Proceedings
Immunostimulatory activity of aminoalkyl glucosaminide 4-phosphates (AGPs): induction of protective innate immune responses by RC-524 and RC-529 Jory R. Baldridge1, Christopher W. Cluff1, Jay T. Evans1, Michael J. Lacy1, Jeffrey R. Stephens2, Valerie G. Brookshire1, Rong Wang1, Jon R. Ward1, Yvonne M. Yorgensen1, David H. Persing3, David A. Johnson1 Corixa Corporation, Hamilton, Montana, USA Department of Chemistry, Seattle University, Seattle, Washington, USA 3 Corixa Corporation, Seattle, Washington, USA 1
2
Earlier we showed that the structural requirements for adjuvanticity among the aminoalkyl glucosaminide 4-phosphate (AGP) class of synthetic immunostimulants may be less strict than those for other endotoxic activities, including the induction of nitric oxide synthase in murine macrophages and cytokine production in human whole blood. The known role of nitric oxide and pro-inflammatory cytokines in the activation of host defenses against infection prompted us to examine the ability of certain AGPs to enhance non-specific resistance in mice to Listeria monocytogenes and influenza infections as well as to stimulate the production of pro-inflammatory cytokines in mouse splenocytes, human PBMCs, and human U937 histiocytic lymphoma cells. Intranasal administration of RC-524 or RC-529 to mice 2 days prior to a lethal influenza challenge provided significant protection in each case. Similarly, the intravenous administration of these AGPs induced resistance to L. monocytogenes infection as measured by survival or reduction of bacteria in the spleen. Activation of the innate immune response by AGPs appears to involve activation of Toll-like receptor 4 (TLR4) because RC-524 failed to elicit a protective effect in C3H/HeJ mice which have a defect in TLR4 signaling or induce significant cytokine levels in C3H/HeJ splenocytes. Both AGPs also stimulated pro-inflammatory cytokine release in human cell cultures in a dose-dependent manner.
INTRODUCTION Lipid A, the active principle of bacterial lipopolysaccharide, is well known as a potent stimulator of host defense systems, both as a potent adjuvant for vaccine antigens1 and as an inducer of non-specific resistance (NSR) to infection in animal models.2 However, the profound
Received 19 July 2002 Revised 16 September 2002 Accepted 18 September 2002 Correspondence to: David A. Johnson PhD, Corixa Corporation, 553 Old Corvallis Rd, Hamilton, MT 59840, USA Tel: +1 406 375 2134; Fax: +1 406 363 6129; E-mail:
[email protected] Journal of Endotoxin Research, Vol. 8, No. 6, 2002 DOI 10.1179/096805102125001064
pyrogenicity and lethal toxicity of lipid A have precluded its use in human vaccines as well as its application as a monotherapeutic for enhancing host resistance to infection.1 Recently, we3,4 identified a new class of synthetic vaccine adjuvants structurally related to the major hexa-acyl component 1 present in the low toxicity 3-O-deacylated monophosphoryl lipid A (3-D-MPL) derived from Salmonella minnesota R595 lipid A (Fig. 1 [1]). Known chemically as -aminoalkyl 2-amino-2-deoxy-4-phosphono- -D-glucopyranosides (aminoalkyl glucosaminide 4-phosphates, AGPs) and possessing the general structure 2, the AGPs are synthetic mimetics of compound 1 in which the reducing sugar has been replaced with an N[(R)-3-n-alkanoyloxytetradecanoyl]aminoalkyl aglycon unit (Fig. 1 [2]). Certain members of this class of synthetic © W. S. Maney & Son Ltd
454
Baldridge, Cluff, Evans, Lacy, Stephens, Brookshire et al. OH
O
O
(HO)2PO O O O
O O
n-C 11 H 23 n-C 13H 27
O
O
HO HO
NH
NH
OH
n-C 11H 23 n-C 11H 23
O O
n-C 11 H 23R2O
O O
n-C 11H 23 n-C 15H 31
R1
O
R2O
O
O
OH
O (HO)2PO O O
O
n
NH
O
NH
R 2O
n-C 11H 23
n-C 11H 23
2 AGP generic structure 3 R1=H, R2=n-C9H19CO, n=1 (RC-524) 4 R1=H, R2=n-C13H27CO, n=1 (RC-529)
1 3-D-MPL major component Fig. 1. Chemical structures of synthetic and naturally derived lipid A derivatives.
