No impairment of contextual fear memory consolidation by oxytocin receptor antagonism in male rats

Oxytocin is a peptide released into brain regions associated with the processing of aversive memory and threat responses. Given the expression of oxytocin receptors across this vigilance surveillance system of the brain, we investigated whether pharmacological antagonism of the receptor would impact contextual aversive conditioning and memory. Adult male rats were conditioned to form an aversive contextual memory. The effects of peripheral administration of either the competitive antagonist Atosiban or noncompetitive antagonist L-368,899 were compared to saline controls. Oxytocin receptor antagonism treatment did not significantly impact the consolidation of aversive contextual memory in any of the groups. We conclude that peripheral antagonism of oxytocin signalling did not impact the formation of aversive memory.


Introduction
Memory consolidation is the process where a memory is moved from a labile to a stable state [1].Consolidation is generally addressed at two levels: the synaptic or cellular level, and the systems level [2].The time course of the window of cellular consolidation has been revealed through interventions that result in a reduction in the conditioned response, interpreted as a retrograde amnesia (for review, see [1,3].Therefore, it has been proposed that memories be viewed as active, labile, and subject to disruption or inactive and stable [4].
Oxytocin is most commonly known for its role in maternal and paternal behaviours [5] however there is mounting evidence that it plays a role in both social and non-social behaviours.These include behaviours learned as an individual, as well as in a social context [6][7][8].
In the last decade, oxytocin signalling has emerged as a prominent pharmacological target in the context of anxiety disorders [9][10][11].The control of conditioned fear responses by enhancing oxytocin signalling has been demonstrated to depend on the location of administration.In rats, pre-conditioning administration of oxytocin receptor (OXTR) specific agonists increased freezing behaviour to a conditioned context if administered to the basolateral amygdala, but inhibited conditioned freezing if given into the central nucleus of the amygdala [12].Also in male rats, it was shown that peripheral administration of a competitive oxytocin receptor antagonist Atosiban (d[D-Tyr(Et)2,Thr4]OVT, ATO) impaired contextual fear memory consolidation in a dose-dependent manner [13].Administration of ATO after training specifically impaired long-term memory (LTM), but not short-term memory (STM) nor memory reconsolidation [13].In contrast, another OXTR antagonist L-368,899 only affected socially transmitted fear, or so-called 'emotional contagion' in a dose dependant manner, but left conditioned fear potentiated startle intact in mice [14].
In order to build on these findings, the following set of studies were carried out.In one experiment, a range of doses of ATO were administered following contextual fear conditioning [13].In another, the non-competitive oxytocin receptor antagonist, L-368,899 was administered before contextual fear conditioning.ATO is a competitive OXTR antagonist that can also act on arginine vasopressin receptors (VAR), whereas L-368,899 is a non-competitive selective OXTR antagonist, with better binding affinity for the OXTR versus the related VARs [15,16].As the selectivity of oxytocin receptor antagonists used across studies of aversive conditioning has varied it is important to compare them for consistency and to establish whether antagonising endogenous OXTR signalling can prevent consolidation of contextual fear memories.

Materials and methods
Subjects.Subjects were 64 adult male Lister-Hooded rats (Charles River and Envigo, UK) weighing approximately 250-300g at the start of experiments and were housed in groups in conventional cages (NKP) in a room maintained at 20 -24 • C, 45 -65% relative humidity, 15 -20 air changes per hour with an artificial light cycle of 60 lux between 0800 and 2000h.Cages were provided with lignocel bedding, and standardised enrichment (chewing blocks and tunnels).All animals had access to ad libitum food (Lab Diet 5053) and water except for during behavioural procedures, which were conducted during the rats' dark cycle.The rats were randomly assigned to each group and experimenters were blind to group allocation for subsequent analysis.Subjects were divided into Cohort 1 (n = 40) and Cohort 2 (n =24); the experiment with Cohort 1 was performed at the facilities of the University of Cambridge, while the experiment with Cohort 2 was performed at those of the University of Bristol.The research on Experiment 1 was conducted under Project Licence PA9FBFA9F and Experiment 2 under Project Licence PP5432050, which have been regulated under the Animals (Scientific Procedures) Act 1986 Amendment Regulations 2012 following ethical review by the respective Animal Welfare and Ethical Review Body (AWERB).

