1521-0103/357/2/320–330$25.00 http://dx.doi.org/10.1124/jpet.115.230300THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS J Pharmacol Exp Ther 357:320–330, May 2016Copyright ª 2016 by The American Society for Pharmacology and Experimental Therapeutics

Comparative Characterization of Hepatic Distribution and mRNAReduction of Antisense Oligonucleotides Conjugated withTriantennary N-Acetyl Galactosamine and Lipophilic LigandsTargeting Apolipoprotein Bs

Ayahisa Watanabe, Mado Nakajima, Takeshi Kasuya, Reina Onishi, Naohisa Kitade,Kei Mayumi, Tatsuya Ikehara, and Akira KugimiyaPhysicochemical and Preformulation (A.W.), Bioanalysis (R.O.), and Drug Metabolism and Pharmacokinetics (N.K., K.M.),Research Laboratory for Development, Exploratory Chemistry (M.N.) and Biotechnology-Based Medicine (T.K., T.I, A.K.),Discovery Research Laboratory for Innovative Frontier Medicines, Shionogi & Co., Ltd., Osaka, Japan

Received December 8, 2015; accepted February 22, 2016

ABSTRACTTriantennary N-acetyl galactosamine (GalNAc, GN3) and lipo-philic ligands such as cholesterol and a-tocopherol conjugationsdramatically improve the distribution and efficacy of second-generation antisense oligonucleotides (ASOs) in the whole liver.To characterize ligands for delivery to liver cells based onpharmacokinetics and efficacy, we used a locked nucleic acidgapmer of ASO targeting apolipoprotein B as amodel compoundand evaluated the amount of ASO and apolipoprotein BmRNA inthe whole liver, hepatocytes, and nonparenchymal (NP) cells aswell as plasma total cholesterol after administration of ASOconjugated with these ligands to mice. Compared with uncon-jugated ASO, GN3 conjugation increased the amount (7-fold)and efficacy (more than 10-fold) of ASO in hepatocytes only andshowed higher efficacy than the increased rate of the amount of

ASO. On the other hand, lipophilic ligand conjugations led toincreased delivery (3- to 5-fold) and efficacy (5-fold) of ASO toboth hepatocytes and NP cells. GN3 and lipophilic ligandconjugations increased the area under the curve of ASOs and thepharmacodynamic duration but did not change the half-life inhepatocytes and NP cells compared with unconjugated ASO. Inthe liver, the phosphodiester bond between ASO and theseligands was promptly cleaved to liberate unconjugated ASO.These ligand conjugations reduced plasma total cholesterolcompared with unconjugated ASO, although these ASOs werewell tolerated with no elevation in plasma transaminases. Thesefindings could facilitate ligand selection tailored to liver cellsexpressed in disease-related genes and could contribute to thediscovery and development of RNA interference–based therapy.

IntroductionAntisense oligonucleotides (ASOs) are short synthetic

single-strand nucleotide polymers that are designed to specif-ically hybridize to their target RNA via Watson–Crick basepairing and to prevent expression of the encoded “disease-related” protein product (Yu et al., 2013). Second-generationgapmer ASOs are fully phosphorothioate-modified chimericASOswith a central gap region of 8–14 phosphorothioate DNAnucleotides flanked on either end with 29-modified nucleo-tides, such as two to five 29-O-methoxyethyl RNA (Kimberet al., 2003), locked nucleic acid (LNA) (Grünweller andHartmann, 2007), and amino bridged nucleic acid (Torigoeet al., 2001). ASOs have led to dramatic advances in biologic

stability, which consequently led to longer half-lives (t1/2s) intarget tissues. The gapmer structure efficiently inducescleavage of the target RNA by recruiting endogenous RNaseH as a major pathway to translational inhibition (Grünwelleret al., 2003), resulting in potent pharmacologic actions withfewer side effects. Therefore, ASOs are an emerging class ofdrugs that hold promise for silencing undruggable disease-related genes, thus creating unique opportunities for innova-tive medicines (Sehgal et al., 2013).The liver consists of twomajor types of liver cells: hepatocytes

and nonparenchymal (NP) cells. Disease-related genes causinghyperlipidemia and diabetes are expressed in hepatocytes,which account for a large proportion (approximately 80%) ofliver volume (Sehgal et al., 2013). On the other hand, NP cellsaccount for only a small proportion of liver volume and consist ofseveral cell types, such as sinusoidal endothelial cells, Kupffercells, and hepatic stellate cells, that express disease-relatedgenes causing inflammation (Huang et al., 2013) and fibrosis

This research was supported by Shionogi & Co., Ltd.dx.doi.org/10.1124/jpet.115.230300.s This article has supplemental material available at jpet.aspetjournals.org.

ABBREVIATIONS: Apo-B, apolipoprotein B; ASO, antisense oligonucleotide; AUC, area under the hepatic antisense oligonucleotideamount–time curve; BLQ, below the lower limit of quantification; BSA, bovine serum albumin; Chol, cholesterol; EMCS, 6-maleimidohexanoicacid N-hydroxysuccinimide ester; GN3, triantennary N-acetyl galactosamine; LC-MS/MS, liquid chromatography/tandem mass spectrometry;LNA, locked nucleic acid; NP, nonparenchymal; PCR, polymerase chain reaction; PD, pharmacodynamic; PK, pharmacokinetic; siRNA, smallinterfering RNA; t1/2, half-life; Toc, a-tocopherol.

320

http://jpet.aspetjournals.org/content/suppl/2016/02/23/jpet.115.230300.DC1Supplemental material to this article can be found at:

at ASPE

T Journals on January 9, 2021

jpet.aspetjournals.orgD

ownloaded from

(Fallowfield, 2011). Several studies have been conducted to aidthe development of drug delivery systems to the liver usingligand conjugation. Triantennary N-acetyl galactosamine(GalNAc, GN3) is a high-affinity ligand for the hepatocyte-specific asialoglycoprotein receptor. GN3 conjugation enhancesdelivery of ASO and small interfering RNA (siRNA) to hepato-cytes versus NP cells and increases potency in the whole liver(Biessen et al., 1999; Meade et al., 2014; Nair et al., 2014;Prakash et al., 2014). In addition, conjugations of lipophilicligands such as cholesterol (Chol) and a-tocopherol (Toc) targetendocytosis receptors of the liver via binding of Chol and Toc tolipoproteins and serum albumin (Onishi et al., 2015). Cholconjugation enhances the delivery of ASO and siRNA to NPcells, especially sinusoidal endothelial cells and Kupffer cells,and increases potency in the whole liver (Crooke et al., 1996;Bijsterbosch et al., 2000; Wolfrum et al., 2007). Toc conjugationenhances the delivery of ASOs, siRNA, and DNA/RNA hetero-duplex oligonucleotides to thewhole liver and increases potencyin the whole liver, although its hepatic distribution is unknown(Nishina et al., 2015a,b). Although several studies on hepaticdistribution have been conducted using these ligand conjuga-tions, their efficacies in hepatocytes and NP cells are unknown.In addition, the pharmacokinetic (PK)/pharmacodynamic (PD)durability of ASO conjugated with these ligands in hepatocytesand NP cells was not reported. To promote ligand selectiontailored to the expression cells of disease-related genes, it isimportant to understand the pharmacokinetics and efficacy notonly in thewhole liver but also in hepatocytes andNP cells afteradministration of ASOs conjugated with these ligands.Straarup et al. (2010) reported that a LNA gapmer of ASOs

designed to target an apolipoprotein B (Apo-B) mRNA, whichis important for cholesterol metabolism, remained in the liverfor a long time and improved the plasma total cholesterol levelvia reduction of Apo-B mRNA after administration to mice. Inour study, we used this ASO as a model compound, and wemeasured the amounts of ASO and Apo-B mRNA in the wholeliver, hepatocytes, and NP cells as well as the plasma totalcholesterol after single administration of ASOs conjugatedwith GN3, Chol and Toc to mice. This study aimed to character-ize ligands for delivery to liver cells based on pharmacokineticsand efficacy using ASOs targeting Apo-B.

