|Home | About | Journals | Submit | Contact Us | Français|
Previously, G protein–coupled receptor (GPCR) agonists were tethered from polyamidoamine (PAMAM) dendrimers to provide high receptor affinity and selectivity. Here we prepared GPCR Ligand Dendrimer (GLiDe) conjugates from a potent adenosine receptor (AR) antagonist; such agents are of interest for treating Parkinson’s disease, asthma, and other conditions. Xanthine amine congener (XAC) was appended with an alkyne group on an extended C8 substituent for coupling by Cu(I)-catalyzed click chemistry to azide-derivatized G4 (fourth-generation) PAMAM dendrimers to form triazoles. These conjugates also contained triazole-linked PEG groups (8 or 22 moieties per 64 terminal positions) for increasing water-solubility and optionally prosthetic groups for spectroscopic characterization and affinity labeling. Human AR binding affinity increased progressively with the degree of xanthine substitution to reach Ki values in the nM range. The order of affinity of each conjugate was hA2AAR > hA3AR > hA1AR, while the corresponding monomer was ranked hA2AAR > hA1AR ≥ hA3AR. The antagonist activity of the most potent conjugate 14 (34 xanthines per dendrimer) was examined at the Gi-coupled A1AR. Conjugate 14 at 100 nM right-shifted the AR agonist concentration-response curve in a cyclic AMP functional assay in a parallel manner, but at 10 nM (lower than its Ki value) it significantly suppressed the maximal agonist effect in calcium mobilization. This is the first systematic probing of a potent AR antagonist tethered on a dendrimer and its activity as a function of variable loading.
We have used poly(amidoamine) (PAMAM) dendrimers as polymeric nanocarriers for drugs that bind to G protein–coupled receptors (GPCRs) located on the cell surface.1–3 These receptors are an important area of drug discovery; Agonists and antagonists of rhodopsin-like GPCRs constitute a large fraction (~30%) of the 324 molecular targets of FDA-approved drugs (4). Multivalent GPCR Ligand Dendrimer (GLiDe) conjugates, in which a strategically derivatized agonist or antagonist is covalently tethered from a high molecular weight PAMAM dendrimer, have displayed enhanced pharmacodynamic properties in comparison to the monomeric ligands.2,3,5 One of the objectives of this approach is to favorably alter the pharmacokinetics of the drug in vivo when bound to the nanocarrier. It is also feasible to target binding of a polymeric drug to aggregates of GPCRs, which have been shown to exist and postulated to be major participants in the varied biological effects of a given receptor.6 GPCR dimers, including heterodimers of two different receptor classes, have already been established to have an altered pharmacology compared to homomeric GPCRs.7 We have shown with molecular modeling that GLiDe conjugates of sufficient molecular size can adopt a suitable geometry to bridge multiple binding sites in such receptor aggregates.8
We initially reported the use of PAMAM dendrimers as scaffolds for the presentation of nucleosides and nucleotides that selectively activate adenosine receptors (ARs, particularly the A2A and A3 subtypes) or antagonize P2Y receptors (specifically the P2Y1 subtype) to modulate intracellular signaling,1,2,9 with both of these receptor families being GPCRs. These GLiDe conjugates are stable and biologically active without requiring the release of small molecules or cellular internalization, which are often an integral part of other polymeric drug delivery schemes.10 Recently, we showed that a potent A3AR agonist conjugate displayed cytoprotective effects in a cardiomyocyte culture heterologously expressing the A3AR.11 The present study extends the GLiDe conjugate approach for the first time from AR agonists to multivalent conjugates of AR antagonists. AR antagonists, the naturally occurring xanthines theophylline 1 and caffeine 2 being prototypical examples (Figure 1), are of therapeutic interest in the treatment of Parkinson’s disease, diabetes, asthma, cancer, and other conditions.4,12 The structure activity relationship (SAR) of xanthine derivatives as AR antagonists has been exhaustively explored.13,14 One of the earliest high affinity AR ligand probes was the 8-phenyl derivative xanthine amine congener (3, XAC), which is a moderately selective antagonist of the A1AR subtype in the rat and relatively nonselective in binding to human (h) ARs.15 XAC contains an amine-terminal chain placed at an insensitive site on the pharmacophore and was designed as a functionalized congener,16 i.e. for attachment to larger carrier moieties and reporter groups with retention of the ability to bind potently to ARs.
In this study, we have identified a means of linking XAC to water-soluble PAMAM dendrimers by Cu(I)-catalyzed click chemistry between an alkyne and an azide,17 which we found to preserve the high receptor affinity in this series. The degree of substitution by XAC and by polyethylene glycol (PEG) water-solubilizing chains was systematically varied and the effects on receptor affinity studied. This analysis suggests that each GLiDe conjugate molecule may bind to multiple AR binding sites in vicinity on the cell surface. Furthermore, the ability to introduce prosthetic groups for receptor localization or characterization on the carrier without interfering with the high affinity in receptor binding was shown.
All reactions were carried out under a nitrogen atmosphere. We purchased N-ethyl-N′-dimethylaminopropylcarbodiimide hydrochloride (EDC.HCl) and G4 PAMAM dendrimer (5 wt% solution in methanol) with an ethylenediamine core 6a from Sigma-Aldrich (St. Louis, MO). All other reagents and solvents, except those indicated, came from Sigma-Aldrich. We purchased dialysis membranes (Spectra/Pore Membrane, MWCO 3500, flat width 18 mm) from Spectrum Laboratories (Rancho Dominguez, CA). Imidazole-1-sulfonyl azide hydrochloride was prepared.18 Synthesis of XAC 3 is described elsewhere.14 Active esters IRDye800CW 23 and IRDye700DX 24 were obtained from LI-COR Biosciences (Lincoln, NE). Amicon centrifugal ultrafilters (molecular weight cutoff 3000) were used.