immunostimulants were shown to enhance the production of tetanus toxoid-specific antibodies in mice and augment vaccine-induced cytotoxic T cells (CTLs) against EG.7ova target cells. However, the structural requirements for adjuvanticity appeared to be less strict than those for other endotoxic activities, including the induction of nitric oxide (iNOS) in murine peritoneal macrophages and pro-inflammatory cytokines in human whole blood. For example, RC-524 (Fig. 1 [3]) which possesses C10 acyl groups was approximately 5 times more potent than RC-529 (Fig. 1, [4]; C14 acyl groups) with respect to the induction of TNFand IL- in human whole blood, and exhibited 20 times greater iNOS activity, but elicited lower CTL responses to ovalbumin than RC-529. The potent adjuvanticity of RC529 vis-à-vis RC-524 in certain animal models may relate to physicochemical differences between the two compounds such as solubility and aggregation properties and
OTCBOC
O O O
(PhO)2 PO O
O
OPG
HO NH Troc
not to chain length per se. The clinical utility of RC-529 as a vaccine adjuvant was recently demonstrated in a phase III pivotal trial with a recombinant hepatitis B antigen.4 Because the induction of both iNOS and pro-inflammatory cytokines is believed to play an important role in the activation of host defense mechanisms,5,6 these initial results suggested that certain AGPs would be capable of enhancing host non-specific resistance to bacterial and viral infections. Accordingly, RC-524 and RC-529 were tested for their ability to enhance non-specific resistance in mice to challenge by Listeria monocytogenes, a facultative intracellular parasite, and infectious influenza virus. In addition, the ability of RC-524 and RC-529 to stimulate cytokine production in mouse splenocytes, human peripheral blood mononuclear cells (PBMCs), and human U937 histiocytic lymphoma cells (a premonocytic cell type) was evaluated.
O O R = n-C9H19CO or n-C13H27CO
NH Troc Cl
RO n-C11H23
5 PG = SE or TBDPS
6 HO
O
Ag
O O
O
HO NH Troc 9
Fig. 2. Synthesis of RC-524 and RC-529.
O
NH Aoc
n -C11H23 7
+
O
OR
NH
(PhO)2PO O O RO
OTCBOC O O NH Troc
O
NH
RO
n-C11H23
n-C11H23 8
RC-524 or RC-529
Immunostimulatory activity of aminoalkyl glucosaminide 4-phosphates (AGPs) RESULTS AND DISCUSSION
Table 1. Induction of non-specific resistance to influenza infection
Chemistry RC-524 and RC-529 were prepared in a highly convergent manner from either the 2-(trimethylsilyl)ethyl (SE) or tbutyldiphenylsilyl(TBDPS) glycoside 5 beginning with the conversion of 5 to glycosyl chlorides 6 following a reaction sequence similar to that reported earlier for the preparation of 6 (Fig. 2).7 Koenigs-Knorr coupling of chlorides 6 with 2-[(R)-3-decanoyl or tetradecanoyloxytetradecanoylamino]ethanol (7) in dichloroethane at room temperature in the presence of silver triflate gave exclusively the -glycosides 8 in 75–80% yield.3 Alternatively, compound 8 could be prepared from acetonide 9 via a sequence involving 3-Oacylation and N-deprotection/acylation, followed by acetonide hydrolysis and 4- and 6-hydroxyl functionalization as described earlier for the preparation of 8.8 Reductive cleavage of the trichloroethyl-based protecting groups of 8 with Zn/AcOH, followed by N-acylation with (R)-3decanoyl or tetradecanoyloxytetradecanoic acid9 in the presence of 2-ethoxy-1-ethoxycarbonyl-1,2-dihydroquinoline and hydrogenolysis of the phenyl protecting groups with Adams’ catalyst provided RC-524 and RC-529 in 35–40% overall yield from 8 after chromatographic purification. Bligh-Dyer extraction and lyophilization from aqueous t-butanol containing triethylamine afforded RC-524 and RC-529 as their triethylammonium salts.
