6-(4-Amino-2-butyl-imidazoquinolyl)-norleucine: Toll-like receptor 7 and 8 agonist amino acid for self-adjuvanting peptide vaccine

Abstract

Generally, small peptides by themselves are weak to induce antibody responses. Toll-like receptor (TLR) ligands are attractive candidates of vaccine adjuvants to improve their antigenicity. The covalent conjugation of TLR ligands with antigens to produce self-adjuvanting peptide vaccine is a promising approach. Based on the structure of TLR7/8 ligands, a series of synthetic amino acids 6-imidazoquinolyl-norleucines were synthesized, wherein an imidazoquinoline structure as the TLR7/8 agonistic pharmacophores was constructed on the ε-NH₂ group of Lys. Of them, 6-(4-amino-2-butyl-imidazoquinolyl)-norleucine showed the most potent TLR7 and TLR8 agonistic activities with EC₅₀ values of 8.55 and 106 μM, respectively. Subsequently, mice were immunized with the influenza A virus M2e antigen mixed with or covalently conjugated to the TLR7/8 agonist amino acid, which led to induction of M2e specific antibody productions in the absence of other adjuvant. We successfully developed a novel efficient tool for self-adjuvanting peptide vaccines targeting TLR7/8.

Introduction

Vaccine is one of the most successful and cost-effective medicines for preventing infectious diseases. Subunit vaccine that uses synthetic peptide or protein instead of microbe as an immunogen has several advantages (e.g. high purity of immunogens and low risk of adverse effects) and disadvantages (e.g. weak immunogenicity); therefore they require adjuvant systems for optimal performance. However, limited vaccine adjuvants are currently approved for human use.

The discovery of Toll-like receptors (TLRs) revealed that TLRs sense pathogen-associated molecular patterns such as lipopolysaccharides (TLR4 ligands), lipopeptides (TLR2 ligands), peptidoglycans (TLR2 ligands), flagellin proteins (TLR5 ligands), and nucleotides (TLR3, 7, 8 and 9 ligands). Each TLR–ligand interaction can activate the host innate immunity (Blasius and Beutler 2010; Kawai and Akira 2010). One of the promising applications of TLR ligands is the development of vaccine adjuvant (Duthie et al. 2010). Monophosphoryl lipid A (MPL™, TLR4 agonist) in AS04 (GSK Adjuvant Systems 04, combination with the aluminium salt) is currently registered for use in several countries.

For vaccination purposes, both the adjuvant and antigen must be administered at the same time and through the same delivery route (Cohn and Delamarre 2014). This led to the design of TLR ligand-conjugated peptide vaccines (self-adjuvanting peptide vaccines) (Fujita and Taguchi 2012). Currently, a variety of TLR ligands has been applied for the approach of self-adjuvanting vaccine development. Most of the TLR ligands are biopolymers; therefore, their isolation, synthesis, or chemical modification are difficult for large scale synthesis of vaccine. Meanwhile, Pam3Cys, Pam2Cys, and lipoamino acids as TLR2 agonists have an amino acid scaffold (Zaman and Toth 2013), which can easily conjugate with peptide antigens via amino or carboxyl group during solid phase peptide synthesis (SPPS). The resulting lipopeptides have shown self-adjuvanting property inducing the strong immunogenicity in the absence of an adjuvant. Several adenine-like compounds including imidazoquinoline derivatives [imiquimod and resiquimod (R848)] show TLR7 and TLR8 agonistic activities. Some of these compounds have been studied for TLR7/8 ligand-conjugated antigens in self-adjuvanting vaccine (Wu et al. 2007; Chan et al. 2009; Shukla et al. 2011; Gao et al. 2015). In any case, TLR7/8 ligands were conjugated with the purified or isolated antigens (proteins or peptides) through the ε-amino group of Lys or the terminal amino group. In such manner, if an antigen molecule has multiple potential reactive positions, the number of attached TLR7/8 ligands, positions of attachment, and the homogeneity of the immunogens must be taken into particular considerations. These issues led us to design novel amino acids that have an imidazoquinoline moiety as a TLR7/8-activating pharmacophore on the side chain (Fig. 1) and can be readily utilized as a building block of peptide antigens during SPPS. Hence we synthesized 6-(4-amino-imidazoquinolyl)-norleucines and evaluated their fundamental immunological activity as an adjuvant.

Fig. 1

TLR7/8 ligands and the designed amino acids: 6-(4-amino-imidazoquinolyl)-norleucines (1a–e)

Materials and methods

All of the solvents and reagents used were obtained commercially and used as received unless otherwise noted. Analytical TLC was performed on silica gel on Merck TLC silica gel 60 F ₂₅₄ aluminum sheet. Column chromatography was carried out using Merck silica gel 60 (70–230 mesh ASTM). Melting point was determined on a Round Science RFS-10. ¹H- and ¹³C-NMR spectra were recorded at 297 K on a Varian 400-MR or 600-MR using MeOH-d ₄, CDCl₃ or DMSO-d ₆ as solvents and TMS as an internal standard. Coupling constants are given in Hz. ESI–MS measurement was carried out on an Agilent 1100 Series LC/MSD model in a positive scan mode by direct injection of the compound solution into the MS. Optical rotations were determined with a JASCO P-2200 polarimater. Analytical and semi-preparative RP-HPLC used a Waters Delta 600.

N ^α-Boc-Lys(N ^ε-4-(2-chloro-3-nitro-quinolyl)-OH (2)

To a solution of Boc-Lys-OH (2.46 g, 10 mmol) and TEA (2.8 mL, 20 mmol) in 20 mL of MeOH/DCM (1:1) was added 2,4-dichloro-3-nitroquinoline (2.18 g, 9 mmol). The reaction mixture was stirred at 40 °C for 2 h. After removal of solvent, the residue was extracted with EtOAc. The organic phase was washed with 10 % citric acid and sat. NaCl, and dried over MgSO₄. The solvent was concentrated under vacuum, and Et₂O was added to afford the product 2 (4.14 g, 9.16 mmol, 91.6 %). R _f = 0.53 (CHCl₃/MeOH/AcOH = 20:2:1); mp 128–132 °C; $[\alpha ]_{\text{D}}^{20}$ +21.0° (c 0.1, MeOH); ESI–MS: C₂₀H₂₅ClN₄O₆ (452.15) m/z [M + H]⁺ 453.17 (calcd, 453.15), [M + Na]⁺ 475.17 (calcd, 475.15), [2M + H]⁺ 905.37 (calcd, 905.29), [2M + Na]⁺ 927.36 (calcd, 927.28); ¹H-NMR (400 MHz; CDCl₃): 1.39 (9H, s), 1.43–1.90 (6H, m), 3.30 (2H, br), 4.13 (1H, br), 5.37 (1H, br), 6.18 (1H, br), 7.45 (1H, br), 7.66 (1H, br), 7.77 (1H, br), 8.00 (1H, br); ¹³C-NMR (100 MHz; CDCl₃): 22.8, 28.3, 29.6, 31.9, 44.5, 54.4, 80.4, 119.0, 121.7, 126.6, 127.1, 129.3, 131.9, 142.2, 144.6, 145.9, 156.3, 177.8.

