Alfaxalone does not have long‐term effects on goldfish pyramidal neuron action potential properties or GABAA receptor currents

Anesthetics have varying physiological effects, but most notably alter ion channel kinetics. Alfaxalone is a rapid induction and washout neuroactive anesthetic, which potentiates γ‐aminobutyric acid (GABA)‐activated GABAA receptor (GABAA‐R) currents. This study aims to identify any long‐term effects of alfaxalone sedation on pyramidal neuron action potential and GABAA‐R properties, to determine if its impact on neuronal function can be reversed in a sufficiently short timeframe to allow for same‐day electrophysiological studies in goldfish brain. The goldfish (Carassius auratus) is an anoxia‐tolerant vertebrate and is a useful model to study anoxia tolerance mechanisms. The results show that alfaxalone sedation did not significantly impact action potential properties. Additionally, the acute application of alfaxalone onto naive brain slices caused the potentiation of whole‐cell GABAA‐R current decay time and area under the curve. Following whole‐animal sedation with alfaxalone, a 3‐h wash of brain slices in alfaxalone‐free saline, with saline exchanged every 30 min, was required to remove any potentiating impact of alfaxalone on GABAA‐R whole‐cell currents. These results demonstrate that alfaxalone is an effective anesthetic for same‐day electrophysiological experiments with goldfish brain slices.

Anesthetics have varying physiological effects, but most notably alter ion channel kinetics.Alfaxalone is a rapid induction and washout neuroactive anesthetic, which potentiates c-aminobutyric acid (GABA)-activated GABA A receptor (GABA A -R) currents.This study aims to identify any long-term effects of alfaxalone sedation on pyramidal neuron action potential and GABA A -R properties, to determine if its impact on neuronal function can be reversed in a sufficiently short timeframe to allow for same-day electrophysiological studies in goldfish brain.The goldfish (Carassius auratus) is an anoxia-tolerant vertebrate and is a useful model to study anoxia tolerance mechanisms.The results show that alfaxalone sedation did not significantly impact action potential properties.Additionally, the acute application of alfaxalone onto naive brain slices caused the potentiation of whole-cell GABA A -R current decay time and area under the curve.Following whole-animal sedation with alfaxalone, a 3-h wash of brain slices in alfaxalone-free saline, with saline exchanged every 30 min, was required to remove any potentiating impact of alfaxalone on GABA A -R whole-cell currents.These results demonstrate that alfaxalone is an effective anesthetic for same-day electrophysiological experiments with goldfish brain slices.
In the early 1970s, the anesthetic alfaxalone (3ahydroxy-5a-pregnane-11,20-dione) was first utilized in humans under the name Althesin, and for veterinary use under the name Saffan [1].The original formulation for human use-Althesin, was later discontinued for anesthetic use in the mid-1970s due to multiple reports of anaphylactic reactions in response to one of the components of the drug formulation [2].In more recent years, a newly formulated neuroactive anesthetic compound-alfaxalone-has been approved for use in Canada.
Alfaxalone is a neuroactive steroid that binds to the caminobutyric acid (GABA) activated GABA A receptor (GABA A -R) [3] and the anesthetic acts as a potent coagonist, potentiating the channel opening effect of GABA through prolonging the mean open time of the GABA A -R [4][5][6][7].Alfaxalone can induce rapid anesthesia induction, followed by a rapid recovery period.This is supported by the low partition coefficient value, indicating a relatively low lipophilicity, and therefore an increased likelihood to diffuse out of the cellular membrane [8].Further, the anesthetic induces minimal, cardio-respiratory depression, especially at lower concentrations; however, these effects have been shown to be dose dependent [9,10].Rapid behavioral responses to alfaxalone are associated with low concentrations required to induce sedation, alongside favorable pharmacokinetics, indicating a rapid half-life, and therefore a rapid washout period in mammals [11][12][13][14][15].
Alfaxalone has also been shown to successfully induce effective anesthesia in other animal models including fish [16][17][18][19].However, the analysis of alfaxalone elimination and half-life times in fish are limited, especially when considering brain tissue.Alfaxalone residue clearance in rainbow trout liver, kidney, heart, spleen, and muscle: however, has been quantified to support a general timeframe of alfaxalone elimination from fish tissue.Following sedation induced with 5 mgÁL À1 and maintained with 2 mgÁL À1 alfaxalone, an 84-93% decrease in alfaxalone occurred within the first hour of measurements.Within 2 h of measurements, a reduction of up to 99% of alfaxalone was determined, further illustrating the rapid washout capabilities of the anesthetic [20].
Fish welfare during non-invasive procedures such as transport and handling, and experimentation involving invasive techniques often require light or full anesthesia [21,22].In addition, fish currently represent the second largest group of animals used for research, according to the Canadian Council for Animal Care [23].The rapid induction and washout of alfaxalone could, therefore, play an important role in improving a variety of invasive fish studies including neuroendocrine, energetics, and anoxia tolerance studies in model organism fish species.A fish species of particular interest in anoxia tolerance studies is one of the most hypoxia-tolerant vertebrate species, the crucian carp (Carassius carassius).Oxygen consumption in this species of fish is able to be maintained at normal levels as low as 5% oxygen air saturation [24], and it can survive complete anoxia for months at temperatures close to 0 °C [25,26].Closely related to the crucian carp, is the common goldfish (Carassius auratus) which displays similar anoxia tolerance characteristics.The common goldfish has a half-lethal time of 45 h under anoxic conditions at 5 °C and 22 h at 20 °C and therefore, while not as extremely hypoxia tolerant as the crucian carp, is an inexpensive, abundant, and convenient model organism to help better understand anoxia tolerance mechanisms [25,27].
In contrast to the extremely anoxia-tolerant common goldfish, if the human brain is deprived of an adequate source of oxygen for only 5 min, neuronal damage and cell death results [28].To limit the deleterious excitotoxic cell death which occurs alongside the mammalian response to hypoxia, mechanisms have been identified in anoxia-tolerant animal models including the common goldfish.The first anoxia-tolerant mechanism to be proposed was the metabolic arrest hypothesis [29,30].
Direct calorimetry experiments demonstrated a 70% reduction in heat production, and therefore metabolic rate, during the transition from normoxia to anoxia in the common goldfish [31,32].The "ion channel arrest" hypothesis was then proposed as an anoxia tolerance response in turtle brain by Hochachka [33].Blocking glutamatergic channel activity has been shown to be neuroprotective in mammal hippocampal pyramidal neurons during ischemia [34].The common goldish experiences a 40-50% decrease in NMDA activity following 40 min of acute anoxia [35].Contradictory to glutamatergic channel arrest, the identification of a five fold increase in the concentration of the inhibitory neurotransmitter, GABA, would then form the basis for the "spike arrest" (SA) hypothesis [36].GABA perfusion resulted in the depolarization of pyramidal neuron membrane potential to the GABA reversal potential (E GABA ) through increased presynaptic GABA B -R activity and postsynaptic GABA A -R channel activity, alongside a decrease in action potential (AP) or "spike" activity [37].The most recent hypothesis, the "synaptic arrest" hypothesis, combines the increase in inhibitory channel currents and decrease in excitatory channel currents, central to the two previous theories, into one overarching mechanism [38].
In order to further the understanding of anoxiatolerant mechanisms utilized by the common goldfish brain, decapitation followed by whole-brain removal is required to allow for the telencephalon slice preparation used for electrophysiological analysis.However, studies have identified the presence of nociceptors in bony fish, including the common goldfish, similar in both function and structure to mammal pain receptors.It is, therefore, possible that bony fish experience pain during decapitation [39][40][41].While there is still a large debate regarding whether fish experience pain in a manner similar to mammals, fish display nocifensive responses following the application of noxious stimuli, therefore any means to reduce these responses would be beneficial [42][43][44].In addition, goldfish are anoxia tolerant and therefore are likely alert for minutes after decapitation; therefore, it is imperative that a predecapitation anesthetic be found to minimize any possibility for pain and maximize animal well-being for electrophysiological experimentation.Anesthetic use presents a method of blocking pain perception, commonly used alongside surgical procedures in fish, as long as it doesn't confound electrophysiological experiments [44].
The two most common anesthetic compounds for use in fish include tricaine methanesulphonate (MS-222) and benzocaine.Following uptake through the gills, both compounds function similarly, through impeding the propagation of action potentials toward the central nervous system.The compounds induce anesthetic effects through the blockage of voltagegated sodium channels [41,45].Furthermore, MS-222 has a half-life of 1.5-4 h in the blood, and remains detectable up to 8 h following administration [46], while benzocaine has an elimination time of up to 25 h [47].As a result, the use of both compounds can confound electrophysiological measurements [39,41].
Additional anesthetics include the structurally similar compounds isoeugenol and eugenol or clove oil.Both compounds impede the function of multiple channels involved in the proposed anoxia tolerance pathways.These channels include the NMDA receptor and GABA A -R, which are both key in the synaptic arrest hypothesis, alongside additional sodium, potassium, and calcium channels [41,[48][49][50][51]. Furthermore, eugenol has a half-life of 12.14 h in the blood [52].As a result, high concentrations of either compound introduce a long-term confounding impact on electrophysiological recordings [41,53].
Given that alfaxalone prolongs GABA A -R-mediated inhibitory currents, the anesthetic represents a potential confounding factor for electrophysiological recordings.Considering that the GABA A -R plays an integral role in the synaptic arrest model of anoxia tolerance in the common goldfish, determining a sufficiently short washout period is essential to support its use in brain anoxia tolerance research.
The goal of our study is to determine if alfaxalone sedation will have a sufficiently short washout period from telencephalic brain slices to make same-day electrophysiological investigations in goldfish brain practical.Since it is a GABA A -R agonist it should have no long-term impact on passive and active action potential or GABA A -R current electrophysiological properties.Furthermore, alfaxalone sedation prior to decapitation will substantially improve animal welfare when anoxia-tolerant species are used for electrophysiological experiments.

