|Home | About | Journals | Submit | Contact Us | Français|
Ovarian steroids alter cognitive performance of young individuals. Whether progesterone enhances learning and memory in tasks involving the prefrontal cortex and/or hippocampus in aged mice was investigated. Aged mice received progesterone (10 mg/kg, SC) or vehicle and were tested for cortical and/or hippocampal learning and memory. Progesterone increased spontaneous alterations in the T-maze and time spent exploring novel objects in the object recognition task. Progesterone increased the time mice spent in the quadrant of the water maze where the hidden platform had been during training, increased latencies to crossover to the shock-associated side of the inhibitory avoidance chamber, and increased freezing in the contextual fear conditioning task. Progesterone did not enhance performance in tasks mediated by the amygdala (cued conditioning), striatum (conditioned place preference), or cerebellum (rotarod) in these aged mice. Thus, progesterone improved learning and memory in tasks mediated by the prefrontal cortex and/or hippocampus of aged mice.
Variations in cognitive performance of female rodents over reproductive cycles coincide with changes in levels of hormones, such as 17β-estradiol (E2) and progesterone (P4). Rats in behavioral estrus, compared to diestrus, have higher levels of E2 and P4 and perform better in inhibitory avoidance, trace conditioning, object recognition and placement tasks [9,26,39,41], but not active avoidance or water maze tasks [6,8]. Pregnant rats' performance in the water maze task is enhanced when E2 and P4 levels are escalating (1st and 2nd trimester), not declining (3rd trimester), compared to non-pregnant rats . Thus, changes in endogenous hormones influence cognitive performance.
E2 and P4 administration can improve cognitive performance of rodents. E2 improves cognitive performance of young and aged rodents across a variety of tasks, but these effects are clearly influenced by age and the dosing/regimen utilized . P4 to young, middle-aged, or aged E2-primed rats or mice improves cognitive performance in several tasks [10,12,20,22,27,28].P4, independent of E2, may improve cognitive behavior. Chronic low-dose P4, or acute administration of P4, prior to training in the water maze improves performance of young, ovariectomized (ovx) rats . P4 post-training to ovx rats enhances performance in the object recognition, delayed-non-matching-to-sample, and inhibitory avoidance tasks, when rats were tested 24 h later when P4 levels are at nadir . Thus, physiological levels of P4, independent of E2, may enhance learning and memory of young adult rodents.
In addition to changes in cognitive function, aging is associated with decline in gonadal steroids and may be associated with differences in responses to steroids. Young males typically are not exposed to levels of P4 that young females are, and would not be expected to respond to P4 similarly. We found that duration spent immobile in the forced swim test of young male mice is not reduced by P4 administration to the same extent that it is in age-matched females. However, P4 administration to aged male and aged female mice produce similar reductions in duration spent immobile in this task . Some studies show that E2 to aged mice can improve cognitive performance [1,7,11,21], but whether P4 alone can enhance learning and memory of aged male and female mice is unclear. To test the hypothesis that aged individuals can respond to P4 for its mnemonic effects in the PFC and/or hippocampus, mice that were 18–24 months of age were administered P4 or vehicle and effects on performance in tasks that involve the PFC and/or hippocampus as well as the amygdala, striatum, and cerebellum were investigated.
Methods were pre-approved by the IACUC at University at Albany-SUNY and carried out using adequate measures to minimize pain or discomfort, as outlined in NIH Guide for the Care and Use of Laboratory Animals (#80–23, 1996).
Female and male mice, raised on a C57BL/6 or 129SVEV background (and backcrossed to C57BL/6), were bred in the vivarium at University at Albany-SUNY. Mice were group-housed in a room, with a reversed 12-h light/12-h dark cycle (lights on 08:00 h) and free access to Purina Rodent Chow and tap water in their home cages.
To obviate the possibility of attrition in aged mice following surgery, mice were not ovx or gonadectomized (gdx). We have previously demonstrated that testosterone levels of intact aged male mice (plasma =~0.4±0.1 hippocampus =~0.3±0.1 ng/g) are similar to those observed in young, gdx mice (plasma =~0.3±0.3 ng/ml, hippocampus =~0.4±0.1 ng/g vs. young, intact male mice plasma =~2.7±1.5 ng/ml, hippocampus =~1.0±0.3 ng/g) . Aged female mice have levels in the hippocampus of E2 (~2.0±1.0 ng/g) and the P4 metabolite, 5α-pregnan 3α-ol-20-one (3α,5α-THP; ~2.0±0.5 ng/g) that are more similar to young diestrous (E2 ~10.0±3.0 ng/g; 3α,5α-THP ~0.7±0.5 ng/g) than to young proestrous mice (E2 ~40.5±8.1 ng/g; 3α,5α-THP ~19.4±4.6 ng/g) [12,13]. All mice were randomly assigned to receive SC P4 (Sigma, 10 mg/kg) or propylene glycol vehicle. This P4 regimen produces proestrous-like P4 and 3α,5α-THP levels [13,14].
