Synthesis of dimeric analogs of adenophostin A that potently evoke Ca2+ release through IP3 receptors.

Inositol 1,4,5-trisphosphate receptors (IP3Rs) are tetrameric intracellular channels through which many extracellular stimuli initiate the Ca2+ signals that regulate diverse cellular responses. There is considerable interest in developing novel ligands of IP3R. Adenophostin A (AdA) is a potent agonist of IP3R and since some dimeric analogs of IP3R ligands are more potent than the corresponding monomer; we considered whether dimeric AdA analogs might provide agonists with increased potency. We previously synthesized traizolophostin, in which a simple triazole replaced the adenine of AdA, and showed it to be equipotent to AdA. Here, we used click chemistry to synthesize four homodimeric analogs of triazolophostin, connected by oligoethylene glycol chains of different lengths. We evaluated the potency of these analogs to release Ca2+ through type 1 IP3R and established that the newly synthesized dimers are equipotent to AdA and triazolophostin.

Introduction

Inositol 1,4,5-trisphosphate (IP3, 1, Fig. 1) is an important secondary messenger that evokes Ca2+ release from intracellular stores through its interaction with IP3 receptors (IP3R) in the endoplasmic reticulum.[1] IP3R are large tetrameric proteins, within which IP3 binding to each of the four subunits is required to initiate opening of the Ca2+-permeable channel.[2] High-resolution structures of the IP3-binding core (IBC, residues 224–604) have defined the interactions of IP3 with IP3R.[3] More recently, structures of the N-terminal region (residues 1–604)[4] alongside a structure of the complete IP3R derived from cryo-electron microscopy have begun to suggest how IP3 binding might trigger the opening of the intrinsic pore of IP3R.[5]

Fig. 1

The structures of IP3 (1), adenophostin A (2) and triazolophostin (3).

There is continuing interest in the development of potent agonists and antagonists of IP3R.[6] The fungal metabolite, adenophostin A (AdA, 2, Fig. 1), binds to IP3R with greater affinity than IP3 and it is more potent than IP3 in evoking Ca2+ release.[7] AdA analogs with a nucleobase or base-surrogate are also more potent than IP3.[8] Molecular docking[8j,m,9] and mutation studies[10] suggest that a cation–π interaction between the adenine moiety of AdA and Arg504 within the IBC contributes to the increased affinity of AdA. We recently reported synthesis of a library of active AdA analogs, triazolophostins, by using a click chemistry approach.[11] These potent analogs have substituted triazoles as adenine surrogates. The simplest analog, triazolophostin (3, Fig. 1) was equipotent with AdA.

Multimeric ligands often have greater affinity than monomeric ligands.[12] This can be due to simultaneous binding to more than one binding site or a statistical effect arising from the local increase in ligand concentration.[13] The former is unlikely for IP3R because the orientation of the IP3-binding sites within the tetrameric IP3R is unlikely to allow simultaneous binding of two ligands linked by a short tether.[4b,14]

A few multimeric ligands of IP3R have been reported. Before the location of the IP3-binding sites within IP3R was known, clustered bi- and tetra-dentate analogs of ribophostin (4, Fig. 2A) were synthesized, anticipating that if the spacing between the linked ligands was appropriate they might bind simultaneously to the four IP3-binding sites.[15] However, the potencies of the monomeric and polymeric ligands were rather similar. Several homodimeric[16] and heterodimeric[17] ligands of IP3 (5–10, Fig. 2B), particularly those with short linkers, were shown to bind to IP3R with increased affinity.[13] Very recently, dimers of 2-O-Bt-IP4/IP5 (11, Fig. 2C) were shown to be antagonists of IP3Rs.[18] These results demonstrate that dimeric IP3R ligands can provide useful tools, some of which have greater affinity than the monomeric ligands. We therefore considered whether dimers of AdA might be more potent than AdA.

Fig. 2

The representative structures of (A) ribophostin dimer 4, (B) homo and hetero dimers of IP3 (5–10) and (C) dimers of 2-Bt-IP4/IP5 11.