Proton nuclear magnetic resonance (NMR) spectra were recorded on a Bruker DRX-400 spectrometer with d6-DMSO as a solvent. The chemical shifts are reported in parts per million (ppm) relative to tetramethylsilane (δ 0.0 ppm) or in D2O relative to HOD (4.80 ppm). ESI - High Resolution Mass Spectroscopic (HRMS) measurements were performed on a proteomics optimized Q-TOF-2 (Micromass-Waters) using external calibration with polyalanine and matrix-assisted laser desorption ionization (MALDI) time-of-flight MS experiments on a Waters LCT Premier mass spectrometer at the Mass Spectrometry Facility, National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), NIH. Observed mass accuracies are those expected on the basis of known performance of the instrument as well as the trends in masses of standard compounds observed at intervals during the series of measurements. Reported masses are observed masses uncorrected for this time-dependent drift in mass accuracy. IR spectra were recorded using a PerkinElmer Spectrum One FT-IR spectrometer (applied as a DMSO solution).
G4 PAMAM dendrimer containing an ethylene diamine core was obtained from Sigma-Aldrich, and terminal amines reacted with imidazole-1-sulfonyl azide hydrochloride (3). K2CO3 (3.48 g, 30.15 mmol), CuSO4.5H2O (34.6 mg, 0.163 mmol) and imidazole-1-sulfonyl azide hydrochloride (3.0 g, 1.04 mol) were added to a solution of G4 PAMAM dendrimer 6a, (20 g, 10 w% in methanol, [NH2]=16.3 mmol) in anhydrous methanol (11 mL), and the reaction mixture was stirred at room temperature overnight. The mixture was then dialyzed in deionized water for 7 days (changing water 4 times per day). Lyophilization after dialysis provided the azide-derivatized G4 dendrimer 6b (3.2 g, 58%) as a solid. IR νmax 2110 cm−1. ESI-MS: calcd. 15,611; found 15,545.
The azide-derivatized G4 PAMAM dendrimer 6b (111 mg, 7.1 µmol) and a poly(ethylene glycol)methyl ether derivative of butynoic acid 7 (MW 2000, Aldrich Chemical Co., 114 mg, 57 µmol) were added to a mixture of DMSO (4 mL) and water (4 mL). The solution was treated with freshly prepared aqueous sodium ascorbate (1 M solution, 228 µL), followed by the addition of 7.5% aqueous cupric sulfate (380 µL, 113 µmol). The reaction mixture was stirred at rt overnight, and the product was purified by extensive dialysis in water (changing water 4 times per day for 2 days). The mixture was lyophilized to give compound 8 (153 mg, 67%) containing ~8 equivalents of PEG per dendrimer molecule as a colorless foamy solid. 1H NMR (DMSO-d6, 400 MHz) δ 8.10 (br s), 4.35 (s), 4.13 (s), 3.5 (s), 3.41–3.44 (m), 3.33 (br s), 3.24 (s), 2.67–2.87 (m), 2.26 (br s). ESI-MS: calcd. 31,545; found 32,077.
Compound 9 (74%) containing ~22 equivalents of PEG per dendrimer molecule was synthesized from dendrimer 6b following the same method as for compound 8. 1H NMR (DMSO-d6, 400 MHz) δ 8.03 (br s), 4.35 (s), 4.25 (s), 3.59 (s), 3.39–3.44 (m), 3.32 (br s), 3.24 (s), 2.87 (br s), 2.66 (br s), 2.33 (s), 2.24 (br s). ESI-MS: calcd. 59,545; found 59,855.
6-Heptynoic acid (33 µL, 0.27 mmol) and EDC (68.8 mg, 0.36 mmol) were added to a solution of compound 3 (77 mg, 0.18 mmol) in anhydrous DMF (3 mL) and stirred overnight at rt. Solvent was evaporated and the residue was purified using flash silica gel column chromatography to give compound 4 (69 mg, 72%) as a colorless solid. 1H NMR (CD3OD, 400 MHz) δ 8.05 (d, J = 8.8 Hz, 2H), 7.14 (d, J = 8.8 Hz, 2H), 4.61–4.63 (m, 2H), 4.14 (t, J = 7.2 Hz, 2H), 4.01 (t, J = 7.6 Hz, 2H), 3.41–3.65 (m, 6H), 2.19–2.13 (m, 5H), 1.86–1.70 (m, 1H), 1.69–1.64 (m, 4H), 1.52–1.46 (m, 1H), 1.15 (d, J =6.4 Hz, 1H), 1.17–0.95 (m, 6H). HRMS calculated for C28H37N6O5 (M + H) +: 537.2825; found 537.2823.
The azide-derivatized G4 PAMAM dendrimer 6b (10.5 mg, 0.17 µmol) and a XAC alkyne derivative 4 (0.54 mg, 1.02 µmol) were added to a mixture of DMSO (0.4 mL) and water (0.4 mL). The solution was treated with freshly prepared aqueous sodium ascorbate (1 M solution, 3 µL), followed by the addition of 7.5% aqueous cupric sulfate (2.3 µL, 0.7 µmol). The reaction mixture was stirred at rt overnight, and the product was purified by extensive dialysis in water (changing water 4 times per day for 2 days). The mixture was lyophilized to give compound 10 (6.9 mg, 66%) containing ~22 equivalents of PEG and ~6 xanthine equivalents per dendrimer molecule as a foamy solid. 1H NMR (DMSO-d6, 400 MHz) δ 8.15 (br s), 8.09 (d, J= 8.8 Hz), 7.71–7.97 (m), 7.16 (d, J = 9.2 Hz), 4.54 (s), 4.24–4.36 (m), 4.12 (br s), 4.04 (t, J = 7.6 Hz), 3.81 (t, J = 7.6 Hz), 3.54–3.67 (m), 3.49 (br s), 3.14–3.47 (br m), 2.63–2.71 (br m), 2.07–2.33 (m), 1.42–1.81 (br m), 0.84–0.92 (br m). ESI-MS: calcd. 63,074; found 63,069.