Materials and MethodsChemicals. ASO was purchased from GeneDesign, Inc. (Osaka,

Japan). The ASO sequence is 59-GCattggtatTCA-39 [gapmer oligonu-cleotide with LNA (capital letters) and DNA (lowercase letters); allinternucleoside linkages are phosphorothioated], which has thehighest affinity and potency for Apo-B mRNA (Straarup et al.,2010). Saline was purchased from Otsuka Pharmaceuticals (Tokyo,Japan). RNAlater was purchased from QIAGEN (Hilden, Germany). TEbuffer (pH 8.0) was obtained from Nippon Gene Co., Ltd. (Tokyo, Japan).Collagenasewas purchased fromSigma-Aldrich (St. Louis,MO). All otherreagents and solvents were commercial products of reagent grade.

Synthesis of ASOs Conjugated with GN3 and LipophilicLigands. ASOs (Fig. 1) were synthesized using standard phosphor-amidite chemistrywith commercially available phosphoramiditemono-mers and postsynthetic modification (Alam et al., 2011). For oligosynthesis, cholesteryl-TEG phosphoramidite (Glen Research, Sterling,VA), Τoc-TEG phosphoramidite (Glen Research), and N-MMTr-aminohexyl linker phosphoramidite (Sigma-Aldrich) were used. Post-synthetically, all ASOs were treated with ethanol/28% aqueousammonium hydroxide/40% aqueous methylamine [1:3:1 (v/v)] at room

temperature for 24 hours for cleavage from the support and removal ofprotecting groups. ASOs were purified with C18 Sep-Pak Cartridges(Waters Inc., Milford, MA). Pure fractions of Chol-ASO and Toc-ASOwere desalted using ultrafiltration membrane. GN3 was preparedaccording to a previously reported procedure with slight modifications(Supplemental Fig. 1). For the conjugation of GN3, purified amino-hexyl functionalized ASO was allowed to react with 6-maleimidohex-anoic acid N-hydroxysuccinimide ester in 50 mM sodium phosphatebuffer (pH 8.0) for 24 hours at room temperature to obtain amaleimido-modified ASO. The reaction mixture was diluted with water andpurified with reverse phase chromatography. The maleimido-modifiedASO was then reacted with the thiol-functionalized GN3 unit in awater/dimethylsulfoxide mixture for 2 hours at room temperature toobtain GN3-ASO. The resulting reaction mixture was diluted withwater and purified by reverse phase chromatography. Pure fractions ofGN3-ASO were desalted using ultrafiltration membrane. The purityand mass of the ASOs was determined using ion-pair liquidchromatography/tandem mass spectrometry (LC-MS/MS) analysis(Table 1).

Animal Treatment. Animal care and all experimental procedureswere performed with the approval of the Institutional Animal Care andUse Committee of Shionogi in terms of the 3R (Replacement/Reduction/Refinement) principle. Male C57BL/6J Jcl mice were purchased at age7 weeks from CLEA Japan Inc. (Osaka, Japan). After quarantine for aweek, the mice were acclimated for several days in the animalcompartment. The mice were used for the experiments at age 8 to 9weeks (bodyweight: 22–24 g). During the acclimation and experimentalperiods, themice were placed under the conditions of room temperatureof 20–26°C, relative humidity of 30%–70%, and light for 12 hours [light(8:00–20:00)/dark (20:00–8:00)] and they were allowed free access to tapwater and solid laboratory food (CE-2; CLEA Japan Inc.).

ASO was dissolved in saline, which was used as the vehicle andadministered intravenously or subcutaneously to mice (n5 3 per timepoint) at single doses under isoflurane anesthesia conditions. In-travenous administration was well suited for evaluating Chol-ASOand Toc-ASO because ASOs conjugated with lipophilic ligands did notwork after subcutaneous administration (data not shown). For theintravenous administration study, unconjugated ASO, Chol-ASO, and

Fig. 1. Structures of ASO conjugated with ligands. (A) ASO conjugatedwith Chol. (B) ASO conjugated with Toc. (C) ASO conjugated with GN3.

Hepatic ASO Delivery Using Ligand Conjugation 321

at ASPE

T Journals on January 9, 2021

jpet.aspetjournals.orgD

ownloaded from

Toc-ASO were administered to mice at 0.2 and 1 mg as unconjugatedASO/kg, respectively. For the subcutaneous administration study,unconjugated ASO was administered to mice at 0.2 and 1 mg asunconjugated ASO/kg, and GN3-ASOwas administered tomice at 0.1,0.2, and 1 mg as unconjugated ASO/kg.

Blood samples were collected via the inferior vena cava with asyringe attached to a needle containing heparin under isofluraneanesthesia at 1, 3, 7, and 14 days after administration. After bloodcollection, whole liver samples were collected, weighed as the totalamount of liver, and flash frozen on dry ice to determine the amount ofASOandApo-BmRNA.The blood sampleswere centrifuged at 1600� gfor 10 minutes at 4°C. The isolated plasma and whole liver sampleswere stored frozen at 280°C until analysis. In addition to the aboveadministration, unconjugated ASO was subcutaneously administeredtomice at a high dose (5 and 10mg as unconjugated ASO/kg), which ledto amaximum effect as described previously (Straarup et al., 2010), andthe plasma and whole liver samples were collected at 1, 3, 7, 14, 21, and28 days after administration.

Protocol for Cell Fractionation Experiments. ASO was ad-ministered to mice under the above-mentioned conditions. Mice wereeuthanized at 1, 3, 7, and 14 days after administration. Mouse liverwas perfused in accordance with a previously reported procedure(Graham et al., 1998; Yu et al., 2001). Briefly, the mouse portal veinwas catheterized under isoflurane anesthesia. The liver was perfusedwith perfusion buffer (10mMHEPES, 142mMNaCl, and 6.7mMKCl,pH 7.4) and the inferior vena cava was cut for drainage. The liver wassubsequently perfused with collagenase and then removed entirelyand diluted in perfusion buffer containing 1% bovine serum albumin(BSA). For separation of hepatocytes and NP cells, the cell suspensionwas spun at 50 � g. The resulting hepatocyte pellet was washed withperfusion buffer containing 1% BSA and spun twice. After a finalwashing, the resulting pellet was used as a sample of hepatocytes. Thesupernatant, fromwhich the hepatocytes had been removed, was spunat 500 � g to pellet all NP cells, and the resulting pellet was used as asample of NP cells. After final pelleting, the hepatocytes and NP cellswere resuspended in perfusion buffer containing 1%BSA. Because thehepatocytes are considerably larger than NP cells, the enrichment cellfactions were confirmed by observing the forward and side scatterdisplay using a flow cytometer according to previous reports (Grahamet al., 1998; Yu et al., 2001; Prakash et al., 2014). Cell purities weremore than 95% free of other contaminating cell types for hepatocytesand were 95% free of hepatocytes for NP cells, respectively. Thedetermine the amount of ASO and Apo-B mRNA, hepatocytes and NPcell samples were collected, weighed as the total amount of liver, flashfrozen on dry ice, and stored frozen at 280°C until analysis.

