Background

Hypoxia is a main characteristic of gestational sleep apnea and a culprit for other obstetric complications, such as gestational diabetes and preeclampsia [1,2,3,4]. Gestational sleep apnea affects 8–26% of pregnancies [5,6,7,8], and is associated with impaired fetal development underlying fetal central nervous system (CNS) dysfunction, growth restriction, and organ damage [9,10,11,12,13]. Sex plays a role in risk from gestational complications, such as gestational sleep apnea [10, 11]. In pregnancies that experience complications, there is a sex dependent risk for neonatal fatality and morbidities (e.g., survival rates, growth restriction, metabolic dysfunctions) [14,15,16,17]. In addition to adverse perinatal outcomes, gestational hypoxia can increase the risk for chronic disorders, such as neurodevelopmental disorders in offspring later in life [18, 19].

Experimental rodent models of gestational hypoxia have identified multiple hypoxia-sensitive behavioral indices in offspring, such as gestational hypoxia-induced impairments in social function, cognitive function, anxiety-like and repetitive behaviors [20,21,22,23]. Sex is an important factor in the vulnerability of offspring to gestational hypoxia, in which adult male offspring have greater social, cognitive, and anxiety-like dysfunctions than adult female offspring [22,23,24], but there is a paucity of studies on the pubertal consequences of gestational hypoxia. Puberty is a critical period for social, endocrine, and cognitive maturations [25, 26]. It is important to note that many studies examining gestational hypoxia utilized sustained hypoxic exposure for either greater than 15 gestational days or a single exposure lasting between two to eight hours on one gestational day [21,22,23,24], rather than a chronic intermittent hypoxia (CIH) exposure that models the fragmented hypoxic episodes present in sleep apnea [9]. Notably, a recent study demonstrated CIH exposure for 12 days during mid to late gestation impaired memory and social functions during puberty and adulthood in male rat offspring but had no impact on female offspring [20]. Nonetheless, incorporating a shorter CIH exposure that is limited to late gestation would model the majority of sleep apnea cases that present in the last trimester of human pregnancy [6, 8]. It is unknown whether a short exposure to chronic intermittent hypoxia during late gestation would impair social, anxiety-like, or cognitive function.

Social, anxiety-like, repetitive and cognitive impairments are associated with hippocampal dysfunction [27,28,29,30]. The hippocampus is sensitive to hypoxic insults that can result in behavioral deficits [31,32,33]. Hippocampal-mediated dysfunctions have been linked with age-related changes in neurotransmitters and circulating hormones. Prior studies have found hippocampal dysfunction associated with differences in glutamatergic, dopaminergic, and serotonergic pathways [34,35,36]. Further, hippocampal function is sensitive to steroid hormone levels, with impairments observed following dysregulation of testosterone [31, 37], estradiol [38, 39], and corticosterone [37, 40].

Since gestational hypoxia can have a long-lasting impact on offspring physiology, brain function, and behavior, it is important to understand the influence of sex and age on late gestational hypoxia-associated chronic outcomes. To examine these relationships, we used a CIH protocol during gestational days 15–19 in rats to model late gestational sleep apnea [9]. We chose this model of gestational CIH as the incidence of gestational sleep apnea is higher during late gestation [6, 8], during which time fetal CNS development is occurring [41, 42]. Using this CIH protocol, we previously examined ultrasonic vocalizations (USVs) and open field locomotor activity in offspring at different ages (puberty and young adulthood) to examine the impact of late CIH exposure during gestational days 15–19 on fetal neurodevelopment [9]. We found that gestational CIH had no impact on motor function (distance traveled or rearing) or stress-associated 22 kHz USVs, but CIH did alter positive affect-associated 50 kHz USVs in male and female offspring [9]. Importantly, USVs are dependent on multiple maturation domains (e.g., motor function, social function, and anxiety-like behavior [43,44,45,46,47]), indicating that gestational CIH could impact multiple maturation domains. Therefore, in addition to our previous studies examining USVs and locomotor activity, we examined the effects of gestational CIH on social, cognitive, repetitive, and anxiety-like behaviors displayed by offspring. Within the same cohort of offspring, we present here offspring behaviors that include center arena behaviors, marble burying, social function, and spatial memory. We hypothesized that late gestational CIH would produce sex- and age-specific social, cognitive, repetitive, and anxiety-like impairments in offspring.

Methods

Animals

This study was part of a larger study that examined the impact of late-stage gestational hypoxia on maternal and offspring physiology [9]. All experiments were conducted using timed pregnant Long-Evans rats (aged 8–10 weeks, Charles River, Wilmington, MA). Dams arrived at animal facilities on gestational day (GD) 5–7, in which the sperm plug was observed on GD0. Dams were single housed in a 12 h:12 h light/dark cycle with lights on at 0900 h and were provided food and water ad libitum. Dams were allowed to habituate for 4 days prior to the initiation of the chronic intermittent hypoxia (CIH) protocol that was performed during late gestation (GD 15–19) prior to delivery (GD 22–23). Maternal and fetal biometric analyses are reported in our prior publication [9]. Following delivery, the sex of the offspring was determined via anogenital distance and confirmed prior to weaning. Litters were reduced to 8 pups/litter and, when possible, to equal number of males and females. Weaning occurred on postnatal day (PND) 28. Approximately 4 male and 4 female pups in each litter were randomly assigned to either pubertal (PND 40–45) or young adult (PND 60–65) offspring experimental groups. Based on gestational exposure to hypoxia or normoxia (room air), offspring were included in the following treatment groups. Puberty: Male Normoxic (n = 12), Male CIH (n = 16), Female Normoxic (n = 10), Female CIH (n = 11); Young adult: Male Normoxic, (n = 13) Male CIH (n = 13), Female Normoxic (n = 10), Female CIH (n = 9).

