Amphoteric liposomes enable systemic antigen-presenting cell-directed delivery of CD40 antisense and...

ARTHRITIS & RHEUMATISMVol. 60, No. 4, April 2009, pp 994–1005DOI 10.1002/art.24434© 2009, American College of Rheumatology

Amphoteric Liposomes Enable SystemicAntigen-Presenting Cell–Directed Delivery of CD40 Antisense

and Are Therapeutically Effective in Experimental Arthritis

Evangelos Andreakos,1 Una Rauchhaus,2 Athanassios Stavropoulos,1 Gerold Endert,2

Volkmar Wendisch,3 Amina S. Benahmed,4 Stavros Giaglis,1 James Karras,5 Sam Lee,5

Hans Gaus,5 C. Frank Bennett,5 Richard O. Williams,6 Paschalis Sideras,1 and Steffen Panzner2

Objective. Mediation of RNA interference by oli-gonucleotides constitutes a powerful approach for thesilencing of genes involved in the pathogenesis of in-flammatory disease, but in vivo application of thistechnique requires effective delivery to immune cellsand/or sites of inflammation. The aim of the presentstudy was to develop a new carrier system to mediatesystemic administration of oligonucleotides to rheuma-toid arthritis (RA) joints, and to develop an antisenseoligonucleotide (ASO)–based approach to interfere withCD40–CD154 interactions in an experimental model ofRA.

Methods. A novel liposomal carrier with ampho-

teric properties, termed Nov038, was developed andassessed for its ability to systemically deliver an ASOdirected against CD40 (CD40-ASO). Male DBA/1 micewith collagen-induced arthritis were treated withNov038-encapsulated CD40-ASO, and the effects oftreatment on various parameters of disease activity,including clinical score, paw swelling, lymph node re-sponses, and inflammatory cytokine production in thejoints, were assessed.

Results. Nov038 was well tolerated, devoid ofimmune-stimulatory effects, and efficacious in mediat-ing systemic oligonucleotide delivery to sites of in-flammation. In mice with collagen-induced arthritis,Nov038 enabled the therapeutic administration ofCD40-ASO and improved established disease, whileunassisted CD40-ASO was ineffective, and anti–tumornecrosis factor � (anti-TNF�) treatment was less effec-tive in this model. Nov038/CD40-ASO efficacy was at-tributed to its tropism for monocyte/macrophages andmyeloid dendritic cells (DCs), resulting in rapid down-regulation of CD40, inhibition of DC antigen presenta-tion, and reduction in collagen-specific T cell responses,as well as decreased levels of TNF�, interleukin-6(IL-6), and IL-17 in arthritic joints.

Conclusion. Amphoteric liposomes represent anovel carrier concept for systemic and antigen-presentingcell–targeted oligonucleotide delivery with clinical ap-plicability and numerous potential applications, in-cluding target validation in vivo and inflammatorydisease therapeutics. Moreover, Nov038/CD40-ASOconstitutes a potent alternative to monoclonal antibody–based approaches for interfering with CD40–CD40Linteractions.

Rheumatoid arthritis (RA) is a chronic inflam-matory disease of unknown etiology that is characterized

Supported by the federal government of Saxony Anhalt (grant0504/00125 to Novosom AG). Dr. Andreakos’ work was supported bythe Greek Secretariat for Research and Technology (grant 05AKMON105).

1Evangelos Andreakos, PhD, Athanassios Stavropoulos, BSc,Stavros Giaglis, PhD, Paschalis Sideras, PhD: Biomedical ResearchFoundation, Academy of Athens, Athens, Greece; 2Una Rauchhaus,PhD, Gerold Endert, PhD, Steffen Panzner, PhD: Novosom AG,Halle, Germany; 3Volkmar Wendisch, PhD: Chiroblock GmbH,Wolfen, Germany; 4Amina S. Benahmed, PhD: Coley PharmaceuticalGmbH, Dusseldorf, Germany; 5James Karras, PhD (current address:Altair Therapeutics, San Diego, California), Sam Lee, PhD, HansGaus, PhD, C. Frank Bennett, PhD: ISIS Pharmaceuticals, Carlsbad,California; 6Richard O. Williams, PhD: Imperial College of Science,Technology and Medicine, London, UK.

Drs. Rauchhaus, Endert, and Panzner are coinventors ofNov038 and Nov038/CD40-ASO; all benefits from the patent of thisinvention are assigned to Novosom AG. Drs. Karras and Bennett ownstock or stock options in ISIS Pharmaceuticals. Dr. Panzner owns stockof stock options in Novosom AG.

Address correspondence and reprint requests to EvangelosAndreakos, PhD, Biomedical Research Foundation, Academy ofAthens, 4 Soranou Ephessiou Street, 11527 Athens, Greece (e-mail:[email protected]); or to Steffen Panzner, PhD, NovosomAG, Weinbergweg 22, 06120 Halle, Germany (e-mail: [email protected]).

Submitted for publication June 30, 2008; accepted in revisedform January 7, 2009.

994

by synovial hyperplasia, accumulation of inflammatorycells such as macrophages, T cells, B cells, and plasmacells, and progressive destruction of joint structures.Major progress in understanding the pathogenesis ofRA and in designing new therapeutic strategies wasrealized with the development of monoclonal antibody(mAb) technology for the selective targeting of inflam-mation mediators such as tumor necrosis factor �(TNF�), interleukin-1 (IL-1), and IL-6 (1). Previousstudies demonstrated the pivotal role of these cytokinesin both the inflammatory and the tissue-destructiveprocesses in RA, and their discovery has revolutionizedthe way in which RA patients are treated today (2).However, there are still patients whose condition doesnot respond to these new therapeutic options, andcertainly there is still no cure for this disease. A newtherapeutic approach that is being intensively pursued ismodulation of T cell function through the disruption ofcostimulatory molecule interactions such as CD28–CD80/86 and CD40–CD40L between T cells andantigen-presenting cells (APCs).

In the clinic, the use of a fusion protein betweenCTLA-4 and human IgG1-Fc (CTLA-4Ig; abatacept)that blocks CD28–CD80/86 interactions demonstratedsignificant efficacy in treating RA patients, even inpatients whose condition was refractory to anti-TNF�treatment; this fusion protein has recently been ap-proved as a second-line therapy for RA (2). In contrast,the use of a CD40L mAb resulted in unexpected throm-boembolic events attributed to the mAb-based ap-proach, which halted its clinical development (3,4). Thiswas despite its early efficacy demonstrated in the treat-ment of patients with systemic lupus erythematosus(SLE). Thus, alternative strategies to target CD40–CD40L interactions in vivo are being pursued.

