Immunotherapeutic activity of a conjugateof a Toll-like receptor 7 ligandChristina C. N. Wu*, Tomoko Hayashi*, Kenji Takabayashi†‡, Mojgan Sabet†, Donald F. Smee§, Donald D. Guiney†,Howard B. Cottam*, and Dennis A. Carson*¶

*The Rebecca and John Moores Cancer Center and †Department of Medicine, University of California at San Diego, La Jolla, CA 92093-0820; and §Institutefor Antiviral Research, Department of Animal, Dairy, and Veterinary Science, Utah State University, 5600 Old Main Hill, Logan, UT 84322-5600

Contributed by Dennis A. Carson, December 29, 2006 (sent for review December 15, 2006)

The immunotherapeutic activity of Toll-like receptor (TLR) activa-tors has been difficult to exploit because of side effects related tothe release and systemic dispersion of proinflammatory cytokines.To overcome this barrier, we have synthesized a versatile TLR7agonist, 4-[6-amino-8-hydroxy-2-(2-methoxyethoxy)purin-9-ylm-ethyl]benzaldehyde (UC-1V150), bearing a free aldehyde thatcould be coupled to many different auxiliary chemical entitiesthrough a linker molecule with a hydrazine or amino groupwithout any loss of activity. UC-1V150 was covalently coupled tomouse serum albumin (MSA) at a 5:1 molar ratio to yield a stablemolecule with a characteristically altered UV spectrum. Comparedwith the unconjugated TLR7 agonist, the UC-1V150/MSA was a 10-to 100-fold more potent inducer of cytokine production in vitro bymouse bone marrow-derived macrophage and human peripheralblood mononuclear cells. When administrated to the lung, theconjugate induced a prolonged local release of cytokines at levels10-fold or more higher than those found in serum. Under the sameconditions, the untethered TLR7 ligand induced quick systemiccytokine release with resultant toxicity. In addition, two pulmo-nary infectious disease models were investigated wherein micewere pretreated with the conjugate and then challenged witheither Bacillus anthracis spores or H1N1 influenza A virus. Signif-icant delay in mortality was observed in both disease models withUC-1V150/MSA-pretreated mice, indicating the potential useful-ness of the conjugate as a localized and targeted immunothera-peutic agent.

drug delivery � influenza � innate immunity

The Toll-like receptors (TLRs) are a set of conservedcellular receptors that play an important role in the

recognition of microbial pathogens and in initiating the hostinnate immune response. These receptors recognize distinctmolecular components of invading pathogens, such as cell wallstructures and nucleic acids. The discovery that endogenousligands as well as synthetic small molecules can activate certainTLR pathways has generated tremendous interest in thedevelopment of new therapeutics for diseases related to theimmune response. TLR ligands control the activation ofantigen-presenting cells, in particular dendritic cells, by trig-gering their maturation program, including up-regulation ofthe expression of HLA and costimulatory molecules andsecretion of proinf lammatory cytokines, such as TNF-�, IL-6,IL-12, and IFN-� (1). Recently, we reported that certainderivatives of guanine (2) and adenine can activate immunecells via TLR7 and can inhibit the replication of hepatitis Cvirus replicons in hepatocytes (3). However, the in vitroimmunotherapeutic activities of TLR7 ligands have beendifficult to translate in vivo, because systemic TLR activationcan induce a rapid and potentially toxic cytokine syndrome (4,5). Accordingly the major in vivo applications of TLR7 ligandshave been as topically applied antiviral or antitumor agents oras immune adjuvants injected intramuscularly in small quan-tities (6, 7). Previously, we found that small molecule agonistsof TLR7 enhanced the ability of macrophages to control