Immunology The induction of innate immunity by RC-524 and RC529 was examined in two distinct murine models in which aqueous formulations of the compounds were administered to mice 1–2 days prior to infectious disease challenge. The AGPs were first examined for their ability to provide immunoprophylaxis to a lethal challenge of infectious influenza virus by intranasal (i.n.) administration of both the virus and the AGP (Table 1). In this assay, BALB/c mice were pretreated with 20 mg of RC524 or RC-529 2 days prior to aerosol administration of a lethal dose of infectious influenza A/HK/68. The mice were observed for weight loss, clinical signs of illness, and mortality for 21 days following the viral challenge. The percent survival for RC-524- and RC-529-treated mice was 91% and 75%, respectively. Both groups of AGP-treated mice exhibited only mild disease with little weight loss as compared to the control groups in which all the mice died. Dose response experiments with RC524 showed significant protection ( 90% survival) at doses as low as 1 ng/mouse (lowest dose tested), but with a concomitant rise in clinical symptoms (disease index 1.5 for 1 ng dose, see Table 1). The protective NSR response was greatest ( 70% survival) when the
455
Group/treatment No treatment TEOA vehicle RC-524 RC-529
Influenza challenge Total Disease Survival weight (g) index (%) 1269 1287 2504 2342
2.6 2.8 0.8 1.3
0 0 91 75
BALB/c mice (12/group) received 20 mg of AGP in 20 ml of 0.2% aqueous triethanolamine (TEOA) by intranasal administration 2 days prior to influenza challenge. On day 0, mice were challenged with approximately 2 x LD50 infectious influenza A/HK/68 by intranasal delivery. The mice were observed for 21 days post viral challenge. Total weight for each group of 12 mice after 21 days. The disease index is a subjective score of observed clinical symptoms following influenza challenge. Mice are scored on the basis of ruffled fur, labored breathing and hunched posture. The maximum score of 3 reflects severe disease.
Table 2. Induction of non-specific resistance to Listeria monocytogenes
Treatment TEOA vehicle RC-524 RC-529
L. monocytogenes challenge Log10 PD50b protectiona (mg) 0 1.8 0.5
– 0.2 > 10
BALB/c mice (5/group) received 1 mg of AGP in 100 ml of 0.2% aqueous TEOA by i.v. administration 2 days prior to i.v. challenge with 1 x 105 L. monocytogenes. CFUs in spleens of test and control groups were determined 2 days later by serial dilutions of splenic homogenates on tryptic soy agar plates. Log10 protection was calculated by subtracting CFUs/spleen (log10 value) in test group from CFUs/spleen (log10 value) in control group. b BALB/c mice (5/group) received 0.1–10 mg of AGP in 100 ml of 0.2% aqueous TEOA by i.n. administration 1 day prior to i.n. challenge with 1 x 105 L. monocytogenes. Survival was monitored for 21 days. The 50% protective dose or PD50 was interpolated from the survival curves. a
AGP was administered between 5 days prior to challenge up to the day of challenge; AGP administration postinfluenza challenge was ineffective (data not shown). A second mouse model was used to evaluate the mediation of systemic non-specific resistance to challenge by Listeria monocytogenes, administering both bacteria and AGP either intravenously or intranasally (Table 2). In one experiment, BALB/c mice were injected i.v. with 1 mg of RC-524 or RC-529 2 days prior to L. monocytogenes
Baldridge, Cluff, Evans, Lacy, Stephens, Brookshire et al.
Table 3. IL-10 induction in mouse splenocyte cultures IL-10 induction (pg/ml) C57BL/6 C3H/HeJ splenocytes splenocytes
Treatment TEOA vehicle RC-524 RC-529 E. coli DNA
67 183 203 117
78 75 80 143
Single representative experiments. Mouse splenocytes (2 x 106 cells/well) were cultured for 24 h at 37°C in a CO2 incubator in 1 ml RPMI 1640 media containing 10% fetal bovine serum, 100 mg/ml gentamycin, 2 mM glutamine, and 20 mM HEPES with 0.2 mg of AGP originally solubilized at a concentration of 1 mg/ml in 0.2% aqueous TEOA. IL-10 was quantitated by ELISA. E. coli DNA was used as a positive control to verify cytokine production capabilities in both assays.
challenge given intravenously. As shown in Table 2, RC524 and RC-529 provided approximately 1.8 and 0.5 log10 units of protection, respectively, corresponding to a 60fold and 3-fold reduction in the average number of colony forming units (CFUs) per spleen in the AGP-treated mice as compared to vehicle-treated mice. The kinetics of protection were similar to those observed in the influenza model: minimal reduction in spleen CFUs (log10 value < 0.1) was observed when RC-524 was administered more than 5 days prior to L. monocytogenes challenge or when administered post-challenge (data not shown). The ability of RC-524 to protect mice against a lethal i.n. challenge of L. monocytogenes was also evaluated in BALB/c mice (Table 2). In this experiment, various doses
RC-524
1500 1000 500
RC-529
15000
10000
5000
0
0 10
.0 0
00
0 2.
