Neurosteroid Analogues. 15. A Comparative Study of the Anesthetic and GABAergic Actions of Alphaxalone, Δ¹⁶-Alphaxalone and Their Corresponding 17-Carbonitrile Analogues

Abstract

Alphaxalone, a neuroactive steroid containing a 17β-acetyl group, has potent anesthetic activity in humans. This pharmacological activity is attributed to this steroid’s enhancement of γ-amino butyric acid-mediated chloride currents at γ-amino butyric acid type A receptors. The conversion of alphaxalone into Δ¹⁶-alphaxalone produces an analogue that lacks anesthetic activity in humans and that has greatly diminished receptor actions. By contrast, the corresponding 17β-carbonitrile analogue of alphaxalone and the Δ¹⁶-17-carbonitrile analogue both have potent anesthetic and receptor actions. The differential effect of the Δ¹⁶-double bond on the actions of alphaxalone and the 17β-carbonitrile analogue is accounted for by a differential effect on the orientation of the 17-acetyl and 17-carbonitrile substituents.

Alphaxalone (1a, (3α,5α)-3-hydroxypregnane-11,20-dione, Chart 1) is a steroid that has potent anesthetic activity [1,2]. By contrast, Δ¹⁶-alphaxalone (2a) does not [3,4]. Explanations for the dramatic influence of the Δ¹⁶-double bond on anesthetic activity were initially focused on how this structural modification altered the behavior of these two steroids in lipids [5–19].

Chart 1.

Although the different behaviors of steroids 1a and 2a in lipids are likely factors contributing to the difference in the anesthetic activity of the two compounds, it has been hypothesized that a pharmacophore-based differential interaction of the two compounds with GABA_A receptors is also important [20]. The Δ¹⁶ double bond found in steroid 2a has effects on both the conformation of the steroid D-ring and the free rotation of the 17-acetyl group [10,20]. A study of analogues of steroids 1a and 2a without the 11-ketone group and the 17-acetyl group (1b and 2b) found that these analogues share similarly weak potency as GABA_A receptor modulators. Thus, it was proposed that the effect of the Δ¹⁶ double bond on the steroid D-ring conformation is not important. Instead, this structural modification was hypothesized to fix the orientation of the C-17 acetyl group in a position that is not favorable for its interaction with the receptor thus explaining the diminished activities of steroid 2a [20]. However, steroids 1b and 2b do not have a C-17 substituent and therefore do not directly address the effect that the Δ¹⁶ double bond has on the orientation of a C-17 substituent in three dimensional space. Additionally, steroids 1b and 2b lack the 11-ketone group found in steroids 1a and 2a and any effect that this substituent might have on anesthetic activity is not addressed by these analogues.

We have prepared the C-17 carbonitrile analogues (3,4; Chart 1) of steroid 1a and Δ¹⁶-steroid 2a and compared the actions of compounds 1a, 2a, 3 and 4 on GABA_A receptor function. Unlike the previously prepared compounds 1b and 2b, which are both relatively weak modulators of GABA_A receptors because they lack a hydrogen bond acceptor group at C-17, steroids 3 and 4 are both strong modulators. Steroids 3 and 4 also contain the 11-ketone group which is present in steroids 1a and 2a, but not in steroids 1b and 2b, so that any potentially confounding effect caused by the absence of this group is avoided. We report that the conversion of steroid 3 into steroid 4 results in only a slight loss of potency for enhancement of GABA-mediated currents at GABA_A receptors, a slight increase in the IC₅₀ value for allosteric displacement of [³⁵S]-TBPS from the picrotoxin site on GABA_A receptors and a slight decrease in anesthetic potency as measured by LRR and LSR in tadpoles. These results further refine the previous hypothesis regarding the effect that a Δ¹⁶ double bond has on the activity of steroids that modulate GABA_A receptors, and identify a Δ¹⁶ analogue with high activity at these receptors.