Bacillus anthracis in vitro (8). However, to develop a safe andeffective immunotherapeutic agent for pulmonary infection,we needed a TLR7 agonist that would activate cytokineproduction in the lung but without causing systemic cytokinerelease. One potential effective strategy involved the stableconjugation of the agent to a macromolecule, such as a proteinor a polymer (9), that would restrict systemic absorption;promote local drug uptake into endosomes, where TLR7resides (10); and perhaps deliver the immunotherapeutic tospecific cells or organs (11, 12). In this regard, a recent reportshowed that conjugation of HIV Gag protein to a low-molecular-weight TLR7/8 agonist yielded a vaccine that elic-ited broad-based adaptive immunity in nonhuman primates(13). In the present study, we examined the effects of conju-gation of a nonimmunogenic carrier protein to a modifiedadenine-based TLR7 agonist on the magnitude and durationof innate immune activation in vitro and in vivo. Because amouse model was selected for study in vivo, a covalentconjugate of a TLR7 agonist and mouse serum albumin (MSA)connected through a versatile bifunctional linker molecule wasprepared and characterized. The conjugate was 10- to 100-foldmore potent than the free drug. When administrated to thelung, it induced profound local cytokine release, with minimalsystemic dispersal. Mice that were pretreated by intrapulmo-nary delivery of the conjugate and then challenged with eitherB. anthracis spores or H1N1 inf luenza A virus had a delayedand attenuated course of infection.

ResultsSynthesis and Conjugation of UC-1V150 to MSA. UC-1V150 (8) wassynthesized in our laboratory in seven steps from 2,6-dichloropurine (Fig. 1). The free aldehyde group on the benzylmoiety of UC-1V150 enabled us to couple the agonist to manydifferent auxiliary chemical entities, including proteins, oligo-nucleotides, aromatic molecules, lipids, viruses, and cells,through a linker molecule that contained a hydrazine or aminogroup. UC-1V150 was covalently coupled to MSA first mod-ified with a succinimidyl 6-hydrazino-nicotinamide acetonehydrazone (SANH) linker to yield a stable molecule with acharacteristically altered UV spectrum (Fig. 2). The UC-1V150/MSA conjugate was identified by a UV absorption peakat 342 nm due to hydrazone formation, whereas SANH alone

Author contributions: C.C.N.W., T.H., D.D.G., H.B.C., and D.A.C. designed research;C.C.N.W., T.H., K.T., M.S., and D.F.S. performed research; C.C.N.W. analyzed data; andC.C.N.W., H.B.C., and D.A.C. wrote the paper.

The authors declare no conflict of interest.

Abbreviations: TLR, Toll-like receptor; MSA, mouse serum albumin; BMDM, bone-marrow-derived macrophages; BALF, bronchial alveolar lavage fluid; i.t., intratracheal(ly); i.n.,intranasal(ly); SANH, succinimidyl 6-hydrazino-nicotinamide acetone hydrazone.

‡Deceased August 4, 2006.

¶To whom correspondence should be addressed at: Department of Medicine, University ofCalifornia at San Diego, 3855 Health Sciences Drive, La Jolla, CA 92093-0820. E-mail:dcarson@ucsd.edu.

© 2007 by The National Academy of Sciences of the USA

3990–3995 � PNAS � March 6, 2007 � vol. 104 � no. 10 www.pnas.org�cgi�doi�10.1073�pnas.0611624104

Dow

nloa

ded

by g

uest

on

Aug

ust 2

4, 2

021

absorbed at 322 nm. Quantification of UC-1V150 moleculesconjugated per MSA was extrapolated from a standard curveof UC-1V150-SANH (Fig. 3). We consistently obtained theUC-1V150/MSA conjugates at a ratio of �5:1. The biologicalstudies reported here were done by using 5:1 UC-1V150/MSA.

Potent in Vitro and in Vivo Cytokine Release in Response to UC-1V150/MSA Conjugates. Incubation of bone-marrow-derived macro-phages (BMDM) with UC-1V150 alone stimulated cytokinerelease (Fig. 4). When conjugated to MSA, similar or higherlevels of cytokines were detected with a 10-fold lower equivalentconcentration of the pharmacophore (Fig. 4). Experiments withTLR transformants, performed as described previously, con-firmed that UC-1V150, similar to the compound lacking thealdehyde modification (UC-1V136), was a specific TLR7 agonist

(3). After i.v. injection into mice, UC-1V150 induced serumcytokine levels that peaked at �2 h after injection and thenquickly declined to near background levels (data not shown).Comparison of the cytokine production profiles of UC-1V150versus the UC-1V150/MSA 2 h after i.v. injection at variousdosages demonstrated that the MSA conjugate enhanced thepotency by 10- to 100-fold (Fig. 5). Sera from saline or MSAcontrol groups revealed little or no detectable cytokine levels(data not shown).

