p-(4-Azipentyl)-propofol: A Potent Photoreactive General Anesthetic Derivative of Propofol

Abstract

We synthesized 2,6-Diisopropyl-4-[3-(3-methyl-3H-diazirin-3-yl)-propyl]-phenol (p-(4-azipentyl)-propofol), or p-4-AziC5-Pro, a novel photoactivable derivative of the general anesthetic propofol. p-4-AziC5-Pro has an anesthetic potency similar to propofol. Like propofol, the compound potentiates inhibitory GABA_A receptor current responses and allosterically modulates binding to both agonist and benzodiazepine sites, assayed on heterologously expressed GABA_A receptors. p-4-AziC5-Pro inhibits excitatory current responses of nACh receptors expressed in Xenopus oocytes and photoincorporates into native nACh receptor-enriched Torpedo membranes. Thus p-4-AziC5-Pro is a functional general anesthetic that both modulates and photoincorporates into Cys-loop ligand-gated ion channels, making it an excellent candidate for use in identifying propofol binding sites.

Introduction

Propofol is the most widely used intravenous general anesthetic for induction and maintenance of anesthesia.¹ Its advantages include both rapid-onset and rapid-recovery from anesthesia accompanied with relatively few of the common side effects of general anesthetics, such as post-operative nausea and vomiting. Many lines of evidence suggest that one of the main pharmacological actions for general anesthetics is the potentiation of the type A γ-amino butyric acid (GABA_A) receptor’s inhibitory response.^2–7 GABA_A receptors are a member of the Cys-loop super-family of pentameric ligand gated ion channels that also includes glycine, serotonin (5-HT₃), and nicotinic acetylcholine (nACh) receptors. In addition to intravenous and volatile anesthetics, GABA_A receptors are modulated by a number of other classes of drugs, including: neurosteroids, benzodiazepines, and ethanol.^{8, 9}

Identifying the binding sites of these different classes of drugs on the GABA_A receptor is of major interest for understanding their mechanisms of action and for further drug development. Although mutagenesis and genetic experiments have been successful for identifying residues involved in the action of general anesthetics, the use of affinity labels is more likely to identify residues involved in a binding pocket. Experiments using a photoreactive analog of etomidate ([³H]azietomidate) identified two residues (αMet236 and βMet286) as contributing to an intersubunit etomidate binding site, only one of which had been implicated by earlier mutagenesis experiments.¹⁰ Both propofol and etomidate can potentiate the GABA_A receptor’s current response and can directly activate the receptor, activities which correlate with the drugs’ in vivo anesthetic potency. The similarity in these drugs’ mechanisms of action, lead to the question of whether or not they act at the same site on the GABA_A receptor. A series of [³H]azietomidate photolabeling assays on bovine brain GABA_A receptors supports the idea that etomidate and propofol may act through different sites.¹¹ Recent crystallographic studies on a bacterial homolog of the pentameric receptors from Gloebacter violaceus (GLIC) have identified an intrasubunit binding site for propofol and volatile anesthetics, in contrast to the intersubunit binding site identified for etomidate in the GABA_A receptor.^{10, 12}

The development of propofol-based photoaffinity reagents is critical for identifying residues that interact with this important drug in mammalian receptors. One such reagent has been synthesized, 2-isopropyl-5-[3-(trifluoromethyl)-3H-diazirin-3-yl]phenol (m-TFD-propofol), a propofol analog with an alkyl-diazirinyl group placed at the meta-position of the phenyl ring.¹³ To further investigate the nature of propofol’s binding site, we have now synthesized 2,6-diisopropyl-4-[3-(3-methyl-3H-diazirin-3-yl)-propyl]-phenol (p-(4-azipentyl)-propofol), or p-4-AziC5-Pro ( 6 ), an analog that keeps the core propofol structure intact by adding the reactive diazirine group at the para-position of the phenol ring (Scheme 2). Here, we describe its synthesis and initial pharmacologic characterization. It functions as a general anesthetic and interacts with the GABA_A and nACh receptor ligand-gated ion channels. Importantly, the p-4-AziC5-Pro-derivative differs from the previously described photoreactive propofol (m-TFD-propofol) in the location of the reactive group, opening the possibility of exploring the binding site in more detail.

Scheme 2.

Results

Synthesis of Diazirine Analogs of Propofol

A large number of photoaffinity reagents based on 1-aryl-1-trifluoromethyldiazirines have been described in the literature.¹⁴ Based on the existing QSAR of propofol’s anesthetic activity, the introduction of the trifluoromethyl diazirine moiety on the aromatic ring of propofol could be best achieved at the 4-position, since substitutions at this position do not significantly alter the anesthetic properties of propofol.¹⁵ An alternative strategy involves the removal of one of the iso-propyl groups and modification of the 3-position.¹³

As our first synthetic target we selected 4-trifluoromethyldiazirynyl propofol (1, 4-TFD-Pf) (Scheme 1). The intermediate 5 was synthesized in a sequence of Friedel-Crafts acylation, phenol protection with dimethyl-tert-butylsilyl group, and assembly of the diazirine group by a routine procedure. However, the final deprotection of the diazirine 5 failed to produce a stable product. We have tried without success a variety of deprotecting conditions, including TBAF, TBAF-acetic acid, HF, pyridine-HF, TsOH and HCl. The rapid decomposition of the target compound 1 is most likely due to the electronic effect of the hydroxy group in the para-position stabilizing the benzylic carbene. Indeed, syntheses of 4-hydroxyphenyl diazirine have been previously reported, but no detailed synthetic procedures were provided.¹⁶ In contrast, the isomeric meta-hydroxyphenyl diazirines are stable.¹⁷

Scheme 1.

To avoid this source of instability we redesigned the photoactivatable analog of propofol to separate the diazirine moiety from the phenyl ring. In the new analog 6 (Scheme 2), the photoaffinity label is linked to the propofol core via a short non-polar spacer. The synthetic expediency and the necessity to maintain a nonpolar environment at the bottom of the ring suggested a three carbon linker.

The synthesis of the diazirine 6 shown in Scheme 2 started with an acylation of propofol with succinic anhydride to give the key intermediate 7. The next steps involved esterification of the carboxyl group and an unusual direct reduction of the ketoester 8 into the acid 11. Interestingly, such a reduction has not been previously reported, and the model NaBH₄ reduction of the keto group in 4-acetylpropofol stops at the stage of a corresponding alcohol. We hypothesize that the initially formed alcohol 9 is first converted into the lactone 10, and that the activation of the hydroxyl group by an ester enables further reduction of the C-O bond. Thus, the ketone group in 8 is converted into a methylene group in 11 via a one-step procedure and under very mild conditions. The carboxylic acid 11 was next converted into the corresponding ketone 12 using a Weinreb procedure¹⁸, and the subsequent diazirination by the modified method of Church and Weiss¹⁹ followed by deprotection of the phenol afforded the target diazirine 6.

For the purpose of radiolabeling of 6 with tritium, an alternative synthesis was developed (Scheme 3), in which the introduction of the label was achieved by the reduction of the ketone (e.g. step iii in Scheme 2), performed at a later stage of synthesis. Thus, the keto group in the ester 8 was protected as 1,3-dioxolane to give the ketal 14. The subsequent two carbon elongation to the ketone 15 and assembly of the diazirine 16 was performed analogously as described in Scheme 2. The keto group was deprotected with TFA and the phenol protected with TBDMS to form the diazirine 18. The protection of phenol with the TBDMS group was necessary to minimize decomposition of the reducing agent (NaBH₄ or NaBT₄) by the acidic phenolic OH group during the radiolabeling reduction step. The reduction of the ketone 18 with sodium borohydride followed by desilylation of the phenol afforded the alcohol 19, and the subsequent further reduction with triethylsilane and desilylation produced the final product 6. The tritiation at the level of 12 Ci/mmol was achieved by substituting NaBT₄ for NaBH₄ in step viii-a.

Scheme 3.