Behavioural procedures 2.1.1. Contextual fear conditioning
Experiment 1 Rats were housed in groups of four.40 rats were individually placed in the conditioning contexts as reported previously (MedAssociates, USA, as per [18]), which they had not previously been habituated to.Two minutes after being placed into the context, rats received three-foot shocks at 0.5mA for 0.5 s with two minutes interval between each shock.After the last foot shock, they remained in the conditioning box for a further two minutes, resulting in a ten-minute total conditioning session.The rats were then removed from the conditioning boxes and immediately injected with Atosiban (ATO) or Saline, and then returned to their home cage.Three hours after contextual fear conditioning, rats were returned to the conditioning chambers for 90 s as a STM test.48 hours after contextual fear conditioning, rats were placed back into the conditioning context for a long-term memory test.
Experiment 2 Rats were housed in groups of three.24 rats were injected with L-368,899 or saline one hour prior to contextual fear conditioning.During contextual fear conditioning, rats were individually placed in the conditioning boxes, as above, which they had not previously been habituated to.Conditioning intervals and shock parameters were exactly as above for Experiment 1.Three hours after contextual fear conditioning, rats were placed back into the conditioning chambers for a STM test, which lasted 90s in duration.24 hours after contextual fear conditioning, rats were placed back into the conditioning chambers for a LTM test.

Drug administration
Atosiban (ATO, A3480, Merck) was made into a 1mg/mL stock solution, dissolved in Saline, and i.p. injected at 2mL/kg at either: 0.001μg/mL, 0.01μg/mL, 0.1μg/mL, or 1μg/mL directly after conditioning.L-368,899 (Tocris Bioscience) was made into a 1mg/mL stock solution dissolved in saline and i.p. injected at 5mg/1ml/kg 1 hour before conditioning [19].L-368,899, a non-competitive antagonist, was administered before contextual fear conditioning in order to inhibit activation of the receptor during the conditioning session and investigate any effect of OXT-R antagonism on acquisition of conditioned responding.On the other hand, ATO, a competitive antagonist, was administered immediately after fear conditioning in order to block consolidation [13,17].
Ultrasonic Vocalisation (USV) Analysis.An Ultrasound Microphone (Metris, Netherlands for Experiment 1, Avisoft Bioacoustics for Experiment 2) was placed through above hole in the middle of the operant box roof about 30 cm above the context floor.The microphones were sensitive to frequencies of 15-125 kHz.Vocalization was recorded and analysed using the Sonotrack (Experiment 1) or Avisoft Bioacoustics software (Experiment 2).For Experiment 2, due to equipment failure during conditioning the calls of two rats failed to be recorded.These were excluded from analysis from that session.Call detection was provided by an automatic threshold-based algorithm filtered by frequency range.The number of USV calls detected and total calling time were analysed for the 22kHz range calls only.Behavioural Scoring.Every session was recorded using CCTV cameras (Spy Camera model CMOSNC76).Behavioural scoring for freezing was measured throughout the sessions.Freezing was defined as immobility with a characteristic vigilant posture [20].The number of seconds the rat spent freezing was scored and the percentage time the rat spent freezing was calculated for each phase of the test.
Statistical analysis.Statistical analysis and graphing was performed in GraphPad Prism (version 10.1.0,San Diego, CA, USA).Data is presented at mean ± standard error of the mean (SEM) throughout and ⍺ was 0.05 for all tests such that results of p < 0.05 were taken as statistically significant.Power analyses were performed for sample size using G*power version 3.1 (⍺=0.05;power set to 0.8).The percentage of time spent freezing in each experiment was analysed for significance using repeated measures two-way analysis of variance (ANOVA) followed by Tukey's multiple comparisons test, or one-way ANOVA followed by Tukey's multiple comparison test or Dunnett's multiple comparisons test where required.Fisher's exact tests or Chi-squared tests were used to determine if there were significant differences in the proportions of rats categorised as 'vocalisers' or not per group.

