaceutics 2014, 11, 1424−1434

Molecular Pharmaceutics

Article

Quantification of siRNA in Plasma and Tissue. siRNA concentrations in the plasma, liver, hepatocyte, and subcellular fractionations were determined using stem-loop RT-PCR as described in ref 5. Liver and hepatocyte samples were homogenized in 500 μL of Qiazol lysis reagent (Qiagen, Valencia, CA) per 50 mg liver punch or 20E6 hepatocyte pellets by a Genogrinder2000 (Spex Certiprep Inc., Metuchen, NJ) at 1300 strokes/min for 5 min at 4 °C. Samples were subsequently incubated at 37 °C for 30 min. Aliquots of the tissue homogenate, plasma or subcellular fractions were diluted 1:100 to 1:1000 in Tris-TE pH 8.0 buffer. The RT primer sequence for the ApoB siRNA was 5′-GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACAACTTTA-3′. The annealing step was performed in a 16 μL reaction containing 125 nM stem-loop RT primer and 10 μL of diluted tissue/plasma samples. The reaction mixture was incubated at 95 °C for 5 min, 80 °C for 2 min, 70 °C for 2 min, 60 °C for 2 min, and 45 °C for 2 min with a 4 °C hold thereafter. RT-PCR was performed using the TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems) per the product protocol with 0.5 and 8 μL per reaction of MultiScribe RT enzyme and RT-primer annealed tissue/plasma sample respectively. Real-time qPCR was performed in a 20 μL reaction volume using 7900 Fast Real-Time PCR System (Applied Biosystems). The forward primer sequence was 5′GGCGCGATTTCAGGAATTGTT-3′ (1.5 μM final concentration), and the reverse primer sequence was 5′-AGTGCAGGGTCCGAG-3′ (0.75 μM final concentration). The probe sequence was 5′-6FAM-CTGGATACGACAACTTTAACAMGBNFQ-3′ (200 nM final concentration). Primers were obtained through Sigma and probes through Applied Biosystems. siRNA concentrations were calculated according to the linear regression between Ct value and log siRNA concentration from standards generated in corresponding matrix. Quantitation of siRNA Associated with RISC. The Ago2 immunoprecipitation assay was performed to quantify RISC associated siRNA as described in ref 17 with minor modifications. In brief, frozen liver samples were homogenized in 100 μL of 0.5% Triton X-100 lysis buffer per 50 mg of tissue using a Genogrinder2000 at 1200 strokes/min for 2 min at 4 °C. The Ago2 pull-down from the liver lysate was immunoprecipitated with anti-rat Ago2 antibody (Wako Laboratory Chemicals, Richmond, VA) and Dynabeads Protein G (Invitrogen, Carlsbad, CA). The amounts of GS siRNA and miR16 that coimmunoprecipitated with Ago2 were quantified by stem-loop RT-PCR as described above. To determine the amount of miR16 bound to Ago2, serial dilutions of synthetic RNA oligonucleotides representing miRNAs including miR16 (mirVana miRNA reference panel V9.1, Applied Biosystems/ Ambion, Austin, TX) were spiked into the Cell-to-Ct lysis and stop buffer (Applied Biosystems) used in the final elution step during Ago2 immunoprecipitation. Concentrations of siRNA and miR16 in each sample were determined by calculating the linear regression between Ct value and log siRNA concentration from siRNA standards. Immunoprecipitation efficiency could be variable depending on the assay conditions (not uncommon for the immunoprecipation assays). To minimize the batch-to-batch variations, Ago2-siRNA levels were normalized to endogenous miR16 associated Ago2 and presented as copy number ratio of siRNA GS to miR16. Nonspecific siRNA binding was also evaluated with normal rat