Results and discussion

As the synthesis of AdA dimers is challenging, we decided to make oligoethylene glycol-tethered dimers of triazolophostin (Fig. 3). We envisaged that use of click reaction[19] with a linker connected to alkyne at both termini would ensure both formation of triazole and link the two monomers in one step. Previous studies suggested that short linkers were most likely to improve the affinity of homodimers.[13] We therefore selected spacers smaller than hexaethylene glycol. The linkers 14a–d were synthesized by slightly modifying previously reported procedures.[20] The oligoethylene glycols were first co-evaporated with toluene and then treated with sodium hydride in the presence of excess propargyl bromide to get dipropargyl polyethylene glycols 14a–d in good to excellent yields. The azide 13 was synthesized from glucose and xylose by several protection–deprotection reactions followed by phosphorylation as reported earlier.[11] The azide 13 was then treated with dialkynyl polyethylene glycols 14a–d in the presence of Cu(i) catalyst to get fully protected triazolophostin dimers 15a–d in good yields. The debenzylation of protected triazolophostin dimers 15a–d was carried out using transfer hydrogenolysis in the presence of palladium and cyclohexene under reflux condition and the products were purified by ion-exchange chromatography to yield dimers 12a–d, in excellent yields (Scheme 1).

Fig. 3

The structure of dimeric analogs of triazolophostin 12.

Scheme 1

Synthesis of triazolophostin dimers. Reagents and conditions: (a) ref. 11; (b) Cu, CuSO4, H2O : BuOH (1 : 1, v/v), rt, 24 h; (c) Pd(OH)2/C, cyclohexene, MeOH : H2O (10 : 1, v/v), 80 °C. 4 h; (a), n = 2; (b), n = 3; (c), n = 4; (d), n = 6.

The dimeric ligands 12a–d were screened for their abilities to evoke Ca2+ release through IP3R (Table 1, Fig. 4). All four dimers were full agonists of IP3R, more potent than IP3, but similar in their potency to AdA and the monomer, triazolophostin. The similar potencies of 12a–d irrespective of their tether length suggest that these ligands might be interacting with IP3R1 in monodentate fashion.

Table 1

Responses of IP3R1 to IP3 (1), monomer (3) and its dimeric analogs 12a–d

Ligand	pEC50	EC50 (nM)	EC50 w.r.t. 1 b	Max. response (%)	n H
IP3 (1)	6.72 ± 0.12	190.5	1	69 ± 3	1.40 ± 0.16
Monomer (3)	7.86 ± 0.17	13.8	13.8	65 ± 1	1.66 ± 0.21
12a	7.83 ± 0.18	14.8	12.9	68 ± 2	1.33 ± 0.12
12b	7.85 ± 0.13	14.1	13.5	66 ± 1	1.89 ± 0.13
12c	7.62 ± 0.11	24.0	7.9	61 ± 3	1.60 ± 0.16
12d	7.84 ± 0.12	14.4	13.2	60 ± 1	1.94 ± 0.47

Maximal Ca2+ release, the half-maximally effective ligand concentration (EC50), –log EC50 (pEC50) and Hill coefficient (n H) are shown as means ± SEM (n = 3).

The EC50 value of each ligand is also shown relative to that for IP3 (1) (EC150/ECanalog50).

Fig. 4

Summary of Ca2+ release from permeabilized DT40-IP3R1 cells evoked by IP3, monomer 3 and its dimeric analogs 12a–d.

Conclusions

In conclusion, based on several previous reports that dimeric IP3R ligands can be more potent than the corresponding monomers, we anticipated that dimers of AdA might have increased potency. We used click chemistry to synthesize dimers of a potent analog of AdA (triazolophostin) linked by spacers of different length. In assays of Ca2+ release through IP3R, the dimeric ligands were no more potent than the corresponding monomer (3). This suggests that whereas dimeric derivatives of IP3 have reduced efficacy but improved affinity,[10,21] dimerization of AdA analogs does not improve their affinity.