Compound 11 (70%) was synthesized from PEG conjugated dendrimer 9 following the same procedure as for compound 10 using 8 equivalents of the XAC alkyne derivative 4. 1H NMR (DMSO-d6, 400 MHz) δ 8.13 (br s), 8.10 (d, J= 8.8 Hz), 7.87–7.93 (m), 7.11 (d, J = 9.2 Hz), 4.55 (s), 4.32–4.36 (m), 4.12 (br s), 4.04 (t, J = 7.6 Hz), 3.85 (t, J = 7.6 Hz), 3.59–3.67 (m), 3.46 (br s), 3.14–3.42 (br m), 2.67–2.86 (br m), 2.05–2.33 (m), 1.51–1.75 (br m), 0.86–0.91 (br m). ESI-MS: calcd. 64,147; found 64,251.
Compound 12 (75%) was synthesized from PEG conjugated dendrimer 9 following the same procedure as for compound 10 using 19 equivalents of the XAC alkyne derivative 4. 1H NMR (DMSO-d6, 400 MHz) δ 8.17 (br s), 8.07 (d, J= 8.8 Hz), 7.69–7.93 (m), 7.11 (d, J = 9.2 Hz), 4.54 (s), 4.30–4.35 (m), 4.13 (br s), 4.02 (t, J = 7.6 Hz), 3.84 (t, J = 7.6 Hz), 3.58–3.68 (m), 3.51 (br s), 3.13–3.48 (br m), 2.62–2.7 (br m), 2.05–2.33 (m), 1.39–1.81 (br m), 0.86–0.92 (br m). ESI-MS: calcd. 70,050; found 70,074.
Compound 13 (78%) was synthesized from PEG conjugated dendrimer 9 following the same procedure as for compound 10 using 23 equivalents of the XAC alkyne derivative 4. 1H NMR (DMSO-d6, 400 MHz) δ 8.13 (br s), 8.09 (d, J= 8.8 Hz), 7.91–7.86 (m), 7.11 (d, J = 9.2 Hz), 4.54 (s), 4.32–4.34 (m), 4.12 (br s), 4.04 (t, J = 7.6 Hz), 3.86 (t, J = 7.6 Hz), 3.58–3.68 (m), 3.44 (br s), 3.14–3.44 (br m), 2.67–2.87 (br m), 2.03–2.33 (m), 1.49–1.75 (br m), 0.86–0.92 (br m). ESI-MS: calcd. 72,197; found 72,320.
Compound 14 (84%) was synthesized from PEG conjugated dendrimer 9 following the same method as for compound 10 using 34 equivalents of the XAC alkyne derivative 4. 1H NMR (DMSO-d6, 400 MHz) δ 8.11 (br s), 8.09 (d, J= 9.2 Hz), 7.90–7.85 (m), 7.11 (d, J = 9.2 Hz), 4.54 (s), 4.02 (t, J = 6.0 Hz), 3.87 (t, J = 7.6 Hz), 3.12–3.71 (br m), 2.67–2.74 (m), 2.30–2.38 (m), 2.10–2.14 (m), 2.05 (t, J = 7.2 Hz), 1.72–1.79 (m), 1.53–1.61 (m), 1.37–1.41 (m), 0.86–0.92 (br m). ESI-MS: calcd. 78,100; found 77,999.
Compound 15 (67%) was synthesized from PEG conjugated dendrimer 8 following the same method as for compound 10 using 4 equivalents of the XAC alkyne derivative 4. 1H NMR (DMSO-d6, 400 MHz) δ 8.14 (br s), 8.08 (d, J= 8.8 Hz), 7.72–7.90 (m), 7.10 (d, J = 8.8 Hz), 4.54 (s), 4.34 (br s), 4.13 (br s), 4.02 (t, J = 7.6 Hz), 3.86 (t, J = 7.5 Hz), 3.60–3.67 (m), 3.49 (br s), 3.08–3.44 (br m), 2.67–2.87 (br m), 2.22–2.33 (br m), 1.48–1.68 (br m), 0.86–0.92 (br m). ESI-MS: calcd. 34,223; found 34,377.
Compound 16 (72%) was synthesized from PEG conjugated dendrimer 8 following the same method as for compound 10 using 17 equivalents of the XAC alkyne derivative 4. 1H NMR (DMSO-d6, 400 MHz) δ 8.15 (br s), 8.09 (d, J= 8.8 Hz), 7.81–7.86 (m), 7.11 (d, J = 8.8 Hz), 4.54 (s), 4.34 (s), 4.19 (t, J = 7.5 Hz), 4.04 (t, J = 7.6 Hz), 3.52 (br s), 3.14–3.44 (br m), 2.62–2.74 (m), 2.03–2.33 (m), 1.39–1.75 (br m), 0.89–0.92 (br m). ESI-MS: calcd. 40,662; found 40,461.
Compound 17 (80%) was synthesized from PEG conjugated dendrimer 8 following the same method as for compound 10 using 26 equivalents of the XAC alkyne derivative 4. 1H NMR (DMSO-d6, 400 MHz) δ 8.16 (br s), 8.09 (d, J= 8.8 Hz), 7.86–7.88 (m), 7.11 (d, J = 8.8 Hz), 4.54 (s), 4.02 (t, J = 6.8 Hz), 3.89 (t, J = 6.7 Hz), 3.51 (br s), 3.14–3.32 (m), 2.10–2.14 (m), 2.04 (t, J = 7.2 Hz), 1.72–1.77 (m), 1.51–1.61 (m), 1.35–1.43 (m), 0.81–0.92 (br m). ESI-MS: calcd. 45,492; found 45,223.