Pubertal and young adult offspring were housed with 1–2 littermates of the same sex to accommodate for uneven distribution of male or female offspring. Male and female offspring were housed in separate rooms in our animal facility. Offspring were housed in rooms on a 12 h:12 h reverse light/dark cycle where lights were off at 0900 h. Reverse lighting allowed behavioral testing to be conducted during the offspring’s active phase of the circadian cycle. Food and water were provided ad libitum for all rats. To acclimatize the offspring to operator handling and reduce stress responses during behavior testing, offspring were handled daily, beginning approximately 10 days prior to the start of behavior testing. At the conclusion of behavior testing, the offspring were anesthetized with 2–3% isoflurane and euthanized via decapitation during the active phase of the circadian cycle between 0900 and 1100 h on PND 48 (puberty) or PND 66 (young adult). All experiments were conducted in agreement with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and the ARRIVE guidelines. These protocols were approved by the Institutional Animal Care and Use Committee of the University of North Texas Health Science Center.

Chronic intermittent hypoxia protocol

Timed pregnant female rats were assigned to receive CIH (n = 8) or room air normoxic (NORM; n = 8) treatments for 8 h starting at 0900 h during their sleep phase of the circadian cycle on GD 15–19. Four days prior to the initiation of the CIH protocol, the home cages (clear plastic containers with sufficient bedding and nesting materials for the dams) of the timed pregnant females were placed into Oxycycler chambers (76.2 × 50.8 × 50.8 cm, BioSpherix, Lacona, NY, USA) to acclimatize the dams to the chambers under NORM conditions. The CIH protocol consists of oxygen reduction from 21% (room air) to 10% oxygen, then returned to 21% oxygen in 6-min cycles with 10 cycles/hour over 8 h/day during the sleep phase of the rat circadian cycle (lights on) for a period of 5 days, as previously described [9]. Following the end of the CIH protocol (GD 19), the cages containing the dams were left in the Oxycycler chambers with the chambers open to room air, and the dams were not disturbed until 24 h following delivery for sexing of the pups.

Behavioral tasks

Behavioral studies were conducted either during puberty (PND 40–45) or young adulthood (PND 60–65) over the course of 1 week during the active phase of their circadian cycles, from 0945 to 1700 h. This cross-sectional design, in which rats were only tested either during puberty or young adulthood, was performed to prevent learning due to prior experience with these tasks (i.e., test battery effects). Notably, test battery effects have been reported in open field tests [48,49,50] and Morris water maze [48, 49], which are behavior tests used in this study. Therefore, to avoid this confound, one behavioral task per index of interest was performed and rats were assessed at either puberty or young adulthood. The order of the behavior tests was randomized. Male offspring were tested at least 1 h before female offspring to avoid potential confounding effects of pheromones on behavior. All testing equipment (e.g., marbles, arenas, tanks) were thoroughly cleaned with 70% ethanol between each rat. All behavior studies were conducted under red lighting and recorded for later analysis by an investigator blinded to treatment groups. Behavior tests were used to assess hippocampal-dependent behaviors, such as repetitive behaviors (marble burying test) [51, 52], social function (social behaviors, social disengagement, social withdrawal, aggression) [27, 53, 54], anxiety-like behaviors (center entries and center duration in an open field, exploration) [27, 29], and spatial learning and memory (Morris water maze) [27, 51]. Methods and results for USVs quantified from this test are reported in Wilson et al. [9].

Marble burying test – repetitive behaviors

This test was used to observe repetitive behaviors [55, 56]. To conduct this test, the floor of the testing arena (50 × 25x30 cm) was thoroughly covered with 1 cm of rodent bedding litter to allow the rats to easily bury marbles. Twenty marbles of similar color and size (1.5 cm) were positioned in a 4 × 5 grid spaced evenly on one side of the arena base, and a pre-test photograph was taken. At the beginning of the test, a single rat was placed on the opposite side of the arena facing away from the marbles (Additional file 1: Fig. S1A). Each rat was given 10 min to explore and interact with the marbles. Operators visually monitored the experiment and manually recorded behaviors. After 10 min, the rat was removed and the arena was photographed for later quantification of the number of buried marbles. A marble was considered buried if 75% or more of the marble was covered with bedding litter. Post hoc quantification was performed by an experimenter blinded to the treatment of the rats by examining the photographs taken before and after the test.

Social function

To quantify social behaviors, social disengagement, and social withdrawal in a group social setting, rats were placed into a novel arena (50 × 25 × 30 cm) with novel conspecific rats of the same sex for 10 min. Due to the uneven numbers in some of the study groups and to maintain the novelty of the arena, social group settings were either 2 or 3 rats of the same sex. We observed no differences in behaviors due to these social group numbers (Additional file 9: Table S1). The arena was lined with litter bedding (1.5 cm). To allow the rats to retreat from the novel arena, a foil covered semi-cylindrical tube (6.5 × 9 × 20.5 cm) was placed into the arena. Behaviors assessed included: (1) social behaviors (sniffing + following + climbingunder + climbingover = social behavior composite score) [22, 57], (2) social disengagement behavior (percentage of nose-to-nose engagements where the rat was the first to disengage from interaction with a conspecific; Additional file 1: Fig. S1B), (3) social withdrawal (rats that did not perform nose-to-nose interactions with conspecifics), (4) total exploratory behaviors (rearing + digging + tubeclimb + tubesniff + tubechew) [58, 59], (5) aggressive behaviors (kicking + biting + dominance posture + boxing + fighting = aggressive behavior composite score) [54, 60,61,62]. Due to technical issues with video recordings during the social behavior testing, footage for 9 young adult females (normoxic n = 4; CIH n = 5) was lost and not available for analysis for social disengagement behavior, social withdrawal, and exploratory behavior.