Animal care and cortical sheet preparation
This study was approved by the University of Toronto Animal Care Committee and conforms to the care and handling of animals as outlined in the Canadian Council on Animal Care's Guide to the Care and Use of Experimental Animals, Vol. 2. Animal use protocol number (20012745).Common goldfish, Carassius auratus, weighing 50-150 g, were held in flowing dechlorinated City of Toronto tap water at 18 °C at the University of Toronto.
Whole-animal alfaxalone exposure (alfaxalone sedated goldfish) was achieved through immersing the goldfish in 500 mL of water containing 0.5 mL of 10 mgÁmL À1 alfaxalone, resulting in a final concentration of 0.01 mgÁmL À1 (AlfaxanÒ Multidose, Jurox Pty.Ltd, Rutherford, NSW, Australia).Once deep anesthesia was achieved, as assessed through the loss of the righting reflex, the animal was decapitated, and the whole brain was removed from the cranium and placed in alfaxalone-free oxygenated ice-cold 4 °C artificial cerebrospinal fluid (aCSF) containing (in mmolÁL À1 ): 20 NaHCO 3 ,118 NaCl, 1.2 MgCl 2 Á6H 2 O, 1.2 KH 2 PO 4 , 1.9 KCl, 10 HEPES-Na, 10 D-glucose and 2.4 CaCl 2 (pH 7.6, adjusted with HCl); osmolarity 285-290 mOsm.Both lobes of telencephalon were then dissected from the rest of the brain in the chilled aCSF.Na€ ıve goldfish tissue was defined as tissue harvested from nonalfaxalone exposed goldfish and alfaxalone sedated goldfish tissue was used to define tissue harvested from wholeanimal alfaxalone sedated goldfish as described above.Decapitation and whole brain removal remained the same for both groups.Each telencephalon lobe was glued to a sectioning plate on a vibratome slicing chamber using cyanoacrylate glue (Krazy Glue) and was filled with alfaxalone-free 4 °C aCSF.Isolated goldfish telencephalon was cut into 400-300 lm slices using a VT1200S vibratome.Slices were lifted out of the chamber and were stored in vials of alfaxalone-free aCSF for no longer than 48 h [54].Slice washing began following slice preparation.
For goldfish slices, cell-attached 1-20 GΩ seals were obtained using the blind-patch technique [55].To achieve a GΩ seal, the recording electrode was advanced toward the cell using a PCS-6000 motorized manipulator (Burleigh, Newton, NJ, USA) until the square-wave pulse was abruptly decreased, at which point a slight negative pressure was applied to form a seal.To break into the cell, a soft pulse of negative pressure was applied to break through the cell membrane, while potential was voltage-clamped to À55 mV.Once the whole-cell configuration was established, cells were given at least 2 min to acclimate to experimental conditions before access resistance was measured, which normally ranged from 20 to 30 MΩ. Patches were discarded if access resistance varied by > 25% over the course of an experiment.Data was collected at 5-10 kHz using a MultiClamp 700B digital amplifier, a CV-7B head stage, and a Digidata 1550B digitizer (Molecular Devices, Sunnyvale, CA, USA) and then collected and stored on computer using CLAMPEX 11.2 software (Molecular Devices).A liquid junction potential (LJP) was accounted for and is experimentally measured between the aCSF and the pipette solution and supported by LJP calculations using a generalized version of the Henderson equation (Clampex junction potential calculator; Molecular Device) [37,56].

