Peripheral site ligand conjugation to a non-quaternary oxime enhances reactivation of nerve agent-inhibited human acetylcholinesterase

Toxicology Letters 206 (2011) 54–59

Contents lists available at ScienceDirect

Toxicology Letters journal homepage: www.elsevier.com/locate/toxlet

Peripheral site ligand conjugation to a non-quaternary oxime enhances reactivation of nerve agent-inhibited human acetylcholinesterase Martijn C. de Koning ∗ , Marco van Grol, Daan Noort TNO, Department of Chemical Toxicology, P.O. Box 45, 2280 AA Rijswijk, The Netherlands

a r t i c l e

i n f o

Article history: Available online 8 April 2011 Keywords: Acetylcholinesterase Non-quaternary oxime Nerve agent Peripheral site Reactivation

a b s t r a c t Commonly employed pyridinium-oxime (charged) reactivators of nerve agent inhibited acetylcholinesterase (AChE) do not readily pass the blood brain barrier (BBB) because of the presence of charge(s). Conversely, non-ionic oxime reactivators often suffer from a lack of reactivating potency due to a low affinity for the active site of AChE. It was therefore hypothesized that an extra contribution in affinity may be achieved by covalently connecting a peripheral site ligand (PSL) to a non-ionic reactivator, which may result in a higher reactivation potency of the total construct. This validity of this approach, which proved successful for charged pyridinium oximes in earlier work, is now further exemplified with the covalent linkage of a neutral PSL via a spacer to a non-ionic and otherwise almost non-reactivating ␣-ketoaldoxime. It is demonstrated that the linkage of the PSL resulted in a remarkable increase in reactivation potency of the hybrid compounds. Although the molecules reported here are still inefficient reactivators compared to the current pyridinium oximes, the presented approach holds promise for the future design and synthesis of non-ionic oxime reactivators with improved BBB penetration and may be suited as well for non-oxime reactivators thus further widening the scope in the ongoing search for broad-spectrum reactivators. © 2011 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Acetylcholinesterase (AChE) is a serine hydrolase which has an active site (or A-site) at the bottom of a long narrow gorge and a second substrate binding site at the entrance of the gorge, called peripheral anionic site (PAS or P-site) (Masson et al., 1997; Sussman et al., 1991; Szegletes et al., 1999). The A-site catalyzes the hydrolysis of the neurotransmitter acetylcholine (ACh) at cholinergic synapses at a rate close to the diffusion-controlled limit (Quinn, 1987). The action of AChE can be interrupted by organophosphate (OP) compounds, such as insecticides (e.g. chlorpyrifos) and nerve agents (e.g. sarin, soman, tabun, VX, etc.). Those agents arrest ACh hydrolysis by the permanent phosphorylation of the key serine residue in the A-site (Marrs, 1993). As a result, ACh levels accu-

Abbreviations: AChE, acetylcholinesterase; ACh, acetylcholine; ACN, acetonitrile; AmONO, isomamylnitrite; A-Site, active site; BBB, blood brain barrier; CNS, central nervous system; DCE, dichloroethane; DCM, dichloromethane; DMF, dimethylformamide; EtOAc, ethylacetate; hAChE, human acetylcholinesterase; HPLC, high performance liquid chromatography; LC, liquid chromatography; MS, mass spectrometry; n-BuOH, n-butanol; NMR, nuclear magnetic resonance; OP, organophosphate; PAS, peripheral anionic site; P-Site, peripheral site; PSL, peripheral site ligand; TFA, trifluoroacetic acid; TMSCl, trimethylsilylchloride; UV, ultraviolet absorption. ∗ Corresponding author. Tel.: +31 88 8661320; fax: +31 88 8666938. E-mail address: [email protected] (M.C. de Koning). 0378-4274/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.toxlet.2011.04.004

mulate which leads to overstimulation of the cholinergic receptors and, as a consequence, seizures and ultimately death. Reactivation of OP-inhibited AChE can be accomplished by the timely administration of certain oximes (Worek et al., 2007), such as HI-6, Obidoxime and 2-PAM (Fig. 1) that due to their nucleophilicity mediate cleavage of the covalent bond between the phosphorus atom of the OP-inhibitor and the serine residue of the enzyme. One drawback of the current oximes is that there is no broad spectrum reactivator that sufficiently reactivates AChE inhibited by the various types of OPs (Bajgar et al., 2007). The mono-quaternary oxime 2-PAM is the drug of choice for a number of countries (Luo et al., 2007). On the other hand, the bis-quaternary oximes (i.e. HI-6 and Obidoxime) are generally considered more effective than mono-quaternary oximes, however, whereas HI-6 reactivates sarin- and VX-inhibited AChE it is outperformed by obidoxime for tabun-poisoning (Lundy et al., 2006). A second major drawback of these (bis-)quaternary oximes is that these oximes are permanently charged and therefore show low penetration (e.g. about 10% for 2PAM; Sakurada et al., 2003) of the blood brain barrier (BBB). As a result, the current oximes show only limited activity in the central nervous system (CNS). Thus, much effort has been directed in the past to the development of A-site targeting, non-quaternary (uncharged) oximes (Fig. 2) to improve BBB penetration (Bedford et al., 1986a,b; Benschop et al., 1970, 1979; Beznosko et al., 1977; Degorre et al., 1988; Kenley et al., 1981, 1984, 1985). Unfortunately, the majority of the compounds tested showed unsatisfactory reac-

M.C. de Koning et al. / Toxicology Letters 206 (2011) 54–59

55

Fig. 1. Currently employed oximes in the treatment of patients intoxicated by nerve agents or OP-pesticides.