Open field center—anxiety-like behavior

Anxiety-like behaviors were assessed using a novel open field arena (40.64 × 40.64 × 38.1 cm) with the bi-directional main field bar (San Diego Instruments Photobeam Activity System (PAS)-Open field arena). Rats were allowed 10 min in the novel open field arena to explore. Anxiety-like behaviors can be assessed by the time spent (duration) or entries into the center of the open field [63]. To quantify these behaviors, a center zone was created within the PAS system to denote the number of entries (frequency) to the zone and time spent (duration) within the zone versus time spent around the edge [31]. Locomotor behaviors quantified from this test are reported in Wilson et al. [9].

Morris water maze—spatial memory

To examine spatial memory and learning, the Morris water maze test was used according to our published protocols [31]. On the first day of the Morris water maze, the rats were trained to swim in a pool filled with opaque water (23–25 °C) to a visible platform approximately 1 cm above the surface of the water and remain on the platform for 20 s until removed by the operator. The following three days (day 2–4 of testing) consisted of 3 trials/day with a 10 min inter-trial interval per rat to train the rats to locate a submerged target platform (learning phase). For each trial, a rat was placed into the pool at a randomly assigned point, equidistant from the target. Rats were allowed 90 s to locate the target with a trial ending when either the rat located the target and climbed onto it, or 90 s had passed. Once the target was located, each rat was allowed to sit on the platform and observe visual cues placed on the walls to aid in the formation of spatial memory for 20 s. After the 20 s passed, the rat was removed from the water maze and placed into a carrier to dry and wait for the next trial. Rats that did not locate the hidden target were guided to the target by means of the operator tapping on the target until the rat swam to and climbed onto the platform. Twenty seconds were then provided for the rat to observe spatial cues and rest, and then the rat was returned to the carrier. The target remained in the same location throughout all 3 days of training. On day 5, each rat was administered a probe trial to test for spatial memory retention. During the probe trial, the underwater platform was removed. Each rat was placed into the water at one of the pre-determined random entry points and allowed 30 s to swim to the target location and search for the platform. At the end of 30 s, the platform was returned to its original location and the rat was allowed 20 s to sit on the platform to reduce stress. Latency and pathlength to the target were recorded using ANY-maze software (v. 5.14, Stoelting Co.). Latency and pathlength to the target during the probe trial were used as indicators of spatial memory retention, latency on each day leading up to the probe trial was used to measure learning. A learning index was generated by averaging the latency to the target of the all trials (3/day) of days 2 through 4 to examine performance during the learning phase [64].

Sample collection

To collect tissue and blood samples, rats were anesthetized with isoflurane (2–3%) and decapitated during the first 2 h of the rats’ active phase of the circadian rhythm [9, 31, 65]. Trunk blood was collected in EDTA tubes, and then centrifuged at 2000×g for 10 min at 4 °C to collect plasma. Plasma was stored at − 80 °C. Each brain was quickly removed, flash frozen in 2-methlylbutane, and sliced into 1-mm coronal sections using a brain matrix (RBM-4000C, ASI Instruments, Warren, MI). Using blunt 20-guage needles attached to 1 ml syringes for brain region microdissection [9, 31, 65,66,67], the dentate gyrus (DG) and CA1 region of the dorsal hippocampus (− 5.30 mm from Bregma) were isolated according to Paxinos and Watson’s brain atlas [68]. Micro-dissected brain samples were placed into microcentrifuge tubes to be stored at − 80 °C until protein analysis.

Hormone assays

Steroid hormones were extracted from plasma via acetonitrile preparation [69]. Briefly, plasma samples were diluted 2:3 in HPLC grade acetonitrile, then vortexed at room temperature. After 10 min, the plasma samples were centrifuged at 17,000×g at 4 °C for 5 min. Supernatants were collected and dried by vacuum centrifugation. All samples were reconstituted in the appropriate assay buffer prior to use in ELISA assay according to manufacturer’s instructions. Commercially available, competitive immunoassay ELISA kits were used for quantitative determination of circulating testosterone (ADI-900-065, Enzo Life Sciences Inc., Farmingdale, NY), corticosterone (ADI-900-097, Enzo Life Sciences Inc., Farmingdale, NY), and estradiol (RTC009R, BioVendor, Czech Republic) according to manufacturer's instructions [70, 71]. The intra-assay coefficient of variation for testosterone ELISA was 7.8% with an inter-assay coefficient of variation at 9.3%. The sensitivity of this testosterone ELISA was 5.67 pg/ml at the 2-standard deviation (s.d.) confidence limit. The intra-assay coefficient of variation for corticosterone ELISA was 6.6% with an inter-assay coefficient of variation at 7.8%. The sensitivity of the corticosterone ELISA was 26.99 pg/ml at the 2 s.d. confidence limit. The intra-assay coefficient of variation for estradiol ELISA was 6.1% with an inter-assay coefficient of variation at 7.0%. The sensitivity of estradiol ELISA was 2.5 pg/ml at the 2 s.d. confidence limit. Reported values are mean ± s.d.