N ^α-Boc-Lys(N ^ε-4-(3-amino-2-chloro-quinolyl)-OH (3)

Nitro compound 2 (1.37 g, 3.0 mmol) was hydrogenated in the presence of Pt/C in MeOH for 3 h at room temperature. After removal of Pt/C by filtration and following evaporation, the resulted amine 3 was used for the next reaction immediately. R _f = 0.42 (CHCl₃/MeOH/AcOH = 20:2:1), ESI–MS: C₂₀H₂₇ClN₄O₄ (422.17) m/z [M + H]⁺ 423.17 (calcd, 443.17).

General procedure for the synthesis of 2-substituted imidazoquinoline derivatives

To a solution of 1,2-diamine component 3 (3.0 mmol) in toluene (10 mL) was added the corresponding orthoester (7.5 mmol). The reaction mixture was stirred at 80 °C for 3–18 h, and the reaction mixture was concentrated in vacuo. In the case of 4d–e syntheses, the residue was once dissolved in MeOH/THF (1:1, 2 mL) and 1 M NaOH (3.0 mmol) was added at 0 °C. The stirring was kept for 2 h at room temperature. The reaction mixture was concentrated under vacuum. The residue was diluted with CHCl₃. The organic layer was washed with 10 % citric acid and sat. NaCl, and dried over Na₂SO₄. The organic solvent was concentrated under reduced pressure to afford the crude product, followed by column chromatography over SiO₂ gel using CHCl₃/MeOH as eluent solvent systems to obtain pure product 4.

N ^α-Boc-6-(1-(4-chloro-imidazoquinolyl))Nle-OH (4a)

Yield 1.10 g, 2.55 mmol, 85.0 %; R _f = 0.49 (CHCl₃/MeOH/AcOH = 20:2:1); mp 107–109 °C; $[\alpha ]_{\text{D}}^{20}$ +69.0° (c 0.1, MeOH); ESI–MS: C₂₁H₂₅ClN₄O₄ (432.16) m/z [M + H]⁺ 433.17 (calcd, 433.16), [M + Na]⁺ 455.19 (calcd, 455.15), [M + K]⁺ 471.12 (calcd, 471.12), [2M + Na]⁺ 887.30 (calcd, 887.30); ¹H-NMR (400 MHz; CDCl₃): 1.22–2.16 (15H, m), 4.39 (1H, br), 4.55–4.70 (2H, m), 5.29 (1H, d, J = 7.2 Hz), 7.60–7.78 (2H, m), 8.08–8.19 (3H, m); ¹³C-NMR (100 MHz; CDCl₃): 21.7, 28.3, 29.0, 31.6, 47.7, 52.9, 80.2, 117.5, 120.2, 127.3, 128.3, 130.1, 133.8, 134.2, 143.6, 144.1, 155.5, 174.6.

N ^α-Boc-6-(1-(4-chloro-2-methyl-imidazoquinolyl))Nle-OH (4b)

Yield 1.20 g, 2.69 mmol, 67.2 %; R _f = 0.54 (CHCl₃/MeOH/AcOH = 20:2:1); mp 173–180 °C; $[\alpha ]_{\text{D}}^{20}$ +75.0° (c 0.1, MeOH); ESI–MS: C₂₂H₂₇ClN₄O₄ (446.17) m/z [M + H]⁺ 447.17 (calcd, 447.17), [2M + H]⁺ 893.34 (calcd, 893.33), [2M + Na]⁺ 915.33 (calcd, 915.33); ¹H-NMR (400 MHz; CDCl₃): 1.44 (9H, s), 1.50–2.10 (6H, m), 2.68 (3H, s), 4.36–4.50 (3H, m), 5.29 (1H, d, J = 7.6 Hz), 7.60–7.67 (2H, m), 8.01–8.15 (2H, m); ¹³C-NMR (100 MHz; CDCl₃): 14.7, 23.3, 29.2, 30.2, 33.3, 47.2, 53.8, 81.1, 118.1, 120.6, 128.1, 128.8, 130.8, 133.6, 135.5, 143.7, 144.7, 153.4, 156.6, 175.7.

N ^α-Boc-6-(1-(4-chloro-2-ethyl-imidazoquinolyl))Nle-OH (4c)

Yield 962 mg, 2.09 mmol, 69.5 %; R _f = 0.69 (CHCl₃/MeOH/AcOH = 20:2:1); mp 161–163 °C; $[\alpha ]_{\text{D}}^{20}$ +30.0° (c 0.1, MeOH); ESI–MS: C₂₃H₂₉ClN₄O₄ (460.19) m/z [M + H]⁺ 461.20 (calcd, 461.20), [M + Na]⁺ 483.18 (calcd, 483.18), [2M + H]⁺ 921.38 (calcd, 921.38); ¹H-NMR (400 MHz; CDCl₃): 1.43 (9H, s), 1.49 (3H, t, J = 7.6 Hz), 1.62–2.10 (6H, m), 3.03 (2H, q, J = 7.6 Hz), 4.37 (1H, q, J = 6.4 Hz), 4.53 (2H, t, J = 7.2 Hz), 5.37 (1H, d, J = 7.6 Hz), 7.72 (2H, dd, J = 7.4, 4.4 Hz), 8.13 (1H, d, J = 7.2, 4.4 Hz), 8.63 (1H, br); ¹³C-NMR (100 MHz; CDCl₃): 11.8, 20.9, 22.3, 28.3, 29.4, 32.3, 45.9, 52.9, 80.0, 115.0, 118.3, 121.3, 124.6, 128.1, 128.3, 128.4, 129.0, 142.9, 155.5, 157.0, 174.4.

N ^α-Boc-6-(1-(4-chloro-2-propyl-imidazoquinolyl))Nle-OH (4d)

Yield 832 mg, 1.76 mmol, 58.5 %; R _f = 0.71 (CHCl₃/MeOH/AcOH = 20:2:1); mp 164–167 °C; $[\alpha ]_{\text{D}}^{20}$ +73.0° (c 0.1, MeOH); ESI–MS: C₂₄H₃₁ClN₄O₄ (474.20) m/z [M + H]⁺ 475.22 (calcd, 475.22), [2M + H]⁺ 949.42 (calcd, 949.41); ¹H-NMR (400 MHz; CDCl₃): 1.07 (3H, t, J = 7.4 Hz), 1.45 (9H, s), 1.50–2.09 (8H, m,), 2.97 (2H, dt, J = 7.5, 1.8 Hz), 4.36 (1H, q, J = 6.6 Hz), 4.50 (2H, dt, J = 7.5, 3.0 Hz), 5.19 (1H, d, J = 8.0 Hz), 7.62–7.67 (2H, m), 8.07–8.17 (2H, m); ¹³C-NMR (100 MHz; CDCl₃): 14.0, 21.9, 22.5, 28.3, 29.3, 29.6, 32.4, 45.9, 52.8, 80.3, 117.4, 119.7, 127.0, 128.0, 130.0, 133.3, 134.6, 144.0, 143.8, 155.6, 155.8, 174.7.