tivation of OP-inhibited AChE. Although these molecules were designed to possess a sufficiently reactive oxime group, the main reason underlying the loss in reactivation potency is the absence of charge, which reduces the affinity of the oximes for the A-site of AChE (Kuca et al., 2006). In more recent contributions, attempts were made to improve the BBB penetration of the charged PAM-, HI-6- and obidoxime derivatives by increasing their lipophilicity by replacing the methyl group in 2-PAM by more lipophilic alkyl groups (Ohta et al., 2006; Okuno et al., 2008) or by synthesizing fluorinated derivatives (Jeong et al., 2009a,b). We have recently disclosed an alternative approach in the design and synthesis of novel reactivators (De Koning et al., 2011). This approach entailed the covalent linkage of a reactivating pyridinium oxime to a neutral ligand with affinity for the second substrate binding site in AChE, the P-site (1a–1c in Fig. 3). These conjugates were expected to bind the (inhibited) AChE by a dual site binding mode. It was demonstrated that conjugates 1a–1c not only showed higher affinity for human AChE (hAChE) but also showed enhanced reactivation potency compared to the corresponding Asite targeting parent reactivator 1d. It was therefore hypothesized that the extra contribution in affinity by the PSL may compensate for the loss in affinity associated with the use of a non-quaternary oxime (as in 2a–2c in Fig. 3) leading to enhanced reactivation potency of the hybrid compound. We here report the synthesis and biological evaluation of non-quaternary conjugates 2a–2c and we demonstrate that these conjugates have an enhanced reactivation potency towards nerve agent inhibited hAChE compared to the parent, non-quaternary oxime 2d. 2. Materials and methods 2.1. General All starting compounds (4-hydroxyacetophenon, 4-methoxyacetophenon) and reagents were purchased from Aldrich and used as received. Solvents were obtained

from Biosolve and used as received. Analytical LC was conducted on an AKTA system using an Alltima C18 analytical column (5 ␮m particle size, flow: 1.0 mL/min). Absorbance was measured at 214 nm and 254 nm. Solvent system: A: 5% ACN, 0.1% TFA; B: 80% ACN, 0.1% TFA. Gradients of B were applied over 20 min unless otherwise stated. HPLC purifications were conducted on the AKTA system supplied with a semi-preparative Alltima C18 column (5 ␮m particle size, running at 5 mL/min). UV measurements for the preparation of the biological samples were carried out using a Jasco V-560 UV–VIS spectrometer operating at 308 nm. LC/electrospray tandem mass spectrometric analyses for obtaining structural information were conducted on a Q-TOF hybrid instrument equipped with a standard Z-spray electrospray interface (Micromass, Altrincham, UK) and an Alliance, type 2690 liquid chromatograph (Waters, Milford, MA, USA). The chromatographic hardware consisted of a pre-column splitter (type Acurate; LC Packings, Amsterdam, The Netherlands), a sixport valve (Valco, Schenkon, Switzerland) with a 10 or 50 ␮L injection loop mounted and a PepMap C18 (LC Packings) column (15 cm × 1 mm I.D., 3 ␮m particles). A gradient of eluents A (H2 O with 0.2% (v/v) formic acid) and B (acetonitrile with 0.2% (v/v) formic acid) was used to achieve separation. The following gradient was applied: 0–5 min, 100% solution A, flow 0.1 mL/min; 5–60 min, 100% A to 30% A, flow 0.6 mL/min. The flow delivered by the liquid chromatograph was split pre-column to allow a flow of approximately 6 ␮L/min through the column and into the electrospray MS interface. MS/MS product ion spectra were recorded in positive ion mode using a cone voltage between 15 kV and 30 kV and a collision energy between 11 eV and 25 eV, with argon as the collision gas (at an indicated pressure of 10−4 mbar). Other mass spectrometric analyses were carried out on a TSQ Quantum Ultra mass spectrometer (Finnigan, Thermo Electron Corporations, San Jose, USA) equipped with an Acquity Sample Manager and Binary Solvent Manager (Waters, Milford, USA). For LC–MS experiments, the liquid chromatograph was connected to the mass spectrometer source via the Sample Manager equipped with a 10 ␮L loop and an Acquity BEH C18 column (1.7 ␮m particles, 1 mm × 100 mm; Waters, Milford, USA). The liquid chromatography system was run with a 25 min linear gradient from 100% A to A/B 55.5/45.5 v/v (A: 0.2% formic acid in water; B: 0.2% formic acid in acetonitrile) at a flow rate of 0.09 mL/min. The TSQ Quantum Ultra mass spectrometer was operated with a spray voltage of 3 kV, a sheath gas pressure of 41 A.U., aux gas pressure of 2 A.U. and a capillary temperature of 350 ◦ C. Positive electrospray product ion spectra were recorded at an indicated collision energy of 15–20 eV, using argon as the collision gas at a pressure of 0.002 mbar. GC–MS spectra were recorded using an HP6890 gas chromatograph with 5973N MSD (Agilent). The gas chromatograph was equipped with a Factorfour VF-5MS column (Varian, 50 m × 0.32 mm ID, film thickness 0.25 ␮m). Helium was used as carrier gas with a flow-rate of 1.0 mL/min. The splitless injected sample volume was

Fig. 2. Examples of non-quaternary oximes synthesized in the past with the aim of improving blood brain barrier penetration.