A new, highly promising approach for the selec-tive silencing of genes is RNA interference (RNAi)technology mediated by antisense DNA or small inter-fering RNA (siRNA) oligonucleotides, which providesan alternative to mAb strategies and which is applicableto intracellular targets (5). Despite its potential forcontributing to the understanding of disease mecha-nisms and the development of novel therapeutics, RNAihas only rarely been used in RA or inflammatorydiseases in general. This is due to the lack of effectivedelivery technologies for the systemic and targeteddelivery of oligonucleotides to inflammatory diseasecells or sites in vivo. To facilitate systemic delivery ofoligonucleotides, numerous approaches have beentested, mostly for liver targets, and the systems thatprevail today are based on cationic liposomes or poly-

mers. Their development has been helped by the easyand reversible complex formation between oligonucleo-tide and carrier and by the efficient delivery of theircargo into cells (6).

However, when applied in vivo, cationic lipo-somes or polymers face tolerability issues, due to aggre-gate formation in serum and related toxicity (7), or arelimited in their biodistribution and cannot penetrateorgans or reach inflammatory disease sites (8). Theselimitations in biodistribution and tolerability have beenaddressed through PEGylation of liposomes (9,10), butthis manipulation reduces the cellular uptake of thecarrier (11). In addition, cationic liposomes, PEGylatedor not, are strong inducers of Toll-like receptor (TLR)responses when combined with oligonucleotides(8,12,13). Diffusible PEGylation (14,15) and cleavablePEGylation (16) are the most recent advances that havebeen used to address these issues, but this still carries aheritage burden from its predecessor in terms of theTLR response and cell damage. Neutral or anionicliposomes are rarely used. Although they are safe for invivo application, they are associated with intrinsic prob-lems, such as low encapsulation of the payload due tolack of electrostatic attraction between antisense oligo-nucleotide (ASO) and the lipid membrane and ineffi-cient delivery of their cargo (17).

In this study, we developed a new liposomalcarrier, termed Nov038, for the systemic delivery ofoligonucleotides to inflammatory disease sites. Nov038is unique among oligonucleotide carriers in that it has anisoelectric point of 6.3 and exhibits amphoteric proper-ties. At low pH, Nov038 is cationic, thus facilitatingeffective encapsulation of oligonucleotides, whereas atneutral pH, Nov038 is anionic, a property that makes thesystem blood-compatible and eliminates the need forPEGylation. Once taken up by cells, Nov038 becomesexposed to the lowered pH of the endosomes, whichleads to charge loss and delivery of the oligonucleotidepayload.

We show that Nov038 is safe, well-tolerated inthe circulation, devoid of nonspecific immune-stimulatory effects, and effective for systemic oligonucleo-tide delivery in vivo. We also demonstrate that Nov038enables the systemic and targeted delivery of CD40-specific ASO (CD40-ASO) to myeloid dendritic cells(DCs) and macrophages and achieves potent geneknockdown effects in vivo. In experimental arthritis inmice, Nov038 delivers CD40-ASO to arthritic joints andimproves established disease, whereas unassisted CD40-ASO treatment is ineffective in this model. Therefore,amphoteric liposomes constitute a new carrier technol-

AMPHOTERIC LIPOSOME THERAPY IN EXPERIMENTAL ARTHRITIS 995

ogy for the systemic and targeted delivery of RNAioligonucleotides to inflammatory disease cells or sites.Moreover, Nov038/CD40-ASO provides a novel non–antibody-mediated and presumably safer approach totherapeutically target CD40–CD40L interactions in vivo.

MATERIALS AND METHODS

Production of amphoteric liposomes. Synthesis of�-(3�-O-cholesteryloxycarbonyl)-�-(N-ethylmorpholine)-succinamide (MoChol), production of amphoteric and cationicliposomes, quality control, and measurements of zeta (surface)potentials were performed as described elsewhere (detailsavailable upon request from the corresponding author).

Oligonucleotides. All ASOs used were fully modifiedwith a phosphorothioate backbone. CD40-ASO targets intronVI of the mouse CD40 gene and interferes with the productionof messenger RNA (mRNA) at the pre-mRNA stage. Thesequences used were as follows: for CD40-Cy5.5 (Cy5.5),5�-CCCAGTCAGTGTTCCTGC-3�; for Alexa 488–ASO(Alexa 488–Amino C6), 5�-CCTTCCCTGAAGGTTCCTCC-3�; for CD40-ASO (ISIS 117898), 5�-CCCAGTCAGTGTTC-CTGC-3�; and for scrambled-sequence CD40-ASO (SCR-ASO) (ISIS 141923), 5�-CCTTCCCTGAAGGTTCCTCC-3�(underlines indicate the 2�-O-methoxy-ethyl–modified nucleo-tides).

Encapsulation of oligonucleotides into liposomes.Oligonucleotides or the TLR-7/8 ligand ORN-R0006 wasencapsulated into liposomes, and their immunostimulatorypotentials were assessed as described elsewhere (see supple-mental materials, available upon request from the correspond-ing author).

Biodistribution studies. Biodistribution of amphotericliposome–delivered CD40-ASO was determined using near-infrared imaging (Licor Odyssey, Lincoln, NE) and a Cy5.5-labeled CD40 antisense. Fluorescence-activated cell sorter(FACS) analysis of Alexa 488–labeled oligonucleotide andliquid chromatography mass spectrometry (LC-MS) quantita-tion of oligonucleotide in tissue were also used to assess thebiodistribution at the single-cell level and in a quantitativemanner.

Induction, treatment, and assessment of collagen-induced arthritis (CIA). Male DBA/1 mice (10–12 weeks old)were immunized with 200 �g of type II collagen emulsified inFreund’s complete adjuvant (Difco, Detroit, MI) by a singleintradermal injection at the base of the tail. Starting on day 14after immunization, mice were examined daily for the onsetand progression of clinical arthritis. By day 28, more than 80%of mice had developed the disease. Arthritis severity wasassessed using a clinical scoring system, as follows: 0 � normal,1 � slight swelling and/or erythema, 2 � pronounced edema-tous swelling, and 3 � joint rigidity. Each limb was graded, toyield a maximum clinical score of 12 per mouse. The thicknessof the affected hind paws was also monitored using microcali-pers (Kroeplin, Munich, Germany). After a followup period of10 days post–arthritis onset, mice were killed, and the affectedlimbs were examined by histology.

Treatment was administered intravenously (IV) on thefirst day of clinical signs of arthritis (pronounced edema) and

repeated on days 3 and 5 after arthritis onset. CD40-ASO or itsscrambled control encapsulated in Nov038 was administered ata dose of 4 mg/kg. Control arthritic mice received injections IVof vehicle (saline). All assessments were performed by observ-ers (EA and AS) who were blinded to the treatment received.On day 6 after arthritis onset, lymph node (LN) cultures,isolation of CD11c�CD11b� myeloid DCs, and mixed lym-phocyte reactions with C57BL/6 splenocytes were performed(see supplemental information, available upon request fromthe corresponding author). Sera were also analyzed for hema-tologic parameters, using an automated blood analyzer.

Histopathologic analysis. Ten days after arthritis on-set, hind paws were removed postmortem, fixed in formalin,and decalcified in Calci-Clear (National Diagnostics, Atlanta,GA). Paws were embedded in paraffin, and saggital sectionswere cut and stained with hematoxylin and eosin. Histopatho-logic assessment of arthritis was carried out in a blindedmanner, using a scoring system as follows: 0 � normal, 1 �minimal synovitis without cartilage/bone erosion, 2 � synovitiswith some marginal erosion but joint architecture maintained,and 3 � severe synovitis and erosion with loss of normal jointarchitecture.