Tissue sample preparation, electrophoresis, and western blots

For protein analysis, frozen tissue samples were thawed in RIPA buffer (VWR, cat #N653) containing (per 0.5 ml): 2.5 µl Halt™ protease and phosphatase inhibitor (Thermo Scientific, cat #78,442), 1 µl 0.5 M ethylenediaminetetraacetic acid (EDTA, Sigma-Aldrich), and 1 µl 0.5 mM dithiothreitol (DTT, Sigma-Aldrich) for homogenization, as previously described [9, 67, 72,73,74]. Total protein concentration levels in the homogenate were determined using Pierce BCA Protein Assay (Thermo Fisher, cat #23225). Samples were denatured with β-mercaptoethanol and boiled at 100 °C for 5 min. Equal volumes of denatured tissue samples containing 30 µg protein were loaded into a Bio-Rad 4–15% polyacrylamide gel. The gel then underwent electrophoresis at 25 milliamps (mA) in a tris–glycine running buffer followed by overnight transfer at 4 °C onto a PVDF membrane at 50 mA. Following 30 min washing, membranes were blocked for 30 min with 5% nonfat milk in tris-based saline (TBS)-Tween (TBST) at room temperature. Membranes were then transferred to 1% nonfat milk TBST solutions containing specific primary antibodies and incubated overnight at 4 °C. Primary antibodies used on CA1 hippocampal micro-punched tissue included DAT-HRP (Santa Cruz, SC-32259 1:5000), MAO-A (Abcam, ab126751 1:1000), NR2A (NeuroMab, N327A/38 1:500). For DG tissue samples, primary antibodies for doublecortin (Santa Cruz, sc-271390 1:500) and EGR-1 (Santa Cruz, sc-515830 1:500) were used. Afterwards, membranes were washed in 10-min increments for 30 min, and then incubated in 1% milk TBST- secondary antibody solutions including HRP-conjugated horse anti-mouse (Cell Signaling, 7076P2 1:2500) or HRP-conjugated goat anti-rabbit (Cell Signaling, 7074P2 1:5000) at room temperature for 1 h. For protein normalization, we used beta-actin primary antibody (GeneTex, GTX 629630 1:3000), which was incubated for 1 h at room temperature. Protein bands were visualized using West Pico (Thermoscientific, cat #34580) or West Fempto (Thermoscientific, cat #34095) enhanced chemiluminescence detection using a Syngene G-Box Chemi XRQ system with GeneSys software (version 1.5.2.0) as previously described [74]. NIH Image J software (version 1.48v) was used to quantify band densitometry of protein of interest normalized to beta actin band densitometry. The following proteins of interest were quantified: (1) glutamatergic NMDA receptors (NR2A) that mediate excitatory neuronal synaptic activity and are associated with spatial memory and social behaviors [75, 76]; (2) dopamine transporter (DAT), a monoamine neurotransmitter transporter [77] that can impact memory function, anxiety, and social function [78,79,80]; (3) monoamine oxidase-A (MAO-A), an enzyme that catalyzes monoamines (e.g., serotonin, norepinephrine, and dopamine) and is associated with memory and social function [80,81,82]; (4) early growth response protein 1 (EGR-1), a transcription factor expressed in numerous cell types that acts to mediate cellular activity [83, 84], with hippocampal expression associated with memory, anxiety and neuropsychiatric disorders [85, 86]; and (5) doublecortin (DC), a neurogenesis marker associated with memory, anxiety and social function [87,88,89,90].

Statistical analysis

Significance was defined as p ≤ 0.05. Statistical analyses were conducted in IBM SPSS (SPSS v. 29.0.0, IBM, 2022). Data distribution was tested using the Shapiro–Wilk test. Data with non-Gaussian distribution were normalized by square root transformation through the SPSS transformation function (x = sqrt(x)). Outliers greater than 2 standard deviations from the mean were removed from analysis. Since social withdrawal was exhibited by a low number of animals, Fisher’s exact test was used to determine the impact of gestational CIH on social withdrawal [91]. For all other behavior tests, ELISAs, and protein analyses, the main effects and interactions were examined using two-way ANOVAs with the factors of treatment and sex within either the puberty or young adult groups, wherein we provide the F values, degrees of freedom, p-values, and η2 (measure of effect size). To examine the effect of age, two-way ANOVAs with the factors of treatment and age within either male or female groups were performed. Following ANOVA, a Fisher’s LSD posthoc test was used to determine specific group differences. Results are presented as mean ± S.E.M. unless otherwise indicated.

Results

Late gestational CIH induced social impairments in female offspring without affecting aggressive or exploratory behaviors

We observed that gestational CIH significantly reduced social behaviors displayed by pubertal rats (F1, 43 = 4.934; p = 0.032; η2 = 0.093; Fig. 1A), especially in pubertal females (p ≤ 0.05). However, this effect of gestational CIH on social behaviors was not maintained in young adult rats (Fig. 1B). No effects of sex or age were observed on social behaviors.

Fig. 1
figure 1

Social and exploratory behaviors. Gestational CIH decreased social behaviors in female rats during puberty (A) but not during young adulthood (B). No effect of gestational CIH was observed on male rats during puberty (A) or young adulthood (B). No effect of sex was observed on social behaviors (A, B). No effect of gestational CIH or sex was observed in exploratory behavior (C, D). Analyzed by Two-way ANOVA with Fisher’s LSD multiple comparisons tests. ANOVA significance indicated by: ** = CIH; Post-hoc significance indicated by: # versus normoxic female; p ≤ 0.05

Full size image

We observed no effects of gestational CIH or sex on exploratory behavior in pubertal or young adult rats (Fig. 1C, D). We found increased exploratory behavior by young adult females compared to pubertal females (F1, 27 = 5.001; p = 0.034; η2 = 0.156; Additional file 2: Fig. S2A). This effect was also observed in young adult males compared to pubertal males (F1, 44 = 11.943; p = 0.001; η2 = 0.200; Additional file 2: Fig. S2B).

No effects of gestational CIH or sex on social disengagement were observed in pubertal rats (Fig. 2A). In contrast, there was a significant interaction between gestational CIH and sex on social disengagement in young adult rats (F1, 26 = 14.472; p < 0.001; η2 = 0.307; Fig. 2B), in which gestational CIH increased social disengagement in young adult females. Further, late gestational CIH increased social disengagement in pubertal and young adult females (F1, 20 = 7.540; p = 0.014; η2 = 0.319; Additional file 3: Fig. S3A). No effect of CIH or age was observed in male offspring (Additional file 3: Fig. S3B). Late gestational CIH increased social withdrawal in pubertal females (p ≤ 0.05; Fig. 2C). No males displayed social withdrawal.

