2.1 Animals and experimental design

In this study, 128 adult male Wistar rats (250–300 g each) were employed. The animals were kept in a controlled environment with a 12-h light–dark cycle, maintained at a constant temperature of 25 ± 1°C, and provided unrestricted access to both food and water. The Animal Ethics Committee of Kerman University of Medical Sciences reviewed and granted approval for all procedures involving animal experimentation (IR.KMU.AH.REC.1400.209).

In all groups receiving drug interventions (ASA, BM, or CM), drugs were injected every 2 days for 16 days (de Cristóbal et al., 2001; Rajizadeh et al., 2019; Tsai et al., 2018).

Behavioral tests were performed on the 8 groups, 16 days after the start of experiments in the following order: On the 16th day of the experiment, the open field test and novel object recognition test were performed. Subsequently, on Days 17 and 18, training sessions and probe trials of the Morris water maze were conducted. Finally, on Days 18 and 19, the passive avoidance learning test was carried out.

At the end of the experiment, the animals were subjected to deep anesthesia using ketamine (80 mg/kg) and xylazine (10 mg/kg), and their brains were extracted following decapitation. The right hippocampus was isolated and preserved at −80°C until it was processed in the Western blotting analysis for the assessment of brain-derived neurotrophic factor (BDNF) and cerebral dopamine neurotrophic factor (CDNF) proteins and the determination of oxidative stress markers. Furthermore, the left hippocampus was dissected for histological evaluation. In order to avoid the potential impact of the passive avoidance learning test on oxidative stress markers in this study, we utilized 64 individual rats that had exclusively undergone the drug treatment protocol for groups of our study without any behavioral testing being conducted on them. Figure 1 depicts an illustration of the experimental timeline.

2.2 Extraction, cultivation, and expansion of MSCs from adipose tissue

Human subcutaneous adipose tissue was acquired from patients who underwent elective liposuction surgery at the Department of Neurosurgery, Shahid Bahonar Hospital, Kerman, Iran. The procedures for harvesting adipose tissue and preparing human adipose–derived MSCs received approval from the ethical review boards of Kerman University of Medical Sciences. All patients or their authorized representatives provided informed written consent and were informed that the surgery would involve the collection of adipose tissue. The isolation of MSCs was carried out following procedures documented in the existing scientific literature (de Ugarte et al., 2003; Nakagami et al., 2005). Under aseptic conditions, adipose tissues were cut into small pieces before being subjected to digestion in a solution containing 4 mg of Collagenase Type I with a final concentration of 0.1% (Invitrogen Gibco). This digestion process was conducted with gentle agitation for 15 min at 37°C. Following digestion, the mixture was diluted with 4 mL of culture medium, specifically Dulbecco's modified Eagle's medium (DMEM) containing 15% fetal bovine serum (FBS). Subsequently, it was subjected to centrifugation at 1500 revolutions per minute (rpm) for 15 min to segregate the cellular fraction (pellet) from the adipocytes. The supernatant was discarded, and the cellular pellet was subsequently filtered through a 200-μm nylon mesh to eliminate undigested tissues. Afterward, the isolated cells were cultured in DMEM High Glucose supplemented with 15% FBS (Gibco), along with 100 U/mL of penicillin and 100 μg/mL of streptomycin. These cells were then incubated at a temperature of 37°C in an atmosphere with 5% CO₂. The initial medium change was carried out approximately 2 days after the start of the culture, during which the non-adherent cells were removed. Subsequently, this medium change process was repeated every 48–72 h. When the MSCs reached a confluence of 80%–90%, they were subjected to trypsinization using 0.05% trypsin (Sigma) and 0.02% Ethylenediaminetetraacetic acid (EDTA) for the purpose of initiating a new passage. The cells were then cultured until they reached Passage 2 (Akhondzadeh et al., 2020; Shamsara et al., 2018).

2.3 Characterization of MSCs through flow cytometry analysis

To verify the extraction of MSCs from human abdominal fat, cell assessment was carried out using flow cytometry. During the second passage, a total of 100,000 cells were marked with specific antibodies, including CD34-FITC, CD45-FITC, CD105-FITC, and CD90-FITC (Thermo Fisher). These cells were then incubated for 30 min in a refrigerator. Following two washes with phosphate buffered saline (PBS), the cells were fixed in a solution consisting of PBS with 1% paraformaldehyde and 1% FBS. The analysis of the specific fluorescent labeling was conducted using WinMDI 2.9 software (Mehrabadi et al., 2020; Sheibani et al., 2015).

