1. Neuroscience
Download icon

Repeated social defeat stress enhances glutamatergic synaptic plasticity in the VTA and cocaine place conditioning

  1. Claire E Stelly
  2. Matthew B Pomrenze
  3. Jason B Cook
  4. Hitoshi Morikawa  Is a corresponding author
  1. University of Texas, United States
Research Article
  • Cited 16
  • Views 1,564
  • Annotations
Cite this article as: eLife 2016;5:e15448 doi: 10.7554/eLife.15448

Abstract

Enduring memories of sensory cues associated with drug intake drive addiction. It is well known that stressful experiences increase addiction vulnerability. However, it is not clear how repeated stress promotes learning of cue-drug associations, as repeated stress generally impairs learning and memory processes unrelated to stressful experiences. Here, we show that repeated social defeat stress in rats causes persistent enhancement of long-term potentiation (LTP) of NMDA receptor-mediated glutamatergic transmission in the ventral tegmental area (VTA). Protein kinase A-dependent increase in the potency of inositol 1,4,5-triphosphate-induced Ca2+ signaling underlies LTP facilitation. Notably, defeated rats display enhanced learning of contextual cues paired with cocaine experience assessed using a conditioned place preference (CPP) paradigm. Enhancement of LTP in the VTA and cocaine CPP in behaving rats both require glucocorticoid receptor activation during defeat episodes. These findings suggest that enhanced glutamatergic plasticity in the VTA may contribute, at least partially, to increased addiction vulnerability following repeated stressful experiences.

https://doi.org/10.7554/eLife.15448.001

eLife digest

Daily stress increases the likelihood that people who take drugs will become addicted. A very early step in the development of addiction is learning that certain people, places, or paraphernalia are associated with obtaining drugs. These ‘cues’ – drug dealers, bars, cigarette advertisements, etc. – become powerful motivators to seek out drugs and can trigger relapse in recovering addicts. It is thought that learning happens when synapses (the connections between neurons in the brain) that relay information about particular cues become stronger. However, it is not clear how stress promotes the learning of cue-drug associations.

Stelly et al. investigated whether repeated episodes of stress make it easier to strengthen synapses on dopamine neurons, which are involved in processing rewards and addiction. For the experiments, rats were repeatedly exposed to a stressful situation – an encounter with an unfamiliar aggressive rat – every day for five days. Stelly et al. found that these stressed rats formed stronger associations between the drug cocaine and the place where they were given the drug (the cue). Furthermore, a mechanism that strengthens synapses was more sensitive in the stressed rats than in unstressed rats. These changes persisted for 10-30 days after the stressful situation, suggesting that stress might begin a period of time during which the individual is more vulnerable to addiction.

The experiments also show that a hormone called corticosterone – which is released during stressful experiences – is necessary for stress to trigger the changes in the synapses and behavior of the rats. However, corticosterone must work with other factors because giving this hormone to unstressed rats was not sufficient to trigger the changes seen in the stressed rats. Future experiments will investigate what these other stress factors are and how they work together with corticosterone.

https://doi.org/10.7554/eLife.15448.002

Introduction

Humans with a history of stressful or traumatic experiences are more prone to develop substance use disorders (Sinha, 2008). Adverse experience recruits the hypothalamic-pituitary-adrenal (HPA) axis stress response, culminating in release of glucocorticoids that enables the body to cope with insults to homeostasis (Munck et al., 1984). In rodent models, repeated activation of the stress response typically disrupts learning and cognition [e.g., spatial learning (Conrad et al., 1996), working memory (Mizoguchi et al., 2000), and cognitive flexibility (Liston et al., 2006)]. In contrast to these deficits, prior stress enhances the learning of Pavlovian cue-outcome associations driven by rewarding stimuli, assessed with conditioned place preference (CPP) (Kreibich et al., 2009; Burke et al., 2011; Chuang et al., 2011), or aversive/stressful stimuli, assessed with fear conditioning (Conrad et al., 1999; Sandi et al., 2001; Suvrathan et al., 2014). These effects of stress may have arisen from evolutionary pressure to rapidly acquire information predicting food, shelter, and predator threat during periods of duress. Augmented Pavlovian reward learning mechanisms in stressed individuals may also heighten susceptibility to addiction, as acquisition of cue-drug associations is a crucial early step in drug use, and powerful, enduring memories of drug-associated cues trigger craving and relapse as recreational use progresses to addiction (Hyman et al., 2006). However, it is not clear how repeated stressful experience promotes the learning of cue-drug/reward associations, as repeated stress is generally detrimental to synaptic plasticity underlying learning and memory unrelated to stressful events (Kim and Diamond, 2002; Joels et al., 2006; Schwabe et al., 2012; Chattarji et al., 2015).

The mesolimbic dopamine system originating in the ventral tegmental area (VTA) is critical for reward processing. VTA dopamine neurons tonically fire action potentials (APs) at 1–5 Hz, while responding to unexpected rewards with phasic burst firing (2–10 APs at 10–50 Hz). These dopamine neuron responses are hypothesized to drive the learning of Pavlovian cue-reward associations (Tsai et al., 2009; Darvas et al., 2014). Intriguingly, over the course of repeated cue-reward pairing, dopamine neurons acquire a conditioned burst response to reward-predictive cues, which is thought to encode the positive motivational valence of those cues and to invigorate reward-seeking behavior (Schultz, 1998; Berridge et al., 2009; Bromberg-Martin et al., 2010).

Glutamatergic inputs activating NMDA receptors (NMDARs) drive the transition from tonic firing to bursting in dopamine neurons (Overton and Clark, 1997; Zweifel et al., 2009; Wang et al., 2011); therefore, potentiation of cue-driven NMDAR inputs may contribute to the acquisition of conditioned bursting. Indeed, NMDAR-mediated transmission undergoes long-term potentiation (LTP) when cue-like glutamatergic input stimulation is repeatedly paired with reward-like bursting in dopamine neurons (Harnett et al., 2009). LTP induction requires amplification of burst-evoked Ca2+ signals by preceding activation of group I metabotropic glutamate receptors (mGluRs; more specifically mGluR1) coupled to the generation of inositol 1,4,5-triphosphate (IP3). Here, IP3 receptors (IP3Rs) detect the coincidence of IP3 generated by glutamatergic input activating mGluRs and burst-driven Ca2+ entry. IP3 enhances Ca2+-induced activation of IP3Rs by promoting access to the stimulatory Ca2+ sites, thereby promoting Ca2+-induced Ca2+ release from intracellular stores (Taylor and Laude, 2002). In this study, we demonstrate that repeated social defeat stress (1) enhances NMDAR LTP in the VTA via an increase in IP3 sensitivity of IP3Rs and (2) promotes acquisition of cocaine CPP in behaving rats, and both of these effects require glucocorticoid action during defeat stress.

Results

Repeated social stress increases mGluR-dependent facilitation of burst-evoked Ca2+ signals

NMDAR LTP induction requires mGluR/IP3-induced facilitation of burst-evoked Ca2+ signals (Harnett et al., 2009). Therefore, we first examined the effect of the group I mGluR agonist DHPG (1 µM; 5-min perfusion) on burst-evoked Ca2+ signals, assessed by the size of Ca2+-activated SK currents (termed burst IK(Ca)) in control and stressed animals. Rats were unhandled, handled, or socially defeated (at the end of the dark cycle) for 1, 5, or 10 consecutive days, and VTA slices were prepared 1–2 days after the final handling/defeat session. The magnitude of DHPG effect on IK(Ca) was significantly larger in animals that underwent 5 or 10 days of defeat stress compared to unhandled and handled controls, whereas a single defeat session failed to alter the DHPG effect (Figures 1A and B). There was no significant difference between unhandled and handled controls. The effect of stress plateaued by 5 days, as comparable enhancement of DHPG effect was observed after 10-day defeat. Basal burst IK(Ca) was consistent across groups (Figure 1C), suggesting no alterations in AP-evoked Ca2+ influx. DHPG-induced inward currents, which are independent of Ca2+ signaling (Guatteo et al., 1999), were not affected (Figure 1D); thus the stress-induced increase in IK(Ca) facilitation results from changes in IP3 signaling downstream of mGluRs.

mGluR-dependent facilitation of burst-evoked Ca2+ signals is enhanced after repeated social defeat.