UC-1V150/MSA Conjugates Provide Prolonged and Localized Pulmo-nary Activity. To ensure adequate delivery of the TLR7 agoniststo the respiratory system, we initially delivered the drugsdirectly into the trachea. Substantial cytokine induction wasfound in bronchial alveolar lavage f luid (BALF) extractedfrom mice treated intratracheally (i.t.) with UC-1V150/MSA(Fig. 6 Left), whereas serum cytokines were very low and nearbackground levels in the same animals (Fig. 6 Right). Inmarked contrast, similar levels of cytokine were observed inboth BALF and sera of mice injected i.t. with small-moleculeTLR7 agonists, which sometimes induced behavioral changes,such as hair standing on end and shivering, suggestive of acytokine syndrome (Table 1). Subsequent studies with UC-1V150 showed that intranasal (i.n.) delivery also inducedselective cytokine production in the BALF, probably due todrug aspiration. Accordingly, i.n. administration was used toevaluate the UC-1V150 conjugates in two infectious animalmodels of pneumonitis. Mice pretreated i.n. with UC-1V150/MSA one day before infection with B. anthracis spores had anextended mean survival of 7.5 days compared with 5 days incontrol mice (P � 0.025) (Fig. 7A). In contrast, no significantdifference was observed in mice treated with either saline, theequivalent amount of MSA, or with UC-1V150 alone. Thesedata confirmed that the UC-1V150 conjugate, but not the freedrug, had intrapulmonary immunotherapeutic activity. In an-other study, BALB/c mice were pretreated i.n. with theUC-1V150/MSA conjugate 1 day before inf luenza virus infec-tion (H1N1 strain). The mean survival of the treated mice wasextended to 11.5 days compared with 7 days in untreatedcontrols (P � 0.0001) (Fig. 7B). Together these results suggestthat conjugation of the TLR7 agonist to MSA enhanced itspotency and reduced its toxicity after local delivery to therespiratory tract.

DiscussionThe compound UC-1V150 is one of the most potent and versatilesynthetic small-molecule TLR7 ligands yet discovered because(i) it is active at nanomolar concentrations; (ii) it can be coupledto a variety of macromolecules with enhancement of activity insome cases; and (iii) its pharmacokinetic properties can bechanged by modification of the auxiliary groups. The TLR7-protein conjugate UC-1V150/MSA was characterized as havingapproximately five small molecules covalently linked to eachMSA protein molecule. The conjugate retained TLR7 agonistactivity and indeed was both more potent and had a longerduration of action, compared with the free monomeric drug.Moreover, we showed that this conjugate could be deliveredeffectively to the respiratory system by i.n. or i.t. administration.Drug delivery by i.n. proved to be effective in two mouse modelsof infectious disease, a bacterial infection and a viral infection.When considering delivery to the respiratory system, a poten-tially important advantage of preparing the TLR7 agonists asconjugates of macromolecules is that systemic side effects may beavoided by confining the immunostimulatory activity to the localmucosal environment.

The macromolecular conjugate would be expected to beabsorbed into the systemic circulation more slowly than the freedrug and, indeed, may be avidly scavenged by resident macro-

Fig. 1. Synthesis of UC-1V150. Reagents and conditions were as follows:�-bromo-p-tolunitrile, K2CO3/dimethylformamide, 25°C (step a); NH3-MeOH,60°C (step b); CH3OCH2CH2CH2OH, 100°C (step c); Br2/CH2Cl2, 25°C (step d);CH3ONa/CH3OH, reflux (step e); lithium N,N�-diemthylenediamino aluminumhydride/THF, 0°C (step f); and HCl, 25°C (step g).

Fig. 2. Conjugation of UC-1V150 to MSA. Covalent linkage is formed via alinker, SANH, between the UC-1V150 and MSA containing an amino group.

Wu et al. PNAS � March 6, 2007 � vol. 104 � no. 10 � 3991

IMM

UN

OLO

GY

Dow

nloa

ded

by g

uest

on

Aug

ust 2

4, 2

021

phages and dendritic cells expressing TLR7. Accordingly, theconjugate should mitigate the type of severe side effects thathave been associated with systemic delivery of TLR7/8 agonists.�The UC-1V150/MSA conjugate may also provide beneficialimmunotherapeutic activity when administered to mucosal sites,such as the genitourinary and gastrointestinal tracts, for thecontrol of infectious, allergic, or malignant diseases. The mac-romolecular carrier of the TLR7 agonist may also provide animproved approach for selective delivery of the immunothera-peutic to a specific organ or tissue. For example, the lipidconjugates of UC-1V150 can be incorporated into liposomes ofdifferent size and composition, whereas protein conjugates ofthe TLR7 agonist may target different dendritic cell subsets.Differences in the intracellular trafficking of the UC-1V150conjugate may induce distinct patterns of cytokine production,analogous to the effects observed with TLR9-activating oligo-nucleotides (14).

