For the in vivo study, adult Sprague–Dawley rats were obtained from the Ruakura Animal Research Centre, Hamilton, New Zealand, with approval from the Ruakura Animal Ethics Committee. For the in vitro cortical slice study, adult mice (C57/BL6/129SV) were obtained from a breeding colony at Waikato University, Hamilton, New Zealand, with approval from the Waikato Animal Ethics Committee.
The methods were divided into two parts.
Part 1: correlating in vivo hypnotic potency with cortical slice electrophysiology
In vivo analysis of anaesthetic hypnotic potency
The in vivo analysis of the hypnotic potency of 21 ketamine analogue compounds has been reported in part elsewhere (Harvey et al. 2015; Jose et al. 2013). The results form part of a wider screening investigation pursuing the development of ketamine-ester analogues with rapid offset characteristics via hydrolysis of pharmacologically active ester groups. This set of compounds included seven non-ester entities (including ketamine), but for simplicity we will retain the term “ketamine-esters” to describe the collective group. For the purpose of this study, these ketamine-ester variants provided a range of structurally similar compounds with varying hypnotic potencies—which could be correlated with their effect on cortical slice field potential activity.
The methodology of drug design, synthesis and testing has been detailed previously (Harvey et al. 2015; Jose et al. 2013). In brief; adult female Sprague–Dawley rats (n = 3 per agent) were non-traumatically restrained and the marginal vein of the tail cannulated. One of 21 ketamine-ester analogues was delivered at 10 mg/ml via a minibore extension tube secured to the tail. Weight-adjusted infusions were administered at 20 mg/kg/min initially and continued until the animal lost both its ability to maintain righting, and attenuated its withdrawal response to firm digital pressure on the forepaw. Thereafter, the infusion rate was reduced to 6.7 mg/kg/min and adjusted in an up-and-down fashion to maintain dorsal recumbency for 10 min, before cessation. The dose (mg/kg) to loss of righting was adopted as a measure of effective hypnotic potency.
Cortical slice electrophysiology
Cortical slice preparation
Cortical slices were prepared from adult mice of either sex. The animals were anaesthetised with carbon dioxide prior to decapitation and brain dissection. The cerebrum was placed into ice-cold carbogenated (95 % O2; 5 % CO2) artificial cerebrospinal fluid (aCSF) containing: 92.7 mM NaCl, 3 mM KCl, 19 mM MgCl2, 0 mM CaCl2, 1.2 mM NaH2PO4, 24 mM NaHCO3 and 25 mM d-glucose (Nowak and Bullier 1996). Coronal slices (400 µM) were cut between Bregma −1 to −5 mm on a vibratome (Campden Instruments, UK) and transferred to a holding bath with carbogenated aCSF containing zero magnesium (124 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1.25 mM NaH2PO4, 26 mM NaHCO3 and 10 mM d-glucose). The slices were left undisturbed for at least an hour prior to recording at room temperature (approximately 28 °C).
Electrophysiology recording parameters and experimental procedure
One slice at a time was transferred to a recording bath (Tissue Recording System, Kerr Scientific Instruments, New Zealand) perfused with carbogenated zero-magnesium aCSF at a gravity-fed flow rate of 6.0 ml/min. Removal of magnesium ions from the aCSF activates the cortical tissue, resulting in spontaneous field potential activity resembling short seizure-like events (SLEs) that can be recorded unabated for several hours (Voss and Sleigh 2010). Field potentials were recorded using Teflon-coated (50 µm) tungsten electrodes, referenced to a silver/silver-chloride electrode located in the recording bath. Up to four recording electrodes were positioned equidistant apart in the cerebral cortex, with no particular cortical location targeted. The data was recorded with a 1000× gain, low- and high pass filtered at 1000 and 1.0 Hz respectively (Model 1800 AC amplifier, A-M Systems, USA) and sampled at a frequency of 5000 samples/second (Power 1401, Cambridge Electronic Designs, UK). Recordings were saved for analysis using Matlab (Version 7.3.0.267 (R2006b), The Mathworks Inc., Natick, MA, USA).
Testing ketamine variants in cortical slices
Recordings were made from 24 slices from 6 animals. SLE activity was recorded for at least 10 min to achieve a baseline. Thereafter, one of the ketamine-ester test agents was perfused at 45 µg/ml for 20 min followed by drug washout for 20 min with drug-free zero-magnesium aCSF. All agents were tested at the same dose. On eight occasions, two or three agents were tested in the same slice, in which case sufficient time was allowed for SLE activity to return to baseline levels before perfusing the next drug. Where multiple electrodes were positioned in the same slice, each channel was considered an independent recording on the condition that SLE activity was not coupled between locations. The basis of this proviso is that neocortical SLE activity can be generated from multiple independent locations within the same slice, just as if the slice was physically sectioned between recording locations (Voss et al. 2012).
