Postnatal treadmill exercise alleviates cognitive deficits in offspring exposed to gestational diabetes mellitus via reactivation of the hippocampal Wnt/β-catenin signaling pathway
Article information
Abstract
Gestational diabetes mellitus (GDM) exposes the developing fetal brain to hyperglycemia and has been linked to adverse neurocognitive outcomes in offspring; however, effective postnatal strategies and underlying mechanisms remain incompletely defined. Here, we tested whether early-life aerobic exercise rescues hippocampus-dependent memory impairment and canonical Wnt/β-catenin signaling deficits in offspring born to dams with streptozotocin-induced GDM. Pregnant C57BL/6 mice received streptozotocin (40 mg/kg, gestational day 7) or saline. Male offspring were weaned and assigned to sedentary control or treadmill exercise (15 m/min, 60 min/day, 5 days/wk, 2 weeks), generating four groups: saline control (Sal-CON, n=12), saline exercise (Sal-Ex, n=12), GDM control (GDM-CON, n=12), and GDM exercise (GDM-Ex, n=12). Spatial learning and reference memory were assessed using the Morris water maze, and spatial working memory was evaluated with a T-maze task. Hippocampal expression of Wnt3, β-catenin, and glycogen synthase kinase-3β (GSK-3β) was quantified by Western blotting. GDM offspring exhibited impaired hippocampal memory performance, evidenced by a reduced hidden-to-visible platform latency ratio in the Morris water maze and loss of novelty preference in the T-maze. These behavioral deficits were accompanied by suppression of canonical Wnt signaling, with decreased hippocampal Wnt3 and β-catenin and increased GSK-3β in GDM-CON compared with controls. Notably, postnatal treadmill exercise significantly improved both reference and working memory and normalized Wnt/β-catenin pathway components in GDM offspring. Collectively, these findings indicate that early-life aerobic exercise mitigates GDM-associated cognitive vulnerability, at least in part, by reactivating hippocampal Wnt/β-catenin signaling.
INTRODUCTION
Gestational diabetes mellitus (GDM) is defined as glucose intolerance that is first recognized during pregnancy (Will and Crellin, 2023). It is one of the most common metabolic complications of gestation and affects approximately 10% of pregnancies worldwide (Eades et al., 2024). As a result, millions of fetuses are exposed each year to intrauterine hyperglycemia and its metabolic consequences. Beyond perinatal complications such as macrosomia and neonatal hypoglycemia, growing evidence indicates that GDM can have lasting effects on offspring brain development. The hippocampus, a region critical for learning and memory, is particularly vulnerable to maternal metabolic imbalance. Studies in animal models have shown that gestational hyperglycemia induces oxidative stress (Damasceno et al., 2002), disrupts synaptic proteins (Hami et al., 2017; Luo et al., 2022), and promotes neuronal apoptosis in the developing hippocampus (Piazza et al., 2019). These findings suggest that maternal-fetal metabolic disequilibrium can subtly interfere with normal brain development and contribute to cognitive or behavioral impairments later in life.
Among the molecular pathways that regulate neural development and plasticity, the canonical Wnt/β-catenin signaling pathway plays a pivotal role. Activation of Wnt signaling stabilizes β-catenin, which then translocates to the nucleus and drives the transcription of genes essential for neuronal proliferation, differentiation, and synapse formation (Budnik and Salinas, 2011). Proper regulation of this pathway is indispensable for hippocampal neurogenesis and the maintenance of long-term potentiation, a cellular mechanism underlying learning and memory (Varela-Nallar and Inestrosa, 2013). Conversely, overactivation of glycogen synthase kinase-3β (GSK-3β) or upregulation of Wnt inhibitors such as Dickkopf-1 (Dkk1) can suppress this pathway and result in synaptic degeneration and cognitive decline (Galli et al., 2014). Therefore, impaired Wnt/β-catenin signaling may represent a key molecular link between maternal diabetes and hippocampal dysfunction in the offspring.
