Anxiety and task performance changes in an aging mouse model

Abstract

Due to the increasing focus on aging as an important risk factor for many serious diseases and an emphasis on animal models that have translational value, an increasing number of animal models are being aged. Animal behavior tests can be used to assess effects of aging in mouse models. Female mice begin exhibiting anxiety-like behaviors at 12 months of age which become more serious at 24 months, while males exhibit no age-induced anxiety-like behaviors. Males and females equally demonstrate a failure of daily task performance at 24 months. Despite these cognitive changes, the mice do not show changes in gross motor function. These results suggest cognitive impairment in non-genetically modified aging mice.

1. Introduction

With improvements in living conditions and healthcare across the world, many countries, including the United States, are experiencing a substantial increase in senior population [1]. Aging is a primary risk factor for many of the diseases that result in mortality or morbidity for many seniors, including cancer [2–4], type 2 diabetes [5,6], cardiovascular disease [7,8], stroke [9,10], Parkinson’s disease [11,12], and other neurodegenerative diseases, including Alzheimer’s disease (AD) [13–17]. As medical technologies improve, many of these diseases become less deadly, but no less prevalent [18–20], which leaves many seniors vulnerable to reduced quality of life.

Given the significant impact of these diseases on human health, substantial efforts have been made to create translational models of diseases. These efforts allow for research on disease mechanisms and drug research. The most accurate models of these diseases recapitulate the effects of aging that are seen in humans. This requires the animal models to be aged.

However, models of naturally aging animals are somewhat limited, resulting in flawed comparisons and inappropriate metrics being used when studying genetic models of disease. C57/B16 mice are the most common background for creating genetic strains used for disease research [21]. Thus, it is necessary to establish a baseline to which these models can be compared.

Aging is considered to be an aberrant physiological state in which many biological systems fail to operate at a high level; this may be the underlying reason for aging as a risk factor for dementia and behavioral change such as cognitive decline. Aging is associated with the accumulation of toxic metabolites. Toxin accumulation during aging can be identified in advanced glycation end products (AGEs), which accumulate over a lifetime due to glycolysis and impaired glucose metabolism [22,23]. Individuals with high levels of AGEs are more at risk for developing type 2 diabetes, and diabetic patients with the highest levels of AGEs are more at risk for complications [24–26]. Intervening into the accumulation of AGEs prior to the development of diabetes could be considered an aging intervention rather than a diabetes intervention. AGE reduction, now a novel and promising drug target, is only possible due to careful assessments of molecular changes during aging.

By assessing behavioral changes associated with aging, the present study provides concrete metrics for researchers assessing interventions in the aging process that can be used in addition to traditional survival curves. New metrics of aging will allow researchers more flexibility in conforming to new animal use guidelines, which discourage survival curves as well as flexibility in the experimental time course.

This study evaluates behavioral changes in C57 male and female mice as they age by assessing gross locomotion, free locomotion patterns, anxiety-like behaviors, and daily task performance, demonstrating age-related behavioral changes associated with anxiety and deterioration in the ability to perform daily tasks.

2. Methods

Body weight:

Body weight measurements were made before behavioral experiments were run in the force plate actometer.

Open field:

Open field was conducted using a force plate actometer developed as described [27,28]. Briefly, the apparatus consists of a 30 cm by 30 cm box with acrylic walls and lid. Four force transducers operating at 200 samples per second collect force data that is then processed. Mice were acclimated to the procedure room for 30 min prior to the experiment beginning. Mice were placed in the apparatus and left alone for the 30-min test. Locomotion was based on total movement throughout the apparatus. Anxiety was measured by determining time spent in the center of the arena [29,30]. The center was defined by dividing the arena into two zones, center and outer, of equal area.

Analysis of force actometer data:

Force data is processed in Matlab following Fowler et al. (2001) in that the force data is analyzed for “center of force” following equation 1:

$$X_c=\frac{(x_1{f}_1+x_2{f}_2+x_3{f}_3+x_4{f}_4)}{(f_1+f_2+f_3+f_4)}$$

Where X_c is the location in the x coordinate of the mouse at the given time. The same is performed for the y Cartesian coordinate. This provides location data at a rate of 200 samples per second. This volume of data includes a significant portion of noise from locomotive movements of the mouse that do not affect the mouse’s location. A two-stage process was used to eliminate noise and to provide more accurate estimates of location and distance.

Location data is analyzed for erroneous data, defined as location data that locates the mouse outside of the box. These points are eliminated by replacing the value with the value of the previous data point. Once the data is processed for erroneous values, it is smoothed by a moving average that is 100 samples long, eliminating much of the vibrational noise of the system. This averaging method has no ill effects on mouse location, but reduces the length of the analyzed data stream by 100 samples (0.5 s), as the first average cannot be performed until there are 50 samples and the last average can only be performed until there are 50 samples remaining. A 100-sample moving average was chosen to reduce the effects of individual values while still maintaining a high sampling rate to accurately locate the mouse.