Real-time polymerase chain reaction (PCR) analysis.RNA from the liver and paws of arthritic mice were analyzedby quantitative real-time PCR. (Primer sequences and ampli-cons are available upon request from the corresponding au-thor.) Changes in gene expression level were calculated by the2��Ct method.

Statistical analysis. Student’s t-test for parametricdata and the Mann-Whitney U test for nonparametric datawere used. Active treatment was compared with saline treat-ment (control). P values less than 0.05 were considered signif-icant.

RESULTS

Design of amphoteric liposomes and develop-ment of Nov038. We designed and synthesized MoChol,a weak cationic amphiphile derived from cholesterol(Figure 1A). The pKa of the novel compound wasexperimentally determined in liposomes (18) in which apKa of 6.5 was observed. The combination of equalamounts of MoChol and cholesterylhemisuccinate(CHEMS), which showed a pKa of 5.8 (19), providedamphoteric character with an isoelectric point of �6.3(Figure 1B).

In contrast to cationic carrier systems that bindoligonucleotides at their surface, amphoteric liposomesneed to sequester these materials, because the surfacecharge becomes anionic at neutral pH. We found thatsealed vesicles that stably sequester the soluble dyecarboxyfluorescein were formed from equimolar mix-tures of MoChol and CHEMS in a lipid-dose– andliposome-size–dependent manner (Rauchhaus U: un-published observations), although rapid release of thedye occurred when these liposomes were exposed to

996 ANDREAKOS ET AL

100% human serum. The serum resistance of ampho-teric liposomes could be substantially improved whenneutral lipids, such as 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) or 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE), were added together inthe amphoteric mixture of MoChol and CHEMS, and avalley of stability was identified for mixtures comprising

up to 60 mol% of these charged lipids in combinationwith DOPE and POPC (Figure 1C).

Nov038 (DOPE:POPC of 3:1 with 60 mol% oftotal lipids, and MoChol:CHEMS of 1:1 with 40 mol%of total lipids) was selected due to its optimal stability incombination with a substantial amount of charged lipids.Nov038 itself retains the amphoteric character of its

Figure 1. Design and properties of amphoteric Nov038 liposomes. A, Structure of �-(3�-O-cholesteryl-oxycarbonyl)-�-(N-ethylmorpholine)-succinamide (MoChol), a weak cationic amphiphile derived fromcholesterol. B, Net charge (right axis) of an equimolar mixture of MoChol and cholesterylhemisuccinate(CHEMS) (broken line), and surface potentials (left axis) of Nov038 (squares) and Nov038/CD40–antisense oligonucleotide (ASO) (triangles). C, Serum stability of an array of amphoteric liposomes (grayshaded columns; the solid black column represents Nov038) produced by increasing equimolar amountsof CHEMS and MoChol (left bottom axis) blended into neutral lipids with different ratios of1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE) (right bottom axis). Carboxyfluorescein was encapsulated into amphotericliposomes as an aqueous phase marker, and release of dye was monitored in the presence of serum. A 3:1combination of DOPE to POPC substantially improved the serum stability of amphoteric liposomes. D,Serum stability of Nov038. An anionic derivative of Nov038 (POPC:DOPE:CHEMS 45:15:40) (left) andNov038 itself (middle), but not a weak cationic derivative (POPC:DOPE:MoChol 45:15:40) (right),retained their particle size when contacted with serum albumin. Particle size in buffer is indicated by thebroken line, while particle size in albumin solution is indicated by the solid line. E and F, Directcomparison of the immunomodulatory effects of the Toll-like receptor-7/8 (TLR-7/8) agonist ORN-R0006formulated with either Nov038 or 1,2-dioleoyl-3-trimethylammonium propane (DOTAP). TLR-7–mediated levels of interferon-� (IFN�) (E) and TLR-8–mediated levels of tumor necrosis factor � (TNF�)(F) were determined in human peripheral blood mononuclear cells by enzyme-linked immunosorbentassay. Cells were stimulated with the indicated concentrations of ORN-R0006 formulated in Nov038 orDOTAP. After 24 hours, supernatants were harvested, and IFN� and TNF� were measured byenzyme-linked immunosorbent assay. Bars show the mean � SEM results from 3 donor samples.


charged constituents, but the charge reversal is morepronounced in the presence of ASO (Figure 1B), aphenomenon that needs further investigation. We con-firmed particle growth and strong aggregation withalbumin for a strongly cationic derivative of Nov038 inwhich both CHEMS and MoChol were replaced by1,2-dioleoyl-3-trimethylammonium propane (DOTAP)(results not shown). The same observation, although toa lesser extent, was made for a weakly cationic derivativeof Nov038 in which MoChol was the only charged lipid(Figure 1D). Therefore, even the small fraction ofMoChol that is still protonated at neutral pH is suffi-cient to promote aggregation. In contrast, authenticNov038, or a derivative thereof lacking MoChol but notCHEMS, fully retained its particle size after incubationwith albumin (Figure 1D). We were able to produceNov038 within tight specifications of size, surface poten-tial, and drug:lipid ratio. Particle specifications weremaintained irrespective of the presence or absence ofoligonucleotides (results available upon request fromthe corresponding author).

Nov038 was observed to be well-tolerated in thecirculation. We demonstrated that Nov038 effectivelyprevented ORN-R0006, an immune-stimulatory oligo-nucleotide recognized by TLR-7/8, from inducinginterferon-� (IFN�) and TNF� production from humanperipheral blood mononuclear cells (PBMCs). This wasin contrast to the high levels of IFN� and TNF�produced when human PBMCs were cultured withORN-R0006 complexed in cationic liposomes based onDOTAP (Figures 1E and F) (20). In addition, there wasno increase in the levels of the liver enzymes aspartateaminotransferase/serum glutamic oxaloacetic transami-nase (AST/SGOT) or alanine aminotransferase/serumglutamic pyruvic transaminase (ALT/SGPT) and nochanges in alkaline phosphatase (ALP) levels in theserum, despite repeated administration of Nov038/CD40(results available upon request from the correspondingauthor).

