Using a Murine Model of Psychosocial Stress in Pregnancy as a Translationally Relevant Paradigm for Psychiatric Disorders in Mothers and Infants

The mechanisms underlying increased susceptibility to neuropsychiatric disorders in mothers and infants following adverse maternal exposures in the peripartum period remain largely unknown. Substantial maternal physiological alterations occur during pregnancy and the transition to the postpartum period, including several neuroendocrine adaptations that are hypothesized to be critical not only for healthy offspring neurodevelopment but also for preserving maternal mental health^1,2. At the level of the maternal hypothalamic pituitary adrenal (HPA) axis, adaptions in both circadian and stress-induced levels of glucocorticoid release are observed, including a more flattened rhythm of diurnal HPA axis activity and dampened HPA axis response to acute stressors^3,4,5. Given that enhanced HPA axis activity is reported in a subset of women with postpartum affective dysregulation, including increased levels of circulating glucocorticoids and inhibited negative feedback^6,7,8, exposure to stressors that result in increased postpartum stress reactivity and prevent maternal HPA axis adaptions are thought to increase susceptibility to neuropsychiatric disorders.

To elucidate the effects of stress on affective dysregulation in mothers and infants, several rodent models of stress in the peripartum period have been generated. A majority of these models are characterized by the application of physical stressors that result in homeostatic challenges and alterations in dam physiological status⁹, such as chronic restraint stress¹⁰ and swim stress during gestation¹¹, or postpartum shock exposure¹². Although these paradigms have been shown to result in the emergence of postpartum depressive-like behaviors and alterations in maternal care^10,11,12, they have been limited by their inability to accurately capture the psychosocial nature of stressors commonly experienced by human mothers. This becomes particularly important when attempting to reveal the neuroendocrine consequences of chronic stress in the peripartum period, given that processing of different types of stressors is thought to be mediated by varying neural networks orchestrating HPA axis activation⁹.

In order to overcome this limitation, several groups have designed stress paradigms employing psychosocial insults or a combination of physical and psychosocial stressors. The maternal separation model, where dams are separated from her pups for several hours per day during the postpartum period^13,14, and the chronic social stress model, where the dams are exposed to a male intruder in the presence of their litters^15,16, have been able to reproduce the emergence of abnormalities in maternal care and depressive-like phenotypes associated with physical stress paradigms. The chronic ultramild stress paradigm, where pregnant female mice are exposed to a variety of psychosocial insults, including cage tilt and overnight illumination, as well as substantial physiological insults, such as restraint stress and food restriction, has further revealed exposure to a mixed nature of stressors results in abnormalities in maternal behavior, including impairments in maternal aggression, as well as dysregulation in the circadian activity of the HPA axis^17,18. Consistent with these results, an alternating restraint stress and overcrowding model during gestation results in elevations in postpartum maternal circadian corticosterone levels as well as alterations in maternal care, although no differences are observed in HPA axis re-activity following postpartum exposure to novel acute insults¹.

An expansion of this work, generating a gestational stress paradigm that employs multiple psychosocial insults presented in an unpredictable fashion and minimizes the use of physiological stressors. Studies have previously shown this chronic psychosocial stress paradigm (CGS) results in the development of maternal HPA axis dysfunction, including enhanced stress reactivity in the early postpartum period¹⁹. These changes are associated with abnormalities in maternal behavior, including alterations in the quality of maternal care received by pups, and the emergence of anhedonic and anxiety-like behaviors¹⁹, features consistent with perinatal mood and anxiety disorders^20,21. Furthermore, offspring weight gain reduces during the postnatal period following in-utero exposure to CGS¹⁹, suggesting CGS may have persistent negative programming effects in future generations.

The goal in developing the CGS paradigm was to primarily utilize clinically relevant stressors, which accurately capture the type, intensity, and frequency of insults often associated with neuroendocrine dysregulation and the development of perinatal mood and anxiety disorders. Here, the study provides a detailed protocol of how to subject pregnant female mice to CGS, as well as downstream assessments that can be used to test the validity of the model.

All animal experiments described were approved by the Animal Care and Use Committee at Cincinnati Children's Medical Center and were in accordance with the National Institutes of Health guidelines. Ad libitum access to standard rodent chow and water was provided at all times to mice, including during the CGS paradigm. Mice were housed on a 14 h/10 h light-dark cycle (lights on 06:00 h) unless otherwise specified (i.e., exposure to lights overnight).