(A) Example traces (left) and summary time graph (right) illustrating the facilitating effect of DHPG (1 µM) on burst IK(Ca) in neurons from unhandled rats (traces not shown), rats handled for 5 days, and rats that underwent social defeat for 1, 5, or 10 days. (B) Summary bar graph showing the magnitude of DHPG-induced burst IK(Ca) facilitation (unhandled: 20 cells from 12 rats, handled: 20 cells from 13 rats, 1 day defeat: 19 cells from 11 rats, 5 day defeat: 21 cells from 13 rats, 10 day defeat: 19 cells from 10 rats; F4,94 = 6.19, p<0.001, one-way ANOVA). *p<0.05, **p<0.01 (Bonferroni post hoc test). (C) The size of basal burst IK(Ca) was not altered by social defeat. (D) DHPG-induced inward currents were not affected by social defeat. Inset: Example traces of DHPG-induced currents from unhandled and 5-day defeated rats (5-min DHPG perfusion at the horizontal bar). (E) Summary graph depicting DHPG effect on burst IK(Ca) after different intervals following 5-day social defeat. Data in 6–7 weeks old unhandled and 1–2 day interval groups were from those in panels AD (6–7 weeks old unhandled: 20 cells from 12 rats, 1–2 day interval: 21 cells from 13 rats, 10-day interval: 17 cells from 8 rats, 30-day interval: 16 cells from 9 rats, 10–11 weeks old unhandled: 15 cells from 7 rats).

https://doi.org/10.7554/eLife.15448.003

Next, to examine the persistence of repeated stress effect, the interval between the last social defeat session and recording was prolonged to 10 and 30 days. Although stress-induced enhancement displayed gradual recovery, DHPG effect was still elevated after 30 days compared to age-matched controls (Figure 1E). Subsequent electrophysiology experiments were performed in 5-day defeated rats (with 1–2 day interval) and controls (unhandled and handled controls combined).

Protein kinase a mediates IP3R sensitization in socially defeated animals

To directly examine alterations in IP3 signaling, we applied different concentrations of IP3 (expressed in µM·µJ; see Methods and materials) into the cytosol using flash photolysis of caged IP3, and IP3R-mediated Ca2+ release was assessed by flash-evoked SK currents(IIP3) (Figure 2A). The average IP3 concentration-response curve displayed a leftward shift in defeated rats compared to controls (Figure 2B). Accordingly, the average EC50 valuewas significantly smaller in the defeated group (Figure 2C). Maximal IIP3 amplitude did not differ between groups (Figure 2D), indicating a change in the potency, but not the efficacy, of IP3 in eliciting Ca2+ release.

PKA activity maintains increased IP3R sensitivity in socially defeated rats.

(A) Example traces of IIP3 evoked by different concentrations of IP3 (2000, 6000, 16000, 48000, and 140000 µM·µJ) in control and defeated rats. (B) Averaged IP3 concentration-response curves from control and defeated rats. IIP3 amplitudes were normalized to the maximal value (estimated from fit to a logistic equation) in each cell. Recordings were made with normal internal solution or with PKI (control: 12 cells from 8 rats, defeat: 12 cells from 7 rats, control + PKI: 15 cells from 9 rats, defeat + PKI: 14 cells from 8 rats; treatment (defeat/PKI): F3,196 = 4.88, p<0.01; IP3 concentration: F4,196 = 1214, p<0.001; treatment × IP3 concentration: F12,196 = 4.42, p<0.001; mixed two-way ANOVA). ***p<0.001 vs. control (Bonferroni post hoc test). Lines represent logistic fit to the averaged data in each group. (C) Summary bar graph depicting the average EC50 values (EC50 determined in each cell) in the 4 groups shown in (B) (defeat: F1,49 = 5.11, p<0.05; PKI: F1,49 = 5.11, p<0.05; two-way ANOVA). *p<0.05 (Bonferroni post hoc test). (D) The maximal IIP3 amplitude was not affected by social defeat experience or PKI during recording.

https://doi.org/10.7554/eLife.15448.004

Protein kinase A (PKA)-dependent phosphorylation of IP3Rs increases their IP3 sensitivity (Wagner et al., 2008). To determine the involvement of PKA in stress-induced IP3R sensitization, the effect of a selective peptide inhibitor of PKA, PKI-(6–22)-amide (PKI; 200 µM, loaded into the cytosol via the whole-cell pipette for >15–20 min after break-in), was tested. PKI reversed the leftward shift in IP3 concentration-response curve, and thus the decrease in EC50 value, in stressed animals, while having no significant effect on maximal IIP3 amplitude (Figures 2B–D). PKI had no effect in the control group, suggesting low basal PKA activity in non-stressed animals. It should be noted that PKI eliminated the difference in IP3 potency between the two groups. These data indicate that repeated social stress sensitizes IP3Rs via a PKA-dependent mechanism.

Repeated social stress enhances NMDAR LTP

We next examined whether repeated social defeat affects NMDAR LTP induction, which requires mGluR/IP3-dependent facilitation of burst-evoked Ca2+ signals and is gated by PKA (Harnett et al., 2009). Application of a low concentration of IP3 preceding APs can effectively facilitate IK(Ca) (Cui et al., 2007; Ahn et al., 2010; Bernier et al., 2011). Thus, the LTP induction protocol consisted of applying a low concentration of IP3 (250 µM·µJ) 50 ms prior to simultaneous pairing of a burst with a brief train of synaptic stimuli (Figure 3A), the latter being necessary to activate NMDARs at stimulated synapses at the time of burst for LTP induction (Harnett et al., 2009; Whitaker et al., 2013). This induction protocol produced little LTP in control animals, while large LTP was induced in defeated animals (Figure 3B and C). IP3 application, which caused little IIP3 by itself, caused facilitation of burst IK(Ca) (assessed immediately before LTP induction), which was significantly larger in cells from defeated animals (Figure 3D). Furthermore, the magnitude of LTP was positively correlated with that of IP3-induced facilitation of IK(Ca) across neurons from both groups (Figure 3E). Robust LTP was induced in control rats when a higher IP3 concentration (500 µM·µJ), which produced larger IK(Ca) facilitation, was used during induction (Figure 3—figure supplement 1). These results suggest that the enhanced LTP in defeated rats is a consequence of increased IP3R sensitivity enabling greater facilitation of burst-evoked Ca2+ signals.

Figure 3 with 1 supplement see all
NMDAR-mediated transmission is more susceptible to LTP induction after social defeat.

(A) Example experiments to induce NMDAR LTP in neurons from control and defeated rats. Time graphs of NMDAR EPSCs are shown with example traces at times indicated by numbers (baseline: gray, post-induction: black). The LTP induction protocol (IP3-synaptic stimulation-burst combination; illustrated in top inset) was delivered after 10-min baseline recording (at arrow). (B) Summary time graph of baseline-normalized NMDAR EPSCs in LTP experiments (control: 7 cells from 7 rats, defeat: 7 cells from 7 rats). (C) Summary of NMDAR LTP magnitude in control and defeated rats (t12 = 3.93, **p<0.01, unpaired t-test). (D) Example traces (left; from the defeated rat shown in A) and summary (right) of IK(Ca) facilitation by IP3 assessed before LTP induction (t12 = 4.65, ***p<0.001, unpaired t-test). (E) The magnitude of NMDAR LTP is plotted versus the magnitude of IP3-induced facilitation of IK(Ca). Dashed line is a linear fit to all data points from both control and defeated rats.

https://doi.org/10.7554/eLife.15448.005

It has been reported that repeated stress alters NMDAR expression in certain brain areas (Fitzgerald et al., 1996; Yuen et al., 2012; Costa-Nunes et al., 2014; Chattarji et al., 2015). We found that bath application of NMDA (10 µM) produced comparable inward currents in control and defeated rats (Figure 4); thus repeated defeat stress caused no significant changes in global NMDAR-mediated excitation.