Electrophysiological identification and measurement of action potential parameters
Pyramidal neurons were studied and characterized based on electrophysiological properties.In current clamp mode when current was injected, the more common pyramidal neurons exhibited spike frequency adaptation in response to sustained current, this is not seen in stellate neurons (Fig. 1A).The less abundant, rapidly firing stellate cell action potentials were demonstrated to have smooth rising and falling sections followed by a voltage undershoot, as depicted by Fig. 1B [57].
Action potential threshold (APth) was determined by current-clamping cells and injecting currents in 10 pA increments in a stepwise manner from sub-threshold for 500 ms until a spike was elicited.Threshold was recorded at the point at which a sharp elevation in voltage was observed.The full spike amplitude was measured from the point of the APth to the spike tip, while the half-amplitude spike width was measured as the time elapsed between the two-points of the half-amplitude on the spike.The rising time of a spike was calculated as the time elapsed between 10% of the full spike amplitude to 90% of the full spike amplitude, while the decay time was calculated as 90% of the full spike amplitude to 10% of the full spike amplitude.Whole-cell conductance was measured by voltageclamping cells and generating a voltage-ramp, through increasing voltage from À120 to À50 mV for 150 ms.Data measurements for all parameters were made with CLAMPFIT software (Axon Instrument) [56].

Impact of acute alfaxalone application on evoked naive tissue GABA A -R current
To initiate a GABA A -R current, the same methodology for whole-cell recordings was used with na€ ıve goldfish telencephalon slices, neurons were voltage clamped at a holding potential of À80 mV, and 2 mM GABA was applied for 1-2 s [54].These changes resulted in large outward GABA A -R currents that were easily detected and differentiated from other currents.From the GABA A -R currents produced, decay time was measured as the time elapsed between 90% and 10% of the peak amplitude, area under the curve was measured as the integrated area between the measured current and baseline, the peak amplitude was measured, and the baseline holding current was measured, using CLAMPFIT software (Axon Instrument).Electrophysiological properties were normalized to whole-cell capacitance.Tissue slices were then perfused with oxygenated aCSF and 1 lM alfaxalone for 15 min where the same protocol was repeated to measure GABA A -R current decay time, integrated area under the curve, peak amplitude, and baseline holding current. 1 lM alfaxalone was produced through adding 2 lL of stock alfaxalone solution into 60 mL of control aCSF.This procedure was repeated utilizing oxygenated aCSF without added alfaxalone (control solution) as a negative control and 100 lM picrotoxin, a general GABA A -R antagonist, as a positive control to elicit a shift in baseline holding current [58,59].Alfaxalone-free stock solution (vehicle, a generous gift from Jurox) had no significant impact when applied alone (data not shown) (n = 4).

GABA A -R reversal potential determination
To determine the E GABA , cells were perfused with 1 lM of the voltage-gated sodium channel inhibitor, tetrodotoxin (TTX), for approximately 5 min to eliminate the generation of action potentials.Current-voltage (I/V) curves were then constructed through voltage-clamping na€ ıve goldfish tissue pyramidal neurons and generating a voltage-ramp, through increasing voltage from À90 to À30 mV for 150 ms and measuring the corresponding current under control/baseline conditions and approximately 2-3 s following GABA perfusion.

Impact of whole-animal alfaxalone sedation on evoked GABA A -R current
Large outward GABA A -R currents were initiated using the same methodology as stated prior.GABA A -R current decay time, integrated area under the curve, and peak amplitude were determined using the same methodology as stated prior.Decay time, integrated area under the curve, peak amplitude, and baseline holding current were then normalized to whole-cell capacitance.The protocol was repeated with telencephalon slices obtained from alfaxalone sedated goldfish at 2, 3, and 4-6 h following alfaxalone application.The timepoints were binned so that any measurement performed between 1.5 and 2.5 h was considered 2 h following whole-animal alfaxalone sedation, any measurement performed between 2.5 and 3.5 h was considered 3 h following whole-animal alfaxalone sedation, and any measurement performed between 3.5 and 6.5 h was considered 4-6 h following whole-animal alfaxalone sedation.Measurements made on the 30-min mark were placed in the later time group.Following slice preparation, occurring 1.5 h after alfaxalone sedation, slices were washed every 30 min, with the aCSF housing the tissue slices being replaced with fresh alfaxalone-free aCSF.

Statistics
Welch's t-tests were conducted to compare AP electrophysiological properties between treatments.AP traces which did not immediately return to baseline or hyperpolarize were excluded as decay time could not be determined accurately.Paired t-tests were conducted to compare 0-min GABA A -R active and passive electrophysiological property values to 15-min perfusion values, within treatment groups.Welch's t-tests were conducted to compare absolute property values and the relative change in GABA A -R active electrophysiological property values following 15-min perfusion between treatment groups.A one-way ANOVA was used to compare the absolute holding current values and relative changes in holding current values following 15-min perfusion between treatment groups, followed by pairwise Welch's t-tests.Welch's t-tests were conducted to compare GABA A -R electrophysiological property values at various timepoints following whole-animal alfaxalone sedation, to na€ ıve/control values.N values represent a recording from a single cell from a single telencephalic slide, from a single goldfish.P < 0.05 was considered statistically significant.All statistical analyses were performed using R.