56

M.C. de Koning et al. / Toxicology Letters 206 (2011) 54–59

Fig. 3. Designed P-site ligand–oxime conjugates with varying spacer length. Compounds 1a–1c showed enhanced affinity and higher reactivation potency compared to 1d (De Koning et al., 2011). Compounds 2 are subject of investigation in this work. 1 ␮L. The temperature program used was 40 ◦ C (1 min), 10 ◦ C/min to 280 ◦ C (10 min). 1 H NMR and 13 C NMR spectra were recorded on a Varian (Palo Alto, CA, USA) Mercuryplus spectrometer operating at 400 MHz and 100 MHz, respectively. Chemical shifts (ı) are given in ppm relative to tetramethyl silane (ı 0 ppm). 2.2. Synthesis procedures 2.2.1. Synthesis of compound 2a Compound 7a (639 mg, 1.4 mmol) was coevaporated with DCE and dissolved in dry DCM (0.7 mL). This solution was put under argon gas and cooled to −10 ◦ C. Next, a solution of TMSCl (0.195 mL, 1.54 mmol) in DCM (0.2 mL) was carefully added, followed by the careful addition of a solution of isoamylnitrite (0.227 mL, 1.68 mmol) in DCM (0.5 mL). After addition of these reagents the ice bath was removed and the solution was stirred for 1.5 h. Then, the solution was concentrated and purified by HPLC. Yield: 98 mg, 0.196 mmol (14%). 1 H NMR (400 MHz, CDCl3 ): ıH 1.85 (2H, m, CH2 ); 1.95 (2H, m, CH2 ); 2.05 (2H, m, CH2 ); 2.8 (2H, m, CH2 ); 3.2 (2H, m, CH2 ); 3.55 (2H, m, CH2 ); 3.8 (2H, m, CH2 ); 3.95 (1H, s, CH); 4.15 (2H, m, CH2 ); 5.4 (1H, s, CH); 6.8 (2H, d, 2 × CHarom ); 7.25 (10H, m, 10 × CHarom ); 7.95 (2H, d, 2 × CHarom ); 8.1 (1H, bs, CH NOH), 9.9 mi, 10.45 ma (1H, bs, NOH). 13 C NMR (100 MHz, CDCl3 ): ı (ppm) 27, 49, 57, 65.5, 66, 67, 69, 81, 114, 127, 128, 128.5, 129, 132, 142, 147, 161, 161.5, 163, 187. [M+H]+ : 503.3 (calc 503.2). HPLC: 5–80% CH3 CN/H2 O + 0.1% TFA Rt = 16.1 min (single major peak). 2.2.2. Synthesis of compound 2b Starting from 7b, a similar procedure was followed as described for 2a. Yield 54%. 1 H NMR (400 MHz, CDCl3 ): ı (ppm) 1.8–2.2 (4H, m, 2 × CH2 ); 3.2 (4H, m, 2 × CH2 ); 3.4 (2H, m, CH2 ); 3.4–2.9 (8H, multiple peaks, 4 × CH2 ); 3.9 (1H, s, CH); 4.15 (2H, m, CH2 ); 5.4 (1H, s, CH); 6.9 (2H, d, 2 × CHarom ); 7.2–7.4 (10H, m, 10 × CHarom ); 8.0 (2H, d, CHarom ); 8.1 (1H, s, CH N); 10.6 (1H, bs, N–OH). 13 C NMR (100 MHz, CDCl3 ): ı (ppm) 27, 39, 57, 65, 65.5, 68, 69, 70.05, 71, 82, 114, 137, 138, 139, 139.5, 142, 148, 161, 162, 163, 187. [M+H]+ : 547.1 (calc 547.3). HPLC: 5–80% CH3 CN/H2 O + 0.1% TFA Rt = 16.0 min (single major peak). 2.2.3. Synthesis of compound 2c Starting from 7c a similar procedure was followed as described for 2a. Yield 35%. 1 H NMR (400 MHz, CDCl3 ): ı (ppm) 1.8–2.1 (4H, m, 2 × CH2 ); 3.15 (4H, m, 2 × CH2 ); 3.25–3.8 (multiple peaks; 14H, 7 × CH2 ), 3.85 (1H, s, CH); 4.1 (2H, m, CH2 ); 5.4 (1H, s, CH); 6.9 (2H, d, CHarom ); 7.1–7.4 (10H, m, 10 × CHarom ); 8.0 (2H, d, 2 × CHarom ); 8.05 (1H, s, CH N); 8.8/10.4 (ma/mi, 1H, N–OH). 13 C NMR (100 MHz, CDCl3 ): ı (ppm) 27, 49, 56, 65, 66, 68, 68.5, 70, 71, 82, 114, 127, 128, 128.5, 129, 132, 142, 148, 161, 161.5, 163, 187. [M+H]+ : 591.4 (calc 591.3). HPLC: 5–80% CH3 CN/H2 O + 0.1% TFA Rt = 15.96 min (single major peak). 2.2.4. (4-Methoxyphenyl)-oxo-ethanaloxime (2d) Compound 2d was prepared as previously described (Mohammed and Nagendrappa, 2003). [M+H]+ : 180.1 (calc.180.16). HPLC: 5–80% CH3 CN/H2 O + 0.1% TFA. Rt = 15.89 min; 1 H NMR (400 MHz, CDCl3 ): ıH 3.8 (3H, s, OMe); 6.9 (2H, d, 2 × CHarom ); 8.0 (1H, s, CH); 8.1 (2H, d, 2 × CHarom ).