Summary time graph depicting inward currents induced by 1-min perfusion of NMDA (10 μM) in VTA dopamine neurons from control (8 cells from 3 rats) and 5 day defeated rats (6 cells from 2 rats).
https://doi.org/10.7554/eLife.15448.007

Repeated stress appears to differentially modulate tonic firing of VTA dopamine neurons recorded in vivo, as both an increase and decrease have been reported with different stress paradigms (Cao et al., 2010; Valenti et al., 2012; Tye et al., 2013). However, repeated social defeat failed to alter tonic firing measured in ex vivo slices (Figure 5A). Furthermore, the amplitude of hyperpolarization-activated cationic currents (Ih), which contribute to intrinsic dopamine neuron pacemaker activity (Neuhoff et al., 2002), was not affected (Figure 5B).

Tonic firing is unaltered by social defeat.

(A) Example traces (left) and summary (right) of tonic firing frequency in VTA neurons from control and defeated rats (control: 15 cells from 3 rats, 5 defeats: 9 cells from 3 rats; t22 = 0.066, p=0.95, unpaired t-test). In these experiments, loose-patch recordings (<20 MΩ seal) were made using pipettes filled with 150 mM NaCl to monitor tonic pacemaker firing. (B) Example traces (left; voltage step depicted at bottom) and summary (right) of Ih currents recorded in cells from animals that underwent control procedures or 1, 5, or 10 days of defeat (data were obtained from the same cells shown in Figures 1A–D; F4,94 = 2.01, p=0.10, one-way ANOVA).

https://doi.org/10.7554/eLife.15448.008

Glucocorticoid receptor activation is necessary but not sufficient for stress-induced IP3R sensitization

A major consequence of stress-induced HPA axis activation is the secretion of glucocorticoids (corticosterone in rodents) into the blood (Munck et al., 1984). Thus, we sought to determine whether corticosterone, which readily crosses the blood-brain barrier, is involved in the increase in IP3R sensitivity with repeated stress. Corticosterone activates both glucocorticoid receptors (GRs) and mineralocorticoid receptors (MRs); however, MRs are typically saturated by circadian fluctuations in corticosterone, while lower-affinity GRs are activated by levels attained with stress (Joels and de Kloet, 1994). Therefore, the role of GRs was examined by treating rats with the antagonist mifepristone (40 mg/kg, i.p.) or vehicle 30 min prior to each defeat session. The DHPG effect on burst IK(Ca) was significantly enhanced by repeated defeat in the vehicle-treated group, while there was no effect of stress in the mifepristone-treated group (Figure 6A). Thus, blockade of GRs during defeat sessions prevented IP3R sensitization. Next, rats were treated with corticosterone (2.5, 5, or 15 mg/kg, i.p.; at the end of the dark cycle) for 5 days (no social defeat). The lowest dose (2.5 mg/kg) produces elevation in blood corticosterone concentration comparable to that evoked by a moderate stressor (Graf et al., 2013), while higher doses (≥10 mg/kg) have been used to simulate severe stress levels (Akirav et al., 2004). None of the tested doses significantly altered the DHPG effect on burst IK(Ca) (Figure 6B). Together, these results show that GR signaling is necessary, but not sufficient, for IP3R sensitization.

Stress-induced, but not cocaine-induced, IP3R sensitization is prevented by GR blockade.

(A) Example traces (left) and summary (right) of DHPG-induced burst IK(Ca) facilitation in neurons from animals that were injected with vehicle or mifepristone before undergoing control handling or social defeat sessions (vehicle + control: 8 cells from 5 rats, vehicle + defeat: 10 cells from 4 rats, mifepristone + control: 10 cells from 5 rats, mifepristone + defeat: 11 cells from 4 rats; defeat × mifepristone: F1,35 = 4.56, p<0.05, two-way ANOVA). *p<0.05 (Bonferroni post hoc test). (B) Summary bar graph showing that repeated corticosterone treatment (once daily for 5 days) failed to affect DHPG-induced burst IK(Ca) facilitation (vehicle: 8 cells from 5 rats, 2.5 mg/kg: 11 cells from 5 rats, 5 mg/kg: 10 cells from 4 rats, 15 mg/kg: 12 cells from 7 rats). (C) Summary bar graph demonstrating that mifepristone pretreatment failed to block the increase in DHPG effect resulting from repeated cocaine treatment (10 mg/kg, i.p., once daily for 5 days) (saline: 17 cells from 7 rats, cocaine: 16 cells from 7 rats, mifepristone + saline: 21 cells from 8 rats, mifepristone + cocaine: 16 cells from 6 rats; cocaine: F1,66 = 11.4, p<0.01, two-way ANOVA). *p<0.05 (Bonferroni post hoc test).

https://doi.org/10.7554/eLife.15448.009

It has been shown that repeated psychostimulant treatment sensitizes IP3Rs and enhances NMDAR LTP (Ahn et al., 2010). As addictive drugs, including psychostimulants, stimulate corticosterone secretion (Armario, 2010), we next examined the role of GR signaling in psychostimulant-induced IP3R sensitization. Rats were treated with mifepristone (40 mg/kg, i.p.) 30 min prior to injection of cocaine (10 mg/kg, i.p.) or saline for 5 days. The effect of DHPG on burst IK(Ca) was significantly larger in cocaine-treated animals, which was not affected by mifepristone pretreatment (Figure 6C). Thus, GR signaling is not involved in psychostimulant-induced IP3R sensitization.

Repeated social stress promotes learning of cocaine-associated cues in a GR-dependent manner

Next, the effect of social defeat stress was tested on acquisition of cocaine CPP, in which animals learn to associate a particular context with drug reward. Acquisition of psychostimulant CPP is inhibited by mGluR1 or NMDAR antagonist in the VTA, while CPP expression is attenuated by NMDAR antagonist, but not by mGluR1 antagonist, in the VTA (Whitaker et al., 2013), supporting the potential role of NMDAR LTP in driving CPP. Rats underwent stress or control procedures for 5 days, then underwent 1-day CPP conditioning with cocaine (5 mg/kg, i.p.). It should be noted that a single psychostimulant treatment does not cause IP3R sensitization (Ahn et al., 2010; Whitaker et al., 2013). Stressed rats displayed robust preference for the cocaine-paired side after 1-day conditioning, while unhandled and handled controls showed small preference (Figure 7A). The 1-day CPP score was significantly larger in stressed rats compared to unhandled and handled controls (Figure 7B). Control rats developed significant cocaine CPP comparable to that observed in stressed rats after 3-day conditioning with the same dose of cocaine (Figure 7—figure supplement 1). These data suggest that repeated defeat experience promotes the rate of learning of cocaine-associated cues.

Figure 7 with 3 supplements see all
Social defeat promotes cocaine-induced CPP via a GR-dependent mechanism.

(A) Summary of changes in the preference for the cocaine-paired side following 1-day conditioning in unhandled, handled, and defeated rats (unhandled: t7 = 2.51, p<0.05; handled: t7 = 1.90, p=0.10; defeat: t6 = 11.0, p<0.001; paired t-test). (B) Summary of 1-day cocaine CPP scores in unhandled, handled, and defeated rats (F2,20 = 25.2, p<0.001, one-way ANOVA). ***p<0.001 (Bonferroni post hoc test). (C) Summary of changes in the preference for the cocaine-paired side following 1-day conditioning in rats pretreated with vehicle or mifepristone 30 min prior to social defeat or handling sessions (vehicle + defeat: t8 = 5.30, p<0.001; mifepristone + defeat: t8 = 1.90, p=0.09; mifepristone + handled: t7 = 0.95, p=0.37; paired t-test). (D) Summary of 1-day cocaine CPP scores in the 3 groups shown in panel C (F2,23 = 4.90, p<0.05, one-way ANOVA). *p<0.05 (Bonferroni post hoc test). None of the treatments in this figure affected the overall activity level during the pretest (Figure 7—figure supplement 2).

https://doi.org/10.7554/eLife.15448.010

Finally, we asked whether GR signaling, which is necessary for IP3R sensitization, also plays a role in promoting cocaine CPP. As in the electrophysiology experiments, rats were treated with mifepristone (40 mg/kg, i.p.) or vehicle 30 min before each social defeat session. An additional group received mifepristone followed by control handling procedure. We found that mifepristone suppressed cocaine CPP in stressed rats to a level comparable to that observed in mifepristone-treated controls (Figure 7C and D). Therefore, GR activation during stress is required for CPP enhancement.