Improvement of oligonucleotide delivery in vivoand targeting of oligonucleotides to immune organsand inflamed joints by Nov038. We examined whetherNov038 could be used for the systemic delivery ofoligonucleotides to tissues of interest. Cy5.5-labeling offirst-generation phosphorothioate CD40-ASO and near-infrared imaging analysis revealed that Nov038-encapsulated CD40-ASO (Nov038/CD40-ASO) local-ized to the liver, spleen, kidneys, and LNs (Figure 2A)(see also supplemental information, available upon re-quest from the corresponding author), whereas therewas little accumulation in other organs, such as the lung,

which is a major site associated with toxicity of cationicliposomal formulations (21–23). Penetration of Nov038/

Figure 2. Biodistribution of Nov038 in healthy and arthritic mice. A,Mice with collagen-induced arthritis (CIA) and naive mice (each n �3) were treated with Nov038/Cy5.5-labeled CD40–antisense oligo-nucleotide (ASO), and organs were collected after 8 hours. Organswere frozen, and 10-�m sections were mounted on coverslips. Localconcentrations of Cy5.5-labeled material were quantified using a LicorOdyssey near-infrared imager. B, Mice (n � 3) were treated with asingle dose of 4 mg/kg unlabeled Nov038/CD40-ASO, and organs werecollected after 8 hours. ASO was eluted from the organ samples andquantified with a liquid chromatography mass spectrometry detectionsystem (24). C, Arthritic and nonarthritic mice on day 1 of arthritiswere treated with Nov038/Cy5.5–CD40-ASO, and local accumulationin healthy and inflamed paws was analyzed histopathologically andcorrelated with clinical scoring. Circles indicate dilation of the capil-lary endothelium at the inflamed sites, and values in the cornersindicate the relative fluorescence levels. D, Mice were treated withNov038/Alexa 488–labeled CD40-ASO, and peripheral blood or or-gans were collected after 4 hours. Single-cell suspensions were ob-tained, and the percentage of transfected mononuclear cells wasdetermined by fluorescence-activated cell sorter analysis. Bars showthe mean and SEM. ILN � inguinal lymph nodes.

998 ANDREAKOS ET AL

CD40-ASO was deep and reached the entire liver pa-renchyma and red pulp of the spleen.

A similar pattern of biodistribution was observedwhen LC-MS (24) was used to quantitate unlabeled,second-generation CD40-ASO in absolute amounts inmajor tissues. Approximately 65% of the injected doseof Nov038/CD40-ASO was recovered as intact, full-length oligonucleotide from the liver, spleen, and kid-neys. The highest concentrations were achieved in thespleen, followed by the liver and the kidneys. Becausebiodistribution studies with radiolabeled Nov038 do notsupport the efficacy of carrier localization or Nov038-mediated delivery to the kidney (Endert G: unpublishedobservations), the signal in the kidney probably reflectsthe released or nonencapsulated oligonucleotide thataccumulates in that organ (25). Indeed, unformulatedCD40-ASO, when injected into mice, mostly accumu-lated in the kidney, with the liver and spleen also takingup some oligonucleotide. In a side-by-side comparison,there was a 20-fold increase in accumulation of intactoligonucleotide in the spleen when Nov038-assisted de-livery, as compared with unassisted (unformulated) oligo-nucleotide delivery, was used (Figure 2B).

Importantly, Nov038/CD40-ASO also accumu-lated in inflamed sites, such as arthritic (but not normal)joints in the same mouse, presumably due to the dilationof the capillary endothelium at the inflamed sites (Fig-

ure 2C). However, this represented only a small fractionof the total Nov038/CD40-ASO administered, since itsgross biodistribution was not affected (Figure 2A).

Further analysis by FACS using Alexa 488–labeled oligonucleotide revealed that encapsulation ofCD40-ASO in Nov038 resulted in substantial accumula-tion in mononuclear cells in the peripheral blood, spleen,liver, and inguinal LNs (Figure 2D). These findings dem-onstrate that Nov038 substantially improves delivery ofits cargo oligonucleotide to immune organs and sites ofinflammation in vivo.

Amelioration of experimental arthritis in mice byNov038-mediated CD40-ASO delivery. The safety pro-file and effectiveness of Nov038 in delivering CD40-ASO in vivo, together with its specific uptake in in-flamed joints, prompted us to study its therapeutic use inCIA, an experimental model of RA in mice. Remark-ably, systemic administration of second-generation,2�MOE-modified CD40-ASO in Nov038 at a dose of 4mg/kg potently ameliorated established disease. Within24 hours after administration, treatment improved theclinical score of the paws of arthritic mice and reducedthe number of joints affected over the 10-day period offollowup (Figure 3A). Paw swelling in the arthritic joints,as measured by microcalipers, was also profoundly in-hibited by the treatment (Figure 3B).

Encapsulation of CD40-ASO into Nov038 was

Figure 3. Effect of Nov038 encapsulation on CD40–antisense oligonucleotide (ASO) therapeutic efficacy in vivo. The clinicalscore for arthritis (A, C, and E) and the extent of hind paw swelling (B, D, and F) were determined in mice treated on days 1,3, and 5 after arthritis onset in the following groups: Nov038/CD40-ASO (n � 14), Nov038/scrambled-sequence ASO (SCR-ASO)(n � 8), or saline (n � 13) (A and B); free CD40-ASO (n � 6), Nov038/CD40-ASO (n � 14), empty Nov038 (n � 7), Nov038with nonencapsulated CD40 (CD40�Nov038) (n � 10), or saline (n � 13) (C and D); Nov038/CD40-ASO (n � 14), anti–tumornecrosis factor � (anti-TNF�) neutralizing antibody (n � 6), dexamethasone (n � 7), or saline (n � 13) (E and F). Bars showthe mean and SEM. � � P � 0.05; �� P � 0.01; �� P � 0.001, versus saline-treated control group, by Student’s t-test.


necessary for therapeutic efficacy, as was shown whenadministration of free CD40-ASO at 4 mg/kg or nonen-capsulated CD40-ASO mixed with empty Nov038 hadno effect on either the clinical score or paw swelling(Figures 3C and D). Specificity of the CD40-ASOtreatment was confirmed in experiments using the SCR-ASO or empty Nov038, neither of which had any effecton disease progression or severity (Figures 3A and B).

Importantly, Nov038/CD40-ASO treatment wasfound to be more effective than anti-TNF� antibodytreatment, a gold standard for the treatment of RA (1).On a molar basis, the formulated oligonucleotide was atleast as potent as dexamethasone phosphate, one of themost efficacious treatments for this model. Nov038/CD40-ASO showed a rapid onset of action and acted asfast as the glucocorticoid within the resolution of ourdata (Figures 3E and F).

Reduction in inflammation and joint damagefollowing Nov038-mediated CD40-ASO delivery. InCIA, mice are known to develop extensive inflammationwith pannus formation and erosive changes in theiraffected joints that can be optimally detected 10 daysafter disease onset (26). We therefore performed his-topathologic analysis of the arthritic joints. We foundthat Nov038/CD40-ASO administration potently inhib-ited inflammatory cell infiltration into the joints and

substantially reduced pannus formation, whereas at thesame time it protected joints from erosive changesassociated with cartilage and bone damage (Figure 4A).This was in contrast to the effects of free CD40-ASO,Nov038/SCR-ASO, nonencapsulated CD40-ASO ad-ministered in combination with empty Nov038, or emptyNov038, none of which had any effect at the doses tested(Figure 4A). Differences between the groups were sta-tistically significant when multiple joints were scored(26) using indices of histologically determined inflam-mation and erosion (Figures 4B and C).