Fig. 2
figure 2

Social disengagement. No effect of gestational CIH or sex was observed on social disengagement in pubertal rats (A). Gestational CIH only increased percentage of young adult female rats that exhibited social disengagement (B). Social withdrawal was significantly associated with gestational CIH only in pubertal females (C). Social disengagement analyzed by Two-way ANOVA with Fisher’s LSD multiple comparisons tests (A, B). Relationship between social engagement and CIH measured by Fisher’s exact test (C). Significance indicated by: ** = CIH, *** = interaction; p ≤ 0.05

Full size image

Since developmental dysfunction and altered steroid hormones can impact aggression in pubertal and young adult rats [54, 60,61,62], aggressive behaviors were recorded. Aggressive behaviors include kicking, biting, dominance posture, boxing and fighting [60,61,62]. Offspring displayed little to no aggressive behaviors, regardless of gestational CIH, sex, or age (Additional file 10: Table S2).

Late gestational CIH increased repetitive behaviors only in pubertal females

In pubertal rats, there was a significant effect of gestational CIH on marble burying (F1, 49 = 7.215; p = 0.010; η2 = 0.118; Fig. 3A), where gestational CIH exposed pubertal females buried more marbles than normoxic pubertal females (p ≤ 0.05). There was also a significant effect of sex on marble burying (F1, 49 = 7.205; p = 0.010; η2 = 0.118; Fig. 3A) in pubertal rats, where males buried more marbles than females. Conversely, in young adulthood the significant effect of CIH (F1. 45 = 5.204; p = 0.028; η2 = 0.082; Fig. 3B) is reversed, with fewer marbles buried by offspring exposed to gestational CIH, specifically by females (p ≤ 0.05). Similar to puberty, males during young adulthood (F1, 45 = 14.761; p < 0.001; η2 = 0.232; Fig. 3B) buried more marbles than females. There was a significant interaction between gestational CIH and age in females (F1, 36 = 7.322; p = 0.010; η2 = 0.161; Additional file 4: Fig. S4A), in which gestational CIH increased repetitive behaviors only in pubertal females. Age also affected marble burying in males, as young adult males buried more marbles than pubertal males (F1, 50 = 13.669; p < 0.001; η2 = 0.202; Additional file 4: Fig. S4B).

Fig. 3
figure 3

Repetitive behaviors. Gestational CIH increased marble burying in pubertal females (A) and young adult females (B). Males buried more marbles than females, regardless of age (A, B). Normalized by square-root transformation (A, B). Analyzed by Two-way ANOVA with Fisher’s LSD multiple comparisons tests. ANOVA significance indicated by: * = sex, ** = CIH; Post-hoc significance indicated by: # versus normoxic female, p ≤ 0.05

Full size image

Late gestational CIH did not induce anxiety-like behaviors

No effect of gestational CIH was observed on center duration in pubertal offspring (Fig. 4A) or young adult offspring (Fig. 4B). Pubertal males spent more time in the center of the open field arena than females during puberty (F1, 45 = 8.759; p = 0.005; η2 = 0.153; Fig. 4A), in which normoxic males spent more time than normoxic or gestational CIH exposed females (p ≤ 0.05). Young adult males spent more time in the center than females (F1, 41 = 5.736; p = 0.021; η2 = 0.113; Fig. 4B), wherein gestational CIH exposed males spent more time in the center than normoxic or gestational CIH exposed females (p ≤ 0.05). There was a significant effect of age on time spent in the center of the open field, where young adult females (F1, 36 = 29.056; p < 0.001; η2 = 0.436; Additional file 5: Fig. S5A) and males (F1, 50 = 25.264; p < 0.001; η2 = 0.311; Additional file 5: Fig. S5B) spent more time than pubertal rats. Gestational CIH exposed young adult males spent more time in the center than either normoxic males or gestational CIH exposed pubertal males (p ≤ 0.05). We observed no effects of gestational CIH or sex on the number of center entries displayed by offspring (Fig. 4C, D). Similar to center duration, we observed that aging is associated with increased center entries, as shown by more entries performed by young adult females (F1, 36 = 42.528; p < 0.001; η2 = 0.531; Additional file 5: Fig. S5C) and young adult males (F1, 50 = 17.729; p < 0.001; η2 = 0.255; Additional file 5: Fig. S5D) compared to pubertal rats.

Fig. 4
figure 4

Anxiety-like behaviors. Females spent less time in the center of the open field compared to males, irrespective of age or gestational CIH (A, B). No effect of gestational CIH or sex was observed on center entries, regardless of age (C, D). Normalized by square-root transformation (C, D). Analyzed by Two-way ANOVA with Fisher’s LSD multiple comparisons tests. ANOVA significance indicated by: * = sex; Post-hoc significance indicated by: # versus normoxic female, ### versus CIH female; p ≤ 0.05

Full size image

Late gestational CIH only impacted spatial memory function in pubertal male rats

We observed sex differences in pathlength to the target, in which pubertal females exhibited shorter pathlength to the target than pubertal males (F1, 27 = 4.495; p = 0.043; η2 = 0.129; Fig. 5A). However, gestational CIH did not affect pathlength to the target in pubertal (Fig. 5A) or young adult offspring (Fig. 5B). No sex differences in spatial memory were observed in young adult rats. No effect of age on spatial memory was observed in females (Additional file 6: Fig. S6A). In male offspring, the effects of gestational CIH on spatial memory were dependent on age. Specifically, gestational CIH increased pathlength to the target in pubertal males, but decreased pathlength in young adult males (F1, 27 = 9.375; p = 0.005; η2 = 0.242; Additional file 6: Fig. S6B). Although we observed significant effects of gestational CIH, sex, and age on spatial memory-associated pathlength to target, we did not observe any effects on latency to the target during the probe trial (Additional file 11: Table S3).