The preparation of compounds is shown in Scheme 1. The 3α-hydroxyl group of commercially available steroid 5 was protected as the MOM derivative yielding steroid 6 in quantitative yield (Scheme 1). Conversion of steroid 6 into the Δ¹⁶-17-carbonitrile 8 was achieved by cyanation of the intermediate Δ¹⁶-17-triflate 7 in 89% yield. Removal of the MOM group under acidic conditions gave Δ¹⁶-steroid 4 in an isolated yield of 93%. Catalytic hydrogenation of Δ¹⁶-steroid 4 gave a quantitative yield of steroid 3 which was prepared previously by a different route as described in the patent literature [21]. Steroid 2a (Chart 1) was prepared in 32% yield from its commercially available 3β-hydroxysteroid epimer 9 by a Mitsunobu reaction [22].

Scheme 1^a.

^aReagents: a) MOMCl, Hunig’s base, CH₂Cl₂, room temperature; b) PhN(SO₂CF₃)₂, KHMDS, THF, −78 °C; c) NaCN, CuI, Pd(PPh₃)₄, MeCN; d) MeOH/CH₂Cl₂ (4:1), conc. HCl, room temperature; e) H₂ (30 psi), Pd/C (5%), EtOAc; f) i: DEAD, PPh₃, TFA, PhCO₂Na, THF; ii: NaHCO₃, aqueous MeOH.

The crystal structures of compounds 1a and 2a were previously unreported and were determined in this study. The conformations of these compounds in the solid state are shown in Figure 1 and are consistent with the solution conformations deduced from previous NMR studies [10] and molecular mechanics calculations [20].

Figure 1.

X-ray crystal structures of steroids 1a and 2a. The conformation of the 17-acetyl group in each steroid in the solid state is the same as deduced from solution NMR experiments [10] and molecular mechanics calculations [20].

The potency of compounds 1a, 2a, 3 and 4 for allosteric displacement of [³⁵S]-TBPS from the picrotoxin binding site on GABA_A receptors is reported in Table 1. The IC₅₀ values measured for steroids 1a and 2a (226 ± 24 nM and 2,220 ± 260 nM, respectively) are similar to literature values [20] (1a, 303 ± 37 nM; 2a, 2,956 ± 239 nM). Steroid 3 displaces [³⁵S]-TBPS with essentially the same potency as steroid 1a. This is as expected since a 17β-carbonitrile group has previously been shown to produce neurosteroid analogues with high potency for [³⁵S]-TBPS displacement [23]. The Δ¹⁶-steroid 4 is a weaker displacer of [³⁵S]-TBPS by a factor of two. This is in marked contrast to the ten-fold loss of potency observed when steroid 1a was converted into Δ¹⁶-steroid 2a.

Table 1.

Inhibition of [³⁵S]-TBPS binding by steroids 1a, 2a, 3 and 4.

Compound	IC50 (nM)a	nHill
1a	226 ± 24	1.10 ± 0.11
2a	2,220 ± 260	1.24 ± 0.14
3	190 ± 18	1.14 ± 0.11
4	361 ± 58	1.00 ± 0.14

^aResults are from duplicate experiments performed in triplicate. Error limits are calculated as standard error of the mean. Methods were as reported previously [27].