Total distance is calculated between location points by summing the distance between points as follows:

$$D_i=\sqrt{{(x_i-x_{i+1})}^2+{(y_i-y_{i+1})}^2}$$

Where D_i is the distance between point i and i + 1, x_i/y_i is the initial location, and x_i + 1/y_i + 1 is the next location.

Time in center is calculated by summing the total number of points within the inside box (5 cm away from the edge) and comparing it with the number of points in the outside box.

Nesting experiment:

Nesting was performed according to protocol [31]. Briefly, mice were given 3 g of compressed cotton nestlet. In group housing, mice were given a new nestlet 2 h before the dark cycle for 4 subsequent days. On day five, mice were individually housed with a new nestlet and allowed to build a nest. Nests were scored 16 h later on a scale of 1–5 based on published protocols [32]. Untorn nestlets were additionally weighed for further analysis.

Mice:

All studies were performed in accordance to the National Institutes of Health guidelines for animal care and were done with the approval and oversight of the Institutional Animal Care and Use Committee of the University of Kansas. Mice ages are from 3, 12, 20–24 months, including both genders.

Statistical analysis:

All statistics were performed using StatView. Age comparisons were performed using a one-way ANOVA with Tukey’s post hoc test. Sex comparisons were performed using Student’s t-test. Motion patterns were analyzed via linear regression using analysis of intercept.

3. Results

3.1. Cross locomotion, patterns of motion, and body weight

First, novel open field was used to assess total distance traveled in a 30-min period. Between ages, there were no significant changes in the total distance traveled (p = 0.83) (Fig. 1A). Males and females were analyzed separately, and no significant differences were observed (p = 0.14, p = 0.11) (Fig. 1B and C). Distance traveled in five-minute increments was then assessed to determine whether patterns of exploration and rest were different with age. When analyzed for sex and age, no significant differences were seen (p = 0.21) (Fig. 1D).

Fig. 1.

Male and female mice maintain similar patterns and levels of general locomotion. A. Total distance traveled, males and females combined. B. Total distance traveled, males alone. C. Total distance traveled, females alone. D. Total distance traveled in five-minute intervals. E. Body mass at time of experiment. No significance was found in this group. n = 10–26 per group ***p < 0.001.

Finally, we used force plate measurements to assess changes in body weight with age. This was done as a metric of health in animals as they age, as there are well-established parameters for patterns of healthy weight gain in mice, as identified by Jackson laboratories. These mice maintained healthy body weights throughout the duration of this experiment (Fig. 1E). This data indicates that as mice age from 3 to 24 months, they do not experience gross health changes.

3.2. Anxiety-like behaviors in aging mice

Anxiety-like behaviors were measured via time spent in the center of a novel open field apparatus. As mice are naturally explorative animals, open field exploration gives a reliable and more physiological assessment of anxiety than other common tests [30]. There was no change in motion patterns for male or female mice (Fig. 2A). When first assessed, we determined there was a significant decrease of time in the center of the arena in both 12- and 24-month-old mice (p = 0.006) (Fig. 2B). Upon further analysis, we determined that male mice aged 12 and 24 months did not show a significant decrease in time in the center (p = 0.83) (Fig. 2C). However, female mice showed very significant decreases in time spent in the center (p = 0.0001) (Fig. 2D).

Fig. 2.

Female mice show significant anxiety-like behavior with age.

A. Cumulative time in center in five-minute increments. B. Percentage of time in center, males and females combined. C. Percentage of time in center, males alone. No significance was found in this group. D. Percentage of time in center, females alone. *p < 0.05 vs 3 months, ***p < 0.001 vs 3 months #p < 0.001 vs 12 months n = 10–26 per group n. s. = no significance.

To determine whether the different times spent in the center was because of different motion patterns, we assessed time spent in the center in five-minute increments. Linear regression of this data revealed that while the intercepts were significantly different (p = 0.005), indicating differences in time, the slopes were not significantly different (p = 0.27). This indicates that the overall motion patterns are the same, despite having different total times.

3.3. Daily task performance in aging mice

The nesting paradigm assesses daily task performance based on the quality of a nest built by an individual mouse. Young, 3-month-old mice build nearly perfect nests. By 12 months of age, the nests are less proficiently built (p = 0.0001). From 12 months to 24 months, mice trend towards significantly worse nest building (p = 0.12) (Fig. 3A). Due to significant sex differences in the open field paradigm, sex differences in nest building were analyzed. There were no significant differences between male and female mice at any age point (p = 0.50, p = 0.31, p = 0.20) (Fig. 3B–D). Untorn nesting material was recorded by weight (Fig. 3D). No 3-month-old mouse failed to completely tear up their nestlet, compared with 6.25% of 12-month-old mice and 70.83% of 24-month-old mice. The increase in nestlet left untorn was statistically significant (p < 0.001) at 24 months when compared to mice at either 3 or 12 months of age. Representative nests are shown in Fig. 3E.

Fig. 3.

Nesting activity reduces with age, but is not sex-dependent.