Experiments were conducted in two cohorts of aged mice (described below). Although it was beyond the scope of the present experiment to include experimental groups of young male and female mice of the same background strain, benchmark data of what is typically observed in young control mice (3–6 months old) in the measures utilized in our laboratory are included in the Results.
Gonadally-intact C57BL/6 mice (20–24 months old) female (n = 10) or male (n = 10) mice received P4 or vehicle (n = 5/grp), 1 h prior to testing in the T-maze.
Gonadally-intact 129SVEV X C57BL/6 mice (18–22 months old) female (n = 8) or male (n = 8) mice received P4 or vehicle (n = 3–4/grp), immediately after training in each of the tasks, except the T-maze. Mice were tested once a week until performance in each task was evaluated.
Prior to the testing, mice were handled/habituated for 1 week . The primary measures for each task are described below [13-16]. Motor and/or sensory responses were recorded during testing and no differences were noted. An observer uninformed of the experimental conditions of mice recorded the data.
This spontaneous alternation task examines spatial working memory, which involves the PFC and hippocampus. Training involves a single forced-choice trial with one goal arm blocked. Testing is for 13 free-choice trials or 15 min. Mice are placed in the start arm, must enter the left or right open goal arm and return to the start position, and then enter the opposite arm for each trial. The percentage of correct/alternating choices is recorded.
Object recognition involves the PFC and hippocampus. During training, mice are placed in an open field with 2 identical objects and the duration of time spent actively exploring them is recorded for 180 s. The testing period occur red 240 min after training. Mice are exposed to one familiar and one novel object and time spent exploring the objects are recorded for 180 s. The percentage of time spent with the novel as a function of total time exploring the novel and familiar objects is recorded.
This spatial task is mediated primarily by the hippocampus. Mice are habituated by swimming freely for 1 min. Then mice are trained in 12 trials, organized into 3 blocks of 4 trials with a 30-min inter-trial interval. To assess retention, mice are tested in a 60-s probe trial, in which the platform is not available for escape. The time spent in the quadrant where the platform had been located during training is recorded.
This spatial task is mediated primarily by the hippocampus. Mice are habituated to the chamber for 60 s and then placed in the light side of the chamber. After 60 s, a door is lifted. Mice receive a footshock (2 s, 0.5 mA) when they crossover to the dark side of the chamber. 24 h later, mice are returned to the light side of the chamber. The door is lifted and latency of mice to cross to the shock-associated side of the chamber is recorded.
This assesses effects on memory mediated by the hippocampus and amygdala. Mice are habituated for 4 min to the chamber. During training, a tone is sounded for 10 s, followed by a footshock (2 s, 0.35 mA). There are 3 training trials of the tone-shock pairing. 24 h later, mice are tested in contextual (hippocampus-dependent, with training chamber, no tone) and cued (amygdala-dependent; with training tone, new chamber) conditions. Freezing behavior is observed for 8 min.
CPP assesses hedonic effects, mediated by the striatum/nucleus accumbens. A typical CPP procedure was utilized , including modifications for mice in the flooring (smooth vs. mesh). Mice were habituated to the entire chamber for 30 min on two days. On the third day, baseline preferences for each chamber were determined. Mice received injections of vehicle before being placed on their preferred side or P4 on their non-preferred side for 8 trials (30 min each, once per day). On Day 12, mice were tested for their preference. Spending more time in the non-preferred side of the chamber indicates a CPP to P4.
Motor coordination of mice was assessed using the Accurotor Roto-Rod Apparatus (AccuScan Instruments) . Mice were habituated to the task (30 s maximum latency for 3 consecutive trials). Mice were administered P4 or vehicle and then tested in two sessions, separated by an hour interval, consisting of the fixed (20 rpm) and accelerated speed trials (0–20 rpm; 2 trials per session with 180 s max latencies). Performance is determined by the latency to fall.
Two-way analyses of variance (ANOVAs) were utilized to evaluate effects of hormone condition (P4 or vehicle) and sex (female or male). The α level for statistical significance was P < 0.05, and a trend was P < 0.10. Where appropriate, ANOVAs were followed by Fisher's post hoc tests.