Pharmacological Properties

Solubility and Partition Coefficient

The saturated aqueous solubility of 6 in 0.01M Tris/HCl, pH 7.4 was 73 ± 4 μM (all errors herein are standard deviations unless otherwise stated). At a concentration close to saturation, the compound was stable at 4°C in the aforementioned buffer, for at least 2.5 days and eluted as a single fraction at 81.5% acetonitrile when analyzed using reverse phase HPLC. The octanol/water partition coefficient was estimated to be 15,600 ± 1200. For comparison, the saturated concentration of pure propofol in water is 1 ± 0.02 mM, and the octanol/water partition coefficient is 4,300 ± 280.²⁰

Anesthetic Activity

The anesthetic potency of 6 was assayed in tadpoles. The concentration-dependence of loss of righting reflexes (LoRR) was examined in groups of five animals at seven concentrations between 1.0 and 20 μM (ten animals per concentration). Anesthesia was reversible, the animals recovering in fresh solution overnight. The one death was not at the highest concentration, so was likely unrelated to the agent. Only animals that fully recovered in fresh water overnight were included in the analysis. The EC50 determined with 69 animals was 3.2 ± 0.55 μM and the slope was 1.7 ± 0.4. Thus, 6 is a general anesthetic with potency comparable to that of propofol, which has an EC50 of 2.2 ± 0.2 μM.²⁰

Enhancement of GABA_A Receptor Channel Function

Human α1β2γ2L GABA_A receptor currents (at EC₅) expressed in Xenopus oocytes were potentiated with increasing amounts of 6 (Figure 1A) with an EC₅₀ of 4.5 ± 2.7 μM, compared to 5.2 ± 2.0 μM for propofol (Figure 1B). At maximal concentrations of 6 (100 μM), current responses were enhanced ~20-fold (Figure 1A), similar to ~15-fold seen with 30 – 100 μM propofol (Figure 1B).

Figure 1.

GABA_A receptor current enhancement. (A, B) Two electrode voltage clamp was used to measure currents from oocytes injected with human α1β2γ2L GABA_A receptor subunits. Currents were elicited with 3 μM GABA (~EC₅) in the presence of increasing amounts of 6 (A) or propofol (B). Solid lines represent the nonlinear least-squares fit of the data. For 6: EC₅₀ = 4.5 ± 2.7 μM, nH = 1.4 ± 1.3, Imax = 22 ± 6% (4 oocytes). For propofol: EC₅₀ = 5.2 ± 2.1 μM, nH = 1.7 ± 1.2, Imax = 16 ± 3% (2 oocytes). Insets show current traces elicited by 3 μM GABA (black bars) plus increasing amounts of drug (1, 3, 10, 30, 100 μM 6 and 1, 3, 10, 30, 100, 300 μM propofol) (gray bars). (C, D) GABA concentration response curves shift to the left in the presence of drugs. For plotting, currents were normalized to I_GABA,max and [GABA] was normalized to the control GABA EC₅₀ for each experiment to minimize the effect of the spread of EC₅₀’s of the different oocytes.³⁸ The data was fit as described above and in the Experimental Section, for 6 (C): EC₅₀ = 0.99 ± 0.11, nH = 1.0 ± 0.1 for GABA alone (solid symbols) and EC₅₀ =0.12 ± 0.03, nH = 0.9 ± 0.2 in the presence of 6 μM 6 (open symbols) (3 oocytes). For propofol (D): EC₅₀ = 1.35 ± 0.2, nH = 0.9 ± 0.1 for GABA alone (solid symbols) and EC₅₀ = 0.12 ± 0.04, nH = 1.1 ± 0.3 in the presence of 4.4 μM propofol (open symbols) (4 oocytes).

GABA concentration response curves (+/− drugs) were determined at a fixed concentration of anesthetic equal to twice the EC₅₀ for tadpole anesthesia, a concentration generally considered to represent clinical anesthesia, and were shifted to the left ~9-fold by 6 μM 6 (Figure 1C) and ~11-fold by 4.4 μM propofol (Figure 1D).

Like propofol, 6 directly activated GABA_A receptor currents in the absence of GABA (data not shown). The direct activation data were also fit with the nonlinear least-squares regression to the Hill equation and 6 produced currents of 10 ± 3 % of the GABA response with an EC₅₀ of 15 ± 3 μM (3 oocytes), whereas propofol activated the receptor 45 ± 10% of the maximal GABA response with an of EC₅₀ of 50 ± 5 μM (2 oocytes).

Modulation of Ligand-binding to GABA_A Receptors

To assess allosteric action, the effects of 6 were compared to those of propofol for modulating the binding of 1.5 nM [³H]flunitrazepam (benzodiazepine site) or 3 nM [³H]muscimol (agonist site) to α1β3γ2 GABA_ARs (Figure 2).

Figure 2.

[³H]flunitrazepam and [³H]muscimol binding to GABA_A receptors. The specific [³H]flunitrazepam (1.5 nM) and [³H]muscimol (3 nM) binding to membrane suspensions of α1β3γ2 GABA_A receptors heterogeneously expressed in HEK293 cells was evaluated at various concentrations of 6or propofol. The percent enhancement of binding of three measurements are plotted as mean ± standard deviations.

Specific binding of the benzodiazepine, [³H]flunitrazepam, was enhanced by both agents between 0.1 and 40 μM (Figure 2A). Above this concentration, propofol caused a rapid decrease in binding up to 400 μM whereas 6 caused no further change. Compound 6 was less efficacious, causing a ~150% enhancement compared to ~200% for propofol. The concentration dependence and efficacy for enhancement of binding of the agonist [³H]muscimol by both agents was very similar to that exhibited by [³H]flunitrazepam, although propofol’s efficacy was even higher than 6’s and its effect reached a maximum at somewhat lower concentrations (Figure 2B). Again propofol caused a sharp decrease in binding above 30 μM, but 6 did not.

Because the action of propofol was biphasic we did not attempt to fit the data to a function. In the case of 6, the enhancement is quite small when compared to the errors, but the data could be fit to a mass action curve with half effect concentrations of ~ 0.5 μM.

Inhibition of Torpedo nACh Receptors

In oocytes expressing Torpedo nACh receptors, currents elicited by 10 μM ACh (~EC₂₀) were inhibited by increasing amounts of 6 with an IC₅₀ = 4.0 ± 1.2 μM compared to an IC₅₀ of 7.3 ± 1.4 μM for propofol (Figure 3). The difference in the shapes of the current traces (insets in Figure 3), with increasing amounts of compound, suggest that 6 equilibrates more slowly than propofol or possibly stabilizes the desensitized state. This hypothesis will be addressed in the future with rapid perfusion patch clamp kinetic studies.

Figure 3.

Torpedo nACh receptor current inhibition. Oocytes expressing wild type Torpedo nACh receptors were tested with 10 μM ACh (~EC₂₀) and then with 10 μM ACh plus increasing amounts of propofol or the propofol derivative. The current traces in the top right inset show one oocyte’s current response to 10 μM ACh plus increasing amounts of propofol (0.1, 0.3, 1, 3, 10, 30, 100, 300 μM). The effect of 6 on 10 μM ACh currents is shown in the bottom left inset (1, 3, 10, 30, 100 μM). Nonlinear least-squares analysis of the curves yielded: for propofol (open symbols), IC₅₀ = 7.3 ± 1.4 μM, nH = 0.8 ± 0.1 (3 oocytes); for 6 (solid symbols), IC₅₀ = 4.0 ± 1.2, nH = 0.9 ± 0.2 (3 oocytes). Currents were normalized to the 10 μM ACh response.

Photoincorporation into Torpedo nACh Receptor-enriched Membranes

Photoincorporation was assessed by SDS-PAGE followed by fluorography (Figure 4A) and liquid scintillation counting of excised gel bands (Figure 4B). For nACh receptor-rich membranes photolabeled in the absence of other drugs, there was photoincorporation in each nACh receptor subunit, with the α–subunit labeled most prominently, along with photolabeling of voltage dependent anion channel (VDAC) and the Na⁺/K⁺-ATPase α–subunit. In the absence of agonist, drugs that bind with high affinity with the nACh receptor ion channel either in the resting closed channel state (tetracaine) or in the desensitized state (phencyclidine) had little effect on nACh receptor subunit photolabeling. However, propofol alone increased α subunit photolabeling by ~70%. For nACh receptors desensitized in the presence of agonist (carbamylcholine), α–subunit photolabeling was increased by 100% compared to the resting state. Small additional effects on [³H]6 photoincorporation were observed when propofol (~20% increase) or phencyclidine (~20% decrease) were added to the carbamylcholine-treated samples. For the other nACh receptor subunits, the presence of agonist, propofol, or channel blockers modulated photolabeling by <20%.

Figure 4.

Torpedo nACh receptor photoincorporation. nACh receptor-rich membranes were photolabeled with 1.7 μM [³H]6 as described in the Experimental Section, and the polypeptides were then resolved by SDS-PAGE. (A) Polypeptides were visualized by Coomassie Blue stain (lane 1, a representative lane), and ³H distribution was determined by fluorography (lanes 2–9, 14 day exposure) for aliquots photolabeled in the absence of additional drugs (lanes 2 & 9), in the presence of 100 μM tetracaine (lane 3), 100 μM phencyclidine (lanes 4 & 7), 300 μM propofol (lanes 5 & 6), and in the presence of 1 mM carbamylcholine (Carb, lanes 6–8). Indicated on the left are the stained bands corresponding to the nACh receptor subunits (α, β, γ, & δ), rapsyn (Rsn), the α subunit of the Na⁺/K⁺-ATPase, calelectrin (37K), and the voltage dependent anion channel (VDAC).³⁹ (B) Quantification by liquid scintillation counting of the ³H incorporation into the gel bands containing the nACh receptor subunits.

