Active avoidance under social conditions recruits the anterior cingulate cortex in male and female rats.

doi:10.21203/rs.3.rs-3750422/v1

Download PDF

Article

Active avoidance under social conditions recruits the anterior cingulate cortex in male and female rats.

https://doi.org/10.21203/rs.3.rs-3750422/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this older preprint version

Read the latest preprint version →

Actively avoiding potential threats is necessary for survival. Most research has focused on the behavioral and neurobiological processes when individuals avoid potential threats alone, under solitary conditions. Therefore, little is known about how social context affects active avoidance. Using a modified version of the platform-mediated avoidance task, we investigated whether the presence of a social partner attenuates freezing responses and promotes greater avoidance learning compared to avoidance learned under solitary conditions. Rats spent a similar percentage of time avoiding during the conditioned tone under both conditions; however, rats trained under social conditions exhibited greater freezing during the tone and lower pressing for food reward compared to solitary rats. Under solitary conditions, we observed greater avoidance in female compared to male rats, which was not present in rats under social conditions. To gain greater mechanistic insight, we optogenetically inactivated glutamatergic projection neurons in the anterior cingulate cortex (ACC) following avoidance training. Photoinactivation of ACC neurons impaired avoidance expression under social conditions both in the presence and absence of the partner. Under solitary conditions, photoinactivation of ACC delayed avoidance in males but blocked avoidance in females. Our findings suggest that avoidance is mediated by the ACC, regardless of social context, and may be dysfunctional in those suffering from trauma-related disorders. Furthermore, sex differences in prefrontal circuits mediating active avoidance warrant further investigation, given that females experience a higher risk of developing anxiety disorders.

Biological sciences/Neuroscience/Learning and memory/Operant learning

Biological sciences/Neuroscience/Motivation

Biological sciences/Neuroscience/Learning and memory/Fear conditioning

Active avoidance of harm can become maladaptive when it interferes with daily activities. Excessive avoidance is a hallmark feature in those suffering from anxiety disorders [1]. Preclinical studies have investigated the behavioral and neural mechanisms of active avoidance using the platform-mediated avoidance (PMA) task, in which a tone-signaled shock can be avoided by stepping onto a platform, while forgoing sucrose reward [2, 3]. However, prior research using the PMA task has only been conducted when animals learn alone, under solitary conditions, leaving much unknown about how social context may alter avoidance behaviors and their underlying neural mechanisms.

Prior studies on social transmission of fear [4–6] and observational fear conditioning (Allsop et al., 2018; Jeon et al., 2010) and fear extinction [7, 8] have established that fear can be learned from a social partner. Although a few older studies have reported observational learning of active avoidance [9–11], it remains unknown whether avoidance is enhanced in the presence of a social partner.

The prelimbic cortex (PL) is key for avoidance behavior that is acquired during PMA under solitary conditions [12]. PL signals the basolateral amygdala (BLA) and ventral striatum (VS) to bidirectionally control avoidance, allowing the animal to make appropriate decisions while facing competing motivational drives [13]. Studies have demonstrated that the anterior cingulate cortex (ACC) plays a prominent role during social transmission of fear [5] and observational fear learning [14–16]. Therefore, the goals of the current study were twofold: determine the behavioral differences in male and female rats that underwent PMA under social or solitary conditions and determine whether ACC activity is necessary for the expression of avoidance using optogenetic inactivation.

Subjects

187 Adult Sprague Dawley (n = 93 females, n = 94 males) rats were purchased (Charles River Laboratories, Wilmington, MA) or bred in-house and weighed 280–420 g at the start of experiments. Subjects were same-sex housed in groups of 2–3 rats per cage and maintained on a 12 hr reverse light cycle and handled as previously described [12].

Rats were restricted to 16–18 g/day of standard laboratory rat chow to maintain 85% of their target weight and trained to lever-press for sucrose pellets (BioServ, Flemington, NJ) on a variable interval schedule of reinforcement (VI-30 s). Rats were trained to a criterion of > 10 presses/min for females and > 15 presses/min for males prior to surgical and/or behavioral procedures. All procedures were approved by the Institutional Animal Care and Use Committee of Kansas State University in compliance with the National Institutes of Health guidelines for the care and use of laboratory animals.

Surgery

For optogenetic experiments, rats were anesthetized under isoflurane and infused with viral vectors in the ACC (+ 1.0 mm AP; ± 0.50 mm ML; -2.0 mm DV to bregma, at a 0˚angle) with 5–6 µL bilaterally (flow rate: 0.05-0.06uL/min). The syringe remained in place for an additional 10 min to reduce backflow. Optical fibers (length, 6 mm; 0.22 NA; 200 nm core from RWD life sciences, Dover, DE) targeted the ACC (+ 1.0 mm AP, ± 3.0 mm ML, -3.0 mm DV, at a 15° angle) and anchored to the skull with cement (C&B-Metabond, Parkell, Brentwood, NY; Ortho Acrylic, Lang Dental, Wheeling, IL). Rats were administered an analgesic (Meloxicam, 1mg/Kg or Flunixin, 1–2 mg/Kg) subcutaneously, and triple antibiotic was applied around the surgical incision. Rats recovered for a minimum of 3 weeks prior to behavioral training to allow for sufficient viral expression (6 weeks before first laser test).

Behavior

For solitary PMA, rats were trained as previously described [2, 12]. Rats were conditioned with a pure tone (30 s, 4 kHz, 75 dB) co-terminating with a scrambled footshock (2 s, 0.4 mA). The inter-trial interval (ITI) averaged 3 min. An acrylic square platform (14.0 cm each side, 0.33 cm tall) located in the opposite corner of the food dish allowed rats to escape shock. The platform was fixed to the floor and present during all training stages. Rats were trained for 10 days with nine tone-shock pairings per day. A VI-30 schedule was maintained across all training and test sessions.