00
00
0 0.
40
00
0 0.
08
00
0 0.
01
60
4 0.
00
32
3
06
01 00
00 0.
AGP (mg/ml)
01 3 0. 00 06 4 0. 00 32 0 0. 01 60 0 0. 08 00 0 0. 40 00 0 2. 00 00 10 0 .0 00 00
0
0 0.
RC-524
RC-529
IL-6 (pg/ml)
IL-1 (pg/ml)
2000
of RC-524 were administered intranasally 1 day prior to i.n. L. monocytogenes challenge. The PD50 for RC-524 was determined to be 0.2 mg, but doses as low as 0.5 mg/mouse led to 100% survival in RC-524-treated groups. In contrast, only 20% of the mice survived after pretreatment with 10 mg of RC-529, indicating that RC524 is at least 50 times more potent than RC-529 in this model. To determine whether the NSR activity of RC-524 is due to its ability to activate Toll-like receptor 4 (TLR4), doseresponse studies were conducted with RC-524 in LPShyporesponsive C3H/HeJ mice possessing a defect in TLR4,10 the putative receptor for LPS in mice. Virtually no reduction in CFUs in the spleen (log10 value < 0.1) was observed in C3H/HeJ mice injected i.v. with RC-524 up to a dose of 10 mg (highest dose tested) 1 day prior to i.v. L. monocytogenes challenge (data not shown), indicating that activation of the innate immune response by RC-524 likely involves TLR4 signaling. RC-524 and RC-529 were also evaluated for their ability to stimulate IL-10 in mouse splenocytes (Table 3). The level of IL-10 production was similar for RC-524 and RC529 in C57BL/6 splenocytes. No enhancement of IL-10 levels (or IFN- and IL-6, data not shown) was produced by either AGP in the LPS hyporesponsive strain, providing additional support for the involvement of the TLR4 signaling pathway in AGP-induced cell activation. Earlier, we observed a strong correlation between AGP acyl chain length and the induction of TNF- and IL-1 in human whole blood cultures.3 However, whole blood contains many cell types, including both polymorphonuclear and mononuclear white blood cells. To better characterize the activation of mononuclear cells present in human whole blood by the AGPs, human PBMCs
0. 00
456
AGP (mg/ml)
Fig. 3. Cytokine induction in human PBMCs. PBMCs were isolated from normal healthy donors by density-gradient sedimentation over Ficoll-Hypaque. PBMCs (5 x 10 5 cells/well) were cultured in a 96-well plate for 10 h at 37°C in a CO 2 incubator in 0.2 ml of RPMI 1640 medium containing 10% fetal bovine serum, 50 mg/ml streptomycin, 4 mM glutamine, 50 U/ml penicillin, and 0.05 mM 2-mercaptoethanol with the appropriate concentration of AGP in 0.2% aqueous TEOA. Supernatants were harvested and stored at –80°C. Cytokine levels were determined using the R&D Fluorokine MultiAnalyte Profiling (MAP) system and a Luminex-100 cytometer.
Immunostimulatory activity of aminoalkyl glucosaminide 4-phosphates (AGPs)
457
1600
250
IL-6 (pg/ml)
IL-1 (pg/ml)
1400 200
RC-524 150
RC-529
100
1200
RC-524
1000
RC-529
800 600 400
50
200
100,000
AGP (ng/ml)
10,000
1000
100
10
1
0.1
0.01
10,000
100,000
AGP (ng/ml)
1000
100
10
1
0.1
0
0.01
0
Fig. 4. Cytokine induction in human U937 histiocytic lymphoma cells. U937 cells (3 x 105 cells/well) were incubated with PMA (10–8 M) for 24 h. Following a 20 h incubation with the appropriate concentration of AGP in 0.2% aqueous TEOA, media were harvested and cytokine levels were determined with R&D Fluorokine MAP assay and a Luminex-100 cytometer.