Action potential generation and structure analysis
To generate APs, 500 ms current steps in 10 pA increments were injected into cells, and voltage traces were obtained until APs were generated, at which point current injection was halted.Voltage traces containing one or multiple APs were then isolated for analysis.Current clamp recordings were conducted in naive/ control tissue and tissue derived following alfaxalone sedation, as shown in Fig. 1A,C respectively, where pyramidal cell voltage traces were characterized by low frequency APs.AP traces lacking accommodation were identified as stellate neurons (Fig. 1B) and were not included in the present analysis.Passive voltage traces were generated to monitor neuronal membrane potential changes over time.

Electrophysiological property analysis
To obtain AP electrophysiological characteristics from na€ ıve/control tissue and tissue derived following wholeanimal alfaxalone sedation, data were analyzed using the program Clampfit.Peak amplitude, half-width, rise time, decay time, membrane potential, action potential frequency and threshold potential were determined as summarized in Table 1.No statistical differences were found between control tissue and tissue derived following alfaxalone sedation for any of the electrophysiological properties, as demonstrated through Fig. 2A-E.
To determine whole-cell conductance values for the obtained alfaxalone and control AP traces, voltage clamp recordings were conducted through a voltage ramp protocol, increasing voltage from À120 mV to À50 mV, while measuring corresponding current.Voltage and current measurements were plotted against each other in a I/V curve, linear regression of the I/V curves was used to generate a line of best fit, and the slope of the linear regression line was used to determine whole-cell conductance values as summarized in Table 1.Demonstrated in Fig. 2F, no statistical differences were found between control and alfaxalone sedated goldfish tissue whole-cell conductance values.

GABA A -R current generation and characterization
To generate a whole-cell GABA A -R current trace, a GABA A -R current was elicited through clamping the cell voltage at À80 mV and perfusing 2 mM GABA onto the tissue slice for 1-2 s.This process was repeated after 15 min, and 30 min using the same whole-cell patch to generate additional GABA A -R currents and confirm the long-term stability of the wholecell patch, as shown in Fig. 3A.Whole-cell patch recordings remained stable up to 30 min, as indicated by Fig. S1.Whole-cell patch stability deceased beyond 30 min, however few recordings remained stable up to an hour.To confirm this was a GABA A -R current, the antagonist picrotoxin was perfused into the recording chamber 15 min before GABA was applied.Following perfusion of aCSF containing 100 lM picrotoxin for 15 min, large GABA A -R currents could not be generated (Fig. 3B), resulting in a significantly reduced GABA A -R current peak amplitude normalized to whole-cell capacitance (À3.27AE 0.489 pA per pF vs. À0.368AE 0.118 pA per pF, P = 0.00501, n = 5) (Fig. 3C).

GABA A reversal potential confirmation
To confirm the E GABA value, a similar voltage clamp protocol was utilized to clamp the cell at various voltages between À80 mV to 30 mV in the presence of TTX, while measuring cell current.Whole-cell GABA A -R currents were elicited through perfusion of 2 mM GABA for 1-2 s.GABA A -R currents with reversed directionality were recorded at potentials depolarized relative to E GABA as demonstrated by Fig. 4A.Peak amplitude increased approximately linearly relative to holding potential as indicated by Fig. 4B.
To further confirm that the obtained current traces resulted from GABA A channel activity, I/V curves were generated using a voltage ramp protocol in na€ ıve goldfish tissue.These recordings were taken passively at baseline and following the steady perfusion of 2 mM GABA for 1-2 s, at the peak amplitude of the GABA A -R whole-cell current.The baseline current values were then subtracted from the peak amplitude values to obtain a GABA-specific I/V curve, as demonstrated by Fig. 4C.The intersection between GABA-specific I/V curve and the x-axis represented the E GABA .E GABA was calculated to be À55.49AE 4.485 mV (n = 6), as indicated by Fig. 4D.

Analysis of acute alfaxalone application onto na€ ıve tissue
To confirm patch stability and determine the impact of alfaxalone on GABA A -R currents, GABA A -R currents were elicited in na€ ıve tissue as described above, as shown in Fig. 5A.While maintaining the same wholecell patch, the cell was then perfused with either control aCSF or aCSF containing 1 lM alfaxalone for 15 min, and the same protocol as mentioned prior was utilized to generate an additional GABA A -R trace as shown in Fig. 5A.An alfaxalone concentration of 1 lM was utilized based on the results of our doseresponse curve analysis, as indicated by Fig. S2.Current decay time, area under the curve, and peak amplitude at 0 min and following 15 min of control aCSF perfusion or the acute application of alfaxalone were determined using the program Clampfit and were normalized to whole-cell capacitance.
A statistically significant increase in decay time was found following 15 min of acute alfaxalone application, in comparison to 0-min recordings, as indicated by   To confirm the presence of tonic GABA A -R activity, holding current required to maintain a cellular membrane potential of À80 mV was measured under control/baseline conditions and following the application of the general GABA A -R antagonist, picrotoxin, in na€ ıve tissue as shown in Fig. 6A.Following the application of 100 lM picrotoxin, a statistically significant shift in holding current (0.218 pA per pF AE 0.0763pA per pF, n = 5, P = 0.0460) was found, as shown in Fig. 6B,C.
To determine any passive impact of alfaxalone on the holding current required to maintain a cellular membrane potential of À80 mV, the holding current was measured under control conditions and following acute application of 1 lM alfaxalone onto na€ ıve tissue for 15 min.As a positive control, holding current was measured under control conditions and following the acute application of 100 lM picrotoxin.Holding current was determined using the program Clampfit and was normalized to whole-cell capacitance.No statistically significant difference in holding potential was found between 0-and 15-min recordings under normal aCSF control conditions (À0.778 AE 0.166 pA per pF to À0.888 AE 0.198 pA per pF, P = 0.0599, n = 5) and following the acute application of alfaxalone (À1.41 AE 0.228 pA per pF to À1.56 AE 0.182 pA per pF, P = 0.530, n = 6), as indicated by Fig. 7A.A statistically significant shift in holding current (À1.42 AE 0.256 pA per pF to À1.14 AE 0.225 pA per pF, P = 0.0110, n = 6) was found between control and 100 lM picrotoxin application treatment groups as indicated by Fig. 7A.
Using a one-way ANOVA, no statistically significant difference in the absolute holding current values were found at 15 min between the three conditions (P = 0.0985, n = 5-6).Using a one-way ANOVA, a statistically significant difference in the change in holding current was found at 15 min between the three conditions (P = 0. 0.00530, n = 5-6).Following a pairwise analysis, a statistically significant increase in the change holding current was found between control aCSF conditions and picrotoxin treatment (À0.11AE 0.0424 pA per pF vs. 0.273 AE 0.0693 pA per pF, P = 0.00447, n = 5-6), as shown in Fig. 7B.Further, no statistically significant difference was found in the change in holding current between control aCSF conditions and alfaxalone treatment (À0.11AE 0.0424, P = 0.886, n = 5-6), or between alfaxalone treatment and picrotoxin treatment (P = 0.262, n = 6), as shown in Fig. 7B.