A. Nesting score, males and females combined. B. Nesting score, males alone. C. Nesting score, females alone. D. Weight of untorn nestlet, males and females combined. E. Representative images of nests. *p < 0.05 vs 3-month group **p < 0.001 vs 3-month groups #p < 0.001 vs 12 months n = 10–26 per group. (replace representative figure for 24 months).

4. Discussion

We analyzed sex differences on several behavioral tests as mice age. Given new research requirements to include female animals, it is likely that these differences will become more important. Historically, researchers have underestimated the differences between males and females and chosen to only study males.

In the open field paradigm, male and female mice maintained similar general locomotion and motion patterns at all age points. Many behavioral tests, such as elevated plus maze, object exploration, and social recognition tests, rely on the premise that mice are not experiencing difficulty moving or different explorative patterns. However, these behavioral tests don’t allow for an easy assessment of general locomotion patterns, as demonstrated here. Using a reliable method of identifying gross motor defects will improve the accuracy of other behavioral tests and ensures that behavioral tests as mice age will likely reveal useful, disease-relevant data rather than artifacts of the aging process.

Female mice demonstrate higher levels of anxiety-like behavior with aging. This is of particular importance to research in many aging diseases, such as Alzheimer’s disease, which are more common in women. Knowing that male mice do not naturally display increases in anxiety-like behavior is also important, given that aging diseases such as Parkinson’s disease and Type2 Diabetes are more common in men.

Daily task performance is a useful metric of working and long-term memory. Mice build nests for thermoregulation, meaning that it is not a sex-specific behavior, and is learned as a natural behavior when the mice are pups. A high-quality nest will consist of all of the available material shredded, organized into a region less than ¼ the area of the cage, and built up high enough to cover the mouse entirely. Three-month-old mice, whether male or female, rarely fail to build nests fitting this description. However, by 12 months of age, many mice (8/13) fail to utilize the entire nestlet, and none build nests of sufficient height. This is indicative of a failure of memory or performance of a nesting task learned as pups. This trend is continued in 24-month-old mice, where only 6/23 mice built proper nests. Based on data from the open field paradigm showing that 24-month-old mice are capable of similar gross motor, we interpret this to mean that aging mice start to show subtle memory failures.

Mitochondrial failure with age and neurodegeneration have been demonstrated in several models [33–35]. Specifically, with age there is a decline in mitochondrial respiratory chain complex activity and an increase in mutations of mitochondrial genes [36–40]. This leads both to a loss of ATP production as well as an increase in reactive oxygen species (ROS) [35,41–43]. Loss of mitochondrial function is even more pronounced at the synaptic level, where losses of energy can lead to synaptic loss [44–47]. This may be responsible for the behavioral and cognitive changes that we see here and will be investigated in further studies.

Studies on mitochondrial failure have been conducted in an attempt to find drug targets for aging and neurodegeneration. One such target is Cyclophilin D, a mitochondrial matrix protein responsible for initiating calcium and ROS-induced apoptosis. Reduction in Cyclophilin D by genetic prevents aging-induced reactive oxygen species increases, improves mitochondrial respiration, and increases calcium buffer capacity. Furthermore, loss of Cyclophilin D protects neurons from amyloid beta peptide (Aβ)-mediated death and mitochondrial perturbation and restores learning and memory in an Alzheimer’s disease mouse model [48–53].

Mitochondrial matrix protease, presequence protease (PreP), is important for degrading targeting sequences on the thousand proteins imported into the mitochondria. Although many proteases are known to be capable of degrading amyloid beta in the cytosol, such as Insulin Degrading Enzyme, PreP is the only mitochondrial protease capable of degrading amyloid-beta plaques. Previous research has identified a pathway in which amyloid beta is trafficked to the mitochondria, where it can be detected before cytosolic amyloid beta accumulations have formed [54]. PreP activity has been shown to be reduced with aging and AD- or amyloid beta-affected brain, which negatively correlated with increased levels of ROS and oxidative stress [55,56]. Thus, mitochondrial perturbation could contribute importantly to the age-related dementia and behavioral changes such as anxiety and defects in learning and memory.

Assessing aging in animals has benefits beyond the more accurate creation of disease models. The molecular cause of aging has been the subject of many hypotheses, notably the Free-Radical Theory of Aging, proposed by Denham Harman in the 1950s [57]. Harman proposed that at its molecular root, aging is caused by the accumulation of radicals that cause cumulative oxidative stress and damage [58]. While this may be the most popular theory of aging, some studies point to the system being more complex than a simple imbalance of radicals and antioxidants. In some models, overexpression of radical scavengers such as catalase or superoxide dismutase does not extend lifespans [59]. Additionally, the decrease of certain radicals has been demonstrated to extend lifespans [60].

It is likely that measuring lifespan alone will be insufficient to determine the effect of radicals and antioxidants on aging. Using the behavioral baseline established here, it will be important to define the effects of radicals and radical scavengers on these parameters. This will allow us to assess changes more minutely than survival curves.

In summary, data presented in here will be useful in designing successful experiments for aging models of diseases as well as the molecular underpinnings of aging itself. It is suggested that female mice will demonstrate more anxiety-like behaviors when assessing the effects of disease and drugs on the anxiety levels of mice.