N ^α-Boc-6-(1-(2-butyl-4-chloro-imidazoquinolyl))Nle-OH (4e)

Yield 876 mg, 1.79 mmol, 59.5 %; R _f = 0.71 (CHCl₃/MeOH/AcOH = 20:2:1); mp 138–143 °C; $[\alpha ]_{\text{D}}^{20}$ +68.0° (c 0.1, MeOH); ESI–MS: C₂₅H₃₃ClN₄O₄ (488.22) m/z [M + H]⁺ 489.23 (calcd, 489.23), [M + Na]⁺ 511.21 (calcd, 511.21), [2M + H]⁺ 977.45 (calcd, 977.45), [2M + Na]⁺ 999.44 (calcd, 999.43); ¹H-NMR (400 MHz; CDCl₃): 0.96 (3H, t, J = 7.2 Hz), 1.50–2.09 (19H, m), 2.98 (2H, dd, J = 7.6, 2.2 Hz), 4.35 (1H, q, J = 6.5 Hz), 4.50 (2H, m), 5.18 (1H, d, J = 7.6 Hz), 7.62–7.66 (2H, m), 8.07–8.18 (2H, m); ¹³C-NMR (100 MHz; CDCl₃): 13.8, 22.5, 22.6, 27.2, 28.3, 29.6, 30.5, 32.4, 45.9, 52.8, 80.3, 117.3, 119.7, 127.0, 127.7, 130.0, 133.2, 134.6, 143.3, 143.8, 155.6, 156.0, 174.6.

General procedure for the 4-azidation

NaN₃ (130 mg, 3.0 mmol) was added to a solution of 4-chloro compounds 4 (1.5 mmol) in DMF (1 mL). The reaction mixture was stirred at 50 °C overnight, and then diluted with EtOAc. The organic layer was washed with 10 % citric acid and sat. NaCl, dried over Na₂SO₄, and concentrated under reduced pressure to afford the crude product, followed by column chromatography over SiO₂ gel using CHCl₃/MeOH (50:1) as eluent solvent system to obtain pure product 5.

N ^α-Boc-6-(1-(4-azido-imidazoquinolyl))Nle-OH (5a)

Yield 571 mg, 1.30 mmol, 87.0 %; R _f = 0.29 (CHCl₃/MeOH/AcOH = 20:2:1); mp 114–121 °C; $[\alpha ]_{\text{D}}^{20}$ +50.0° (c 0.1, MeOH); ESI–MS: C₂₁H₂₅N₇O₄ (439.20) m/z [M + H]⁺ 440.20 (calcd, 440.20), [M + Na]⁺ 462.18 (calcd, 462.19), [2M + H]⁺ 879.40 (calcd, 879.40), [2M + Na]⁺ 901.38 (calcd, 901.38); ¹H-NMR (400 MHz; MeOH-d ₄): 1.32–2.10 (15H, m), 4.07 (1H, q, J = 5.6 Hz), 4.75 (2H, t, J = 7.0 Hz), 7.80–7.90 (2H, m), 8.31–8.40 (2H, m), 8.70 (1H, d, J = 6.4 Hz); ¹³C-NMR (100 MHz; MeOH-d ₄): 23.8, 28.7, 30.5, 32.3, 44.0, 54.5, 80.5, 116.6, 119.1, 123.8, 127.5, 129.7, 130.2, 145.0, 146.5, 158.1, 176.0.

N ^α-Boc-6-(1-(4-azido-2-methyl-imidazoquinolyl))Nle-OH (5b)

Yield 530 mg, 1.17 mmol, 90.1 %; R _f = 0.33 (CHCl₃/MeOH/AcOH = 20:2:1); mp 126–130 °C; $[\alpha ]_{\text{D}}^{20}$ +29.0° (c 0.1, MeOH); ESI–MS: C₂₂H₂₇N₇O₄ (453.21) m/z [M + H]⁺ 454.22 (calcd, 454.22), [M + Na]⁺ 476.21 (calcd, 476.20), [2M + Na]⁺ 929.43 (calcd, 929.41); ¹H-NMR (400 MHz; DMSO-d ₆): 1.34 (9H, s), 1.45–1.95 (6H, m), 2.70 (3H, s), 3.87 (1H, m), 4.63 (2H, t, J = 7.4 Hz), 7.03 (1H, d, J = 8.0 Hz), 7.74–8.77 (4H, m); ¹³C-NMR (100 MHz; DMSO-d ₆): 13.7, 22.5, 28.0, 28.8, 30.3, 45.4, 53.0, 77.9, 114.8, 117.6, 122.0, 124.6, 127.5, 128.3, 128.3, 129.0, 143.2, 152.7, 155.4, 173.9.

N ^α-Boc-6-(1-(4-azido-2-ethyl-imidazoquinolyl))Nle-OH (5c)

Yield 638 mg, 1.36 mmol, 91.0 %; R _f = 0.43 (CHCl₃/MeOH/AcOH = 20:2:1); mp 145–149 °C; $[\alpha ]_{\text{D}}^{20}$ +24.0° (c 0.1, MeOH); ESI–MS: C₂₃H₂₉N₇O₄ (467.23) m/z [M + H]⁺ 468.24 (calcd, 468.24), [M + Na]⁺ 490.22 (calcd, 490.22), [2M + H]⁺ 935.46 (calcd, 935.46), [2M + Na]⁺ 957.45 (calcd, 957.45); ¹H-NMR (400 MHz; CDCl₃): 1.41 (9H, s), 1.47 (3H, t, J = 7.2 Hz), 1.70–2.05 (6H, m), 3.01 (2H, q, J = 7.5 Hz), 4.36 (1H, q, J = 6.4 Hz), 4.49 (2H, t, J = 7.0 Hz), 5.46 (1H, d, J = 6.8 Hz), 7.62–7.70 (2H, m), 8.06 (1H, d, J = 8.0 Hz), 8.53 (1H, d, J = 7.2 Hz); ¹³C-NMR (100 MHz; CDCl₃): 11.7, 20.8, 22.3, 28.3, 29.4, 32.4, 45.9, 53.0, 79.9, 114.8, 118.1, 121.3, 124.3, 128.0, 128.3, 128.9, 142.8, 155.6, 156.9, 174.7.

N ^α-Boc-6-(1-(4-azido-2-propyl-imidazoquinolyl))Nle-OH (5d)

Yield 680 mg, 1.41 mmol, 94.0 %; R _f = 0.50 (CHCl₃/MeOH/AcOH = 20:2:1); mp 168–172 °C; $[\alpha ]_{\text{D}}^{20}$ +23.0° (c 0.1, MeOH); ESI–MS: C₂₄H₃₁N₇O₄ (481.24) m/z [M + H]⁺ 482.24 (calcd, 482.25), [M + Na]⁺ 504.22 (calcd, 504.23), [2M + H]⁺ 963.48 (calcd, 963.50), [2M + Na]⁺ 985.46 (calcd, 985.48); ¹H-NMR (400 MHz; CDCl₃): 1.08 (3H, t, J = 7.2 Hz), 1.43 (9H, s), 1.65–2.16 (8H, m), 2.97 (2H, t, J = 7.6 Hz), 4.38 (1H, q, J = 6.4 Hz), 4.55 (2H, t, J = 7.0 Hz), 5.36 (1H, d, J = 7.2 Hz), 7.72 (2H, m), 8.15 (1H, d, J = 8.0 Hz), 8.65 (1H, d, J = 7.2 Hz); ¹³C-NMR (100 MHz; CDCl₃): 14.0, 21.1, 22.3, 28.3, 29.2, 29.5, 32.4, 46.0, 52.8, 80.0, 115.0, 118.3, 121.3, 124.6, 128.2, 128.3, 128.4, 128.9, 142.9, 155.5, 156.0, 174.4.