Alfaxalone washout time
To determine the washout period of alfaxalone, GABA A -R currents were measured as described above in naive/control tissue and tissue harvested from wholeanimal alfaxalone sedation at various time-points following anesthetic administration, as shown in Fig. 8A.Current decay time, area under the curve, and peak amplitude were determined using the program Clampfit and were normalized to whole-cell capacitance.As shown in Fig. 8B, a statistically significant increase (P = 0.00757) in decay time was found between control (531 AE 99.3 ms per pF, n = 12) conditions and 2 h following alfaxalone application (1570 AE 283 ms per pF, n = 8).This statistically significant increase in decay time was reversed at 3 h (757 AE 164 ms per pF, n = 7, P = 0.264) and 4-6 h following alfaxalone sedation (606 AE 99.1 ms per pF, n = 9, P = 0.597).
To determine whether alfaxalone had any long-term impact on GABA A -R mediated tonic currents, holding current was measured in control tissue and alfaxalonetreated tissue at various time-points following anesthetic administration.Holding currents were determined using the program Clampfit and were normalized to wholecell capacitance.As shown in Fig. 8E, no statistically significant difference in holding current was found between control (À2.

Alfaxalone has no long-term impact on action potential properties
In this study, the impact of whole-animal alfaxalone sedation on action potential electrophysiological characteristics was investigated using whole-cell patch   Similar rise time, decay time, and half-width time in both alfaxalone sedated and control goldfish tissue was expected, considering the mechanism in which the anesthetic induces sedation.While alfaxalone is defined as a GABA A agonist, the electrophysiological properties, rise time, decay time, and half-width time, are determined by the action of voltage-sensitive fast sodium currents and outward potassium currents [61].Alfaxalone has not been found to alter Na + and K + channel structure or conductance following its application, and therefore our results support the proposed mechanism regarding alfaxalone binding and action [3].

Prolonged GABA A -mediated currents and their antagonistic elimination
In order to analyze GABA A -R current electrophysiological properties, it was important to first confirm the long-term stability of whole-cell patch-clamp recordings following the perfusion of GABA multiple times over an extended period of time.The stability of whole-cell patch recordings up to 30 min was confirmed following the generation of two GABA A -R currents, at two timepoints separated by 15 min.These large GABA A -R currents were the result of extrasynaptic GABA A -R activity, as indicated by the length of the measured decay times.The average decay time of synaptic inhibitory postsynaptic currents in rat Fig. 8. GABA A -R potentiation induced through whole-animal alfaxalone sedation reversed at 3 h following application.(A) GABA A -R current following GABA application in control conditions, 2 h following alfaxalone application, and 3 h following alfaxalone application.GABA A -R current (B) decay time (C) area under the curve (D) peak amplitude and (E) holding current in control conditions (n = 12) and at 2 h (n = 8), 3 h (n = 7), and 4-6 h (n = 9), following whole-animal alfaxalone sedation.Each point represents data from a separate experiment.Welch's t-test indicated significant difference from control conditions (*P < 0.05, **P < 0.01).hippocampal neurons was found to be 41 AE 9 ms [7], in comparison to the GABA A -R currents produced in this study, which demonstrated a much larger average decay time of 531 AE 99.3 ms per pF in na€ ıve tissue.Stability was dependent on a holding potential of À80 mV, and the use of physiologically relevant concentrations of Cl À , rather than a high intracellularchloride solution to mediate large outward GABA A -R currents.The large negative cell holding potential allowed for visible outward chloride currents, while preventing the opening of voltage-dependent Na + and NMDA receptor channels [62].The resulting phasic GABA A -R currents followed the expected biphasic decay response caused by the fast and slow deactivation phases [63,64].Desensitization at the singlechannel level is described as the process where the GABA A channel is unable to open despite the presence of GABA.In this desensitized state, the GABA A -R can retain GABA, and if GABA remains bound to the receptor following this refractory period, the channel is able to immediately open once again [64][65][66].The fast components of GABA A -R desensitization correspond to receptors displaying a rapid refractory period, allowing these components to remain bound to GABA and to immediately reopen following the completion of this refractory period [67].As such, the fast components of GABA A -R desensitization can prolong whole-cell GABA A -R currents, causing the slow phase of deactivation [64,66].Additionally, few recordings depicted a GABA A -R current including a deactivation period which eventually reached a plateau, hyperpolarized relative to the baseline.The recorded current measurement would then approach baseline over a much longer period of time.Decay time was defined as the time between 90% and 10% of the peak amplitude to account for this occurrence; however, some recordings included a plateau that significantly deviated from the baseline.While these recordings were not included in the overall comparison, their occurrence can be explained through two possible mechanisms.GABA spillover into extracellular space causes the delayed activation of perisynaptic GABA A -Rs, thus resulting in a prolonged phasic GABA A -R response [64,68].Additionally, to accompany the phasic GABA response, slowly activated, lowamplitude, tonic receptors also play a role, especially in the extrasynaptic response to GABA.Explained by their limited desensitization, tonic receptors exhibit GABA responses that can last seconds to minutes, and also exhibit chronic activation [64,69].The presence of tonic GABA A -Rs was first identified through the use of GABA A antagonists bicuculline and gabazine, which both reduced the holding current required to maintain a desired membrane potential [70].The activation of either extra-or perisynaptic GABA receptors may have contributed to the prolonged return to baseline through the spillover of GABA into the extracellular space, thus representing a limitation to our study.