1. Preparing for timed matings

1. At least 2 weeks prior to setting up timed matings, house the adult female mice together in a standard mouse cage (18.4 cm x 29.2 cm x 12.7 cm), four mice per cage. Label each female mouse with a specific ID number via an ear tag.
   NOTE: C57BL6 female mice with no prior pregnancy and between 3 to 6 months of age were used for this protocol.
2. At least 1 week prior to setting up timed matings, individually house the adult male mice to be used for mating.

2. Setting up timed matings

1. Set up the timed matings at 18:00 h. Take two female mice and place them inside a cage that holds an individually housed male mouse. Separate the timed matings the following morning by 08:00 h.

3. Checking the copulatory plug, designated as gestational day 0.5 (G0.5)

1. Immediately after separating timed matings, check for the presence of a copulatory plug in the female mice. The presence of a copulatory plug will mark G0.5. Allow the mouse to hold the wire grid inside the cage and gently lift it by the tail to visualize the vaginal opening.
   NOTE: The presence of a copulatory plug indicates sexual activity has occurred but does not guarantee a pregnancy. When attempting to calculate the number of experimental mice needed, expect 50% of mice to be plugged from timed matings and a pregnancy to plug incidence of 60%-70%.
2. Use simple visual examination to identify the presence of a copulatory plug (an opaque whitish hardened mass within or slightly protruding from the vaginal opening). If the copulatory plug is not easily identified by simple visual examination, gently insert a blunt end probe into the vaginal opening. Identify the plugs located further back in the vagina by the resistance of probe insertion.
3. Separate the female mice with copulatory plugs and group house in standard mouse cages, 3 to 4 mice per cage.

4. Preparing for CGS paradigm

1. Randomly assign cages housing female mice with copulatory plugs into two groups on G5.5: Control and CGS group. Attempt to randomize cages to have an approximately equal number of mice per group. Transfer the mice to clean standard mouse cages and label with a "do not disturb" sign. Designate these cages as "home cages" for mice to place them at the end of each stressor.
2. Designate a separate room in the mouse facility to perform the CGS paradigm. Design a 11-day stressor regimen, which runs from G6.5 to G17.5, to utilize each of the 7 day stressors [exposure to foreign objects (marbles or legos), predator odor exposure (dirty rat bedding), 30° cage tilt, frequent changes of bedding, bedding removal, movement on shaker] twice per day, and to utilize each of the 3 night stressors (overnight lights on, cage mate change, exposure to wet bedding) overnight in a random fashion. For a possible sample schedule and schematic of experiments described below, see Figure 1.
   NOTE: Each day stressor should fall within the light cycle of mice (lights on 06:00 h-20:00 h), and last 2 h, with at least a 2 h break in between stressors. Each night stressor should be set up at the beginning of the dark cycle (lights off 20:00 h) and separated at the start of light cycle (lights on 06:00 h).