Recent studies support this hypothesis. Offspring born to diabetic or high-sucrose-fed dams show reduced expression of Wnt ligands, including Wnt2 and Wnt3a, and decreased β-catenin levels in the hippocampus. At the same time, the expression of the Wnt antagonist Dkk1 and microglial activation are elevated. These molecular changes are accompanied by impaired spatial learning and memory performance (He et al., 2017). A proinflammatory intrauterine environment, characterized by elevated cytokine levels and oxidative stress, may therefore suppress hippocampal neurogenesis and synaptic signaling. This cascade of events ultimately contributes to long-lasting cognitive deficits in offspring exposed to GDM.
Aerobic exercise is widely recognized as a potent enhancer of hippocampal plasticity and cognitive function. Exercise stimulates neurogenesis, increases neurotrophic factors such as brain-derived neurotrophic factor and insulin-like growth factor-1, reduces inflammation, and improves metabolic homeostasis (Cotman et al., 2007). Importantly, it also activates Wnt/β-catenin signaling. Exercise increases the expression of Wnt3 and LRP6 while suppressing GSK-3β and Dkk1 activity (Bayod et al., 2014). In diabetic and neurodegenerative models, treadmill training restores β-catenin stability, enhances hippocampal neurogenesis, and improves performance in learning and memory tasks (Bayod et al., 2014). These findings suggest that activation of the Wnt/β-catenin pathway may be one of the primary molecular mechanisms underlying the neuroprotective and cognitive benefits of exercise. Despite these insights, there is currently no direct evidence that postnatal exercise can restore Wnt/β-catenin signaling and cognitive function in offspring exposed to GDM.
We investigated the effects of moderate-intensity treadmill exercise on hippocampal Wnt/β-catenin signaling and spatial memory in male offspring of mice with streptozotocin (STZ)-induced GDM. This study aims to provide foundational evidence that early-life physical exercise can modulate molecular pathways related to neuroplasticity and counteract neurodevelopmental impairments caused by maternal diabetes.
MATERIALS AND METHODS
Experimental animals
A total of 36 C57BL/6 mice (12 males and 24 females), aged 6 weeks, were utilized for this study. The mice were housed in a controlled animal facility under optimal conditions (temperature maintained at 22°C±1°C, humidity at 55%±3%, and 12-hr light/dark cycle). They received ad libitum access to food and water throughout a 2-week environmental acclimation period preceding the experiment. All experimental procedures, including behavioral testing, were completed. All animals were sacrificed exactly 48 hr after the final intervention. Anesthesia was induced via isoflurane inhalation. Brains were immediately extracted following anesthesia. To ensure the integrity of the tissue for subsequent molecular analysis, the hippocampus was rapidly dissected using a spatula on a chilled stainless-steel block. The dissected hippocampal tissue was then flash-frozen in liquid nitrogen and stored at −70°C until analysis. All experimental procedures conducted in accordance with the guidelines of the Institutional Animal Care and Use Committee of Gyeongsang National University (GNU-240426-M0094).
GDM induction and grouping
Following the acclimation phase, the mice were subjected to mating, with a ratio of one male and two females. Once pregnancy was confirmed, the pregnant female mice were randomly segregated into two primary groups: the saline (Sal, n=12) group and the GDM (GDM, n=12) group, with each mouse housed individually.
GDM was successfully induced on gestational day 7 by administering a single intraperitoneal injection of STZ (Sigma-Aldrich, USA) solution (40 mg/kg) dissolved in normal saline to the GDM group (Kim et al., 2014). Diabetes was confirmed following induction using the following procedure. The control group (Sal group) received an equivalent volume of normal saline upon STZ administration. Two days postinjection, diabetes confirmation was performed by measuring the 2-hr fasting blood glucose levels collected from the tail. All glucose measurements were conducted using the Accu-Chek Active blood glucose meter (Roche Diabetes Care, Germany). A reading exceeding 300 mg/dL was designated as confirmed diabetic status (GDM status).