566
Alfaxalone selectively binds to GABA A -R channels, therefore it was also important to ensure that the GABA A -R currents were the result of GABA A -R activity, as opposed to one of the other three GABA-R subtypes, including GABA B and GABA C [71].Our study confirmed that the recorded GABA A -R currents were evoked through GABA A -mediated channel activity, using three separate methods of confirmation.First, by measuring whole cell conductance at baseline and immediately following the perfusion of GABA, the E GABA was identified through interpolating the point at which the two traces intersected.In na€ ıve goldfish tissue, an E GABA value of À55.5 AE 4.49 mV was calculated, similar to the previously calculated value of À53.80 AE 2.63 mV, thus providing support that the currents were likely the result of GABA A -R channel activity [54].Second, this reversal potential was then confirmed, as current directionality reversed as expected after passing through E GABA and producing a GABA A -R current at holding potentials depolarized relative to E GABA .As a third and final method of confirmation, the negative control experiment was performed by eliminating the GABA A -R-mediated response to GABA application with the agonist picrotoxin.Perfusion of GABA A -R antagonist, gabazine in turtle cortical sheets was previously found to completely eliminate the GABA A -R-mediated response to GABA perfusion in a similar manner [37].
Beyond the classification of the GABA-R channel subtype, the GABA A -R can be further broken town into individual subtypes, composed of different units.Within the human genome 19 GABA A subunits have been identified, divided into classes (a, b, c, q, h, e, p and d) and their respective isoforms (a1-6, b1-3, c1-3 and q1-3) [72].Utilizing these 19 subunits, upwards of 500 subtypes of functional GABA A -R channels are estimated to exist [64,73].Furthermore, cells can co-express multiple GABA A -R subunit isoforms, and therefore subtypes, dependent on cellular location [74].The future identification and analysis of prominent GABA A subtypes which are expressed in the common goldfish pyramidal neurons at the single channel level would allow for a further understanding of the composition of whole-cell GABA A -R currents.Furthermore, a better understanding of prominent subunits expressed within goldfish pyramidal neurons would help explain the whole-cell response to alfaxalone, as the anesthetic produces variable allosteric modification dependent effects on the GABA A subunits present.Alfaxalone induces the most effective modulation on the a1b1c2L GABA A -R isoform [75].