Discussion

Synthesis

We have synthesized 6, a novel, photoreactive anesthetic with potency similar to propofol. Initial attempts to synthesize a propofol analog with the photoreactive group at the para position, 4-TFD-propofol (Scheme 1), were unsuccessful. Therefore, we designed another strategy for obtaining a propofol derivative. To avoid the instability caused by the electronic effect of the hydroxy group in the para-position stabilizing the benzylic carbene, we redesigned the photoactivatable analog of propofol by separating the diazirine moiety from the phenyl ring by linking to the propofol core via a short non-polar, three carbon spacer. The reactive aliphatic diazirine forms a carbonium ion that reacts preferentially with acidic side chains and with nucleophilic residues (tyrosine and methionine), but not with aliphatic side chains. Successful synthesis of another propofol derivative, m-TFD-propofol, with the photoreactive group linked directly to the benzene group at the meta position has been reported, suggesting that it is possible to synthesize a number of photoreactive propofol analogs that vary in the location, size, and type of the reactive group yet maintain their anesthetic properties.¹³ Such an array of compounds will allow for a set of tools to aid in identifying and characterizing the residues involved in propofol’s binding sites.

General Anesthetic Properties

Compound 6 acts as a reversible general anesthetic in tadpoles and was approximately equipotent to the parent general anesthetic, propofol. Propofol was initially discovered during a screen of 97 alkylphenols in mice and rabbits showing that 2,6-dialkylphenols, particularly 2,6-di-sec-alkyl phenols, were the most potent general anesthetics, while more sterically hindered analogues were inactive as anesthetics.²¹ The anesthetic activity of 6 is consistent with later studies showing that substitutions at the para-position are usually well tolerated.^{22, 23} The potency of 6 is predicted by its octanol/water partition coefficient, in line with the empirical Meyer-Overton Rule.

Effects on GABA_A Receptors

The apparent affinity of GABA-elicited currents was enhanced by 6 with a potency similar to that of propofol. At twice their general anesthetic EC₅₀ concentrations in tadpoles, 6 (6 μM) and propofol (4.4 μM) shifted the GABA-induced current concentration-response curve to the left by comparable amounts of 9– and 11–fold, respectively (Fig. 1C & D). This occurred without any dramatic systematic change in the maximum current elicited at high concentrations of GABA.

The concentration-dependence of potentiation of currents elicited by a subsaturating EC₅ concentration of GABA for both agents was comparable. Their EC₅₀s were similar and comparable to their general anesthetic value (Table 1). The efficacy of the compounds was also similar (Fig. 1A & B). The concentration-dependence of potentiation is consistent with the results observed by Krasowski et al in their study of 27 propofol derivatives where the introduction of the trifluoromethyl diazirine moiety on the aromatic ring of propofol could be best achieved at the 4-position, since substitutions at this position do not significantly alter the anesthetic properties of propofol.² In addition, they noted that maximum potentiation of GABA currents varied from 60 – 240%, with propofol exhibiting the greatest efficacy. In a study of 14 active propofol derivatives, two that were para-substituted had 2.5 and 4-fold lower efficacy than propofol for potentiating currents elicited by low concentrations of GABA.^{2, 15} In another study, out of 14 propofol derivatives substituted in the para-position, only 4 enhanced GABA currents but with efficacy similar to propofol’s.²²

Table 1.

Agent	General Anesthesia LoRREC50 (μM)	EC50 for EC5 enhancement Of GABAA currents (μM)	Enhancement of EC5 GABA currents at LoRREC50	Direct Activation	Shift in GABA Dose response curve	Enhancement of [3H] muscimol binding	Enhancement of [3H] flunitrazepam binding
Propofol	2.2c 2.8b 1.9d	5.2a 1.9d	4xb 5xa	30–50%a 67%d	11a	2.4xa	2xa
6	3.2 a	4.5a	8xa	10%a	9a	1.6xa	1.6xa
m-TFD-propofol	3.1b	ND	1.4xb	Noneb	ND	ND	ND

^aThis work.

^bFrom Hall et al, 2010.¹³

^cFrom Tonner and Miller, 1995.³⁷

^dFrom Krasowski et al, 2001.²

ND = not reported

In addition to enhancing GABA-induced currents, propofol alone is a very effective partial agonist, eliciting currents up to 30 – 70% of the maximal GABA response.^{2, 22} In the studies discussed above, the active para–substituted agents all directly activated and with efficacies 30 to 100% of propofol’s. Compound 6 also directly activated currents through the GABA_A receptor to 10% of the maximal GABA response. In contrast, m-TFD-propofol at 30 μM was reported to have no direct agonist activity.¹³ While halogen and aromatic 4-substitution appear to preserve efficacy for potentiation and activation²², the weak efficacy of the propofol photolabel derivatives adds to the small body of evidence suggesting that alkyl groups at the 3-, 4- & 5-positions result in low efficacy.²

Like propofol, 6 enhanced the binding of [³H]muscimol and [³H]flunitrazepam to HEK293 cell membranes expressing α1β3γ2 GABA_A receptors. The subunits are arranged reading clockwise as α1β3α1β3γ2. Muscimol binds at the two agonist binding sites in the α–β subunit interfaces, whereas flunitrazepam binds in the γ–α subunit interface. Both sites are in the receptor’s extracellular ligand–binding domain. Compound 6 potentiated both muscimol and flunitrazepam binding to similar extents, 70 and 80%, respectively with half effect concentrations of ~0.5 μM, well below the EC₅₀ for general anesthesia. Propofol was more efficacious and had a similar concentration–dependence except that at higher concentrations it caused a sharp decrease in [³H]ligand binding.

Effects on nACh Receptors

The only readily available and abundant biological source for a Cys-loop receptor is the electric organ from the Torpedo ray, thus making the nACh receptor the most convenient member of this family for assessing a new photolabel. In this work, we photolabeled the nACh receptor in the presence of various drugs that have known sites of action and allosteric effects, thus enabling an assessment of photolabeling efficacy and specificity. The nACh receptor has five subunits (2α,1β,1γ,1δ) that exhibit homologous secondary structure. A large extracellular agonist-binding domain is followed by a transmembrane domain of four α-helices, M1 – M4, with a large intracellular loop separating M3 and M4. The five M2 helices line the central ion-conducting pore and are the target of many drugs.

Compound 6 most efficiently photolabeled the nACh receptor α-subunit and that photolabeling was enhanced when the receptor was in the desensitized state (Fig. 4). For photolabeling at 1.7 μM [³H]6, the 3000 cpm of carbamylcholine-enhanced [³H] photoincorporation in the nACh receptor α-subunit indicates photolabeling of 0.2% of α-subunits, an efficiency of incorporation similar to that seen for [³H]TDBzl-etomidate or [³H]azietomidate²⁴ and sufficient to allow for identification of photolabeled nACh receptor amino acids.^{25, 26} This photolabeling pattern was similar to that seen for azioctanol²⁷ and azietomidate.²⁵ Thus, carbamylcholine, an agonist that stabilizes the desensitized state, significantly increased the photoincorporation of 6 into the nACh receptor α-subunit. Propofol also enhanced α-subunit photolabeling in both the absence and presence of agonist. The propofol enhancement of α–subunit photolabeling in the absence of agonist may occur because propofol can stabilize the nACh receptor in the desensitized state. However, in the presence of agonist the receptor is already desensitized, so that when propofol further enhances photoincorporation a second distinct for propofol site may be invoked.

We also examined the effects of two known nACh receptor channel blockers on [³H]6 photolabeling; tetracaine, which binds with high affinity to the channel when the nACh receptor is in the resting, closed channel state ²⁸, and phencyclidine, which binds with high affinity when nACh receptors are in the desensitized state.^{29, 30} The 6 binding site appears distinct from that of tetracaine, which had no effect on photolabeling in the resting state. However the partial inhibition by phencyclidine of photolabeling suggests that there may be some 6 binding within the nACh receptor ion channel in the desensitized state. Further photolabeling studies on a larger scale are required to directly identify the nACh receptor amino acids that are photolabeled.

On the nACh receptor, 6 inhibited currents with the same potency and efficacy as propofol. The current traces observed with increasing amounts of 6 suggest a role for desensitization in its inhibitory response, consistent with the photolabeling results that show increased labeling in the desensitized state.