Determination of the Amount of ASO and Apo-B mRNA inthe Whole Liver, Hepatocytes, and NP Cells and Plasma TotalCholesterol and Transaminases. To determine the amount of ASOconjugated with GN3 and lipophilic ligands, whole liver samples at day 1after administration of 1mg/kgwere examinedbyLC-MS/MS. In thewholeliver samplesatday1, theamountsof intactChol-ASOand Toc-ASOwerealready lower than those of unchanged ASO (Supplemental Fig. 2)and the intact GN3-ASO was not detected (Supplemental Fig. 3).Therefore, we determined the unconjugated ASO, which is an active

form, for the whole liver, hepatocytes, and NP cell samples afteradministration of unconjugatedASO,Chol-ASO,Toc-ASO,andGN3-ASO.

Concentrations of ASO in the whole liver, hepatocytes, and NP cellsamples were determined by LC-MS/MS using API5000 (SCIEX,Framingham City, MA). In brief, whole liver samples (50–100 mgtissue), hepatocytes, and NP cell samples (approximately 0.5 ml)were homogenized in extraction buffer (0.5% Nonidet P-40, 25 mMethylenediaminetetraacetic acid, 100 mM sodium chloride, 25 mMtris hydroxymethyl aminomethane, and 0.5 mg/ml proteinase K, pH8.0) and incubated at 37°C for 0.5 hours. To prepare the standard,the blankwhole liver, hepatocytes, andNP cell samples spikedwith thestandard solutions were homogenized as described above. ASO in theabove homogenate samples was extracted by a liquid–liquid extrac-tion method using phenol/chloroform [1:1 (v/v)], followed by solid-phase extraction (10 mg, Oasis HLB; Waters Inc.). The eluent wasevaporated to dryness under nitrogen and reconstituted in TE buffer(pH 8.0)/methanol [9:1 (v/v)]. Five microliters of reconstituted solutionwas injected into the LC-MS/MS system. Chromatographic separa-tionswere performed at a flow rate of 0.4ml/min on aCadenzaCD-C18HT (2.0 � 50 mm, 3 mm; Imtakt Corporation, Kyoto, Japan) for 4minutes. A binary gradientwas used to perform the separations.Mobilephase A consisted of 10 mM di-isopropylamine and 25 mM 1,1,1,3,3,3-hexafluoro-2-propanol inwater, andmobile phase B consisted of 10mMdi-isopropylamine and 25 mM 1,1,1,3,3,3-hexafluoro-2-propanol inmethanol/acetonitrile (1:1). The column temperature was maintainedat 60°C using a column heater. Mass determination was performed inmultiple reaction monitoring mode using a mass-to-charge ratio of719.9/94.9 for unconjugated ASOs. The calibration standards rangedfrom 0.03 to 30mg/g for whole liver samples and from 0.003 to 3 mg/g forhepatocytes and NP cell samples. The precision and accuracy of thecalibration standards was within 20%.

Total RNAwas extracted fromwhole liver, hepatocytes, andNP cellsamples using the RNeasy 96 Universal Tissue Kit (QIAGEN). Onemicrogram of total RNAwas converted to cDNA using SuperScript IIIFirst-Strand Synthesis SuperMix for quantitative reverse-transcription polymerase chain reaction (PCR) (Invitrogen, Carlsbad,CA), according to the manufacturer’s instructions. Quantitative PCRwas performed with SYBR Premix Ex Taq II (Takara Bio, Otsu,Japan) using an Applied Biosystems 7500 Real-Time PCR System(Applied Biosystems, Foster City, CA). Apo-B mRNA levels werenormalized to glyceraldehyde-3-phosphate dehydrogenase and arepresented relative to the saline control. Plasma total cholesterol wasdetermined by the colorimetric method with Pureauto S CHO-N(Sekisui Medical Co., Ltd., Tokyo, Japan). Alanine transaminaseand aspartate aminotransferase as plasma transaminases were de-termined by the colorimetric method with plasma transaminase CII-test Wako kits (Wako Pure Chemical Industries, Ltd., Osaka, Japan).

Data Analysis. All data representmeans6 S.D. except for PK andPD parameters. Amounts of ASO (micrograms per liver) in the wholeliver, hepatocytes, and NP cell samples were calculated by using theconcentration of ASO and weight of the samples.

PK analysis of the averaged hepatic amount of ASO per time wasperformed with Phoenix WinNonlin 6.0 software (Certara USA,Inc., St. Louis, MO) based on a noncompartment model with uniform

TABLE 1Summary of ASO sequence and analytical data

ASO Motif Sequence (59 to 39) MassCalculated

MassObserved UV Purity

%

Unconjugated ASO 2-8-3 GCattggtatTCA 4324.6 4324.2 96.0Chol-ASO 2-8-3 Chol-GCattggtatTCA 5081.6 5083.0 91.0Toc-ASO 2-8-3 Toc-GCattggtatTCA 5024.6 5025.0 98.0GN3-ASO 2-8-3 GN3-GCattggtatTCA 5024.6 5025.0 98.0

Capital letters indicate LNA nucleotides; lowercase letters indicate DNA nucleotides. The ASOs are fullyphosphorothioated modified, whereas the linkers to ligands are phosphodiesters.

322 Watanabe et al.

at ASPE

T Journals on January 9, 2021

jpet.aspetjournals.orgD

ownloaded from

weighting. The area under the hepatic amount of ASO–time curve (AUC)was calculated by the trapezoidal rule. The elimination t1/2 wascalculated by linear regression using logarithmic values of appropriate

points. The other PK parameters were the maximum hepatic amount ofASO (Cmax) and time to reach the maximum hepatic amount of ASO(Tmax). A hepatic amount of ASO below the lower limit of quantification

Fig. 2. ASO distribution in whole liver, hepatocytes, and NP cells. Mice (n = 3 per group) were injected intravenously with a single administration ofunconjugated ASO, Chol-ASO, and Toc-ASO (0.2 and 1 mg as unconjugated ASO/kg) and subcutaneously with a single administration of unconjugatedASO (0.2 and 1 mg/kg) and GN3-ASO (0.1, 0.2, and 1 mg as unconjugated ASO/kg). Mice were euthanized at day 1 (A–C) and day 3 (D–F) afteradministration. The amounts (micrograms per liver) of ASOwere determined for the whole liver (A and D), hepatocytes (B and E), andNP cells (C and F).All data are expressed as means6 S.D. P values were calculated with the Dunnett test (*P , 0.05; **P , 0.01 versus unconjugated ASO group at samedose and route). iv, intravenously; ND, not detected; sc, subcutaneously.

TABLE 2Differential accumulation of ASO between hepatocytes and NP cells in liverDifferential distribution is given as the percentage of the whole liver. All data are expressed as means 6 SD.

