Structure-Activity Relationships for the Anaesthetic and Analgaesic Properties of Aromatic Ring-Substituted Ketamine Esters

A series of benzene ring substituted ketamine N-alkyl esters were prepared from the corresponding substituted norketamines. Few of the latter have been reported since they have not been generally accessible via known routes. We report a new general route to many of these norketamines via the Neber (oxime to α-aminoketone) rearrangement of readily available substituted 2-phenycyclohexanones. We explored the use of the substituents Cl, Me, OMe, CF3, and OCF3, with a wide range of lipophilic and electronic properties, at all available benzene ring positions. The 2- and 3-substituted compounds were generally more active than 4-substituted compounds. The most generally acceptable substituent was Cl, while the powerful electron-withdrawing substituents CF3 and OCF3 provided fewer effective analogues.


Introduction
Racemic ketamine (1, Figure 1) is an effective and widely used anaesthetic/analgaesic [1], and tiletamine (2) is a thiophene analogue widely used as an anaesthetic in veterinary medicine [2].  The sedative and analgaesic effects of 1 have commonly been attributed to its non-competitive antagonism of the calcium channel pore of the N-methyl-D-aspartate (NMDA) receptor [3,4], although this has recently been called into question [5]. Compared to opioid-type pain-relieving drugs, ketamine has the major advantages of no immediate respiratory depression or hyperalgesic The sedative and analgaesic effects of 1 have commonly been attributed to its non-competitive antagonism of the calcium channel pore of the N-methyl-D-aspartate (NMDA) receptor [3,4], although this has recently been called into question [5]. Compared to opioid-type pain-relieving drugs, ketamine has the major advantages of no immediate respiratory depression or hyperalgesic effects, and an absence of longer-term effects such as increased tolerance [6]. The primary drawback of 1 is its substantial psychotogenic effects, which have recently been attributed to its blockade of GluN2C-containing NMDA receptors [7]. These detrimental properties are exacerbated by its relatively long elimination half-life, which means that patients can be exposed to prolonged hallucinogenic events as levels of drug slowly decline. To control these effects ketamine is frequently co-administered with respiratory depressant hypnotic drugs like midazolam or propofol, but these can markedly reduce its clinical safety [8]. In an alternative approach, we have recently shown, in a rat infusion model [9], that alkyl ester derivatives of ketamine (e.g. 5a, 5b) are effective short-term anaesthetics/analgaesics. They minimise psychotomimetic side effects during recovery by undergoing very rapid metabolism by tissue esterases to the corresponding, much more polar, and inactive acids [10]. Ester side chains (CH 2 ) 2 CO 2 i Pr and (CH 2 ) 4 CO 2 Me were particularly suitable [9,11].
We now extend these structure-activity studies to include analogues with Cl, Me, OMe, CF 3 and OCF 3 substituted at each available position on the benzene ring, together with the unsubstituted ring and 2-F variants (Table 1). Such substituents can potentially greatly affect drug binding to target proteins through the lipophilic, electronic, and steric changes that they have on the molecule. They collectively cover a wide range of lipophilic and electronic properties while keeping the steric effect broadly similar [12]. For further comparison we also included the corresponding esters (20a, 20b) of the veterinary thiophene analogue tiletamine. effects, and an absence of longer-term effects such as increased tolerance [6]. The primary drawback of 1 is its substantial psychotogenic effects, which have recently been attributed to its blockade of GluN2C-containing NMDA receptors [7]. These detrimental properties are exacerbated by its relatively long elimination half-life, which means that patients can be exposed to prolonged hallucinogenic events as levels of drug slowly decline. To control these effects ketamine is frequently co-administered with respiratory depressant hypnotic drugs like midazolam or propofol, but these can markedly reduce its clinical safety [8]. In an alternative approach, we have recently shown, in a rat infusion model [9], that alkyl ester derivatives of ketamine (e.g. 5a, 5b) are effective short-term anaesthetics/analgaesics. They minimise psychotomimetic side effects during recovery by undergoing very rapid metabolism by tissue esterases to the corresponding, much more polar, and inactive acids [10]. Ester side chains (CH2)2CO2 i Pr and (CH2)4CO2Me were particularly suitable [9,11]. We now extend these structure-activity studies to include analogues with Cl, Me, OMe, CF3 and OCF3 substituted at each available position on the benzene ring, together with the unsubstituted ring and 2-F variants (Table 1). Such substituents can potentially greatly affect drug binding to target proteins through the lipophilic, electronic, and steric changes that they have on the molecule. They collectively cover a wide range of lipophilic and electronic properties while keeping the steric effect broadly similar [12]. For further comparison we also included the corresponding esters (20a, 20b) of the veterinary thiophene analogue tiletamine.  and 2-F variants (Table 1). Such substituents can potentially greatly affect drug binding to target proteins through the lipophilic, electronic, and steric changes that they have on the molecule. They collectively cover a wide range of lipophilic and electronic properties while keeping the steric effect broadly similar [12]. For further comparison we also included the corresponding esters (20a, 20b) of the veterinary thiophene analogue tiletamine.