After birth, the offspring remained with their respective dams for a 28-day suckling period. Subsequently, the male offspring were weaned and allocated into four definitive experimental groups for the intervention phase: saline control (Sal-CON, n=12), saline exercise (Sal-Ex, n=12), GDM control (GDM-CON, n=12), and GDM exercise (GDM-Ex, n=12) (Fig. 1).
Schematic diagram of the experimental design and timeline. STZ, streptozotocin; GDM, gestational diabetes mellitus; Sal-CON, saline control; Sal-Ex, saline exercise; GDM-CON, gestational diabetes mellitus control; GDM-Ex, gestational diabetes mellitus exercise.
Treadmill exercise protocol
The exercise groups (Sal-Ex and GDM-Ex) underwent a 2-week rigorous training regimen, performing forced locomotion on a motorized treadmill (Daejong Instrument Industry, Korea) 5 times per week. Each daily session consisted of 60 min of running at 15 m/min. This core exercise period was preceded and followed by 5-min warm-up and cool-down phases, respectively, which were executed at a reduced speed of 8 m/min.
Morris water maze
The Morris water maze (MWM) test for spatial learning and memory was initiated 4 days prior to the completion of the 2-week intervention period spanned four consecutive days. The protocol was conducted based on previously established methods (Kim et al., 2020). The setup utilized a circular pool (90-cm diameter; 45-cm height) filled with 26°C±1°C water, housing a fixed, submerged platform (6-cm diameter, 1 cm below the surface). Spatial acquisition (day 1 to day 3) involved daily trials where the platform was rendered invisible using nontoxic white paint. Escape latency was measured for up to 60 sec; animals failing to find the platform were guided to it and allowed to rest for 15 sec. The probe test (day 4) was conducted for 1 min with the platform removed, quantifying the time spent in the target quadrant. Data acquisition utilized the SMART 3.0 video tracking system (Panlab, Spain).
T-maze test
The T-maze test was performed during the final days of the intervention period, preceding animal sacrifice, to evaluate spatial working memory. The procedures were modified from previous study (Baek et al., 2025). The maze, constructed from black acrylic, consisted of one starting arm and two goal arms (left and right). During the training phase, one goal arm (right) was blocked while the other arm (left) was open, and mice were allowed to freely explore the accessible route for 10 min. Twenty-four hr later, in the test phase, animals were again placed in the starting arm with both goal arm open, and arm entries were manually counted for 10 min. An entry was recorded only when the animal’s entire body, excluding the tail, entered an arm. After each trial, the T-maze was disinfected with 70% ethanol to eliminate olfactory cues. Preferences for the new route was expressed as the percentage of entries into the new arm relative to the total number of arm entries [(new arm entries/total entries)×100]. A difference score was also calculated as the percentage of new arm entries minus the percentage of old arm entries (new–old).
Western blotting
Hippocampal tissues were homogenized in ice-cold PRO-PREP protein extraction solution (iNtRON Biotechnology, Korea) using chilled 50-mm stainless-steel beads for 1 min at 50 Hz with a TissueLyser II homogenizer (Qiagen, Germany). Protein concentration was determined using the Bradford assay with bovine serum albumin as a standard. Equal amounts of protein (20 μg per lane) were mixed with Laemmli sample buffer, boiled at 100°C for 5 min, and separated on 12% sodium dodecyl sulfate–polyacrylamide gel (Bio-Rad, USA).
After electrophoresis, the gels were activated under ultraviolet light and total protein was visualized using the Stain-Free imaging system to confirm uniform protein migration and loading prior to transfer. Proteins were then transferred onto polyvinylidene difluoride membranes (Merck Millipore, USA) using a wet-transfer apparatus at 120 V for 1 hr at 4°C. Following transfer, the membranes were blocked for 1 hr at room temperature in 5% (w/v) skim milk dissolved in TBS-T (20 mM Tris-HCl, 150 mM NaCl, 0.05% TWEEN-20).