The effects of compounds 1a, 2a, 3 and 4 on the GABA-mediated chloride currents of rat α₁β₂γ_2L GABA_A receptors expressed in Xenopus laevis oocytes are reported in Table 2. There is a close correlation of electrophysiology results with the [³⁵S]-TBPS binding results. Steroid 1a produces a concentration dependent increase in chloride current that at a steroid concentration of 10 μM is about 20-fold higher than the control response in the absence of steroid 1a. By contrast, the Δ¹⁶-steroid 2a produces only about 2-fold concentration-dependent maximum enhancement of GABA responses. Steroid 3 produced an electrophysiological response that was essentially equal to that of steroid 1a. The Δ¹⁶-steroid 4 also gave a response that was essentially the same as that of steroid 1a and, more significantly, much greater than the response of Δ¹⁶-steroid 2a. Steroids 1a, 3 and 4 directly gated a small but significant chloride current in the absence of added GABA. The current directly gated by steroid 2a was not significant. To further reduce any variation in responses due to the fact that different preparations of oocytes were used in the electrophysiological experiments for the different analogues, all four steroids were tested at the same concentration on the same ooctyes (Figure 2). When evaluated in this way, steroid 4 did give a somewhat lower increase in chloride current than steroid 3. However, the effect of steroid 4 continued to be far greater than the effect of steroid 2a. The anesthetic effects of compounds 1a, 2a, 3 and 4 are summarized in Table 3. The potency of the anesthetic effects closely correlated with the effects found in the previous two bioassays. Steroids 1a and 3 as well as Δ¹⁶-steroid 4 all had similar EC₅₀ values for tadpole LRR (EC₅₀ ~1 μM) and LSR (EC₅₀ ~5.5 μM). The Δ¹⁶-steroid 2a was not effective (EC₅₀ >10 μM) in causing either LRR or LSR.

Table 2.

Modulation of rat α₁β₂γ_2L GABA_A receptor function by steroids 1a, 2a, 3 and 4.

Compound	oocyte electrophysiologya
	0.1 μM	1 μM	10 μM	(gating) 10 μM
1a	2.91 ± 0.57	4.70 ± 1.11	19.64 ± 4.04	0.11 ± 0.02
2a	0.94 ± 0.04	0.97 ± 0.05	1.87 ± 0.14	0.08 ± 0.07
3	1.12 ± 0.03	4.59 ± 0.42	21.14 ± 2.14	0.14 ± 0.03
4	1.49 ± 0.44	4.07 ± 1.09	23.75 ± 3.61	0.21 ± 0.04

^aThe GABA concentration used for the control response was 2 μM. Each compound was evaluated on at least four different oocytes at the concentrations indicated, and the results reported are the ratio of currents measured in the presence/absence of added compound. Gating represents direct current gated by 10 μM compound in the absence of GABA, and this current is reported as the ratio of compound only current/2 μM GABA current. Error limits are calculated as standard error of the mean (N ≥ 4). Methods were as reported previously [27].

Figure 2.

Direct comparison of the ability of steroids 1a, 2a, 3 and 4 to modulate GABA_A receptor-mediated chloride currents at compound concentrations of 1 μM and 10 μM. The compounds were evaluated on the same oocytes expressing recombinant rat α₁β₂γ_2L receptors. (A) Sample currents from an oocyte clamped to –70 mV and exposed transiently to GABA alone and then GABA plus each of the steroids. (B) Summary of effects of steroids on GABA responses. The current mediated by GABA alone is set to one as the control value and indicated by the dotted line. Potentiation is calculated as R2/R1, where R2 is the response in the presence of a steroid and R1 is the response to GABA alone. Error limits are calculated as standard error of the mean for n ≥ 4. Significance levels are as follows: steroids 1a and 2a at 1 μM and at 10 μM, P < 0.005; steroids 3 and 4 at 1 μM and at 10 μM, P < 0.005; steroids 1a and 3 at 1 μM, P < 0.05 and at 10 μM, P not significant; steroids 2a and 4, at 1 μM and at 10 μM, P < 0.005.

Table 3.

Effects of steroids 1a, 2a, 3 and 4 on tadpole righting and swimming reflexes

Compound	Tadpole LRRa EC50 (μM)	Tadpole LRR nHill	Tadpole LSRa EC50 (μM)	Tadpole LSR nHill
1a	1.12 ± 0.14	–3.38 ± 2.28	5.48 ± 0.11	–33 ± 0c
2a	> 10	–	Noneb	–
3	0.72 ± 0.11	–1.49 ± 0.26	5.48 ± 0.12	–33 ± 0c
4	1.04 ± 0.14	–1.77 ± 0.38	5.48 ± 0.12	–33 ± 0c

^aError limits are calculated as standard error of the mean (N = 10 animals at each of five or more different concentrations). Methods were as reported previously [27].