Alfaxalone has a short-term impact on GABA A -R current properties
The short-term alfaxalone-induced GABA A -R current potentiation was explored through acutely perfusing alfaxalone onto na€ ıve tissue and comparing values at 0 min, or baseline values, to 15 min of anesthetic application.Alfaxalone had a significant impact on increasing GABA A -R current decay times and area under the curve relative to 0-min values, however no significant change in peak amplitude was found.These results were consistent when comparing the change in decay time and area under the curve at 15 min following acute alfaxalone application in comparison to 15 min of control aCSF perfusion.However, when considering absolute property values at 15 min of acute alfaxalone or control aCSF perfusion, only a significant increase in decay time was found.In addition, alfaxalone-free stock aCSF solution induced no significant impact.Consistent with our results, the alfaxalone solvent, 2hydroxypropyl-b-cyclodextrin, has been shown to induce no significant impact when applied alone, following dilution by a factor of 1000 [76].Stock alfaxalone was diluted by a factor of at least 20 000 in our study.Overall, our findings confirm that alfaxalone induces short-term potentiation of the GABA A -R channel, supporting previous studies which identified that the same concentration of acute alfaxalone treatment (1 lM) prolonged the mean open-time of GABA A -R channels [6,77].Studies in cultured rat hippocampal neurons have shown a similar increase in decay time for spontaneous and induced GABA-mediated inhibitory postsynaptic currents [7,78].Furthermore, the study by Harrison et al. [78] presented similar results when comparing decay time, peak amplitude, and area under the curve before and during alfaxalone application.Paralleling our findings, an increase in inhibitory post synaptic current decay time and charge transfer (area under the curve) was found during acute alfaxalone application and peak amplitude either remained the same or slightly decreased.
Additionally, no significant change in holding current was seen following the perfusion of 1 lM alfaxalone for 15 min onto na€ ıve tissue, indicating that alfaxalone did not potentiate tonic GABA A -R currents.As alfaxalone induces differing modulation depending on the GABA A subunits present [75] this result could be explained by the subunit composition of extracellular tonic GABA A receptors, which differ from phasic GABA A receptors [64,79].Alternatively, it could be explained by the long duration of tonic receptor modulation, which occurs over the span of seconds to minutes.The subtypes involved in tonic GABA A receptors have a high affinity for GABA and incomplete desensitization, leading to steady, low-level activation.The increased GABA A -R open channel probability and stabilization of the open state induced by alfaxalone may have a limited effect on tonic receptors which maintain a steady open state [64,80].
In contrast to alfaxalone application, a statistically significant increase in holding current was seen following the perfusion of 100 lM picrotoxin, a noncompetitive synaptic, peri-synaptic and extra-synaptic tonic GABA A -R channel blocker [58].This result is consistent with studies in the western painted turtle, which demonstrated a similar shift in holding current following the application of the general GABA A -R channel blocker bicuculline methiodide [81].Bicuculline application is known to induce a similar shift in holding current as picrotoxin [59].Furthermore, the significant increase in holding current following picrotoxin perfusion supports the presence of extrasynaptic tonic GABA A -R activity in the common goldfish pyramidal neurons.
Alfaxalone potentiated decay time and area under the curve but the effect saturated at concentrations of approximately 1 lM or greater, as indicated by Fig. S2.Additionally, peak amplitude demonstrated a poor fit to the Hill equation, with minimal potentiation at any alfaxalone concentration, which supports our finding that acute alfaxalone application has no significant impact on potentiating GABA A -R current peak amplitude at concentrations between 0.1-1.5 lM.Successful whole-cell patch clamp recordings at a 1 lM alfaxalone concentration falls within the range of concentrations (0.1-1 lM) used in previous experiments in cultured rat hippocampal neurons where a similar increase in GABA A -R decay times and charge transfer (area under the curve) was found [78].Analysis of electrophysiological properties following the acute application of alfaxalone onto na€ ıve tissue was limited at higher anesthetic concentrations.Alfaxalone concentrations greater than 1.5 lM destabilized whole-cell patch recordings beyond approximately 10 min of acute perfusion.Alfaxalone increases liposome membrane fluidity in a concentration-dependent manner, with membrane fluidity showing a positive correlation with anesthetic concentration [82].This disruption to membrane stability may explain patch destabilization at greater alfaxalone concentrations and could set the upper limits of the dose-response curve.Furthermore, previous studies analyzing lengthened inhibitory post synaptic current decay times in the presence of high concentrations (10 lM) of alfaxalone used paired recordings at much shorter intervals (1 min) than our study [78].Stable whole cell patches following the application of acute alfaxalone treatment at concentrations greater than 1 lM were possible in our study at timepoints less than 10 min.However, in our study we used 15 min between GABA A -R current recordings to ensure washout of GABA from the bath and for alfaxalone to induce its potentiating effect [7,58].
Alfaxalone has no long-term impact on GABA A -R current properties The long-term impact of whole-animal alfaxalone sedation on GABA A -R current electrophysiological properties was explored in brain tissue from animals sedated with alfaxalone and then following a series of in vitro tissue washes every 30 min.Our results indicate that 3 h after whole-animal alfaxalone sedation, was sufficient to reverse any significant increase in GABA A -R current decay time, peak amplitude, and area under the curve, induced by alfaxalone.Additionally, no significant difference in holding current values was found between na€ ıve tissue and all timepoints following whole-animal alfaxalone sedation.In the western painted turtle, unpublished results by Suganthan et al. showed a similar, rapid washout of alfaxalone following whole-animal sedation.Replacing the aCSF housing the tissue slices every 30 min for 3 h was sufficient to reverse the impact of whole-animal alfaxalone sedation on GABA A -R current electrophysiological properties.Additionally, these results further demonstrate that both the acute application of alfaxalone onto na€ ıve tissue and whole-animal alfaxalone sedation result in no significant impact on extra-synaptic GABA A -R mediated tonic-currents.The potentiation of GABA A -R currents in tissue from whole-animal alfaxalone sedation 2 h was similar to values following the acute application of alfaxalone onto na€ ıve tissue.GABA A -R current decay time and area under the curve both increased but peak amplitude was found to increase only following alfaxalone sedation.These results, in combination, support previous research which identified that alfaxalone potentiates GABA-induced activation of the GABA A -R channel [6,77,83,84].
Overall, our findings suggest that alfaxalone is an anesthetic compatible with electrophysiological experimentation in brain tissue following 3 h of in vitro washing after whole-animal alfaxalone sedation.While our results support the reversal of the potentiated GABA-induced Cl À current response, a limitation of our study involves the analysis of any long-term impact on general cell structure or function induced by whole-animal alfaxalone sedation.As stated previously, alfaxalone induces structural changes upon allosteric binding to the GABA A -R [3] and increases membrane fluidity at high extracellular concentrations [82].It is therefore possible that the anesthetic may alter cellular properties in ways which remain undetected utilizing our current GABA A -R-focused methodology.Furthermore, alfaxalone is a synthetic analogue to the compound -Allopregnanolone, which contributes to increased neurogenesis, neuroplasticity, and neuroprotection [85,86].Both compounds are capable of activating human pregnane X receptors (h-PXR) in the brain, and alfaxalone anesthesia induces similar effects as Allopregnanolone, including the secretion of mature brain-derived neurotrophic factor and subsequently inducing neuroprotection [87].PXR expression is highly conserved and plays an important role in head kidney-mediated detoxification within rainbow trout, therefore it is possible that PXR is expressed in the goldfish brain [88,89].It is unknown whether PXR expression is significant in goldfish pyramidal neurons, and whether its activation in any region of the brain by alfaxalone could confound electrophysiological recordings.Future studies should focus on identifying the role of PXR in the goldfish brain, and whether its activation could cause alternative long-or short-term impacts on electrophysiological recordings.

Conclusion
Alfaxalone has been described as an anesthetic with a fast induction and fast washout in many animal models including fish.In this study, our results demonstrate that whole-animal alfaxalone sedation has no long-term impact on action potential properties or GABA A -receptor currents.While whole-animal alfaxalone sedation had no significant impact on action potential properties, a 3 h washout was required to reverse the alfaxalone-mediated potentiation of GABA A -R channel currents.Our results also confirm that alfaxalone induces short-term alterations of GABA A -R current electrophysiological properties, specifically at 2 h following whole-animal alfaxalone sedation and following short-term acute alfaxalone perfusion onto naive tissue.This study suggests that the short washout period from telencephalic brain slices associated with alfaxalone sedation, will make same-day electrophysiological investigations in goldfish brain practical.The use of an anesthetic alongside future goldfish electrophysiological experimentation will introduce a significant improvement to fish wellbeing.Furthermore, these results indicate that alfaxalone could represent an anesthetic compatible with electrophysiological studies in all fish species.Further experimentation in additional animal models will help confirm the generalizability of the rapid anesthetic washout following whole-animal alfaxalone sedation.