Chemistry
The synthesis of the compounds of Table 1 from the corresponding norketamines is straightforward as we have previously demonstrated [9]. (Scheme 1).

Chemistry
The synthesis of the compounds of Table 1 from the corresponding norketamines is straightforward as we have previously demonstrated [9]. (Scheme 1). However, few analogues of norketamine with substituents other than a 2-Cl in the aromatic ring have been reported; only the unsubstituted compound 21 [13] and the 4-Cl (22) and 4-Br (23) [14] analogues. The 3-OMe (25) and 3-OH (26) derivatives have also been characterised, but only as metabolites of methoxetamine (24) [15] (Figure 2). However, few analogues of norketamine with substituents other than a 2-Cl in the aromatic ring have been reported; only the unsubstituted compound 21 [13] and the 4-Cl (22) and 4-Br (23) [14] analogues. The 3-OMe (25) and 3-OH (26) derivatives have also been characterised, but only as metabolites of methoxetamine (24) [ We initially sought to prepare the required new substituted norketamines by the published method for norketamine itself [13] that we had used previously [9], but this was not successful, probably due to the lower nucleophilicity of ammonia compared with methylamine in that process. The use of more nucleophilic precursor reagents (N-methylhydrazine, 4-methoxybenzylamine) was also not successful. We therefore developed a new general route to many of these required We initially sought to prepare the required new substituted norketamines by the published method for norketamine itself [13] that we had used previously [9], but this was not successful, probably due to the lower nucleophilicity of ammonia compared with methylamine in that process. The use of more nucleophilic precursor reagents (N-methylhydrazine, 4-methoxybenzylamine) was also not successful. We therefore developed a new general route to many of these required norketamines, via the Neber (oxime to α-aminoketone) rearrangement [16] of substituted 2-phenycyclohexanones 27a-27p via the hydrazines 28a-28p and hydrazinium salts 29a-29p to give the required norketamines 21, 22 and 30b,c,e-i, k-p (Scheme 2). We initially sought to prepare the required new substituted norketamines by the published method for norketamine itself [13] that we had used previously [9], but this was not successful, probably due to the lower nucleophilicity of ammonia compared with methylamine in that process. The use of more nucleophilic precursor reagents (N-methylhydrazine, 4-methoxybenzylamine) was also not successful. We therefore developed a new general route to many of these required norketamines, via the Neber (oxime to α-aminoketone) rearrangement [16] of substituted 2phenycyclohexanones 27a-27p via the hydrazines 28a-28p and hydrazinium salts 29a-29p to give the required norketamines 21, 22 and 30b,c,e-i, k-p (Scheme 2). The known 4-methoxynorketamine analogue 30j could not be prepared by the Neber rearrangement, presumably because of the powerful electron-donating and/or inductive effects of this substituent para to the reaction centre, and was prepared instead by the method described by Sato et al [17].
In a further exploration of the nature of the aromatic ring in these esters, we also prepared the thiophene-based tiletamine ester analogues 20a and 20b. Tiletamine itself (2: Figure 1) is a wellknown veterinary animal anaesthetic [2], considered to have a similar mechanism of action to ketamine. The required nortiletamine was prepared by the method of Sato et al. [17] (Scheme 3). The known 4-methoxynorketamine analogue 30j could not be prepared by the Neber rearrangement, presumably because of the powerful electron-donating and/or inductive effects of this substituent para to the reaction centre, and was prepared instead by the method described by Sato et al [17].
In a further exploration of the nature of the aromatic ring in these esters, we also prepared the thiophene-based tiletamine ester analogues 20a and 20b. Tiletamine itself (2: Figure 1) is a well-known veterinary animal anaesthetic [2], considered to have a similar mechanism of action to ketamine. The required nortiletamine was prepared by the method of Sato et al. [