Discussion

Repeated stressful experience leads to metaplasticity, i.e., experience-dependent changes in the capacity of synapses to undergo activity-dependent plasticity (Abraham, 2008), in different brain areas (Kim and Diamond, 2002; Joels et al., 2006; Schwabe et al., 2012; Chattarji et al., 2015). The present study demonstrates that repeated social defeat facilitates the induction of LTP of NMDAR-mediated transmission in VTA dopamine neurons while causing no alterations in global NMDAR-mediated excitation or intrinsic firing activity. Importantly, socially defeated animals display enhanced acquisition of cocaine CPP, a form of Pavlovian conditioning that requires NMDAR-dependent bursting in the VTA (Zweifel et al., 2009; Wang et al., 2011; Whitaker et al., 2013).

Repeated social defeat results in increased sensitivity of IP3Rs, which serve as a coincidence detector of presynaptic activity (causing mGluR-dependent IP3 generation) and postsynaptic bursting (driving Ca2+ influx) during NMDAR LTP induction (Harnett et al., 2009). Inhibition of PKA completely reversed the increase in the potency of IP3, indicating the role of PKA-dependent phosphorylation in stress-induced IP3R sensitization, as has been suggested in previous studies demonstrating similar changes following repeated drug exposure (Ahn et al., 2010; Bernier et al., 2011). It is of note that dopamine neurons in the substantia nigra pars compacta, in contrast to VTA neurons recorded in the present study, display significant PKA-dependent regulation of IP3-induced Ca2+ signaling and NMDAR LTP induction in control rats, which cannot be further enhanced by repeated drug exposure (Harnett et al., 2009; Ahn et al., 2010).

It has been reported that repeated social defeat (10 days) in mice leads to long-lasting (>4 weeks) alterations in gene expression and behavior (e.g., reduced social contact), with little recovery unless treated with antidepressants (Berton et al., 2006; Tsankova et al., 2006). In the present study, mGluR/IP3 action on burst-evoked Ca2+ signals remained elevated for 10–30 days following the 5-day defeat paradigm in rats, although displaying gradual decline during the 30-day stress-free period. It remains to be determined how this recovery is affected by different stress paradigms (e.g., duration or type/severity) and treatments following stress experience.

Mouse studies have also shown that individual animals display different susceptibility to repeated social defeat when assessed by the degree of social avoidance, which correlates with biochemical and physiological changes in the mesolimbic system (Krishnan et al., 2007; Cao et al., 2010). In particular, these studies observed hyperactivity of VTA dopamine neurons, assessed in vivo or ex vivo, associated with an increase in Ih, after 10-day defeat in susceptible mice. However, these parameters were not affected by 5-day social defeat in the current study conducted in rats.

How could repeated stress lead to increased PKA activity regulating IP3Rs in the VTA? Stress, including social defeat, promotes bursting in a subpopulation of VTA dopamine neurons, causing increased dopamine transients in the nucleus accumbens (Anstrom et al., 2009; Brischoux et al., 2009). Bursting also releases dopamine locally from the soma and dendrites, activating somatodendritic D2 autoreceptors (Beckstead et al., 2004). Chronic stimulation of Gi-coupled receptors, such as D2 receptors, is known to upregulate the cAMP-PKA pathway (Hyman et al., 2006), which may contribute to stress-induced IP3R sensitization. Interestingly, intra-VTA blockade of NMDARs during each social defeat episode, which would suppress dopamine neuron bursting, has been shown to prevent repeated stress-induced increases in cocaine self-administration (Covington et al., 2008).

GR signaling during defeat sessions is necessary for the enhancement of IP3R sensitivity. Stress-induced activation of the mesolimbic dopamine system is regulated by glucocorticoids (Marinelli and Piazza, 2002). Recent evidence implicates GRs expressed in projection areas, not in the VTA, in long-term glucocorticoid regulation of dopamine neuron activity (Ambroggi et al., 2009; Butts et al., 2011; Barik et al., 2013). Furthermore, glucocorticoids can also enhance synthesis of corticotropin-releasing factor (CRF), a major stress-related neuropeptide, and activation of CRF neurons in brain areas providing major CRF inputs to the VTA (Makino et al., 1994; Rodaros et al., 2007; Kolber et al., 2008). GR blockade may therefore attenuate CRF-induced excitation of dopamine neurons during stress (Kalivas et al., 1987; Ungless et al., 2003; Wanat et al., 2008; Holly et al., 2015). We found that GR activation alone is not sufficient for IP3R sensitization. Thus, the potential GR mechanisms described above may act to amplify glutamatergic input-driven bursting activity during stress episodes, likely further enhanced by stress-induced activation of noradrenergic inputs stimulating dopamine neurons via α1 adrenergic receptors (Grenhoff et al., 1995; Paladini et al., 2001; Morilak et al., 2005), thereby enabling large local dopamine release in the VTA. In this regard, it is interesting that repeated cocaine treatment was capable of causing similar enhancement of mGluR/IP3 action in a GR-independent manner. Dopamine levels in the VTA caused by cocaine alone are likely sufficient to induce D2-mediated upregulation of the cAMP-PKA pathway.

Increased IP3R sensitivity drives the enhancement of NMDAR LTP induction in socially defeated animals. Our recent study demonstrated the involvement of L-type Ca2+ channels (LTCCs) in NMDAR LTP (Degoulet et al., 2015). Although glucocorticoid-induced upregulation of LTCCs has been reported in the hippocampus and amygdala (Karst et al., 2002; Chameau et al., 2007), pharmacological activation of these channels does not enhance NMDAR LTP in dopamine neurons (Degoulet et al., 2015); thus changes in LTCCs are unlikely to play a role in LTP enhancement.

CPP experiments showed that repeated social defeat promoted acquisition of the preference for contextual cues paired with cocaine experience, in accordance with previous studies demonstrating enhanced drug CPP following a period of repeated stress (Kreibich et al., 2009; Burke et al., 2011). Blockade of the critical components regulating NMDAR LTP induction (i.e., NMDARs, group I mGluRs, PKA, or LTCCs) in the VTA during conditioning has been shown to suppress CPP acquisition (Harnett et al., 2009; Ahn et al., 2010; Whitaker et al., 2013; Degoulet et al., 2015). In the present study, systemic GR blockade during defeat episodes prevented both the enhancement of the LTP induction mechanism and that of cocaine CPP acquisition, consistent with the potential role of NMDAR plasticity in this form of Pavlovian learning. However, enhanced CPP acquisition observed in defeated rats may well be caused by an increase in the primary rewarding action of cocaine itself. The relative contribution of these two possibilities, i.e., enhanced learning mechanism vs. enhanced cocaine reward, remains to be determined.

It is well known that repeated stress impairs LTP of AMPAR-mediated transmission in the hippocampus (Foy et al., 1987; Shors et al., 1989), an effect that requires GR activation during stress (Xu et al., 1998). LTP is similarly impaired in the prefrontal cortex (Goldwater et al., 2009). By contrast, repeated stress leads to enhancement of AMPAR LTP in the lateral amygdala, which underlies Pavlovian fear conditioning driven by stressful/aversive stimuli (Rodriguez Manzanares et al., 2005; Suvrathan et al., 2014). Alterations in the function/expression of NMDARs are implicated in these forms of metaplasticity, as NMDARs play a key role in AMPAR LTP induction (Kim and Diamond, 2002; Chattarji et al., 2015). Here, we described a distinct form of stress-induced metaplasticity in the VTA, i.e., enhancement of mGluR/IP3-dependent NMDAR LTP, which may, at least in part, contribute to the enhanced drug reward-based Pavlovian learning. This may illuminate a key mechanism by which stressful experience increases vulnerability to addiction, a chronic relapsing disorder perpetuated by memories of drug-associated stimuli.

Materials and methods

Animals

Sprague-Dawley rats (Harlan Laboratories, Houston, Texas) were housed in groups of 2–3 on a 12 hr light/dark cycle with food and water available ad libitum. All procedures were approved by the University of Texas Institutional Animal Care and Use Committee.