Fig. 5
figure 5

Spatial memory. Pubertal males had longer pathlength to target during Morris water maze probe trial (A), with the sex difference observed primarily in gestational CIH exposed pubertal males. No effect of sex or gestational CIH was observed on pathlength in young adult rats (B). Analyzed by Two-way ANOVA with Fisher’s LSD multiple comparisons tests. ANOVA significance indicated by: * = sex; Post-hoc significance indicated by: ### = versus CIH female; p ≤ 0.05

Full size image

There were no sex differences or effects of gestational CIH on learning (latency to target) in offspring (Fig. 6A, B). However, we did observe faster learning (shorter latency to target) in young adult males on day 2 of training compared to pubertal males (F1, 27 = 12.109; p = 0.002; η2 = 0.290; Fig. 6B) or young adult females (F1, 26 = 6.824; p = 0.015; η2 = 0.187), though this was not observed on any other days of testing. This effect of age on learning (i.e., shorter latency to target) was not observed in female offspring (Fig. 6A). There were no effects of sex or CIH on the learning index in pubertal rats (Fig. 6C). However, we did observe that young adult males had a shorter learning index than young adult females (F1, 26 = 6.154; p = 0.020; η2 = 0.180; Fig. 6D). No effects of age were observed on the learning index, regardless of sex.

Fig. 6
figure 6

Spatial learning. No effect of gestational CIH on latency to target was observed in females, regardless of age (A). Age improved latency to target during day 2 of learning in males with no effect of age observed on other days (B). Learning index was not affected by gestational CIH or sex during puberty (C). Males had shorter learning index than females during young adulthood (D). Analyzed by Two-way ANOVA with Fisher’s LSD multiple comparisons tests. ANOVA significance indicated by: * = sex; ** = age; Post-hoc significance indicated by: # = versus normoxic female, ## = versus normoxic male; ### = versus CIH female; p ≤ 0.05; P = Puberty; YA = Young Adult

Full size image

Late gestational CIH did not impact expressions of proteins associated with social and cognitive function within the dorsal hippocampus

Of the proteins that we examined, gestational CIH did not impact protein expression (Table 1; Additional file 7: Fig. S7; Additional file 8: Fig. S8), indicating that gestational CIH exposure from GD 15–19 does not impact cellular function in the dorsal hippocampus of offspring. Although we did not observe any effects of late gestational CIH on dorsal hippocampal cellular activity, there were sex– and age-dependent effects on DAT, NMDA receptors, and EGR-1 protein expression (Table 1). Sex only affected DAT expression in the CA1, where young adult males had significantly higher expression compared to young adult females (F1, 20 = 8.122, p = 0.010, η2 = 0.277). We also observed age differences: (1) decreased DAT (F1, 20 = 4.484, p = 0.047, η2 = 0.179) and decreased NR2A (F1, 20 = 6.716, p = 0.017, η2 = 0.251) in the CA1 of young adult females compared to pubertal females, (2) decreased EGR-1 (F1, 20 = 17.013, p < 0.001, η2 = 0.450) in the dentate gyrus of young adult females compared to pubertal females, and (3) decreased EGR-1 (F1, 20 = 15.800, p < 0.001, η2 = 0.424) in the dentate gyrus of young adult males compared to pubertal males.

Table 1 Expression of proteins associated with social and cognitive function within the dorsal hippocampus
Full size table

Late gestational CIH differentially affected circulating corticosterone levels in females

We observed sex differences in circulating testosterone and corticosterone levels, as well as an effect of gestational CIH on corticosterone levels (Table 2). No effect of gestational CIH, sex, or age was observed on plasma estradiol levels (Table 2). No effect of gestational CIH was observed on plasma testosterone levels. However, we observed expected sex differences in testosterone levels: (1) higher testosterone levels in pubertal males compared to pubertal females (F1, 32 = 30.166; p < 0.001; η2 = 0.480) and (2) higher testosterone levels in young adult males compared to young adult females (F1, 35 = 41.037; p < 0.001; η2 = 0.526). Testosterone levels also increased with age in females (F1, 31 = 10.739; p = 0.003; η2 = 0.235) and males (F1, 36 = 31.029; p < 0.001; η2 = 0.462). The testosterone [92, 93] and estradiol levels [94,95,96,97] observed in this study are consistent with other studies.

Table 2 Plasma concentration of steroid hormones
Full size table

To examine the impact of gestational CIH on circulating hormones associated with the stress response, we quantified plasma corticosterone levels (Table 2). Gestational CIH increased circulating corticosterone in pubertal rats (F1, 31 = 5.407; p = 0.027; η2 = 0.094), particularly in female offspring (p ≤ 0.05). A sex difference was also observed in pubertal rats, with higher plasma concentrations of corticosterone observed in female offspring than male offspring (F1, 31 = 19.247; p < 0.001; η2 = 0.338). In young adulthood, the effect of gestational CIH on circulating corticosterone levels was reversed, with lower corticosterone observed (F1, 35 = 5.313; p = 0.027; η2 = 0.061), specifically in young adult females (p ≤ 0.05). The sex difference in corticosterone levels was maintained into young adulthood, with higher levels of corticosterone observed in young adult female offspring than young adult male offspring (F1, 35 = 43.840; p < 0.001; η2 = 0.501). In female offspring, an interaction between age and gestational CIH was observed, with plasma corticosterone levels increasing in gestational CIH exposed pubertal offspring but decreasing in gestational CIH exposed young adult offspring (F1, 33 = 8.343; p = 0.007; η2 = 0.172). No effect of age on corticosterone levels was observed in male offspring.