One potential problem that has been observed with drugsconjugated to proteins is the development of antibodies againstthe low-molecular-weight hapten-like portion of the molecule.However, UC-1V150, unlike the TLR7/8 vaccine conjugatesstudied earlier, has a simple adenine-like structure that isunlikely to induce hypersensitivity reactions. Indeed, we have notobserved anti-UC-1V150 antibodies after administration of theprotein conjugates, except after repeated administration of akeyhole limpet hemocyanin carrier in complete Freud’s adjuvant(unpublished data).

New agents for the prevention and treatment of inf luenzavirus infections are being sought, particularly with the spreadof highly pathogenic strains from Asia. Morbidity and mor-tality from commonly circulating strains is high each year.Treatment of the infection can be accomplished by approvedantiviral drugs, which are moderately effective if started early.Enhancement of the immune system is also being investigatedas a strategy that could accelerate protective antiviral re-sponses, especially in immune compromised hosts. Along theselines, the immunomodulatory agent UC-1V150/MSA conju-gate was tested against an inf luenza A (H1N1) virus infectionin mice. It has been our experience that treatment of inf luenzavirus and bacterial infections with immune modulators is oftennot effective. It is possible that systemic immune activation viaTLR signaling does not create a local cytokine and chemokinegradient required to mobilize immune cells to the site ofinfection. In support of this hypothesis, the unconjugatedUC-1V150, which is rapidly absorbed through the mucosa,failed to protect mice from B. anthracis infection, whereas theUC-1V150 conjugate was effective.

B. anthracis has become an agent of bioterrorism. A rapidresponse against microbial pathogens is critical for effectivebiodefense. In general an antibody or cellular immune re-

sponse may protect against these pathogens; however, gener-ating these protective responses quickly requires prior expo-sure to specific antigens for each organism. Although it isknown that inf luenza virus engages TLR7 (15), bacterialanthrax most likely can engage TLR2, TLR4, and TLR9. Inaddition to being a common signaling intermediary for theTLRs, MyD88 has also been shown to be necessary forresistance to infection in a mouse model of anthrax (16).Because the UC-1V150 conjugate works effectively as anadjuvant against infections that use different pathways, it canbe applied as a biodefense strategy that would not need bespecific to the antigens of a particular microbe and that wouldbe useful in mixed as well as single-agent attacks.

Materials and MethodsChemistry of UC-1V150. The synthesis of UC-1V150 is depicted inFig. 1, and the preparation of the indicated compounds 2–8 wasas follows.Compound 2: 4-(2,6-dichloropurin-9-ylmethyl)benzonitrile. Sixteen mil-limoles of 2,6-dichloro-9H-purine (compound 1) was dissolved in50 ml of dimethylformamide with potassium 50 mmol of car-bonate added, and the mixture was stirred at ambient temper-ature for 16 h after adding 22 mmol of �-bromo-p-tolunitrile.After filtration to remove insoluble inorganic salts, the filtratewas poured into 1,500 ml of water and extracted with ethylacetate (2 � 400 ml), dried over magnesium sulfate, andevaporated to yield a residue that was subjected to flash silica gelchromatography using 1:2:10 ethyl acetate/acetone/hexanes.[Yield, 3.33 g (69%).] UV, NMR, and MS were consistent withstructure assignment.Compound 3: 4-(6-amino-2-chloropurin-9-ylmethylbenzonitrile. Com-pound 2 (1.9 g) was placed in a steel reaction vessel, and 80 mlof 7 M methanolic ammonia was added. The sealed vessel washeated at 60°C for 12 h and cooled in ice, and the solid productwas filtered off. (Yield, 1.09 g.) UV, NMR, and MS wereconsistent with assigned structure.Compound 4: 4-[6-amino-2-(2-methoxyethoxy)purin-9-ylmethyl]benzoni-trile. The sodium salt of 2-methoxyethanol was first generated bydissolving 81 mg of sodium metal in 30 ml of 2-methoxyethanol

�Pockros P., Tong M., Wright T. (2003) Gastroenterology 124(Suppl. 1):76 (abstr.).