Resident-intruder social defeat paradigm

Request a detailed protocol

Twelve week-old male resident rats were vasectomized and pair-housed with 6 week-old females. Residents (used for ~8–10 months) were screened for aggression (biting or pinning within 1 min) by introducing a male intruder to the home cage. Intruders and controls were young males (4–5 weeks old at the beginning) housed in groups of 2–3. For defeat sessions, residents and intruders were taken to a darkened procedure room at the end of the dark cycle. Intruders were introduced to residents’ home cages after removing females. Following 5 min of direct contact, a perforated Plexiglass barrier was inserted for 25 min to allow sensory contact. For repeated defeat, intruders underwent one session daily with a novel resident. Handled controls were taken to a darkened procedure room and placed in novel cages for 30 min. Unhandled controls remained undisturbed in the colony. Intruders and controls were housed separately.

In vivo drug treatments

Request a detailed protocol

All drug and vehicle solutions were administered via i.p. injections (1 ml/kg). Mifepristone and corticosterone (both from Tocris Bioscience, Ellisville, Missouri) were dissolved in 30% propylene glycol plus 1% Tween-20 in 0.9% saline. Cocaine-HCl (Sigma-Aldrich, St. Louis, Missouri) was dissolved in 0.9% saline.

Electrophysiology

Request a detailed protocol

Midbrain slices were prepared and recordings were made in the lateral VTA located 50–150 μm from the medial border of the medial terminal nucleus of the accessory optic tract, as in our previous studies (Ahn et al., 2010; Whitaker et al., 2013; Degoulet et al., 2015). Tyrosine hydroxylase-positive neurons in this area (i.e., lateral part of the parabrachial pigmented nucleus) largely project to the ventrolateral striatum (Ikemoto, 2007) and show little VGluT2 coexpression (Trudeau et al., 2014). Putative dopamine neurons in the lateral VTA were identified by spontaneous firing of broad APs (>1.2 ms) at 1–5 Hz in cell-attached configuration and large Ih currents (>200 pA; evoked by a 1.5 s hyperpolarizing step of 50 mV) in whole-cell configuration (Ford et al., 2006; Lammel et al., 2008; Margolis et al., 2008). Cells were voltage-clamped at –62 mV (corrected for –7 mV liquid junction potential).

A 2 ms depolarizing pulse of 55 mV was used to elicit an unclamped AP. For bursts, 5 APs were evoked at 20 Hz. The time integral of the outward tail current, termed IK(Ca) (calculated after removing the 20 ms window following each depolarizing pulse; expressed in pC), was used as a readout of AP-evoked Ca2+ transients, as it is eliminated by TTX and also by apamin, a blocker of Ca2+-activated SK channels (Cui et al., 2007).

Flash photolysis

Request a detailed protocol

Cells were loaded with caged IP3 (50–400 µM; generous gift from Dr. Kamran Khodakhah) through the recording pipette. A UV flash (~1 ms) was applied with a xenon arc lamp driven by a photolysis system (Cairn Research, Faversham, UK). The UV flash was focused through a 60× objective onto a ~350 μm area surrounding the recorded neuron. Photolysis of caged compounds is proportional to the UV flash intensity; therefore, the concentration of IP3 was defined as the product of caged IP3 concentration in the pipette (µM) and flash intensity (µJ) measured at the focal plane of the objective (expressed in µM·µJ).

NMDAR LTP experiments

Request a detailed protocol

Synaptic stimuli were delivered with a bipolar tungsten electrode placed ~50–100 μm rostral to the recorded neuron. To isolate NMDAR EPSCs, recordings were performed in DNQX (10 µM), picrotoxin (100 µM), CGP54626 (50 nM), and sulpiride (100 nM) to block AMPA/kainate, GABAA, GABAB, and D2 dopamine receptors, and in glycine (20 µM) and low Mg2+ (0.1 mM) to enhance NMDAR activation. NMDAR EPSCs were monitored every 20 s. The LTP induction protocol consisted of photolytic application of IP3 (250 µM·µJ) 50 ms prior to the simultaneous delivery of synaptic stimulation (20 stimuli at 50 Hz) and a burst (5 APs at 20 Hz), repeated 10 times every 20 s. LTP magnitude was determined by comparing the average EPSC amplitude 30 min post-induction with the average EPSC amplitude pre-induction (each from a 5 min window).

Place conditioning

Request a detailed protocol

A CPP box (Med Associates, St. Albans, Vermont) consisting of two distinct compartments separated by a small middle chamber was used for conditioning. One compartment had a mesh floor with white walls, while the other had a grid floor with black walls. A discrete cue (painted ceramic weight) was placed in the rear corner of each compartment (black one in the white wall side, white one in the black wall side; Figure 7—figure supplement 3) for further differentiation. One day after undergoing repeated stress or control procedures, rats were pretested for initial side preference by exploring the entire CPP box for 15 min. The percentage of time spent in each compartment was determined after excluding the time spent in the middle chamber. Rats with initial side preference >60% were excluded. Starting the next day, rats were subjected to 1-day or 3-day conditioning, in which they were given a saline injection in the morning and confined to one compartment, then in the afternoon given cocaine (5 mg/kg) and confined to the other compartment (10 min each). Compartment assignment was counterbalanced such that animals had, on average, ~50% initial preference for the cocaine-paired side. A 15 min posttest was performed 1 day after the last conditioning session. The CPP score was determined by subtracting the preference for the cocaine-paired side during pretest from that during posttest. The experimenter performing CPP experiments was blind to animal treatments.

Data analysis

Request a detailed protocol

Data are expressed as mean ± SEM. Statistical significance was determined by Student's t-test or ANOVA followed by Bonferroni post hoc test. Normality of data distribution was confirmed by Kolmogorov-Smirnov test. The difference was considered significant at p<0.05.

References

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
    Drugs of abuse and stress increase the expression of GluR1 and NMDAR1 glutamate receptor subunits in the rat ventral tegmental area: common adaptations among cross-sensitizing agents
    1. LW Fitzgerald
    2. J Ortiz
    3. AG Hamedani
    4. EJ Nestler
    (1996)
    Journal of Neuroscience 16:274–282.
  28. 28
  29. 29
  30. 30
  31. 31
  32. 32
  33. 33
    Group I metabotropic glutamate receptors mediate an inward current in rat substantia nigra dopamine neurons that is independent from calcium mobilization
    1. E Guatteo
    2. NB Mercuri
    3. G Bernardi
    4. T Knöpfel
    (1999)
    Journal of Neurophysiology 82:1974–1981.
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
    Neurochemical and behavioral effects of corticotropin-releasing factor in the ventral tegmental area of the rat
    1. PW Kalivas
    2. P Duffy
    3. LG Latimer
    (1987)
    The Journal of Pharmacology and Experimental Therapeutics 242:757–763.
  41. 41
  42. 42
  43. 43
  44. 44
  45. 45
  46. 46
  47. 47
  48. 48
  49. 49
  50. 50
  51. 51
    Chronic stress induces impairment of spatial working memory because of prefrontal dopaminergic dysfunction
    1. K Mizoguchi
    2. M Yuzurihara
    3. A Ishige
    4. H Sasaki
    5. DH Chui
    6. T Tabira
    (2000)
    The Journal of Neuroscience 20:1568–1574.
  52. 52
  53. 53
  54. 54
    I(h) channels contribute to the different functional properties of identified dopaminergic subpopulations in the midbrain
    1. H Neuhoff
    2. A Neu
    3. B Liss
    4. J Roeper
    (2002)
    The Journal of Neuroscience 22:1290–1302.
  55. 55
  56. 56
  57. 57
  58. 58
  59. 59
  60. 60
    Predictive reward signal of dopamine neurons
    1. W Schultz
    (1998)
    Journal of Neurophysiology 80:1–27.
  61. 61
  62. 62
  63. 63
  64. 64
  65. 65
  66. 66
  67. 67
  68. 68
  69. 69
  70. 70
  71. 71
  72. 72
  73. 73
  74. 74
  75. 75
  76. 76
  77. 77
  78. 78

Decision letter

  1. Naoshige Uchida
    Reviewing Editor; Harvard University, United States

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

Thank you for submitting your article "Repeated social defeat stress enhances synaptic plasticity that drives learning of drug-associated cue valence" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by a Reviewing Editor, Naoshige Uchida, and a Senior Editor. One of the three reviewers has agreed to share his identity: Jochen Roeper.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

This study found that repeated social defeat stress enhances long-term potentiation of NMDA receptor-mediated glutamate transmission in the ventral tegmental area (VTA). The authors also found that protein kinase A-dependent sensitization of IP3-mediated calcium release plays a key role in this process, and that this plasticity is triggered by stress-related glucocortcoids (GR). The authors also present evidence indicating that cocaine-induced conditioned place preference (CPP) conditioning is enhanced after repeated social defeat stress. The referees found this study of significance and timely, addressing important questions relevant to the mechanism underlying drug addiction. The manuscript is well written and the data are generally sound. However, there are two major issues upon which we would like to see your response.

