Milk fat globule membrane ameliorates depressive-like behaviors in chronic unpredictable mild stress rats by modulating the microbiota-gut-brain axis

Abstract

Depression is one of the common psychiatric disorders, and it has been reported that the imbalance in the microbiota-gut-brain (MGB) axis contributes to the pathogenesis of depression. Milk fat globule membrane (MFGM) can impact the gut-brain axis by regulating the intestinal flora and metabolite production. The aim of this study was to investigate whether MFGM could ameliorate depressive-like behaviors induced by chronic unpredictable mild stress (CUMS) and further elucidate the potential mechanism through a fecal microbiota transplantation (FMT) experiment. Male Sprague-Dawley rats were provided with an MFGM diet for 5 weeks after the induction with CUMS. Depressive-like behaviors were assessed, and the levels of neurotransmitters, neuroendocrine hormones, microbiota, short-chain fatty acids (SCFAs), and tight junction proteins, including occludin and zonula occludens-1 (ZO-1), were measured. It was revealed that MFGM could alleviate the depressive-like behaviors in CUMS rats. MFGM up-regulated the expression of occludin and ZO-1 and ameliorated intestine pathological changes in CUMS rats. Moreover, MFGM increased the levels of 5-hydroxytryptamine (5-HT), dopamine (DA), and norepinephrine and decreased the levels of neuroendocrine hormones in CUMS rats. Furthermore, it was confirmed that the concentrations of SCFAs, DA, 5-HT, and tight junction proteins significantly increased in the recipient rats that were inoculated with the fecal microbiota from the rats after MFGM treatment. These findings demonstrated that MFGM could alleviate depressive-like behaviors in CUMS rats and was possibly associated with modulation of the gut microbiota and up-regulation of SCFAs and monoamine neurotransmitters.

INTRODUCTION

Depression is one of the common psychiatric disorders characterized by persistent depressed mood, dysphoria, impaired motivation, and other symptoms. It significantly contributes to the global disease burden, affecting approximately 350 million individuals [1, 2]. The pathogenesis of depression is complex, including the neurotransmitter hypothesis, neuroendocrine disorder, immune response, neuroinflammation, and gut microbiota dysbiosis [3,4,5]. Prolonged use of treatment may induce side effects such as suicidality, self-perception deprivation, addiction, and weight gain [6]. Thus, there is an urgent need to discover and identify novel products with better antidepressant efficacy and fewer side effects.

As revealed in increasing studies, the intestinal flora is closely related to depression [7, 8]. It has been demonstrated that depression may cause intestinal flora dysbiosis and decrease beneficial bacteria, including Lactobacillus and Firmicutes [9]. Animal studies indicate that intestinal flora alteration and intestinal barrier impairment may increase the pro-inflammatory substances in the circulation, exacerbating systemic inflammation and oxidative stress, which lead to the development of depression [10, 11]. Furthermore, amending the imbalanced microbiota with a probiotic may ameliorate depression-like behaviors [11]. Recently, it has been reported that the imbalance in the microbiota-gut-brain (MGB) axis contributes to the pathogenesis of depression [12]. Thus, explorations based on the MGB axis might provide novel insights into the onset of depression, thereby facilitating the development of new antidepressant products.

As a complex tri-layer structure in milk, the milk fat globule membrane (MFGM) is mainly composed of phospholipids and proteins, constituting over 90% of its composition. It has been found that the protein in MFGM is beneficial to brain development, metabolic response, and gut health [13,14,15]. MFGM can influence the gut-brain axis via regulation of the intestinal flora and metabolite production [16, 17]. Moreover, phospholipids, especially the sphingomyelin (SM) in MFGM, can also regulate the intestinal flora composition [18, 19]. These results indicate that MFGM might affect depression by regulating the intestinal flora [20, 21]. Based on these findings, it can be hypothesized that MFGM could attenuate depression behaviors through modulation of the MGB axis. In this study, two experiments were performed to elucidate this hypothesis. The chronic unpredictable mild stress (CUMS) model of depression was constructed and treated with an MFGM diet to investigate the effect of MFGM on depression-like behaviors in CUMS rats. Moreover, a fecal microbiota transplantation (FMT) experiment was performed to further verify the potential mechanism of effect of MFGM on improving the depression-like behaviors in CUMS rats.