Conclusions

Compound 6 is a novel, photoreactive derivative of the anesthetic propofol. In most respects it acts like propofol. It causes general anesthesia at a similar concentration, and modulates and activates heterologously expressed GABA_A receptors. This propofol derivative also functions as a photoaffinity agent, photoincorporating into nACh receptors in a pharmacologically specific fashion. Compared to a previously developed photoactivable propofol derivative, m-TFD-propofol, 6 has a different photoselectivity for amino acid residues. The two photoactivable derivatives of propofol have similar GABA_A receptor pharmacology, but they differ in detail and a comparison may allow for a more detailed mapping of binding sites just as it has in the case of a pair of etomidate derivatives containing an aliphatic diazirine, azi-etomidate, and an aromatic diazirine, TDBzl-etomidate.³¹

Experimental Section

Materials

[³H]Muscimol (3-hydroxy-5-aminomethylisoxazole, [Methylene-³H(N)]) (22.46 Ci/mmol), and [³H]flunitrazepam (flunitrazepam, [Methyl-³H]) (75.7 Ci/mmol) were from Perkin Elmer (Waltham, MA). GABA (γ-aminobutyric acid), flurazepam dihydrochloride, Propofol (2,6-diisopropylphenol, 97%) and all other chemicals and reagents, as well as solvents, including anhydrous THF, were purchased from Sigma–Aldrich (St. Louis, MO) and were used as received, without further purification. cDNAs for the α1, β2 and γ2L subunits of human GABA_A receptors in pCDM8 vectors were gifts from Dr Paul J. Whiting (Merck Sharp & Dohme Research Labs, Essex, UK).

Analytical Chemistry

¹H, ¹³C and ¹⁹F NMR spectra were recorded on a Bruker Avance spectrometer at 400 MHz, 100 MHz and 376 MHz respectively, unless otherwise noted, with TMS as an internal standard. HRMS experiments were performed with Q-TOF-2TM (Micromass). TLC was performed with Merck 60 F254 silica gel plates. Purity of the final compound was assessed by HPLC analysis with a Synergy Hydro-Rp column (4 um, 4.60x150 mm) using methanol-water with 0.05% trifluoroacetic acid; gradient applied was 50% MeOH at the start to 100% MeOH over 18 min, then isocratic. Elution was monitored by UV at 280 nm (aromatic absorbance). The analysis indicated purity greater than 97%.

4-Trifluoroacetylpropofol (2)

Aluminum chloride (4.00 g, 30.0 mmol) was suspended in dichloromethane (150 mL) and cooled to −48°C. Trifluoroacetic anhydride (3.06 mL, 4.62 g, 22 mmol) in dichloromethane (20 mL) was added dropwise to the stirred suspension. The resulting suspension was stirred at −48°C for 30 min, a solution of 2,6-diisopropylphenol (3.56 g, 20.0 mmol) in dichloromethane (20 mL) was added dropwise. The reaction mixture was stirred at the same temperature for 3 h and allowed to warm to room temperature overnight. The reaction mixture was poured onto a mixture of ice and 2.0 M HCl (150 mL each), stirred for 1h, the layers were separated and organic phase was washed with H₂O (2x100 mL), saturated NaHCO₃ and brine (100 mL each), dried over MgSO₄, and concentrated. The residue was chromatographed on silica gel using ethyl acetate – hexanes, 1:10, as eluent to yield pure 2 as a low-melting solid (3.18 g, 58%). ¹H NMR (360 MHz, CDCl₃): δ 7.81 (s, 2H, H_arom), 3.26 (sept, 2H, J = 6.8 Hz, CH-i-Pr), 1.24 (d, 12H, J = 6.8 Hz, CH₃). ¹³C NMR (90.6 MHz, CDCl₃): δ 179.2 (q, ²J_C-F = 33.5 Hz, C=O), 158.5, 135.8, 126.7, 117.6, 117.6 (q, ¹J_C-F = 291.0 Hz), 30.0, 22.1. ¹⁹F NMR (338.8 MHz, CDCl₃): δ −71.7 (s). HRMS (ESI): Calcd for C₁₄H₁₆F₃O₂ [M−H] (273.1108). Found: 273.1111.

O-tert-Butyldimethylsilyl-4-trifluoroacetylpropofol (3)

The mixture of the ketone 2 (1.00 g, 3.65 mmol), tert-butyldimethylsilyl chloride (660 mg, 4.38 mmol), imidazole (497 mg, 7.30 mmol) and dry DMF (5 mL) was stirred for 12h at 20°C. This mixture was added to hexane (50 mL), washed with water (50 mL) and the aqueous layer was extracted with hexane (2x20 mL). The combined organic phases were washed with water (3x30 mL), dried over MgSO₄ and concentrated under vacuum. Column chromatography on silica (1 to 3% of ether in hexanes as eluent) afforded the protected ketone 3 as a colorless solid (1.02 g, 27%, mp. 41–44°C). ¹H NMR (CDCl₃): δ 7.84 (unresolved q, 2H, ⁵J = 0.9 Hz, CH_arom), 3.33 (m, 2H, J = 6.9 Hz, CH-i-Pr), 1.22 (d, 12H, J = 6.9 Hz, CH₃-i-Pr), 1.06 (s, 9H, tert-Bu), 0.27 (s, 6H, Si-CH₃). ¹³C NMR (CDCl₃): δ 179.5 (q, ²J_C-F = 34.1 Hz, C=O), 156.6, 140.3, 126.6 (unresolved q, J_C-F = 2.1 Hz), 123.8, 117.0 (q, ¹J_C-F = 291.8 Hz), 26.8, 26.0, 23.0, 19.0, −3.2 (CH₃-Si). ¹⁹F NMR (CDCl₃): δ −70.6 (s). HRMS (ESI): Calcd. for C₂₀H₃₂F₃O₂Si [M+H] 389.2218. Found: 389.2129.

1-[4-(tert-Butyldimethylsilyloxy)-3,5-diisopropyl-phenyl]-2,2,2-trifluoro-ethanone oxime (4)

The ketone 3 (1.00 g, 2.57 mmol), hydroxylamine hydrochloride (357 mg, 5.15 mmol), and pyridine (0.31 mL, 3.9 mmol) were refluxed with stirring in methanol (4 mL) for 2 hours. The resultant clear solution was partitioned between water and 10% ethyl acetate in hexanes (40 mL each), the aqueous layer was extracted with the same solvent (2x20 mL), the combined organic phases were washed with water (3x20 mL), brine (20 mL), and dried over MgSO₄. Evaporation and chromatography on silica (ethyl acetate – hexanes, 1:20) afforded oxime 4 as a mixture of two isomers with 1:0.74 ratio as a low-melting solid (999 mg, 96%). ¹H NMR (CDCl₃): δ 9.49 (s, 1.7H, NOH), 7.38 (s, 2H, CH_arom), 7.23 (s, 1.5H, CH_arom), 3.42 – 3.29 (m, ~3.5H, CH-i-Pr), 1.21 (d, 21H, J = 6.8 Hz, CH₃-i-Pr), 1.08 (2 s, 16H, tert-Bu), 0.26 and 0.24 ( each s, 10.5H, CH₃-Si). ¹⁹ F NMR (CDCl₃): δ −62.3 (s, 1.26F), −65.7 (s, 1.74F). HRMS (ESI): Calcd for C₂₀H₃₃F₃NO₂Si [M+H] 404.2227. Found: 404.2233.

3-[4-(tert-Butyldimethylsilyloxy)-3,5-diisopropyl-phenyl]-3-trifluoromethyl-3H-diazirine (5)

The oxime 4 (100 mg, 0.248 mmol) and p-toluenesulphonyl chloride (95 mg, 0.50 mmol) were dissolved in dry dichloromethane (1 mL), cooled with stirring on an ice bath and pyridine (0.2 mL) was added dropwise. After 1 hour at 0°C the reaction mixture was warmed to room temperature and stirred for a further 24 h, diluted with ether (3 mL), washed with water (2x1 mL) and saturated NaH₂PO₄ (1 mL), dried over MgSO₄ and concentrated. The resultant colorless oil was dissolved in methanolic ammonia (7 M, Aldrich) and the resultant yellow solution was stirred at room temperature for 2 days. The reaction mixture was then concentrated under vacuum, dissolved in methanol/triethylamine mixture (10:1, 3 mL) and cooled to 0°C with stirring. Finely dispersed iodine (63 mg, 0.248 mmol) was added in small portions over 15 min until the mixture remained persistently brown. The excess iodine was quenched with 5% Na₂S₂O₃ (1 mL), the reaction mixture was diluted with water (15 mL) and extracted with DCM (3x5 mL). The combined organic phases were washed with water and brine (10 mL each) and dried over Na₂SO₄. Evaporation under vacuum and chromatography on silica gel (ethyl acetate – hexanes, 1:100) afforded the diazirine 5 (32 mg, 32%) as a low-melting solid. ¹H NMR (CDCl₃): δ 6.85 (s, 2H, H_arom), 3.30 (sept, 2H, J = 6.9 Hz, CH-i-Pr), 1.16 (d, 12H, J = 6.9 Hz, CH₃-i-Pr), 1.03 (s, 9H, tert-Bu), 0.20 (s, 6H, Si-CH₃). ¹³C NMR (CDCl₃): δ 150.6, 140.0, 122.4 (q, ¹J_C-F = 274.8 Hz), 122.1, 121.8, 28.6 (q, ²J_C-F = 40.0 Hz), 26.7, 26.0, 23.1, 18.9, −3.3 (Si-CH₃). ¹⁹F NMR (CDCl₃): δ −65.2 (s) HRMS (ESI): Calcd for C₂₀H₃₁F₃N₂OSi 400.2158. Found: 373.2178 (M⁺-N2+H⁺, C₂₀H₃₂F₃OSi, 373.2180).