Membranes were incubated overnight at 4°C with primary antibodies. The antibodies used included Wnt3 (Abcam, UK), a key Wnt signaling ligand involved in promoting neurogenesis and synaptic plasticity; GSK-3β (Santa Cruz Biotechnology, USA), a pivotal kinase that regulates synaptic plasticity and promotes β-catenin degradation; and β-catenin (Cell Signaling Technology, USA), the central effector protein of the canonical Wnt pathway essential for gene transcription related to neuronal proliferation and synapse function. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (Abcam), a housekeeping protein, was used as a loading control.
After washing 3 times with TBS-T, membranes were incubated for 1 hr with horseradish peroxidase-conjugated secondary antibodies (anti-mouse or anti-rabbit IgG as appropriate). Immonoreactive bands were visualized using an enhanced chemiluminescence substrate and captured with a ChemiDoc MP Imaging System (Bio-Rad). Densitometric quantification was performed using AlphaEase FC software (Alpha Innotech, USA), and expression levels were normalized to GAPDH.
Statistical analysis
Statistical analyses and data visualization were conducted using GraphPad Prism ver. 9.0.0 (GraphPad Software, USA). Maternal fasting blood glucose levels were compared using a Student t-test. For the MWM, escape latencies for the hidden versus exposed platforms within each experimental group were analyzed using a paired t-test, while the hidden-to-exposed latency ratios were compared across groups using a one-way analysis of variance (ANOVA) followed by Tukey post hoc test. In the T-maze test, the number of entries into the new arm versus the old arm within each group was compared using a paired t-test, and the difference scores (the percentage of entries into the new arm minus the percentage of entries into the old arm) across groups were evaluated using a one-way ANOVA followed by Tukey post hoc test. Protein expression levels obtained via Western blotting were analyzed using a one-way ANOVA followed by Tukey post hoc test. Statistical significance was set at P<0.05, and all data are presented as mean±standard deviation.
RESULTS
Treadmill exercise alleviates spatial learning and reference memory deficits in GDM offspring
The successful induction of maternal hyperglycemia was first confirmed by maternal blood glucose monitoring, which revealed that the glucose concentration in the STZ group was significantly higher than that in the saline group (P<0.0001) (Fig. 2). To assess the resulting cognitive impact on the offspring and the potential therapeutic efficacy of treadmill exercise, the MWM test was utilized. Qualitative analysis of the swimming track plots in Fig. 3A demonstrated that while offspring from the saline-treated groups (Sal-CON and Sal-Ex) exhibited localized and efficient search patterns focused on the target area, the GDM-CON offspring displayed disorganized and diffuse navigation paths, which were notably improved in the exercise-treated group (GDM-Ex). Regarding the quantitative data in Fig. 3B, although the absolute escape latencies for the hidden and exposed platforms within each experimental group showed no statistically significant differences, the hidden-to-exposed latency ratio served as a more critical and sensitive indicator of hippocampal reference memory. As shown in Fig. 3C, the GDM-CON exhibited a significantly lower latency ratio compared to the Sal-CON (P<0.05), confirming a profound spatial memory deficit resulting from intrauterine hyperglycemia. Crucially, the treadmill exercise intervention successfully alleviated this impairment, as the latency ratio in the GDM-Ex was significantly higher than that in the GDM-CON (P<0.05) and was found to be comparable to the ratios observed in the Sal-CON.
Confirmation of gestational diabetes mellitus (GDM) induction through maternal fasting blood glucose levels. Blood glucose concentrations were measured in pregnant mice 2 days after the single intraperitoneal injection of either saline or streptozotocin (STZ, 40 mg/kg) on gestational day 7. Statistical analysis was performed using the Student t-test. Values are presented as mean±standard deviation. ****P<0.0001 versus the saline group.