Experimental section

General methods

The chemicals were purchased from commercial sources and used as received. The TLC plates were visualized under UV light and by dipping plates into either phosphomolybdic acid in MeOH or sulphuric acid in ethanol, followed by heating. All NMR experiments were carried out on a 500 MHz NMR spectrometer and at room temperature. Tetramethylsilane (TMS, δ 0.0 ppm) or the solvent reference (CDCl3, δ 7.26 ppm; D2O, δ 4.79 ppm) relative to TMS were used as the internal standard. The data are reported as follows: chemical shift in ppm (δ) (multiplicity [singlet (s), doublet (d), doublet of doublet (dd), triplet (t), quartet (q), and multiplet (m)], coupling constants [Hz], integration and peak identification). All NMR signals were assigned on the basis of 1H NMR, 13C NMR, COSY and HMQC experiments. 13C NMR spectra were recorded with complete proton decoupling. Carbon chemical shifts are reported in ppm (δ) relative to TMS with the respective solvent resonance as the internal standard. The concentration of the compounds for 1H NMR was 5 mg per 0.5 mL and for 13C NMR it was 20 mg per 0.5 mL for protected compounds and 5–7 mg per 0.5 mL for final compounds in case of 1H and 13C NMR. Modified Brigg's phosphate assay[22] was employed to quantify each triazolophostin 12a–d. Silica gel 230–400 mesh was used to perform flash column chromatography.

General procedure for syntheses of fully protected triazolophostin dimers

To a solution of azide 13 (0.144 mmol) and dialkynyl PEG 14a–d (0.072) in H2O/BuOH (1/1, v/v, 2 mL) was added Cu (0.036 g, 0.57 mmol) and CuSO4 (8 mg, 0.028 mmol) and stirred at room temperature for 24 h. The reaction was monitored by TLC. When the TLC showed complete disappearance of the azide 13, the mixture was filtered through a Celite bed and was partitioned between ethyl acetate and water. The organic layer was washed with brine. The organic layer was dried over anhyd. sodium sulphate, filtered and concentrated under reduced pressure. The residue thus obtained was purified by flash column chromatography using a mixture of acetone, diethyl ether and petroleum ether (4 : 2 : 15 v/v/v) as eluent to get pure 15a–d as a colourless gum.

Protected triazolophostin dimer 15a

Click reaction of azide 13 (0.2 g, 0.144 mmol) with diyne 14a (0.011 g, 0.072 mmol) gave the protected dimer 15a (0.18 g, 85%) as a colourless gum. 1H NMR (500 MHz, CDCl3) δ: 3.47–3.57 (m, 18H, H-2′′, H-4′′, H-6′′A, and DEG-H), 3.73–3.75 (m, 2H, H-5′′), 4.20–4.23 (m, 2H, PhCH 2), 4.27–4.32 (m, 8H, H-5′A and PhCH 2), 4.30–4.45 (m, 10H, H-3′, H-4′, H-5′B, H-6′′B and PhCH 2), 4.57–4.59 (m, 2H, PhCH 2), 4.63–4.66 (m, 4H, PhCH 2), 4.68–4.73 (m, 6H, PhCH 2), 4.80–4.93 (m, 16H, H-3′′, H-4′′ and PhCH 2), 5.11 (d, 2H, J = 3.2 Hz, H-1′′), 5.26–5.28 (m, 2H, H-2′) 6.24 (d, 2H, J = 5.0 Hz, H-1′), 7.00 (d, 4H, J = 7.0 Hz, Ar-H), 7.05–7.19 (m, 82H, Ar-H), 7.26 (d, 4H, J = 7.0 Hz, Ar-H), 7.60 (s, 2H, H-5); 13C NMR (125 MHz, CDCl3) δ: 64.2, 68.3, 69.1, 69.2, 69.3, 69.5, 69.6, 69.7, 69.9, 70.1, 70.4, 71.9, 73.3, 73.5, 74.1, 78.0, 78.5, 82.8, 90.1, 95.7, 121.6, 127.6, 127.7, 127.9, 128.0, 128.1, 128.3, 128.4, 128.5, 135.2, 135.7, 135.8, 136.1, 137.3, 137.5, 138.0, 145.2; 31P NMR (202.4 MHz, CDCl3) δ: –1.484, –1.928, –2.146; HRMS (ESI) mass calcd for C158H166N6O39P6 [M]+ 2956.9616, found 2956.9620.