Down-regulation of CD40 in arthritic joints, re-duction in proinflammatory cytokines, and inhibition ofjoint-specific T cell responses by Nov038/CD40-ASO.We next set out to identify the potential mechanisms ofaction of Nov038/CD40-ASO in vivo. Using quantitativereal-time PCR, we found that Nov038/CD40-ASO rap-idly decreased CD40 levels in arthritic joints. Comparedwith saline treatment, a single administration of Nov038/CD40-ASO down-regulated CD40 levels by �40%, andtriple administration resulted in �70% inhibition ofCD40 mRNA production (P � 0.05) (Figure 5A). Incontrast, administration of Nov038/SCR-ASO had noeffect.

The down-regulation of CD40 in arthritic jointswas accompanied by a significant reduction in the ex-

Figure 4. Histopathologic assessment of the joints of arthritic mice. Arthritic mice were treated with free CD40–antisense oligonucleotide (ASO) (n � 9), empty Nov038 (n � 7), Nov038 with nonencapsulated CD40(CD40�Nov038) (n � 9), Nov038/CD40-ASO (n � 10), Nov038/scrambled-sequence ASO (SCR-ASO) (n � 13), orsaline control (n � 8) and then were killed on day 10 after arthritis onset. A, Representative images of the proximalphalangeal joints of the hind feet, stained with hematoxylin and eosin (original magnification 40). Severe synovitis(sy), inflammatory pannus tissue (p), and erosions of cartilage and bone (e) are shown. B and C, Inflammation (B) anderosion (C) histology scores for the hind paws of treated mice. Bars show the mean and SEM. �� P � 0.01 versussaline-treated control group, by Mann-Whitney U test.

1000 ANDREAKOS ET AL

pression of TNF� (mean � SEM fold decrease 74 �9%; P � 0.01 versus saline-treated controls), IL-6 (folddecrease 88 � 20%; P � 0.05 versus saline-treatedcontrols), and IL-17 (fold decrease 77 � 14%; P � 0.05versus saline-treated controls) (Figure 5B), all of whichare major inflammatory cytokines involved in the patho-genesis of RA (27). At the same time, administration ofNov038/CD40-ASO profoundly reduced joint (type IIcollagen)–specific T cell responses in the draining LNs,whereas administration of its scrambled control had noeffect (Figure 5C). Since T cell activation is dependent

on effective presentation of type II collagen by APCs,this observation suggests that Nov038/CD40-ASO alsocompromises APC function via down-regulation ofCD40. Therefore, Nov038/CD40-ASO treatment inter-

Figure 6. Tropism of Nov038 for myeloid dendritic cells (DCs), andeffect of treatment on DC function and T cell responses. A and B,Fluorescence-activated cell sorter (FACS) analysis of peripheral blood(A) and splenocytes (B) from mice treated with Nov038-encapsulatedor free Alexa 488–labeled antisense oligonucleotide (ASO) (mix of 2:8labeled ASO with functional CD40-ASO), showing the presence offluorescence in CD11chighCD11b� myeloid DCs and CD11c�CD11b�macrophages. C, Reduction of CD40 protein abundance monitored byFACS in CD11b� cells from the spleens of mice treated once withNov038/CD40-ASO; expression of CD40 was analyzed 24 hours aftertreatment. The reduction in CD11b� cells is �1 order of magnitude.D, Mixed lymphocyte reaction of CD11c�CD11b� myeloid DCs fromsaline-treated (n � 6) or Nov038/CD40-ASO–treated (n � 4) DBA/1arthritic mice. Myeloid DCs were sorted by FACS and cultured withallogeneic splenocytes from C57BL/6 mice. Bars show the mean andSEM 3H-thymidine incorporation. � � P � 0.05 versus saline-treatedcontrol group, by Mann-Whitney U test.

Figure 5. Effect of Nov038/CD40–antisense oligonucleotide (ASO)on the expression of CD40 and inflammatory cytokines in arthriticjoints, and joint-specific T cell responses in draining lymph nodes.A, The hind feet from arthritic mice were analyzed for CD40 mRNAexpression after single or triple administration of Nov038/CD40-ASO,Nov038/scrambled-sequence ASO (SCR-ASO), or saline. B, Thehind feet from arthritic mice were analyzed for expression of tumornecrosis factor � (TNF�), interleukin-6 (IL-6), and IL-17 after tripleadministration of Nov038/CD40-ASO or saline. The mRNA levels inthe treated groups are expressed as the fold difference from saline-treated controls. C, Proliferation of draining lymph node cells wasassessed from day 6 after arthritis onset in arthritic mice treated withNov038/CD40-ASO, Nov038/SCR-ASO, or saline, in medium alone orafter culture in the presence of type II collagen. Bars showthe mean and SEM results from 5 mice per group. � � P � 0.05; �� P � 0.01, versus saline-treated control group, by Student’s t-test orMann-Whitney U test.


feres with both the inflammatory and the autoimmuneprocesses that occur in this model.

Delivery of CD40-ASO to macrophages and my-eloid DCs and down-regulation of DC antigen-presentingfunction by Nov038. FACS analysis revealed thatNov038 delivers its cargo to monocyte/macrophages(CD11c�CD11b�) and myeloid DCs (CD11chighCD11b�B220�), which are important target cells for both theinduction and progression of autoimmune and inflam-matory diseases. Thus, monocyte/macrophages and my-eloid DCs from the peripheral blood (Figure 6A), spleen(Figure 6B), and liver (results not shown) all wereobserved to take up Nov038-encapsulated CD40-ASOlabeled with Alexa 488, whereas T cells (CD3�), B cells(CD19�), or plasmacytoid DCs (CD11clow/intermediate

CD11b�B220�) showed no uptake (Figures 6A and B).This resulted in down-regulation of CD40 expression inCD11b� macrophages/myeloid DCs (Figure 6C).

These findings provide direct evidence for amode of action of Nov038/CD40-ASO that starts frominhibition of the target gene expression and subse-quently involves reduction of proinflammatory geneexpression in the joints and, more importantly, inhibi-tion of joint-specific T cell responses in draining LNs.Indeed, sorted splenic myeloid DCs (CD11chighCD11b�B220�) from Nov038/CD40-ASO–treated mice had ab-rogated APC capacity compared with saline-treatedmice (Figure 6D), demonstrating that Nov038/CD40-ASO–mediated inhibition of the autoimmune reaction isdue to the selective targeting of the antigen-presentingfunction of myeloid DCs.

DISCUSSION

Oligonucleotides mediating RNAi constitute apowerful approach for the silencing of genes involved inthe pathogenesis of inflammatory disease, both forunderstanding disease mechanisms and for developingnovel therapeutics. However, their in vivo applicationrequires effective delivery to immune cells and/or sitesof inflammation (28). Recent efforts to facilitate sys-temic oligonucleotide delivery to inflammation sites invivo have included the use of neutral liposomes (17) oratelocollagen (29) in experimental allergic dermatitis,cationic liposomes in experimental arthritis (30), andantibody-targeted stabilized nanoparticles in experimen-tal colitis (31).