N ^α-Boc-6-(1-(4-azido-2-butyl-imidazoquinolyl))Nle-OH (5e)

Yield 690 mg, 1.39 mmol, 92.9 %; R _f = 0.58 (CHCl₃/MeOH/AcOH = 20:2:1); mp 136–144 °C; $[\alpha ]_{\text{D}}^{20}$ +23.0° (c 0.1, MeOH); ESI–MS: C₂₅H₃₃N₇O₄ (495.26) m/z [M + H]⁺ 496.26 (calcd, 496.27), [M + Na]⁺ 518.25 (calcd, 518.25), [2M + H]⁺ 991.53 (calcd, 991.52), [2M + Na]⁺ 1013.51 (calcd, 1013.50); ¹H-NMR (400 MHz; CDCl₃): 0.97 (3H, t, J = 7.2 Hz), 1.40–2.10 (19H, m), 2.99 (2H, t, J = 9.6 Hz), 4.36 (1H, q, J = 6.8 Hz), 4.55 (2H, t, J = 6.8 Hz), 5.35, (1H, d, J = 7.6 Hz), 7.74 (2H, dd, J = 6.4, 3.6 Hz), 8.16 (1H, dd, J = 6.4, 3.6 Hz), 8.67 (1H, br); ¹³C-NMR (100 MHz; CDCl₃): 14.8, 23.3, 23.5, 28.0, 29.3, 30.4, 30.7, 33.4, 46.9, 53.8, 81.0, 116.0, 119.4, 122.2, 125.7, 129.2, 129.2, 129.3, 129.9, 144.0, 156.5, 157.1, 175.4.

General procedure for azide reduction using Zn/NH₄Cl

Zinc powder (975 mg, 15 mmol, 30 eq.) was added to a solution of azide 5 (1.0 mmol) in 4 mL of dioxane/sat. NH₄Cl (1:1). The resultant mixture was stirred at 70 °C for 48 h. Upon completion of the reaction (monitored by TLC) and following hot filtration, the filtrate was concentrated in vacuo and H₂O (20 mL) was added to the residue. The precipitate was isolated by hot filtration, washed with cold MeOH and dried under vacuum.

N ^α-Boc-6-(1-(4-amino-imidazoquinolyl))Nle-OH (6a)

Yield 385 mg, 930 μmol, 93.0 %; R _f = 0.51 (CHCl₃/MeOH/AcOH = 8:3:1), 0.21 (CHCl₃/MeOH/AcOH = 20:2:1); mp 109–116 °C; $[\alpha ]_{\text{D}}^{20}$ +32.0° (c 0.1, MeOH); ESI–MS: C₂₁H₂₇N₅O₄ (413.21) m/z [M + H]⁺ 414.21 (calcd, 414.21); ¹H-NMR (400 MHz; DMSO-d ₆): 1.23–1.90 (15H, m), 3.84 (1H, br), 4.56 (2H, br), 6.67 (2H, br), 7.27 (1H, t, J = 7.2 Hz), 7.44 (1H, t, J = 7.2 Hz), 7.62 (1H, d, J = 8.0 Hz), 8.01 (1H, d, J = 8.0 Hz), 8.19 (1H, s); ¹³C-NMR (100 MHz; DMSO-d ₆): 23.8, 29.5, 30.8, 32.9, 47.7, 55.3, 78.8, 116.1, 121.7, 122.5, 127.3, 127.9, 129.3, 132.8, 144.0, 153.3.

N ^α-Boc-6-(1-(4-amino-2-methyl-imidazoquinolyl))Nle-OH (6b)

Yield 393 mg, 921 μmol, 92.1 %; R _f = 0.53 (CHCl₃/MeOH/AcOH = 8:3:1), 0.26 (CHCl₃/MeOH/AcOH = 20:2:1); mp 150–154 °C; $[\alpha ]_{\text{D}}^{20}$ +40.0° (c 0.1, MeOH); ESI–MS: C₂₂H₂₉N₅O₄ (427.22) m/z [M + H]⁺ 428.23 (calcd, 428.23); ¹H-NMR (400 MHz; DMSO-d ₆): 1.33–1.86 (15H, m), 2.64 (3H, s), 3.85 (1H, dt, J = 8.8, 4.4 Hz), 4.53 (2H, t, J = 7.6 Hz), 7.05 (1H, d, J = 8.0 Hz), 7.93 (1H, t, J = 8.0 Hz), 7.93 (1H, t, J = 7.2 Hz), 7.93 (1H, d, J = 7.2 Hz), 8.17 (1H, d, J = 8.0 Hz), 9.01 (2H, br); ¹³C-NMR (100 MHz; DMSO-d ₆): 13.6, 22.6, 28.2, 28.8, 30.4, 45.5, 53.1, 78.0, 112.5, 118.5, 121.4, 124.3, 125.0, 129.4, 133.9, 134.7, 148.9, 153.2, 162.3, 174.1.

N ^α-Boc-6-(1-(4-amino-2-ethyl-imidazoquinolyl))Nle-OH (6c)

Yield 398 mg, 902 μmol, 90.2 %; R _f = 0.56 (CHCl₃/MeOH/AcOH = 8:3:1), 0.31 (CHCl₃/MeOH/AcOH = 20:2:1); mp 159–162 °C; $[\alpha ]_{\text{D}}^{20}$ +42.0° (c 0.1, MeOH); ESI–MS: C₂₃H₃₁N₅O₄ (441.24) m/z [M + H]⁺ 442.25 (calcd, 442.25); ¹H-NMR (400 MHz; DMSO-d ₆):1.25–1.85 (18H, m), 2.91 (2H, q, J = 7.2 Hz), 3.85 (1H, dt, J = 8.8, 4.0 Hz), 4.44 (2H, t, J = 6.8 Hz), 6.63 (2H, br), 7.25 (1H, t, J = 7.2 Hz), 7.40 (1H, t, J = 7.6 Hz), 7.59 (1H, d, J = 8.4 Hz), 7.98 (1H, d, J = 8.0 Hz); ¹³C-NMR (100 MHz; DMSO-d ₆): 12.5, 20.4, 23.2, 28.7, 29.8, 32.39, 45.2, 54.1, 78.2, 115.1, 120.5, 121.9, 126.1, 126.6, 126.9, 132.9, 144.4, 152.0, 154.5, 155.9, 174.8.

N ^α-Boc-6-(1-(4-amino-2-propyl-imidazoquinolyl))Nle-OH (6d)

Yield 400 mg, 879 μmol, 87.9 %; R _f = 0.69 (CHCl₃/MeOH/AcOH = 8:3:1), 0.38 (CHCl₃/MeOH/AcOH = 20:2:1); mp 165–171 °C; $[\alpha ]_{\text{D}}^{20}$ +52.0° (c 0.1, MeOH); ESI–MS: C₂₃H₃₁N₅O₄ (455.25) m/z [M + H]⁺ 456.26 (calcd, 456.26); ¹H-NMR (400 MHz; DMSO-d ₆): 1.00 (3H, t, J = 7.4 Hz), 1.35 (9H, s), 1.41–1.85 (8H, m), 2.87 (2H, d, J = 7.4 Hz), 3.84 (1H, dt, J = 8.4, 4.8 Hz), 4.46 (2H, t, J = 7.6 Hz), 6.65 (2H, br), 6.99 (1H, br), 7.25 (1H, t, J = 7.2 Hz), 7.40 (1H, t, J = 6.8 Hz), 7.59 (1H, d, J = 8.4 Hz), 7.98 (1H, d, J = 8.0 Hz); ¹³C-NMR (100 MHz; DMSO-d ₆): 14.3, 21.4, 23.2, 28.7, 28.8, 29.9, 31.0, 45.2, 53.9, 78.3, 115.1, 120.5, 121.9, 126.1, 126.7, 126.9, 132.8, 144.4, 152.0, 153.4, 156.0, 174.8.