Protected triazolophostin dimer 15b

Click reaction of azide 13 (0.2 g, 0.144 mmol) with diyne 14b (0.016 g, 0.072 mmol) gave the protected dimer 15b (0.185 g, 86%) as a colourless gum. 1H NMR (500 MHz, CDCl3) δ: 3.44–3.57 (m, 22H, H-2′′, H-4′′, H-6′′A, and TEG-H), 3.75 (bs, 2H, H-5′′), 4.21–4.30 (m, 10H, H-5′A and PhCH 2), 4.42–4.43 (m, 10H, H-3′, H-4′, H-5′B, H-6′′B and PhCH 2), 4.57–4.59 (m, 2H, PhCH 2), 4.64–4.66 (m, 6H, PhCH 2), 4.68–4.73 (m, 4H, PhCH 2), 4.84–4.92 (m, 16H, H-3′′, H-4′′ and PhCH 2), 5.11 (bs, 2H, H-1′′), 5.27 (bs, 2H, H-2′) 6.24 (d, 2H, J = 5.0 Hz, H-1′), 7.00–7.25 (m, 90H, Ar-H), 7.61 (s, 2H, H-5); 13C NMR (125 MHz, CDCl3) δ: 63.2, 67.3, 68.0, 68.3, 68.4, 68.6, 68.7, 69.1, 69.4, 70.9, 72.3, 75.7, 75.8, 75.9, 76.9, 77.5, 81.7, 89.0, 94.7, 120.6, 126.7, 127.0, 127.2, 127.4, 134.2, 134.6, 135.1, 136.3, 136.5, 137.0, 144.2; 31P NMR (202.4 MHz, CDCl3) δ: –1.486, –1.935, –2.155; HRMS (ESI) mass calcd for C160H170N6O40P6 [M]+ 3000.9879, found 3000.9877.

Protected triazolophostin dimer 15c

The reaction of azide 13 (0.2 g, 0.144 mmol) with diyne 14c (0.019 g, 0.072 mmol) gave the protected dimer 15c (0.175 g, 81%) as a colourless gum. 1H NMR (500 MHz, CDCl3) δ: 3.54–3.67 (m, 26H, H-2′′, H-4′′, H-6′′A, and TetraEG-H), 3.84 (bs, 2H, H-5′′), 4.30–4.32 (m, 2H, PhCH 2), 4.37–4.39 (m, 8H, H-5′A and PhCH 2), 4.48–4.53 (m, 10H, H-3′, H-4′, H-5′B, H-6′′B and PhCH 2), 4.66–4.68 (m, 2H, PhCH 2), 4.73–4.74 (m, 4H, PhCH 2), 4.78–4.82 (m, 6H, PhCH 2), 4.92–4.94 (m, 10H, H-3′′, H-4′′ and PhCH 2), 4.97–5.03 (m, 6H, PhCH 2), 5.20 (bs, 2H, H-1′′), 5.36 (bs, 2H, H-2′) 6.34 (d, 2H, J = 5.0 Hz, H-1′), 7.09–7.34 (m, 90H, Ar-H), 7.75 (s, 2H, H-5); 13C NMR (125 MHz, CDCl3) δ: 64.2, 68.3, 69.1, 69.15, 69.2, 69.3, 69.39, 69.5, 69.5, 69.6, 69.8, 69.9, 70.4, 70.5, 70.55, 71.9, 73.3, 73.5, 76.7, 82.8, 95.7, 121.6, 127.6, 127.7, 127.78, 127.8, 127.9, 128.0, 128.1, 128.3, 128.37, 128.4, 128.5, 128.55, 128.6, 135.2, 136.1, 136.2, 137.3, 137.5, 138.0; 31P NMR (202.4 MHz, CDCl3) δ: –1.468, –1.908, –2.138; HRMS (ESI) mass calcd for C162H174N6O41P6 [M]+ 3045.0141, found 3045.0131.