Compound 18 (84%) was synthesized from PEG conjugated dendrimer 8 following the same method as for compound 10 using 37 equivalents of the XAC alkyne derivative 4. 1H NMR (DMSO-d6, 400 MHz) δ 8.18 (br s), 8.08 (d, J= 9.2 Hz), 7.86–7.88 (m), 7.11 (d, J = 8.8 Hz), 4.54 (s), 4.02 (t, J = 6.0 Hz), 3.87 (t, J = 6.8 Hz), 3.12–3.62 (br m), 2.67–2.74 (m), 2.33 (br s), 2.10–2.13 (m), 2.05 (t, J = 6.8 Hz), 1.72–1.77 (m), 1.53–1.61 (m), 1.37–1.43 (m), 0.88–0.92 (br m). ESI-MS: calcd. 51,931; found 52,035.
A solution of Alexa Fluor 488 alkyne 19 (0.5 mg, 0.64 µmol) in water (0.5 mL) and freshly prepared sodium ascorbate (1 M, 5 µL) were added to a solution of the XAC-dendrimer conjugate 18 (8.2 mg, 0.16 µmol) in DMSO (0.5 mL). Then, a 7.5% aqueous cupric sulfate (6 µL, 113 µmol) was added to the reaction mixture, which was stirred overnight at rt. The dendrimer derivative was purified by extensive dialysis in water and lyophilization of the resulting solution gave compound 19 (5.9 mg, 66%) as an orange colored solid. ESI-MS: calcd. 57,065; found 57,079.
6-Heptynoic acid (11 µL, 0.09 mmol), HATU (31 mg, 0.08 mmol) and diisopropyl ethylamine (14 µL, 0.08 mmol) were added to a solution of 4-(2-aminoethyl)benzene-1-sulfonyl fluoride (Aldrich, 15 mg, 0.06 mmol) in anhydrous DMF (1 mL). The reaction mixture was stirred overnight at rt. Solvent was evaporated and the residue was purified using flash silica gel column chromatography (hexane:ethyl acetate=2:1) to give compound 21 (15.2 mg, 78%) as a colorless solid. 1H NMR (CD3OD, 400 MHz) δ 8.00 (d, J = 8.4 Hz, 2H), 7.60 (d, J = 8.0 Hz, 2H), 3.52–3.48 (m, 2H), 2.98 (t, J = 6.8 Hz, 2H), 2.32 (t, J = 7.2 Hz, 2H), 2.23–2.20 (m, 2H), 1.75–1.44 (m, 5H). HRMS calculated for C15H19NO3SF (M + H) +: 312.1070; found 312.1073.
Benzene-sulfonyl fluoride alkyne derivative 21 (0.5 mg, 1.6 µmol) and freshly prepared sodium ascorbate (1 M, 3 µL) were added to a solution of XAC-dendrimer conjugate 20 (5.6 mg, 0.1 µmol) in DMSO (0.4 mL). Then, a 7.5% aqueous cupric sulfate (4 µL, 0.9 µmol) was added to the reaction mixture, which was stirred overnight at rt. The dendrimer derivative was purified by extensive dialysis in water and lyophilization of the resulting solution gave compound 22 (4.2 mg, 73%) as an orange colored solid. ESI-MS: calcd. 58,945; found 59,155.
A solution of IR dye 800CW NHS ester 23 (0.5 mg, 0.42 µmol) in DMF (0.3 mL) was added to a solution of compound 14 (2.66 mg, 0.05 µmol) in DMF (1 mL) followed by triethylamine (10 µL, 0.42 µmol) and stirred overnight at rt. The dendrimer derivative was purified by extensive dialysis in water, and lyophilization of the resulting solution gave compound 25 (2.13 mg, 76%) as a green colored foamy solid. ESI-MS: calcd. 81,943; found 82,258.
A solution of IR dye 700DX NHS ester 24 (0.5 mg, 0.25 µmol) in DMF (0.3 mL) was added to a solution of compound 14 (2.32 mg, 0.03 µmol) in DMF (1 mL) followed by sodium tetraborate buffer (20 µL, 0.25 µmol), the reaction mixture was stirred overnight at rt. The dendrimer derivative was purified by extensive dialysis in water, and lyophilization of the resulting solution gave compound 26 (2.19 mg, 83%) as a green colored foamy solid. ESI-MS: calcd. 88,511; found 89,063.
Penicillin-Streptomycin-Glutamine and Hygromycin B were purchased from Invitrogen (Carlsbad, CA). [3H]R-N6-(2-Phenylisopropyl)adenosine ([3H]R-PIA, 42.6 Ci/mmol) was obtained from Moravek Biochemicals (Brea, CA). [125I]4-Amino-3-iodobenzyl-5’-N-methylcarboxamidoadenosine ([125I]I-AB-MECA, 2200 Ci/mmol) and [3H]-2-[p-(2-carboxyethyl)phenylethylamino]-5’-N-ethylcarboxamidoadenosine ([3H]CGS21680, 40.5 Ci/mmol) were purchased from PerkinElmer (Waltham, MA). DMEM/F12 medium, 1 M Tris–HCl (pH 7.5) and G418 Sulfate were purchased from Mediatech, Inc. (Herndon, VA). Calcium assay kit was from Molecular Devices (Sunnyvale, CA). All other reagents were from standard sources and are of analytical grade.
We prepared test compounds as 0.1 mM stock solutions in DMSO and stored them at 4°C. Adenosine deaminase (25.3 U/mg) was purchased from Worthington Biochemical Corporation (Lakewood, NJ.)