Discussion

The major findings of this study are the effects of late gestational hypoxia on offspring are dependent on sex and age. In pubertal female offspring, we observed social function impairments, increased repetitive behaviors, and corticosterone dysregulation. These CIH-induced social impairments and corticosterone dysregulation in pubertal females were sustained into young adulthood. In contrast, late gestational CIH induced a different behavioral phenotype in pubertal males in which cognitive dysfunction (increased pathlength) was observed in the absence of anxiety-like behaviors, aggression, and altered exploratory behaviors. The only long-term effect of gestational CIH exposure in male offspring was impaired 50 kHz USVs in young adult males that were identified in our previous study [9].

This study is the first to examine the effects of late gestational CIH on social function, anxiety, and cognitive functions in rat offspring. Further, this is the only study to use Long-Evans rats, as prior studies used either Sprague Dawley or Wistar rats (Additional file 12: Table S4). We chose to use Long-Evans rats in our studies, as this rat strain exhibits increased behaviors (activity, cognitive function, exploration), greater stress reactivity, and increased sensitivity to hypoxia compared to other rat strains [71, 98, 99]. Additionally, Long-Evans rats are commonly used to study prenatal brain development [100,101,102]. Brain development during pregnancy occurs in three stages (Additional file 12: Table S4): (1) stage 1 (GD 1–10) in which the neural tube is formed and is comparable to the first three weeks of human gestation [103, 104]; (2) stage 2 (GD 10–15) in which the establishment of cortical and subcortical brain regions occurs and is comparable to the first 2 months of human gestation [41, 105, 106]; and (3) stage 3 (GD 15–22) in which cortical and subcortical brain maturation occurs and is comparable to the last 7–8 months of human gestation [41, 104, 107]. We have previously shown that short-term, late gestational CIH (GD 15–19) had sex- and age-specific effects on nigrostriatal pathway maturation in pubertal female and young adult male offspring [9]. Impairment of the nigrostriatal pathway is associated with multiple neuropsychiatric disorders, such as cognitive dysfunction [108], autism spectrum disorder [108, 109], and mood disorders [108, 109]. Therefore, here we examined the impact of late gestational CIH on social function, anxiety, and cognitive function in offspring.

Prior studies investigating the impact of gestational hypoxia in rodent offspring (Additional file 12: Table S4) have found equivocal findings that range from social impairments in male offspring with no effects in female offspring [20, 24] to social impairments in female offspring with no effects in male offspring [22, 23] or social impairments in both sexes [21]. Reports of sex differences in cognition due to gestational hypoxia have been observed, with impairment observed in male offspring [20, 110], female offspring [111], or neither sex [21, 22, 112]. These behavioral differences in offspring exposed to gestational hypoxia may be due to the timing or the type of hypoxic insult, as they ranged in hypoxia exposure duration (1–21 days), hypoxia cycles (intermittent, sustained hypoxia), and intensity of lowered oxygen concentration (13%-5%). Our study utilized CIH (10 hypoxia cycles/hour/for 8 h each day at 10% O2) during late gestation (GD 15–19). Although no other study has used gestational intermittent hypoxia exposure targeting brain development stage 3, prior studies using a sustained hypoxia protocol during this time have been conducted (Additional file 12: Table S4). Specifically, sustained hypoxia during this time period (GD 15–22) impaired social behaviors displayed by young adult female offspring and not young adult male offspring [22, 23], had no impact on anxiety-like behaviors in young adult male and female offspring [22, 113], and decreased repetitive behaviors in young adult female offspring [23]. However, there were differences between these prior studies and our current study, in which sustained hypoxia impaired cognitive function in young adult female offspring [111], decreased repetitive behaviors in young adult male offspring [22, 23], and increased anxiety-like behaviors in young adult male offspring [22, 23], indicating that the type of hypoxic exposure during this critical period of brain development is important.

This study is the first to examine the impact of sex as a biological variable in behavior displayed by pubertal offspring exposed to short-term hypoxic stress during late gestation. We observed several sex- and age-associated developmental effects on brain and behavior induced by short-term CIH during late gestation. A recent study by Vanderplow et al. [20] showed that long-term CIH exposure from GD 10–21 induced functional effects in male offspring (impaired memory and social function in pubertal and adult males) but not female offspring. In contrast, our results using a short-term CIH protocol during late gestation (GD 15–19) found sex-specific effects in offspring that were dependent on age. Pubertal male offspring exposed to late gestational hypoxia exhibited cognitive impairment and suppressed 50 kHz USV calls [9], and this suppressed USV call production was sustained into young adulthood [9]. We observed a different phenotype in females. Pubertal female offspring exposed to late gestational hypoxia exhibited increased social dysfunction, increased repetitive behaviors, corticosterone dysregulation, and impaired 50 kHz USVs [9], with only social dysfunction and corticosterone dysregulation present during young adulthood. Although social dysfunction, repetitive behaviors, and corticosterone dysregulation have been associated with anxiety-like behavior in rats [114,115,116,117,118], we did not observe the presence of other anxiety-associated behaviors (e.g., open field center entries and duration, exploratory behavior, or stress-associated 22 kHz USVs [9, 45, 119]).