MATERIALS AND METHODS

Animals

Based on our previous study [22], the behavior test alteration rates were 20% and 50% for our control and CUMS groups, respectively. With a significance level (α) of 0.05 (two-tailed) and statistical power (1-β) of 0.90, the sample size calculation for detecting this for 30% difference in proportions between the groups using a two-sample proportions test indicated a minimum requirement of 8 rats per group. Considering the possibility of animal death during the experiment, 10 rats per group were used in the stage 1 experiment. Meanwhile, for the FMT experiment (stage 2), a sample size of 6 rats per group was selected, which was consistent with the sample size reported as effective in our previous methodology [22]. Therefore, a total of 50 adult male Sprague-Dawley (SD) rats (10 weeks old, body weight 270–300 g, HFK Bioscience Co., Ltd., Beijing, China) were used in this study, including 38 rats in stage 1 and 12 rats in stage 2. All rats were housed under pathogen-free conditions with a temperature of 22–24°C, humidity of 55 ± 10%, and a 12 hr light/dark cycle and were provided with unrestricted access to food and water. The animal study was approved by the Animal Ethical and Welfare Committee of Tianjin Nankai Hospital (NKYY-DWLL-2020-180).

Animal grouping and treatment

In the first experiment, after adaptive feeding, 38 rats were equally randomized into four groups: non-stressed plus saline on control feed (control, n=10), CUMS procedure plus saline on control feed (CUMS, n=10), CUMS plus saline on MFGM feed (CUMS-MFGM, n=10), and CUMS procedure plus fluoxetine on control feed (CUMS-Flu, n=8).

CUMS is a widely used behavioral model leading to depression-related behaviors through disruption of hypothalamic-pituitary-adrenal (HPA) axis homeostasis. CUMS procedures were induced as previously described [22]. Briefly, rats underwent 6 weeks of daily exposure to two randomized stressors from among food deprivation for 24 hr, water deprivation for 24 hr, swimming in 4°C water for 5–10 min, reversed light/dark cycle for 24 hr, 60 Hz noise for 1 hr, wet bedding for 8 hr, tail pinching for 2 min, cage shaking for 15 min, and physical restraint for 2 hr. Meanwhile, to ensure unpredictability and avoid the adaptability of rats to stressors, no stressors were repeated within 3 days.

All diets, including the control and MFGM diets, were administered to the rats in pellet form. Both the control and MFGM diets were formulated according to the AIN-76A standard (American Institute of Nutrition, 1977). The AIN-76 diet is commonly used for adult rodents because it can maintain sufficient energy and nutrition for rodents, and it is suitable for research in fields such as tumors, metabolic diseases, and the microbiome. In specific experiments, the compositions of feeds are often adjusted based on the AIN-76 diet according to the different experimental aims [23, 24]. MFGM was sourced from Fonterra Co-operative Group (Auckland, New Zealand), and Guangdong Medical Laboratory Animal Center was commissioned for production and processing. The effective dose of MFGM in the rats was 150 mg/kg/day. MFGM was the main difference in composition between the control diet and the MFGM-enriched diet [25]. The MFGM we used was separated from milk and then processed through concentration and drying steps to produce the MFGM feed. The MFGM mainly comprised phospholipids (100 g MFGM contained 6.3 g phospholipids). Detailed information about the MFGM composition and the MFGM diet are shown in Tables 1 and 2.

Table 1. Composition of phospholipids in the milk fat globule membrane (MFGM) diet.

Ingredient	Contents (%)
Phosphatidylcholine	26.4
Phosphatidylethanolamine	26.6
Phosphatidylinositol	9.3
Phosphatidylserine	12.7
Sphingomyelin	25.0
Total	100.0

Table 2. Compositions of the control and milk fat globule membrane (MFGM) diet formulations.