Biology
The compounds were evaluated for their ability to anaesthetise rats when administered by continuous intravenous infusion, as reported previously [9]. Compounds were administered to initially deliver 20 mg/kg/min (weight-adjusted flow) to achieve a pedal withdrawal reflex score (PWR = 1), then titrated to maintain loss of righting reflex (LORR) for 10 min. Three rats were used in each study, with each group of rats also acting as their own ketamine control. Data were collected on the total dose of drug (mg/kg), to achieve LORR and a PWR = 1, and on the time (in seconds, from cessation of the infusion) to recovery of righting reflex (RORR) (recovery from the hypnotic anaesthesia effect). Given the complexity of the experimental protocol, the total dose for LORR (Table  1) is very consistent, with ranges of only 1.5-fold within each group. The consistency of the postsedation recovery times are expectedly lower, with ranges of about 2.5-fold. Average data for LORR and RORR are given in Table 1. During recovery the rats were monitored for multiple signs of behavioural dysfunction (see biology section for details) and the sum of these scores (from 0 to 4, with 0 being no effect and 4 being severe dysfunction) is given in Table 2. IC50 values for inhibition of the NMDA receptor were conducted by Eurofins PanLabs, Taiwan.

Biology
The compounds were evaluated for their ability to anaesthetise rats when administered by continuous intravenous infusion, as reported previously [9]. Compounds were administered to initially deliver 20 mg/kg/min (weight-adjusted flow) to achieve a pedal withdrawal reflex score (PWR = 1), then titrated to maintain loss of righting reflex (LORR) for 10 min. Three rats were used in each study, with each group of rats also acting as their own ketamine control. Data were collected on the total dose of drug (mg/kg), to achieve LORR and a PWR = 1, and on the time (in seconds, from cessation of the infusion) to recovery of righting reflex (RORR) (recovery from the hypnotic anaesthesia effect). Given the complexity of the experimental protocol, the total dose for LORR (Table 1) is very consistent, with ranges of only 1.5-fold within each group. The consistency of the post-sedation recovery times are expectedly lower, with ranges of about 2.5-fold. Average data for LORR and RORR are given in Table 1. During recovery the rats were monitored for multiple signs of behavioural dysfunction (see biology section for details) and the sum of these scores (from 0 to 4, with 0 being no effect and 4 being severe dysfunction) is given in Table 2. IC 50 values for inhibition of the NMDA receptor were conducted by Eurofins PanLabs, Taiwan.