Protected triazolophostin dimer 15d

The reaction of azide 13 (0.2 g, 0.144 mmol) with diyne 14d (0.026 g, 0.072 mmol) gave the protected dimer 15d (0.185 g, 82%) as a colourless gum. 1H NMR (500 MHz, CDCl3) δ: 3.53 (bs, 34H, H-2′′, H-4′′, H-6′′A, and HEG-H), 3.74 (bs, 2H, H-5′′), 4.23–4.28 (m, 10H, H-5′A and PhCH 2), 4.42 (bs, 10H, H-3′, H-4′, H-5′B, H-6′′B and PhCH 2), 4.56–4.58 (m, 2H, PhCH 2), 4.65–4.71 (m, 10H, PhCH 2), 4.83–4.91 (m, 16H, H-3′′, H-4′′ and PhCH 2), 5.11 (bs, 2H, H-1′′), 5.27 (bs, 2H, H-2′) 6.24 (d, 2H, J = 5.0 Hz, H-1′), 6.99–7.24 (m, 90H, Ar-H), 7.62 (s, 2H, H-5); 13C NMR (125 MHz, CDCl3) δ: 64.2, 69.1, 69.16, 69.2, 69.3, 69.4, 69.5, 69.6, 69.7, 69.8, 69.9, 70.0, 70.4, 70.5, 71.9, 73.3, 73.5, 82.8, 95.7, 127.5, 127.8, 127.7, 127.75, 127.76, 127.8, 127.9, 128.0, 128.1, 128.2, 128.3, 128.4, 128.46, 128.49, 128.5, 128.6, 135.2, 136.1, 137.3, 137.5, 138.0; 31P NMR (202.4 MHz, CDCl3) δ: –1.482, –1.919, –2.168; HRMS (ESI) mass calcd for C166H182N6O43P6 [M]+ 3133.0665, found 3133.0669.

General procedure for syntheses of triazolophostin dimers 12a–d

The protected triazolophostin dimers 15a–d (0.15–0.175 g, 0.05–0.055 mmol) were treated with cyclohexene (3 mL) and Pd(OH)2 (20% on carbon, 50 mg) in a mixture of methanol and water (9 : 1 v/v, 10 mL) at 80 °C for 4 h. The reaction mixture was then cooled, filtered through a membrane filter, washed successively with methanol and water. The combined filtrate was evaporated under reduced pressure. The crude product thus obtained was purified by ion-exchange column chromatography on Q-Sepharose matrix using 0–1.0 M TEAB as eluent to get pure triazolophostin dimers 12a–d.

Triazolophostin dimer 12a

The global debenzylation of 15a (0.15 g, 0.05 mmol) gave 46 mg (69%) of triazolophostin dimer 12a as a white hygroscopic solid: 1H NMR (500 MHz, D2O) δ: 3.63–3.65 (m, 8H, DEG-H), 3.70–3.83 (m, 12H, H-5′A, H-2′′, H-6′′ and DEG-H), 4.09–4.10 (m, 2H, H-5′′), 4.41 (bs, 2H, H-4′), 4.48 (bs, 2H, H-5′B), 4.62–4.65 (m, 6H, H-3′, H-3′′ and H-4′′), 5.16 (bs, 2H, H-2′), 5.24 (bs, 2H, H-1′′), 6.36 (bs, 2H, H-1′), 8.22 (s, 2H, H-5); 13C NMR (125 MHz, D2O) δ: 60.1, 60.7, 62.8, 68.8, 69.4, 70.5, 71.5, 72.8, 73.7, 76.4, 77.9, 83.8, 90.9, 97.9, 124.3, 144.1; 31P NMR (202.4 MHz, D2O) δ: 3.504, 3.583, 4.301; HRMS (ESI) mass calcd for C32H58N6O39P6 [M]+, 1336.1165, found: 1336.1169.