For social partner PMA, rats were conditioned with the same tone, footshock, and ITI parameters, but with another rat located opposite from a perforated plexiglass barrier in a reconfigured shuttle-box operant chamber which allowed rats access to their own lever, food dish, and platform. Learner Rats were paired with partners that were either previously trained in PMA (Trained Partners) or were naïve to the task (Learner Rat) prior to social partner PMA training. Partners were same-sex and non-cagemates. Rats underwent PMA training with their partner across all daily sessions. After 10 days of social partner PMA, rats underwent an additional session in the absence of their partner on Day 11 (9 tone-shock presentations).

Following 10 days of PMA training, rats in the solitary optogenetic group underwent a test of avoidance expression (2 tones presented without shock), [see 12, 13]. Rats in the social partner optogenetic group underwent two expression tests, one in the presence and one in the absence of their partner. Expression tests were counterbalanced to prevent any order effects.

Viruses

The adeno-associated viruses (AAVs; serotype 5) were obtained from the University of North Carolina Vector Core (Chapel Hill, NC). Viral titers were 4 x 10¹² particles/mL for archaerhodopsin (AAV5:CaMKIIα::eArchT3.0-eYFP), and 3 x 10¹² particles/mL for enhanced yellow fluorescent protein (eYFP) control (AAV5:CaMKIIα::eYFP). Rats expressing eYFP control were used to control for changes due to laser-induced heating of tissue [17]. The CaMKIIα promoter was used to enable transgene expression favoring pyramidal neurons [18] in cortical regions [19–21] (Jones et al., 1994; Van der Oever et al., 2013; Warthen et al., 2016). Viruses were housed in a -80°C freezer until the day of infusion.

Laser delivery

ACC neurons were bilaterally illuminated using a DPSS green laser (532 nm, constant, 10–12 mW at the optical fiber tip; OptoEngine, Midvale, UT). The laser was activated at tone onset during Tone 1 of the test and persisted throughout the 30 s tone presentation. Laser light passed through a shutter/coupler (200 nm, SRS, Stanford, CA), patch cord (200 nm core, ThorLabs, Newton NJ), rotary joint (200 nm core, 1x2, Doric Lenses, Quebec city, Canada), dual patch cord (0.22 NA, 200 nm core, ThorLabs or RWD life sciences), and optical fibers targeting ACC. Rats were familiarized with dummy patch cords prior to tests.

Open field task

Locomotion was automatically assessed (ANY-Maze, Stoelting Co, Wood Dale, IL) in an open field arena (90 cm diameter) during 30 s laser off and 30 s laser on time periods. A 6 min acclimation period preceded laser illumination. Speed and distance traveled were used to assess locomotion, and time in center was used to assess anxiety.

Pressing test

Lever-pressing was assessed with a VI-30 schedule and began with a 60 s acclimation period, followed by 7 Laser On (30 s) trials and 6 60 s Laser Off (60 s) intervals. The number of lever presses was compared during Laser Off and Laser On periods within subjects using a paired-t-test.

Histology

After experiments, rats were deeply anesthetized with sodium pentobarbital (450 mg/kg i.p.) and transcardially perfused with 0.9% saline followed by a 10% buffered formaldehyde. Brains were removed and stored in 30% sucrose for cryoprotection for at least 72 h before sectioning. Coronal sections were cut (40 µm), mounted on slides and analyzed for viral expression and optical fiber placement.

Data Collection and Analysis

Behavior was recorded with digital video cameras and quantified using ANY-Maze software (Stoelting, Wood Dale, IL). The number of shocks avoided was calculated as the rat spending at least 1.75 s on the platform during the 2 s shock period of each tone presentation. Shock reactivity was calculated using the maximum speed of each rat during exposure to the first shock on Day 1 of PMA.

Multilevel regressions were performed to assess significant differences in several behaviors observed during PMA. For percentage of time spent on the platform and time spent freezing during the tone (with the maximum time, 30 seconds, included as the model weight parameter), multilevel binomial logistic regressions were performed to account for individual differences and proportion data being bounded by zero and one [22]. Model parameter estimates are available in Supplementary Tables S1 (freezing) and S2 (platform time). For number of shocks avoided and number of ITI presses (30 s prior to tone onset), multilevel negative binomial regressions were performed to account for individual differences and positively skewed count data. Model parameter estimates are available in Supplementary Tables S3 (lever presses) and S4 (shocks avoided). All models included Group Type (social vs solitary), Sex, and Session (day of training) as fixed effects and Individual and Tone Number as random effects. To investigate effects of Partner Type (Learner Rat vs. Trained Partner), additional models were run only on the social partner PMA data, with Partner Type added as a predictor variable instead of Group Type. Parameter estimates for these models are available in Supplementary Tables S1-S8.

All analyses for experiments in Figs. 1–3 were conducted in R (version 4.2.1), using the lme4 library, version 1.1–30 [23]. The emmeans library, version 1.8.0 [24] was used to calculate post-hoc Tukey tests and estimated marginal means from each model (all reported means model estimates).

For optogenetic manipulations, repeated-measures ANOVA, followed by post-hoc Tukey analyses, or Student’s two-tailed t tests were used where appropriate using Prism (Graphpad, La Jolla, CA), or JMP (SAS, Cary, NC) software. Aspects of operant box schematic were created with Biorender.com.

Platform-mediated avoidance (PMA) under social conditions increases freezing and decreases pressing compared to PMA under solitary conditions.

Previous studies of PMA under solitary conditions have shown that freezing decreases and avoidance increases as rats progress through training [2, 25]. To investigate how the presence of another rat can affect avoidance, several behaviors were compared across PMA training under social or solitary conditions. One group of rats was trained in social partner PMA, in which rats underwent training simultaneously with another rat (Fig. 1A, left) and another group of rats was trained in solitary PMA, in which they learned avoidance alone (Fig. 1A, right). Both groups underwent training for 10 days as previously described [2, 12]. We utilized multilevel regressions to describe behavior observed during PMA under solitary or social conditions (see Methods and Supplemental Information on regression models and parameter estimates). Across 10 days of training, both social (purple) and solitary (blue) groups showed similar levels of avoidance, as measured by the percentage of time spent on the platform during the tone (Fig. 1B) and the average number of shocks avoided on each day (Fig. 1C). There was no significant effect of Group Type for time on platform (social vs. solitary; z = -1.252, p = 0.211). or number of shocks avoided (z = -0.552, p = 0.581).