N ^α-Boc-6-(1-(4-amino-2-butyl-imidazoquinolyl))Nle-OH (6e)

Yield 411 mg, 876 μmol, 87.6 %; R _f = 0.69 (CHCl₃/MeOH/AcOH = 8:3:1), 0.45 (CHCl₃/MeOH/AcOH = 20:2:1); mp 144–147 °C; $[\alpha ]_{\text{D}}^{20}$ +52.0° (c 0.1, MeOH); ESI–MS: C₂₃H₃₁N₅O₄ (469.27) m/z [M + H]⁺ 470.27 (calcd, 470.27), [M + Na]⁺ 492.26 (calcd, 492.26); ¹H-NMR (400 MHz; DMSO-d ₆): 0.95 (3H, t, J = 7.2 Hz), 1.30–1.85 (19H, m), 2.90 (2H, t, J = 7.6 Hz), 3.86 (1H, dt, J = 8.8, 4.8 Hz), 4.47 (2H, t, J = 7.6 Hz), 6.65 (2H, br), 7.00 (1H, br), 7.27 (1H, t, J = 7.6 Hz), 7.42 (1H, t, J = 7.6 Hz), 7.61 (1H, d, J = 8.4 Hz), 8.00 (1H, d, J = 8.0 Hz); ¹³C-NMR (100 MHz; DMSO-d ₆): 14.3, 22.4, 23.2, 26.6, 28.6, 29.9, 30.1, 31.0, 45.2, 53.8, 78.3, 115.1, 120.4, 121.8, 126.2, 126.7, 126.8, 132.8, 144.4, 152.0, 153.5, 155.9, 174.8.

General procedure for the Boc-deprotection

TFA (38 μL, 500 μmol, 10 eq.) was added to 50 μmol of N ^α-Boc-protected amino acid derivatives 6 at 0 °C. The stirring was kept for 2 h at room temperature. After addition of cold Et₂O, the resulted precipitate was filtrated and dried under vacuum. Purity was measured by analytical RP-HPLC on cosmosil^® PBr (4.6 × 250 nm). The solvents for HPLC were as follows: A, 0.05 % TFA in water; B, 0.05 % TFA in acetonitrile. The column was eluted at a flow rate of 1 mL/min with a linear gradient of 90 % A to 90 % A for 5 min and 90 % A to 55 % A in 35 min; the retention time (t _R) is reported in minutes.

2TFA·H-6-(1-(4-amino-imidazoquinolyl))Nle-OH (1a)

Yield 24 mg, 44.3 μmol, 88.7 %; RP-HPLC: t _R = 22.4 min; $[\alpha ]_{\text{D}}^{20}$ +55.0° (c 0.1, H₂O); ESI–MS: C₁₆H₁₉N₅O₂ (313.1539) m/z [M + H]⁺ 314.1613 (calcd, 314.1617); ¹H-NMR (600 MHz; DMSO-d ₆): 1.40–1.95 (6H, m), 3.70 (1H, br), 4.56 (2H, t, J = 7.0 Hz), 7.50 (1H, t, J = 7.2 Hz), 7.64 (1H, t, J = 7.2 Hz), 7.73 (1H, d, J = 8.0 Hz), 8.02 (2H, br), 8.40 (1H, d, J = 8.0 Hz), 8.42 (1H, s), 8.77 (2H, br); ¹³C-NMR (150 MHz; DMSO-d ₆): 26.6, 34.2, 35.1, 51.9, 57.6, 118.2, 121.4, 123.3, 127.0, 131.5, 134.3, 138.7, 150.2, 155.2, 163.3, 163.5.

2TFA·H-6-(1-(4-amino-2-metyl-imidazoquinolyl))Nle-OH (1b)

Yield 24 mg, 43.2 μmol, 86.4 %; RP-HPLC: t _R = 23.6 min; $[\alpha ]_{\text{D}}^{20}$ +55.0° (c 0.1, H₂O); ESI–MS: C₁₇H₂₁N₅O₂ (327.1695) m/z [M + H]⁺ 328.1774 (calcd, 328.1773); ¹H-NMR (600 MHz; DMSO-d ₆): 1.45–1.89 (6H, m), 2.72 (3H, s), 3.81 (1H, br), 4.57 (2H, t, J = 7.5 Hz), 7.56 (1H, t, J = 7.8 Hz), 7.70 (1H, t, J = 7.5 Hz), 7.81 (1H, d, J = 7.8 Hz), 8.15 (1H, br), 8.19 (1H, d, J = 7.8 Hz), 9.01 (2H, br); ¹³C-NMR (150 MHz; DMSO-d ₆): 13.5, 21.4, 28.7, 29.7, 45.3, 51.9, 112.4, 116.0, 118.0, 121.2, 124.2, 124.7, 129.1, 134.5, 148.8, 153.0, 158.1.

2TFA·H-6-(1-(4-amino-2-ethyl-imidazoquinolyl))Nle-OH (1c)

Yield 26 mg, 45.7 μmol, 91.3 %; RP-HPLC: t _R = 27.3 min; $[\alpha ]_{\text{D}}^{20}$  +65.0° (c 0.12, H₂O); ESI–MS: C₁₈H₂₃N₅O₂ (341.1852) m/z [M + H]⁺ 342.1931 (calcd, 342.1930); ¹H-NMR (600 MHz; DMSO-d ₆): 1.39 (3H, t, J = 6.0 Hz), 1.40–1.90 (6H, m), 3.02 (2H, q, J = 7.4 Hz), 3.88 (1H, br), 4.58 (2H, t, J = 7.5 Hz), 7.57 (1H, t, J = 7.8 Hz), 7.72 (1H, t, J = 8.1 Hz), 7.82 (1H, d, J = 8.4 Hz), 8.19 (1H, d, J = 8.4 Hz), 8.23 (2H, br), 9.07 (2H, br); ¹³C-NMR (150 MHz; DMSO-d ₆): 11.4, 20.6, 21.3, 28.8, 29.6, 44.9, 51.7, 112.4, 118.5, 121.4, 124.3, 129.3, 133.8, 134.7, 148.8, 157.2, 158.5, 171.1.

2TFA·H-6-(1-(4-amino-2-propyl-imidazoquinolyl))Nle-OH (1d)

Yield 25 mg, 42.8 μmol, 85.7 %; RP-HPLC: t _R = 27.5 min; $[\alpha ]_{\text{D}}^{20}$  +43.0° (c 0.094, H₂O); ESI–MS: C₁₉H₂₅N₅O₂ (355.2008) m/z [M + H]⁺ 356.2094 (calcd, 356.2086); ¹H-NMR (600 MHz; DMSO-d ₆): 1.05 (3H, t, J = 7.2 Hz), 1.45–1.90 (8H, m), 2.97 (2H, t, J = 7.2 Hz), 3.89 (1H, br), 4.58 (2H, t, J = 7.5 Hz), 7.56 (1H, t, J = 7.8 Hz), 7.73 (1H, t, J = 7.8 Hz), 7.82 (1H, d, J = 8.4 Hz), 8.19 (1H, d, J = 7.8 Hz), 8.26 (2H, br), 9.07 (2H, br); ¹³C-NMR (150 MHz; DMSO-d ₆): 13.7, 20.5, 21.3, 28.2, 28.9, 29.6, 44.9, 51.6, 112.5, 115.9, 117.9, 124.4, 133.8, 134.5, 148.8, 156.1, 158.4, 158.7, 170.9.