ASO

Differential Distribution in Hepatocytes and NP Cells

Day 1 Day 3

Hepatocytes NP Cells Hepatocytes NP Cells

%

Unconjugated ASO (i.v. dosing) 30 6 5 55 6 2 27 6 7 60 6 9Chol-ASO 35 6 2 55 6 3 27 6 5 67 6 17Toc-ASO 37 6 4 75 6 16 32 6 2 36 6 3Unconjugated ASO (s.c. dosing) 46 6 9 61 6 3 52 6 23 26 6 4GN3-ASO 54 6 3 22 6 2 74 6 2 15 6 5

Mice (n = 3 per group) were injected intravenously with a single administration of unconjugated ASO, Chol-ASO, andToc-ASO and subcutaneously with a single administration of unconjugated ASO and GN3-ASO (1 mg as unconjugatedASO/kg). Mice were euthanized at days 1 and 3 after administration.

Hepatic ASO Delivery Using Ligand Conjugation 323

at ASPE

T Journals on January 9, 2021

jpet.aspetjournals.orgD

ownloaded from

(BLQ) was represented as BLQ. BLQ values were treated as zero. ThePD parameters were as follows: maximum reducing mRNA (percentageof control) in whole liver, hepatocytes, and NP cells; maximum reducingplasma total cholesterol (in milligrams per decaliter) in plasma (PDEmax); and time to reach maximum PD effect (PD Tmax).

TheDunnett testwas used to compare values of the ligand conjugationgroupwith those of unconjugated ASO group at the same dose and route.P values , 0.05 were considered significantly different. Statisticalprobability in figures and figure legends is expressed as *P , 0.05 and**P , 0.01.

ResultsGN3and Lipophilic Ligand Conjugations Enhance

Delivery of ASO to the Liver and Change the Distribu-tion of ASO to Hepatocytes and NP Cells. In the wholeliver, hepatocytes, and NP cells at day 1, the amount of ASO

after intravenous administration of unconjugated ASO wascomparable with after subcutaneous administration of uncon-jugated ASO (Fig. 2, A–C). In the whole liver at day 1, bycomparison between the same doses, the amounts of ASO afteradministration of Chol-ASO and Toc-ASO were 3.5-fold and4-fold higher than after intravenous administration of uncon-jugated ASO (Fig. 2A). Furthermore, the amount of ASO afteradministration of GN3-ASO was 6-fold higher than aftersubcutaneous administration of unconjugated ASO. In thehepatocytes at day 1, the amounts of ASO after administra-tion of Chol-ASO and Toc-ASO were 4-fold and 5-fold higherthan after intravenous administration of unconjugated ASO(Fig. 2B). Furthermore, the amount of ASO after adminis-tration of GN3-ASO was 7-fold higher than after subcutane-ous administration of unconjugated ASO. In the NP cells atday 1, the amounts of ASO after administration of Chol-ASO

Fig. 3. Apo-B mRNA in whole liver, hepatocytes, and NP cells and plasma total cholesterol and transaminases. Mice (n = 3 per group) were injectedintravenously with a single administration of unconjugated ASO, Chol-ASO, and Toc-ASO (0.2 and 1 mg as unconjugated ASO/kg), and subcutaneouslywith a single administration of unconjugated ASO (0.2 and 1mg/kg) and GN3-ASO (0.1, 0.2, and 1mg as unconjugated ASO/kg). Mice were euthanized atday 3 after administration. Reduction of Apo-B mRNA (percentage of control) was determined for the whole liver (A), hepatocytes (B), and NP cells (C).Mouse plasma was assayed for plasma total cholesterol (D), alanine transaminase (E), and aspartate aminotransferase (F). All data are expressed asmeans 6 SD. P values were calculated with the Dunnett test (*P , 0.05; **P , 0.01 versus unconjugated ASO group at same dose and route). ALT,alanine transaminase; AST, aspartate aminotransferase; iv, intravenously; sc, subcutaneously.

324 Watanabe et al.

at ASPE

T Journals on January 9, 2021

jpet.aspetjournals.orgD

ownloaded from

and Toc-ASO were 3-fold and 4-fold higher than after intra-venous administration of unconjugated ASO (Fig. 2C). On theother hand, the amount of ASO after administration of GN3-ASOwas the same or slightly higher than after subcutaneousadministration of unconjugated ASO.As the differential distribution between hepatocytes andNP

cells, the fraction of NP cells after administration of unconju-gated ASO, Chol-ASO, and Toc-ASO was the same or higherthan those of hepatocytes, whereas the fraction of hepatocytesafter administration of GN3-ASO constituted a vast majorityof the whole liver (Table 2). In addition, the accumulation ofASO in the liver after administration of ASO conjugated withGN3 and lipophilic ligands at day 3 showed the same tendencyas that of day 1 (Fig. 2, D–F).GN3and Lipophilic Ligand Conjugations Improve

the Efficacy of ASO in the Liver, Hepatocytes, andNP Cells. In the whole liver, hepatocytes, and NP cells atday 3, the efficacy of ASO after intravenous administration ofunconjugated ASO was comparable with after subcutaneousadministration of unconjugated ASO (Fig. 3, A–C). In the wholeliver, by comparison between different doses (0.2 and 1 mg/kg),the efficacies of ASO after administration of Chol-ASOand Toc-ASO were more than 5-fold higher than after in-travenous administration of unconjugated ASO (Fig. 3A).

Furthermore, the efficacy of ASO after administration of GN3-ASO was more than 10-fold higher than after subcutaneousadministration of unconjugated ASO by comparison betweendifferent doses (0.1 and 1 mg/kg). In the hepatocytes, theefficacies of ASO after administration of Chol-ASO and Toc-ASO were more than 5-fold higher than after intravenousadministration of unconjugated ASO (Fig. 3B). Furthermore,the efficacy of ASO after administration of GN3-ASOwas morethan 10-fold higher than after subcutaneous administration ofunconjugated ASO. In the NP cells, the efficacies of ASO afteradministration of Chol-ASO and Toc-ASO were 5-fold higherthan after intravenous administration of unconjugated ASO(Fig. 3C). On the other hand, the efficacy of ASO afteradministration of GN3-ASO was less than twice that aftersubcutaneous administration of unconjugated ASO.Similar to the mRNA reduction, Chol-ASO and Toc-ASO

showed 5-fold enhancement and GN3-ASO showed 10-foldenhancement for reducing plasma total cholesterol comparedwith unconjugated ASO (Fig. 3D). These ASOs were welltolerated with no elevation in plasma transaminases (Fig. 3,E and F). In addition, the efficacy of ASO after administra-tion of ASO conjugated with GN3 and lipophilic ligands atday 7 showed the same tendency as that of day 3 (Supple-mental Fig. 4).

Fig. 4. Correlation analysis for theamount of ASO and the reduction ofApo-B mRNA in the whole liver, hepa-tocytes, and NP cells and plasma totalcholesterol. Mice (n = 3 per group) wereinjected intravenously with a singleadministration of unconjugated ASO,Chol-ASO, and Toc-ASO (0.2 and 1 mgas unconjugated ASO/kg) and subcuta-neously with a single administration ofunconjugated ASO (0.2 and 1 mg/kg)and GN3-ASO (0.1, 0.2, and 1 mg asunconjugated ASO/kg). Mice were eu-thanized at day 3 after administration.(A–C) The amount of ASO and Apo-BmRNA were assayed in the whole liver(A), hepatocytes (B), and NP cells (C).(D) Mouse plasma was assayed forplasma total cholesterol. The dashedline represents the extrapolation of theobserved unconjugated ASO. All dataare expressed as means 6 S.D. iv,intravenously; sc, subcutaneously.