Interestingly, rats trained under social conditions showed greater freezing compared to rats trained under solitary conditions (Fig. 1D). A multilevel regression showed a significant effect of Group Type predicting time freezing (z = 4.068, p < 0.0001), with social rats freezing more than solitary rats. In addition, solitary rats pressed significantly more than social rats (z = -6.304, p < 0.0001; Fig. 1E). Therefore, PMA training under social conditions enhanced freezing responses while decreasing food-seeking, with little effect on avoidance, compared to PMA training under solitary conditions.

Solitary PMA reveals sex differences that are not present during social partner PMA.

Most PMA studies have used only male rats [2, 12, 13, 25–28]. However, recent studies have included both sexes [29–31]. To investigate sex differences in PMA under social or solitary conditions, we included both male and female rats in all experiments. Data in Fig. 2A-H is the same data as in Fig. 1 but separated by males (teal) and females (salmon). Post-hoc Tukey tests on the previous regression models showed that males and females under social conditions showed no significant differences in avoidance (Fig. 1A; z = 1.308, p = 0.191), number of shocks avoided (Fig. 2B; z = 1.590, p = 0.112), or freezing (Fig. 2C; z = -0.673, p = 0.501). However, males showed significantly increased pressing compared to females (Fig. 2D; z = -3.112, p = 0.0019).

During solitary PMA, females avoided significantly more than males, as measured by time on platform (Fig. 2E; z = 3.174, p = 0.0015;) and number of shocks avoided (Fig. 2F; z = 2.949, p = 0.0032). This effect was not due to shock reactivity (Fig. 2F, inset, t-test, t₍₄₇₎ = 1.465, p = 0.150). Post-hoc Tukey tests on the regression model showed no significant differences in freezing between males and females (Fig. 2G, z = 1.903, p = 0.0571), but males pressed more than females (Fig. 2H, z = -3.044, p = 0.0023). Altogether, these results suggest that sex differences are suppressed during social partner PMA compared to solitary PMA.

The significant effects of Group Type (social vs. social) suggests that there are differences between groups in freezing and pressing regardless of sex differences. To confirm this, we directly compared behaviors in females between Group Type and males between Group Type using contrast tests on the multilevel regression models. Post-hoc Tukey tests identified significant differences in freezing between social and solitary females (z = 4.77, p < 0.0001) and between social and solitary males (z = 7.449, p < 0.0001; Supplementary Fig. 1C and G). Similarly, significant differences were found in pressing between social and solitary females (z = -4.069, p < 0.0001), and between social and solitary males (z = -3.623, p = 0.0003; Supplementary Fig. 1D and H). Collectively, freezing was enhanced and pressing was diminished during social partner PMA, regardless of sex.

Previous PMA experience of a social partner does not affect acquisition but alters avoidance and freezing in the absence of the partner.

To determine whether the previous experience of a social partner affects avoidance, rats were paired with same-sex non-cagemate partners that had previously undergone PMA (Learner Rat with a Trained Partner) or were naïve to the task at the start of PMA (Learner Rat with another Learner Rat). We compared the same behaviors (time on platform, number of shocks avoided, time freezing, and pressing) across the 10 days in Learner Rats trained with a Trained Partner (purple) or Learner Rats trained with another Learner Rat (yellow). A multilevel regression found no significant effect of Partner Type on avoidance (Fig. 3B; z = -1.548, p = 0.122), freezing (Fig. 3D; z = -1.168, p = 0.243), number of shocks avoided (Fig. 3C; z = 0.391, p = 0.695), or pressing (Fig. 3E; z = 0.218, p = 0.828). Altogether, partner types learn social partner PMA in a similar fashion.

We were next interested in whether the absence of the partner would alter behaviors during PMA. We compared the same behavioral measures of avoidance, freezing, and pressing in the presence of their partner (on Day 10) versus in the absence of their partner (on Day 11) using contrast comparisons on the previous multilevel regression models. Learner Rats with either Partner type spent significantly more time on the platform on Day 11 than Day 10 (Fig. 3F, Trained Partner z = -31.104, p < 0.0001; Learner Rat z = -7.669, p < 0.0001). Learner rats trained with either Partner Type avoided more shocks on Day 11 than Day 10 (Fig. 3G, Trained Partner z = -3.455, p = 0.0005; Learner Rat z = -2.519, p = 0.0118).

There was also a significant increase in freezing from Day 10 to Day 11 for Learner Rats with a Trained Partner (Fig. 3H, z = -3.921, p < 0.0001). There were no significant differences in freezing between groups on Day 11 (with a Learner rat: z = -0.158; p = 0.874; with a Trained Partner z = 1.172; p = 0.241). There were small significant differences in pressing between Day 10 and 11 for Learner Rats with either Partner Type (Fig. 3I; Trained Partner z = -15.022, p < .0001; Learner Rat Partner z = -17.218, p < .0001) but no significant differences in pressing between groups on Day 11 (with a Learner rat: z = -1.423; p = 0.1547; with a Trained Partner: z = -1.065; p = 0.2871).

We next assessed whether there were any sex differences (regardless of Partner Type) that depended on partner presence. Both males and females spent significantly more time on the platform on Day 11 than Day 10, and females avoided significantly more than males when their partner was absent (Day 11), regardless of Partner Type during training (Supplementary Fig. 2A). Females also avoided significantly more shocks (Supplementary Fig. 2B) and showed a small significant increase in freezing for females from Day 10 to 11 (Supplementary Fig. 2C). Taken together, partner absence augments fear and avoidance responses, especially in females, when avoidance is learned socially.