Fig. 3. Quantification of UC-1V150 molecules conjugated to MSA. (A) Different concentrations of UC-1V150-SANH were subjected to spectrophotometric analysis.(B) The readout of UC-1V150-SANH was used to derive a standard curve. (C) Conjugation of UC-1V150 to MSA was indicated by a UV absorption peak at 342 nm dueto hydrazone formation. The concentration of UC-1V150/MSA used in this analysis at 0.25 �g was based on MSA concentration, which is equivalent to 3.63 �M MSA.The concentration of UC-1V150 in conjugate was calculated as (0.6901 � 0.0056)/0.0366 � 18.7 �M; thus, the UC-1V150:MSA ratio is 18.7/3.63 � 5.2.

Fig. 4. In vitro cytokine release in response to UC-1V150/MSA conjugates.Murine BMDM were treated with UC-1V150 or UC-1V50/MSA conjugates atvarious concentrations as indicated. Culture supernatants were harvested 24 hlater, and cytokine levels were measured by immunoassay. The results are arepresentative of at least two separate experiments in triplicate per treatment.

3992 � www.pnas.org�cgi�doi�10.1073�pnas.0611624104 Wu et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 2

4, 2

021

with heat, and then 1.0 g of compound 3 dissolved in 300 ml ofmethoxyethanol was added with heat. The reaction mixture washeated for 8 h at 115°C bath temperature and concentrated invacuo to near dryness; the residue was then partitioned betweenethyl acetate and water. Flash silica gel chromatography of theorganic layer by using 5% methanol in dichloromethane gave 763mg of product. NMR was consistent with structure assignment.Compound 5: 4-[6-amino-8-bromo-2-(2-methoxyethoxy)purin-9-ylmethyl]-benzonitrile. Compound 4 (700 mg) was dissolved in dichlo-romethane (400 ml) and bromine (7 ml) was added dropwise.The mixture was stirred overnight at room temperature andextracted first with 2 liters of a 0.1 M aqueous sodium thiosulfatesolution, then with 500 ml of aqueous sodium bicarbonate(saturated). The residue from the organic layer was chromato-graphed on silica gel by using 3% methanol in dichloromethaneto yield 460 mg of bromo product. NMR, UV, and MS wereconsistent with structure assignment.Compound 6: 4-[6-amino-8-methoxy-2-(2-methoxyethoxy)purin-9-ylmethyl]benzonitrile. Sodium methoxide was generated by reac-tion of 81 mg of sodium metal in 30 ml of dry methanol and

combined with a 700-mg solution of compound 5 dissolved in drydimethoxyethane, and the temperature raised to 100°C. Afterovernight reaction, the mixture was concentrated in vacuo, andthe residue was chromatographed on silica by using 5% methanolin dichloromethane. (Yield, 120 mg.) NMR was consistent withstructure assignment.Compound 7: 4-[6-amino-8-methoxy-2-(2-methoxyethoxy)-purin-9-ylmethyl]benzaldehyde. Compound 6 (100 mg) was dissolved in 3 mlof dry THF and cooled to 0°C under argon. The reducing agent,lithium N,N�-(dimethylethylenediamino)-aluminum hydride,used to convert the nitrile to the aldehyde function was preparedessentially as previously described (17). A 0.5 M solution in dryTHF was prepared, and 0.72 ml of it was added to the reactionflask. The mixture was stirred at 0–5°C for 1 h, quenched byaddition of 3 M HCl, extracted with ethyl acetate followed bydichloromethane, and then concentrated in vacuo to yield 85 mg.NMR was consistent with structure assignment.Compound 8: 4-[6-amino-8-hydroxy-2-(2-methoxyethoxy)purin-9-ylmethyl]benzaldehyde (UC-1V150). Compound 7 (800 mg) wascombined with 504 mg of sodium iodide and 40 ml of aceto-nitrile, and 0.5 ml of chlorotrimethylsilane was slowly added.The mixture was heated at 70°C for 3.5 h, cooled, and filtered.The solid product was washed with water and then ether toyield 406 mg. NMR, UV, and MS were consistent withstructure assignment.