We cultured Chinese hamster ovary (CHO) cells stably expressing the recombinant hA1AR, hA2BAR, and hA3AR and human embryonic kidney (HEK) 293 cells stably expressing the hA2AAR in DMEM and F12 (1:1) supplemented with 10% fetal bovine serum (FBS), 50 U/mL penicillin, 50 µg/mL streptomycin.19 In addition, we added 800 µg/mL Geneticin to the hA2AAR media and 500 µg/mL Hygromycin B to the hA1AR, hA2BAR, and hA3AR media. After harvesting the cells, we centrifuged them at 250 × g for 5 min at 4°C. The pellet was resuspended in 50 mM Tris-HCL buffer (pH 7.5), containing 10 mM MgCl2. The suspension was homogenized them with an electric homogenizer for 10 sec, and was then re-centrifuged at 20,000 × g for 30 min at 4°C. The resultant pellet was homogenized again and resuspended in buffer mentioned above in the presence of 3 U/mL adenosine deaminase and finally pipetted them into 1 mL vials, and then stored them at −80°C until we conducted the binding experiments. We measured the protein concentration with a BCA Protein Assay Kit from Pierce Biotechnology (Rockford, IL).20
Binding assays using standard AR ligands were performed using the general methods reported.21–23 Each tube in the binding assay contained 50 µL of increasing concentrations of the test ligand in Tris-HCl buffer (50 mM, pH 7.5) containing 10 mM MgCl2, 50 µL of the appropriate agonist radioligand, and 100 µL membrane suspension, added sequentially. For the hA1AR (20 µg protein/tube), we used the radioligand [3H]R-PIA (0.2 nM, precise final concentration is calculated for each experiment). For the hA2AAR (20 µg protein/tube), we used the radioligand [3H]CGS21680 (10 nM). For the hA3AR (20 µg protein/tube), we used the radioligand [125I]IAB-MECA (0.89 nM). We determined nonspecific binding with a final concentration of 10 µM unlabeled NECA diluted with the buffer. We incubated the mixtures at 25°C for 60 min in a shaking water bath. We terminated binding reactions by filtration through Brandel GF/B filters under a reduced pressure with an M-24 cell harvester (Brandel, Gaithersburg, MD). We washed filters three times with 3 mL 50 mM ice-cold Tris-HCl buffer (pH 7.5). We then placed filters for hA1AR and hA2AAR binding in scintillation vials containing 5 mL Hydrofluor scintillation buffer and counted with a PerkinElmer Liquid Scintillation Analyzer (Tri-Carb 2810TR). We counted filters for hA3AR binding with a Packard Cobra II γ-counter (PerkinElmer).
CHO cells expressing the A1AR or A2BAR were seeded in 24-well plates and incubated at 37 °C overnight. The following day the medium was removed and replaced with DMEM containing 50 mM HEPES, 10 µM rolipram, 3 U/mL adenosine deaminase, and increasing concentrations of agonists. For measurements at the A1AR, forskolin (10 µM) was added 30 min after the addition of agonists and incubated for another 15 min. Antagonists were added 20 min before the addition of agonists. The medium was removed, and the cells were lysed with 200 µL of 0.1 M HCl. 100 µL of the HCl solution was used in the Sigma Direct cAMP Enzyme Immunoassay following the instructions provided with the kit. The results were interpreted using a BioTek ELx808 Ultra Microplate reader (BioTek, Winooski, VT) at 405 nm.
CHO cells expressing the A1AR cells were grown overnight in 100 µL media in 96-well flat bottom plates at 37 °C at 5% CO2 or until 80–90% confluency. The calcium assay kit (Molecular Devices, Sunnyvale, CA) was used as directed without washing cells, and with probenecid added to the loading dye at a final concentration of 2.5 mM to increase dye retention. Cells were loaded with 50 µL dye with probenecid to each well and incubated for 60 min at rt. The compound plate was prepared using dilutions of various compounds in Hanks Buffer (pH 7.4). Samples were run on a FLIPRTETRA (Molecular Devices) in duplicate at rt. Antagonists were added 20 min before the addition of agonists. Cell fluorescence was measured (excitation at 485 nm; emission at 525 nm) following exposure to agonists. Increases in intracellular calcium are reported as the maximum fluorescence value after exposure minus the basal fluorescence value before exposure.
We initially measured the concentrations of the dendrimer-ligand complexes by the concentration of the dendrimer, not the attached ligand. We determined EC50 and apparent Ki values with Prism software (GraphPad, San Diego, CA); they are presented as mean ± SE. The number of xanthine moieties on a given dendrimer that are accessible to receptors is uncertain; the data are therefore shown as apparent Ki values. Later the same data were converted to concentration of the xanthine moieties for graphic presentation. We repeated all experiments at least three times.
We prepared generation 4 (G4) PAMAM dendrimer conjugates of the potent, nonselective hAR antagonist XAC 3 (Scheme 1, Table 1, 9–18) to act as multivalent ligands of the receptors. The peripheral groups of the precursor dendrimer 6a were modified to contain predominantly azido groups in place of the 64 terminal primary amines by reacting with imidazole-1-sulfonyl azide hydrochloride in methanol to give compond 6b.3,18 The same approach of preparing the PAMAM dendrimer for click chemistry by substitution of amino groups with azides was taken in the synthesis of conjugates of AR agonists and P2Y1 receptor antagonists.3,9 Mass spectral data of 6b indicated that 8 amino groups out of the 64 remained unreacted, and the presence of azido groups was confirmed by a strong IR peak at 2110 cm−1. The presence of these residual amines was later used for synthetic advantage in the coupling of additional prosthetic groups. Prior to coupling of the pharmacophoric species, the dendrimer was appended with PEG moieties by a Cu(I)-catalyzed click cyclization using an alkyne-derivatized PEG 7 (end-capped by a methyl ether and a butynoyl ester) to yield the water-soluble, and mainly uncharged PAMAM derivatives 8 and 9, having 8 or 22 PEG moieties per dendrimer, respectively. Thus, two degrees of substitution with PEG of the 64-terminal groups of the G4 dendrimer were compared.