In the present study, we assessed a novel lipo-somal carrier system for systemic and targeted deliveryof oligonucleotides to inflammatory disease sites. Theuniqueness of this system is its amphoteric, charge-

reversible character, achieved by the combination of anovel, weakly cationic lipid, MoChol, which providescationic charge at low pH, and CHEMS, which isanother pH-sensitive but anionic amphiphile. Unlikecationic liposomes, optimized amphoteric carriers suchas Nov038 do not aggregate in the serum. Serum ag-gregation and rapid interaction of cationic liposomeswith endothelial surfaces limit their in vivo application,because it results in broad distribution but limitedpenetration of such carriers into tissues and also causestolerability problems (8). In contrast, the amphotericcarrier Nov038 penetrates beyond endothelia and accu-mulates in the liver, spleen, and, very importantly,sites of inflammation such as arthritic joints, whichrepresent known sites of absorption for anionic orneutral liposomes.

Of note, the favorable biodistribution propertiesof Nov038 were achieved without PEGylation, address-ing a longstanding dilemma in the development ofliposomal oligonucleotide carriers. Delivery in Nov038results in enhanced overall absorption of intact CD40-ASO and changes the organ distribution of the activeingredient. This was most pronounced in the spleen(20-fold increase), a major immune organ that does notaccumulate substantial amounts of unformulated CD40-ASO but shows the highest local concentrations ofoligonucleotide when delivered in Nov038.

Nov038 is well-tolerated in vivo, with no obvioustreatment-related effects on the appearance or behaviorof the animals and no increase in levels of ALT, AST, orALP in the serum. Nov038 is also devoid of nonspecificimmune-stimulatory effects, even when immunostimula-tory oligonucleotides, such as the TLR-7/8 ligand ORN-R0006 (20), are used. This is a dramatic advantage overcationic liposomes that are based on DOTAP, whichinduce a strong release of inflammatory cytokines in thesame setting, or PEGylated and transiently PEGylatedcationic liposomes, which have been described recently(12,13,32). To minimize immunogenicity of the cationiccarrier–oligonucleotide assembly, the selection of non-immunostimulatory siRNA molecules or the circumven-tion of APCs have been suggested (14,15). It is likelythat nonspecific immune stimulation and triggering ofTLR activation is due to oligonucleotide presentation onthe outer surface of liposomal carriers, a property that isintrinsic to all cationic carriers but absent in amphotericcarriers such as Nov038, which escape immune surveil-lance even when targeting macrophages and DCs.

In terms of efficacy, Nov038-delivered CD40-ASO achieves potent gene knockdown effects in vivo,even at sites of inflammation, as demonstrated by the


depletion of CD40 levels in arthritic joints. In contrast,unassisted CD40-ASO or Nov038-delivered SCR-ASOhad no effect. In CIA, Nov038 encapsulation renders anotherwise-ineffective dose of CD40-ASO therapeuticallyeffective. Systemic administration of Nov038/CD40-ASO successfully treated established arthritis in ourmodel by reducing the clinical score, the extent of pawswelling, and the number of joints affected. Histologi-cally, Nov038/CD40-ASO reduced inflammatory cellinfiltration and prevented cartilage and bone damage inthe affected joints. In contrast, unformulated CD40-ASO or Nov038/SCR-ASO had no effect on any of theseparameters.

This is, to our knowledge, the first study demon-strating therapeutic efficacy of CD40–CD154 blockadein experimental arthritis in mice, although prophylacticinterference with this pathway has been previously re-ported (33,34). Remarkably, we found, in a direct com-parison, that Nov038/CD40-ASO was as rapid and ef-fective as dexamethasone phosphate in improvingdisease and far more effective than TNF�-blockingantibodies, a gold standard for RA therapy. Furtherstudies will be needed to determine the therapeuticbenefit of Nov038/CD40-ASO treatment in the longterm and in later stages of the disease process.

The efficacy of systemic Nov038/CD40-ASOtreatment may be explained by the directed delivery ofCD40-ASO to APCs, namely, monocyte/macrophagesand myeloid DCs. This points to phagocytosis as themechanism of uptake and is consistent with the fact thatplasmacytoid DCs, which are phagocytically less activethan myeloid DCs (35), and lymphocytes are not trans-fected with this carrier. Immune cell–directed deliveryprovides significant advantages for oligonucleotide-based therapeutics, both in terms of the concentrationand dose needed for efficacy and the possible sideeffects and toxicity, factors that limit the systemic use ofalternative oligonucleotide carriers. Recent reports havestrongly suggested the use of antibody-conjugated oligo-nucleotides (36) or targeting ligands (31) to achievetransfection of primary blood cells after systemic admin-istration. The effective cell-specific targeting of ampho-teric liposomes without the need for ligandation greatlyfacilitates the technical development of this new class ofcarriers. It is envisaged that further manipulation of thelipid composition of amphoteric liposomes will altertheir tropism, thus enabling oligonucleotide delivery todifferent cell types.

Several pathogenic processes may be affected byNov038/CD40-ASO treatment. First, treatment deliversCD40-ASO to arthritic joints either directly through

accumulation of Nov038 in inflamed joints or indirectlythrough trafficking of macrophages and DCs to thejoints. Either way, Nov038/CD40-ASO successfully in-hibits CD40 expression in arthritic joints and reduces thelocal production of TNF�, IL-6, and IL-17, which arecritical pathogenic mediators of RA. The reduction inTNF� and IL-6 expression is likely to be a directconsequence of CD40 down-regulation. T cells frompatients with RA express abundant CD40L (37), andCD40–CD40L interactions have been shown to regulatemacrophage activation and production of inflammatorycytokines such as TNF� and IL-6 locally in the joints(38–40). Moreover, Nov038 delivers CD40-ASO to my-eloid DCs, major regulators of the immune systemprocesses leading to immunity or tolerance.

The importance of CD40–CD40L interactions inT cell activation, differentiation, and tolerance (41) andthe importance of effective DC antigen presentation(42,43) are well established. Nov038/CD40-ASO–mediated down-regulation of CD40 in DCs may there-fore account for the down-regulated collagen-specificT cell responses in draining LNs in the joints and thereduced presence of IL-17–producing T cells in arth-ritic joints. Indeed, CD11chighCD11b� myeloid DCsfrom Nov038/CD40-ASO–treated mice were poor stim-ulators of allogeneic splenocyte proliferation comparedwith that in saline-treated mice, demonstrating thatNov038/CD40-ASO treatment severely compromisesmyeloid DC APC function.