2.4 Assessment of MSCs through a differentiation assay

The differentiation capability of MSCs into osteocyte and adipocyte lineages was evaluated during the second passage:

2.5 Preparation of MSCs’ conditioned medium

CM collection utilized adipose tissue–derived MSCs from the fourth passage. When the cells reached approximately 80% confluence, they were cultured for a duration of 48 h in serum-free DMEM. Subsequently, the collected conditioned media was subjected to centrifugation at 1800 rpm for 20 min at 4°C to eliminate cell debris. It was then filtered through a 0.22 μm syringe filter (Mehrabadi et al., 2020). Subsequently, supernatants from the MSCs were gathered. The collected media were concentrated by subjecting them to centrifugation at 6000 g using an Amicon Ultra-10 device from (Millipore Corporation) (Egashira et al., 2012). The protein concentration of MSC–CM was determined to be within the range of 700–1400 μg/mL using the Pierce bicinchoninic acid (BCA) protein assay kit (Thermo Fisher Scientific). Following concentration, the media were stored at −80°C for future use.

2.6 Hippocampal damage induction

To conduct the surgical procedure for hippocampal damage, the animals underwent deep anesthesia via an IP injection of ketamine (100 mg/kg) and xylazine (20 mg/kg). Subsequently, they were positioned in the skull-flat orientation on a rat stereotaxic instrument (Stoelting). The hair on the surface of the skull was shaved, and an incision was made to reveal the skull. Two openings were drilled into the skull at precise coordinates to access the CA1 region of the hippocampus. These coordinates were (3.8 mm dorsal to the bregma, 2.4 mm deep from the surface of the dura, and ±2.2 mm laterally) (Hanie et al., 2024). Hippocampal damage in the experimental model was induced bilaterally through a single direct injection of 3 μL of a 0.01% solution of EB from (CinnaGen), into sterile 0.9% saline using a 22-G needle (Rajizadeh et al., 2019). Figure 2a,b shows the different regions of the hippocampus and the path of the needle throughout the cortex to the hippocampus (Figure 2a,b).

2.7 Behavioral tests

After 16 days of the experiment and the administration of drug agents, the rats underwent the following behavioral test. An investigator conducted the test on the animals in a blinded manner, and data were collected using a video image motion analyzer.

2.8 Assessment of hippocampal oxidative stress markers

After completing the behavioral tests, the rats were deeply anesthetized through IP injections of ketamine (80 mg/kg) and xylazine (10 mg/kg). After making sure the animal was anesthetized, it was sacrificed by the cervical dislocation method. The skin and muscles of the neck and head were dissected. Afterward, using a special guillotine, the animal's head was separated. The scalp was cut with scissors along the sagittal suture. The muscles were removed. By placing the tip of the scissors in the foramen magnum and repeating this procedure from the nasal side, the brain and meninges were slowly exposed. The removal of the meninges was done carefully. Following the removal of the brain, the hippocampal tissue was dissected and subsequently homogenized in cold PBS (with a pH of 7.4). The specimens were subjected to centrifugation at 10,000 times the force of gravity (at 4°C). Afterward, the resulting supernatants were gathered and preserved at −80°C until they were ready for subsequent examination.

The concentration of malondialdehyde (MDA) was determined using the thiobarbituric acid reactive substance method (Zarin et al., 2019). Superoxide dismutase (SOD) and total antioxidant capacity (TAC) levels were measured using commercially available assay kits (KIAZIST Life Sciences) employing the colorimetric method. To determine the protein content in the tissue, the BCA Protein Quantification Kit (Parstous) was employed (Aghaei et al., 2023; Rajizadeh et al., 2023).

2.9 Western blot analysis

Hippocampus tissue specimens were homogenized with the use of a lysis buffer, and subsequently, total protein extract was obtained through centrifugation at 15,000 rpm for 5 min. The protein content in the resulting supernatants was quantified using Bradford's method. Lysates containing 60 μg of protein each were loaded onto a 12% SDS–PAGE gel and subsequently transferred to a polyvinylidene difluoride (PVDF) membrane (Chemicon Millipore Co.). The membranes were blocked using a 2% blocking reagent from the Electrochemiluminescence Advanced Kit (Amersham Bioscience Co.) and were then incubated separately with primary antibodies overnight. Following three washes with Tris-buffered saline with Tween 20, the blots were subjected to incubation with a secondary antibody, specifically rabbit IgG horseradish peroxidase conjugated (dilution of 1/3000, Cell Signaling Technology Co.), for a duration of 1 h at room temperature. Additionally, the bands that displayed reactivity were identified using a chemiluminescence kit reagent (Amersham Bioscience Co.). Quantification of the blots was carried out using the ImageJ software. Rabbit anti-β-Actin (dilution of 1/1000, Cell Signaling Technology Co.) served as the internal control (Akhondzadeh et al., 2020; Bejeshk et al., 2023).