Hepatic ASO Delivery Using Ligand Conjugation 325

at ASPE

T Journals on January 9, 2021

jpet.aspetjournals.orgD

ownloaded from

PK/PD Analysis of ASO Conjugated With GN3 andLipophilic Ligands. Figure 4 shows the correlation analy-sis between the amount of ASO and efficacy in the whole liver,hepatocytes, and NP cells at day 3. These results indicatedthat the efficacies of Chol-ASO and Toc-ASO improve inproportion to the increase of the amount of ASO as well asunconjugated ASO, whereas the efficacy of GN3-ASO im-proves more than the increased rate of the amount of ASO inthe whole liver and hepatocytes. The efficacy of Apo-BmRNAreduction after administration of GN3-ASO at 1 mg/kgreached the maximal level because it was equal to or higherthan that of unconjugated ASO at a high dose (10 mg/kg),which gave the maximum effect (Supplemental Table 3)(Straarup et al., 2010). In addition, the efficacy of plasmatotal cholesterol after administration of ASO conjugatedwith GN3 and lipophilic ligands at 1 mg/kg reached themaximal level because reduction of Apo-B mRNA expressiondoes not affect high-density lipoprotein, which constitutespart of the plasma total cholesterol, and the maximum effectof the reduction does not lead to zero plasma total cholesterol(Straarup et al., 2010). In NP cells, the efficacies of Chol-ASOand Toc-ASO improved in proportion to the increase in theamount of ASO as well as unconjugated ASO, whereas theefficacy of GN3-ASO did not show a dose dependence onApo-B mRNA reduction.PK/PD Durability of ASO Conjugated With GN3 and

Lipophilic Ligands. After single administration of ASOsconjugated with GN3 and lipophilic ligands at 1 mg/kg, theASOs were slowly eliminated from the whole liver afteradministration of all ASOs (t1/2: 3–5 days; Tmax: day 1) (Fig.5A; Table 3). After the Tmax of ASO, Apo-B mRNA and plasmatotal cholesterol began decreasing, and their PD Tmax valueswere observed at day 3 (Fig. 6A; Table 4). In the hepatocytes,the ASOs were slowly eliminated from the hepatocytesafter administration of all ASOs (t1/2: 3–7 days; Tmax: day 1)(Fig. 5B; Table 3). After the Tmax of ASO, Apo-B mRNA andplasma total cholesterol began decreasing, and their PD Tmax

values were observed at days 3–7 (Fig. 6B; Table 4). In the NPcells, the ASOs were slowly eliminated from the NP cellsafter administration of all ASOs (t1/2: 2–4 days; Tmax: days1–3) (Fig. 5C; Table 3). The durability of the PD effect of Chol-ASO and Toc-ASO was better than that of unconjugated ASO,whereas GN3-ASO showed the same degree of PD effectdurability as unconjugated ASO. After the Tmax of ASO, Apo-BmRNA and plasma total cholesterol began decreasing, and theirPDTmax valueswere observed at days 3–7 (Fig. 6C; Table 4). Thet1/2 of whole liver, hepatocytes, and NP cells for each ASO wasnearly equal. In addition, the elimination and efficacy of ASOafter administration of ASO conjugated GN3 and lipophilicligands at 0.2 mg/kg indicated the same tendency as that of 1mg/kg (Supplemental Fig. 5; Supplemental Tables 1 and 2).

DiscussionIn this study, we evaluated the amount of ASO and Apo-B

mRNA in the whole liver, hepatocytes, and NP cells andrevealed an effect of GN3 and lipophilic ligands on pharma-cokinetics and efficacy. These ligands were characterized ashaving the following characteristics. First, GN3 conjugationenhances the delivery of ASO only to hepatocytes and showshigher efficacy than the increased rate of the amount of ASO.Lipophilic ligand conjugations enhance the delivery of ASO to

both hepatocytes and NP cells, show high efficacy, and lead toan increase in the amount of ASO. Second, GN3 and lipophilicligand conjugations increase the AUC of ASO but do notchange the t1/2 in hepatocytes and NP cells. GN3 conjugation

Fig. 5. The amounts of ASO in the whole liver, hepatocytes, and NP cells.Mice (n = 3 per group) were injected intravenously with a singleadministration of unconjugated ASO, Chol-ASO, and Toc-ASO and subcu-taneously with a single administration of unconjugated ASO and GN3-ASO(1mg as unconjugatedASO/kg).Micewere euthanized at days 1, 3, 7, and 14after administration. The amounts (micrograms per liver) of ASO weredetermined for whole liver (A), hepatocytes (B), andNP cells (C). All data areexpressed as means 6 S.D. iv, intravenously; sc, subcutaneously.

326 Watanabe et al.

at ASPE

T Journals on January 9, 2021

jpet.aspetjournals.orgD

ownloaded from

increases the PD effect duration more than the increased rateof ASO exposure in hepatocytes alone, whereas lipophilicligand conjugation increases the PD effect duration in bothhepatocytes and NP cells.To analyze ASO, we used LC-MS/MS because of its high

sensitivity and selectivity. Hybridization enzyme-linked im-munosorbent assay, fluorescence labeling, and radiolabeling

cannot distinguish full-length ASO (unchanged form) fromits metabolites and degradation products, resulting in over-estimation of the ASO concentration (Yu et al., 2004; Cenet al., 2012). Therefore, we considered LC-MS/MS to beappropriate for investigating the hepatic distribution ofASO. We confirmed that the amounts of intact Chol-ASOand Toc-ASO were already lower than those of unchanged

TABLE 3PK parameters after administration of ASOs in miceValues are given in micrograms per liver for Cmax and in micrograms per liver � day for AUCinf.

ASO Matrix t1/2 Tmax Cmax AUCinf

d mg

Unconjugated ASO (i.v. dosing) Whole liver 3.3 1.0 1.1 5.7Hepatocytes 4.2 1.0 0.3 2.0NP cells 3.0 1.0 0.6 3.1

Chol-ASO Whole liver 4.8 1.0 4.0 32.9Hepatocytes 6.9 1.0 1.4 14.9NP cells 2.7 3.0 2.4 16.9

Toc-ASO Whole liver 4.2 1.0 4.5 36.8Hepatocytes 6.6 1.0 1.7 18.6NP cells 2.6 1.0 3.4 23.8

Unconjugated ASO (s.c. dosing) Whole liver 2.8 1.0 1.3 6.4Hepatocytes 2.7 1.0 0.6 3.1NP cells 3.2 1.0 0.8 3.1

GN3-ASO Whole liver 4.2 1.0 7.2 37.1Hepatocytes 4.2 1.0 3.9 22.0NP cells 3.4 1.0 1.6 7.7

Mice (n = 3/group) were injected intravenously with a single administration of unconjugated ASO, Chol-ASO, andToc-ASO and subcutaneously with a single administration of unconjugated ASO and GN3-ASO (1 mg as unconjugatedASO/kg). Mice were euthanized at 1, 3, 7, and 14 days after administration. The amounts (micrograms per liver) ofASO were determined in whole liver, hepatocytes, and NP cells.