Atosiban did not produce any significant effects on ultrasonic vocalisation at test
USVs were measured during conditioning, the STM test, and the LTM test.All rats from each group produced 22kHz USV 'alarm calls' during contextual fear conditioning.There were no significant differences in the number of vocalisers from each group during the STM test (χ2 (4, n =40) = 4.267, p =0.3711).Additionally, there were no significant differences in the number of vocalisers from each group during the LTM test (χ2(4, n =40) = 3.542, p =0.4716).There were no significant differences in the duration of vocalisation between any of the groups during conditioning, STM, or LTM tests (one-way ANOVA, Fgroup(

L-368,899 did not block consolidation of contextual fear conditioning
During conditioning, each group acquired equivalent freezing to the context, suggesting antagonism of OXTR by L-368,899 did not prevent expression of conditioned freezing (Two-way ANOVA, FCS(3, 88) = 79.50,p < 0.0001, Fig. 2b).There was no significant difference in the percentage of time spent freezing between L-368,899 and Saline during the STM nor LTM tests (Two-tailed t-test, p =0.5, Fig. 2c, Two-tailed ttest, p =0.15, Fig. 2d).

L-368,899 did not produce any significant effects on number of vocalisers or duration of vocalisation
Fisher's exact test revealed that there was no significant difference in the number of vocalisers between the saline and L-368,899 groups during conditioning (n=10/11 vocalisers per group, p > 0.9999), suggesting no acute effect of OXTR antagonism on 22kHz alarm call production.Fisher's exact tests also revealed that there was no significant difference in the number of vocalisers between groups during the STM nor LTM tests (p > 0.9999 [STM test], p=0.6668 [LTM test]).There were no significant differences in the duration of vocalisation between groups during conditioning, STM, or LTM tests (t-test, p=0.5762; p=0.6038; p=0.8230, respectively).