Conjugation of UC-1V150 to MSA. The benzaldehyde group onUC-1V150 enabled facile coupling to a variety of auxiliarymolecules with free amino or hydrazine groups by using wellestablished conjugation chemistry. To prepare a conjugate thatcould be administrated to mice and conveniently monitoredspectrophotometrically, we chose SANH as a linker and MSA(Sigma, St. Louis, MO) as the auxiliary molecule (Fig. 2).Initially, the amino groups on MSA derivatized with SANHaccording to standard procedures (18). To remove excess SANH,the reaction mixture was loaded on a NAP-10 column equili-brated with PBS and eluted with PBS. Then the UC-1V150 wasadded to the modified proteins at a 40-fold molar excess indimethylformamide and incubated at 20°C overnight. UnreactedUC-1V150 was removed by size-exclusion gel filtration as de-scribed above. The lyophilized conjugate was stable for at least6 months at �20°C.

Mice. Female C57BL/6 mice (5–6 wk of age) were obtainedfrom Harlan West Coast (Germantown, CA), and female A/Jmice (6–8 wk of age) were purchased from The JacksonLaboratories (Bar Harbor, ME). A/J mice were used forinfection with the Sterne strain of B. anthracis (19). The micewere bred and maintained under standard conditions in theUniversity of California at San Diego Animal Facility, whichis accredited by the American Association for Accreditation ofLaboratory Animal Care. All animal protocols received priorapproval by the Institutional Review Board. For the H1N1inf luenza study, female BALB/c mice (16–18 g) were obtainedfrom Charles River Laboratories (Wilmington, MA) and main-tained in the American Association for Accreditation of

Fig. 5. In vivo efficacy of UC-1V150/MSA conjugates. C57BL/6 mice were administered various amounts of UC-1V150 or UC-1V150/MSA via the tail vein as indicated.Sera were collected 2 h later, and cytokine levels were determined by multiplex immunoassay. Each group had four mice. The error bars indicate the SEM.

Fig. 6. Sustained in vivo local activity of UC-1V150/MSA conjugates withoutsystemic effect. C57BL/6 mice were anesthetized and administered i.t. with 3nmol of UC-1V150/MSA. At the indicated time points, mice were killed andboth BALF and sera were collected for multiplex immunoassay of cytokinelevels. The data were combined from two separate experiments with at leastsix mice per group. The results show the mean values � SEM.

Wu et al. PNAS � March 6, 2007 � vol. 104 � no. 10 � 3993

IMM

UN

OLO

GY

Dow

nloa

ded

by g

uest

on

Aug

ust 2

4, 2

021

Laboratory Animal Care-accredited Laboratory Animal Re-search Center of Utah State University.

In Vitro Stimulation of BMDM. BMDM were isolated from variousstrains of mice and grown as previously described (20) andwere seeded in 96-well plates at a density of 5 � 104 cells perwell. Compounds were added to 10-day-old cultures at a finalconcentration ranging from 0.01 to 10 �M or as otherwiseindicated. After 24 h of incubation, culture supernatants werecollected and assayed for cytokine inductions by either sand-wich ELISA (BD Pharmingen, San Diego, CA) (20) or mul-tiplex Luminex (Austin, TX) assay using the Beadlyte MouseMultiCytokine customized kit (Upstate, Charlottesville, VA,and eBiosciences, San Diego, CA), according to the manufac-turer’s instructions.

Administration of Compounds to Mice. Female age-matchedC57BL/6 mice were injected with 100 �l of saline solution contain-

ing UC-1V150 or UC-1V150/MSA, each containing the equivalentof 0.38–38 nmol of the pharmacore via the tail vein. For intrapul-monary administration, mice were anesthetized with i.p. Avertinsolution and shaved around the neck area. The trachea wereexposed with a small incision and injected with 50 �l of salinesolution containing various amounts of UC-1V150/MSA or theunconjugated drug. After recovery and at different time points,serum and BALF were collected and analyzed for IL-6, IL-12p40,IFN-�, RANTES, and MCP-1 by Luminex assay. In other exper-iments, mice were anesthetized with an intramuscular ketamine/xylene solution and administered the same amount of UC-1V150/MSA in i.t. doses of 50 �l or i.n. doses of 20 �l. Because similarcytokine levels were observed in the BALF 24 h after administra-tion by either method, the more convenient i.n. route was used ininfectious model studies.