	Control	MFGM
Ingredient (g)
Casein	140.00	76.01
L-Cystine	1.80	1.81
Corn starch	495.69	494.68
Maltodextrin	125.00	125.35
Sucrose	100.00	100.28
Cellulose	50.00	50.14
Soybean oil	40.00	24.07
t-BHQ	0.01	0.01
Mineral mix	35.00	35.10
Vitamin mix	10.00	10.03
Choline bitartrate	2.50	2.51
MFGM ingredient	0.00	80.02
Total	1,000.00	1,000.00
Macronutrient energy (Kcals/kg)
Protein	494	494
Carbohydrate	2,883	2,883
Fat	450	450
Kcals (%)
Protein	13%	13%
Carbohydrate	75%	75%
Fat	12%	12%
Total	100%	100%

Mineral Mix: calcium, phosphonium, potassium, magnesium, manganese, barium, sodium, iron, zinc, copper, iodine, chlorine, sulfur. Vitamin Mix: vitamin A, vitamin D, vitamin E, vitamin K, vitamin B₁, vitamin B₂, vitamin B₃, pantothenic acid, vitamin B₆, vitamin B₁₂, folic acid, choline.

t-BHQ: tertiary butylhydroquinone.

Figure 1a shows the detailed experiment design. First, after a week of adaptive feeding, the behavioral test was performed to ensure that there was no difference in behavior at baseline among the groups, and body weight was measured every week. Then, all groups except the control group were treated to establish the CUMS model. After 6 weeks of stimulation, the behavioral test was carried out to ensure that the depression model was established. Throughout the period of the experiment, all rats received the control diet. Finally, during the intervention process, the CUMS-MFGM group received the MFGM diet, while the other groups continued to receive the control diet. Fluoxetine, a selective serotonin reuptake inhibitor (SSRI) antidepressant, has become the first-line treatment for patients with depression due to its favorable safety profile, fewer side effects and greater tolerance [26, 27]. Therefore, fluoxetine was administered to the CUMS-Flu group as a positive control in this study. The dosage of fluoxetine was 10 mg/kg/day. The other groups were provided with saline (1 mL/day) via oral gavage.

Fig. 1.

Effects of milk fat globule membrane (MFGM) on depressive-like behavior and brain histopathology in chronic unpredictable mild stress (CUMS) rats. (a) Experimental design for stage 1. (b) Weight gain of rats during the experiment. (c) Total distance traveled in the open-field test during weeks 0, 6, and 11. (d) Total activity time in the open-field test during weeks 0, 6, and 11. (e) Distance in the close arms in the elevated-plus test during weeks 0, 6, and 11. (f) Time in the closed arms in the elevated-plus test during weeks 0, 6, and 11. (g) Evaluation of histological changes in the brain in rats by hematoxylin and eosin (HE) staining (scale bar=100 μm). (h) Evaluation of histological changes in the brain in rats by Nissl staining. Arrows indicate depleted epithelial cells (scale bar=100 μm). Data are expressed as the mean ± SD (CUMS-Flu group, n=8; other groups, n=10). *p<0.05, compared with control group. ^#p<0.05, compared with CUMS group. behavior tests, open-field test and elevated-plus test; SC: sample collecting.

Preparation of fecal bacterial solution

Fresh fecal transplants were pooled from the donor rats in the CUMS and the CUMS-MFGM groups, respectively. Before FMT, the fecal samples were removed from a −80°C freezer and allowed to thaw for 10–15 min at room temperature. Subsequently, they were put into sterile EP tubes with sterilized tweezers on an ultra-clean bench and then immediately diluted 40-fold in phosphate-buffered saline (PBS) [28]. The supernatant was collected and used for fecal microbiota transplantation after centrifugation at 100 × g for 5 min at 4°C.