Photoinactivation of ACC impairs avoidance under social conditions.

Previous studies have linked ACC activity with social learning [15, 16, 32]. We therefore reasoned that ACC activity would be necessary for social partner PMA. To assess this, we used an optogenetic approach to test if inactivation of ACC neurons would impair avoidance under social conditions. ArchT-eYFP (AAV5:CaMKIIa::eArchT3.0-eYFP) or eYFP control was expressed in ACC gluatamatergic projection neurons [18–21]. Following viral infusion, surgical placement of optical probes, and a 4–5 week period to allow viral expression, rats were trained in social partner PMA (Fig. 4A). Histological analysis confirmed that expression of ArchT-eYFP was largely confined to the ACC (including anatomical areas Cg1 and Cg2) with minimal spread to prelimbic cortex (PL; Fig. 4A, bottom left).

Following 10 days of social partner PMA (Fig. 4A, middle), Learner rats underwent two avoidance expression tests: one in the presence and another in the absence of their partner (Fig. 4A, top and bottom right, respectively). Green laser was presented concurrently with the first 30 s tone (Laser ON), but not during Tone 2 (Laser OFF). A 2-way repeated measures ANOVA comparing time on platform in ArchT-eYFP and eYFP controls during Tone 1 on the last day of training, and Tones 1 (Laser ON) and 2 (Laser OFF) of test with the partner present revealed a significant main effect of trial (F_(2,38) = 13.88, p < .001) and interaction between trial and AAV (F_(2,38) = 3.70, p = 0.034), but not a significant main effect of AAV (F_(1,19) = 1.81, p = 0.194; data collapsed across partner type and sex). Post-hoc Tukey tests revealed a significant decrease in time on platform between Tone 1 on the last day of training and Tone 1 of test in the ArchT-eYFP group (p = 0.002), but not in the eYFP control group (p = 0.478; Fig. 4B). When comparing avoidance latency across the tone + laser trial between ArchT-eYFP and eYFP control rats, there was no significant difference (Fig. 4C, t₍₁₉₎ = 1.122, p = 0.276), nor across the timecourse of avoidance, as measured in 3 s bins of the tone period (Fig. 4D, repeated measures ANOVA: F_(9,126) = 0.891, p = 0.535). Finally, photoinactivation had no effect on freezing (t₍₁₉₎ = 1.683, p = 0.109) nor on suppression of bar pressing (t₍₁₉₎ = 0.607, p = 0.551) in the presence of the partner (Fig. 4E).

In the absence of the partner, a 2-way repeated measures ANOVA comparing time on the platform in ArchT-eYFP and eYFP controls during Tone 1 of the last day of training, and Tones 1 and 2 of test revealed a significant main effect of trial (F_(2,40) = 21.79, p < .001) and interaction between trial and AAV (F_(2,40) = 4.21, p = .022), but not a significant main effect of AAV (F_(1,20) = 1.01, p = 0.328). Post-hoc Tukey tests revealed a significant decrease in time on the platform between Tone 1 on the last day of training and Tone 1 of test in the ArchT-eYFP group (p < .001), but not in the eYFP control group (p = 0.098; Fig. 4F). Photoinactivation had no effect on avoidance latency (Fig. 4G, t₍₂₀₎ = 0.539, p = 0.596). When comparing the timecourse of avoidance between ArchT-eYFP and eYFP control rats, photoinactivation significantly reduced avoidance throughout the tone (Fig. 4H, F_(9,135) = 3.09, p = 0.002)). Finally, photoinactivation had no effect on freezing (t₍₂₀₎ = 2.036, p = 0.0552) but a significant decrease on suppression of bar pressing (t₍₂₀₎ = 3.225, p = 0.0042) in absence of the partner (Fig. 4I). Altogether, photoinactivation of ACC somata during the tone impaired avoidance expression during the presence and absence of the partner.

Photoinactivation of ACC delays avoidance under solitary conditions in males but blocks avoidance in females.

To determine whether ACC activity is necessary for avoidance under solitary conditions, we photoinactivated these neurons during an expression test following solitary PMA. Similar to the experiment in Fig. 4, rats were infused with either ArchT-eYFP or eYFP control and subsequently underwent solitary PMA (Fig. 5A). A 2-way repeated measures ANOVA comparing time on the platform in ArchT-eYFP and eYFP controls during Tone 1 of the last day of training, and Tones 1 and 2 of test showed a significant main effect of trial (F_(2,70) = 18.49, p < 0.001), but no main effect of AAV (F_(1,35) = 2.83, p = 0.102), or their interaction (F_(2,70) = 0.555, p = 0.577; Fig. 5B). Post-hoc Tukey tests revealed a significant decrease in time on the platform between Tone 1 on the last day of training and Tone 1 of test in the ArchT-eYFP group (p < .0.001) but also in the eYFP control group (p = 0.004). ArchT-eYFP rats showed a significantly longer avoidance latency (mean: 12.3 s) compared to eYFP controls (mean: 7.34 s, t₍₃₇₎ = 2.132, p = 0.0397, Fig. 5C). Analysis of the timecourse of avoidance showed no differences between ArchT-eYFP and eYFP control rats (Fig. 5D, repeated measures ANOVA, F_(9,279) = 1.25, p = 0.265). However, female ArchT-eYFP rats showed significantly less avoidance compared to female eYFP controls (p = 0.0010), whereas male ArchT-eYFP and male eYFP control rats did not show any significant differences (p = 0.189, Fig. 5E). Interestingly, ArchT-eYFP males showed a delay in avoidance compared to ArchT-eYFP females, who showed a near complete impairment of avoidance across the 30 s tone (comparing light and dark green lines across both graphs in Fig. 5E; repeated measures ANOVA, F_(9,162) = 3.76, p < 0.001, post hoc Tukeys, all p’s < 0.001 after 15 s). Photoinactivation had no effect on freezing (t₍₃₇₎ = 0.607, p = 0.548) or suppression of bar pressing (t₍₃₁₎ = 1.023, p = 0.314; Fig. 5F). Thus, photoinactivation of ACC somata blocked avoidance in females, but delayed avoidance in males.