Attempted desilylation of diazirine 5

In a typical experiment, TBAF (1 mL, 1 M in THF) was cooled to 0°C and mixed with acetic acid (mixtures from 1 to 5 eq. AcOH were used). Diazirine 5 (5.0 mg, 0.012 mmol) was dissolved in this TBAF-AcOH-THF mixture with stirring and the reaction was monitored by TLC. After completion, the reaction mixture was poured onto crushed ice (10 mL) and extracted with ether (2x5 mL). After aqueous work-up, TLC indicated that the initially formed compound decomposed into a complex mixture of several new products.

4-(4-Hydroxy-3,5-diisopropyl-phenyl)-4-oxo-butyric acid (7)

Finely powdered succinic anhydride (6.73 g, 67.3 mmol) was added in small portions to a stirring suspension of AlCl₃ (29.9 g, 224 mmol) in dry dichloromethane (300 mL). The reaction mixture was stirred at room temperature for 30 min, cooled to −78°C and the solution of 2,6-diisopropylphenol (10 g, 56 mmol) in dichloromethane (40 mL) was added dropwise over 30 min. The resultant yellowish suspension was allowed to warm up overnight. The progress of the reaction was monitored with ¹H NMR. The reaction mixture was cooled to −10°C, poured onto crushed ice (500 g) and warmed up to room temperature with stirring (1 h). Organic and aqueous phases were separated, the aqueous layer was extracted with dichloromethane (4x100 mL), the combined organic phases were washed with brine (200 mL) and dried over MgSO₄. Concentration followed by drying under vacuum gave a dark-brown oil which was added to a solution of KOH (12 g, 214 mmol) in methanol/water 1:1 (100 mL) and stirred for 2 hours. The resultant solution was diluted with water (0.5 L), acidified to pH 10 with HCl, and washed with ether (3x100 mL). The aqueous layer was acidified to pH 3 and extracted with dichloromethane (5x100 mL). The combined extracts were washed with water and brine (100 mL each), dried over MgSO₄ and concentrated. Crystallization from ethyl acetate – hexane, 1:9, gave the acid 7 as off-white crystals (3.00 g, 19%), mp. 146 – 147.5°C. ¹H NMR (CDCl₃): δ 12 – 9 (brs, 1H, COOH), 7.77 (s, 2H, CH_arom), 5.6 – 5.3 (brs, 1H, ArOH), 3.33 (t, 2H, J = 6.6 Hz, CH₂), 3.18 (sept, 2H, J = 6.9 Hz, CH-i-Pr), 2.82 (t, 2H, J = 6.6 Hz, CH₂), 1.31 (d, 12H, J = 6.9 Hz, CH₃). ¹³C NMR (CDCl₃): δ 197.1 (C=O), 178.7 (COOH), 154.8, 133.7, 129.3, 124.4 (CH_arom), 32.8 (CH₂), 28.3 (CH₂), 27.2 (CH-i-Pr), 22.5 (CH₃-i-Pr). HRMS (ESI): Calcd for C₁₆H₂₇O₃ [M+H] 279.1591. Found: 279.1594.

Methyl 4-(4-hydroxy-3,5-diisopropylphenyl)-4-oxobutanoate (8)

Methanol (40 mL) was cooled to −60°C and thionyl chloride (2.4 mL, 33 mmol) was added dropwise over 5 min. The resultant solution was slowly warmed to 0°C and the ketoacid 7 (3.00 g, 10.8 mmol) was added at once. The reaction mixture was left stirring overnight at room temperature. The resulting clear solution was diluted with dichloromethane (100 mL), cooled to −5°C and stirred with cold water (100 mL) for 5 min. The layers were separated, the aqueous phase was extracted with dichloromethane (2x30 mL), the combined organic phases were washed with water (2x50 mL), brine (50 mL), dried over MgSO₄ and concentrated. Column chromatography on silica gel (ethyl acetate – hexanes, 1:10 to 1:5) afforded the ketoester 8 as colorless crystals, yield 2.84 g (90%), mp. 74 – 76°C. ¹H NMR (CDCl₃): δ 7.77 (s, 2H, CH_arom), 5.82 (s, 1H, ArOH), 3.73 (s, 3H, OCH₃), 3.33 (t, 2H, J = 6.7 Hz, CH₂), 3.19 (sept, 2H, J = 6.9 Hz, CH-i-Pr), 2.78 (t, 2H, J = 6.6 Hz, CH₂), 1.29 (d, 12H, J = 6.9 Hz, CH₃). ¹³C NMR (CDCl₃): δ 197.3 (C=O), 173.8 (COOMe), 155.0, 133.8, 129.3, 124.3 (CH_arom), 51.8 (OCH₃), 33.0 (CH₂), 28.2 (CH₂), 27.1 (CH-i-Pr), 22.6 (CH₃-i-Pr). HRMS (ESI): Calcd for C₁₇H₂₃O₄ [M−H] 291.1602. Found: 291.1616.

4-(4-Hydroxy-3,5-diisopropyl-phenyl)-butyric acid (11)

Into a stirring solution of the ketoester 8 (1.00 g, 3.4 mmol) in ethanol (40 mL) was added an excess of sodium borohydride in several portions (total 1.287 g, 34 mmol) over 1 hour. The reaction was quenched with acetone (10 mL) and water (200 mL). The reaction mixture was washed with dichloromethane (3x30 mL), acidified with HCl to pH 4 and extracted with dichloromethane (5x30 mL). The acidic extract was washed with water and brine (50 mL each) and dried over MgSO₄. Concentration by crystallization from hexane afforded acid 11 as colorless crystals (638 mg, 71%), mp. 95–96°C. ¹H NMR (CDCl₃): δ 11.5-9.0 (br.s., 1H, CO₂H), 6.87 (s, 2H, CH_arom), 5.0-4.4 (br.s., 1H, ArOH), 3.16 (sept., 2H, J = 6.9 Hz, CH-i-Pr), 2.62 (t, 2H, J = 7.5 Hz, CH₂), 2.41 (t, 2H, J = 7.5 Hz, CH₂), 1.96 (quint., 2H, J = 7.5 Hz, CH₂), 1.28 (d, 12H, J = 6.9 Hz, CH₃-i-Pr). NMR ¹³C: (CDCl₃) δ 180.2 (C=O), 148.2, 133.8, 133.1, 123.4 (CH_arom), 34.9 (CH₂), 33.6 (CH₂), 27.2 (CH-i-Pr), 26.7 (CH₂), 22.8 (CH₃-i-Pr). HRMS (ESI): Calcd for C₁₆H₂₃O₃ [M−H] (263.1653). Found: 263.1662.