2TFA·H-6-(1-(4-amino-2-butyl-imidazoquinolyl))Nle-OH (1e)

Yield 28 mg, 46.9 μmol, 93.8 %; RP-HPLC: t _R = 29.6 min; $[\alpha ]_{\text{D}}^{20}$ −44.0° (c 0.1, H₂O); ESI–MS: C₂₀H₂₇N₅O₂ (369.2165) m/z [M + H]⁺ 370.2239 (calcd, 370.2243); ¹H-NMR (600 MHz; DMSO-d ₆): 0.97 (3H, t, J = 7.2 Hz), 1.41–1.88 (10H, m), 2.98 (2H, d, J = 7.8 Hz), 3.89 (1H, br), 4.59 (2H, t, J = 7.6 Hz), 7.57 (1H, t, J = 7.8 Hz), 7.72 (1H, t, J = 7.5 Hz), 7.83 (1H, d, J = 8.4 Hz), 8.10–8.26 (3H, m), 8.90 (2H, br); ¹³C-NMR (150 MHz; DMSO-d ₆): 13.7, 21.3, 22.0, 26.1, 28.9, 29.2, 29.6, 44.9, 51.6, 112.5, 118.5, 121.4, 124.3, 124.9, 129.3, 133.7, 134.5, 148.7, 156.3, 158.1, 170.9.

N ^α-Fmoc-6-(1-(4-amino-2-butyl-imidazoquinolyl))Nle-OH (7)

The amino acid 1e (2.35 mmol) and NaHCO₃ (652 mg, 7.76 mmol) were dissolved in H₂O/THF (20 mL). The solution of Fmoc-OSu (872 mg, 2.59 mmol) in THF (2 mL) was added at 0 °C. The reaction mixture was stirred for 30 min at 0 °C and then overnight at room temperature. After evaporation, the residue was dissolved in CHCl₃, and washed with 10 % citric acid, and sat. NaCl. The organic layer was dried over Na₂SO₄ and concentrated in vacuo. After addition of Et₂O, the resulted precipitate was filtrated and dried under vacuum. Yield 1.29 g, 2.18 mmol, 92.6 %; R _f = 0.65 (CHCl₃/MeOH/AcOH = 20:2:1); mp 141–145 °C; [$[\alpha ]_{\text{D}}^{20}$ −25.4° (c 0.11, MeOH); ESI–MS: C₃₅H₃₇N₅O₄ (591.28) m/z [M + H]⁺ 592.29 (calcd, 592.29), [2M + H]⁺ 1183.59 (calcd, 1183.58); ¹H-NMR (400 MHz; CDCl₃) 0.96 (3H, t), 1.40–2.15 (10H, m), 2.86 (2H, t, J = 6.8 Hz), 4.19–4.48 (6H, m), 6.19 (1H, d, J = 4.4 Hz), 7.22–7.92 (13H, m);¹³C-NMR (100 MHz; CDCl₃): 14.8, 23.2, 23.5, 28.0, 30.6, 30.7, 33.4, 47.0, 48.3, 56.2, 67.4, 113.8, 120.6, 120.9, 121.4, 125.7, 125.9, 126.1, 126.2, 127.9, 128.6, 129.9, 135.4, 135.8, 142.2, 150.0, 145.2, 150.9, 156.9, 157.1, 179.3.

Solid phase peptide synthesis

Fmoc-amino acids and resins were purchased from Watanabe Chemical Industries, Ltd, Japan. Following side protecting groups of Fmoc amino acid derivatives were used: Gln(Trt), Cys(Trt), Asp(OtBu), Asn(Trt), Glu(OtBu), Arg(Pbf), Ser(tBu), Thr(tBu), and Trp(Boc). Peptide synthesis was started with 500 mg of Fmoc-Asp(OtBu)-Wang resin (loading value: 0.20 mmol/g) on 0.1 mmol scale, or 45 mg of Rink-amide PEG MBHA resin (loading value: 0.55 mmol/g) on 0.025 mmol scale. Fmoc-deprotection was performed using 20 % piperidine in DMF (2 × 5 min and 1 × 20 min). Each amino acid coupling reaction was carried out using 4.1 equimolars of amino acid derivative, 4.0 equimolars of PyBOP, and 5.0 equimolars of DIPEA. After assemble of peptide, the resins were treated with reagent K at room temperature for 2 h. After addition of cold Et₂O, the yielded precipitate was centrifuged and re-dissolved in 50 % MeCN in H₂O. After lyophilization, 15 mg of each were purified by semipreparative RP-HPLC on cosmosil^® 5C₁₈-AR-II (20 × 250 nm) and the purity was measured on cosmosil^® 5C₁₈-AR-II (4.6 × 250 nm). The solvents for HPLC were as follows: A, 0.05 % TFA in water; B, 0.05 % TFA in acetonitrile. The column was eluted at a flow rate of 1 mL/min with a linear gradient of 80 % A to 80 % A for 5 min and 80 % A to 45 % A in 35 min; the retention time (t _R) is reported in minutes. The purified peptide was lyophilized to give amorphous powder.

M2e

RP-HPLC: t _R = 22.4 min; ESI–MS: MW 2623.16 g/mol, [M + 2H]²⁺ m/z 1312.57 (calcd, 1312.58), [M + 3H]³⁺ m/z 875.39 (calcd, 875.39).

Peptide 8

RP-HPLC: t _R = 23.9 min; ESI–MS: MW 2973.39 g/mol, [M + 2H]²⁺ m/z 1467.69 (calcd, 1487.70), [M + 3H]³⁺ m/z 992.13 (calcd, 992.13), [M + 4H]⁴⁺ m/z 745.33 (calcd, 744.34).

Peptide 9

RP-HPLC: t _R = 22.6 min; ESI–MS: MW 2973.39 g/mol, [M + 2H]²⁺ m/z 1487.69 (calcd, 1487.69), [M + 3H]³⁺ m/z 992.15 (calcd, 992.13), [M + 4H]⁴⁺ m/z 744.61 (calcd, 744.35).

Peptide 10

RP-HPLC: t _R = 25.8 min; ESI–MS: MW 3324.59 g/mol, [M + 2H]²⁺ m/z 1663.19 (calcd, 1663.29), [M + 3H]³⁺ m/z 1109.147 (calcd, 1109.197), [M + 4H]⁴⁺ m/z 832.112 (calcd, 832.148).

Circular dichroism (CD) spectroscopy

CD spectra were recorded at room temperature using a JASCO J-820 spectropolarimeter (JASCO, Tokyo, Japan). Two scans were collected for each sample over a wavelength range of 190–260 nm. Peptide solutions were prepared in 10 % 2,2,2-trifluoroethanol (TFE) in water with peptide concentration ranging from 99 to 102 μM. Band intensities are expressed as molar ellipticities, [θ]_m × 10³ (deg cm²/dmol).