Discussion
Given prior findings we predicted a deficit in conditioned freezing at LTM following oxytocin receptor antagonist administration with contextual fear conditioning [13].Contrast to our expectations, the two different oxytocin receptor antagonists, Atosiban and L-368,899, produced no effects on consolidation of contextual fear learning as measured by conditioned freezing or the production of 22kHz class of USVs.
Pisansky et al. reported no differences in fear conditioned startle in mice injected with two doses of L-368,899 prior to acquisition, which the data here supports.In contrast, Rasie Abdullahi et al. (2018) reported that ATO had a dose-dependent effect specifically on contextual conditioning LTM.We tried to conceptually reproduce those experimental conditions but did not find the same conclusions.The levels of conditioned freezing obtained herein were comparable, so it is unlikely that it is simply a matter of differential strength of conditioning.A different strain of rats were studied, Wister versus the Lister Hooded strain herein, and all were adult males.Reports have generally noted that Lister hooded are more active than Wister rats and there are differential monoamine-related gene expression levels in the prefrontal cortex between the strains [21,22].It is possible that differential responses to drugs could be due to underlying strain differences, as such has been reported between Lister Hooded and Wister rats for object memory tasks, despite drug-free performance being comparable [23].Furthermore, the majority of work on has focused on male rodents including the experiments herein, so whether the role of oxytocin in responding to conditioned threat cues varies by sex would be pertinent to ascertain given the potential clinical relevance of aversive memories to psychiatric disorders and the clinical application of OXTR antagonists in prevention of pre-term labor.
Atosiban is an antagonist for VAR as well as OXTR, due to their closeness in chemical homology [24].In fact, Atosiban has been shown to have a higher affinity for vasopressin receptors rather than oxytocin receptors, and it was found that low doses of vasopressin/oxytocin peptides enhanced LTM of a passive avoidance task in cuttlefish; [25,26].The balance between vasopressin and oxytocin signaling in the central nucleus of the amygdala is likely to be critical in the control of downstream conditioned responses [27,28].Interestingly, rats with inherited low levels of arginine vasopressin in the central nervous system displayed deficits in contextual fear conditioning [29].Therefore, it would be interesting in future work to delineate the relative contribution of oxytocin or vasopressin signaling on aversive memory.On the other hand, L-368,899 crosses the blood brain barrier and has a higher binding affinity for OXTR than VAR [30,31].The dose used herein was documented to impair the benefit of oxytocin on a stress-induced phenotype and hippocampal plasticity in male Sprague-Dawley rats [32], yet in our hands it did not affect conditioned freezing acquisition, STM nor LTM.
OXTR are G-protein coupled receptors (GPCRs).While Atosiban has been shown to act as an antagonist on the OXTR G q coupled receptors, it also enhanced G i -mediated pathways [33].Thus, it is possible Atosiban acts not to reduce OXTR signaling in general, but instead to bias it towards G i signaling.Unlike Atosiban, L-368,899 was shown to be a full antagonist of all G protein associated signaling at oxytocin receptors [34].Both Atosiban and L-368,899 were shown to not recruit β-arrestins, which block further G protein-initiated signaling and inhibit GPCRs [34].Atosiban and L-368,899 not recruiting β-arrestins could potentially contribute to why consolidation was not blocked.We administered the drugs peripherally.Atosiban concentration in plasma of rats rapidly drops after intravenous administration [35], and has been demonstrated to effectively blunt central effects of OXT on neurotransmission relevant to learning such as dopamine levels after systemic administration [36].However, L-368,899 is argued to have much better penetration across the blood-brain barrier and accumulates in brain regions implicated in aversive conditioning such as the amygdala [30].
While oxytocin administration has been shown to affect USVs in mothers and pups [37,38], and in social contexts [39], peripheral treatment of an oxytocin receptor antagonist (d(CH 2 ) 5 [Tyr-NH 9   2 9 ]OVT) did not affect the rate of alarm call USVs in response to social isolation.This has been interpreted as evidence for exogenous oxytocin's effect on the emission of USVs in a social context; however, an oxytocin receptor antagonist had no effect on USV emission when animals are tested in isolation [40].The USVs herein were recorded after L-368,899 was administered and no significant effects on duration nor number of rats making alarm calls was detected.Additionally, ATO did not have an impact on USV production.It is possible that OXTR antagonists may not have reliable effects on USVs in an individual context.Overall, ATO and L-368,899 did not impair STM nor LTM following contextual fear conditioning.

Fig. 1 .
Fig. 1.Effects of ATO on freezing and vocalisation during Conditioning, STM and LTM.a) Drug was administered immediately following conditioning.b) Percentage of time each rat in each prospective group spent freezing during conditioning.c) Percentage of time each rat spent freezing during the 90s STM test three hours after conditioning.d) Percentage of time each rat spent freezing during the 90s LTM test 48h after conditioning.e) Total duration each rat spent emitting 22kHz USVs during conditioning (ten minutes).f) Total duration each rat spent emitting USVs during the STM test (90s).g) Total duration each rat spent emitting USVs during the LTM test (90s).n = 8/group throughout.

Fig. 2 .
Fig. 2. Effects of L-368,899 on Conditioning, STM, and LTM tests.a) Experimental timeline.b) Percentage of time each rat spent freezing during conditioning one hour post injection.c) Percentage of time each group spent freezing during the 90s STM test.d) Percentage of time each group spent freezing during the 90s LTM test.e) Total duration each rat spent emitting 22kHz USVs during conditioning (ten minutes).f) Total duration each rat spent emitting USVs during the STM test (90s).g) Total duration each rat spent emitting USVs during the LTM test (90s).n = 12 per group throughout.
ENC received funding from Boehringer Ingelheim Pharma GmbH & Co. for a project unrelated to this work.All authors declare no conflicts of interest.CRediT authorship contribution statement Emily R Sherman: Writingreview & editing, Writingoriginal draft, Visualization, Investigation, Formal analysis, Data curation.Jialu Li: Investigation, Formal analysis, Writingreview & editing.Emma N Cahill: Writingoriginal draft, Writingreview & editing, Methodology, Project administration, Funding acquisition, Investigation, Data curation, Formal analysis, Supervision, Conceptualization.