Antibiotic treatment and FMT experiment

After adaptive feeding, 12 male SD rats received an oral antibiotic regimen consisting of ampicillin, kanamycin, metronidazole, neomycin (all at 0.25 mg/day), and vancomycin (0.125 mg/day) for a duration of 14 consecutive days to eliminate the microbiota and establish an antibiotic-mediated microbiota depletion rat model. Then, the rats were randomly divided into those receiving fresh fecal samples from CUMS-donor rats (FMT+S, n=6) and those receiving fresh fecal samples from MFGM-donor rats (FMT+SM, n=6). All the rats were provided with 0.2 mL/day of the corresponding fresh fecal bacterial solution by gavage for 2 weeks [29].

During the experiment, if a rat was not in good condition for subsequent experiments, it was euthanized with 50% CO₂ per minute to shorten the time of death and reduce its suffering. Finally, after intervention and behavioral assessments, all rats were euthanized with 50% CO₂ per minute, and blood samples were collected from the abdominal aorta. The supernatant of the blood was collected and stored at −80°C. The brain and colon tissues were dissected and kept at −80°C for subsequent biochemical tests. Moreover, to determine the changes in the intestinal flora and short-chain fatty acids (SCFAs) caused by the FMT experiments, fecal samples were collected freshly from the colon.

Open-field test (OFT)

The rats subjected to the OFT were allowed to freely explore an open field arena for 5 min to assess exploratory and locomotor activities. A classic open field was selected as the testing apparatus, and the OFT was performed in a black square box (L 100 × W 100 × H 50 cm). The bottom of this square box was divided into 16 squares of equal size. The four central grids constituted the central area, and the outer grids constituted the peripheral area. Each experimental rat was placed individually in a fixed position within the center area, and their performance was monitored. Moreover, the ANY-maze system was used to automatically record the time spent in the center and peripheral areas and the distance traveled within the areas.

Elevated-plus maze test (EPM)

The EPM apparatus consisted of four arms (L 50 × W 10 cm) connected by a common 10 × 10 cm center area. Two oppositely facing arms were open, while the other two facing arms were enclosed with walls 40 cm high. During the experiment, the rats were allowed to freely explore the area for 5 min. Their movement distance and time spent in closed arms were recorded by a video tracking system (ANY-maze software, Stoelting, Wood Dale, IL, USA). The apparatus was cleaned with 75% alcohol and completely dried after the test was completed for each rat.

Histopathological analysis

For hematoxylin and eosin (H&E) staining procedures, both brain and colon tissue sections underwent standardized pretreatment as in our previous study [22].

Western blot analysis

Western blot was performed to analyze tight junction (TJ) protein expression in the brain and colon of these rats. Specifically, appropriate amounts of tissues were suspended in ice-cold RIPA lysis buffer and broken up using a hand-held homogenizer (SparkJade, Shandong, China). Subsequently, the tissues were lysed on ice for 30 min and then homogenized on ice using an ultrasound homogenizer. Next, the tissues were centrifuged at 14,000 rpm at 4°C for 15 min, and their supernatants were collected to measure the concentration of proteins using a bicinchoninic acid (BCA, SparkJade, Shandong, China) assay. Following electrophoretic separation via 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transfer to 0.45 μm polyvinylidene fluoride membranes, the membranes underwent blocking with 5% nonfat milk with Tris-buffered saline with Tween 20 (TBST) at room temperature for 1 hr. Then, they were incubated with primary antibodies-occludin (1:2000, bs-10011R, Bioss, Beijing, China) and zonula occludens-1 (ZO-1; 1:2000, bs-34023R, Bioss) at 4°C and gently shaken overnight. The membranes were then washed with TBST (3 times; 6 min each) and incubated with secondary antibodies for 1 hr at room temperature. GAPDH was used as a loading control to analyze the relative protein quantity. After washing with TBST (3 times; 6 min for each), immunoreactive proteins were visualized with an ECL Western Blotting Detection System (KeyGEN BioTECH, Nanjing, China) and a G:BOX Chemi XR5 imaging machine.