TLR7 and TLR8 agonist assays

Human embryonic kidney cells (HEK293) stably transfected with either human TLR7 gene (HEK-Blue™-hTLR7 cells) or human TLR8 gene (HEK-Blue™-hTLR8 cells) and secreted embryonic alkaline phosphatase (SEAP) reporter gene were purchased from InvivoGen (San Diego, CA). HEK293 cells were cultured in DMEM supplemented with 10 % (v/v) heat-inactivated fetal bovine serum (Sigma), 2 mM l-glutamine (Nacalai), penicillin 10,000 IU/mL, streptomycin 10,000 mg/mL (Nacalai), Normocin 100 mg/mL (Nacalai), and 30 μg/mL Blasticidin (Nacalai), 100 μg/mL Zeocin (Nacalai). HEK293 cells were incubated at a density of ~40,000 cells/mL in a volume of 180 μL/well in 96-well. The amino acids (1a–e) were dissolved in endotoxin-free water (Otsuka) at a concentration of 1 mM. All of the stock solutions were stored at −30 °C until they were used. Each of the five amino acids diluted in endotoxin free water at appropriate dilutions were added to 96-well plates in triplicates in a total volume of 20 μL. SEAP was assayed spectrophotometrically using an alkaline phosphatase-specific chromogen (present in HEK-detection medium as supplied by the vendor) at 650 nm (SpectraMax^® M5, SoftMax Pro 5.2). Each experiment was repeated three times with similar results observed and the standard deviation shown above each sample represents three replicates in one experiment. Graph was created using GraphPad Prism software version 6.0.

Mice subcutaneous immunization and collection of sera

Immunization and the following collection of sera were carried out by BioGate Co., Ltd. (Japan). Female BALB/c mice (4- to 6-week-old) were used for immunization. Mice (n = 4/group) were injected subcutaneously on days 0, 7, 14, and 21 with 50 μg of immunogens in a total volume of 50 μL of sterile-filtered PBS. Blood was collected from the tail artery of each mouse prior to each injection and 7 days after the last immunization. The blood was left to clot at 37 °C for 1 h, and then centrifuged for 10 min at 3000 rpm to remove clots. Sera were then stored at −20 °C.

ELISA

M2e-specific serum antibodies were detected by ELISA. Ninety-six-well plates (Nunc Maxisorp) were coated with M2e (10 μg/mL) diluted in carbonate/bicarbonate coating buffer (0.05 mM, pH 9.6) and incubated for 1 h at 37 °C. The plates were washed with PBS 0.05 % Tween 20 (PBST) and blocked with PBST containing 5 % skim milk overnight at 4 °C. After washing of the plates, the primary mouse sera diluted in PBST and 1 % skim milk were added to the plates in triplicate at appropriate dilutions (starting at 1:330 concentration with threefold dilutions) and incubated for 1 h at 37 °C. The plates were washed and incubated with a horseradish peroxidase conjugated goat antimouse IgG Ab (Abcam ^® ab97265) for 1 h at 37 °C. The serum antibodies were detected by developing with freshly prepared O-phenylenediamine (OPD) solution (OPD in 0.05 M phosphate citrate buffer) for 15 min, reactions were stopped using 2 M H₂SO₄, and the absorbance was measured at λ = 490 nm on an ELISA plate reader (SpectraMax^® M5, SoftMax Pro 5.2). A positive response was defined with optical density more than three standard deviations greater than the mean absorbance control wells containing control mouse serum. GraphPad Prism software version 6.0 was used for creating a graph and statistical analysis. The statistical analysis of antibody responses between groups was performed using a one-way analysis ANOVA with Student’s t test.

Result and discussion

Design and synthesis

The primary attractiveness of imidazoquinoline compound as TLR7/8 ligands for vaccine adjuvant is its molecular weight. Furthermore, imiquimod has been approved for the treatment of actinic keratosis, external genital warts, and basal cell carcinoma (Schön and Schön 2008; Smits et al. 2008). According to the studies that examined the structure–activity relationship of imidazoquinolines, the amino group at the C-4 position is essential for agonist activity. The size of the alkyl group at the C-2 position influences the potency of the TLR7/8 agonist (Shukla et al. 2010; Schiaffo et al. 2014). While, several chemical modifications at the N-1 position of imidazoquinoline and N-9 position of adenine do not lose the TLR7 and/or TLR8 agonistic activities. Therefore, the self-adjuvanting vaccine studies using imidazoquinolines or adenine-like compounds, antigens were covalently linked at the N-1 or N-9 position of these heterocyclic compounds via a spacer (Wu et al. 2007; Chan et al. 2009; Shukla et al. 2011; Gao et al. 2015). The N-1 position was modified for the design of the amino acid (Fig. 1). Lys was chosen as the basic structure because that the ε-amino group can be used as the structural component in the heterocyclic compound, and the n-butyl unit of the side chain could act as a spacer in order to maintain distance between the imidazoquinoline moiety on the side chain and the peptide antigen (Scheme 1).

Scheme 1

Synthesis scheme for 6-(4-amino-imidazoquinolyl)-norleucines

To begin with, the ε-amino group of Boc-Lys-OH was conjugated with 2,4-dichloro-3-nitroquinoline to give 2. The selective nitro group reduction was performed with Pt-C in a mixed methanol-DCM solvent system for 3 h to yield 1,2-diamine 3. Imidazoquinoline was subsequently formed by the reaction between 3 and orthoesters with various alkyl chains. For longer alkyl chains (propyl and butyl) at the C-2 position, the cyclisation reaction required a longer duration, which resulted in ethyl ester formation. An additional saponification step was required to obtain 4. The chloro group at the C-4 position was converted to an azido group by treatment of NaN₃ in DMF at 50 °C. The conditions of the azido group reduction were evaluated using Pd–C, Pt-C, Zn powder, or Staudinger reaction. The completion of the reduction required high temperature and a long duration because of ring-chain tautomerism (tautomeric-tetrazole formation) (Goswami et al. 2005; Sonar et al. 2010; Abdou et al. 2012). Eventually treatment with a large excess of Zn powder in the presence of NH₄Cl yielded the desired amino acid in a N _α-Boc-protected form. Finally, the Boc group of 6 was removed by TFA to yield 6-(4-amino-imidazoquinolyl)-norleucines 1a-e. The overall yields from Boc-Lys-OH were 44.3–63.0 % (in 6 steps). Furthermore, N _α-Fmoc protected amino acid 7 was synthesized from 1e by the general method and was used for Fmoc-SPPS.

In order to evaluate the correlation between self-adjuvanting efficacy and the attachment position (N- and/or C-terminus) of the peptide epitope, the amino acid 1e was covalently attached to the peptide antigen M2e (Fig. 2). The highly conserved extracellular domain of the influenza A virus M2 protein (Fiers et al. 2009) was selected as the principle antigen component of the vaccine in this study. The M2e sequence (SLLTE VETPI RNEWG CRCND SSD) was assembled by conventional Fmoc-SPPS on Wang resin. After construction of the M2e sequence, N _α-Boc-protected amino acid 6e was coupled to the N-terminus of M2e using PyBOP and DIPEA in DMF. In the case of C-terminal modification, N _α-Fmoc-protected amino acid 7 was loaded on Rink-amide PEG MBHA resin. After the initial loading, the resin was treated with Boc₂O as capping and then the antigen sequence was assembled by conventional Fmoc-SPPS. All peptides were cleaved from the solid supports by reagent K and the crude products were purified using RP-HPLC prior to mouse immunization.