Finally, photoinactivation had no effect on spontaneous bar-pressing between ArchT-eYFP and eYFP controls (t-test, t₍₂₄₎ = 0.343, p = 0.738; ArchT-eYFP n = 11, M = 4.7, eYFP n = 13, M = 4.8). Photoinactivation also had no effect on locomotion in an open field (t₍₄₁₎ = -1.22, p = 0.232; ArchT-eYFP n = 17, M = 1.1 m, eYFP n = 24,M = 0.97 m), nor on anxiety levels, as both groups spent similar amounts of time in the center of the open field (t₍₄₁₎ = -1.35, p = 0.184; ArchT-eYFP n = 17, M = 3.3 s, eYFP n = 24, M = 2.0 s).

We developed a behavioral task to assess the acquisition of PMA under social conditions and compared learning between male and female rats trained in PMA under social or solitary conditions. We found that females avoided more than males, whereas males pressed more for sucrose reward than females during PMA training, and this persisted, even in the absence of the social partner. Sex differences were more prominent in solitary vs. social rats. Thus, a social context of aversive learning appears to suppress sex differences, causing female rats to “act more like” male rats. We also found that, regardless of sex, freezing was enhanced under social vs. solitary conditions. In addition, when rats underwent social partner PMA, the prior experience of their partner did not affect behavior during training. However, removing their partner increased avoidance in females and freezing in both sexes, especially if their partner had previous PMA experience. We next found that ACC activity was necessary for the expression of avoidance in rats that learned under social and solitary conditions, but in different ways. In social partner PMA, ACC activity was necessary for the expression of avoidance both in the presence and absence of the partner. In solitary PMA, ACC activity modulated avoidance expression in a sex-dependent manner, such that photoinactivating ACC had a milder effect of delaying avoidance in males, but a stronger effect of blocking avoidance in females. Below, we highlight some unexpected findings from our study along with possible explanations and future research directions.

Our surprising finding that freezing increased during social partner PMA compared to solitary PMA (Fig. 1C) suggests that Learner Rats associated their partner with danger. This agrees with studies showing that rats can associate another rat with a shock US [33]. A rat that is freezing to a CS can also become a signal for danger, in addition to the CS itself [34]. It is also possible that Learner Rats increased freezing during social partner PMA due to detecting lack of movement from their partner, similar to previous studies showing that social transmission of fear can be propagated by silence caused by cessation of movement [35]. In the current study, it is likely that Learner Rats perceived their partner as a negative stimulus rather than a positive stimulus that would have buffering fear responses. This agrees with studies showing that partner presence during a traumatic event does not reduce fear responses, but rather a partner present after a traumatic event does reduce fear responses [8]. Future studies using social partner PMA can address this question by testing whether presenting a partner after PMA training (under solitary conditions) would decrease freezing and facilitate extinction of avoidance.

We also observed lower press rates under social compared to solitary conditions, suggesting that Learner Rats may be less motivated for food reward when another rat is present. This may be due to Learner Rats investigating their partner to obtain sensory information about PMA. Studies have found that during social learning, rodents use auditory and visual cues [36] as well as olfactory cues [37, 38] to gather information about the environment. In the current study, Learner Rats may observe their partner avoid, causing them to also avoid or hear an alarm vocalization emitted by their partner during the tone, but these were not measured. Future studies should aim to investigate whether social partner PMA relies on visual, auditory, or olfactory cues between partners.

Studies investigating social interactions and social learning have pointed to the anterior cingulate cortex (ACC) as a key region across many species [16]. Previous research suggests that ACC is necessary for the integration of social cues and information about aversive stimuli, which are mediated by connections between the ACC and basolateral amygdala (BLA) [39]. The ACC is a highly interconnected cortical brain region, projecting densely to the PL [40, 41] and BLA [42–44]. The ACC may signal to the PL information about the social partner during PMA. The ACC also comprises part of the value processing network together with the BLA, which processes observed emotion and social interactions [16, 32]. ACC neurons may signal the BLA during PMA under social conditions, thereby allowing Learner rats to use social information to avoid danger.

We found that during PMA under solitary conditions, ACC photoinactivation had a mild effect of delaying avoidance in males but had a stronger effect of blocking avoidance in female, suggesting that ACC is recruited differentially in male and female rats. This suggests that female avoidance circuits rely more heavily on ACC, whereas male avoidance circuits may rely more on other nodes of the avoidance circuit such as PL [12, 13]. Future electrophysiological studies could determine ACC correlates of behavior during PMA and how ACC activity may differ under social partner vs. solitary PMA. For example, Partner Type may alter ACC activity during social partner PMA, similar to studies showing that dominance status changes neuronal activity during social interactions [45].

Excessive avoidance is a key symptom in several neuropsychiatric illnesses including PTSD [46, 47], OCD [48], social anxiety disorder (SAD) [49], depression [50], and autism spectrum disorder (ASD) [51]. The human homolog of the rat ACC, Brodmann areas 24a/b [16], has been implicated in these disorders [OCD: 52, 53], [SAD: 54], [depression: 55], [ASD: 56]. Interestingly, a meta-analysis reported that the ACC was one of seven regions with disrupted functional connectivity across different anxiety disorders [57]. Therefore, findings of the current study are consistent with clinical studies that the ACC is an area of interest for developing treatments to resolve symptoms of these neuropsychiatric disorders.

The PMA task has previously been used as a model to study extinction-based treatment for PTSD and OCD [25, 26]. Therefore, modifying the PMA task to include a social context could be used as an animal model of behavior to study other anxiety disorders such as SAD, in which social context is a key factor. Finally, our findings that females avoided more than males and showed a stronger impairment of avoidance when ACC was optogenetically inactivated suggest that ACC may be a key factor in understanding why females are at a greater risk of an anxiety disorder.