Postnatal treadmill exercise alleviates spatial learning and reference memory deficits in gestational diabetes mellitus (GDM) offspring. (A) Representative swimming track plots during the Morris water maze task. (B) Escape latency to reach hidden and exposed platforms. Paired t-test analysis revealed no statistically significant differences (N. S.). in absolute escape latencies within each experimental group. (C) Ratio of hidden-to-exposed escape latency, representing spatial reference memory performance. Sal-CON, saline control; Sal-Ex, saline exercise; GDM-CON, gestational diabetes mellitus control; GDM-Ex, gestational diabetes mellitus exercise. Statistical analysis for the ratio was performed using one-way analysis of variance followed by Tukey post hoc test. Values are presented as mean±standard deviation. *P<0.05.
Treadmill exercise improves spatial working memory performance in GDM offspring
To evaluate the impact of maternal hyperglycemia on spatial working memory, the T-maze spontaneous alternation test was conducted. As shown in Fig. 4A, offspring from the saline-treated control groups (Sal-CON and Sal-Ex) exhibited a significant preference for the new route, with the percentage of entries into the new arm being significantly higher than those into the old arm (P<0.001 and P<0.0001, respectively). In contrast, the GDM-CON showed no significant difference between the entries into the new and old arms, indicating a lack of preference and a corresponding deficit in spatial working memory. However, treadmill exercise effectively restored this cognitive preference, as the GDM-Ex displayed a significantly higher percentage of entries into the new arm compared to the old arm (P<0.001). Quantitative analysis of the difference score in Fig. 4B, which represents the percentage of new arm entries minus the percentage of old arm entries, revealed that the score in the GDM-CON was significantly lower than that of the Sal-CON (P<0.05) and Sal-Ex (P<0.001). Importantly, the treadmill exercise intervention successfully rescued this impairment, with the difference score in the GDM-Ex being significantly higher than that in the GDM-CON (P<0.01) and restored to levels comparable with normal control performance.
Postnatal treadmill exercise improves spatial working memory performance in gestational diabetes mellitus (GDM) offspring. (A) Percentage of entries into the old versus new goal arms during the T-maze test. Paired t-test analysis confirmed that the saline-control (Sal-CON, P<0.001), saline-exercise (Sal-Ex, P<0.0001), and GDM-exercise (GDM-Ex, P<0.001) exhibited a significant preference for the new arm. (B) Difference scores calculated as the percentage of new arm entries minus the percentage of old arm entries, representing spatial working memory efficiency. Sal-CON, saline control; Sal-Ex, saline exercise; GDM-CON, gestational diabetes mellitus control; GDM-Ex, gestational diabetes mellitus exercise. Statistical analysis for the difference score was performed using one-way analysis of variance followed by Tukey post hoc test. Values are presented as mean±standard deviation. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.
Treadmill exercise modulates the hippocampal Wnt/β-catenin signaling pathway in GDM offspring
To investigate the molecular mechanisms underlying the cognitive rescue effects of exercise, we analyzed the expression of key proteins in the canonical Wnt/β-catenin pathway using Western blot analysis (Fig. 5). Fetal exposure to hyperglycemia significantly suppressed the hippocampal Wnt signaling cascade in the offspring; the protein expression levels of Wnt3 and β-catenin in the GDM-CON were significantly lower than those in the Sal-CON (P< 0.05). Conversely, the expression of GSK-3β, a key inhibitor that promotes the degradation of β-catenin, was significantly higher in the GDM-CON compared to the saline control groups (P<0.05). Crucially, postnatal treadmill exercise intervention successfully mitigated these prenatal insults by restoring the molecular balance within the hippocampus. The GDM-Ex exhibited significantly higher levels of Wnt3 and β-catenin, along with a significantly lower level of GSK-3β activity compared to the sedentary GDM-CON (P<0.05). These results indicate that treadmill exercise alleviates GDM-induced cognitive deficits by reactivating the Wnt/β-catenin signaling pathway, which is essential for maintaining neurogenesis and synaptic plasticity.