^bNone is defined as no loss of behavioral reflex at the highest concentration tested (10 μM).

^cAll tadpoles retained LSR at 3 μM and all tadpoles lost LSR at 10 μM. Steep slopes for LSR dose–response curves are commonly observed for anesthetics in this bioassay.

This study was performed to gain a better understanding of how anesthetic steroid analogues containing a Δ¹⁶ double bond interact with GABA_A receptors. As mentioned previously, the presence of the Δ¹⁶ double bond in steroid 2a has multiple structural effects. It affects the conformation of the steroid D-ring, eliminates free rotation of the 17-acetyl group about the C-17, C-20 bond and reorients this group in three-dimensional space. By contrast, a Δ¹⁶ double bond has fewer structural effects when the C-17 substituent is a carbonitrile group. Although the effect on the conformation of the steroid D-ring caused by a Δ¹⁶ double bond is the same for both the 17-acetyl and 17-carbonitrile groups, loss of free rotation of the substituent is only a factor for the 17-acetyl group. Additionally, the orientation of the 17-carbonitrile group along a vector passing midway through the C-14, C-15 bond and through C-17 (Figure 3) is not affected for the 17- carbonitrile group as it is for the 17-acetyl group.

Figure 3.

Partial structure of a steroid showing a vector that passes midway through the C14–C15 bond and C17. A hydrogen bond acceptor group at C-17 oriented along this vector is predicted to have high activity. A significant dispacement of the 17-acetyl carbonyl group to the left side of this vector occurs upon introduction of a Δ¹⁶ double bond into steroid 1a (see Figure 1), but a similar displacement does not occur for the 17-carbonitrile group of steroid 4.

Although it appears that the 17-acetyl substituent of steroid 2a does not interact favorably with the GABA_A receptor, it is not clear what conformation the 17β-acetyl substituent has when steroid 1a is bound to the GABA_A receptor. This substituent may not be in the minimum energy conformation found in calculations, solution and the solid state (Figure 1). Indeed, in a different study that utilized steroids containing ring constrained C-17 substituents, evidence for the importance of placing a hydrogen bond acceptor group above C-17 and along the vector shown in Figure 3 for obtaining high activity was described [24]. Rotation of the 17β-acetyl group of steroid 1a would allow this group to obtain such an orientation. A 17β-carbonitrile group also fulfills this structural requirement and introduction of a Δ¹⁶ double bond does not greatly displace this group to either side of the vector shown in Figure 3, although it does place the carbonitrile group in a vertical position that is intermediate between that of 17α and 17β substituents. Apparently, this intermediate positioning of the 17-carbonitrile group is of only minor significance since steroids 3 and 4 have similar biological activities. A somewhat larger loss of activity would not have been too surprising since steroids having a 17α-carbonitrile group are ineffective as modulators of GABA_A receptor function [25].

In conclusion, we have found that the potent GABAergic actions of steroids containing a 17β-carbonitrile instead of a 17β-acetyl group are not greatly affected by introduction of a Δ¹⁶-double bond. These results, obtained with analogues that are more similar to alphaxalone and Δ¹⁶-alphaxalone than those studied previously, support and refine the earlier hypothesis which proposed that the loss of activity for Δ¹⁶-alphaxalone is likely due to the negative consequences that the Δ¹⁶-double bond has on the positioning of the 17-acetyl group, not to conformational effects on the steroid D-ring [20].