Enzyme-linked immunosorbent assay (ELISA) analysis

ELISA was used to detect the levels of neurotransmitters and their metabolites in serum and brain tissue. Excised brain tissue was placed into PBS solution (at a weight-volume ratio of 1:9), and then the sample was fully homogenized with a hand-held homogenizer. Finally, the supernatant was collected by centrifugation at 15,000 rpm for 15 min. Serum and brain tissue concentrations of neurotransmitters and their metabolites were determined using ELISA kits (Meimian Industrial Co., Ltd., Jiangsu, China) following the manufacturer’s instructions. The protein concentration was determined by BCA protein detection kit (SparkJade, Shandong, China).

16S rRNA sequencing

Sequencing of 16S rRNA was conducted using the NovaSeq 6000 platform. The total genomic DNA from these samples was extracted using the cetyltrimethylammonium bromide (CTAB) and sodium dodecyl sulfate (SDS) method. According to the selection of the sequenced region, the V3–V4 variable regions were amplified by polymerase chain reaction (PCR) using specific primers with barcoding and high-fidelity DNA polymerase. Then, the PCR products were resolved by 2% agarose gel electrophoresis for further detection. Subsequently, the PCR products were mixed at an appropriate ratio. Next, the PCR product mixture was purified with an AxyPrep DNA Gel Extraction Kit (Axygen). Following library construction with a NEBNext^® Ultra^™ DNA Library Prep Kit for Illumina (New England Biolabs, Ipswich, MA, USA), sequencing libraries were subjected to quality control assessments using a Qubit^® 2.0 Fluorometer (Thermo Fisher Scientific, Waltham, MA, USA) and Agilent 2100 Bioanalyzer system (Agilent, Santa Clara, CA, USA). The raw sequencing data underwent adapter trimming, quality filtering, and chimera removal to mitigate the impact of low-quality reads (Q-score <20). Subsequently, operational taxonomic units (OTUs) were clustered, and taxonomic classification was performed using these processed sequences. The Silva (version 138) database was used in this study.

Extraction and quantification of SCFAs

The composition and concentrations of SCFAs in the fecal samples were determined by gas chromatography-mass spectrometry (GC-MS; 7890B GC System and 5977B GC/MSD, Agilent). First, fecal samples were thawed on ice, and 30 mg of feces were placed into a 2 mL glass centrifuge tube. Then, 900 µL of 0.5% phosphoric acid was added to the tube, followed by shaking and mixing for 2 min and centrifugation at 14,000 g for 10 min. Subsequently, 800 µL of the supernatant was collected, and an equal amount of ethyl acetate was added to the tube for extraction, followed by shaking and mixing for 2 min. After centrifugation at 14,000 g for 10 min, 600 µL of the upper organic phase was collected, and 4-methylvaleric acid with a final concentration of 500 μM was added as an internal standard. The mixture was fully mixed for further detection. The gas chromatography conditions were as follows: An Agilent DB-WAX capillary column (30 m × 0.25 microns × 0.25 μM) was applied. The initial temperature was 90°C, and the temperature was increased to 120°C in 10°C/min increments and then increased to 150°C in 5°C/min increments. Finally, the temperature was increased to 250°C in 25°C/min increments and held for 2 min. The carrier gas was helium, with a flow rate of 1.0 mL/min. Instrument parameters were configured as follows: inlet 250°C, split ratio 10:1, injection volume 1 µL, and detector temperature 250°C.

Chromatographic parameters, including peak areas and retention times, were extracted by MSD ChemStation software, with subsequent construction of calibration curves for the determination of SCFA concentrations.

Statistics analysis

Data are expressed as the mean ± standard deviation (mean ± SD), and the statistical analyses were performed using IBM SPSS Statistics version 24.0. Differences among the various groups were statistically evaluated using either independent t-tests or one-way analysis of variance (ANOVA), further followed by the least significant difference (LSD) test. Principal Coordinates Analysis (PCoA) based on weighted and unweighted UniFrac metrics was used to assess the variation of bacterial composition among the different groups. A p-value of less than 0.05 was considered indicative of statistical significance.