Immunotherapy using antigen-specific ap-proaches such as costimulation blockade has attractedconsiderable attention for the treatment of autoimmu-nity and transplant rejection (44). Recent efforts havefocused on targeting the CD40–CD40L interaction witholigonucleotides using a CD40 siRNA in vitro (45), andCD40 antisense was used for topical administration invivo in an animal model of colitis (46) and transplantrejection (47). In humans, targeting costimulation with ahumanized mAb against CD40L has shown efficacy inclinical trials of SLE and transplant rejection (3,48).However, further development of this approach wasdiscontinued because of toxic thromboembolytic eventsthat were believed to be caused either by direct inter-action of the anti-CD40L mAb with platelet Fc receptorsand related effector mechanisms or by inhibition ofthe CD40L–�3 platelet integrin interaction (4,8). In oursystem, we achieved cell-specific down-regulation ofCD40 in macrophages and DCs. Thus, targeted deliveryof CD40-ASO using Nov038 may provide a more selec-tive and presumably safer therapeutic alternative for


interfering with CD40–CD40L interactions in inflamma-tory diseases such as RA.

In summary, we demonstrate that amphotericliposomes constitute a novel carrier concept for systemicand APC-targeted oligonucleotide delivery, with clinicalapplicability and numerous potential applications, in-cluding target validation in vivo and inflammatory dis-ease therapeutics. The ability of Nov038 to accumulatein inflammatory disease sites and to transfect myeloidDCs and macrophages in vivo without affecting plasma-cytoid DCs or lymphocytes permits the use of RNAitechnology for the subtle regulation of the immunesystem in conditions in which extensive or prolongedimmune and inflammatory responses are involved.Therefore, this study has wider implications both for thedevelopment of oligonucleotide-based therapeutics andfor immunomodulation therapies for inflammatory andautoimmune diseases such as RA.

ACKNOWLEDGMENTS

We thank Charoula Xirakia, Anastasia Apostolidou,Anna Agapaki, Arianna Gavriil, Ulrike Samulowitz, andRainer Curdt for technical assistance. We also thank BrettMonia, Gene Hung, and Marc Feldmann for helpful discus-sions.

AUTHOR CONTRIBUTIONS

Drs. Andreakos and Panzner had full access to all of the datain the study and take responsibility for the integrity of the data and theaccuracy of the data analysis.Study design. Andreakos, Rauchhaus, Karras, Bennett, Williams,Sideras, Panzner.Acquisition of data. Andreakos, Rauchhaus, Stavropoulos, Endert,Wendisch, Benahmed, Giaglis, Lee, Panzner.Analysis and interpretation of data. Andreakos, Rauchhaus, Stavro-poulos, Endert, Lee, Gaus, Williams, Panzner.Manuscript preparation. Andreakos, Stavropoulos, Lee, Gaus,Bennett, Williams, Sideras, Panzner.Statistical analysis. Andreakos, Stavropoulos, Giaglis, Lee, Gaus.

REFERENCES

1. Andreakos ET, Foxwell BM, Brennan FM, Maini RN, FeldmannM. Cytokines and anti-cytokine biologicals in autoimmunity:present and future. Cytokine Growth Factor Rev 2002;13:299–313.

2. Smolen JS, Aletaha D, Koeller M, Weisman MH, Emery P. Newtherapies for treatment of rheumatoid arthritis. Lancet 2007;370:1861–74.

3. Grammer AC, Slota R, Fischer R, Gur H, Girschick H, YarboroC, et al. Abnormal germinal center reactions in systemic lupuserythematosus demonstrated by blockade of CD154-CD40 inter-actions [published erratum appears in J Clin Invest 2007;117:835].J Clin Invest 2003;112:1506–20.

4. Kawai T, Andrews D, Colvin RB, Sachs DH, Cosimi AB. Throm-boembolic complications after treatment with monoclonal anti-body against CD40 ligand. Nat Med 2000;6:114.

5. Scherer LJ, Rossi JJ. Approaches for the sequence-specific knock-down of mRNA. Nat Biotechnol 2003;21:1457–65.

6. Akhtar S, Benter IF. Nonviral delivery of synthetic siRNAs in vivo.J Clin Invest 2007;117:3623–32.

7. Chien PY, Wang J, Carbonaro D, Lei S, Miller B, Sheikh S, et al.Novel cationic cardiolipin analogue-based liposome for efficientDNA and small interfering RNA delivery in vitro and in vivo.Cancer Gene Ther 2005;12:321–8.

8. Santel A, Aleku M, Keil O, Endruschat J, Esche V, Fisch G, et al.A novel siRNA-lipoplex technology for RNA interference in themouse vascular endothelium. Gene Ther 2006;13:1222–34.

9. Meyer O, Kirpotin D, Hong K, Sternberg B, Park JW, WoodleMC, et al. Cationic liposomes coated with polyethylene glycol ascarriers for oligonucleotides. J Biol Chem 1998;273:15621–7.

10. Anwer K, Kao G, Proctor B, Rolland A, Sullivan S. Optimizationof cationic lipid/DNA complexes for systemic gene transfer totumor lesions. J Drug Target 2000;8:125–35.

11. Shi N, Zhang Y, Zhu C, Boado RJ, Pardridge WM. Brain-specificexpression of an exogenous gene after i.v. administration. ProcNatl Acad Sci U S A 2001;98:12754–9.

12. Mui B, Raney SG, Semple SC, Hope MJ. Immune stimulation bya CpG-containing oligodeoxynucleotide is enhanced when encap-sulated and delivered in lipid particles. J Pharmacol Exp Ther2001;298:1185–92.

13. Semple SC, Harasym TO, Clow KA, Ansell SM, Klimuk SK, HopeMJ. Immunogenicity and rapid blood clearance of liposomescontaining polyethylene glycol-lipid conjugates and nucleic acid.J Pharmacol Exp Ther 2005;312:1020–6.

14. Morrissey DV, Lockridge JA, Shaw L, Blanchard K, Jensen K,Breen W, et al. Potent and persistent in vivo anti-HBV activity ofchemically modified siRNAs. Nat Biotechnol 2005;23:1002–7.

15. Zimmermann TS, Lee AC, Akinc A, Bramlage B, Bumcrot D,Fedoruk MN, et al. RNAi-mediated gene silencing in non-humanprimates. Nature 2006;441:111–4.

16. Hatakeyama H, Akita H, Kogure K, Oishi M, Nagasaki Y, KihiraY, et al. Development of a novel systemic gene delivery system forcancer therapy with a tumor-specific cleavable PEG-lipid. GeneTher 2007;14:68–77.

17. Klimuk SK, Semple SC, Nahirney PN, Mullen MC, Bennett CF,Scherrer P, et al. Enhanced anti-inflammatory activity of a lipo-somal intercellular adhesion molecule-1 antisense oligodeoxy-nucleotide in an acute model of contact hypersensitivity. J Phar-macol Exp Ther 2000;292:480–8.

18. Heyes J, Palmer L, Bremner K, MacLachlan I. Cationic lipidsaturation influences intracellular delivery of encapsulated nucleicacids. J Control Release 2005;107:276–87.

19. Hafez IM, Cullis PR. Cholesteryl hemisuccinate exhibits pHsensitive polymorphic phase behavior. Biochim Biophys Acta2000;1463:107–14.

20. Forsbach A, Nemorin JG, Montino C, Muller C, Samulowitz U,Vicari AP, et al. Identification of RNA sequence motifs stimu-lating sequence-specific TLR8-dependent immune responses.J Immunol 2008;180:3729–38.