Testing propofol and etomidate in cortical slices
In addition to the ketamine variants, dose response characteristics of propofol and etmoidate, two established general anaesthetics, were quantified in cortical slices. For propofol, 28 recordings were made from 20 slices (10 animals) and for etomidate, 18 recordings were made form 14 slices (9 animals). Following at least 10 min of baseline SLE recording, one or other drug was perfused at 3 sequential doses (28, 56 and 84 µM for propofol and 8, 16 and 24 µM for etomidate). Each dose was applied for 30 min in a step-wise manner until either the maximum dose was reached or the SLE frequency reduced to <50 % of that established during baseline. Thereafter, washout with zero-magnesium aCSF was continued for 40 min.
Data analysis
For the ketamine-esters, the drug effect on SLE frequency, length and amplitude was quantified as the mean percent change in each parameter from baseline relative to the 15–20 min period towards the end of drug perfusion. This period represents the time at which the drug was at peak concentration within the slice bath, taking into consideration the initial wash-in period. The in vivo hypnotic potency [the dose (mg/kg) to loss of righting] for each variant was related to its effect on SLE frequency, amplitude and length. Linear regression was used to quantify the relationship between in vivo hypnotic potency and change in each SLE parameter.
For the propofol and etomidate experiments, SLE frequency was normalised to baseline and averaged over three 30 min time periods corresponding to each drug dose (offset by 10 min to the start of each dose to allow for drug equilibration in the perfusion bath) and a washout period at the end of the recording. If the second or third dose was not delivered (because SLE frequency had already reduced by more than 50 %), for that recording the frequency was assumed to be zero for the analysis of those time periods.
Part 2: testing astrocytic metabolic inhibition on cortical slice SLE activity
The methods for cortical slice preparation and electrophysiological recording of zero-magnesium SLE activity were as described above.
Two astrocyte metabolic inhibitors were tested, fluoroacetate and aminoadipic acid. Fluoroacetate was delivered in three concentrations, 1, 5 and 10 mM. The pH of the 5 mM solution was 7.52 and was not adjusted. The pH of the 10 mM solution was 7.74 and was adjusted to 7.58 with 0.1 M HCl. The osmolarity of the 10 mM solution was adjusted to the equivalent of 5 mM by reducing the NaCl concentration in solution by 5 mM. All fluoroacetate concentrations were run for 45 min, followed by drug washout for 40 min. The effects were similar across all concentrations, therefore the data was pooled. For statistical analysis, SLE amplitude, length and frequency were averaged for each slice over three broad epochs: the 7 min period prior to drug delivery; the 45 min period of drug delivery; and the first 20 min period of drug washout. Because the data was not normally distributed (Kolmogorov and Smirnov test), the three epochs were compared statistically using non-parametric repeated measures ANOVA (Friedman test).
Aminoadipic acid was delivered in 3 concentrations, 0.5 (n = 17 from 3 animals), 1.0 (n = 13 from 2 animals) and 5 mM (n = 8 from 2 animals). The pH of the 1 mM solution was 7.54 and was not adjusted. The pH of the 5 mM solution was 7.2 and was adjusted to 7.58 with 0.1 M NaOH for three slices. The effect was qualitatively identical whether pH was adjusted or not and the data was therefore pooled. The 0.5 mM solution was run for 15 min, the 1 mM solution for 10 min and the 5 mM solution for 7 min, followed by drug washout. SLE amplitude, length and frequency were averaged for each slice over seven 100 s epochs, one during baseline recording 3 min before start of drug infusion and six sequential epochs from the start of drug infusion. This enhanced time resolution was necessary because aminoadipic acid had multiple effects both within and between the three concentrations tested. Because the data was not normally distributed (Kolmogorov and Smirnov test), the epochs were compared statistically using non-parametric repeated measures ANOVA (Friedman test). Only statistical comparison to the baseline epoch is reported. In three cases for the 5 mM dose, recording was terminated immediately after SLE activity ceased. It was assumed in these cases that SLE activity would have continued suppressed for the remaining epochs under analysis, in keeping with the data from the slices in which recording was continued.
Combined astrocytic metabolic inhibition and anaesthetic delivery
To confirm whether the anaesthetic effect on SLE frequency persisted during astrocytic metabolic inhibition, in two recordings each for propofol and etomidate, slices were pretreated with 0.5 mM aminoadipic acid for 15 min before anaesthetic perfusion (84 and 24 µM, respectively). When SLE frequency had reduced to at least half of the baseline frequency, aminoadipic acid and anaesthetic were washed out with drug-free zero-magnesium until SLE activity returned.