2.2.5. Synthesis of compound 4a A suspension of 4-hydroxyacetophenon (3a, 2.0 g, 14.7 mmol), K2 CO3 (3.0 g, 22.0 mmol) and KI (0.24 g, 1.47 mmol) in bis(2-chloroethyl)ether (14 mL, 81.0 mmol) was stirred at 110 ◦ C for 24 h. After cooling the mixture was filtrated and the filtrate was concentrated. The residue was taken up in EtOAc (100 mL) and washed with 10% NaOH (150 mL) and brine (100 mL). After drying the organic layer (MgSO4 ), filtration and evaporation, the product was purified using column chromatography (25–40% EtOAc/hexane). Yield: 2.85 g, 11.76 mmol (80%). HPLC: 5–80% CH3 CN/H2 O + 0.1% TFA Rt = 19.00. [M+H]+ : 243.1 (calc. 243.07). 1 H NMR (400 MHz, CDCl3 ): ı (ppm) 2.4 (3H, s, CH3 ); 3.5 (4H, m, 2 × CH2 ); 3.65 (2H, t, CH2 ); 3.7 (2H, t, CH2 ); 3.75 (2H, t, CH2 ); 4.08 (2H, t, CH2 ); 6.82 (2H, d, 2 × CHarom ); 7.8 (2H, d, 2 × CHarom ). 13 C NMR ı (ppm) 26, 43, 67, 69, 72, 114, 130, 162, 196. 2.2.6. Synthesis of compound 4b A suspension of 4-hydroxyacetophenon (3a, 1.37 g, 10.0 mmol), 1,2-bis(2chloroethoxy)ethane (2.32 mL, 12.4 mmol), K2 CO3 (2.18 g, 15.8 mmol) and KI (0.17 g, 1.0 mmol) in dry n-butanol (20 mL) was stirred for 24 h at 100 ◦ C. After cooling the mixture was filtered and the filtrate was concentrated in vacuo. The residue was dissolved in EtOAc (50 mL) and extracted with 10% NaOH (60 mL) and 10% brine (60 mL). The organic layer was dried (MgSO4 ), filtered and concentrated. The crude product was purified using column chromatography (25–40% EtOAc/hexane). Yield: 1.20 g, 4.18 mmol (42%). HPLC: 5–80% CH3 CN/H2 O + 0.1% TFA Rt = 18.93 min. 1 H NMR (400 MHz, CDCl3 ): ı (ppm) 2.6 (3H, s, Me); 3.62 (2H, t, CH2 ); 3.7 (2H, m, CH2 ); 3.74 (4H, m, 2 × CH2 ); 3.88 (2H, t, CH2 ); 4.2 (2H, t, CH2 ); 7.0 (2H, d, 2 × CHarom ); 8.0 (2H, d, 2 × CHarom ). GC–MS: [M]+ : 286 (calc. 286.1). 2.2.7. Synthesis of compound 4c A suspension of 4-hydroxyacetophenon (3a, 1.35 g, 9.92 mmol), bis[2-(2chloroethoxy) ethyl] ether (2.40 mL, 12.1 mmol), K2 CO3 (2.05 g, 14.8 mmol) and KI (0.23 g, 1.39 mmol) in dry n-butanol (20 mL) was stirred for 24 h at 100 ◦ C. After cooling the mixture was filtered and the filtrate was concentrated in vacuo. The residue was dissolved in EtOAc (50 mL) and extracted with 10% NaOH (60 mL) and 10% brine (60 mL). The organic layer was dried (MgSO4 ), filtered and concentrated. The crude product was purified using column chromatography (25–70% EtOAc/hexane). Yield: 1.56 g, 4.72 mmol (47%). HPLC: 5–80% CH3 CN/H2 O + 0.1% TFA Rt = 18.45 min. 1 H NMR (400 MHz, CDCl3 ): ı (ppm) 2.5 (3H, s, Me); 3.58 (2H, t, CH2 ); 3.64 (6H, m, 3 × CH2 ); 3.7 (4H, m, 2 × CH2 ); 3.84 (2H, t, CH2 ); 4.16 (2H, t, CH2 ); 7.0 (2H, d, 2 × CHarom ); 8.0 (2H, d, 2 × CHarom ). GC–MS: [M]+ : 330 (calc. 330.1). 2.2.8. Synthesis of compound 5a 4a (185 mg, 0.76 mmol) was dissolved in 1.0 mL DCM. TMSCl (91 mg, 0.84 mmol) was added and the mixture was placed under an inert atmosphere. After cooling to −19 ◦ C isoamylnitrite (107 mg, 0.91 mmol) was carefully added. After addition, the mixture was stirred for another hour and concentrated. The crude product was purified using column chromatography (20–40% EtOAc/hexane). Yield: 27 mg, 0.098 mmol (13%). HPLC: 5–80% CH3 CN/H2 O + 0.1% TFA Rt = 11.58 min. [M+H]+ : 272.1 (calc. 272.06). 1 H NMR (400 MHz, CDCl3 ): ı (ppm) 3.66 (2H, t, CH2 ); 3.83 (2H, t, CH2 ); 3.91 (2H, t, CH2 ); 4.2 (2H, t, CH2 ); 7.0 (2H, d, 2 × CHarom ); 8.0 (2H, d, 2 × CHarom ); 9.2 (1H, s, OH). 13 C NMR (100 MHz, CDCl3 ): ı (ppm) 43 (CH2 –Cl); 68 (CH2 –O); 70