RESULTS

MFGM improved depressive-like behaviors and the morphological changes of the hippocampus in CUMS rats

Male SD rats were subjected to the CUMS paradigm for 11 weeks and received the MFGM treatment in the sixth week (Fig. 1a). All rats gained weight continuously during the experiment (Fig. 1b).

The ability to perform self-motivated and exploratory activities was tested using the OFT and EPM. In the OFT, there were no significant differences in total distance traveled or movement time among the four groups at baseline (p>0.05). Compared with the control group, the total distance traveled and activity time of the other groups were reduced at the 6th week (p<0.05). Under CUMS conditions, the performance of the rats administered MFGM was reversed at the 11th week (p<0.05; Fig. 1c, 1d).

In the EPM test, the rats in the MFGM intervention group spent significantly less time and had a significantly increased travel distance in the closed arms compared with the CUMS group (p<0.05; Fig. 1e, 1f). These data indicated that the MFGM intervention could improve the ability of CUMS rats to perform self-motivated and exploratory activities.

As shown in Fig. 1g, neurons in the hippocampal CA1, DG regions, and cortex of the control rats were numerous, normal in shape, and closely arranged and had uniformly stained cytoplasm and normal space between nerve cells and the surrounding brain interstitium. In contrast, neurons had contracted nuclei, and indistinct nucleoli in CUMS rats were arranged irregularly. Notably, MFGM treatment significantly attenuated these CUMS-induced histopathological alterations, preserving neuronal architecture relative to the untreated CUMS group. It was found that the amounts of Nissl bodies in the hippocampus and cortex were higher in the control rats and MFGM-treated rats (Fig. 1h).

MFGM improved the levels of neurotransmitters and neuroendocrine hormones in CUMS rats

CUMS stimulation significantly reduced the concentrations of 5-hydroxytryptamine (5-HT), dopamine (DA) and norepinephrine (NE) in the serum and brain tissues of rats (p<0.05). Furthermore, the MFGM intervention significantly elevated brain neurotransmitter metabolite concentrations, including those of 5-hyroxyindole acetic acid (5-HIAA), dihydroxyphenylacetic acid (DOPAC), and homovanillic acid (HVA) (p<0.05). The MFGM intervention reversed the CUMS-induced decrease in neurotransmitters and their metabolites (p<0.05; Fig. 2a–2i).

Fig. 2.

Effects of milk fat globule membrane (MFGM) on the levels of neurotransmitter and neuroendocrine hormones in chronic unpredictable mild stress (CUMS) rats. (a and b) Levels of 5-hydroxytryptamine (5-HT) in serum and the brain. (c) Level of 5-hyroxyindole acetic acid (5-HIAA) in the brain. (d and e) Levels of dopamine (DA) in serum and the brain. (f) Level of dihydroxyphenylacetic acid (DOPAC) in the brain. (g and h) Level of norepinephrine (NE) in serum and the brain. (i) Level of homovanillic acid (HVA) in the brain. (j) Level of corticotropin-releasing hormone (CRH) in serum. (k) Level of corticosterone (CORT) in serum. (l) Level of adrenocorticotropic hormone (ACTH) in serum. Results are presented as the mean ± SD (in serum, n=8 for the CUMS-Flu group and n=10 for the other groups; in the brain, n=5 for all groups). *p<0.05, compared with control group. ^#p<0.05, compared with CUMS group.

Compared with the control group, the serum levels of corticotropin-releasing hormone (CRH), corticosterone (CORT), and adrenocorticotropic hormone (ACTH) were significantly increased in the CUMS group (p<0.05; Fig. 2j–2l). The serum levels of CRH, CORT, and ACTH in the CUMS-MFGM group and CUMS-Flu group were reduced compared with the CUMS group (p<0.05).

MFGM ameliorated the CUMS-induced alteration of the intestinal microbiota and increased SCFA levels

The richness and diversity indices (Chao1 richness, Shannon diversity, and Simpson diversity) significantly decreased in the CUMS group compared with the control group, while they increased in the CUMS-MFGM group (p<0.05; Fig. 3a–3d). PCoA analysis showed that the CUMS group was separated from the other groups, while the CUMS-MFGM group and CUMS-Flu group were clustered more closely together (Fig. 3e).