Fig. 1 .
Fig. 1.Induced action potential recordings.AP traces generated from 500 ms current steps in 10 pA increments from a na€ ıve tissue (A) pyramidal neuron and (B) stellate neuron, and (C) a pyramidal neuron from goldfish tissue following whole-animal alfaxalone sedation.
Fig. 5B (795 AE 176 ms per pF to 1120 AE 194 ms per pF, P = 0.0175, n = 6), representing a 51.8 AE 16.9% change.A statistically significant difference in area under the curve was found following 15 min of acute alfaxalone application, in comparison to 0-min recordings, as indicated by Fig. 5C (À28 AE 5.38 nA*ms per pF to À35.4 AE 5.62 nA*ms per pF, P = 0.0221, n = 6), representing a 34.6 AE 11.4% change.No statistically significant (À2.5 AE 0.50 pA per pF to À2.65 AE 0.502 pA per pF, P = 0.569 n = 6) difference in peak amplitude was found following 15 min of acute alfaxalone application, in comparison to 0-min recordings as shown in Fig. 5D.A statistically significant increase in absolute decay time values was found between 15 min of control aCSF perfusion and 15 min of acute alfaxalone perfusion, as shown in Fig. 5B (540 AE 126 ms per pF vs. 1120 AE 194 ms per pF, P = 0.0360 n = 5-6).No statistically significant difference in the absolute area under the curve (À28.7 AE 4.83 nA*ms per pF vs. À35.4AE 5.62 nA*ms per pF, P = 0.386 n = 5-6) or peak amplitude (À1.83 AE 0.334 pA per pF vs. À2.65 AE 0.502 pA per pF, P = 0.211 n = 5-6) values was found between 15 min of control aCSF perfusion and 15 min of acute alfaxalone perfusion, as shown in Fig. 5C,D respectively.A statistically significant difference in the change in decay time was found between 15 min of control aCSF perfusion and 15 min of acute alfaxalone perfusion, as shown in Fig. 5E (3.58 AE 6.98 ms per pF vs. 324 AE 92.8 ms per pF, P = 0.0181 n = 5-6).Further, a statistically significant difference in the change in area under the curve was found between 15 min of control aCSF perfusion and 15 min of acute alfaxalone perfusion, as shown in Fig. 5F (À0.661AE 1.89 nA*ms per pF vs. À7.44AE 2.27 nA*ms per pF, P = 0.0475 n = 5-6).No statistically significant difference in the change in peak amplitude following 15 min of perfusion was found between control and acute alfaxalone conditions, as shown in Fig. 5G (À0.223AE 0.152 pA per pF vs. À0.148AE 0.243 pA per pF, P = 0.853 n = 5-6).

Fig. 3 .
Fig. 3. Whole-cell GABA A -R current characterization and stability.(A) Whole-cell GABA A -R currents taken 15 min apart.(B) GABA A -R current trace following GABA application in control conditions, following 15 min of picrotoxin perfusion, and after 15 of aCSF perfusion recovery.(C) GABA A -R current peak amplitude following GABA perfusion under control conditions and following 15 min of picrotoxin perfusion.Picrotoxin perfusion began immediately following the completion of the initial control GABA A -R current.Each point represents data from a separate experiment (n = 5).Paired t-test indicated significant difference from baseline/control conditions ( ## P < 0.01).

Fig. 4 .
Fig. 4. GABA A -R reversal potential identification through generation of wholecell GABA A -R mediated currents.(A) GABA A -R currents at various holding potentials between À80 mV and 30 mV, demonstrating reverse directionality at potentials above the GABA A channel reversal potential.(B) Whole-cell currentvoltage relationship depicting GABA A -R current peak amplitude relative to holding potential.Peak amplitude increased linearly with holding potential.The xintercept represents E GABA .Values are means AE SEM (n = 5).(C) GABA-specific whole-cell current-voltage relationships taken in na€ ıve goldfish tissue following exposure to 2 mM GABA (n = 6).(D) Average E GABA value calculated through the x-intercepts of GABA-specific wholecell current-voltage relationships.E GABA was calculated as mean AE SEM (n = 6).
recordings.It was shown that whole-animal alfaxalone sedation has no significant impact on any of the evoked action potential properties analyzed in this study, including rise time, decay time, half-width time, peak amplitude, threshold potential, alongside membrane potential and whole cell conductance.Previous average membrane potential and whole-cell conductance values for the common goldfish were found to 564 FEBS Open Bio 14 (2024) 555-573 ª 2024 The Authors.FEBS Open Bio published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies.

Fig. 5 .
Fig. 5. GABA A -R potentiation following acute application of alfaxalone onto naive tissue.(A) GABA A -R currents following GABA application at 0 min under control aCSF conditions and following 15 min of acute alfaxalone perfusion.Absolute (B) decay time, (C) area under the curve, and (D) peak amplitude at 0 min under control aCSF conditions and following 15 min of constant control aCSF (Con) or 1 lM alfaxalone (Alfax) perfusion onto na€ ıve tissue.Change in (E) decay time, (F) area under the curve, and (G) peak amplitude at baseline and following 15 min of constant control aCSF or 1 lM alfaxalone perfusion onto na€ ıve tissue.Each point represents data from a separate experiment (n = 5-6).Note negative y axes in (C), (D), (F), and (G) indicate increasing currents.Paired t-tests indicated a significant difference from 0min recordings ( # P < 0.05).Welch's t-tests indicated a significant difference in absolute property values or the change in property values between 15-min control aCSF and acute alfaxalone recordings (*P < 0.05).

Fig. 6 .
Fig. 6.Picrotoxin induces a positive shift in holding current following application onto na€ ıve tissue.(A) Positive shift in holding current following the application of general GABA A -R antagonist picrotoxin at a holding potential of À80 mV.Absolute change (B) and difference (C) in holding current following the application of 1 lM alfaxalone and 100 lM picrotoxin.Each point represents data from a separate experiment (n = 5).Paired t-test indicated significant difference from baseline holding current ( # P < 0.05).

Fig. 7 .
Fig. 7. Alfaxalone induces no significant change in holding current following application onto na€ ıve tissue.(A) Absolute and (B) change in holding current following the application of control aCSF (Con), 1 lM alfaxalone (Alfax), and 100 lM picrotoxin (Ptx).Each point represents data from a separate experiment (n = 5-6).Paired ttests indicated a significant difference from 0-min recordings ( # P < 0.05).Pairwise Welch's t-tests indicated a significant difference in absolute holding current or change in holding current between 15-min recordings (**P < 0.01).

FEBS
Open Bio 14 (2024) 555-573 ª 2024 The Authors.FEBS Open Bio published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies.

568FEBS
Open Bio 14 (2024) 555-573 ª 2024 The Authors.FEBS Open Bio published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies.

Table 1 .
Control and alfaxalone treated tissue AP electrophysiological property summary values.Data are mean AE SEM (n = 9-20).