Essential revisions:

1) It was pointed out that the CPP experiment is difficult to interpret without further control data. Specifically, the current protocol (1 day conditioning with 5mg/kg cocaine) does not induce CPP robustly in control animals (Figure 7A, B). This may be because the dose of cocaine used, together with the other experimental conditions, is not rewarding at all, or because conditioning was not enough. Depending on these possibilities, the interpretation of the result can be very different. If the authors claim that social defeat stress enhanced learning, the authors should show that this dose of cocaine is rewarding, and extended conditioning actually causes CPP in control mice. Thus, we would like to see the data indicating that positive CPP can be observed in control animals with the same dose of cocaine.

2) The authors interpret the result as "enhanced learning of associated cue valence", however, it is possible that the increased CPP is due to an enhanced rewarding effect of cocaine. Please interpret the results more carefully so that it becomes clearer that the latter interpretation is plausible.

Reviewer #1:

This paper is well written and effectively discusses a series a thoughtful experiments surrounding the authors' central question. Namely, that GR receptors inDA neurons of the VTA shape changes in synaptic strength following an ethologically relevant model of stress and this is accompanied by enhanced place conditioning for cocaine. However, there are major issues that should be addressed prior to publication.

Perhaps the most serious issue, in my opinion, is the interpretation of the CPP data. Here the authors observe a CPP in stressed rats after a single conditioning trial, however, a similar CPP is not developed in control animals.

1) Without a positive control (i.e. an observed CPP in controls), these data are not interpretable. I understand the journal's mandate to not require additional experiments, but this is one of those cases where this is critical. Are CPPs generally observed after a single conditioning session with a 5mg/kg IP dose? If not, the authors should extend the conditioning protocol until a CPP can be observed in all animals.

2) The authors interpret CPP as "enhanced learning of associated cue valence", however, it could very well also be just enhanced reward of cocaine. CPP technically measures reward, so it is unclear why the authors did not acknowledge the possibility (in my opinion, strong likelihood) that social defeat stress actually enhances cocaine reward rather than Pavlovian associations. Please revise thoroughly to account for this alternative possibility.

3) The authors overstate the involvement of dopamine in learning – this is a well-supported theory, but certainly only one of several. In the Introduction the authors state, "…dopamine neuron responses drive the learning of Pavlovian cue-reward associations", this should be adjusted to reflect that this is a theory, perhaps, " dopamine neuron responses are hypothesized to drive the learning…".

4) The involvement of LTP in the CPP behavioral data is too speculative, for example in the first paragraph of the subsection “Repeated Social Stress Promotes Learning of Cocaine-Associated Cues in a GR-Dependent Manner”; please revise. Specifically, there is not a single piece of empirical evidence that the observed LTP "promotes" the Pavlovian associations that give to rise to CPP. This should be thoroughly re-written, to reflect the findings as they actually are. Namely, that LTP enhancement accompanies social defeat stress (or it could in fact be an epiphenomenon of the defeat). To show unambiguously that it "promotes" the behavioral effects observed, would require a whole new set of experiments, which fall outside of the scope of this study.

5) Animal/group numbers were poorly described. Please clarify these as well as the degrees of freedom for all experiments.

Reviewer #2:

This is an excellent study which has discovered that chronic social defeat stress in rats leads to a persistent enhancement of in vitro NMDA-LTP in DA VTA neurons, without changing intrinsic firing properties. The Morikawa lab has previously discovered and carefully characterized this type of plasticity. Here, they also show that PKA-dependent sensitization of IP3-mediated calcium release is a key component of the underlying signalling process and that this plasticity is triggered by stress-related glucocortcoids (GR). Importantly, they give also some evidence that the GR-dependent enhanced NMDA-LTP might be relevant for potentiating cocaine-induced condition place-preference learning. The behavioral data do however not prove that NMDA-LTP itself in causal for the learning improvement, but it a likely candidate.

While the Discussion covers well the changes in signaling processes induced by stress that might converge on enhanced NMDA-LTP, it avoids a very important issue regarding the different outcome of chronic social defeat stress (resilient and non-resilient) and the correlative and even causal link to changes in DA VTA neuronal firing. In contrast, this study sees no changes in DA VTA firing, which might imply that these were highly resilient rats. It would be important to provide additional data on the post-stress behaviors of the rats used, e.g. reduced social approach and/or anhedonia that have been associated with altered in vivo and vitro firing pattern of DA VTA – enhanced mean frequencies and increased bursting. Interestingly, the molecular mechanism identified here would be a very good candidate for inducing these activity pattern in DA VTA neurons of non-resilient post-defeat animals. In addition to providing this information, the authors should more actively engage with this prominent literature.

Reviewer #3:

In general, this is an interesting manuscript. The data is sound, and the central message is timely and of significance.

1) If recruitment of Ca2+-activated K+ currents is enhanced after stress, then there may be a change in firing pattern and waveform (i.e. AHP) of glutamate-evoked spiking of DA neurons, but not tonic as observed. Did the authors consider examining this measure?

2) Is the enhancement of IP3 sensitivity after stress similar for all Gq-coupled GPCRs? Or is only mGluR1 function increased due to the IP3 enhancement? This may have important implications for understanding functional consequences.

3) An additional experiment with a more direct link between the cocaine CPP and changes in IP3 signaling in VTA DA neurons after stress would help strengthen the study.

[Editors' note: further revisions were requested prior to acceptance, as described below.]

Thank you for submitting your article "Repeated social defeat stress enhances synaptic plasticity that drives learning of drug-associated cue valence" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by a Reviewing Editor, Naoshige Uchida, and a Senior Editor. One of the three reviewers has agreed to share his identity: Jochen Roeper.(Reviewer #2).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

While the reviewers appreciated that the addition of the new conditioned place preference (CPP) experiment has greatly improved the manuscript, reviewer #1 remained unsatisfied by the authors' interpretations.

Essential revisions:

Specifically, the authors should discuss, in a more balanced way, the possibility that cocaine exposures enhanced rewarding effects of cocaine not merely the learning of it. Additionally, the referee remained unsatisfactory regarding the discussion on the link between the LTP and CPP behavioral data. The discussion must be more explicit that a causal demonstration will be required to establish the link.

The reviewers agreed that these are interpretational issues and do not require additional experiments. Nonetheless, the manuscript will greatly benefit from discussing these in a more careful way. For more details, please refer to the referee's comments that are copied below:

Reviewer #1:

The current version of the manuscript is improved by the addition of the new CPP results. However, issues still remain with the authors' interpretation of their data that make this manuscript incomplete.

1) I remain unsatisfied that the CPP protocol utilized here measured learning of cocaine-associated cues and not cocaine reward.

The authors now acknowledge that CPP measures reward in the eighth paragraph of the Discussion, however, this is quickly overshadowed by a qualifying statement which is contradictory:

"Enhanced CPP acquisition observed in defeated rats might be due to an increase in the "primary rewarding" action of cocaine. Although this possibility cannot be ruled out, it should be noted that socially defeated rats displayed no significant increases in the overall activity level during the cocaine conditioning sessions (Figure 7—figure supplement 1), suggesting that the "locomotor stimulating" action of cocaine was not altered by repeated defeat experience."

This statement is very misleading. Are the authors suggesting that because stressed mice did not show enhanced cocaine locomotor activity compared to control mice, that they likely did not have enhanced cocaine reward? If in fact this is what the authors are trying to convey, then I would surmise that they are equating locomotor activation with reward, which has repeatedly been shown not to be the case (for example, see PMID: 22197517).