M.C. de Koning et al. / Toxicology Letters 206 (2011) 54–59

57

Scheme 1. Synthetic route towards target hybrid compounds 2a–2c and reference compounds 2d and 5a–5c. Reagents and conditions: (a) Cl(CH2 CH2 O)n CH2 CH2 Cl (neat or in n-butanol), K2 CO3 , KI, T, 24 h; (b) trimethylsilylchloride, isoamylnitrite, dichloromethane; (c) 6 (1 equiv.), K2 CO3 (5 equiv.), KI (1 equiv.), dimethylformamide, 90 ◦ C/24 h. (CH2 –O); 72 (CH2 –O); 115 (2 × CH); 131 (C); 133 (2 × CH); 149 (C); 164 (C O); 187 (CHNOH). 2.2.9. Synthesis of compound 5b An identical procedure was followed as described for 5a using 4b (133 mg, 0.46 mmol). Purification: 10–20% EtOAc/hexane Yield: 64 mg, 0.203 mmol (44%). HPLC: 5–80% CH3 CN/H2 O + 0.1% TFA Rt = 11.46 min. [M+H]+ : 316.1 (calc. 316.09). 1 H NMR (400 MHz, CDCl3 ): ı (ppm) 3.62 (2H, t, CH2 ); 3.72 (2H, m, CH2 ); 3.75 (2H, m, CH2 ); 3.90 (2H, t, CH2 ); 4.2 (2H, t, CH2 ); 7.0 (2H, d, 2 × CHarom ); 8.0 (2H, d, 2 × CHarom ); 9.2 (1H, s, OH). 13 C NMR (100 MHz, CDCl3 ): ı (ppm) 43 (CH2 –Cl); 68 (CH2 –O); 70 (CH2 –O); 71 (CH2 –O); 72 (CH2 –O); 115 (2 × CH); 131 (C); 133 (2 × CH); 149 (C); 164 (C O); 187 (CHNOH). 2.2.10. Synthesis of compound 5c An identical procedure was followed as described for 5a using 4c (513 mg, 1.55 mmol). Purification: 50–60% EtOAc/hexane Yield: 141 mg, 0.392 mmol (25%). HPLC: 5–80% CH3 CN/H2 O + 0.1% TFA Rt = 11.24 min. [M+H]+ : 360.2 (calc. 360.11). 1 H NMR (400 MHz, CDCl3 ): ı (ppm) 3.53 (2H, t, CH2 ); 3.60–3.70 (10H, m, 5 × CH2 ); 3.80 (2H, t, CH2 ); 4.1 (2H, t, CH2 ); 7.0 (2H, d, 2 × CHarom ); (2H, d, 2 × CHarom ); 9.2 (1H, s, OH). 13 C NMR (100 MHz, CDCl3 ): ı (ppm) 43 (CH2 –Cl); 68 (CH2 –O); 70 (CH2 –O); 72 (CH2 –O); 73 (CH2 –O); 115 (2 × CH); 131 (C); 133 (2 × CH); 149 (C); 164 (C O); 187 (CHNOH). 2.2.11. Synthesis of compound 7a K2 CO3 (307 mg, 2.23 mmol) and KI (123 mg, 0.74 mmol) were added to a solution of 4a (180 mg, 0.74 mmol) and PSL 6 (198 mg, 0.74 mmol) in DMF (2 mL). The mixture was heated to ca. 90 ◦ C for 24 h. After cooling, CHCl3 (15 mL) was added and the mixture was extracted with brine (2× 30 mL) and water (2× 30 mL). After drying (MgSO4 ), filtration and concentration the crude product was purified using column chromatography (eluent: 5% CH3 OH/CHCl3 ). Yield: 280 mg, 0.59 mmol (80%). HPLC: 5–80% CH3 CN/H2 O + 0.1% TFA Rt = 19.34 min. [M+H]+ : 474.3 (calc.474.3). 1 H NMR (400 MHz, CDCl3 ): ı (ppm) 1.7 (2H, m, CH2 ); 1.9 (2H, m, CH2 ); 2.2 (2H, m, CH2 ); 2.5 (3H, s, CH3 ); 2.6 (2H, t, CH2 ); 2.8 (2H, m, CH2 ); 3.4 (1H, m, CH); 3.7 (2H, t, CH2 ); 3.8 (2H, t, CH2 ); 4.2 (2H, t, CH2 ); 5.5 (1H, s, CH); 6.9 (2H, d, 2 × CHarom ); 7.3 (10H, m, 10 × CH); 7.9 (2H, d, 2 × CHarom ). 2.2.12. Synthesis of compound 7b Identical procedure as described for 7a using 250 mg, 0.94 mmol of 4b. Purification: 5% CH3 OH/CHCl3 . Yield: 380 mg, 0.72 mmol (75%). HPLC: 5–80% CH3 CN/H2 O + 0.1% TFA Rt = 15.58 min. [M+H]+ : 518.3 (calc.518.3). 1 H NMR (400 MHz, CDCl3 ): ı (ppm) 1.8 (2H, m, CH2 ); 1.9 (2H, m, CH2 ); 2.4 (2H, m, CH2 ); 2.5 (3H, s, CH3 ); 2.6 (2H, t, CH2 ); 2.8 (2H, m, CH2 ); 3.5 (1H, m, CH); 3.7 (6H, m, 3 × CH2 ); 3.8 (2H, t, CH2 ); 4.1 (2H, t, CH2 ); 5.5 (1H, s, CH); 7,0 (2H, d, 2 × CHarom ); 7.3 (10H, m, 10 × CH); 7.9 (2H, d, 2 × CHarom ). 2.2.13. Synthesis of compound 7c Identical procedure as described for 7a using 340 mg, 1.03 mmol of 4c. Purification: 5% CH3 OH/CHCl3 . Yield: 503 mg, 0.90 mmol (8%). HPLC: 5–80%

