STREStrogen: The Dichotomous Role of Estrous Cycle on Anxiety‐like Behavior is Dependent upon Stress History

Abstract

Stress‐induced psychiatric disorders are some of the most debilitating conditions worldwide and occur more often in women than men. Increased risk coincides with the reproductive years when circulating estrogen (E) levels peak, alluding to a critical role of ovarian hormones in susceptibility. Human stress studies support this association as E is linked to negative mood and anxiety during social stressors. However, preclinical research often reports opposing findings of protective effects of E in non‐stressed subjects, indicating reduced anxiety‐like behavioral responses. Yet, rodent studies often neglect to acknowledge the latent effects of stress on E‐induced behavioral differences, like the negative impact of a traumatic experience that perpetuates future interactions. Thus, this divergence between clinical and preclinical findings limits the capability of identifying adequate pharmacological targets for use in naturally cycling females. Therefore, we investigated if prior exposure to stress in intact adult female rats sensitized behavioral outcomes during anxiogenic tasks and whether this was contingent upon the current phase of estrous. Rats were exposed for 15 mins/day over 5 consecutive days to either control handling (CON) or witness stress (WS), a modified social defeat paradigm whereby a female observes social defeat from within a protected compartment of the aggressive resident’s cage. Forty‐eight hours after the 5th and final CON/WS exposure, rats underwent elevated plus maze (EPM) testing, and then acoustic startle response (ASR) testing the next day. Vaginal lavage was obtained immediately following EPM and ASR to confirm estrous phase. Analyses were conducted in a 2 x 2 fashion: stress exposure x estrous phase, in which metestrus/diestrus (M/D) signified putatively low levels of E, and proestrus/estrus (P/E) signified high levels. Results revealed a prominent stress exposure x estrous phase interaction for percentage of time spent in the closed and open arms of EPM. First, CON rats demonstrated reduced avoidance of the open arms when tested during P/E vs M/D. This is in line with studies reporting an anxiolytic effect of E. However, our findings in WS rats supports the notion that E is a critical component of sensitizing anxiety‐like responses in females with a history of stress. Specifically, WS rats demonstrated greater avoidance of open arms in the EPM during P/E vs M/D. Importantly, we also found that not all tests are impacted by cycle stage. Females with a history of stress demonstrated a sensitized startle amplitude compared with CON rats, independent of estrous phase, when tested for magnitude of the innate startle response. Further investigations into E receptor expression and subsequent signaling mechanisms in stress‐sensitive brain regions are necessary to determine how a history of stress modifies behavioral outcomes in the presence of high circulating E. These data can then elucidate novel therapeutic targets important for decreasing the high incidence of psychological disorders in this susceptible population.