In our control animals that were not exposed to gestational CIH, we observed sex and age effects on repetitive behavior, anxiety-like behaviors, and corticosterone. Female rats displayed less repetitive behaviors (fewer marbles buried), increased anxiety-like behaviors (shorter open field center duration), and increased circulating corticosterone compared to male rats. Although the literature is equivocal on sex differences in anxiety-associated behaviors [22, 23, 113], multiple studies have shown that females exhibit higher corticosterone than males [120,121,122,123]. We found that aging decreased anxiety-like behaviors and increased exploratory behaviors in both males and females, consistent with other studies [124]. We also found sex-differences, in which we observed increased anxiety-like behaviors in females consistent with increased anxiety-like behavior previously reported in open field, light–dark box, and social interactions [58, 124]. However, studies have also shown that females exhibit decreased anxiety-like behavior in elevated plus mazes [58], indicating that testing conditions may impact this behavior. Indeed, it has been shown that several factors can influence anxiety-like behaviors, such as the testing apparatus (elevated plus maze, open field, social interactions, light–dark box) [58, 124]. Strain differences may mediate the equivocal findings on sex differences in center duration in an open field [58, 124, 125]. Prior studies that examined behaviors within the center of an open field range from no differences in center duration in Sprague Dawley rats [58] to increased center duration by Wistar female rats [125] and Long-Evans male rats [124]. Our study using Long-Evans rats is consistent with previous studies that showed increased center duration in Long-Evans rats that increased with age [124]. However, regardless of age, offspring in this study spent little time in the center of the open field, which could be an effect of arena size [125, 126].

Pubertal female offspring had shorter pathlengths in the Morris water maze than pubertal male offspring during the probe spatial memory trial, an effect that was not sustained into young adulthood. Prior studies have shown that females can exhibit shorter pathlength and latency, depending on experimental conditions (e.g., training duration and room lighting) [127]. Further, pubertal females have been observed to outperform pubertal males in hippocampal function tests, which may have contributed our observed sex differences [128, 129]. We observed no sex differences in the learning index by pubertal rats. In contrast, young adult males exhibited a shorter learning index compared to females, indicating better spatial learning performance. This sex difference in learning index is consistent with reports of improved learning performance in Morris water maze by males compared to females [130, 131].

In addition to behavioral effects of late gestational hypoxia, we examined glutamatergic (NR2A), dopaminergic (DAT), and serotonergic (MAO-A) associated proteins in the CA1 region of the dorsal hippocampus, as these have been associated with social and cognitive impairments [109, 132,133,134,135]. We also examined markers for cellular activity (EGR-1) and neurogenesis (doublecortin) within the dentate gyrus region of the dorsal hippocampus, as cellular activity and neurogenesis within the dentate gyrus is important in mediating spatial memory [136, 137] and social function [87,88,89,90]. Late gestational CIH did not impact protein expression within the CA1 (i.e., NR2A, DAT, MAO-A), the dentate gyrus (i.e., EGR-1, doublecortin), or the substantia nigra (i.e., calpain enzymatic activity, caspase-3 enzymatic activity, tyrosine hydroxylase), regardless of sex or age of the offspring [9]. While we observed expected fragments of NR2A that range from ~ 90 kDa to ~ 70 kDa (Neuromab), we showed no calpain cleavage of NR2A (~ 115 kDa [138, 139]), similar to no effect of gestational CIH on calpain cleavage of spectrin [9]. Future studies need to be conducted to examine the electrophysiological properties (long-term potentiation and long-term depression) of dorsal hippocampal neurons and neurogenesis (bromodeoxyuridine, BrdU), as dysfunction in these cellular properties have been linked with behavioral impairments [140, 141].

The dopamine system has been associated with anxiety-like behaviors [142, 143] and repetitive behaviors [144]. We observed decreased hippocampal DAT in young adult females compared to young adult males. Along with this sex difference, we observed that young adult female rats had decreased hippocampal DAT, NR2A, and EGR-1 compared to pubertal female rats. There are a paucity of studies investigating hippocampal DAT expression, as prior studies using Wistar, F344, and Sprague Dawley rats showed low DAT expression in the hippocampus [145,146,147]. Similarly, the data on sex differences in marble burying behavior is unclear. Previous marble burying studies were conducted using C57BL/6 mice, Wistar rats, and Sprague Dawley rats [148,149,150,151], which found either no sex difference [148, 149], female bias [150], or male bias in marble burying behavior [151]. Few studies on marble burying behavior incorporate both males and females [152].

Limitations

This study has several strengths that include a broad behavior battery that allows us to examine social, cognitive, and motor functions, along with USVs. However, there are some limitations, such as the lack of multiple behavior tests to examine the same functional outcome. The behavioral battery used in this study included (1) marble burying, (2) social behavior, (3) open field- anxiety-like assessments, (4) Morris water maze, (5) USVs [9], and (6) open field-locomotor assessments [9]. These behavior tests were chosen to examine late gestation maturation domains and behavioral functions linked with USVs (e.g., motor, social, and anxiety-like functions [43,44,45,46,47]), while avoiding a testing battery confound [48,49,50]. Therefore, we are unable to broadly generalize these findings. Another limitation of this study is that we did not conduct vaginal smears for estrous cycle staging (the gold standard) to delineate if females were in high- or low-estrogen physiological states [153, 154]. We chose not to conduct vaginal smears to ensure that we treated both males and females the same and to avoid the risk of increased corticosterone and spatial memory deficits that could occur in response to vaginal lavage [155].

Perspectives and significance

Hypoxia-associated pregnancy complications can increase the risk for chronic impairments in offspring, which are dependent on the sex and age. In response to late gestational CIH, we observed both sex- and age-dependent effects. Both pubertal male and female offspring exposed to gestational hypoxia exhibited decreased USVs [9], but the behavioral phenotype associated with this suppression in USVs was dependent on sex. In pubertal female offspring, gestational hypoxia induced social function impairments, repetitive behaviors, and corticosterone dysregulation but had no effect on cognitive function. In pubertal male offspring, gestational hypoxia induced cognitive dysfunction but had no effect on social behaviors. Some of these CIH induced behaviors were maintained into adulthood such as 1) sustained social impairments and corticosterone dysregulation in young adult females and 2) USV call production impairments in young adult males [9]. Importantly, the results from our study indicate that short-term hypoxia-associated pregnancy complications during late gestation can have long-term effects on offspring that impact social, endocrine, and cognitive maturation.