5-[4-(tert-Butyldimethylsilyloxy)-3,5-diisopropyl-phenyl]-pentan-2-one (12)

A mixture of acid 11 (300 mg, 1.135 mmol), N,O-dimethylhydroxylamine hydrochloride (221 mg, 2.27 mmol) and (benzotriazol-1-yloxy)tris(dimethylamino)phosphonium hexafluorophosphate (753 mg, 1.7 mmol) was stirred in dry dichloromethane (3 mL) under argon for 5 min at room temperature, cooled on an ice bath, and triethylamine (0.47 mL, 3.4 mmol) was added dropwise. The reaction mixture was left at room temperature overnight, diluted with diethyl ether (20 mL), washed with water (2x15 mL), 0.5 M HCl, 10% Na₂CO₃ and brine (10 mL each) and dried over MgSO₄. Concentration under vacuum afforded crude Weinreb amide as a reddish solid. The above was added to a solution of tert-butyldimethylsilyl chloride (257 mg, 1.7 mmol) and imidazole (155 mg, 2.27 mmol) in dry DMF (0.5 mL) and stirred for 24 h. The resulting emulsion was partitioned between ether and water (15 mL each), the organic layer was washed with water, 0.1 M HCl, saturated NaHCO₃ and brine (10 mL each) and dried over MgSO₄. The solution was filtered, concentrated and the residue dried under vacuum. This product was dissolved in dry THF (3 mL) under argon, cooled with a dry ice – acetone bath and a solution of methyllithium (2.1 mL, 1.6 M in ether, 3 eq.) was added dropwise. After stirring at −78°C for 2 hours the reaction mixture was quenched with a mixture of methanol and ethyl acetate (0.5 mL each) in THF (2 mL), followed by saturated aqueous NH₄Cl (10 mL). The reaction mixture then was warmed up to room temperature, extracted with ether (2x10 mL), and the combined organic phase was washed with water (2x10 mL), brine (10 mL) and dried over MgSO₄. Concentration followed by chromatography on silica gel (ethyl acetate – hexanes 1:50) afforded pure 12 as a colorless oil (337 mg, 79%). ¹H NMR (CDCl₃): δ 6.84 (s, 2H, H_arom), 3.29 (sept, 2H, J = 6.9 Hz, CH-i-Pr), 2.56 (t, 2H, J = 7.5 Hz, CH₂), 2.47 (t, 2H, J = 7.5 Hz, CH₂), 2.14 (s, 3H, CH₃-C=O), 1.90 (quint, 2H, J = 7.5 Hz, CH₂), 1.17 (d, 12H, J = 6.9 Hz, CH₃-i-Pr), 1.04 (s, 9H, tert-butyl), 0.20 (s, 6H, Si-CH₃). ¹³C NMR (CDCl₃): δ 209.2 (C=O), 147.2, 138.8, 134.3, 123.3 (CH_arom), 53.4, 43.2, 34.9, 30.0, 26.5, 26.1, 25.6, 23.5, 18.9, −3.3 (CH₃-Si). HRMS (ESI): Calcd for C₂₃H₄₁O₂Si [M+H] 377.2870. Found: 377.2861.

3-{3-[4-(tert-Butyldimethylsilyl)-3,5-diisopropyl-phenyl]-propyl}-3-methyl-3H-diazirine (13)

The ketone 12 (50 mg, 0.133 mmol) was dissolved in a mixture of methanol-ether (1:1, 0.5 mL) and cooled to −40°C. A solution of ammonia in methanol (1 mL, ~7 M) was added dropwise with stirring. The reaction mixture was warmed up to −30°C and stirred at this temperature for 6 h. The solution of hydroxylamine-O-sulfonic acid (23 mg, 0.20 mmol) in methanol (0.4 mL) was added dropwise, and the resulting clear solution was left at −20°C for 5 days with occasional stirring (the reaction progress was monitored by TLC). After completion, the reaction mixture was evaporated in vacuo, dissolved in a mixture of methanol (2 mL) and triethylamine (0.06 mL, 0.4 mmol) and cooled to −10°C. A saturated solution of iodine (34 mg, 0.133 mmol) in dichloromethane (1 mL) was added dropwise until the mixture remained persistently light-brown in color. The excess iodine was quenched with 5% solution of Na₂S₂O₃ (0.5 mL), the mixture was diluted with water (15 mL) and extracted with ether (3x10 mL). The combined organic phases were washed with water and brine (10 mL each) and dried over Na₂SO₄. Concentration and chromatography on silica gel using 1 to 2% ethyl acetate in hexanes as an eluent afforded pure diazirine 13 as a colorless oil (30 mg, 58%). ¹H NMR (CDCl₃): δ 6.83 (s, 2H, H_arom), 3.30 (sept, 2H, J = 6.9 Hz CH-i-Pr), 2.53 (t, 2H, J = 7.4 Hz, CH₂), 1.58 – 1.44 (m, 2H, CH₂), 1.44 – 1.37 (m, 2H, CH₂), 1.18 (d, 12H, J = 6.9 Hz, CH₃-i-Pr), 1.05 (s, 9H, tert-Bu), 1.04 (s, 3H, α-azi-CH₃) 0.21 (s, 6H, Si-CH₃). ¹³C NMR (CDCl₃): δ 147.2, 138.8, 134.5, 123.1 (CH_arom), 35.1 (CH₂), 34.0 (CH₂), 26.5, 26.1, 26.0 (CH₂), 25.8 (CN₂), 23.4, 19.9 (α-azi-CH₃), 18.9 (Si-CMe₃), −3.3 (Si-CH₃). HRMS (ESI): Calcd. for C₂₃H₄₁N₂OSi [M+H] 389.2983. Found: 389.2993.

2,6-Diisopropyl-4-[3-(3-methyl-3H-diazirin-3-yl)-propyl]-phenol (6)

Method A (Scheme 2): Diazirine 13 from the previous step (30 mg, 0.077 mmol) was deprotected with cesium fluoride (101 mg, 0.67 mmol) in ethanol or TBAF-acetic acid (1:2 molar ratio, 1M TBAF in THF + neat acetic acid, 1 mL, overnight) to obtain crude 6. Column chromatography on silica (ethyl acetate – hexanes 1:20) afforded pure diazirine 6 as a colorless oil (19 mg, 52% from ketone 12). ¹H NMR (CDCl₃): δ 6.85 (s, 2H, H_arom), 4.66 (brs, 1H, ArOH), 3.16 (sept, 2H, J = 6.9 Hz, CH-i-Pr), 2.53 (t, 2H, J = 7.5 Hz, CH₂), 1.55-1.45 (m, 2H, CH), 1.45-1.37 (m, 2H, CH₂), 1.29 (d, 12H, J = 6.9 Hz, CH₃-i-Pr), 1.04 (s, 3H, α-azi-CH₃). ¹³C NMR (CDCl₃): δ 148.1, 133.6, 133.5, 123.3 (CH_arom), 35.1 (CH₂), 33.9 (CH₂), 27.2 (CH-i-Pr), 26.2 (α-azi-CH₂), 25.8 (CN₂), 22.8 (CH₃-i-Pr), 19.9 (α-azi-CH₃). HRMS (ESI): Calcd for C₁₇H₂₅N₂O [M−H] 273.1972. Found: 273.1976.

Method B: Compound 6 was also synthesized by another route permitting introduction of tritium label by a reduction of the carbonyl group in the ketone 18 (Scheme 3) The solution of the ketone 18 (100 mg, 0.248 mmol) in absolute ethanol (1 mL) was stirred with sodium borohydride (47 mg, 1.24 mmol) for 1 h at room temperature. After completion of the reaction (TLC), cesium fluoride (188 mg, 1.24 mmol) was added and stirring continued for another 24 h (TLC). Then the reaction mixture was divided between hexane - ethyl acetate (5:1) and water phases (10 mL each), the aqueous phase was additionally extracted with the same solvent (2x5 mL), combined organic phases were washed with brine and dried over MgSO₄. The resulting solution was filtered, concentrated in vacuo and dissolved in a mixture of dichloromethane (3 mL) and triethylsilane (0.5 mL). The reaction mixture was cooled to −30°C, the solution of trifluoroacetic acid (0.4 mL) in dichloromethane (0.6 mL) was added dropwise, and the mixture was warmed up to room temperature (0.5 h). After the usual aqueous work-up, chromatography on silica gel (ethyl acetate – hexanes, 1:20) afforded the target compound 10 (61 mg, 90%) as a colorless oil, identical with that obtained via Scheme 2.

³H-2,6-Diisopropyl-4-[3-(3-methyl-3H-diazirin-3-yl)-propyl]-phenol

The tritium labeled 6 with specific radioactivity of 12 Ci/mmol was synthesized as described above by Vitrax Company, Placenta, CA. This compound was stored at −20°C as a solution in ethanol containing 1 mCi/mL.

Methyl 3-(2-(4-hydroxy-3,5-diisopropylphenyl)-1,3-dioxolan-2-yl)propanoate (14)

The solution of the ketoester 8 (2.00 g, 6.84 mmol), ethylene glycol (5 mL), trimethyl orthoformiate (2 mL) and p-toluenesulfonic acid monohydrate (66 mg, 0.34 mmol) in dichloromethane (5 mL) was stirred at room temperature for 48 h (the reaction was monitored by ¹H NMR). The reaction was quenched with triethylamine (1 mL), and the resulting solution was partitioned between ether and water (40 mL each). The organic phase was washed with water, brine (20 mL each), dried over MgSO₄, evaporated and chromatographed on silica (ethyl acetate – hexanes, 1:10, 1% triethylamine) to afford dioxolane 14 as a viscous colorless oil, that crystallized upon storage in a refrigerator (1.43, 62%, m.p. 84–86°C). ¹H NMR (CDCl₃): δ 7.14 (s, 2H, H_arom), 4.88 (brs, 1H, ArOH), 4.06 – 3.96 (m, 2H, OCH₂), 3.86 – 3.76 (m, 2H, OCH₂), 3.67 (s, 3H, OCH₃), 3.16 (sept, 2H, J = 6.9 Hz), 2.50 – 2.42 (m, 2H, CH₂), 2.28 – 2.20 (m, 2H, CH), 1.29 (d, 12H, J = 6.9, Me₂CH). ¹³C NMR (CDCl₃): δ 174.1 (C=O), 149.7, 133.8, 133.3, 120.8 (CH_arom), 109.8 (O-C-O), 64.6 (OCH₂), 51.5 (OCH₃), 35.6 (CH₂), 28.7 (CH₂), 27.3 (CH-i-Pr), 22.7 (CH₃-i-Pr). HRMS (ESI): Calcd for C₁₉H₂₉O₅ [M+H] 337.2010. Found: 337.2014.