Experimental Section

General Methods

Solvents were either used as purchased or dried and purified by standard methodology. Extraction solvents were dried with anhydrous Na₂SO₄ and after filtration, removed on a rotary evaporator. Flash chromatography was performed using silica gel (32–63 μm) purchased from Scientific Adsorbents (Atlanta, GA). Melting points were determined on a Kofler micro hot stage and are uncorrected. FT-IR spectra were recorded as films on a NaCl plate. NMR spectra were recorded in CDCl₃ at ambient temperature at 300 MHz (¹H) or 74 MHz (¹³C). Purity was determined by TLC on 250 μm thick Uniplates^™ from Analtech (Newark, DE). All pure compounds (purity >95%) gave a single spot on TLC. Elemental analyses were performed by M-H-W Laboratories (Phoenix, AZ). Steroids 1, 5 and 9 were purchased from Steraloids (Newport, RI).

(3α,5α)-3-Hydroxypregn-16-ene-11,17-dione (2a)

Steroid 9 (75 mg, 0.23 mmol) in anhydrous THF (0.5 mL) was added to a stirred solution of DEAD (0.15 mL, 0.34 mmol, 40% in toluene) and at room temperature TFA (22 μL, 0.29 mmol) and then solid PPh₃ (90 mg, 0.34 mmol) were added. After stirring the reaction for 10 min PhCO₂Na (50 mg, 0.35 mmol) was added and the reaction was stirred overnight. Since after this time a large amount of unreacted steroid 9 was detected by TLC, additional DEAD (60 μL, 0.14 mmol), PPh₃ (38 mg, 0.14 mmol) and PhCO₂Na (22 mg, 0.15 mmol) were added. The reaction was stirred for another 20 h and volatiles were removed under reduced pressure. The product ester was separated from starting material by column chromatography on silica gel (40% EtOAc in hexanes). The inverted benzoate ester (50 mg) was then hydrolyzed by refluxing overnight with NaHCO₃ (60 mg, 0.71 mmol) in MeOH (10 mL). Volatiles were removed and the crude product was extracted with CH₂Cl₂. The combined organic layers were washed with water, then brine and dried. The crude product was further purified by column chromatography on silica gel (30–50% EtOAc in hexanes). Pure compound 2a (24 mg, 32%) had: mp 253–54 °C; lit [26] mp 243–44 °C; [α]_D²⁰ +71.2 (c = 1.20, CHCl₃); IR 732, 919, 1000, 1368, 1434, 1666, 1703, 2857, 2923, 3407 cm⁻¹; ¹H-NMR: δ 0.82 (s, 3H), 1.02 (s, 3H), 2.27 (s, 3H), 3.02 (d, 1H, J = 12.6 Hz), 4.04 (br s, 1H), 6.75 (m, 1H); ¹³C-NMR δ 10.88, 17.27, 26.88, 27.76, 28.90, 30.87, 31.72, 32.39, 35.16, 35.29, 36.01, 39.07, 48.45, 53.99, 55.71, 65.71, 66.18, 144.40, 153.03, 195.93, 210.30.

(3α,5α,17β)-3-Hydroxy-11-oxoandrostane-17-carbonitrile (3)

Steroid 4 (30 mg, 0.10 mmol) was dissolved in EtOAc (3 mL) and 5% Pd/C (10 mg, 0.08 mmol) was added. The hydrogenation flask was then evacuated and filled with H₂ gas three times. The compound was hydrogenated for 3 h at 30 psi. The catalyst was filtered through Celite® and washed with CH₂Cl₂. Product 3 was obtained in a quantitative yield (30 mg) and had: mp 175–76 °C; lit [21] mp 256–262 °C; [α]²⁰ _D +71.5 (c = 1.24, CHCl₃); IR 732, 917, 1044, 1166, 1447, 1705, 2238, 2859, 2924, 3491 cm⁻¹; ¹H NMR δ 0.88 (s, 3H), 1.01 (s, 3H), 2.03–2.33 (m, 3H), 2.46–2.58 (m, 2H), 4.05 (m, 1H); ¹³C NMR δ 10.82, 15.12, 23.89, 26.89, 27.62, 28.79, 30.75, 32.54, 35.12, 35.73, 36.94, 38.78, 39.13, 47.26, 53.61, 54.66, 63.98, 66.07, 119.99, 208.34. Anal. Calcd for C₂₀H₂₉NO₂: C, 76.15; H, 9.27; N, 4.44. Found: C, 76.32; H, 9.06; N, 4.40.