Fig. 6. Duration of Apo-B mRNA in thewhole liver, hepatocytes, and NP cells andplasma total cholesterol. Mice (n = 3 pergroup) were injected intravenously with asingle administration of unconjugatedASO, Chol-ASO, and Toc-ASO and subcu-taneously with a single administration ofunconjugated ASO and GN3-ASO (1 mg asunconjugated ASO/kg). Mice were eutha-nized at days 1, 3, 7, and 14 after admin-istration. (A–C) Reduction of Apo-B mRNA(percentage of control) was determined forthe whole liver (A), hepatocytes (B), and NPcells (C). (D) Mouse plasma was assayedfor plasma total cholesterol. All data areexpressed as mean 6 SD. iv, intravenously;sc, subcutaneously.

Hepatic ASO Delivery Using Ligand Conjugation 327

at ASPE

T Journals on January 9, 2021

jpet.aspetjournals.orgD

ownloaded from

ASO (Supplemental Fig. 2), and intact GN3-ASO was notdetected (Supplemental Fig. 3) in the whole liver samples atday 1 after administration of 1 mg/kg, in agreement withprevious reports on GN3-ASO (Prakash et al., 2014) and Toc-ASO (Nishina et al., 2015b). These results suggested that thephosphodiester bond between the ASO and these ligands ispromptly cleaved to liberate unconjugated ASO in the livercells, and these ligands serve as prodrugs targeting hepato-cytes and/or NP cells. In general, the amount of ASO correctedby cell counting might be imprecise and may affect evaluationof the distribution to hepatocytes and NP cells because amanual method for cell counting is imprecise (Riley et al.,2002). Therefore, we selected weight correction as a methodfor evaluating the amounts of ASO (micrograms per liver) inhepatocytes and NP cells per liver and were able to obtain theamounts of ASOwith less individual differences. We conducteda quantitative reverse-transcription PCR experiment to eval-uate mRNA in hepatocytes and NP cells and confirmed theexpression of Apo-B mRNA in both hepatocytes and NP cellsaccording to previous reports (Pape et al., 1991; Werner et al.,2015). However, Apo-B is generally used as a marker forhepatocytes, and the expression of Apo-B is extremely low inNP cells. It is difficult to demonstrate the expression of mRNAin NP cells using the ASO targeting Apo-B, suggesting that itis an experimental limitation. Additional studies using ASOtargeting as amarker for NP cells are required to demonstratethe expression of mRNA in NP cells. In general, microscopypathology evaluation is a definitive diagnosis for drug-inducedliver injury. In ASO toxicity studies, mice treated with ASOdemonstrated significantly increased plasma transaminases(alanine aminotransferase and aspartate aminotransferase)levels as well as histopathological evidence of liver necrosisand apoptosis (Swayze et al., 2007; Frazier, 2015). In addition,plasma transaminase levels and liver weight are frequently

used to screen for hepatotoxicity when evaluating ASOtoxicity studies (Straarup et al., 2010; Prakash et al.,2014). In this study, ASOs were also well tolerated with noelevation in liver weight or plasma transaminases (Supple-mental Fig. 4J).The amounts of ASO of whole liver at day 1 after intrave-

nous and subcutaneous administration of unconjugated ASOwere comparable to those reported previously (Straarup et al.,2010). The amount of ASO of NP cells after administration ofunconjugated ASO was slightly higher than those of hepato-cytes. In addition, ASO concentration (the amount of ASO perliver volume) in the NP cells was more than 4-fold higher thanthat of hepatocytes because hepatocytes represent a largerproportion (approximately 80%) of the liver. These resultssuggested that the unconjugated ASO preferentially accumu-lates in NP cells to the same degree reported previously(Prakash et al., 2014). As an additional explanation, theamounts (percentage of dose) of the administered dose ofASOs in the liver are shown in Supplemental Table 4.By comparison between different doses, targeted delivery of

ASO to hepatocytes using GN3 improved the efficacy morethan 10-fold. However, to evaluate the efficacy enhancementusing GN3 in detail, it is important to show the ED50 (potency)of each ASO as described previously (Prakash et al., 2014).According to expectation, the amount and efficacy of ASO inthe hepatocytes after administration of GN3-ASO was 7-foldand more than 10-fold higher than those of unconjugated ASOas found for the whole liver, respectively. Interestingly, theefficacy of GN3-ASO was higher than the increased rate of theamount of ASO in the whole liver as well as hepatocytes (Fig.4). The phosphodiester bond and the length effect of the linkerwere required to improve the efficacy of GN3-ASO in the liver.The length of our linker may be acceptable for biologic efficacyalthough it was longer than a previous report (Prakash et al.,2014). These results were consistent with those of the wholeliver reported previously (Prakash et al., 2014), suggestingthat the GalNAc-conjugation strategy is very useful forhepatocyte-specific delivery at a reduced dose. These mecha-nisms were not directly observed in this study, and additionalwork is needed to clarify them. It is important to clarify thedistribution of GN3-ASO at earlier time points. In the previ-ous report, a major metabolite, which had lost three GalNAcmoieties, already accounted for a large percentage in the liver1 hour after administration, and intact GN3-ASO accountedfor a small percentage (Prakash et al., 2014). UnconjugatedASO accounted for a large percentage in the liver 24 hoursafter administration and then maintained a large percentagein the liver. We could not quantitatively evaluate the distri-bution pattern of GN3-ASO–related substances (intact GN3-ASO, amajormetabolite, unconjugatedASO,minormetabolites,which have lost one or two GalNAc moieties) at earlier timepoints because we do not have the metabolite standards. In aprevious report, a radiolabeled GN3-ASO was used to clarifythe distribution pattern of GN3-ASO at earlier time points(Biessen et al., 1999). After administration of the radiolabeledGN3-ASO, the radioactivity rapidly accumulated in the hepa-tocytes. However, the radioactivity cannot relate to theefficacy because it cannot distinguish unchanged ASO (anactive form) from other GN3-ASO–related substances. Ac-cordingly, we used unlabeled ASO and time points suitablefor unconjugated ASO to relate to the efficacy in this study.Apo-B protein and low-density lipoprotein cholesterol are

TABLE 4Pharmacodynamic parameters after administration of ASOs in mice

ASO Matrixa PD Tmax PD Emaxb

d

Unconjugated ASO (i.v. dosing) Whole liver 3 38.9Hepatocytes 7 29.7NP cells 3 43.3Plasma 3 42.7

Chol-ASO Whole liver 3 2.6Hepatocytes 3 2.5NP cells 7 1.9Plasma 3 17.2

Toc-ASO Whole liver 3 3.9Hepatocytes 3 1.8NP cells 7 1.1Plasma 3 15.7

Unconjugated ASO (s.c. dosing) Whole liver 3 31.6Hepatocytes 3 12.8NP cells 3 45.4Plasma 3 44.7Whole liver 3 0.5Hepatocytes 3 0.5NP cells 3 22.2Plasma 3 26.5

Mice (n = 3 per group) were injected intravenously with a single administration ofunconjugated ASO, Chol-ASO, and Toc-ASO and subcutaneously with a singleadministration of unconjugated ASO and GN3-ASO (1 mg as unconjugated ASO/kg).Mice were euthanized at 1, 3, 7, and 14 days after administration.

aReducing mRNA in the whole liver, hepatocytes, and NP cells; reducing plasmatotal cholesterol in plasma.

bReducing mRNA (% of control) in the whole liver, hepatocytes, and NP cells;reducing plasma total cholesterol (milligrams per decaliter) in plasma.