As such, the interpretation of the data continues to the main problem of the paper. Place conditioning is a test of reward and/or aversion and not a test that provides a clear examination of drug-associated cues without more specific controls in place. Perhaps if the authors had administered drug in the CPP chamber or, better yet, had the animals self-administer the drug in the CPP chamber then this interpretation would be more acceptable. As is, the question they are answering is one of subjective cocaine experience and not of learning cues associated with drug taking.

In fact, the citations the authors employ to justify their use of CPP clearly state that CPP is a measure of reward: "…Pavlovian cue-outcome associations driven by rewarding stimuli, assessed with conditioned place preference (CPP) (Kreibich et al., 2009; Burke et al., 2011; Chuang et al., 2011)…".

Kreibich AS, Briand L, Cleck JN, Ecke L, Rice KC, Blendy JA (2009) Stress-induced potentiation of cocaine reward: a role for CRF R1 and CREB. Neuropsychopharmacology 34:2609-2617.

Burke AR, Watt MJ, Forster GL (2011) Adolescent social defeat increases adult amphetamine conditioned place preference and alters D2 dopamine receptor expression. Neuroscience 197:269-279.

"The effects of adolescent stress on drug preferences in adulthood were studied."

Chuang JC, Perello M, Sakata I, Osborne-Lawrence S, Savitt JM, Lutter M, Zigman JM (2011) Ghrelin mediates stress-induced food-reward behavior in mice. The Journal of clinical investigation 121:2684-2692.

Moreover, the authors refer to the cocaine chamber as cocaine-associated cues, however, these contextual cues are in fact paired with the experience of cocaine and not cocaine administration itself, which, presumably, takes place elsewhere. This should be made clear. It is as if the authors lack some of the fundamental knowledge of Pavlovian conditioning or choose to ignore it to fit the framework of the paper. For example:

"Repeated social defeat enhanced learning of cocaine-associated cues assessed with a CPP paradigm."

2) The authors have done little to assuage initial concerns that the involvement of LTP in the CPP behavioral data is greatly overstated.

The data clearly show that social stress enhances LTP and this is glucocorticoid-dependent. Enhancement of CPP seen in stressed rats is also glucocorticoid-dependent. These are important findings and the experiments are performed elegantly. However, there is no assessment that unambiguously shows that enhanced LTP is required for the observed facilitation of cocaine CPP. As such, there are a large number of instances throughout the manuscript wherein the authors overstate their findings (as was mentioned in the original review).

For example, the title of this manuscript is unsupported by its contents:

"Repeated social defeat stress enhances synaptic plasticity that drives learning of drug- associated cue valence"

The authors do not provide any empirical data to support that LTP drives CPP.

In the Abstract, this trend continues when the authors state:

"Here, we show that repeated social defeat stress in rats results in persistent enhancement of long-term potentiation (LTP) of NMDA receptor-mediated glutamatergic transmission in the ventral tegmental area (VTA), thereby promoting the learning of cocaine-associated contextual cues assessed with a conditioned place preference (CPP) paradigm."

and…

"…plasticity in the VTA as the critical mechanism by which repeated stress increases addiction vulnerability."

Likewise, statements such as the following are far too speculative (i.e., not a single piece of evidence is provided to support this claim):

"Here, mGluR1/NMDAR blockade would suppress CPP acquisition via inhibiting LTP induction at glutamatergic inputs activated by contextual cues of the CPP box, while blocking potentiated NMDAR-mediated excitation at those inputs would interfere with CPP expression".

Reviewer #2:

I am happy with the revised version and believe that all essential points have been addressed adequately.

Reviewer #3:

I understand the reluctance of the authors to add many additional experiments.

With that in mind, I am OK for this article to be accepted for publication, should the other reviewers feel the same.

https://doi.org/10.7554/eLife.15448.014

Author response

Essential revisions:

1) It was pointed out that the CPP experiment is difficult to interpret without further control data. Specifically, the current protocol (1 day conditioning with 5mg/kg cocaine) does not induce CPP robustly in control animals (Figure 7A, B). This may be because the dose of cocaine used, together with the other experimental conditions, is not rewarding at all, or because conditioning was not enough. Depending on these possibilities, the interpretation of the result can be very different. If the authors claim that social defeat stress enhanced learning, the authors should show that this dose of cocaine is rewarding, and extended conditioning actually causes CPP in control mice. Thus, we would like to see the data indicating that positive CPP can be observed in control animals with the same dose of cocaine.

We have performed 3-day CPP conditioning with 5 mg/kg cocaine in both control and socially defeated rats (Figure 7—figure supplement 1). We found that control rats displayed robust CPP comparable to defeated rats when the number of conditioning sessions was increased from one to three.

2) The authors interpret the result as "enhanced learning of associated cue valence", however, it is possible that the increased CPP is due to an enhanced rewarding effect of cocaine. Please interpret the results more carefully so that it becomes clearer that the latter interpretation is plausible.

This possibility is explicitly discussed in the eighth paragraph of the Discussion.

Reviewer #1:

This paper is well written and effectively discusses a series a thoughtful experiments surrounding the authors' central question. Namely, that GR receptors inDA neurons of the VTA shape changes in synaptic strength following an ethologically relevant model of stress and this is accompanied by enhanced place conditioning for cocaine. However, there are major issues that should be addressed prior to publication.

Perhaps the most serious issue, in my opinion, is the interpretation of the CPP data. Here the authors observe a CPP in stressed rats after a single conditioning trial, however, a similar CPP is not developed in control animals.

1) Without a positive control (i.e. an observed CPP in controls), these data are not interpretable. I understand the journal's mandate to not require additional experiments, but this is one of those cases where this is critical. Are CPPs generally observed after a single conditioning session with a 5mg/kg IP dose? If not, the authors should extend the conditioning protocol until a CPP can be observed in all animals.

This issue is discussed above in Essential revisions: item 1.

2) The authors interpret CPP as "enhanced learning of associated cue valence", however, it could very well also be just enhanced reward of cocaine. CPP technically measures reward, so it is unclear why the authors did not acknowledge the possibility (in my opinion, strong likelihood) that social defeat stress actually enhances cocaine reward rather than Pavlovian associations. Please revise thoroughly to account for this alternative possibility.

This issue is addressed in Essential revisions: item 2.

3) The authors overstate the involvement of dopamine in learning – this is a well-supported theory, but certainly only one of several. In the Introduction the authors state, "…dopamine neuron responses drive the learning of Pavlovian cue-reward associations", this should be adjusted to reflect that this is a theory, perhaps, " dopamine neuron responses are hypothesized to drive the learning…".

This sentence was revised as suggested.

4) The involvement of LTP in the CPP behavioral data is too speculative, for example in the first paragraph of the subsection “Repeated Social Stress Promotes Learning of Cocaine-Associated Cues in a GR-Dependent Manner”; please revise. Specifically, there is not a single piece of empirical evidence that the observed LTP "promotes" the Pavlovian associations that give to rise to CPP. This should be thoroughly re-written, to reflect the findings as they actually are. Namely, that LTP enhancement accompanies social defeat stress (or it could in fact be an epiphenomenon of the defeat). To show unambiguously that it "promotes" the behavioral effects observed, would require a whole new set of experiments, which fall outside of the scope of this study.

We have revised the parts describing the potential role of NMDAR LTP in the VTA in CPP (subsection “Repeated Social Stress Promotes Learning of Cocaine-Associated Cues in a GR-Dependent 178 Manner”, first paragraph; Discussion, last paragraph).

5) Animal/group numbers were poorly described. Please clarify these as well as the degrees of freedom for all experiments.

These types of information are all provided in the figure legends.

Reviewer #2:

While the Discussion covers well the changes in signaling processes induced by stress that might converge on enhanced NMDA-LTP, it avoids a very important issue regarding the different outcome of chronic social defeat stress (resilient and non-resilient) and the correlative and even causal link to changes in DA VTA neuronal firing. In contrast, this study sees no changes in DA VTA firing, which might imply that these were highly resilient rats. It would be important to provide additional data on the post-stress behaviors of the rats used, e.g. reduced social approach and/or anhedonia that have been associated with altered in vivo

We have added a paragraph addressing this issue (Discussion, fourth paragraph).

Reviewer #3:

In general, this is an interesting manuscript. The data is sound, and the central message is timely and of significance.