CH3 CN/H2 O + 0.1% TFA Rt = 17.14 min. [M+H]+ : 562.3 (calc.562.3). 1 H NMR (400 MHz, CDCl3 ): ı (ppm) 1.8 (2H, m, CH2 ); 1.9 (2H, m, CH2 ); 2.4 (2H, m, CH2 ); 2.5 (3H, s, CH3 ); 2.6 (2H, t, CH2 ); 2.8 (2H, m, CH2 ); 3.5 (1H, m, CH); 3.55–3.75 (10H, m, 5 × CH2 ); 3.85 (2H, t, CH2 ); 4.1 (2H, t, CH2 ); 5.5 (1H, s, CH); 7.0 (2H, d, 2 × CHarom ); 7.3 (10H, m, 10 × CH); 7.9 (2H, d, 2 × CHarom ). 2.3. Biological evaluation 2.3.1. Sample preparation for biological tests As some of the isolated test compounds (from HPLC) tend to contain some water, or exist as the TFA salt (2a–2c), the weighing of the amounts necessary for biological testing could be quite inaccurate. As an alternative, UV measurements (absorption max at 308 nm) were used to calibrate all the solutions. The reference solution (1 × 10−2 M) was accurately prepared from 4-methoxyisonitrosoacetophenon (2d). Thus, dilution of the stock solution of 2d to 1.0 × 10−5 and subsequent UV measurement gave an absorption of 0.143 at 308 nm (ε = 14300 L mol−1 cm−1 ). Next, solutions of approx. 2 × 10−2 M were prepared by weighing a sufficient amount (typically 5 mg) of the test substances and dissolve them in water. Aliquots of these stock solutions were diluted 1000× and UV absorption of the resulting dilutions was compared to that of the reference 2d solution. Next, the test substance stock solutions were diluted correcting for the difference in UV absorption and, after preparing a new 10−5 M dilution re-checked with UV. This method assumes an equal extinction coefficient for all of the test substances. This is justified, because the extinction coefficient is not expected to change dramatically on variation of the linker length. The UV-measurement of the PSL (4-O-(benzhydryl)-hydroxypiperidine) showed no absorption in the 250–350 nm region at similar concentrations, and thus did not disturb the measurements involving the complete constructs 2a–2c (representative UV spectra are given in the supporting information). 2.3.2. Biological tests The biological tests performed in this study are quite standard and described in detail in the supporting information.

3. Results and discussion 3.1. Synthesis The synthesis of compounds 2a–2c is outlined in Scheme 1 and commenced with the alkylation of 4-hydroxyacetophenon with bischloroethylethers of varying lengths giving chlorides 4a–4c. Conversion of the acetyl groups in 4a–4c into ␣-ketoaldoxime groups (→5a–5c) was effected by reaction with trimethylsilylchloride and isoamylnitrite in dichloromethane in low to moderate yields (Mohammed and Nagendrappa, 2003). Unfortunately, alkylation of PSL 6 (De Koning et al., 2011; Kwon et al., 2007) with

58

M.C. de Koning et al. / Toxicology Letters 206 (2011) 54–59

Fig. 4. Inhibition of human AChE by conjugates 2a–2c and oximes 5a–5c at varying concentrations.

chlorides 5a–5c led to the formation of inseparable mixtures of products. Nevertheless, compounds 5a–5c still served as useful control compounds in the biological evaluation (vide infra). Following a slightly altered synthetic route the conversion of 4a–4c into 7a–7c proceeded smoothly and the products were isolated in high yields after column chromatography. The final reaction of these compounds into the target conjugates 2a–2c could be accomplished using the isoamylnitrite/trimethylsilylchloride method mentioned above. The target conjugates 2a–2c were purified using HPLC and isolated in moderate yields. Finally, the reference oxime 2d was obtained in 50% yield from commercially available 4methoxyacetophenon (3b). All products and intermediates were analyzed to confirm identity and homogeneity using common techniques such as 1 H and 13 C NMR, HPLC, UV spectroscopy and mass spectrometry. 3.2. Biological evaluation Having all requisite compounds in hand, attention was turned to the biological evaluation. An inherent danger of covalently connecting two ligands with affinity for different substrate binding sites of the same enzyme is that, in potential, a very strong (reversible) dual site inhibitor can be obtained, which would be very toxic (Lewis et al., 2002; Manetsch et al., 2004). Thus, the relative binding affinity of conjugates 2a–2c was first evaluated using purified, uninhibited hAChE by measuring their enzyme inhibiting potency. These experiments were carried out by subjecting hAChE (3 nM) to a series of 10-fold dilutions (10−3 M to 10−6 M) of 2a–2c in phosphate buffer pH 7.4. The enzyme activity after 30 min of incubation was measured using the method of Ellman et al. (1961). The results of these experiments are displayed in Fig. 4. The conjugates 2a–2c showed no significant enzyme inhibition at pharmacologically interesting concentrations (10−5 M or less), but at 100 ␮M and 1 mM concentrations enzyme inhibition was about 20% and around 70%, respectively. For comparison, enzyme inhibitory potency of the compounds lacking the PSL (5a–5c) is also displayed. These compounds showed no significant inhibition at all concentrations, thus confirming the contribution to affinity by the PSL 6 in conjugates 2a–2c. It was further concluded that compounds 2a–2c were not strong inhibitors of hAChE and that they should not be discarded for that reason. Next, the reactivating potency of 2a–2c towards sarin-inhibited hAChE was investigated. Briefly, these experiments were carried out by treating hAChE (3 nM) with sufficient amounts of the nerve agent to inhibit at least 85% of the enzyme. Next, the inhibited

Fig. 5. Reactivation of sarin, VX and tabun-inhibited human AChE by conjugates 2a–2c, oximes 5a–5c and reference oxime 2d.

enzyme was subjected to a series of 10-fold dilutions of oximes 2a–2c and incubated for 30 min. The amount of active enzyme present after that period of time was measured using the method of Ellman et al. (1961). Reactivation is given relative to the control (no inhibited enzyme ∼ 100% activity) and the values were corrected for spontaneous reactivation and residual activity after inhibition (measured from inhibited enzyme without oxime). The results of the reactivation experiments carried out with 2a–2c at 10−4 M concentration are depicted in Fig. 5. In addition, the results of similar experiments with fragments of the conjugate (i.e. the parent oxime 2d), as well as the conjugates lacking the PSL (5a–5c) are also displayed. Comparison of these results clearly shows that conjugates 2a–2c possess a remarkably higher reactivation potency (e.g. 36% for 2b) than the parent oximes 2d and 5a–5c (i.e. only about 5% reactivation of sarin-inhibited hAChE). The observations that the reactivation potencies of 2d and 5a–5c are virtually the same and that enhanced reactivation potency is only observed in the compounds containing the PSL moiety 6 further confirm that conjugation of the PSL is essential to obtain molecules with enhanced reactivation potency. Conjugates 2a–2c were also tested for their reactivation potency towards VX- and tabun inhibited hAChE (Fig. 5).1 In the case of VX similar results were obtained (approx. 37% reactivation at 10−4 M) while tabun-inhibited hAChE, which is notoriously difficult to reactivate (Ashani et al., 1995), was not reactivated. No clear influence of the linker length on the reactivation potency has been observed thus far for these molecules. Conjugate 2b seems to perform best with sarin-inhibited hAChE, suggesting an optimal spacer length of n = 2 in compounds 2a–2c. However, with VX-inhibited hAChE no apparent differences in reactivation potency between the compounds 2a–2c were measured. Therefore, the influence of the linker length on the reactivation potency still seems unclear and may be explained by a different binding mode than originally anticipated or by binding of the PSL-moiety to different binding sites within the P-Site area. 4. Conclusion Concluding, it was demonstrated that the covalent connection of a PSL to an otherwise hardly reactivating compound (i.e. 2d) leads to