Results and Discussion
Ketamine has long been of interest for its multiple biological activities and its potential as a non-opioid anaesthetic. While a large number of side-chain analogues are known, relatively few with benzene ring substituents have been reported. The new route that we report here by the Neber rearrangement [16] of readily available substituted 2-phenycyclohexanones gives access to a wide range of benzene-substituted ketamines. We compare two sets of N-alkyl esters as short-acting anaesthetics, exploring Cl, Me, OMe, CF 3 , and OCF 3 benzene ring substituents. Table 1 gives structural and biological data for ketamine (1), two previously-reported [9,10] ester analogues (5a, 5b), two series of esters of novel benzene ring-substituted analogues (4a-19a, 4b-19b) and two similar esters (20a, 20b) of the thiophene-based analogue tiletamine (2, Figure 1) [12].
It has often been stated, [3,4] and equally disputed [5], that the sedative analgaesic and psychotomimetic properties of ketamine are due to interaction with/inhibition of the NMDA receptor. The data acquired in this study for ketamine (1) and the ketamine esters (3a-19a, 3b-19b) provide an opportunity to test these claims.
Of the 34 compounds studied (compound 17b was omitted due to its seizure-inducing effect), 12 showed no sedative or analgaesic activity at doses up to 200 mg/kg. All but one of these compounds (9b) also showed no psychotomimetic properties, as judged by the behavioural dysfunction test. They also showed much weaker inhibition of the NMDA receptor than ketamine (IC 50 0.7 µM), with IC 50 s ranging from 57 to >1000 (average IC 50 350 µM).
In contrast, the 22 actively sedative compounds (including ketamine and the previously reported ester analogues 5a and 5b) had a wide range of IC 50 s for inhibition of NMDA (from 0.7 to >1000 µM) but did have a much lower average IC 50 (167 µM). All but compounds 5a and 17a also showed analgaesic activity (most at potencies <60 mg/kg). This is broadly consistent with some relationships between these properties and NMDA inhibition. The majority of the active compounds were also much less psychotomimetic than ketamine (behavioural dysfunction score 3), but this may be at least in part due to the much faster recovery times for the esters (the main reason for this work). The shorter-chain ester analogue 20a of tiletamine (2) was not active, but the longer-chain analogue 20b was also an effective and relatively potent anaesthetics and analgaesics. However, 20b generated severe dysfunction on awakening (score 4).
In terms of sedative activity structure-activity relationships for the benzene ring substituents, the active (CH 2 ) 4 CO 2 Me series compounds were on average about 2.5-fold more potent than the corresponding (CH 2 ) 2 CO 2 i Pr shorter-chain series, but the ring substituent effects were broadly similar across both series. The 2-and 3-substituted compounds were generally more active than 4-substituted compounds. The active anaesthetic compounds with the shorter (CH 2 ) 2 CO 2 i Pr chain (4a-19a) included all of the 2-substituted examples except 2-Me (6a), making this overall the favoured position for substitution. The 3-Cl, 3-Me and 3-OMe compounds (10a-12a) and the 4-Cl and 4-OMe (15a, 16a) analogues were also active anaesthetics. Overall, the most generally acceptable substituent was Cl, while the non-polar and powerful electron-withdrawing substituents CF 3 and OCF 3 were the least successful. All of the compounds generated very little dysfunction in the rats during recovery (averaged scores of mostly 0 or 1, of short duration), in contrast to ketamine (average score 3 for a prolonged period).
Overall, this study has helped to define the SAR for this series of ketamine esters and provide useful information towards selection of a clinical candidate.

Conclusions
The above results show that the short chain aliphatic ester analogues of ketamine across the range of different benzene ring substituted compounds broadly retain the parent's desirable anaesthetic and analgaesic properties, yet are sufficiently rapidly metabolised to minimise the drawbacks of ketamine in this capacity. The structure activity relationships for the esters were not straightforward, the results suggest the (CH 2 ) 4 CO 2 Me series compounds were on average more active than the corresponding (CH 2 ) 2 CO 2 i Pr shorter-chain series. The 2-and 3-substituted compounds were generally more active than 4-substituted compounds.

Chemistry
All reagents and solvents were obtained from commercial suppliers and were used without further purification. Reactions requiring anhydrous conditions were performed under nitrogen atmospheres. Reactions were monitored by thin layer chromatography (TLC) on preloaded silica gel F254 plates (Merck, Darmstadt, Germany). with a UV indicator. Column chromatography was performed with Merck 230-400 mesh silica gel. 1 H and 13 C NMR spectra were obtained with a Bruker Avance 400 spectrometer (Bruker, Zuerich, Switzerland) at 400 MHz for 1 H and 101 MHz for 13 C spectra. Spectra were obtained in CDCl 3 or (CD 3 ) 2 SO. The chemical shifts are reported in parts per million (δ) downfield using tetramethylsilane (SiMe 4 ) as internal standard. Spin multiplicities are given as s (singlet), d (doublet), dd (double doublet), br (broad), m (multiplet), and q (quartet). Coupling constants (J values) were measured in hertz (Hz). All LC/MS data were gathered by direct injection of methanolic solutions into a Surveyor MSQ mass spectrometer using an atmospheric pressure chemical ionisation (APCI) with a corona voltage of 50 V and a source temperature of 400 • C. High-resolution electrospray ionisation (HRESIMS) mass spectra were determined on a Bruker micrOTOFQ II mass spectrometer (Bruker, Switzerland). Final products were analysed by reverse-phase HPLC (Alltima C18 5 µm column, 150 mm × 3.2 mm; Alltech Associated, Inc., Deerfield, IL, USA) using an Agilent HP1100 equipped with a diode array detector. The mobile phase was 80% MeCN/20% H 2 O (v/v) in 45 mM HCO 2 NH 4 at pH 3.5 and 0.5 mL/min. The purity was determined by monitoring at 272 nm and was ≥95% for final products unless otherwise stated. DCM refers to dichloromethane, DMF refers to N,N-dimethylformamide, EtOAc refers to ethyl acetate, EtOH refers to ethanol.