(3α,5α)-3-Hydroxy-11-oxoandrost-16-ene-17-carbonitrile (4)

Steroid 8 (43 mg, 0.12 mmol) was dissolved in CH₂Cl₂ (0.5 mL) and MeOH (2 mL). Concentrated HCl (0.5 mL) was then added and the reaction was stirred for 4 h at room temperature. The solvents were removed and CH₂Cl₂ was added to dissolve the residue. The CH₂Cl₂ was washed with Na₂CO₃ solution, brine and dried. The crude compound was purified by column chromatography (silica gel, 30–40% EtOAc in hexanes) to yield product 4 (35 mg, 93%) as a white solid: mp 174–76 °C; [α]²⁰ _D +41.9 (c = 1.30, CHCl₃); IR 732, 916, 1045, 1382, 1454, 1593, 1705, 2218, 2860, 2924, 3408 cm⁻¹; ¹H NMR δ 0.86 (s, 3H), 1.02 (s, 3H), 2.42–2.60 (m, 3H), 4.05 (br s, 1H), 6.70 (br s, 1H); ¹³C NMR δ 10.87, 17.56, 27.59, 28.84, 30.81, 32.25, 32.36, 35.21, 35.40, 36.08, 38.96, 50.15, 52.96, 55.28, 65.78, 66.03, 114.83, 125.19, 147.82, 207.98. Anal. Calcd for C₂₀H₂₇NO₂: C, 76.64; H, 8.68, N, 4.47. Found: C, 76.82; H, 8.42; N, 4.47.

(3α,5α)-(3-Methyloxymethyl)oxy-androstane-11,17-dione (6)

To a stirred solution of steroid 5 (100 mg, 0.33 mmol) in anhydrous CH₂Cl₂ (4 mL), diisopropylethyl amine (0.10 mL, 0.57 mmol) was added in a N₂ atmosphere followed by dropwise addition of chloromethyl methyl ether (0.05 mL, 0.66 mmol). The reaction was stirred overnight at room temperature. Volatiles were removed and the product was purified by column chromatography (silica gel, 10–15% EtOAc in hexanes). Product 6 was isolated in a quantitative yield (114 mg) as a white crystalline solid: mp 135–36 °C; [α]²⁰ _D +108.5 (c = 1.18, CHCl₃); IR 1046, 1455, 1710, 1745, 2926 cm⁻¹; ¹H NMR δ 0.82 (s, 3H), 1.04 (s, 3H), 3.36 (s, 3H), 3.82–3.83 (m, 1H), 4.65 (q, 2H, J = 9.4 Hz, J = 6.6 Hz); ¹³C NMR δ 11.12, 14.54, 21.43, 25.86, 27.56, 29.61, 31.25, 31.37, 33.13, 35.61, 35.99, 39.52, 50.33, 50.58, 50.60, 55.18, 64.81, 71.17, 94.45, 209.04, 217.55. Anal. Calcd for C₂₁H₃₂O₄: C, 72.38; H, 9.26. Found: C, 72.50; H, 9.06.