328 Watanabe et al.

at ASPE

T Journals on January 9, 2021

jpet.aspetjournals.orgD

ownloaded from

more relevant measurements of the efficacy of ASO targetingApo-B than total cholesterol. To understand its efficacyaccurately, an additional study is needed with a detailedanalysis of cholesterol.Although lipophilic ligand conjugations increased the deliv-

ery of ASO to the liver, especially NP cells as found in aprevious study (Crooke et al., 1996; Bijsterbosch et al., 2000;Wolfrum et al., 2007; Nishina et al., 2015b), no markedenhancement of efficacy in the whole liver, hepatocytes, andNP cells has been reported. We observed that the efficacyincreased proportionally with the amount of ASO, unlike whatwas found in a previous study (Nishina et al., 2015b). Thephosphodiester bond and the length effect of the linker wererequired to improve the efficacy of the ASO in the liver. In thecase of single-stranded ASO, the length of the linker in thisstudy might be appropriate for effectively cleaving thephosphodiester bond between the ASO and these ligands.The efficacy of the Toc–heteroduplex oligonucleotide wasrecently found to be higher than the increased rate of theamount of ASO in the whole liver (Nishina et al., 2015a).Additional studies of structure-activity relationships arerequired to explore lipophilic ligand conjugations. Regardingthe delivery and efficacy of Chol-ASO and Toc-ASO, it isdifficult to show which of the two is better. Toc-ASO may bebetter than Chol-ASO because Toc is the least toxic and fat-soluble vitamin (Kappus and Diplock, 1992). In contrast withGN3-ASO, the differential distribution rates between hepa-tocytes and NP cells after administration of Chol-ASO andToc-ASO were comparable with unconjugated ASO. Lipo-philic ligand conjugations can enhance delivery to NP cells.NP cells consist of sinusoidal endothelial cells, Kupffer cells,and hepatic stellate cells, which express disease-relatedgenes causing inflammation (Huang et al., 2013) and fibrosis(Fallowfield, 2011). Although our findings represent animportant step in advancing ligand selection tailored to theexpression cells of disease-related genes, additional studiesare required to isolate the sinusoidal endothelial cells andKupffer cells as described previously (Bijsterbosch et al.,2000) and evaluate the efficacy.We selected pharmacological doses (0.1–1 mg/kg) adapted

to the model ASO. We considered that the t1/2 of ASO wasrelatively small because the terminal elimination phase couldnot be adequately evaluated due to limitations of the quanti-fication by LC-MS/MS. In the case of a high dose, the t1/2 ofASOwould be long, as found in our data (10mg/kg) and that ofa previous report (Straarup et al., 2010). In addition, weconsidered that these ligands serve as prodrugs targetinghepatocytes and/or NP cells just after administration of ASOsbecause these ligand conjugations increased the Cmax andAUC but did not change the t1/2. After cleavage of thephosphodiester bond, the stability of the active form (uncon-jugated ASO) was nearly equal in the liver cells, although thet1/2 of hepatocytes showed a long tendency compared with thatof NP cells.PK/PD durability would be informative for understanding

the design of a pharmacological study. The design of a studyusingGN3 conjugation should be based on efficacy because theefficacy of GN3-ASO was higher than the increased rate of theamount of ASO. On the other hand, the design of a study usinglipophilic ligand conjugation should be based on pharmacoki-netics and efficacy because the efficacy of these conjugationswas the same as the increased rate of the amount of ASO.

Lipophilic ligand conjugations are limited to intravenousadministration. On the other hand, GN3 conjugation enablessubcutaneous administration, which facilitates both conve-nience for patients (self-administration) and a reduction ofmedical costs in ASO therapeutics.In conclusion, we demonstrated that GN3 conjugation

enhances the delivery of ASO only to hepatocytes and showsa higher efficacy than the increased rate of the amount ofASO. In contrast with GN3 conjugation, lipophilic ligandconjugations enhance the delivery of ASO to both hepato-cytes and NP cells and show high efficacy equivalent to theincreased rate of the amount of ASO. GN3 and lipophilicligand conjugations increase the AUC of ASO and the phar-macodynamic duration but do not change the t1/2 in hepato-cytes and NP cells compared with unconjugated ASO. In theliver, the phosphodiester bond between ASO and theseligands was promptly cleaved to liberate unconjugatedASO. These findings could facilitate ligand selection tailoredto the expression liver cells of disease-related genes and maycontribute to the successful discovery and development ofRNA interference–based therapy.

Acknowledgments

The authors thankYoichiroNihashi, RyokoOka,KeisukeKusumoto,Kosuke Takemoto, Dr. Masahiro Sakagami, Toru Yanagimoto,Dr. Hiroshi Kamimori, and Shingo Sakamoto (Shionogi & Co.,Ltd.) for support during and discussion related to the experiments.

Authorship Contributions

Participated in research design: Watanabe.Conducted experiments: Watanabe, Kasuya, Onishi, Kitade,

Mayumi.Contributed new reagents or analytic tools: Watanabe, Nakajima,

Ikehara.Performed data analysis: Watanabe, Kasuya, Onishi.Wrote or contributed to the writing of the manuscript: Watanabe,

Nakajima, Kasuya, Onishi, Kugimiya.

References

Alam MR, Ming X, Fisher M, Lackey JG, Rajeev KG, Manoharan M, and Juliano RL(2011) Multivalent cyclic RGD conjugates for targeted delivery of small interferingRNA. Bioconjug Chem 22:1673–1681.

Biessen EA, Vietsch H, Rump ET, Fluiter K, Kuiper J, Bijsterbosch MK, and vanBerkel TJ (1999) Targeted delivery of oligodeoxynucleotides to parenchymal livercells in vivo. Biochem J 340:783–792.

Bijsterbosch MK, Rump ET, De Vrueh RL, Dorland R, van Veghel R, Tivel KL,Biessen EA, van Berkel TJ, and Manoharan M (2000) Modulation of plasma pro-tein binding and in vivo liver cell uptake of phosphorothioate oligodeoxynucleotidesby cholesterol conjugation. Nucleic Acids Res 28:2717–2725.

Cen Y, Li X, Liu D, Pan F, Cai Y, Li B, Peng W, Wu C, Jiang W, and Zhou H (2012)Development and validation of LC-MS/MS method for the detection and quantifi-cation of CpG oligonucleotides 107 (CpG ODN107) and its metabolites in miceplasma. J Pharm Biomed Anal 70:447–455.

Crooke ST, Graham MJ, Zuckerman JE, Brooks D, Conklin BS, Cummins LL, GreigMJ, Guinosso CJ, Kornbrust D, and Manoharan M, et al. (1996) Pharmacokineticproperties of several novel oligonucleotide analogs in mice. J Pharmacol Exp Ther277:923–937.

Fallowfield JA (2011) Therapeutic targets in liver fibrosis. Am J Physiol GastrointestLiver Physiol 300:G709–G715.