1 Some additional reactivation experiments using cyclosarin and paraoxon-etyl (PXE) as well as sarin, VX and tabun were carried out by dr. F. Worek (IPT, Germany). The results of these experiments are given in the supporting information (Fig. S2).

M.C. de Koning et al. / Toxicology Letters 206 (2011) 54–59

a dramatic improvement in reactivation potency. The higher affinity of the conjugates for the free enzyme, as well as the comparison of the reactivation potency of the compounds lacking the PSL support the hypothesis that the enhanced reactivation potency stems from the extra contribution in affinity by the PSL. Although the compounds presented here are still inefficient reactivators when compared to the current standard the presented approach holds promise for the future design of non-quaternary reactivators (e.g. based on the ones depicted in Fig. 2) with potentially improved BBB penetration. It is also not excluded that a broader range of reactivating moieties (e.g. hydroxamic acids) may be employed which would widen the scope for the search for reactivators with broadspectrum activity. Conflicts of interest statement The authors declare that there are no conflicts of interest. Acknowledgments The authors wish to acknowledge the funding from the Dutch Ministry of Defence, the generous gift of purified human acetylcholinesterase by dr. Florian Nachon from IRBA/CRSSA (France) and COL dr. F. Worek from IPT (Germany) for performing additional reactivation experiments using erythrocyte acetylcholinesterase inhibited with VX, tabun, sarin, paraoxon-ethyl and cyclosarin. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.toxlet.2011.04.004. References Ashani, Y., Radic, Z., Tsigelny, I., Vellom, D.C., Pickering, N.A., Quinn, D.M., Doctor, B.P., Taylor, P., 1995. Amino acid residues controlling reactivation of organophosphonyl conjugates of acetylcholinesterase by mono- and bisquaternary oximes. J. Biol. Chem. 270, 6370–6380. Bajgar, J., Fusek, J., Kuca, K., Bartosova, L., Jun, D., 2007. Treatment of organophosphate intoxication using cholinesterase reactivators: facts and fiction. Mini Rev. Med. Chem. 7, 461. Bedford, C.D., Howd, R.A., Dailey, O.D., Miller, A., Nolen III, H.W., Kenley, R.A., Kern, J.R., Winterle, J.S., 1986a. Nonquaternary cholinesterase reactivators. 3. 3(5)-Substituted 1,2,4-oxadiazol-5(3)-aldoximes and 1,2,4-oxadiazole-5(3)thiocarbohydroximates as reactivators of organophosphonate-inhibited eel and human acetylcholinesterase in vitro. J. Med. Chem. 29, 2174–2183. Bedford, C.D., Miura, M., Bottaro, J.C., Howd, R.A., Nolen III, H.W., 1986b. Nonquaternary cholinesterase reactivators. 4. Dialkylaminoalkyl thioesters of ␣-keto thiohydroximic acids as reactivators of ethyl methylphosphonyl- and 1,2,2trimethylpropyl methylphosphonyl-acetylcholinesterase in vitro. J. Med. Chem. 29, 1689–1696. Benschop, H.P., van den Berg, G.R., Van Hooidonk, C., De Jong, L.P.A., Kientz, C.E., 1979. Antidotes to organophosphate poisoning. 2. Thiadiazole-5-carboxaldoximes. J. Med. Chem. 22, 1306–1313. Benschop, H.P., van Oosten, A.M., Platenburg, D.H.J.M., van Hooidonk, C., 1970. Isothiazolecarboxaldoximes and methylated derivatives as therapeutic agents in poisoning by organophosphorus compounds. Med. Chem. 13, 1208–1212. Beznosko, B.K., Voronov, I.B., Feigman, E.E., Krivenchuk, V.E., 1977. On the central action of diethyxime A new cholinesterase reactivator. Russ. Pharm. Toxicol. 40, 154–159. De Koning, M.C., Joosen, M.J.A., Noort, D., van Zuylen, A., Tromp, M.C., 2011. Peripheral site ligand-oxime conjugates: a novel concept towards reactivation of nerve agent-inhibited human acetylcholinesterase. Bioorg. Med. Chem. 19, 588–594.