2.10 Histological studies

Following the completion of behavioral tests, the animals were subjected to deep anesthesia by administering a combination of ketamine (at a dose of 80 mg/kg) and xylazine (at a dose of 10 mg/kg). After cutting the olfactory bulb, the brain was removed from the skull cavity. An incision about 3–4 mm from the frontal pole was made with a razor to separate the frontal cortex and hippocampus. The left hippocampus was meticulously extracted and then placed in a 10% formalin solution for preservation. The specimens were cut into sections after being embedded in optimal cutting temperature using cryomolds. Sequential coronal sections of the hippocampus, each measuring 7 μm in thickness, were meticulously sliced using a cryostat machine (SLEE). These sections were then preserved in a cryoprotectant solution at −20°C until they were ready to be utilized for Cresyl violet staining. A uniform systematic random sampling approach was employed, where every seventh section, with intervals of 350 μm, was placed onto slides coated with gelatin. These slides were then left at room temperature overnight. In each group, 100 fields on the slides were chosen at random. The selection of fields was done by matching them between the two groups based on the adult rat brain atlas (Watson, 2007). The samples were examined using an optical microscope set at a magnification of ×400. Only pyramidal neurons located in the CA1 region of the left hippocampus, which displayed a distinct and visible nucleus as well as a nucleolus, were categorized as viable and undamaged cells (Zangiabadi et al., 2019).

2.11 Statistical analysis

The data underwent statistical analysis using Prism (Version 8). In order to assess the normality of data distribution, the Shapiro–Wilk test or K–S test will be employed. For normally distributed data, parametric tests such as repeated-measures analysis of variance (ANOVA) and one-way ANOVA with Bonferroni correction were used. Otherwise, the nonparametric Kruskal–Wallis test followed by Dunn's multiple comparisons test was utilized. The data were presented using the mean ± standard error of the mean, and a significance level of p < .05 was established as the threshold for statistical significance.

3.1 Demonstration of mesenchymal cells

Confirmation of MSCs derived from adipose tissue was done through flow cytometry analysis. MSCs showed higher mesenchymal markers (CD105 and CD90) and lower hematopoietic markers (CD34 and CD45) (Figure 3). Moreover, to confirm that stem cells are mesenchymal, these cells were differentiated into adipo- and osteo-cells. During the adipogenic differentiation of the cells, the lipid droplets formed were stained red using the Oil Red staining method (Figure 4a). In the Alizarin Red staining method, osteogenic activity areas were observed as red areas (Figure 4b).

3.2 Behavioral tests

3.2.1 Morris water maze

EB group demonstrated a significantly higher traveled distance (Figure 5a) and escape latency time (Figure 5b) in Blocks 2 and 3 for locating the platform over 2 days compared to the equivalent metrics in the control group. Similarly, in EB/BM group, an increase in both the distance traveled and the time taken to locate the hidden platform during Blocks 2 and 3 was observed. Administration of ASA, CM, or co-administration of CM/ASA resulted in a significant improvement in learning and memory as seen by reduced distance traveled and escape latency in Blocks 2 and 3 compared to the EB group (Figure 5a,b). Notably, the EB/CM/ASA group exhibited a reduction in both the traveled distance and escape latency when compared to the EB/ASA group, suggesting a synergistic effect of ASA and CM (Figure 5a,b).

In the probe trial, it was seen that the percentage of time spent in the target quadrant was significantly lower in both the EB and EB/BM groups when compared to the control group (Figure 5c). On the other hand, the administration of ASA and CM in EB/ASA, EB/ASA/BM, EB/CM, and EB/CM/ASA groups resulted in a significant increase in the percentage of target quadrant time compared to the EB group. Furthermore, the EB/CM/ASA group displayed significant increase in the percentage of target quadrant time compared to the EB/ASA group (Figure 5c).

3.2.2 Open field test

When compared to the control group, the time spent in the inner zone and the number of rearing were significantly decreased in EB and EB/BM groups (Figure 6a,d). Moreover, the time spent in the outer zone and the number of grooming were significantly increased in these groups compared to the control group (Figure 6b,c). However, EB/ASA, EB/ASA/BM, EB/CM, and EB/CM/ASA groups showed significantly decreased time spent in the periphery and the number of grooming (Figure 6b,c) as well as increased time spent in the central zone and the number of rearing (Figure 6a,d) compared to the EB group. Nevertheless, there were no significant differences in total traveled distance (Figure 6e) and movement velocity (Figure 6f) among the groups.