(3α,5α)-3-(Methyloxymethyl)oxy-17-(trifluoromethanesulfonyloxy)-androst-16-en-11-one (7)

A solution of steroid 6 (100 mg, 0.29 mmol) in anhydrous THF (1.4 mL) was cooled to −78 °C and KHMDS (0.6 mL, 0.5 M in toluene, 0.3 mmol) was added dropwise. After stirring the reaction for 15 min, a solution of N-phenyltrifluoromethane sulfonimide (125 mg, 0.35 mmol) in THF (2.0 mL) was added dropwise and stirring at −78 °C was continued for 3 h. The reaction was quenched by adding saturated NH₄Cl solution (1 mL). The product was extracted with hexanes and the combined extracts were washed with brine and dried. After solvent removal, the product was further purified by column chromatography (silica gel, 1–2.5% EtOAc in hexanes). Product 7 (91 mg, 91% based on recovery of steroid 6, 27 mg) retained a trace of an aromatic impurity that could not be removed by column chromatography. Attempts to remove it by recrystallization also were not successful. Oily product 7 containing the trace impurity had: [α]²⁰ _D +39.5 ( c = 1.24, CHCl₃); IR 1044, 1143, 1213, 1423, 1632, 1708, 2863, 2926 cm⁻¹; ¹H NMR δ 0.89 (s, 3H), 1.02 (s, 3H), 3.36 (s, 3H), 3.82–3.83 (m, 1H), 4.66 (q, 2H, J = 9.4 Hz, J = 6.9 Hz), 5.65–5.67 (m, 1H); ¹³C NMR δ 11.08, 16.50, 25.92, 27.59, 28.09, 29.66, 31.31, 31.43, 33.26, 35.27, 35.87, 39.72, 46.99, 51.90, 53.60, 55.11, 66.01, 71.25, 94.49, 115.66, 118.48 (q, J_CF = 320.6 Hz), 156.13, 208.16.

(3α,5α)-3-(Methyloxymethyl)oxy-11-oxoandrost-16-ene-17-carbonitrile (8)

Compound 7 (1.39 g, 2.89 mmol) was placed in a 100 mL round bottom flask and the flask was evacuated and filled with N₂. NaCN (193 mg, 1.36 mmol), tetrakis(triphenylphosphine) Pd(0) (239 mg, 0.21 mmol), and CuI (55 mg, 0.29 mmol) were added, followed by addition of anhydrous acetonitrile (40 mL). The yellowish solution was then heated to reflux for 3 h. The flask was cooled and the reaction mixture was filtered through a Celite® pad and washed with EtOAc. After solvent removal, the residue was redissolved in EtOAc and the EtOAc was washed with water, brine and dried. After solvent removal, the crude product was purified by column chromatography (silica gel, 5–10% EtOAc in hexanes) and purified product 8 (92 mg, 89%) was isolated as a white solid: mp 165–67 °C; [α]²⁰ _D +47.3 (c = 1.19, CHCl₃); IR 910, 1045, 1094, 1144, 1366, 1380, 1456, 1587, 1714, 2216, 2871, 2949 cm⁻¹; ¹H NMR δ 0.85 (s, 3H), 1.02 (s, 3H), 2.42–2.60 (m, 3H), 3.36 (s, 3H), 3.80–3.85 (m, 1H), 4.65 (q, 2H, J = 9.7 Hz, J = 6.9 Hz), 6.69 (q, 1H, J = 1.7 Hz); ¹³C NMR δ 11.09, 17.58, 25.94, 27.67, 31.46, 32.24, 32.38, 33.26, 35.40, 35.87, 39.66, 50.18, 52.97, 55.14, 55.36, 65.81, 71.20, 94.52, 114.86, 125.24, 147.81, 207.93. Anal. Calcd for C₂₂H₃₁NO₃: C, 73.91; H, 8.74; N, 3.92. Found: C, 74.12; H, 8.53; N, 3.88.

[³⁵S]-TBPS Binding Methods

The methods used were as described previously [27].

Xenopus Oocyte Electrophysiological Methods

Receptor expression and whole-cell recordings were carried out as described previously [27].

Tadpole Behavioral Methods

The methods used were as described previously [27].

Abbreviations

GABA

γ-amino butyric acid

GABA_A

γ-amino butyric acid receptor type A

[³⁵S]-TBPS

[³⁵S]-t-butylbicyclophosphorothionate

LRR

loss of righting reflex

LSR

loss of swimming reflex