Fig. 3.

Effect of milk fat globule membrane (MFGM) on the intestinal microbiota changes induced by chronic unpredictable mild stress (CUMS). (a) Shannon index. (b) Simpson index. (c) ACE index. (d) Chao index. (e–g) Differences in microbiota flora abundance at the phylum, family, and genus levels among the four groups. (h) Total short-chain fatty acids (SCFAs). (i) Acetic acid. (j) Propitiate acid. (k) Isobutyric acid. (l) Butyric acid. (m) Isovaleric acid. (n) Valeric acid. (o) Hexanoic acid. Results are presented as the mean ± SD (n=6). *p<0.05, compared with control group. ^#p<0.05, compared with chronic unpredictable mild stress (CUMS) group.

The dominant intestinal microbiota across the phylum, family, and genus taxonomic ranks were selected for analysis of relative abundance (Fig. 3e–3g). At the phylum level, the relative abundances of Firmicutes and Actinobacteria decreased, while those of Proteobacteria and Bacteroidetes increased. The Firmicutes/Bacteroidetes ratio showed an upward trend in the CUMS rats. However, MFGM administration alleviated the alterations at the phylum level. There were no statistically significant differences among each group (p>0.05; Fig. 3e).

At the family level, the MFGM intervention reversed the changes in the relative abundances of Ruminococcaceae and Peptostreptococcaceae caused by CUMS (p<0.05; Fig. 3f). At the genus level, the MFGM intervention significantly decreased the relative abundance of Romboutsia (p<0.05) and reversed the changes in Ruminococcaceae UCG-005, Blautia, and Lactobacillus caused by CUMS stimulation (Fig. 3g).

The changes in the concentrations of SCFAs among the groups are shown in Fig. 3h–3o. Compared with the control group, the CUMS group exhibited reduced concentrations of total SCFAs, acetic acid, isobutyric acid, butyric acid, and valeric acid (p<0.05). In contrast, the levels of these SCFAs were increased in the CUMS-MFGM group compared with the CUMS group (p<0.05).

MFGM up-regulated the expression of TJ proteins in CUMS rats

The pathological changes and TJ protein expression in brain and colon tissues were evaluated to further assess the impact of MFGM on the intestinal barrier in CUMS rats. The intestinal villi of the control group exhibited a normal shape and irregular arrangement, without obvious damage (Fig. 4a). Compared with the control group, CUMS rats exhibited structurally disorganized intestinal villi with significant damage, and the intestinal mucosa became thinner. The MFGM intervention ameliorated the damage to the intestinal mucosa induced by CUMS.

Fig. 4.

Effects of milk fat globule membrane (MFGM) on intestinal morphology and barrier function in chronic unpredictable mild stress (CUMS) rats. (a) Evaluation of histological changes in the colon of rats by hematoxylin and eosin (HE) staining. (b–d) Protein expression of ZO-1 and occludin in brain tissue. (e–g) Protein expression of ZO-1 and occludin in colon tissue. Data are expressed as the mean ± SD (n=3). *p<0.05, compared with control group. ^#p<0.05, compared with CUMS group.

CUMS resulted in down-regulation of the expression of ZO-1 and occludin in both the brain and colon tissues (p<0.05). However, the MFGM intervention significantly up-regulated the expression of ZO-1 and occludin in the brain and colon tissues (p<0.05; Fig. 4b–4g).

FMT from MFGM intervention donor rats ameliorated the microbiota and increased SCFA levels in recipient rats

FMT was conducted to further verify the potential mechanism related to the anti-depression effects of MFGM on CUMS rats. Figure 5a illustrates the experimental design. Compared with the FMT+S group, the intestinal flora richness was increased in the FMT+SM group (p<0.05; Fig. 5d, 5e). The altered intestinal flora in the recipient rats demonstrated partial consistency with that of the donor rats.

Fig. 5.