Competing interests

The authors declare that no competing interests exist.

Funding

This study was supported by NIH grants #P20-GM103418 and #P20-GM113109 (subawards to MMD), and the department of Psychological Sciences, College of Arts & Sciences, and the Vice President of Research at Kansas State University.

Author Contributions

Conceptualization, MMD; Research design, MMD, GE; Methodology, MMD; Investigation, SR, CK, LW, AS, GE, MMD; Data analysis, SR, CK, KP, MMD; Supervision of experiments and analysis: SR, MMD; Writing, SR, CK, LW, KP, MMD; Funding Acquisition, MMD.

Acknowledgements

We thank Drs. Fabricio Do Monte, Anthony Burgos-Robles, Debra Bangasser, and Gabriel Gasque for helpful comments on the manuscript. We also thank Tessa Maze, Jenna Thompson, Kelly Krehbiel, Ashvini Wickramasundara, Allison Moser, Allison Drouhard, Emma Wrampe, and Ivy Auletti for technical assistance.

American Psychiatric Association, Diagnostic and statistical manual of mental disorders: DSM-5, 5th ed. Washington, D.C.: American Psychiatric Association, 2013, pp. xliv, 947 p.
C. Bravo-Rivera, C. Roman-Ortiz, E. Brignoni-Perez, F. Sotres-Bayon, and G. J. Quirk, "Neural structures mediating expression and extinction of platform-mediated avoidance" J Neurosci, vol. 34, no. 29, pp. 9736–42, Jul 2014, doi: 10.1523/JNEUROSCI.0191-14.2014.
M. M. Diehl, C. Bravo-Rivera, and G. J. Quirk, "The study of active avoidance: A platform for discussion" Neurosci Biobehav Rev, vol. 107, pp. 229–237, Sep 2019, doi: 10.1016/j.neubiorev.2019.09.010.
A. K. Bruchey, C. E. Jones, and M. H. Monfils, "Fear conditioning by-proxy: social transmission of fear during memory retrieval" Behav Brain Res, vol. 214, no. 1, pp. 80 – 4, Dec 2010, doi: 10.1016/j.bbr.2010.04.047.
C. E. Jones and M. H. Monfils, "Dominance status predicts social fear transmission in laboratory rats" Anim Cogn, vol. 19, no. 6, pp. 1051–1069, Nov 2016, doi: 10.1007/s10071-016-1013-2.
H. B. Brandl, J. C. Pruessner, and D. R. Farine, "The social transmission of stress in animal collectives" Proc Biol Sci, vol. 289, no. 1974, p. 20212158, May 11 2022, doi: 10.1098/rspb.2021.2158.
N. Brill-Maoz and M. Maroun, "Extinction of fear is facilitated by social presence: Synergism with prefrontal oxytocin" Psychoneuroendocrinology, vol. 66, pp. 75–81, Apr 2016, doi: 10.1016/j.psyneuen.2016.01.003.
T. Gorkiewicz et al., "Social buffering diminishes fear response but does not equal improved fear extinction" Cereb Cortex, vol. 33, no. 8, pp. 5007–5024, Apr 04 2023, doi: 10.1093/cercor/bhac395.
W. J. Presley and A. J. Riopelle, "Observational learning of an avoidance response" J Genet Psychol, vol. 95, pp. 251–4, Dec 1959, doi: 10.1080/00221325.1959.10534265.
E. R. John, P. Chesler, F. Bartlett, and I. Victor, "Observation learning in cats" Science, vol. 159, no. 3822, pp. 1489–91, Mar 29 1968, doi: 10.1126/science.159.3822.1489.
J. E. Del Russo, "Observational learning of discriminative avoidance in hooded rats" Animal Learning & Behavior, vol. 3, no. 1, pp. 76–80, 1975/03/01 1975, doi: 10.3758/BF03209103.
M. M. Diehl et al., "Active avoidance requires inhibitory signaling in the rodent prelimbic prefrontal cortex" Elife, vol. 7, May 31 2018, doi: 10.7554/eLife.34657.
M. M. Diehl et al., "Divergent projections of the prelimbic cortex bidirectionally regulate active avoidance" Elife, vol. 9, 10 2020, doi: 10.7554/eLife.59281.
A. Olsson and E. A. Phelps, "Social learning of fear" Nat Neurosci, vol. 10, no. 9, pp. 1095 – 102, Sep 2007, doi: 10.1038/nn1968.
J. Debiec and A. Olsson, "Social Fear Learning: from Animal Models to Human Function" Trends Cogn Sci, vol. 21, no. 7, pp. 546–555, Jul 2017, doi: 10.1016/j.tics.2017.04.010.
A. Burgos-Robles, K. M. Gothard, M. H. Monfils, A. Morozov, and A. Vicentic, "Conserved features of anterior cingulate networks support observational learning across species" Neurosci Biobehav Rev, vol. 107, pp. 215–228, 12 2019, doi: 10.1016/j.neubiorev.2019.09.009.
J. M. Stujenske, T. Spellman, and J. A. Gordon, "Modeling the Spatiotemporal Dynamics of Light and Heat Propagation for In Vivo Optogenetics" Cell Rep, vol. 12, no. 3, pp. 525 – 34, Jul 2015, doi: 10.1016/j.celrep.2015.06.036.
X. B. Liu and E. G. Jones, "Localization of alpha type II calcium calmodulin-dependent protein kinase at glutamatergic but not gamma-aminobutyric acid (GABAergic) synapses in thalamus and cerebral cortex" Proc Natl Acad Sci U S A, vol. 93, no. 14, pp. 7332–6, Jul 1996.
E. G. Jones, G. W. Huntley, and D. L. Benson, "Alpha calcium/calmodulin-dependent protein kinase II selectively expressed in a subpopulation of excitatory neurons in monkey sensory-motor cortex: comparison with GAD-67 expression" J Neurosci, vol. 14, no. 2, pp. 611 – 29, Feb 1994.