The synthetic method for cross-linking the receptor ligand to the nanocarriers 8 and 9 by click chemistry3 required introduction of an alkyne at an insensitive site on the xanthine moiety, and for this purpose the hexynoyl amide derivative of XAC 4 was prepared. We have explored in detail the SAR at ARs at the distal region of the chain extended from the para position of the 8-phenylxanthines.15,16 This is a site for unlimited chain derivatization, and thus the alkyne-functionalized chain was used to provide click products with the azide-functionalized dendrimers.3 We systematically varied the degree of substition of the dendrimer carrier with the GPCR ligand, i.e., the nonselective AR antagonist, ranging from substitution of an average of 4 to 37 of the terminal groups with the xanthine moiety. The efficiency of incorporation of the xanthine was nearly quantitative depending on the equivalents of 4 added to the click reaction, as indicated by mass spectroscopy. A fully XAC-substituted dendrimer was not prepared due to limitation of aqueous solubility, which presented a concern in the absence of tethered PEG groups.
XAC conjugate 18, which contained a random distribution of an average of 37 xanthine moieties and 8 PEG groups, was further derivatized with functional prosthetic groups that were coupled by click chemistry (Scheme 2). A fluorescent prosthetic group for spectroscopic characterization (Alexa Fluor 488) was included in conjugates 20 and 22 (24). Alexa Fluor 488, which has emission and excitation maxima at 495 and 519 nm, respectively, has already been used as a fluorophore in small ligand probes of the A2AAR.25 Conjugate 22 also contained an aryl sulfonyl fluoride (an average of 5 moieties incorporated per PAMAM dendrimer) for affinity labeling of the receptor.26 Aryl sulfonyl fluorides readily react irreversibly with proximal nucleophilic amino acids on a receptor yet are sufficiently stable to be applied in aqueous medium. Two near-infrared (NIR) dyes27 with absorption in the range of 800 nm (25) and 700 nm (26) were included in conjugates containing an average of 34 xanthine moieties per dendrimer. The NIR dyes were amide-linked to the dendrimer 14 using the active ester precursors 23 and 24, leading to an average degree of incorporation of 4 or 6 dye moieties incorporated per PAMAM dendrimer in 25 and 26, respectively, as shown in Scheme 3.
We tested the xanthine-dendrimer conjugates in binding assays at three hAR subtypes; we used standard radioligands21–23 and membrane preparations from CHO cells (hA1AR and hA3AR) or HEK293 cells (A2AAR) stably expressing a hAR subtype.28,29 Binding affinity at the A2BAR was not determined because of the lack of a facile radioligand binding assay, but one representative conjugate was examined in a functional assay at this subtype.19 Binding results shown in Table 1 indicated that receptor affinity was maintained in many of the conjugates. We used a previously reported monomeric antagonist of the ARs, XAC 3,16 and its hexynoyl derivative 4, the synthetic precursor of the conjugates, for comparison in the binding assays. We included the parent dendrimers 8 and 9, which contained a mixture of PEG, azido, and amino terminal groups, as controls; they were inactive or only weakly active in inhibiting the binding of hA1AR and hA2AAR radioligands. The parent dendrimers 8 and 9 at 1 µM inhibited binding at the hA3AR by 72% and 52%, respectively. However, we were not able to plot a full sigmoidal curve in the competition binding experiments to determine a Ki value. The relatively high displacement of radioligand at the hA3AR by the parent dendrimers in high concentration may be related to the presence of aromatic triazolo rings on the surface of the dendrimer.
Inhibition of radioligand binding by dendrimer conjugates 14 and 18 in membranes of CHO cells expressing the hA1 and A3 ARs and HEK-293 cells expressing the hA2AAR is shown in Figure 2B,C. Inhibition of radioligand binding by the xanthine monomer 4 is shown for comparison in Figure 2A. The results were also plotted graphically to illustrate the dependence of the affinity on the degree of substitution (Figure 3). Two methods of calculating the data were used: initially according to the concentration of the dendrimer, which is a single multivalent molecule as was done in previous studies,1,2,5,9 or according to the concentration of the pendant xanthine moieties. The latter method is a mathematical construct, not an authentic equilibrium constant, because the xanthine moieties could not all be available to compete for receptor binding sites, by virtue of spatial limitations.
There was a smooth dependence of the affinity on the degree of substitution of the XAC-derivatized click-linked dendrimers in both series containing 8 or 22 PEG groups out of a total of 64 terminal sites. The more highly substituted dendrimer derivatives were considerably more potent than lightly-substituted conjugates in binding assays at the hA1AR, hA2AAR, and hA3AR. Thus, there was a regular progession toward increased affinity at each of these AR subtypes for 10 – 14 (from 6 to 34 xanthine moieties) in the more heavily PEGylated series and for 15 – 18 (from 4 to 37 xanthine moieties) in the more lightly PEGylated series. The dendrimer conjugates having 8 PEG groups displayed Ki values at the hA2AAR that varied from 33 nM to 4 nM (n = 3), for 15 and 18, respectively. The gain of affinity for the more highly PEGylated dendrimer series (Figures 3C and D) was steeper than the gain in the less highly PEGylated dendrimer series (Figures 3A and B). Thus, the affinity of the most highly substituted conjugate having 22 PEG groups 14 in binding to the ARs was particularly high, with Ki values of 15, 2.6, and 8.9 nM at hA1AR, hA2AAR and hA3AR, respectively. Therefore, the binding affinity at hAR subtypes (inversely related to apparent Ki) of these multivalent conjugates depended mainly on the degree of loading with the pharmacophore and secondarily on the degree of PEG substitution.