5. Performing the CGS paradigm

1. Set up specific stressors on a standard static cage with filtered top and water bottle in the room designated for the CGS paradigm. Prepare the number of static cages needed for the experiment depending on the number of mouse cages designated to undergo CGS during randomization. Before starting each stressor, transfer the mouse cages of the CGS group from the housing room to the CGS room.
   NOTE: Perform the handling/transfer of mice from home cage to experimental cage and back in laminar flow hoods.
2. Apply the following stressors according to the pre-designed regimen (refer to step 4.2).
   1. Exposure to foreign objects (marbles or legos): Place six marbles (14 mm in diameter) or six legos (different shapes, not to exceed 4 cm in height) randomly distributed into a clean static cage with mouse bedding, without including the mouse nestlets. Place the mice together with their home cage counterparts into the static cage with foreign objects for 2 h. Return the mice to their home cage with the same counterparts at the conclusion of the stressor.
      NOTE: Clean the foreign objects after use.
   2. Predator odor exposure (dirty rat bedding): Place 1 cm in depth of fresh dirty rat bedding from female rats into a clean static cage with no mouse bedding, without including the mouse nestlets. Place the mice together with their home cage counterparts into the static cage with dirty rat bedding for 2 h. Return the mice to their home cage with the same counterparts at the conclusion of the stressor.
   3. 30° cage tilt: Place the mice with their home cage counterparts into a clean static cage with mouse bedding, without including the mouse nestlets. Tilt the cage at 30° against the wall for 2 h. Return the mice to their home cage with the same counterparts at the conclusion of the stressor.
   4. Frequent changes of bedding: Place the mice with their home cage counterparts into a clean static cage with mouse bedding, without including the mouse nestlets. Substitute the mouse bedding with clean mouse bedding every 10 min for 2 h. During mouse bedding changes, gently place the mice in a different clean cage to avoid direct contact with the mice. Return the mice to their home cage with the same counterparts at the conclusion of the stressor.
   5. Bedding removal: Place the mice together with their home cage counterparts into an empty clean static cage (with no mouse bedding or nestlets) for 2 h. Return the mice to their home cage with the same counterparts at the conclusion of the stressor.
   6. Movement on shaker: Place the mice with their home cage counterparts into a clean static cage with mouse bedding, without including the mouse nestlets. Place the static cage atop a reciprocal lab shaker set to 140 strokes per min for 2 h. Return the mice to their home cage with the same counterparts at the conclusion of the stressor.
   7. Overnight exposure to lights: Place the mice with their home cage counterparts into a clean static cage with mouse bedding, without including the mouse nestlets. Keep the lights on overnight (20:00 h-06:00 h) to interfere with dark cycle. Return the mice to their home cage with the same counterparts at the conclusion of the stressor.
   8. Cage mate change: Transfer the mouse into a clean static cage with mouse bedding which is being housed by a different group of two female mice (intact females not part of treatment or control group). Keep the mouse in the static cage with unfamiliar cage mates overnight. Return the mouse to its home cage with its specific home cage counterparts at the conclusion of the stressor.
   9. Exposure to wet bedding: Fill the static cage with mouse bedding with clean water kept at 24 °C until bedding is saturated with water. Place the mice together with their home cage counterparts into the static cage with wet bedding overnight. Return the mice to their home cage with the same counterparts at the conclusion of the stressor.
3. During CGS paradigm, keep the control mice undisturbed in their home cages inside the housing room.
4. Replace the used home cages with new home cages on G10.5. On G17.5, at the conclusion of the overnight stressor, single-house all the experimental mice to prepare for parturition and downstream functional assessments.

6. Monitoring the experimental mice during the CGS paradigm

1. Monitor the mice every 1 h during stressor application, except during overnight stressors.
2. Exclude the mice displaying distress signs, including wounds, lethargy, or any physical abnormality from the experiment. Contact the veterinary staff as needed.

7. Measuring the percentage body weight gain during gestation in the experimental mice (optional)

1. On G6.5, weigh the mice individually before the exposure to stressors. On G17.5, at the conclusion of the overnight stressor, weigh the mice individually. Weigh the control mice at the equivalent gestational time points.
2. Measure the percentage body weight gain during gestation by setting the weight of the first day of the CGS paradigm (G6.5) as 100%.

8. Measuring the postpartum relative adrenal gland weights in experimental mice (optional)

1. On postpartum day 2 (PP2), weigh the control and the CGS dams individually. Euthanize the dams by carbon dioxide inhalation followed by cervical dislocation in a fume hood.
2. Place the mice on a dissection plate, sterilize the abdominal area with 70% ethanol, and open the abdominal cavity using scissors to make a vertical cut. Isolate the adrenal glands located adjacent to the anterior pole of the kidneys with forceps, bilaterally. Carefully dissect the fat tissue surrounding the adrenal glands underneath a dissecting microscope.
3. Weigh the bilateral adrenal glands individually. Calculate the relative adrenal gland weights in milligrams per gram (total weight of the right and the left adrenal glands/body weight).

9. Measuring the postpartum hypothalamic pituitary adrenal (HPA) axis activity in the experimental mice (optional)

1. In preparation for HPA axis measurements, euthanize litters to 6 pups per litter on postpartum day 0 (PP0). Use carbon dioxide inhalation, followed by decapitation with surgical scissors as a secondary method of euthanasia.
2. On postpartum day 2 (PP2), individually restrain the control and the CGS dams inside a well-ventilated 50 mL polypropylene conical tube for 20 min. Immediately after restraint stress, remove the mouse from the conical tube and restrain the mouse with the non-dominant hand by holding the loose skin over the shoulders and posterior to the ears to have the skin over the mandible taut.
3. Puncture the submandibular vein with a lancet slightly behind the mandible but anterior to the ear canal. Collect up to 100 µL of maternal blood in a serum separator tube. After sample collection, apply gentle pressure with gauze to the puncture site to stop the bleeding. Return the dams to the home cage once the bleeding stops.
4. Centrifuge the serum separator tube at 21,130 x g for 6 min and carefully remove the serum. Store the serum at -20 °C for later use. Measure the serum corticosterone concentration by an ELISA kit following the manufacturer's protocol.