3.2.3 Novel object recognition

There were no significant differences in discrimination ratio during the training phase of novel object recognition test (Figure 7a). However, EB and EB/BM groups demonstrated significant decreased discrimination ratio during test phase compared with the control group (Figure 7b). Moreover, EB/ASA, EB/ASA/BM, EB/CM, and EB/CM/ASA groups had significantly increased discrimination ratio during test phase compared with the EB group (Figure 7b).

3.2.4 Passive avoidance learning and memory

EB and EB/BM groups showed a significantly increased number of shocks compared with control group (Figure 8a), implying impaired passive avoidance learning. However, EB/ASA, EB/ASA/BM, EB/CM, and EB/CM/ASA groups showed a significantly decreased number of shocks compared with the EB group (Figure 8a), suggesting that these agents reversed learning impairments induced by EB. Furthermore, STL was reduced in EB and EB/BM groups compared with the control group (Figure 8b). However, EB/ASA, EB/ASA/BM, EB/CM, and EB/CM/ASA demonstrated significantly less desire toward the dark chamber compared to the EB group, suggesting that ASA and CM have a beneficial effect in alleviating the harmful impact of EB on passive avoidance memory.

3.3 Oxidative markers

The hippocampal MDA levels in EB and EB/BM rats were observed to be significantly higher than in the control group (Figure 9a). On the other hand, the MDA concentration was significantly reduced to a level comparable to control group in the EB/ASA, EB/ASA/BM, EB/CM, and EB/CM/ASA groups (Figure 9a). Moreover, the EB and EB/BM groups exhibited reduced SOD antioxidant activity compared to the control group (Figure 9b). Reciprocally, the SOD activity in the EB/ASA, EB/ASA/BM, EB/CM, and EB/CM/ASA groups was significantly higher than in the EB group (Figure 9b). Furthermore, the hippocampal TAC levels were significantly increased following EB and EB/BM administrations when compared to the control group (Figure 9c). However, the TAC levels in EB/ASA, EB/ASA/BM, EB/CM, and EB/CM/ASA groups showed significant increase than the EB group (Figure 9c).

3.4 Western blot analysis

Given that the administration of EB is likely to result in damage to the hippocampal regions, we assessed the potential impact of ASA and CM administrations on BDNF and CDNF proteins within the hippocampal tissue samples. EB and EB/BM groups had a decrease in BDNF (Figure 10a) and CDNF (Figure 10b) protein levels in the hippocampus tissue samples compared with the control group. When ASA and/or CM were accompanied with EB, we observed a relative elevated level of BDNF (Figure 10a) and CDNF (Figure 10b) protein levels compared to EB group.

3.5 Histological characteristics of the hippocampus

The hippocampus was divided into four subdivisions, namely, CA1, CA2, CA3, and the dentate gyrus (DG), based on the cytoarchitectural characteristics of its principal cells. Under light microscopic examination, Cresyl violet–stained sections of all groups revealed a hippocampus with a C-shaped structure (Figure 11). The assessment of CA1 area of the hippocampus in experimental groups demonstrated that, in the control group (Figure 12a), CA1 consisted of five to six densely packed layers of small pyramidal cells arranged in a distinct palisade-like pattern. The pyramidal cells displayed vesicular nuclei, and the cytoplasm of the cells exhibited a pale and basophilic appearance. Examination of the EB and EB/BM groups (Figure 12c,d) revealed significant disorganization, scattering, and a reduction in the number of small pyramidal cells within the CA1 region. These cells appeared darker with hyperchromatic, pyknotic nuclei. Following treatment with ASA or CM, there was a notable improvement in the previously mentioned cell death within the hippocampus. In the Cresyl violet–stained sections of the EB group that received ASA, ASA/BM, and CM (Figure 12e–g), there was a preservation of the smallest pyramidal cells in the CA1 area. Most of these cells exhibited vesicular nuclei, and only a few nuclei appeared pyknotic. The combination therapy of ASA with CM resulted in a significant enhancement in the preservation of pyramidal cells, which exhibited less pronounced shrinkage, and there were only a few apoptotic cells with pyknotic nuclei (Figure 12h). In the Sham group (Figure 12b), histological characteristics were similar to those of the control group, and the small pyramidal cells were organized in a distinctive palisade-like pattern. These cells had pale basophilic cytoplasm with vesicular nuclei. There was a significant increase in the number of pancellular necrosis pyramidal CA1 cells in both the EB and EB/BM groups compared to the control group (Figure 13). However, EB/ASA, EB/ASA/BM, EB/CM, and EB/CM/ASA groups demonstrated less pancellular necrosis pyramidal CA1 cells compared to the EB group (Figure 13).