The order of affinity at AR subtypes for each conjugate was generally hA2AAR > hA3AR > hA1AR. Therefore, a slight selectivity of these xanthine-dendrimer conjugates for the hA2AAR in comparison to the other ARs was demonstrated (3 to 6-fold for 14 and 4 to 7-fold for 18). However, the corresponding monomers 3 and 4 displayed affinity in the order of hA2AAR > hA1AR ≥ hA3AR. Therefore, conjugation to the carrier altered the pharmacological profile of the xanthine by relatively increasing the A3AR affinity. It is unknown whether this difference is due to the multivalency or to the chemical nature of the linker and its effect on interaction with the extracellular regions of the receptor.
The association of binding activity of the conjugate with a macromolecular species was confirmed by filtering a solution of compound 14 through a 3000 MW cutoff centrifugal filter. Both the UV absorption of the retained fraction and its binding activity were maintained following ultrafiltration. Thus, the observed activity was due to the intact conjugate, rather than a diffusable small molecular weight impurity.
High affinity in AR binding was maintained in the conjugates containing prosthetic groups for receptor characterization 20, 22, 25, and 26. In fact, the Ki values of these conjugates were nearly the same as those of the precursors, i.e. XAC-conjugates 14 and 18. This was a striking observation considering the structural diversity of the attached prosthetic groups. Compound 22 might have a component of nonequilibrium binding because of the chemically reactive sulfonyl fluoride group, but this was not evident from these binding results. The affinity of the 800 nm NIR dye in 25 was identical to the precursor 14, but the more sterically bulky NIR dye in 26 caused a slight (3-fold) loss of affinity at the A3AR.
We further examined the A1AR antagonist activity of compound 14, one of the most potent conjugates for the AR subtypes in two types of functional assays in stably transfected CHO cells. We did not assay the functional effects at all the ARs; the Gi-coupled A1AR is only representative of the full biological spectrum of the xanthines, and the behavior of the conjugate could be different at each of the AR subtypes. It was found that 14 at 100 nM right-shifted the concentration-response curve to the potent A1AR agonist N6-cyclopentyladenosine (CPA) in the cyclic AMP functional assay (inhibitory) in a parallel manner (Figure 4A). The EC50 values of CPA in the absence and presence of compound 14 at 100 nM were 1.5 ± 0.3 and 45 ± 17.2 nM, respectively. The activation of the A1AR is also reported to stimulate mobilization of intracellular Ca2+ in various cells, which is likely dependent on Gβγ subunits (30). CPA stimulated Ca2+ mobilization with an EC50 of 57.5±16.8 nM (Figure 4B), and compound 14 at a concentration (10 nM) lower than its Ki value both right-shifted the CPA curve and suppressed the maximal effect, suggesting qualitative pharmacological differences between the conjugates and monomers. In contrast, the monomer XAC 3 at 10 nM or 100 nM (Figure 4C) inhibited CPA-induced calcium mobilization in A1AR-expressing CHO cells in a competitive fashion. In the presence of 100 nM XAC, CPA displayed an EC50 of 346±58 nM, corresponding to a 6-fold reduced potency.
A representative conjugate 14 was examined as an antagonist in a functional adenylate cyclase assay at the Gs-coupled hA2BAR expressed in CHO cells.19 The multivalent xanthine at a concentration of 10 nM produced an 18-fold right-shift of the NECA response curve.
In this study, we used the Cu(I)-catalyzed alkyne-azide click reaction as a chemically and biologically effective means of linking a functionalized congener of a potent AR antagonist to PAMAM dendrimers. Click chemistry was previously used to link small molecular AR agonists to dendrimeric structures (30), but this is the first systematic probing of PAMAM dendrimer conjugates of small-molecular AR antagonists. As with the agonist conjugates, high affinity of the GLiDe conjugates was achieved. The efficiency of the click reaction readily allowed a high degree of substitution in the dendrimer derivatives that could be varied by the proportion of PEG-alkyne or XAC-alkyne to dendrimer molecules that was added to the reaction.
The effectiveness of PEGylation of polymeric drug conjugtes to enhance polymer architecture, self-assembly, and bioavailability has been demonstrated.31 In the present conjugates of a hydrophobic xanthine pharmacophore, water solubility was increased and the characteristic aggregation of PAMAM dendrimers was reduced by the presence of PEG groups. We compared different degrees of loading of the GPCR ligand and of PEG groups to the dendrimeric carrier. The ability to bind the ARs was not prevented by attached PEG groups (each approximately 43 oxoethylene units in length), similar to previous findings with AR agonist GLiDe conjugates.32
More highly XAC-substituted dendrimer derivatives prepared by click chemistry were particularly potent in binding to the ARs, with some selectivity for the hA2AAR subtype in comparison to hA1 and hA3ARs. There was also indication of high potency antagonist activity at the hA2B for a representative derivative 14. Nearly nanomolar affinities in binding to three AR subtypes were attained in the most highly XAC-substituted conjugates, with affinity enhancement for the conjugates versus the monomer 4 most pronounced for the A3AR. The phenomenon of progressively increased affinity upon greater xanthine substitution was evident in both series of high and low PEG content. However, the function of the gain in AR affinity with increasing xanthine content did vary somewhat with different degrees of PEG substitution. The conjugates with an average of 22 PEG units, corresponding to a total of 44,000 D added to the molecular weight of the PAMAM, were initially weaker in binding to ARs than the 8 PEG series but gained more rapidly with increasing xanthine substitution. The molecular weights of the highly substituted conjugates 14 and 18 were ~78 and ~53 KD, respectively. Thus, the PEG content of these two conjugates amounted to approximately 56% and 30%, respectively, of the total mass.