10. Measuring the postpartum behavioral changes in the experimental mice (optional)

1. To prepare for the behavioral analysis, cull litters to 6 pups per litter on PP0.
2. Perform analysis of the maternal care fragmentation from PP2 to PP5. On each day, during the light cycle, expose the dams to the testing room for a 5 min habituation period before videotaping the maternal behavior for a 30 min period.
   1. Assess the maternal care fragmentation by measuring the average length of an individual licking/grooming bout and the total number of bouts performed by dams¹⁹.
      NOTE: Licking/grooming behavior is defined as a behavior where the dam is making contact with the pup's body with her tongue, or the pup is being handled by the dam with her forepaws. A bout is defined as an uninterrupted period of time where the dam is engaged in licking/grooming of her pups.
3. Perform analysis of anhedonia via sucrose preference test (SPT) from PP0 to PP6. Expose the dams to one 100 mL bottle of clean water and one 100 mL bottle of 4% sucrose solution in their home cage. Measure the amount of water and sucrose consumed (in mL) daily. Interchange the bottle placement in the home cage. Calculate the sucrose preference using the averages from the last 4 days: preference % = [(sucrose consumption / sucrose + water consumption) x 100].
4. Perform analysis of anxiety-like behavior via elevated zero maze (EZM) on PP8. Place the dams individually on the EZM apparatus consisting of two closed quadrants and two open quadrants elevated from the floor. Allow the dams to explore the maze undisturbed for 5 min. Quantify the time spent in the open quadrant and the number of entries into the open quadrants.

11. Measuring the postnatal offspring weight changes (optional)

1. To prepare for the offspring weight analysis, cull litters to 6 pups per litter on the day of birth (postnatal day 0, PN0).
2. Record the weight of pups on PN0 and at different time points during the postnatal period (PN2, 7, 15, 21).

Exposing the pregnant female mice to CGS results in changes in chronic stress-relevant parameters, including a reduction in body weight gain during pregnancy (Figure 2A) and increased adrenal gland weights in the early postpartum period (Figure 2B)¹⁹. Importantly, exposure to CGS results in postpartum abnormalities in maternal neuroendocrine function. CGS dams exhibit a hyperactive HPA axis as evidenced by the increased serum corticosterone levels following the application of a novel acute insult (Figure 3)¹⁹.

Exposing the pregnant female mice to CGS further results in behavioral abnormalities in the early postpartum period that appear to reflect the emergence of a depressive-like phenotype. CGS dams display alterations in maternal care as reflected by an increase in the degree of fragmentation of maternal signals received by the pups. The average duration of licking/grooming bouts is reduced and associated with an increase in the mean number of bouts following CGS, indicating numerous short episodes of nurturing behavior (Figure 4A,B)¹⁹. Sucrose preference is also depressed in CGS dams when compared to control dams, suggesting the presence of anhedonia (Figure 4C)¹⁹. Lastly, the CGS dams also display increased anxiety-related behaviors as measured by a reduction in the time spent in the open quadrants of the EZM when compared to control dams (Figure 4D)¹⁹.

In the offspring, exposure to CGS in-utero results in decreased weight gain during the postnatal period, from postnatal day 7 to 21, although no changes are observed at birth. This reduction in body weight gain is present in offspring of both sexes (Figure 5)¹⁹. Of note, the CGS paradigm did not have any effect on gestational length, litter size, or sex ratio per litter (data not shown)¹⁹.

Figure 1: Schematic of CGS paradigm and functional assessments for validation. This figure has been modified from Zoubovsky, S.P. et al.¹⁹. Please click here to view a larger version of this figure.

Figure 2: Changes in the chronic stress-related parameters in the dams following CGS exposure. (A) Body weight changes from G6.5-G17.5, Control = 17, CGS = 17. (B) Relative maternal adrenal gland weights at PP2, Control = 20, CGS = 15. Data presented as mean + SEM. *p < 0.05, ****p < 0.0001. This figure has been modified from Zoubovsky, S.P. et al.¹⁹. Please click here to view a larger version of this figure.

Figure 3: Maternal HPA axis measurements following CGS exposure. Maternal serum corticosterone levels measured after 20 min of restraint stress on PP2, Control = 8, CGS = 5. Data presented as mean + SEM. *p < 0.05. This figure has been modified from Zoubovsky, S.P. et al.¹⁹. Please click here to view a larger version of this figure.