4-(2-(4-Hydroxy-3,5-diisopropylphenyl)-1,3-dioxolan-2-yl)butan-2-one (15)

Ester 14 (1.80 g, 5.35 mmol) was dissolved in methanol (20 mL), and a solution of KOH (2.0 g, 36 mmol) and water (10 mL) was added at once. The reaction mixture was stirred at room temperature for 2 h (the progress was monitored by TLC), poured into 100 mL of water and carefully acidified with 10% H₃PO₄ to pH 3. The resulting emulsion was extracted with dichloromethane (5x20 mL), the extracts were combined and added with triethylamine (3 mL), dried over MgSO₄, evaporated and dried at 0.1 mm Hg and 100°C for 1 h. The residual solid was cooled to 0°C (with stirring under argon) and mixed with a dichloromethane solution (20 mL) of triethylamine (2.24 mL, 3 eq.), N,O-dimethylhydroxylamine hydrochloride (0.783 g, 8.03 mmol) and (benzotriazol-1-yloxy)-tris(dimethylamino)phosphonium hexafluorophosphate (3.55 g, 8.03 mmol). The resulting dark-red suspension was stirred overnight. After the completion of the reaction (monitored by TLC, ether-hexane 1:1 with small amount of 7 M ammonia in methanol), the reaction mixture was stirred with water (20 mL) and diluted with dichloromethane (100 mL). The layers were separated. The organic layer was washed with brine (50 mL), dried over MgSO₄, evaporated and filtered through a short silica gel column (eluted with ethyl acetate/hexanes 1:3, 1% triethylamine). The concentration of eluate afforded a dark-red oil (2.5 g). This product was dissolved in anhydrous THF (40 mL), cooled to −78°C with vigorous stirring, and a solution of methyllithium (10 mL, 1.6 M in ether, 3 eq.) was added dropwise. The solidified reaction mixture was warmed up to −20°C for 5 min, cooled to −78°C and quenched with ethyl acetate (10 mL), ethyl acetate – acetic acid 1:1 (5 mL) and water (50 mL). The usual work-up (as above), followed by chromatography on silica gel (ethyl acetate – hexanes, 1:10, 1% triethylamine) afforded ketoketal 15 as a viscous colorless oil (1.07 g, 62%), (crystallized upon storage at 4°C, m.p. 67–72°C). ¹H NMR (CDCl₃): δ 7.12 (s, 2H, H_arom), 5.11 (brs, 1H, ArOH), 4.05 – 3.95 (m, 2H, OCH₂), 3.85 – 3.75 (m, 2H, OCH₂), 3.18 (sept, 2H, J = 6.9 Hz, CH-i-Pr), 2.55 (t, 2H, J =7.5 Hz, CH₂), 2.20 (t, 2H, J =7.5 Hz, CH₂), 2.14 (s, 3H, CH₃C=O) 1.27 (d, 12H, J = 6.9, CH₃-i-Pr). ¹³C NMR (CDCl₃): δ 208.8 (C=O), 149.8, 133.8, 133.5, 120.7 (CH_arom), 110.0 (O-C-O), 64.5 (OCH₂), 38.2 (CH₂), 34.7 (CH₂), 29.8, 27.2, 22.8 (CH₃-i-Pr). HRMS (ESI): Calcd for C₁₉H₂₉O₄ [M+H] 321.2060. Found: 321.2067.

2,6-Diisopropyl-4-(2-(2-(3-methyl-3H-diazirin-3-yl)ethyl)-1,3-dioxolan-2-yl)phenol (16)

Diaziridination of ketoketal 15 was accomplished by a procedure similar to that used in the synthesis of 13 (Scheme 2). The crude product was purified by silica gel chromatography using ethyl acetate – hexanes, 1:10 with 1% of triethylamine as eluent. Starting from 1.00 g of 15, 330 mg of diazirine 16 (32%) was obtained as a colorless oil that slowly crystallized upon storage at 4°C, mp. 30–35°C. ¹H NMR (CDCl₃): δ 7.10 (s, 2H, CH_arom), 4.85 (brs, 1H, ArOH), 4.03 – 3.93 (m, 2H, OCH₂), 3.83 – 3.73 (m, 2H, OCH₂), 3.16 (sept, 2H, J = 6.9 Hz), 1.84 – 1.76 (m, 2H, CH₂), 1.48 – 1.40 (m, 2H, CH₂), 1.29 (d, 12H, J = 6.9, CH-i-Pr), 0.99 (s, 3H, α-azi-CH₃). ¹³C NMR (CDCl₃): δ 149.7, 133.9, 133.3, 120.7 (CH_arom), 109.9 (O-C-O), 64.5 (OCH₂), 34.9 (CH₂), 28.8 (CH₂), 27.3 (CH-i-Pr), 25.7 (CN₂), 22.7 (CH₃-i-Pr), 19.7 (α-azi-CH₃). HRMS (ESI): Calcd for C₁₉H₂₈N₂O₃ [M+H] 333.2173. Found: 333.2178.

1-(4-Hydroxy-3,5-diisopropyl-phenyl)-3-(3-methyl-3H-diazirin-3-yl)-propan-1-one (17)

Diazirine 16 (200 mg, 0.6 mmol) was dissolved in dichloromethane (3 mL), cooled to −20°C and trifluoroacetic acid (0.5 mL) was added dropwise with stirring. The reaction mixture was allowed to warm up to 0°C and water (1 mL) was added; stirring continued for 3 h (TLC was used to monitor the reaction). Water (5 mL) was added, the layers were separated, the aqueous phase was extracted with dichloromethane (2 mL), combined organic phases were washed with water and brine (5 mL each; the product is unstable under basic conditions) and dried over MgSO₄. Evaporation of solvent under vacuum left pure 17 as white crystals (163 mg, 94%), m.p. 87 – 90°C. ¹H NMR (CDCl₃): δ 7.71 (s, 2H, CH_arom), 5.5 – 5.3 (br.s., 1H, ArOH), 3.19 (sept, 2H, J = 6.9 Hz, CH-i-Pr), 2.80 (t, 2H, J = 7.4 Hz, CH₂), 1.87 (t, 2H, J = 7.4 Hz, CH₂), 1.31 (d, 12H, J = 6.9, CH₃-i-Pr), 1.11 (s, 3H, α-azi-CH₃). NMR ¹³C: (CDCl₃) δ 197.6 (C=O), 154.8, 133.7, 129.5, 124.3, 32.1, 28.8, 27.2, 25.7, 22.6, 20.1. HRMS (ESI): Calcd. for C₁₇H₂₃N₂O₂ [M−H] 287.1765. Found: 287.1770.

1-(4-(tert-Butyldimethylsilyloxy)-3,5-diisopropylphenyl)-3-(3-methyl-3H-diazirin-3-yl)propan-1-one (18)

The diazirine 17 (100 mg, 0.347 mmol) was dissolved in the pre-formed mixture of tert-butyldimethylsilyl chloride (157 mg, 1.04 mmol), imidazole (47 mg, 0.69 mmol) and dry DMF (0.5 mL). After stirring for 12 h at room temperature the reaction mixture was partitioned between ether and water (10 mL each), the organic phase was washed with brine, dried over MgSO₄, evaporated under vacuum and chromatographed on silica gel (2 to 5% of ethyl acetate in hexanes as eluent) to afford 18 as a colorless solid (120 mg, 86%), mp. 48–50°C. ¹H NMR (CDCl₃): δ 7.69 (s, 2H, H_arom), 3.32 (sept., 2H, J = 6.9 Hz, CH-i-Pr), 2.81 (t, 2H, J = 7.5 Hz, CH₂), 1.86 (t, 2H, J = 7.5 Hz, CH₂), 1.21 (d, 12H, J = 6.9, CH₃-i-Pr), 1.12 (s, 3H, α-azi-CH₃), 1.04 (s, 9H, tert-butyl), 0.23 (s, 6H, Si-CH₃). ¹³C NMR (CDCl₃): δ 197.6 (C=O), 154.2, 139.5, 130.6, 124.0 (CH_arom), 32.1 (CH₂), 28.8 (CH₂), 26.7, 26.0, 25.6 (CN₂), 23.2 (CH₃-i-Pr), 20.1 (α-azi-CH₃), 18.9 (Si-CMe₃), −3.3 (Si-CH₃). HRMS (ESI): Calcd for C₂₃H₃₉N₂O₂Si [M+H] 403.2775. Found: 403.2377.