Frazier KS (2015) Antisense oligonucleotide therapies: the promise and the chal-lenges from a toxicologic pathologist’s perspective. Toxicol Pathol 43:78–89.

Graham MJ, Crooke ST, Monteith DK, Cooper SR, Lemonidis KM, Stecker KK,Martin MJ, and Crooke RM (1998) In vivo distribution and metabolism of aphosphorothioate oligonucleotide within rat liver after intravenous administration.J Pharmacol Exp Ther 286:447–458.

Grünweller A and Hartmann RK (2007) Locked nucleic acid oligonucleotides: thenext generation of antisense agents? BioDrugs 21:235–243.

Grünweller A, Wyszko E, Bieber B, Jahnel R, Erdmann VA, and Kurreck J (2003)Comparison of different antisense strategies in mammalian cells using lockednucleic acids, 29-O-methyl RNA, phosphorothioates and small interfering RNA.Nucleic Acids Res 31:3185–3193.

Huang H, Chen HW, Evankovich J, Yan W, Rosborough BR, Nace GW, Ding Q,Loughran P, Beer-Stolz D, and Billiar TR, et al. (2013) Histones activate theNLRP3 inflammasome in Kupffer cells during sterile inflammatory liver injury.J Immunol 191:2665–2679.

Hepatic ASO Delivery Using Ligand Conjugation 329

at ASPE

T Journals on January 9, 2021

jpet.aspetjournals.orgD

ownloaded from

Kappus H and Diplock AT (1992) Tolerance and safety of vitamin E: a toxicologicalposition report. Free Radic Biol Med 13:55–74.

Kimber WL, Puri N, Borgmeyer C, Ritter D, Sharov A, SeidmanM, and Ko MS (2003)Efficacy of 2-methoxyethoxy-modified antisense oligonucleotides for the study ofmouse preimplantation development. Reprod Biomed Online 6:318–322.

Meade BR, Gogoi K, Hamil AS, Palm-Apergi C, van den Berg A, Hagopian JC,Springer AD, Eguchi A, Kacsinta AD, and Dowdy CF, et al. (2014) Efficient de-livery of RNAi prodrugs containing reversible charge-neutralizing phosphotriesterbackbone modifications. Nat Biotechnol 32:1256–1261.

Nair JK,Willoughby JL, ChanA, Charisse K, AlamMR,WangQ, HoekstraM, KandasamyP, Kel’in AV, and Milstein S, et al. (2014) Multivalent N-acetylgalactosamine-conjugatedsiRNA localizes in hepatocytes and elicits robust RNAi-mediated gene silencing. J AmChem Soc 136:16958–16961.

Nishina K, Piao W, Yoshida-Tanaka K, Sujino Y, Nishina T, Yamamoto T, Nitta K,Yoshioka K, Kuwahara H, and Yasuhara H, et al. (2015a) DNA/RNA heteroduplexoligonucleotide for highly efficient gene silencing. Nat Commun 6:7969.

Nishina T, Numata J, Nishina K, Yoshida-Tanaka K, Nitta K, Piao W, Iwata R, Ito S,Kuwahara H, and Wada T, et al. (2015b) Chimeric antisense oligonucleotide con-jugated to a-tocopherol. Mol Ther Nucleic Acids 4:e220.

Onishi R, Watanabe A, Nakajima M, Sekiguchi M, Kugimiya A, Kinouchi H, NihashiY, and Kamimori H (2015) Surface plasmon resonance assay of binding propertiesof antisense oligonucleotides to serum albumins and lipoproteins. Anal Sci 31:1255–1260.

Pape ME, Ulrich RG, Rea TJ, Marotti KR, and Melchior GW (1991) Evidence that thenonparenchymal cells of the liver are the principal source of cholesteryl estertransfer protein in primates. J Biol Chem 266:12829–12831.

Prakash TP, Graham MJ, Yu J, Carty R, Low A, Chappell A, Schmidt K, Zhao C,Aghajan M, and Murray HF, et al. (2014) Targeted delivery of antisense oligonu-cleotides to hepatocytes using triantennary N-acetyl galactosamine improves po-tency 10-fold in mice. Nucleic Acids Res 42:8796–8807.

Riley RS, Ben-Ezra JM, Tidwell A, and Romagnoli G (2002) Reticulocyte analysisby flow cytometry and other techniques. Hematol Oncol Clin North Am 16:373–420, vii.

Sehgal A, Vaishnaw A, and Fitzgerald K (2013) Liver as a target for oligonucleotidetherapeutics. J Hepatol 59:1354–1359.

Straarup EM, Fisker N, Hedtjärn M, LindholmMW, Rosenbohm C, Aarup V, HansenHF, Ørum H, Hansen JB, and Koch T (2010) Short locked nucleic acid antisenseoligonucleotides potently reduce apolipoprotein B mRNA and serum cholesterol inmice and non-human primates. Nucleic Acids Res 38:7100–7111.

Swayze EE, Siwkowski AM, Wancewicz EV, Migawa MT, Wyrzykiewicz TK, Hung G,Monia BP, and Bennett CF (2007) Antisense oligonucleotides containing lockednucleic acid improve potency but cause significant hepatotoxicity in animals.Nucleic Acids Res 35:687–700.

Torigoe H, Hari Y, Sekiguchi M, Obika S, and Imanishi T (2001) 29-O,49-C-methylenebridged nucleic acid modification promotes pyrimidine motif triplex DNA forma-tion at physiological pH: thermodynamic and kinetic studies. J Biol Chem 276:2354–2360.

Werner M, Driftmann S, Kleinehr K, Kaiser GM, Mathé Z, Treckmann JW, PaulA, Skibbe K, Timm J, and Canbay A, et al. (2015) All-in-one: advanced prepa-ration of human parenchymal and non-parenchymal liver cells. PLoS One 10:e0138655.

Wolfrum C, Shi S, Jayaprakash KN, Jayaraman M, Wang G, Pandey RK, Rajeev KG,Nakayama T, Charrise K, and Ndungo EM, et al. (2007) Mechanisms and opti-mization of in vivo delivery of lipophilic siRNAs. Nat Biotechnol 25:1149–1157.

Yu RZ, Geary RS, and Levin AA (2004) Application of novel quantitative bio-analytical methods for pharmacokinetic and pharmacokinetic/pharmacodynamicassessments of antisense oligonucleotides. Curr Opin Drug Discov Devel 7:195–203.

Yu RZ, Grundy JS, and Geary RS (2013) Clinical pharmacokinetics of second gen-eration antisense oligonucleotides. Expert Opin Drug Metab Toxicol 9:169–182.

Yu RZ, Zhang H, Geary RS, Graham M, Masarjian L, Lemonidis K, Crooke R, DeanNM, and Levin AA (2001) Pharmacokinetics and pharmacodynamics of an anti-sense phosphorothioate oligonucleotide targeting Fas mRNA in mice. J PharmacolExp Ther 296:388–395.

Address correspondence to: Ayahisa Watanabe, Physicochemical andPreformulation, Research Laboratory for Development, Shionogi & Co., Ltd.,Toyonaka, Osaka 561-0825, Japan. E-mail: ayahisa.watanabe@shionogi.co.jp

330 Watanabe et al.

at ASPE

T Journals on January 9, 2021

jpet.aspetjournals.orgD

ownloaded from