Figure 4: Behavioral changes in the early postpartum period in dams following CGS exposure. (A) Mean duration and (B) number of licking/grooming bouts recorded from PP2-PP5, Control = 17, CGS = 17. (C) Percent sucrose preference in SPT, Control = 17, CGS = 19. (D) Total amount of time spent in open quadrant of EZM during the 5 min period, Control = 17, CGS = 19. Data presented as mean + SEM *p < 0.05, **p < 0.01. This figure has been modified from Zoubovsky, S.P. et al.¹⁹. Please click here to view a larger version of this figure.

Figure 5: Changes in the offspring body weight during postnatal development following in-utero exposure to CGS. The offspring body weight measured from PN0 to PN21, Control = 17 litters, CGS = 17 litters.Data presented as mean + SEM. *p < 0.05, ****p < 0.0001. This figure has been modified from Zoubovsky, S.P. et al.¹⁹. Please click here to view a larger version of this figure.

Exposing the pregnant mice to CGS perturbs postpartum maternal neuroendocrine function, including HPA axis response to novel stressors, and is associated with various behavioral abnormalities relevant to perinatal mood and anxiety disorders. Given that the model employs utilization of an environmental risk factor, higher phenotypic variation is expected than otherwise observed in genetic models²². Nevertheless, results obtained from application of the CGS paradigm can be consistent across research laboratories if care is taken to minimize variables that may confound results.

Critical steps in the protocol include steps related to general husbandry practices, including housing control mice separate from CGS mice, and timed-mating steps. Co-housing control and CGS mice could in itself be a stressful stimulus for the control group and therefore confound neuroendocrine or behavioral results^23,24. Likewise, initiating experiments with pregnant mice shipped from the supplier is not recommended. To maximize the efficiency of timed matings, it is recommended to house adult female mice together at least 2 weeks prior to setting up timed matings to synchronize their estrus cycles. Likewise, using sexually experienced adult male mice and preventing males from mating at least 1 week prior to setting up timed matings will maximize their fertility and increase the potential for successful pregnancies. The timeline of the CGS paradigm must also be carefully followed. Applying these stressors of varying intensity too early in gestation could affect uterine decidualization and inhibit embryo implantation²⁵. Stress exposure during different gestational time windows has also been found to carry varying sex-specific neurodevelopmental disease risk for offspring, where male offspring are significantly more vulnerable than female offspring to stressors during early gestation^26,27. CGS schedule must be designed in such a way that it ensures unpredictability to prevent the development of adaptation mechanisms and acclimation often associated with repeated exposure to predictable stressors²⁸. Lastly, litters should be culled to six pups on the day of birth to ensure comparable conditions across all dams and prevent litter size variabilities from confounding maternal hormone or behavior analysis. Likewise, different cohorts should be employed for neuroendocrine and behavioral assessments to minimize confounding effects of restraint stress and submandibular bleeds on behavior. Different cohorts should also be used for maternal care assessment and analysis of other behavioral parameters to minimize disruption of maternal interaction with the pups.

There are several limitations to the current protocol. The inability to accurately predict the number of pregnant mice prior to the start of the CGS paradigm can pose a considerable financial and animal use burden. Modifications could be made to the protocol to achieve more predictable success with timed matings, including evaluating vaginal cytology to identify mice in estrus stage, where both mating and ovulation typically occur²⁹. Ultrasonographic examination of mice could also be incorporated into the CGS paradigm as an alternative non-invasive technique to accurately identify pregnancies from very early stages of gestation³⁰. The use of special breeding chow, with increased fat content, has also been employed by other groups to improve mating success³¹. However, caution must be taken when instituting changes in diet, given that this could affect maternal stress reactivity and behavior³². In addition, the current protocol has been shown to be effective in wild-type C57BL/6 mice, but modifications of the protocol may be needed for different strains or genetic backgrounds as well as species, for they may have large variations in stress sensitivity, maternal care, and emotional regulation.

Compared to currently existing peripartum stress models, the CGS paradigm proves to be more translationally relevant given the resulting disease-relevant endophenotypes observed, including enhanced maternal stress reactivity and postpartum abnormalities in maternal care, anhedonia, and anxiety. These alterations seem to recapitulate clinical findings associated with perinatal mood and anxiety disorders. Future applications of this model include utilizing the CGS paradigm to identify sex-specific effects of maternal psychosocial stress on offspring brain development and disease susceptibility. Studying the effects of CGS on placental function should be considered, given that dysfunction in key placental functions have been shown to impact fetal brain development³³. Incorporating cross-fostering experiments with the CGS paradigm would further help understand the individual contributions in-utero CGS exposure and associated maternal hormonal milieu changes versus postpartum abnormalities in nurturing behavior play in shaping offspring emotional development.