1) If recruitment of Ca2+-activated K+ currents is enhanced after stress, then there may be a change in firing pattern and waveform (i.e. AHP) of glutamate-evoked spiking of DA neurons, but not tonic as observed. Did the authors consider examining this measure?

The size of the SK current evoked by bursts, termed basal burst-evoked IK(Ca), was not altered by repeated social defeat stress (Figure 1C).

2) Is the enhancement of IP3 sensitivity after stress similar for all Gq-coupled GPCRs? Or is only mGluR1 function increased due to the IP3 enhancement? This may have important implications for understanding functional consequences.

We do think that IP3 signaling evoked by all Gq-coupled receptors would be affected and we agree that this would have important implications. We appreciate this insightful comment.

3) An additional experiment with a more direct link between the cocaine CPP and changes in IP3 signaling in VTA DA neurons after stress would help strengthen the study.

Although directly manipulating IP3 signaling in the VTA in vivo would not be a simple experiment, again we appreciate this insightful comment for our future study.

[Editors' note: further revisions were requested prior to acceptance, as described below.]

Essential revisions:

Specifically, the authors should discuss, in a more balanced way, the possibility that cocaine exposures enhanced rewarding effects of cocaine not merely the learning of it. Additionally, the referee remained unsatisfactory regarding the discussion on the link between the LTP and CPP behavioral data. The discussion must be more explicit that a causal demonstration will be required to establish the link.

The reviewers agreed that these are interpretational issues and do not require additional experiments. Nonetheless, the manuscript will greatly benefit from discussing these in a more careful way. For more details, please refer to the referee's comments that are copied below:

Reviewer #1:

The current version of the manuscript is improved by the addition of the new CPP results. However, issues still remain with the authors' interpretation of their data that make this manuscript incomplete.

1) I remain unsatisfied that the CPP protocol utilized here measured learning of cocaine-associated cues and not cocaine reward.

The authors now acknowledge that CPP measures reward in the eighth paragraph of the Discussion, however, this is quickly overshadowed by a qualifying statement which is contradictory:

"Enhanced CPP acquisition observed in defeated rats might be due to an increase in the "primary rewarding" action of cocaine. Although this possibility cannot be ruled out, it should be noted that socially defeated rats displayed no significant increases in the overall activity level during the cocaine conditioning sessions (Figure 7—figure supplement 1), suggesting that the "locomotor stimulating" action of cocaine was not altered by repeated defeat experience."

This statement is very misleading. Are the authors suggesting that because stressed mice did not show enhanced cocaine locomotor activity compared to control mice, that they likely did not have enhanced cocaine reward? If in fact this is what the authors are trying to convey, then I would surmise that they are equating locomotor activation with reward, which has repeatedly been shown not to be the case (for example, see PMID: 22197517).

We have removed this statement and replaced it with the following statement:

“However, enhanced CPP acquisition observed in defeated rats may well be caused by an increase in the primary rewarding action of cocaine itself. The relative contribution of these two possibilities, i.e., enhanced learning mechanism vs. enhanced cocaine reward, remains to be determined.”

Moreover, the authors refer to the cocaine chamber as cocaine-associated cues, however, these contextual cues are in fact paired with the experience of cocaine and not cocaine administration itself, which, presumably, takes place elsewhere. This should be made clear. It is as if the authors lack some of the fundamental knowledge of Pavlovian conditioning or choose to ignore it to fit the framework of the paper. For example:

"Repeated social defeat enhanced learning of cocaine-associated cues assessed with a CPP paradigm."

This sentence has been rewritten as follows:

“CPP experiments showed that repeated social defeat promoted acquisition of the preference for contextual cues paired with cocaine experience.”

2) The authors have done little to assuage initial concerns that the involvement of LTP in the CPP behavioral data is greatly overstated.

The data clearly show that social stress enhances LTP and this is glucocorticoid-dependent. Enhancement of CPP seen in stressed rats is also glucocorticoid-dependent. These are important findings and the experiments are performed elegantly. However, there is no assessment that unambiguously shows that enhanced LTP is required for the observed facilitation of cocaine CPP. As such, there are a large number of instances throughout the manuscript wherein the authors overstate their findings (as was mentioned in the original review).

For example, the title of this manuscript is unsupported by its contents:

"Repeated social defeat stress enhances synaptic plasticity that drives learning of drug- associated cue valence"

The title has been changed as follows:

“Repeated social defeat stress enhances glutamatergic synaptic plasticity in the VTA and cocaine place conditioning”.

In the Abstract, this trend continues when the authors state:

"Here, we show that repeated social defeat stress in rats results in persistent enhancement of long-term potentiation (LTP) of NMDA receptor-mediated glutamatergic transmission in the ventral tegmental area (VTA), thereby promoting the learning of cocaine-associated contextual cues assessed with a conditioned place preference (CPP) paradigm."

and…

"…plasticity in the VTA as the critical mechanism by which repeated stress increases addiction vulnerability."

The Abstract has been rewritten in two places as follows:

“Notably, defeated rats display enhanced learning of contextual cues paired with cocaine experience assessed using a conditioned place preference (CPP) paradigm.”

“These findings suggest that enhanced glutamatergic plasticity in the VTA may contribute, at least partially, to increased addiction vulnerability following repeated stressful experiences.”

Likewise, statements such as the following are far too speculative (i.e., not a single piece of evidence is provided to support this claim):

"Here, mGluR1/NMDAR blockade would suppress CPP acquisition via inhibiting LTP induction at glutamatergic inputs activated by contextual cues of the CPP box, while blocking potentiated NMDAR-mediated excitation at those inputs would interfere with CPP expression".

This sentence has been removed.

https://doi.org/10.7554/eLife.15448.015

Article and author information

Author details

  1. Claire E Stelly

    1. Department of Neuroscience, University of Texas, Austin, United States
    2. Waggoner Center for Alcohol and Addiction Research, University of Texas, Austin, United States
    Contribution
    CES, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article
    Competing interests
    The authors declare that no competing interests exist.
  2. Matthew B Pomrenze

    1. Waggoner Center for Alcohol and Addiction Research, University of Texas, Austin, United States
    2. Division of Pharmacology and Toxicology, University of Texas, Austin, United States
    Contribution
    MBP, Acquisition of data, Analysis and interpretation of data
    Competing interests
    The authors declare that no competing interests exist.
  3. Jason B Cook

    1. Department of Neuroscience, University of Texas, Austin, United States
    2. Waggoner Center for Alcohol and Addiction Research, University of Texas, Austin, United States
    Contribution
    JBC, Acquisition of data, Analysis and interpretation of data
    Competing interests
    The authors declare that no competing interests exist.
  4. Hitoshi Morikawa

    1. Department of Neuroscience, University of Texas, Austin, United States
    2. Waggoner Center for Alcohol and Addiction Research, University of Texas, Austin, United States
    Contribution
    HM, Conception and design, Analysis and interpretation of data, Drafting or revising the article
    For correspondence
    morikawa@utexas.edu
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-2948-493X

Funding

National Institutes of Health (AA007471)

  • Jason B Cook

National Institutes of Health (DA015687)

  • Hitoshi Morikawa

National Institutes of Health (AA015521)

  • Hitoshi Morikawa

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

This work was supported by NIH grants DA015687 and AA015521. JBC was supported by NIH training grant AA007471. We thank Dr. Kamran Khodakhah for the generous gift of caged IP3 made in his lab at Albert Einstein College of Medicine. We also thank Dr. Michela Marinelli for comments on this manuscript and Dr. Kevin Lominac and Nhi Le for assistance with CPP experiment.

Ethics

Animal experimentation: All animal procedures were approved by the University of Texas Institutional Animal Care and Use Committee (Protocol ID: AUP-2013-00177).

Reviewing Editor

  1. Naoshige Uchida, Harvard University, United States

Publication history

  1. Received: February 22, 2016
  2. Accepted: June 7, 2016
  3. Version of Record published: July 4, 2016 (version 1)

Copyright

© 2016, Stelly et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 1,564
    Page views
  • 362
    Downloads
  • 16
    Citations

Article citation count generated by polling the highest count across the following sources: Scopus, Crossref, PubMed Central.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)