Fig. 2

Structure of self-adjuvanting M2e antigens containing 6-(4-amino-2-butyl-imidazoquinolyl)-norleucine 1e

CD spectra were measured to evaluate the influence of attachment of 1e on the secondary structure of the peptide epitope (Fig. 3). In 10 % TFE/water, the negative maxima at 200 nm were determined, which suggests that all peptides adopted a random structure and the addition of a synthetic amino acid 1e did not affect the secondary structure of antigen M2e.

Fig. 3

CD spectra of Me2 and its derivatives (8–10) in 10 % TFE/water 101 μM M2e (solid line), 102 μM peptide 8 (dash hyphen dotted line), 100 μM peptide 9 (dotted line), and 99 μM peptide 10 (continuous line)

In vitro TLR ligand evaluation

The induction of NF-κB from HEK293 stably transfected with human TLR7 or TLR8 and SEAP reporter gene (InvivoGen, San Diego, CA) was determined for an in vitro TLR ligand evaluation, as shown in Fig. 4. Regarding to evaluation of TLR7 agonistic activity (Fig. 4 left), the strong NF-κB induction was observed in the case of compounds 1d and 1e with propyl and butyl groups at the C-2 position, respectively. However, compounds 1a, 1b, and 1c with H, Me, and Et groups at the C-2 position, respectively showed weak or negligible NF-κB induction. Meanwhile, evaluation of the TLR8 agonistic activity (Fig. 4 right) showed a similar tendency; NF-κB was induced by compounds 1d and 1e with propyl and butyl groups, respectively. This observation is consistent with the reported correlation that the length of the C-2 alkyl group influences TLR7/8 agonistic activities (n-butyl group showing the highest potency) (Shukla et al. 2010; Schiaffo et al. 2014). Taken together, TLR7/8 agonistic activities were retained after. Among the five amino acids, 1e with the C-2 butyl chain exhibited the strongest TLR7 and TLR8 agonistic activities with EC₅₀ values of 8.55 and 106 μM, respectively. These values are approximately 50 times weaker than those of the positive control R848 (TLR7 EC₅₀ = 0.160 μM, TLR8 EC₅₀ = 2.22 μM).

Fig. 4

Agonistic activities of 6-(4-amino-imidazoquinolyl)-norleucines in human TLR7 (left) and TLR8 (right) reporter gene assays

Immunization and self-adjuvanting potency

The TLR7/8 agonist amino acid 1e was selected to investigate its adjuvant potency and self-adjuvanting efficacy when covalently conjugated with the N- and/or C-terminus of the peptide antigen. Mice (n = 4/group) were subcutaneously injected with 50 μg of immunogens (M2e with 1e, peptide 8, 9, or 10) in sterile-filtered PBS (total volume of 50 μL) on days 0, 7, 14, and 21. The ELlSA of the collected serum on day 28 showed significant antigen-specific antibody responses (Fig. 5). In the case of mice immunized with the mixture of M2e and 1e, two different doses of 1e were tested: one equimolar to antigen (10 μg, 16.9 nmol of 1e per mouse) and ten equimolar to antigen (100 μg, 169 nmol of 1e per mouse). The M2e-specific antibody was not detected in the sera of mice immunized with either M2e or the mixture of M2e and one equimolar of 1e. However, immunization with the mixture of the higher dose of 1e and M2e without other adjuvant induced M2e-specific antibody production. This indicates that the free amino acid 1e acts as a vaccine adjuvant. While, immunization with peptide antigens containing 1e induced measurable M2e-specific antibody responses without other adjuvant. At least one equimolar of the imidazoquinoline pharmacophore was enough to elicit self-adjuvanting activity by conjugation. It was clearly observed that the attachment position affects the antibody responses in the case of M2e antigen. Namely, the C-terminal-modified peptide 9 induced a stronger antibody response than the N-terminal-modified peptide 8. Furthermore, the number of attached amino acid 1e also influenced the antibody response. During immunization, the attachment of 1e to both of the termini (peptide 10) led to a synergistic antibody response. Consequently, peptide 10 induced the strongest antibody response compared with the mixture of M2e and 1e (10 eq.). This suggests that the possibility that a reduction in the amount of administrated adjuvant and the number of boosts could be achieved by increasing the number of 1e to be conjugated but not mixed.

Fig. 5

Serum levels of total M2e-specific IgG Ab on day 28. Symbol asterisk indicates a p < 0.05

A variety of TLR ligands have been investigated for the development of self-adjuvanting vaccines (Fujita and Taguchi 2012; Talbot et al. 2010; Khan et al. 2007; Wang et al. 2012; Abdel-Aal et al. 2010; Jackson et al. 2004; Lau et al. 2006; Zeng et al. 2015). Among them, Pam3Cys, Pam2Cys, and lipoamino acids as TLR2 agonists have an amino acid structure, which can be easily conjugated with peptide antigens during SPPS (Abdel-Aal et al. 2010; Jackson et al. 2004; Lau et al. 2006; Zeng et al. 2015). Several structure–activity relationship studies of self-adjuvanting vaccine have shown the attachment positions of the TLR ligand to antigen influence the induction of antibody responses (Abdel-Aal et al. 2010; Zeng et al. 2002, 2015). M2e is the N-terminal ectodomain of M2 protein and it has been suggested that a free N-terminus is important for its antigenicity (Fu et al. 2009). A recent report (Zeng et al. 2015) showed that the antibody response of M2e–Pam2Cys conjugates did not differ significantly between the N-terminal and C-terminal conjugations. In those constructions Pam2Cys was attached on the ε-amino group of an additional Lys residue via a Ser–Ser spacer, and this may reduce the influences on induction of antibody response. When we designed TLR7/8 agonist 1e based on the structure of imidazoquinoline and Lys, it was thought that the length of the n-butyl unit derived from Lys would be long enough to maintain distance between the imidazoquinoline pharmacophore and the antigen. As a matter of results, this study showed that the antibody responses induced by immunization of the M2e derivatives were significantly different; N-terminal modification of M2e by attachment of 1e did not improve antigenicity compared to that of C-terminal modification, even though CD spectra showed no conformational difference between each M2e and its derivatives (peptides 8–10). The addition of a further spacer may remedy its defect. As far as the TLR ligand–conjugation is concerned, controlling the position and number of TLR ligand to be conjugated and the distance between the TLR ligand and antigen are important for optimising the design of self-adjuvanting vaccine.

Conclusion

Based on the structure of imidazoquinoline as TLR7/8 ligands, we successfully developed a novel amino acid—6-(4-amino-2-butyl-imidazoquinolyl)-norleucine 1e—with TLR7 agonist activity (EC₅₀ = 8.55 μM) and moderate selectivity (TLR7/TLR8 = 11.1), and subsequently demonstrated that the attachment of 1e to either or both of the termini of the peptide antigen can be performed by SPPS in a general manner, and led to a poorly antigenic peptide with self-adjuvanting property. As such, the amino acid 1e is an applicable building block in the development of self-adjuvanting peptide vaccine targeting TLR7/8; similar to or addition to the approach that uses TLR2 agonists (Pam2/3Cys). Further, using the amino and carboxyl groups, the amino acid 1e could be explored TLR7/8 ligands.