P4 treatment resulted in a significant increase in spontaneous alternation in both females and males (Fig. 1, top). There was a main effect of treatment (F(1,16) = 21.39, P < 0.01), but no main effect of sex (F(1,16) = 0.09, P = 0.76), and no significant interaction of treatment and sex (F(1,16) = 1.31, P = 0.27). Young ovx or gdx control mice typically demonstrate ~50% spontaneous alternations in this task (which would be considered performing at chance levels).
P4 significantly increased the percentage of the time that mice spent with the novel object over that seen when vehicle was administered post-training (Fig. 1, bottom). There was a main effect of P4 (F(1,12) = 6.61, P = 0.03), but not sex (F(1,12) = 2.36, P = 0.16), and no significant interaction of treatment and sex (F(1,12) = 2.96, P = 0.13). Young ovx or gdx control mice typically spend ~50% of the total time exploring objects, exploring the novel and familiar objects (which would be considered performing at chance levels) .
P4 significantly increased the time that mice spent in the quadrant where the platform had been located during training over that seen when vehicle was administered post-training (Fig. 2, top, left). Male mice tended to perform better than female mice. There was a main effect of P4 treatment (F(1,9) = 58.08, P < 0.01), and there was a trend for an effect of sex (F(1,9) = 3.76, P = 0.08), and no interaction between these variables (F(1,9) = 0.08, P = 0.79). Young control mice typically spend ~20–30 s in the target quadrant .
P4 significantly increased the latencies, over that seen when vehicle was administered, for mice to cross to the dark, shock-associated side of the chamber (Fig. 2, top, right). There was a main effect of P4 (F(1,9) = 15.10, P = 0.004), but not sex (F(1,9) = 2.03, P = 0.19), and no interaction (F(1,9) = 2.72, P = 0.13). Young control mice typically have latencies of 40–50 s to crossover to the dark, shock-associated side of the chamber in this task [11,15].
Post-training administration of P4, but not vehicle, significantly increased the time spent freezing when mice were tested in the same contextual (Fig. 2, bottom, left), but not cued (Fig. 2, bottom, right), environment in which they had been trained. There was a main effect of P4 (F(1,8) = 6.63, P = 0.03), but not sex (F(1,8) = 0.10, P = 0.35), and no interaction between treatment and sex (F(1,8) = 0.03, P = 0.86) for contextual fear conditioning. Young control mice typically spend ~25 (contextual) and ~15 (cued) seconds freezing.
There were no differences due to sex or P4 condition for rotarod performance and mice had similar mean latencies across fixed-speed or accelerated trials. Furthermore, no statistically significant differences were observed for conditioned place preference between groups. Subjects spent similar duration on the chamber associated with P4 administration, irrespective of condition (mean secs±S.E.M.; males: vehicle = 969±0, P4 = 1085±214; females: vehicle = 1002±10, P4 = 1440±68).
Our hypothesis that aged male and female mice would be similarly responsive to P4 to improve memory consolidation was supported. P4 improved cognitive performance of aged mice in several tasks mediated by the PFC and/or hippocampus (T-maze, object recognition, water maze, inhibitory avoidance, contextual fear conditioning), but there was little evidence for effects of P4 in aged mice for performance in tasks mediated by the amygdala (cued fear conditioning), striatum (conditioned place preference), and cerebellum (rotarod). Thus, P4 may have actions in the PFC and/or hippocampus to improve learning and memory of aged mice.
These findings confirm previous findings that P4 may alter cognitive processes and extend these results to demonstrate that aged mice are responsive to P4's effects on memory consolidation [10,12,20,22,27,28]. These findings, that P4 alone enhanced performance of aged mice, imply that P4, independent of E2, has salient mnemonic effects. Here we see that P4 improves performance of aged mice in tasks mediated by the PFC and hippocampus, the T-maze, object recognition, water maze, inhibitory avoidance, and contextual fear conditioning tasks. No differences between groups were observed in tasks mediated by the amygdala (cued fear conditioning), striatum (CPP), and cerebellum (rotarod). This pattern of effects implies that the PFC and/or hippocampus may be targets of P4 for learning and memory in aged mice.