piece of the caudate lobe of the liver (∼0.2−0.4 g) was removed, and the incision site was tightly ligated to prevent leakage and to minimize excessive bleeding during perfusion. Liver was then perfused with a perfusion medium having physiological pH/osmolarity to remove siRNA-LNP trapped in blood vessels as well as that associated with blood cells, then a heparinase (Sigma, St. Louis, MO) containing buffer to loosen the electrostatic charge interaction between siRNA-LNP and extracellular matrix (e.g., positively charged heparin sulfate), followed by a collagenase containing medium to digest liver tissues. Hepatocytes were then isolated by sequential low-speed centrifugations and Percoll (Sigma) washes. The cells were resuspended in Williams E medium containing 4 mM Lglutamine (Sigma) to assess cell viability and yield under microscopy. Generally over 100−200 million rat hepatocytes with cell viability greater than 80−90% were obtained. The excised liver sections and the perfusates collected were flashfrozen and stored at −80 °C until further analysis. Hepatocyte Homogenization, Differential Centrifugation, and Density Gradient Separation. Fresh hepatocytes of ∼10−15 million cells/mL were homogenized in ice-cold homogenization medium (0.25 M sucrose, 1 mM EDTA, 10 mM Tris-HCl in 1× PBS buffer without Ca2+ and Mg2+, pH 7.4) with 1× Proteinase Inhibitor Cocktail (Roche, Indianapolis, IN) using a Dounce homogenizer. Differential centrifugation was conducted for the hepatocyte homogenates as previously described with minor modifications.16 Cell debris and nuclei were removed from the crude hepatocyte homogenate by a low speed centrifugation (1000g for 10 min, 4 °C, Eppendorf 5810R). The resulting post nuclear supernatant (PNS) was centrifuged at 20000g for 30 min (4 °C, Eppendorf 5417R) to pellet out the large intact organelles (P20). Supernatant collected at the 20000g spin (S20) was further centrifuged at 100000g for 60 min (4 °C, Beckman Optima MAX Ultracentrifugation, fixed angle rotor TLA110). The final supernatant (S100) comprises cytosolic soluble components, and the pellet (P100) contains microsomal fraction and some free membrane fragments. A 5−40% OptiPrep (Sigma) gradient was prepared using a GradientMixer (BioComp Instruments, Fredericton, NB, Canada) per the manufacturer’s instruction. A large intact organelle mixture (P20) obtained from the 20000g spin was loaded onto the OptiPrep gradient and centrifuged at 90000g for approximately 20 h at 4 °C using a SW28 swing bucket rotor (Beckman Optima LE-BOK). Following centrifugation, twelve gradient fractions (1.0 mL/each) were collected with a GradientStation (BioComp Instruments) and measured for density by an ISSYS FC6 density meter (ISSYS: Integrated Sensing Systems, Ypsilanti, MI) within 4 h of collection. The remaining hepatocyte homogenate and the subcellular fractionations were flash-frozen and stored at −80 °C until further analysis. Quantification of mRNA in Liver. ApoB mRNA levels in liver were measured by relative quantitative RT-PCR as previously described.4 The PCR reaction was performed on an ABI 7500 instrument with a 96-well FastBlock. The ApoB mRNA levels were normalized to the concentrations of the housekeeping PPIB (NM011149) mRNA, which were determined by RT-PCR using a commercial probe set (Applied Biosystems, Cat. No. Mm00478285_m1). Results were expressed as the ratio of ApoB mRNA/PPIB mRNA. All mRNA data post LNP treatments were further normalized relative to the PBS vehicle control. 1426