Postnatal treadmill exercise restores the hippocampal Wnt/β-catenin signaling pathway in gestational diabetes mellitus (GDM) offspring. (A) Representative Western blot bands for β-catenin, glycogen synthase kinase-3β (GSK-3β), and Wnt3 in the hippocampus. (B) Relative protein expression levels of Wnt3, GSK-3β, and β-catenin normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Sal-CON, saline control; Sal-Ex, saline exercise; GDM-CON, gestational diabetes mellitus control; GDM-Ex, gestational diabetes mellitus exercise. Statistical analysis for protein expression was performed using one-way analysis of variance followed by Tukey post hoc test. Treadmill exercise successfully restored the prenatal GDM-induced downregulation of Wnt3 and β-catenin and inhibited the upregulation of GSK-3β in the offspring hippocampus. Values are presented as mean±standard deviation. ***P<0.001, ****P<0.0001.
Consistent with accumulating epidemiological observations linking GDM exposure to altered neurocognitive outcomes in offspring, our behavioral data indicate impairments in both spatial reference memory and cognitive flexibility. Large human cohort studies have reported lower global or domain-specific cognitive scores in adolescent and adult offspring exposed to GDM, even after partial adjustment for socioeconomic and perinatal confounders.
DISCUSSION
GDM is a prevalent metabolic complication of pregnancy, and global estimates indicate that approximately one in seven pregnancies is affected worldwide, underscoring the scale at which the developing fetal brain may be exposed to hyperglycemia and its downstream biological consequences. In the present study, we demonstrate that intrauterine exposure to maternal hyperglycemia induces persistent impairments in hippocampus-dependent cognitive functions in male offspring and that these behavioral alterations are accompanied by suppression of canonical Wnt/β-catenin signaling in the hippocampus. Importantly, we show that a brief period of moderate-intensity postnatal treadmill exercise is sufficient to normalize both cognitive performance and key molecular components of the Wnt pathway, supporting the concept that adverse neurodevelopmental programming induced by GDM is not fixed and can be modified after birth.
Evidence from human studies suggests that the relationship between GDM exposure and offspring cognition is complex and often modest in magnitude. In adult offspring of women with diet-treated GDM, cognitive performance has been reported to be slightly lower compared with unexposed controls; however, these differences are substantially attenuated after adjustment for socioeconomic, parental, and perinatal confounders (Clausen et al., 2011). Similarly, studies assessing neurocognitive development in children exposed to GDM have identified associations with altered cognitive or behavioral outcomes, but causality remains difficult to establish due to the strong influence of postnatal environment and maternal factors (Nomura et al., 2012). These observations indicate that, while human data raise concern regarding potential neurodevelopmental vulnerability, mechanistic insight requires controlled experimental models.
In this context, animal studies have provided more direct evidence that gestational hyperglycemia can disrupt hippocampal development and function. Prenatal exposure to maternal diabetes has been shown to induce oxidative stress, neuroinflammation, and neuronal disorganization in the offspring hippocampus, leading to impairments in learning and memory tasks (Plagemann, 2005; Rodolaki et al., 2023; Vuong et al., 2017). Our behavioral findings are consistent with this literature, as GDM-exposed offspring displayed deficits in both spatial reference memory and working memory. Notably, although absolute escape latencies in the MWM were not markedly altered, the reduced hidden-to-visible platform latency ratio suggests a selective impairment in spatial information processing rather than generalized motor dysfunction. This interpretation is further supported by the loss of novelty preference in the T-maze, a hippocampus-dependent measure of cognitive flexibility.