21. Scheule RK, St George JA, Bagley RG, Marshall J, Kaplan JM,Akita GY, et al. Basis of pulmonary toxicity associated withcationic lipid-mediated gene transfer to the mammalian lung.Hum Gene Ther 1997;8:689–707.

22. Freimark BD, Blezinger HP, Florack VJ, Nordstrom JL, Long SD,Deshpande DS, et al. Cationic lipids enhance cytokine and cellinflux levels in the lung following administration of plasmid:cationic lipid complexes. J Immunol 1998;160:4580–6.

23. Song YK, Liu F, Chu S, Liu D. Characterization of cationicliposome-mediated gene transfer in vivo by intravenous adminis-tration. Hum Gene Ther 1997;8:1585–94.

24. Yu RZ, Kim TW, Hong A, Watanabe TA, Gaus HJ, Geary RS.Cross-species pharmacokinetic comparison from mouse to man ofa second-generation antisense oligonucleotide, ISIS 301012, tar-


geting human apolipoprotein B-100. Drug Metab Dispos 2007;35:460–8.

25. Geary RS, Watanabe TA, Truong L, Freier S, Lesnik EA, SioufiNB, et al. Pharmacokinetic properties of 2�-O-(2-methoxyethyl)-modified oligonucleotide analogs in rats. J Pharmacol Exp Ther2001;296:890–7.

26. Williams RO. Collagen-induced arthritis in mice: a major role fortumor necrosis factor-�. Methods Mol Biol 2007;361:265–84.

27. Feldmann M, Maini RN. TNF defined as a therapeutic target forrheumatoid arthritis and other autoimmune diseases [publishederratum appears in Nat Med 2003;9:1433]. Nat Med 2003;9:1245–50.

28. Nielsen PE. Systemic delivery: the last hurdle? Gene Ther 2005;12:956–7.

29. Hanai K, Kurokawa T, Minakuchi Y, Maeda M, Nagahara S,Miyata T, et al. Potential of atelocollagen-mediated systemicantisense therapeutics for inflammatory disease. Hum Gene Ther2004;15:263–72.

30. Khoury M, Louis-Plence P, Escriou V, Noel D, Largeau C, CantosC, et al. Efficient new cationic liposome formulation for systemicdelivery of small interfering RNA silencing tumor necrosis factor� in experimental arthritis. Arthritis Rheum 2006;54:1867–77.

31. Peer D, Park EJ, Morishita Y, Carman CV, Shimaoka M. Systemicleukocyte-directed siRNA delivery revealing cyclin D1 as ananti-inflammatory target. Science 2008;319:627–30.

32. Rozema DB, Lewis DL, Wakefield DH, Wong SC, Klein JJ,Roesch PL, et al. Dynamic polyconjugates for targeted in vivodelivery of siRNA to hepatocytes. Proc Natl Acad Sci U S A2007;104:12982–7.

33. Durie FH, Fava RA, Foy TM, Aruffo A, Ledbetter JA, Noelle RJ.Prevention of collagen-induced arthritis with an antibody to gp39,the ligand for CD40. Science 1993;261:1328–30.

34. Kyburz D, Carson DA, Corr M. The role of CD40 ligand andtumor necrosis factor � signaling in the transgenic K/BxN mousemodel of rheumatoid arthritis. Arthritis Rheum 2000;43:2571–7.

35. Liu YJ. IPC: professional type 1 interferon-producing cells andplasmacytoid dendritic cell precursors. Annu Rev Immunol 2005;23:275–306.

36. Song E, Zhu P, Lee SK, Chowdhury D, Kussman S, DykxhoornDM, et al. Antibody mediated in vivo delivery of small interferingRNAs via cell-surface receptors. Nat Biotechnol 2005;23:709–17.

37. MacDonald KP, Nishioka Y, Lipsky PE, Thomas R. FunctionalCD40 ligand is expressed by T cells in rheumatoid arthritis. J ClinInvest 1997;100:2404–14.

38. Sekine C, Yagita H, Miyasaka N, Okumura K. Expression andfunction of CD40 in rheumatoid arthritis synovium. J Rheumatol1998;25:1048–53.

39. Wagner DH Jr, Stout RD, Suttles J. Role of the CD40-CD40ligand interaction in CD4� T cell contact-dependent activation ofmonocyte interleukin-1 synthesis. Eur J Immunol 1994;24:3148–54.

40. Harigai M, Hara M, Kawamoto M, Kawaguchi Y, Sugiura T,Tanaka M, et al. Amplification of the synovial inflammatoryresponse through activation of mitogen-activated protein kinasesand nuclear factor �B using ligation of CD40 on CD14� synovialcells from patients with rheumatoid arthritis. Arthritis Rheum2004;50:2167–77.

41. Quezada SA, Jarvinen LZ, Lind EF, Noelle RJ. CD40/CD154interactions at the interface of tolerance and immunity. Annu RevImmunol 2004;22:307–28.

42. Cella M, Scheidegger D, Palmer-Lehmann K, Lane P, Lanzavec-chia A, Alber G. Ligation of CD40 on dendritic cells triggersproduction of high levels of interleukin-12 and enhances T cellstimulatory capacity: T-T help via APC activation. J Exp Med1996;184:747–52.

43. Andreakos E, Smith C, Monaco C, Brennan FM, Foxwell BM,Feldmann M. I�B kinase 2 but not NF-�B-inducing kinase isessential for effective DC antigen presentation in the allogeneicmixed lymphocyte reaction. Blood 2003;101:983–91.

44. Feldmann M, Steinman L. Design of effective immunotherapy forhuman autoimmunity. Nature 2005;435:612–9.

45. Pluvinet R, Petriz J, Torras J, Herrero-Fresneda I, Cruzado JM,Grinyo JM, et al. RNAi-mediated silencing of CD40 preventsleukocyte adhesion on CD154-activated endothelial cells. Blood2004;104:3642–6.

46. Gao D, Wagner AH, Fankhaenel S, Stojanovic T, Schweyer S,Panzner S, et al. CD40 antisense oligonucleotide inhibition oftrinitrobenzene sulphonic acid induced rat colitis. Gut 2005;54:70–7.

47. Stojanovic T, Scheele L, Wagner AH, Middel P, Bedke J, Lau-tenschlager I, et al. STAT-1 decoy oligonucleotide improvesmicrocirculation and reduces acute rejection in allogeneic ratsmall bowel transplants. Gene Ther 2007;14:883–90.

48. Boumpas DT, Furie R, Manzi S, Illei GG, Wallace DJ, Balow JE,et al. A short course of BG9588 (anti–CD40 ligand antibody)improves serologic activity and decreases hematuria in patientswith proliferative lupus glomerulonephritis. Arthritis Rheum 2003;48:719–27.


Amphoteric liposomes enable systemic antigen-presenting cell-directed delivery of CD40 antisense and...

Documents

Transcript of Amphoteric liposomes enable systemic antigen-presenting cell-directed delivery of CD40 antisense and...