dx.doi.org/10.1021/mp400584h | Mol. Pharmaceutics 2014, 11, 1424−1434

Molecular Pharmaceutics

Article

Cell Signaling Technology) by enhanced chemiluminescence (Super Signal West Femto Substrate solution, Pierce) using a FluorChem8900 imaging system (ProteinSimple, Santa Clara, CA). Ex Vivo Fluorescence Microscopy. Isolated primary hepatocytes in hepatocyte maintenance media cocktail (Invitrogen, Carlsbad, CA) were evaluated for cell viability using Trypan blue exclusion. Hepatocyte suspensions were transferred to a 384-well Aurora imaging plate (Nexus Biosystems, Poway, CA) at 25,000 cells per well and stained with 5 μg/mL Hoechst 33342 dye (Invitrogen) for 20 min. The plate was briefly centrifuged and then immediately imaged using an OPERA confocal microscope (PerkinElmer, Waltham, MA) equipped with a 40× water immersion objective (0.90 numerical aperture) and 60× water immersion objective (1.20 numerical aperture), a digital CCD camera (PCO Imaging, SensiCam QE), and an OPERA CN/QEHSTM software (version 1.8.1). For 40× imaging, a single image plane was captured through the central region of the live cell and Hoechst and VivoTag680 dye channels were detected in 8 image locations per sample. For 60× imaging, a z-stack image series of 14 planes separated by 1.5 μm distance was captured in 6 locations per sample. 60× z-stack images were reconstructed into a 3D image using Fast Deconvolution with a Point Spread Function in Volocity v5.4.1 software (PerkinElmer). Data Analysis. In vivo siRNA pharmacokinetic data in rat plasma and liver were analyzed by a Non-Compartmental Analysis using WinNonlin 5.2.1 (Pharsight Corporation). Plasma siRNA results were also analyzed with a twocompartmental model to estimate the half-lives in α and β phases respectively. Levels of siRNA in liver, hepatocyte, perfusate, subcellular fraction, and RISC complex were converted to percent injected siRNA dose (% ID) (Supporting Information) S1 based on reported physiological parameters in rodents.18

IgG pull down (Santa Cruz Biotechnology, Santa Cruz, CA) and was found to be insignificant (data not shown). Quantitation of Cationic Lipid in Plasma and Tissue. The cationic lipid concentrations in plasma, liver, hepatocyte, and subcellular fractionation samples were determined by LC/ MS−MS in positive ion mode post protein precipitation. Briefly, frozen liver and other tissues were weighed into 20 mL vials and homogenized in a ratio of nine volumes of water per one gram of tissue in the presence of 5−10 steel beads using a GenoGrinder2000 at a minimum of 1300 strokes/min for 5−10 min. A 50 μL aliquot of tissue homogenate, plasma, or subcellular fraction sample was then mixed with 40 μL of diluent (40% methanol/60% n-propanol/10 mM ammonium acetate/0.1% formic acid), 25 μL of internal standard (a historic cationic lipid analogue) in diluent, and 400 μL extraction/ protein precipitation solvent (50:50 methanol:acetonitrile w/ 0.1% formic acid) in a 96-well precipitation plate. The plate was sealed tightly, vortexed, and centrifuged to sediment precipitated proteins. A volume of 150 μL of each supernatant was then transferred to separate wells of a 96-well injection plate, and 5 μL of sample was directly injected and analyzed by LC/MS−MS. Absolute quantification versus standards prepared and extracted from corresponding matrix was performed using a Transcend LX2Multiplexed UPLC with 20 μL injection loops (Thermo Scientific/CTC) coupled to an API 4000 triple quadrupole mass spectrometer interfaced via the Sciex electrospray ion source (ESI) (Applied Biosystems). For each run, a total of 5 μL of sample was injected onto a BDS Hypersil C8 HPLC column (Thermo, 50 × 2 mm, 3 μm) at room temperature. A dual eluent system was used: 95% H2O/5% methanol/10 mM ammonium acetate/0.1%formic acid (A) and 40% methanol/60% n-propanol/10 mM ammonium acetate/ 0.1% formic acid (B). The flow rate was 0.4 mL/min, and the gradient elution profile was as follows: hold at 50% A for 0.25 min, ramp to 0% A at 1.3 min, hold at 0% A for 2.5 min, and then hold at 50% A for 1.75 min. Total run time was 5.5 min. API 4000 source parameters were as follows: CAD, 4; CUR, 20; GS1, 65; GS2, 35; IS, 4000; TEM, 550 °C; CXP, 15; DP, 60; and EP, 10. The assay monitored the precursor ion of each analyte and the corresponding product fragment ion. Protein Measurement and Western Blot Assay. Protein levels in the hepatocyte homogenate and subcellular fractions were determined by a BCA assay per the manufacturer’s protocol (Pierce, Rockford, IL). Western blot assay was conducted for the P20 density gradient fractions to evaluate the enrichment of suborganelle in the gradient fractions. Briefly, samples from the twelve gradient fractions were prepared in NuPAGE LDS sample buffer and NuPAGE reducing agent according to the manufacturer’s instruction (Invitrogen, Grand Island, NY). These samples were denatured by heating for 10 min at 90 °C and resolved by electrophoresis on a 4%−12% NuPAGE Bis-Tris Gel (Invitrogen). MagicMark XP (Invitrogen) was used as protein standard. The polypeptides were transferred to PVDF membranes (Invitrogen, 0.2 μm pore size) using a wet-transfer apparatus (Invitrogen). The membrane was blocked by 10% nonfat milk in TBST buffer containing 0.5% Surfact-Amps20 (Pierce) at 4 °C overnight, and then probed with 1:1000 dilution of primary antibodies against EEA1 (early endosome marker, Cell Signaling Technology, Danvers, MA) and mannose-6-phosphate receptor (late endosome marker, AbCam, Cambridge, MA), respectively, for 1 h at room temperature. The resultant immune complexes were visualized with HRP-conjugated secondary antibodies (1:2000 dilution,