Infection of A/J Mice with B. anthracis Spores. Spores were preparedfrom the Sterne strain of B. anthracis (pXO1�pXO2�) aspreviously described (8, 21). Purified spores were stored inPBS at 1 � 108 to 4 � 108 cfu/ml at 4°C. Before infection, thespores were heated to 65°C for 30 min to initiate germination.A/J mice were anesthetized intramuscularly with ketamine/xylene solution and administered i.n. with 0.75 nmol ofUC-1V150 or UC-1V150/MSA per mouse 1 day before anthraxinfection. Control mice received saline only or saline contain-ing MSA at equivalent amounts as in UC-1V150/MSA. Infec-tion was carried out i.n. with 2 � 105 to 8 � 105 spores of B.anthracis in a 20-�l volume. Survival was observed for 13 days,because the majority of the saline-treated mice died within 3–6days. Results were obtained from eight mice per group.

Infection of BALB/c Mice with Influenza Virus. Influenza A/NewCaledonia/20/99 (H1N1) virus was obtained from the Centers forDisease Control and Prevention (Atlanta, GA). The virus waspropagated twice in Madin Darby canine kidney cells, furtherpassaged seven times in mice to make it virulent, followed byanother passage in cell culture to amplify it. Mice were anesthetizedi.p. with 100 mg/kg ketamine and infected i.n. with virus at �105.0

cell culture infectious doses per mouse in a 50-�l inoculum volume.A single 75-�l i.n. dose of either saline alone or saline containingUC-1V150/MSA to 5 nmol per mouse was given 24 h before virusexposure. Ten infected mice per treated group and 20 placebocontrol animals were followed for survival for 21 days.

Table 1. Local cytokine profile in mice treated with conjugated vs. unconjugated TLR7agonists

Conjugated Unconjugated

Cytokine Time, hBALF,

ng/ml (SEM)Serum,

ng/ml (SEM) RatioBALF,

ng/ml (SEM)Serum,

ng/ml (SEM) Ratio

IL-6 2 0.25 (0.09) 0.28 (0.00) 0.9 0.73 (0.40) 5.00 (1.14) 0.14 1.34 (0.38) 0.06 (0.02) 22.1 1.09 (0.64) 2.93 (1.53) 0.46 3.50 (1.07) 0.14 (0.05) 24.3 3.00 (0.76) 3.95 (1.24) 0.8

24 4.06 (0.41) 0.03 (0.00) 117.2 4.39 (0.84) 1.08 (0.31) 4.1

IL-12p40 2 0.06 (0.00) 0.23 (0.03) 0.2 0.11 (0.09) 0.53 (0.53) 0.24 0.11 (0.00) 0.50 (0.39) 0.2 0.06 (0.03) 0.23 (0.23) 0.36 0.29 (0.04) 0.25 (0.06) 1.1 0.00 (0.00) 2.82 (0.46) 0.0

24 5.05 (3.70) 0.11 (0.00) 45.1 0.01 (0.01) 0.49 (0.47) 0.0

TNF-� 2 1.25 (1.02) 0.08 (0.01) 16.4 2.62 (0.83) 0.52 (0.21) 5.04 9.67 (1.01) 0.02 (0.01) 580.1 1.53 (0.73) 0.32 (0.16) 4.86 8.28 (1.72) 0.02 (0.00) 427.5 2.03 (0.67) 0.39 (0.15) 5.2

24 4.22 (2.80) 0.01 (0.00) 551.8 0.93 (0.47) 0.15 (0.04) 6.2

C57BL/6 mice were administered i.t. UC-1V150/MSA conjugate or unconjugated UC-1V136 at 3 nmole or 500nmole per mouse, respectively. BALF and sera were collected at the indicated time points, and cytokine levels weredetermined by multiplex immunoassay. Mean values from at least three to five mice per group are shown � SEM.

Fig. 7. Preclinical efficacy of UC-1V150/MSA in pulmonary infectious dis-eases. (A) Age-matched female A/J mice were administered i.n. saline only orsaline containing MSA (amount equivalent to UC-1V150/MSA), UC-1V150, orUC-1V150/MSA at 0.75 nmol per mouse 1 day before B. anthracis infection,and survival was followed for 13 days. (B) BALB/c mice were administered i.n.saline or UC-1V-150/MSA at 5 nmol per mouse 1 day before influenza infec-tion. Survival followed up to 21 days. In each model, Kaplan–Meier survivalcurves and log-rank tests were performed to determine significance. At leasteight mice were tested in each group.