Substitution with additional prosthetic groups of varied structure attached either by click or amide-forming reactions did not appreciably alter the binding affinity. For example, two sterically bulky NIR dyes with absorption in the range of 800 nm and 700 nm that are used in imaging in vivo33 were incorporated in conjugates with retention of high receptor binding affinity. For example, conjugate 25 is identical in A3AR affinity to its precursor and contains an average of 4 equivalents of IRdye800CW, which is excited at 785 nm and fluorescence can be measured at 830 nm, which is suitable for whole body imaging in small animals. Such GLiDe conjugates might eventually be useful for in vivo diagnostic imaging of tissue overexpressing ARs, such as occurs in disease states.34 Thus, the ability to substitute the dendrimer with chemically diverse prosthetic groups in this series of AR conjugates allows for considerable structural breadth for probing and possibly targeting of the conjugates to the ARs or to cells having specific markers on the surface.
The advantages of nanocarriers for small molecular weight drugs include enhanced pharmacokinetic and pharmacodynamic properties. Our approach derivatizes a small molecular GPCR ligand for coupling to a carrier, but allows it to remain active while covalently attached.1,15 Internalization is not needed or desired, because the binding sites of the receptors are accessible only from the extracellular medium. Molecular modeling indicated that flexible PAMAM dendrimers are able to spread over the cell surface, and purine conjugates of G3 dendrimers and higher generations can readily adopt a conformation that can bridge multiple GPCR binding sites.8 In the theoretical model of a dendrimer conjugate of an A2AAR agonist bridging adjacent binding sites of an A2AAR homodimer, the distance between the two sites was ~35 Å, which was easily accommodated by two intermediate-spaced arms of the G3 dendrimer in a nonextended conformation. This roughly spherical G3 dendrimer conjugate had a diameter of ~67 Å, and the xanthine moieties of the present G4 conjugates in a nonextended conformation might be expected to achieve an estimated maximum separation of ~80 Å. This distance would be increased by stretching of the dendrimer over a surface.35
GPCRs tend not to be evenly or randomly distributed over the surface of a given cell; we demonstrated this for the hA3AR expressed heterologously in CHO cells (2). In general, GPCRs may exist as complexes with other receptors and associated proteins. The receptors to which a given GPCR pairs or forms a higher-order aggregate can dramatically affect its pharmacology, and the binding of an agonist to both protomers does not have equivalent pharmacological effects.36 Thus, a multivalent GPCR ligand might display unanticipated, complex effects.5 There is a lack of GPCR ligands that are designed specifically to interact with dimeric and oligomeric receptors. We do not have effective ligand tools for separating the effects of one receptor dimeric combination from another. The approach of mixing ligands for two different receptors.3,37,41 promises to be a means of achieving selectivity for receptors in a given tissue, based on association or aggregation of GPCRs, in comparison to another tissue in which the same receptor may be alternatively paired. At this early stage, we do not know the topological requirements for selectively targeting these combinations of receptors, because we are just beginning to detect their existence. Nevertheless, providing tools for studying these receptor aggregates is one of the objectives of this study. The multivalent binding to other types of cell surface receptors has been explored through sytematic structural modification of the carrier polymer.42
There is evidence here for multivalent binding of the same conjugate to more than one AR site. The fact that the AR affinity rose substantially with increased xanthine content, even when calculating IC50 values according to the concentration of the pharmacophore rather than the dendrimer as shown in Figure 3D, suggests that the same GLiDe conjugate molecule may bind to multiple AR binding sites. The interaction of the multivalent GLiDe conjugates can now be studied in cells expressing varying levels of a given AR or with other receptors coexpressed.
In previous studies, we have noted that the pharmacological profile of GLiDe conjugates can differ from that of the monomeric ligands both quantitatively and qualitatively.5,11 For example, the dendrimer conjugates of AR agonists have displayed a shift in the AR subtype selectivity or in the correspondence between the binding and functional potencies.2,11 Here we have detected a qualitative difference in the A1AR antagonistic response between a potent, multivalent xanthine conjugate and its corresponding monomer. The conjugate 14 acted as a competitive antagonist of the Giα-dependent effect of inhibition of adenylate cyclase, but at comparable concentrations acted as a noncompetitive inhibitor of calcium mobilization, which is likely independent of Giα. The functional effects at other AR subtypes were not determined here and will be the subject of further studies.
In conclusion, we have synthesized and characterized pharmacologically a series of multivalent dendrimeric derivatives of a potent strategically-functionalized AR antagonist. GLiDe conjugates provide different quantitative and qualitative pharmacological properties than do monomers. Prosthetic groups designed for therapeutic targeting or reporter groups for diagnostic purposes might be introduced. Various therapeutic applications12 may be proposed for multivalent conjugates of AR antagonists. In vivo studies will be needed to fully characterize these potent ligands biologically and determine their advantage over monomeric drugs.
We thank Dr. John Lloyd and Dr. Noel Whittaker (NIDDK) for the mass spectral determinations, Dr. Martin Garraffo (NIDDK) for FT-IR measurements, and Khai Phan (NIDDK) for pharmacological measurements. We thank Dr. Francesca Deflorian (NIDDK) and Dr. Michael Kilbey III and Dr. Kunlun Hong (Oak Ridge National Laboratory, Knoxville TN) for helpful discussion. This research was supported by the Intramural Research Program of the NIH, NIDDK.
Supporting Information Available: 1H NMR and mass spectra of representative compounds 14 and 18. This material is available free of charge via the Internet at http://pubs.acs.org/BC.