Triazolophostin dimer 12b

The global debenzylation of 15b (0.155 g, 0.051 mmol) gave 51 mg (72%) of triazolophostin dimer 12b as a white hygroscopic solid: 1H NMR (500 MHz, D2O) δ: 3.56–3.60 (m, 12H, TEG-H), 3.69–3.74 (m, 12H, H-5′A, H-2′′, H-6′′ and TEG-H), 4.06 (bs, 2H, H-5′′), 4.36 (bs, 2H, H-4′), 4.44 (bs, 2H, H-5′B), 4.50–4.60 (m, 6H, H-3′, H-3′′ and H-4′′), 5.12 (bs, 2H, H-2′), 5.18 (bs, 2H, H-1′′), 6.31 (bs, 2H, H-1′), 8.18 (s, 2H, H-5); 13C NMR (125 MHz, D2O) δ: 60.1, 60.7, 62.8, 68.8, 69.4, 69.48, 70.4, 71.5, 72.8, 73.7, 76.4, 77.8, 83.8, 90.8, 97.9, 124.3, 144.1; 31P NMR (202.4 MHz, D2O) δ: 3.451 (2 × P), 4.224; HRMS (ESI) mass calcd for C34H62N6O40P6 [M]+, 1380.1427, found: 1380.1420.

Triazolophostin dimer 12c

The global debenzylation of 15c (0.16 g, 0.052 mmol) gave 64 mg (85%) of triazolophostin dimer 12c as a white hygroscopic solid: 1H NMR (500 MHz, D2O) δ: 3.57–3.61 (m, 16H, TetraEG-H), 3.69–3.74 (m, 12H, H-5′A, H-2′′, H-6′′ and TetraEG-H), 4.05 (bs, 2H, H-5′′), 4.37 (bs, 2H, H-4′), 4.44 (bs, 2H, H-5′B), 4.58–4.61 (m, 6H, H-3′, H-3′′ and H-4′′), 5.12 (bs, 2H, H-2′), 5.19 (bs, 2H, H-1′′), 6.32 (bs, 2H, H-1′), 8.19 (s, 2H, H-5); 13C NMR (125 MHz, D2O) δ: 60.1, 60.7, 62.9, 68.9, 69.4, 69.5, 70.5, 71.5, 72.8, 73.7, 76.4, 77.9, 83.8, 90.8, 98.0, 124.3, 144.0; 31P NMR (202.4 MHz, D2O) δ: 3.478 (2 × P), 4.259; HRMS (ESI) mass calcd for C36H66N6O41P6 [M]+, 1424.1690, found: 1424.1699.

Triazolophostin dimer 12d

The global debenzylation of 15d (0.175 g, 0.055 mmol) gave 65 mg (77%) of triazolophostin dimer 12d as a white hygroscopic solid: 1H NMR (500 MHz, D2O) δ: 3.58–3.72 (m, 24H, HEG-H), 3.77–3.81 (m, 12H, H-5′A, H-2′′, H-6′′ and HEG-H), 4.01 (bs, 2H, H-5′′), 4.38–4.48 (m, 4H, H-4′ and H-5′B), 4.58–4.63 (m, 6H, H-3′, H-3′′ and H-4′′), 5.12 (bs, 2H, H-2′), 5.20 (bs, 2H, H-1′′), 6.32 (bs, 2H, H-1′), 8.19 (s, 2H, H-5); 13C NMR (125 MHz, D2O) δ: 60.2, 60.8, 62.9, 68.9, 69.4, 69.5, 70.8, 71.7, 72.6, 73.7, 76.3, 77.4, 83.8, 90.9, 97.9, 124.2, 144.2; 31P NMR (202.4 MHz, D2O) δ: 3.482 (2 × P), 4.258; HRMS (ESI) mass calcd for C40H74N6O43P6 [M]+, 1512.2214, found: 1512.2210.

Biological assay

Ca2+ release from the intracellular stores of saponin-permeabilized DT40 cells expressing only type 1 IP3Rs was measured using a low-affinity Ca2+ indicator (Mag-fluo-4) trapped within the endoplasmic reticulum as described previously.[11] Briefly, Ca2+ uptake was initiated by addition of 1.5 mM MgATP in cytosol-like medium (140 mM KCl, 20 mM NaCl, 1 mM EGTA, 20 mM PIPES, pH 7.0, free [Ca2+] ∼220 nM after addition of ATP) containing p-trifluoromethoxyphenylhydrazone (FCCP) to inhibit mitochondria. After about 120 s, the triazolophostin analogs were added with cyclopiazonic acid (10 μM) to inhibit further Ca2+ uptake. Ca2+ release was assessed 10–20 s after addition of the analog, and expressed as a fraction of the ATP-dependent Ca2+ uptake.