3994 � www.pnas.org�cgi�doi�10.1073�pnas.0611624104 Wu et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 2

4, 2

021

Statistics. Cytokine levels were compared by the Mann–Whitney U test with P � 0.05 to determine statistical signif-icance. Kaplan–Meier survival curves and log-rank tests wereperformed by using Prism software version 4.0c (GraphPad,San Diego, CA) to compare differences in survival.

We thank Drs. Eyal Raz, Maripat Corr, Jongdae Lee, and Guangyi Jinfor helpful advice and Rommel Tawatao, Michael Chan, Andrea Agu-irre, Christine Gray, and Chuong Dang for technical assistance. Thiswork was supported in part by National Institutes of Health GrantsAI56453, AI40682, AR44850, AR07567, and GM23200.

1. Stanley MA (2002) Clin Exp Dermatol 27:571–577.2. Lee J, Chuang T-H, Redecke V, She L, Pitha PM, Carson DA, Raz E, Cottam

HB (2003) Proc Natl Acad Sci USA 100:6646–6651.3. Lee J, Wu CCN, Lee KJ, Chuang T-H, Katakura K, Liu Y-T, Chan M, Tawatao

R, Chung M, Shen C, et al. (2006) Proc Natl Acad Sci USA 103:1828–1833.4. Cristofaro P, Opal SM (2003) Expert Opin Ther Targets 7:603–612.5. Knuefermann P, Nemoto S, Baumgarten G, Misra A, Sivasubramanian N,

Carabello BA, Vallejo JG (2002) Chest 121:1329–1336.6. Ambach A, Bonnekoh B, Nguyen M, Schon MP, Gollnick H (2004) Mol

Immunol 40:1307–1314.7. Hemmi H, Kaisho T, Takeuchi O, Sato S, Sanjo H, Hoshino K, Horiuchi T,

Tomizawa H, Takeda K, Akira S (2002) Nat Immunol 3:196–200.8. Sabet M, Cottam HB, Guiney DG (2006) FEMS Immunol Med Microbiol

47:369–379.9. Takakura Y, Mahato RI, Nishikawa M, Hashida M (1996) Adv Drug Delivery

Rev 19:377–399.10. Sioud M (2006) Trends Mol Med 12:167–176.11. Blakey DC (1992) Acta Oncol 31:91–97.12. Fiume L, Busi C, Mattioli A, Spinosa G (1988) Crit Rev Ther Drug Carrier Sys

4:265–284.

13. Wille-Reece U, Flynn BJ, Lore K, Koup RA, Kedl RM, Mattapallil JJ, WeissWR, Roederer M, Seder RA (2005) Proc Natl Acad Sci USA 102:15190–15194.

14. Rothenfusser S, Tuma E, Endres S, Hartmann G (2002) Hum Immunol63:1111–1119.

15. Barchet W, Krug A, Cella M, Newby C, Fischer JA, Dzionek A, Pekosz A,Colonna M (2005) Eur J Immunol 35:236–242.

16. Hughes MA, Green CS, Lowchyj L, Lee GM, Grippe VK, Smith MF, Jr, HuangLY, Harvill ET, Merkel TJ (2005) Infect Immun 73:7535–7540.

17. Cha JS, Jang SH, Kwon SY (2002) Bull Kor Chem Soc 23:1697–1698.18. Kozlov IA, Melnyk PC, Stromsborg KE, Chee MS, Barker DL, Zhao C (2004)

Biopolymers 73:621–630.19. Kenney RT, Yu J, Guebre-Xabier M, Frech SA, Lambert A, Heller BA,

Ellingsworth LR, Eyles JE, Williamson ED, Glenn GM (2004) J Infect Dis190:774–782.

20. Wu CC, Lee J, Raz E, Corr M, Carson DA (2004) J Biol Chem 279:33071–33078.

21. Guidi-Rontani C, Levy M, Ohayon H, Mock M (2001) Mol Microbiol 42:931–938.

Wu et al. PNAS � March 6, 2007 � vol. 104 � no. 10 � 3995

IMM

UN

OLO

GY

Dow

nloa

ded

by g

uest

on

Aug

ust 2

4, 2

021