Solubility and Partition Properties

The solubility of 6 in 0.01M Tris/HCL, pH 7.4 was determined by stirring excess compound in the buffer for 24h, centrifuging at 10,000g, and analyzing the average fraction area by reverse phase HPLC. To determine octanol/water partition coefficients of 6, the compound was stirred in a two-phase mixture of octanol and water, aliquots were removed from the separated phases and applied to an HPLC Proto 300 C-18 5 μm reverse-phase column (Higgins Analytical, Inc., Mountain View, CA). The UV detector was set with an absorbance wavelength of 220 nm, the mobile phase consisted of 100% acetonitrile, 0.05% TFA with a flow rate of 1 ml/min, and the average fraction area in each phase was estimated.

General Anesthetic Potency

Xenopus laevis tadpoles (Xenopus One, Dextor, Michigan) in the pre-limb-bud stage (1–2 cm in length) were housed in large glass tanks filled with Amquel+ (Kordon, div. of Novalek, Inc, Hayward, CA) treated tap water. Stock solutions of the test compound were made in ethanol. With prior approval of the MGH Subcommittee on Research Animal Care, general anesthetic potency was assessed in the tadpoles as follows. Groups of 5 tadpoles were placed in foil-covered 100 mL beakers containing varying dilutions of the test compound in 2.5 mM Tris HCl at pH 7.4 under low levels of ambient light. The final concentration of ethanol did not exceed 5 mM, a concentration that does not contribute to anesthesia.³² Every 10 minutes tadpoles were individually flipped using the hooked end of a fire-polished glass pipette until a stable response was reached (usually at 40 minutes). Anesthesia was defined as the point at which the tadpoles could be placed in the supine position, but failed to right themselves after 5 seconds (LoRR). All animals were placed in a recovery beaker of Amquel+ treated tap water and monitored for recovery overnight. Each animal was assigned a score or either 0 (awake) or 1 (lost righting reflex), and the individual points were fit to a logistic equation by nonlinear least squares.

Electrophysiology of GABA_A and Torpedo nACh Receptors

With prior approval by the Massachusetts General Hospital Subcommittee on Research Animal Care, oocytes were obtained from adult, female Xenopus laevis (Xenopus One, Dextor, Michigan) and prepared using standard methods as previously described (see below). In vitro transcription from linearized cDNA templates and purification of subunit specific cRNAs was carried out using Ambion mMessage Machine RNA kits and spin columns. For GABA_A receptor studies, oocytes were injected with ~100 ng total mRNA (α1, β2, γ2L) mixed at a ratio of 1:1:2 transcribed from human GABA receptor subunit cDNAs in pCDNA3.1.³³ For Torpedo nACh receptor studies, oocytes were injected with ~25 ng total mRNA mixed at a ratio of 2α:1β:1γ:1δ as previously described.³⁴

All two-electrode voltage clamp experiments were done at room temperature, with the oocyte transmembrane potential clamped at −50 mV and with continuous oocyte perfusion with ND96 (100 mM NaCl, 2 mM KCl, 10 mM Hepes, 1 mM EGTA, 1 mM CaCl₂, 0.8 mM MgCl₂, pH 7.5) at ~2 mL/min. Compound 6 and propofol were dissolved in DMSO at a concentration of 10 mM just prior to use. Stocks were further diluted in ND96 to achieve the desired concentration.

GABA_A and nACh receptor Concentration Response and Current Inhibition Studies

All agents were applied for 15–20 s; oocytes were washed ~3 min between each application. Currents were amplified using an Oocyte Clamp OC-725C amplifier (Warner Instrument Corp), digitized using a Digidata 1322A (Axon Instruments, Foster City, CA), and analyzed using Clampex/Clampfit 8.2 (Axon Instruments) and OriginPro 6.1 software. Concentration response data were fit by nonlinear least squares regression to the Hill (logistic) equation of the general form:

$${\text{I}}_{\text{X}}/{\text{I}}_{\text{GABA},max}=({\text{I}}_{\text{X},max}/{\text{I}}_{\text{GABA},max})\text{x}(1/(1+{{({\text{EC}}_{50}/[\text{X}])}^{\text{n}}}_{\text{H}}))$$

where X is the concentration of the activating ligand, I_GABA,max is the maximally evoked current, EC₅₀ is the concentration of X eliciting half of its maximal effect, and n_H is the Hill coefficient of activation. Inhibition experiments were fit with logistic equations of the form:

$$\text{I}=1-({{[\text{X}]}^{\text{n}}}_{\text{H}}/({{{\text{IC}}_{50}}^{\text{n}}}_{\text{H}}+{{[\text{X}]}^{\text{n}}}_{\text{H}})).$$

Allosteric Regulation of GABA_A Receptor Ligand Binding/Radioligand binding assays

α1β3γ2 GABA_A receptors were expressed in HEK293S–TetR cells. Homogenized cell membranes were prepared as described previously ³⁵ and 200 μg of α1β3γ2 GABA_A receptor membrane protein was resuspended in 500 μL assay buffer (10 mM phosphate buffer (pH 7.4), 200 mM KCl, and 1 mM EDTA). The suspension was equilibrated with radio–ligands (3nM [³H]muscimol or 1.5nM [³H]flunitrazepam) and various concentrations of 6 or propofol at 4°C for 1 h. The nonspecific binding was determined in the presence of 1 mM GABA or 1 mM flurazepam for [³H]muscimol and [³H]flunitrazepam binding, respectively. Next, the suspension was filtered on GF/B glass fiber filters (Whatman, Schleicher & Schuell, Maidstone, United Kingdom) that were pretreated in 0.5% w/v Poly(ethyleneimine) for 1 h. After receptor application, filters were washed under vacuum with 7 mL of cold assay buffer, and dried under a lamp for 30 min. Subsequently, they were equilibrated in Liquiscint (Atlanta, GA) and counted (Tri-Carb 1900, Liquid Scintillation analyzer, Perkin Elmer / Packard, Waltham, MA).

Photoincorporation of [³H]6 into the nACh receptor

nACh receptor-rich membranes, prepared from Torpedo californica electric organs as described ³⁶, were resuspended at 2 mg protein/mL in Torpedo physiological saline (250 mM NaCl, 5 mM KCl, 3 mM CaCl₂, 2 mM MgCl₂, and 5 mM sodium phosphate, pH 7.0) supplemented with 1 mM oxidized glutathione. Aliquots (100 μL, 160 pmol ACh binding sites) were equilibrated in glass test tubes for 30 min with 1.7 μM [³H]6 (12 Ci/mmol) in the absence or presence of other drugs. The samples were then transferred to a 96-well polyvinyl chloride microtiter plate and irradiated on ice for 30 min with a 365 nm UV lamp (Spectronics Corporation EN-16 lamp) at a distance of ~2 cm. After irradiation, the samples were pelleted and solubilized in electrophoresis sample buffer, and equal aliquots were resolved on two 1.5 mm thick, 8% acrylamide/0.33% bis-acrylamide gels. The gels were stained with Coomassie Blue R-250, and one was then treated with Amplify for fluorography and exposed to film (Biomax XAR, Kodak) for 2 weeks. For the second gel, the nAChR subunit bands were excised and quantified by liquid scintillation counting by extracting in [10 % TS-2 tissue solubilizer (Research Products International Corp, Mt Prospect, Il) and 90 % Ecoscint A (National Diagnostics, Atlanta, GA)] for three days to determine the amount of ³H associated with each band.

Abbreviations

ACh

acetylcholine

EC50

concentration required for 50% of full effect

GABA

γ-amino butyric acid

IC50

concentration required for 50% of full inhibitory effect

LoRR

loss of righting reflexes

nH

Hill coefficient

nACh

nicotinic acetylcholine

p-4-AziC5-Pro

p-(4-azipentyl)-propofol or 2,6-Diisopropyl-4-[3-(3-methyl-3H-diazirin-3-yl)-propyl]-phenol

PCP

phencyclidine

Rsn

rapsyn

TBDMS

tertbutyldimethylsilyl

TBPS

tert-butylbicyclophosphorothionate

TFA

trifluoroacetic acid

VDAC

voltage-dependent anion channel