M. C. Van den Oever et al., "Ventromedial prefrontal cortex pyramidal cells have a temporal dynamic role in recall and extinction of cocaine-associated memory" J Neurosci, vol. 33, no. 46, pp. 18225–33, Nov 2013, doi: 10.1523/JNEUROSCI.2412-13.2013.
D. M. Warthen et al., "Activation of Pyramidal Neurons in Mouse Medial Prefrontal Cortex Enhances Food-Seeking Behavior While Reducing Impulsivity in the Absence of an Effect on Food Intake" Front Behav Neurosci, vol. 10, p. 63, 2016, doi: 10.3389/fnbeh.2016.00063.
B. M. Bolker, "Linear and generalized linear mixed models.," in Ecological Statistics: Contemporary Theory and Application, G. A. Fox, Negrete-Yankelevich, S., and Sosa, V.J., Ed., 1 ed: Oxford University Press, 2015, pp. 309–333.
D. Bates, M. Maechler, B. M. Bolker, and S. Walker, "lme4: linear mixed-effects models using Eigen and S4.," R package version 1.1-5 ed, 2014.
R. V. Lenth, P. Buerkner, M. Herve, M. Jung, J. Love, and F. Miguez, "Emmeans: Estimated Marginal Means, aka Least-Squares Means.," Version 1.8. 1–1 ed, 2022.
F. J. Martínez-Rivera, M. J. Sánchez-Navarro, C. I. Huertas-Pérez, B. D. Greenberg, S. A. Rasmussen, and G. J. Quirk, "Prolonged avoidance training exacerbates OCD-like behaviors in a rodent model" Transl Psychiatry, vol. 10, no. 1, p. 212, 07 2020, doi: 10.1038/s41398-020-00892-5.
J. Rodriguez-Romaguera, B. D. Greenberg, S. A. Rasmussen, and G. J. Quirk, "An Avoidance-Based Rodent Model of Exposure With Response Prevention Therapy for Obsessive-Compulsive Disorder" Biol Psychiatry, vol. 80, no. 7, pp. 534 – 40, Oct 2016, doi: 10.1016/j.biopsych.2016.02.012.
F. J. Martínez-Rivera, C. Bravo-Rivera, C. D. Velázquez-Díaz, M. Montesinos-Cartagena, and G. J. Quirk, "Prefrontal circuits signaling active avoidance retrieval and extinction" Psychopharmacology (Berl), Sep 2018, doi: 10.1007/s00213-018-5012-7.
C. Bravo-Rivera, C. Roman-Ortiz, M. Montesinos-Cartagena, and G. J. Quirk, "Persistent active avoidance correlates with activity in prelimbic cortex and ventral striatum" Front Behav Neurosci, vol. 9, p. 184, 2015, doi: 10.3389/fnbeh.2015.00184.
C. J. Gabriel et al., "BehaviorDEPOT is a simple, flexible tool for automated behavioral detection based on markerless pose tracking" Elife, vol. 11, Aug 23 2022, doi: 10.7554/eLife.74314.
M. W. Gongwer et al., "Brain-wide projections and differential encoding of prefrontal neuronal classes underlying learned and innate threat avoidance" J Neurosci, Jul 24 2023, doi: 10.1523/JNEUROSCI.0697-23.2023.
C. J. Halcomb et al., "Sex divergent behavioral responses in platform-mediated avoidance and glucocorticoid receptor blockade" Psychoneuroendocrinology, vol. 159, p. 106417, Oct 20 2023, doi: 10.1016/j.psyneuen.2023.106417.
M. A. Apps, M. F. Rushworth, and S. W. Chang, "The Anterior Cingulate Gyrus and Social Cognition: Tracking the Motivation of Others" Neuron, vol. 90, no. 4, pp. 692–707, May 18 2016, doi: 10.1016/j.neuron.2016.04.018.
L. M. Dawud et al., "A novel social fear conditioning procedure alters social behavior and mTOR signaling in differentially housed adolescent rats" Dev Psychobiol, vol. 63, no. 1, pp. 74–87, Jan 2021, doi: 10.1002/dev.22001.
A. Cruz, M. Heinemans, C. Márquez, and M. A. Moita, "Freezing Displayed by Others Is a Learned Cue of Danger Resulting from Co-experiencing Own Freezing and Shock" Curr Biol, vol. 30, no. 6, pp. 1128–1135.e6, Mar 23 2020, doi: 10.1016/j.cub.2020.01.025.
A. G. Pereira, A. Cruz, S. Q. Lima, and M. A. Moita, "Silence resulting from the cessation of movement signals danger" Curr Biol, vol. 22, no. 16, pp. R627-8, Aug 21 2012, doi: 10.1016/j.cub.2012.06.015.
N. Paraouty, J. A. Charbonneau, and D. H. Sanes, "Social learning exploits the available auditory or visual cues" Sci Rep, vol. 10, no. 1, p. 14117, Aug 24 2020, doi: 10.1038/s41598-020-71005-x.
A. Contestabile, G. Casarotto, B. Girard, S. Tzanoulinou, and C. Bellone, "Deconstructing the contribution of sensory cues in social approach" Eur J Neurosci, vol. 53, no. 9, pp. 3199–3211, May 2021, doi: 10.1111/ejn.15179.
G. Sanchez-Andrade and K. M. Kendrick, "The main olfactory system and social learning in mammals" Behav Brain Res, vol. 200, no. 2, pp. 323 – 35, Jun 25 2009, doi: 10.1016/j.bbr.2008.12.021.
S. A. Allsop et al., "Corticoamygdala Transfer of Socially Derived Information Gates Observational Learning" Cell, vol. 173, no. 6, pp. 1329–1342.e18, 05 2018, doi: 10.1016/j.cell.2018.04.004.
F. Conde, E. Maire-Lepoivre, E. Audinat, and F. Crepel, "Afferent connections of the medial frontal cortex of the rat. II. Cortical and subcortical afferents," J.Comp Neurol., vol. 352, no. 4, pp. 567–593, 1995.
