Acute stress induces long-lasting alterations in the dopaminergic system of female mice

Stress is a risk factor for many neuropsychiatric disorders, and the mesolimbic dopamine (DA) pathway is a crucial node of vulnerability. Despite the high prevalence of stress-related neuropsychiatric disorders in women, preclinical knowledge on the impact of stress on neural circuitry has predominantly been acquired in males. Here, we examine how a non-social stressor impacts the effect of DA neurotransmission on social and reward-related behaviors in female mice. Acute stress exposure attenuated the anti-social effects of photoinhibiting ventral tegmental area (VTA) DA neurons and transformed photoactivation of these cells into an anti-social signal. Fast-scan cyclic voltammetry (FSCV) revealed an enhancement in optogenetically-induced DA release after stress. 60 days after stress, mice showed distinct patterns of intra-cranial self-stimulation of VTA DA neurons. Our results reveal the impact stress exerts on females and show that neural and behavioral changes induced by acute stress exposure are still present months later.


139
To assay social behavior, mice were tested on a 2-day social interaction paradigm. Here, 140 an unfamiliar young female was introduced into the cage of the experimental mouse and VTA DA 141 neuron activity was inhibited in the experimental mouse during one testing session 142 (counterbalanced for order) ( Figure 1D). Consistent with previous reports (Gunaydin et

176
Importantly, dynamic changes during adolescence that influence fear extinction have been 177 reported (Pattwell et al., 2012). We next investigated whether the differences in the remote stress 178 group were related to the duration of time between stress exposure and testing or the 179 developmental stage during initial stress exposure. Thus, we included another group of mice 180 wherein the initial stress exposure was delivered in adulthood rather than adolescence, and kept 181 the duration of 60 days constant. We found that there was no difference between groups wherein 182 the stress exposure period was delivered during adolescence (P28-32) and adulthood (P86-90;

263
Together with our data this suggests that amplified dopaminergic activity promotes social 264 avoidance. This theory can be consolidated with our results employing halorhodopsin-induced 9 inhibition of VTA DA neurons during our social interaction task. Here, stress exposure prevented 266 the social aversion optically triggered under non-stress conditions.

267
Our results go beyond previous literature in several ways, and highlight the exquisite 268 sensitivity of the female dopaminergic system to stress. Further, our novel non-social stress 269 paradigm did not significantly alter baseline responses to social interaction. Many studies report 270 social avoidance after chronic social defeat stress, a model that has great relevance to humans

295
Indeed, stress induces similar long-term adaptations within the VTA-NAc pathway as seen 296 after chronic drug abuse (Nestler 2006;Saal 2003;Ortiz 1996). Likewise, our new 5-day swim 297 stress appears to induce long-lasting adaptations in the VTA-NAc pathway that sensitizes 298 individuals to subsequent manipulations of this system and contributes to behavioral 299 10 abnormalities. It is also interesting to note that the only difference observed between our two 300 stress groups (recent vs. remote) was intra-cranial self-stimulation response rates for VTA DA 301 photostimulation. Considering stress-evoked elevations in drug self-administration dissipate 302 within 24 hours and then re-emerges after a time interval of days to weeks (Haney et al., 1995; 303 Logrip et al., 2012;Lowery et al., 2008), it is possible that the differential reward sensitivity we 304 observed between stress groups may result from a similar stress-mediated time course.

349
In summary, we find that stress experience can produce long-lasting alterations in the

368
Mice were randomly assigned to an exposure group (non-stress, recent stress, or remote stress) 369 and mice housed together were always subjected to the same exposure. Remote stress was 370 performed between P28 and P32 and recent stress between P86 and P90. Behavioral testing 371 occurred around P97 ( Figure 1A). An additional subgroup of females (n=10) were exposed to 372 adult remote stress between P86-P90. Those mice were then tested around P155 together with 373 a small cohort of non-stressed mice (n=8). All mice were naïve before any experimental proce-

447
Following recordings, mice were transcardially perfused with 4% PFA and processed using im-448 munohistochemical techniques (described below). Evoked DA release was quantified by calcu-  to an optical patch cable for at least 3 days before being subjected to any behavioral assay. All 15 behavioral tests were recorded by a video camera located directly above the respective arena.

461
The EthoVision XT video tracking system (Noldus, Wageningen, Netherlands) was used to track 462 mouse location, velocity, and movement of head, body, and tail. All measurements displayed are 463 relative to the center of the mouse body. were temporarily moved to a holding cage and the experimental mouse was allowed to explore 467 its homecage freely for 1 min (habituation). A novel young (3-5 weeks of age) female C57BL/6 468 mouse was then introduced into the cage and the two mice were then allowed to interact freely

574
We thank P. Namburi for sharing custom MATLAB scripts, Erik Douma and the entire Tye

575
Laboratory for helpful support and discussion. We also thank Jordan T. Yorgason (Oregon Health 576 and Science University) for providing custom scripts for fast-scan cyclic voltammetry analyses.

757
Significantly more nose pokes were performed into the active versus the inactive nose-poke port.