59

Degorre, F., Kiffer, D., Terrier, F., 1988. Sulfur derivatives of 2-oxopropanal oxime as reactivators of organophosphate-inhibited acetylcholinesterase in vitro: synthesis and structure-reactivity relationships. J. Med. Chem. 31, 757–763. Ellman, G.L., Courtney, K.D., Anders Jr., V., Feather-Stone, R.M., 1961. A new and rapid colorimetric determination of acetylcholinesterase activity. Biochem. Pharmacol. 7, 88–95. Jeong, H.-C., Kang, N.S., Park, N.-J., Yum, E.K., Jung, Y.-S., 2009a. Reactivation potency of fluorinated pyridinium oximes for acetylcholinesterases inhibited by paraoxon organophosphorus agent. Bioorg. Med. Chem. Lett. 19, 1214–1217. Jeong, H.C., Park, N.-J., Chae, C.H., Musilek, K., Kassa, J., Kuca, K., Jung, Y.-S., 2009b. Fluorinated pyridinium oximes as potential reactivators for acetylcholinesterases inhibited by paraoxon organophosphorus agent. Bioorg. Med. Chem. 17, 6213–6217. Kenley, R.A., Bedford, C.D., Dailey, O.D., Howd, R.A., Miller, A., 1984. Nonquaternary cholinesterase reactivators. 2. ␣-Heteroaromatic aldoximes and thiohydroximates as reactivators of ethyl methylphosphonyl-acetylcholinesterase in vitro. J. Med. Chem. 27, 1201–1211. Kenley, R.A., Bedford, C.D., Howd, R.A., Jackson, S.E., 1985. Reactivation of ethyl methylphosphonylated eel acetylcholinesterase in vitro by 2-PAM H16, and a series of nonquaternary ␣-ketothiohydroximates. Biochem. Pharmacol. 34, 3606–3608. Kenley, R.A., Howd, R.A., Mosher, C.W., Winterle, J.S., 1981. Nonquaternary cholinesterase reactivators dialkylaminoalkyl thioesters of ␣-ketothiohydroximic acids as reactivators of diisopropyl phosphorofluoridate inhibited acetylcholinesterase. J. Med. Chem. 24, 1124–1133. Kuca, K., Jun, D., Musilek, K., 2006. Structural requirements of acetylcholinesterase reactivators. Mini Rev. Med. Chem. 6, 269–277. Kwon, Y.E., Park, J.Y., No, K.T., Shin, J.H., Lee, S.K., Eun, J.S., Yang, J.H., Shin, T.Y., Kim, D.K., Chae, B.S., Leem, J.-Y., Kim, K.H., 2007. Synthesis, in vitro assay, and molecular modeling of new piperidine derivatives having dual inhibitory potency against acetylcholinesterase and A␤1–42 aggregation for Alzheimer’s disease therapeutics. Bioorg. Med. Chem. 15, 6596–6607. ´ Z., Carlier, P.R., Taylor, P., Finn, M.G., Lewis, W.G., Green, L.G., Grynszpan, F., Radic, Sharpless, K.B., 2002. Click chemistry in situ: acetylcholinesterase as a reaction vessel for the selective assembly of a femtomolar inhibitor from an array of building blocks. Angew. Chem. 41, 1053–1057. Lundy, P.M., Raveh, L., Amitai, G., 2006. Development of the bisquaternary oxime HI-6 toward clinical use in the treatment of organophosphate nerve agent poisoning. Toxicol. Rev. 25, 231. Luo, C., Tong, M., Chilukuri, N., Brecht, K., Maxwell, D.M., Saxena, A., 2007. Biochemistry 46, 11771. ´ Z., Raushel, J., Taylor, P., Sharpless, K.B., Kolb, Manetsch, R., Kransinski, A., Radic, H.C., 2004. In situ click chemistry: enzyme inhibitors made to their own specifications. J. Am. Chem. Soc. 126, 12809–12818. Marrs, T.C., 1993. Organophosphate poisoning. Pharmacol. Ther. 58, 51. Masson, P., Legrand, P., Bartels, C.F., Froment, M.-T., Schopger, L.M., Lockridge, O., 1997. Role of aspartate 70 and tryptophan 82 in binding of succinyldithiocholine to human butyrylcholinesterase. Biochemistry 36, 2266–2277. Mohammed, A.H.A., Nagendrappa, G., 2003. A remarkably simple ␣-oximation of ketones to 1,2-dione monooximes using the chlorotrimethylsilane–isoamyl nitrite combination. Tetrahedron Lett. 44, 2753–2755. Ohta, H., Ohmori, T., Suzuki, S., Ikegaya, H., Sakurada, K., Takatori, T., 2006. New safe method for preparation of sarin-exposed human erythrocytes acetylcholinesterase using non-toxic and stable sarin analogue isopropyl pnitrophenyl methylphosphonate and its application to evaluation of nerve agent antidotes. Pharm. Res. 23, 2827–2833. Okuno, S., Sakurada, K., Ohta, H., Ikegaya, H., Kazui, Y., Akutsu, T., Takatori, T., Iwadate, K., 2008. Blood–brain barrier penetration of novel pyridinealdoxime methiodide (PAM)-type oximes examined by brain microdialysis with LC–MS/MS. Toxicol. Appl. Pharm. 227, 8–15. Quinn, D.M., 1987. Acetylcholinesterase: enzyme structure, reaction dynamics, and virtual transition states. Chem. Rev. 87, 955. Sakurada, K., Matsubara, K., Shimizu, K., Shiono, H., Seto, Y., Tsuge, K., Yoshino, M., Sakai, I., Mukoyama, H., Takatori, T., 2003. Pralidoxime iodide (2-PAM) penetrates across the blood–brain barrier. Neurochem. Res. 28, 1401–1407. Sussman, J.L., Harel, M., Frolow, F., Oefner, C., Goldman, A., Toker, L., Silman, I., 1991. Atomic structure of acetylcholinesterase from Torpedo californica: a prototypic acetylcholine-binding protein. Science 253, 872–879. Szegletes, T., Mallender, W.D., Thomas, P.J., Rosenberry, T.L., 1999. Substrate binding to the peripheral site of acetylcholinesterase initiates enzymatic catalysis. Substrate inhibition arises as a secondary effect. Biochemistry 38, 122–133. Worek, F., Eyer, P., Aurbek, N., Szinicz, L., Thiermann, H., 2007. Recent advances in evaluation of oxime efficacy in nerve agent poisoning by in vitro analysis. Toxicol. Appl. Pharmacol. 219, 226–234.