RESULTS In Vivo Potency and Pharmacokinetics of Selected LNPs in Rat. Table 1 summarizes the composition and physical characterization of three LNPs evaluated in this study. Their dose−response (mRNA KD) curves are presented in Figure 1. Potency of the three selected LNPs (mRNA KD) spanned a 10-fold range (ED50 ranging 0.022−0.25 mg/kg) in rat, with highest potency observed for LNP3, followed by LNP2 and LNP1. Since the encapsulated siRNA, assembly protocol, size, and other physical properties of the nanoparticles are comparable for the three assemblies, potency differences are proposed to reflect the impact of LNP vehicle composition on the siRNA delivery efficiency to the target. In the case of LNP1 and LNP2, the 2-fold differences in their ED50 (0.25 mg/kg for LNP1 and 0.12 mg/kg for LNP2) reflect variations in the PEGlipid acyl chain length (C18-PEG-DSA in LNP1 versus C14PEG-DMG in LNP2). In earlier in-house studies changes in acyl chain length led to differences in kinetics of PEG-lipid dissociation from the LNP particles, where there was a rapid loss of C14-PEG-DMG from LNP (∼50% loss in 15 min) in 37 °C serum incubation while the C18-PEG-DSA was stably associated with the LNP for at least 4 h (Supporting Information S2).19 In the case of LNP2 versus LNP3 (ED50 of 0.12 and 0.022 mg/kg respectively) differences in cationic lipid (DLinDMA versus lipid A) would drive the differences in potency. 1427

dx.doi.org/10.1021/mp400584h | Mol. Pharmaceutics 2014, 11, 1424−1434

Molecular Pharmaceutics

Article

ically active siRNA,5 displayed RISC loading kinetics that paralleled the target mRNA KD kinetic profile, with greater RISC loading associated with LNPs of higher potency (Table 2, Figure 2C,D). siRNA Delivery Efficiency to Liver and Hepatocyte Following LNP Administration. At 1 and 24 h following single iv administration of 0.25 mg/kg LNP, rat liver was perfused in situ, followed by tissue digestion and hepatocyte isolation. Quantification of siRNA in the perfusion and wash solutions indicates that a majority of siRNA retained in the liver was associated with hepatocytes, with 100− 150 nm versus