P4's effects on cognitive performance in the present study are not likely to be due to non-specific effects on motor coordination or affective processes. Many of the previous reports on cognitive effects of P4 and/or E2 involved pre-training administration and/or testing when hormone levels were still high. Here we found that cognitive performance of aged mice was improved when P4 was on board prior to training (T-maze), when P4 was administered immediately after training (all tasks but T-maze), when P4 levels would have been high (T-maze, water maze, object recognition), or low (inhibitory avoidance, conditioned fear) during testing. Given that post-training administration of P4 may have produced an anxiolytic effect during consolidation to enhance later performance, it was important to compare whether there were mnemonic effects in tasks with different intervals between training and testing as well as performance in non-mnemonic tasks. In rats, we have previously demonstrated that P4's mnemonic and anti-anxiety effects are temporally independent . In this  and the present study, P4 had mnemonic effects when P4 and 3α,5α-THP levels are elevated (1–4 h post-administration) and after P4 and 3α,5α-THP levels have declined (24 h later). P4 enhanced learning and memory without altering motor and/or sensory performance in these tasks (e.g. swim speed in the water maze, response to footshock in inhibitory avoidance or conditioned fear) or in the rotarod. Furthermore, P4 did not alter performance in a task mediated by the amygdala (a limbic region important for fear/anxiety) or by the striatum (a region important for hedonic effects). Thus, these effects of P4 in the PFC and hippocampus may be due in part to its effects on memory consolidation (rather than changes in affect or motor behavior).
The limitations of the present study need to be considered. It is important to note that only a small number of subjects were utilized in this study. Factors, such as difficulty acquiring aged mice and concerns about their limited physical capacity, heterogeneity in their baseline cognitive performance, and their potential demise, influenced our study. The mice in experiment 1 were slightly older and tested once in a task that was neither considered cognitively, nor physically, difficult. The clear effects in the T-maze provoked us to utilize a repeated-measures design with the slightly younger, aged mice in experiment 2. There were fewer mice in this experiment and attrition, not related to experimental condition, reduced group size in this study. A similar pattern is observed for behavior of aged, vehicle-administered mice in the present study and others [11,16] and those reported for young, ovx mice administered vehicle . In the present studies, aged vehicle-administered mice had modest reductions in performance compared to levels typically observed in young control mice in most tasks in our laboratory, which likely are due to age. Future studies could directly compare these effects in young and aged mice to determine whether differences were due to aging and/or hypogonadism.
Elucidating the potential for steroids to mitigate age-related factors is critical given the increasing proportion of the population that will be aged, age-related decline in cognitive performance, and increases in dementia with aging . Evidence in support of P4 and/or E2's effects on these processes is as follows. Women are twice as likely as are men to develop Alzheimer's Disease (AD) . Hormone therapy (HT) can decrease risk of AD ; however, the potential beneficial effects of HT decline with delay in initiation and/or may be limited to healthier individuals. Improvements in performance with P4, as observed in the animal models utilized in the present experiment, are not observed consistently in women [25,29,32,34]. For example, higher endogenous P4 or P4 administration to young women enhances PFC-related (but not hippocampus/spatial) function [23,32]. Progestins may produce detrimental, disorganizing effects on cognition that may be particularly evident in older and/or impaired individuals, or in more challenging situations [25,34]. Post-menopausal women are typically treated with E2-based HT that are augmented with progestins (not P4 alone) to minimize trophic effects in the endometrium [35,40]. However, about a third of women discontinuing HT do so because of negative mood effects . The two main factors that predict unfavorable response to progestins are the preparation of the HT and prior negative responses to HT. Further investigation of the effects and mechanisms of P4-based HT for performance in cognitive domains is necessary.
Given the differences among women in response to P4 HT, the potential mechanisms of P4 for cognition need to be explored using an animal model. Whether there are interactions with the cholinergic (and/or other neurotransmitter) systems that change with aging and can mitigate steroids' cognitive effects is of interest [5,19]. It is likely that some of the effects are due to actions of 3α,5α-THP. P4 enhances performance in hippocampal tasks and reinstates hippocampal 3α,5α-THP levels of middle-aged wildtype, but not transgenic “AD”, mice . Increasing 3α,5α-THP levels by administering P4 and/or E2 reduces damage in a rat model of AD . In models of degeneration, P4 and 3α,5α-THP, but not medroxyprogesterone acetate (the synthetic progestin used in HT), exert protective effects and increase 3α,5α-THP levels [3,30,36]. Circulating levels of 3α,5α-THP, but not P4, are significantly lower in people with AD or non-AD dementia compared to controls . It may be that deficiencies in 3α,5α-THP may contribute to, or be a consequence of, age-related pathologies. Another intriguing possibility is that P4, and/or 3α,5α-THP altered neurogenesis, and thereby, improved cognitive performance in the present study . These possibilities are currently being investigated.
This research was supported by grants from NIMH (MH06769801) and NSF (IBN03–16083). The assistance of K. Sumida and M. Rhodes is appreciated.