were isolated by density-gradient sedimentation and stimulated with various concentrations of RC-524 and RC-529. After a 10-h incubation period, pro-inflammatory cytokine levels (IL-1 , TNF- , G-CSF, IL-6, IL-8, and IL-10) were determined. At lower concentrations (0.13 ng/ml to 1 mg/ml), RC-524 invariably produced higher levels of cytokines than RC-529, whereas at higher concentrations (> 1 mg/ml) RC-529 was more effective. Two representative dose-response curves (IL1 and IL-6 production) are shown in Figure 3. The lower potency of RC-524 as a mediator inducer at concentrations greater than 1 mg/ml may reflect a toxicity of RC-524 on mononuclear cells not evident in whole blood assays. The ability of RC-524 and RC-529 to induce the production of cytokines (TNF- , IL-1 , IL-6 and IL-10) in phorbol-12-myristate-13-acetate (PMA)-stimulated U937 cells was also examined. Results mirrored those obtained with PBMCs, except that RC-524 was more active than RC-529 throughout an 8-log concentration range. Nevertheless, as in the PBMC assay, RC-524 led to diminished cytokine levels at higher concentration, again suggestive of a dosedependent toxicity of RC-524 on monocytic cells. Representative dose-response curves (IL-1 and IL-6) are shown in Figure 4. Studies comparing the relative endotoxicities of aqueous formulations of RC-524 and RC-529, such as lethal toxicity in mice and pyrogenicity in rabbits, are in progress. CONCLUSIONS It has been demonstrated that members of the AGP class of synthetic immunostimulants are potent inducers of nonspecific resistance against bacterial and viral infections. A
comparison of the NSR and immunostimulant activity of RC-524 and RC-529 shows that AGP acyl chain length significantly affects stimulatory capacity. The results indicate that AGP activation of the innate immune response is mediated by TLR4, a transmembrane protein involved in LPS signaling. Studies are currently underway to characterize further the mechanism of AGPmediated cell activation using HeLa cells transfected with TLR4 and other LPS-binding molecules as well as to investigate further the relationship between AGP structure and the induction of innate immunity. A CKNOWLEDGEMENTS This research was funded in part by a Small Business Innovation Research (SBIR) Phase II grant (2R44 AI41811-02) from the National Institute of Allergy and Infectious Diseases and by the Defense Advanced Research Projects Agency (DARPA) under Contract No. N66001-01-C-8007. Any opinions, findings and conclusions or recommendations expressed herein are those of the authors and do not necessarily reflect the views of DARPA.
REFERENCES 1. Johnson AG. Molecular adjuvants and immunomodulators: new approaches to immunization. Clin Microbiol Rev 1994; 7: 277–289. 2. Alexander C, Rietschel ET. Bacterial lipopolysaccharides and innate immunity. J Endotoxin Res 2001; 7: 167–202. 3. Johnson DA, Sowell CG, Johnson CL et al. Synthesis and biological evaluation of a new class of vaccine adjuvants: Aminoalkyl glucosaminide 4-phosphates (AGPs). Bioorg Med Chem Lett 1999; 9: 2273–2278.
458
Baldridge, Cluff, Evans, Lacy, Stephens, Brookshire et al.
4. Baldridge JR. A synthetic adjuvant, RC-529, moves into the clinic. 3rd Meeting on Novel Adjuvants Currently in/close to Human Clinical Testing WHO/TDR – Fondation Merieux, 2002. 5. Rietschel ET, Kirikae T, Schade FU et al. The Chemical Structure of Bacterial Endotoxin in Relation to Bioactivity. Immunobiology 1993; 187: 169–190. 6. DeForge LE, Kenney JS, Jones ML et al. Biphasic production of IL-8 in lipopolysaccharide (LPS)-stimulated human whole blood. Separation of LPS- and cytokine-stimulated components using anti-tumor necrosis factor and anti-IL-1 antibodies. J Immunol 1992; 148: 2133–2141. 7. Johnson DA, Keegan DS, Sowell CG et al. 3-O-desacyl
monophosphoryl lipid A derivatives: synthesis and immunostimulant activities. J Med Chem 1999; 42: 4640–4649. 8. Johnson DA, Sowell CG. Aminoalkyl glucosamine phosphate compounds and their use as adjuvants and immunoeffectors. 2002; 6,355,257 B1, US. 9. Keegan DS, Hagen SR, Johnson DA. Efficient asymmetric synthesis of (R)-3-hydroxy- and alkanoyloxytetradecanoic acids and method for the determination of enantiomeric purity. Tetrahedron Asymmetry 1996; 7: 3559–3564. 10. Poltorak A, He X, Smirnova I et al. Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: mutations in Tlr4 gene. Science 1998; 282: 2085–2088.