B. F. Jones, H. J. Groenewegen, and M. P. Witter, "Intrinsic connections of the cingulate cortex in the rat suggest the existence of multiple functionally segregated networks" Neuroscience, vol. 133, no. 1, pp. 193–207, 2005, doi: 10.1016/j.neuroscience.2005.01.063.
M. D. Cassell and D. J. Wright, "Topography of projections from the medial prefrontal cortex to the amygdala in the rat" Brain Res Bull, vol. 17, no. 3, pp. 321 – 33, Sep 1986, doi: 10.1016/0361-9230(86)90237-6.
S. Bissière et al., "The rostral anterior cingulate cortex modulates the efficiency of amygdala-dependent fear learning" Biol Psychiatry, vol. 63, no. 9, pp. 821 – 31, May 01 2008, doi: 10.1016/j.biopsych.2007.10.022.
P. L. Gabbott, T. A. Warner, P. R. Jays, P. Salway, and S. J. Busby, "Prefrontal cortex in the rat: projections to subcortical autonomic, motor, and limbic centers," J.Comp Neurol., vol. 492, no. 2, pp. 145–177, 2005.
L. Kingsbury et al., "Correlated Neural Activity and Encoding of Behavior across Brains of Socially Interacting Animals" Cell, vol. 178, no. 2, pp. 429–446.e16, 07 2019, doi: 10.1016/j.cell.2019.05.022.
N. Breslau, "Outcomes of posttraumatic stress disorder," J.Clin.Psychiatry, vol. 62 Suppl 17, pp. 55–59, 2001.
G. J. Asmundson, J. A. Stapleton, and S. Taylor, "Are avoidance and numbing distinct PTSD symptom clusters?" J Trauma Stress, vol. 17, no. 6, pp. 467 – 75, Dec 2004, doi: 10.1007/s10960-004-5795-7.
J. F. McGuire, E. A. Storch, A. B. Lewin, L. H. Price, S. A. Rasmussen, and W. K. Goodman, "The role of avoidance in the phenomenology of obsessive-compulsive disorder" Compr Psychiatry, vol. 53, no. 2, pp. 187 – 94, Feb 2012, doi: 10.1016/j.comppsych.2011.03.002.
S. Wake, C. M. van Reekum, and H. Dodd, "The effect of social anxiety on the acquisition and extinction of low-cost avoidance" Behav Res Ther, vol. 146, p. 103967, Nov 2021, doi: 10.1016/j.brat.2021.103967.
J. L. Trew, "Exploring the roles of approach and avoidance in depression: an integrative model" Clin Psychol Rev, vol. 31, no. 7, pp. 1156–68, Nov 2011, doi: 10.1016/j.cpr.2011.07.007.
A. R. Madipakkam, M. Rothkirch, I. Dziobek, and P. Sterzer, "Unconscious avoidance of eye contact in autism spectrum disorder" Sci Rep, vol. 7, no. 1, p. 13378, Oct 2017, doi: 10.1038/s41598-017-13945-5.
I. Lee, K. M. Kim, and M. H. Lim, "Theta and Gamma Activity Differences in Obsessive-Compulsive Disorder and Panic Disorder: Insights from Resting-State EEG with eLORETA" Brain Sci, vol. 13, no. 10, Oct 10 2023, doi: 10.3390/brainsci13101440.
E. Kosová et al., "Glutamatergic abnormalities in the pregenual anterior cingulate cortex in obsessive-compulsive disorder using magnetic resonance spectroscopy: A controlled study" Psychiatry Res Neuroimaging, vol. 335, p. 111721, Oct 2023, doi: 10.1016/j.pscychresns.2023.111721.
A. J. Jamieson et al., "A brain model of altered self-appraisal in social anxiety disorder" Transl Psychiatry, vol. 13, no. 1, p. 344, Nov 11 2023, doi: 10.1038/s41398-023-02644-7.
L. Alexander, P. C. T. Hawkins, J. W. Evans, M. A. Mehta, and C. A. Zarate, "Preliminary evidence that ketamine alters anterior cingulate resting-state functional connectivity in depressed individuals" Transl Psychiatry, vol. 13, no. 1, p. 371, Dec 01 2023, doi: 10.1038/s41398-023-02674-1.
J. Bartolotti, J. A. Sweeney, and M. W. Mosconi, "Functional brain abnormalities associated with comorbid anxiety in autism spectrum disorder" Dev Psychopathol, vol. 32, no. 4, pp. 1273–1286, Oct 2020, doi: 10.1017/S0954579420000772.
S. Rezaei et al., "Machine learning applied to functional magnetic resonance imaging in anxiety disorders" J Affect Disord, vol. 342, pp. 54–62, Dec 01 2023, doi: 10.1016/j.jad.2023.09.006.
N. Sommet and D. Morselli, "Keep Calm and Learn Multilevel Logistic Modeling: A Simplified Three-Step Procedure Using Stata, R, Mplus, and SPSS," International Review of Social Psychology, 2017, doi: 10.5334/irsp.90.

There is NO conflict of interest to disclose.

RubleDiehlSocialPartnerPMAsuppmat3.pdf

Download PDF

Version 1

posted

You are reading this older preprint version

Read the latest preprint version →

Active avoidance under social conditions recruits the anterior cingulate cortex in male and female rats.

Status:

Version 1

Abstract

Figures

Introduction

Materials and Methods

Subjects

Surgery

Behavior

Viruses

Laser delivery

Open field task

Pressing test

Histology

Data Collection and Analysis

Results

Discussion

Declarations

Competing interests

Funding

Author Contributions

Acknowledgements

References

Additional Declarations

Supplementary Files

Status:

Version 1