At the molecular level, our data identify suppression of canonical Wnt/β-catenin signaling as a plausible mechanistic link between gestational hyperglycemia and offspring cognitive dysfunction. Canonical Wnt signaling is essential not only for embryonic brain development but also for adult hippocampal neurogenesis, synaptic maintenance, and memory formation (Inestrosa and Varela-Nallar, 2015; Lie et al., 2005). Chen et al. demonstrated that regular aerobic exercise rescued down-regulated Wnt/β-catenin signaling and ameliorated age-related cognitive decline in aged rats, indicating that physical activity can restore canonical Wnt pathway components in models of central nervous system decline (Chen et al., 2020). In the present study, GDM offspring exhibited reduced hippocampal expression of Wnt3 and β-catenin together with increased GSK-3β, a kinase that promotes β-catenin degradation and effectively suppresses Wnt-dependent transcription. This molecular profile is consistent with previous reports showing that diabetes during pregnancy adversely affects GSK-3β regulation in the offspring hippocampus (Hami et al., 2015) and with evidence that maternal metabolic disturbances impair synaptogenesis-related markers, such as synaptophysin, in developing brain regions critical for cognition (Plagemann, 2005; Rodolaki et al., 2023).
A central finding of this study is that postnatal aerobic exercise reversed both the behavioral deficits and the molecular suppression of the Wnt pathway observed in GDM offspring. Exercise is a well-established modulator of hippocampal plasticity and has been shown to activate Wnt/β-catenin signaling in the adult brain, including normalization of pathway inhibitors and kinases that constrain β-catenin stability (Bayod et al., 2014; Inestrosa and Varela-Nallar, 2015). In models of metabolic dysfunction, aging, or brain injury, treadmill exercise restores Wnt signaling and improves cognitive outcomes (Bayod et al., 2014; Chen et al., 2020). Consistent with these reports, postnatal treadmill training in offspring born to diabetic dams has been shown to alleviate memory impairment and reduce apoptosis in the hippocampus (Kim et al., 2014). Our findings extend this work by directly linking exercise-induced cognitive recovery to normalization of the canonical Wnt/β-catenin signaling axis in a GDM offspring model.
Although the precise upstream mediators were not examined here, exercise is known to engage convergent signaling pathways (including PI3K–Akt, neurotrophic factor induction, and anti-inflammatory mechanisms) that inhibit GSK-3β activity and stabilize β-catenin (Chen et al., 2020). In addition, Wnt pathway antagonists such as Dkk1 have been shown to drive synapse loss and memory impairment when upregulated in the adult hippocampus, whereas suppression of Dkk1 restores synaptic integrity and cognitive function (Purro et al., 2012; Seib et al., 2013). While our study assessed Dkk1 transcript levels, further investigation at the protein and functional levels will be required to determine whether modulation of extracellular Wnt antagonists contributes to the exercise-induced rescue observed here.
Several limitations should be acknowledged. First, GDM was induced using STZ, which mimics insulin-deficient hyperglycemia and may not fully reproduce the typical insulin-resistant phenotype seen in many human GDM cases. Second, only male offspring were examined, despite evidence that prenatal metabolic programming and neurodevelopmental outcomes can be sexually dimorphic (Bale, 2011). Third, although ratio-based analysis in the MWM reduces confounding by motor factors, additional behavioral measures would further strengthen interpretation. Finally, molecular analyses were limited to selected components of the Wnt pathway; integration of synaptic markers, neurogenesis indices, inflammatory mediators, and epigenetic profiling would provide a more comprehensive mechanistic framework.
In summary, this study demonstrates that gestational hyperglycemia programs hippocampal cognitive vulnerability in offspring in association with suppression of canonical Wnt/β-catenin signaling and that early postnatal exercise is sufficient to reverse both molecular and behavioral abnormalities. These findings support the concept that neural plasticity remains accessible after adverse intrauterine exposure and identify the Wnt/β-catenin pathway as a key molecular node through which postnatal physical activity may mitigate the long-term neurodevelopmental consequences of GDM.
Notes
CONFLICT OF INTEREST
No potential conflict of interest relevant to this article was reported.
ACKNOWLEDGMENTS
This work was supported by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea (NRF-2023S1A5A2A01082966).