Effect of fecal microbiota transplantation (FMT) on the diversity of gut microbiota in recipient rats. (a) Experimental design for stage 2. (b) Shannon index. (c) Simpson index. (d) ACE index. (e) Chao index. (f-h) Differences in microbiota flora abundance at the phylum, family, and genus levels between groups. (i) Total short-chain fatty acids (SCFAs). (j) Acetic acid. (k) Propitiate acid. (l) Isobutyric acid. (m) Butyric acid. (n) Isovaleric acid. (o) Valeric acid. (p) Hexanoic acid. Results are presented as the mean ± SD (n=6). ^#p<0.05, compared with fecal microbiota transplantation (FMT)+S group.

At the phylum level, compared with the FMT+S group, the relative abundance of Proteobacteria decreased significantly in the FMT+SM group (p<0.05), while those of Bacteroidetes and Cyanobacteria increased in this group (p>0.05; Fig. 5f). At the family taxonomic level, the FMT+SM group showed increased relative abundances of Ruminococcaceae and Prevotella compared with the FMT+S group (p<0.05; Fig. 5g). At the genus level, compared with the FMT+S group, the relative abundance of Ruminococcus 1 significantly increased in the FMT+SM group (p<0.05), while those of Romboutsia and Lactobacillus decreased significantly in this group (p<0.05; Fig. 5h).

Compared with the FMT+S group, the total SCFA level increased in the FMT+SM group (p<0.05). More specifically, the levels of acetic acid, propionic acid, butyric acid, isobutyric acid, isovaleric acid, and valeric acid increased (p<0.05; Fig. 5i–5p).

FMT from MFGM intervention donor rats up-regulated the expression of TJ proteins in recipient rats

There was a significant up-regulation of ZO-1 and occludin in the FMT+SM group compared with the FMT+S group in both brain and colon tissues (p<0.05), which suggested that MFGM could improve the barrier function via improvement of the microbiota (Fig. 6).

Fig. 6.

Effect of fecal microbiota transplantation (FMT) on the barrier function in recipient rats. (a–c) Protein expression of ZO-1 and occludin in brain tissue. (d–f) Protein expression of ZO-1 and occludin in colon tissue. Results are presented as the mean ± SD (n=3). ^#p<0.05, compared with fecal microbiota transplantation (FMT)+S group.

FMT from MFGM intervention donor rats regulated the levels of neurotransmitters and neuroendocrine hormones in recipient rats

Figure 7 shows the effect of MFGM on the levels of neurotransmitter and neuroendocrine hormones in the recipient rats. Compared with the FMT+S group, the levels of 5-HT and DA in the brain in the FMT+SM group increased significantly (p<0.05; Fig. 7a, 7b), which indicated that the intestinal flora from the CUMS-MFGM rats could regulate the levels of neurotransmitters.

Fig. 7.

Effects of fecal microbiota transplantation (FMT) on the levels of neurotransmitter and neuroendocrine hormones in recipient rats. (a) Level of 5-hydroxytryptamine (5-HT) in the brain. (b) Level of dopamine (DA) in the brain. (c) Level of corticotropin-releasing hormone (CRH) in serum. (d) Level of corticosterone (CORT) in serum. (e) Level of adrenocorticotropic hormone (ACTH) in serum. Data are expressed as the mean ± SD (n=6). ^#p<0.05, compared with fecal microbiota transplantation (FMT)+S group.

Compared with the FMT+S group, the serum levels of CRH and ACTH decreased in the FMT+SM group (p<0.05; Fig. 7c–7e), indicating that microbiota from CUMS-MFGM rats could suppress the production and release of neuroendocrine hormones.

DISCUSSION

In this study, it was indicated that MFGM could effectively enhance the voluntary activity and exploratory behavior of CUMS rats. The hippocampus is a crucial region responsible for emotional and cognitive functions [30]. As revealed in a previous study, CUMS stimulation induced a disrupted arrangement of nerve cells and decreased cell numbers in the hippocampus, which is consistent with our study [31]. Critically, the present study demonstrated that MFGM could significantly attenuate the hippocampus histopathological damage caused by CUMS. In addition, the results of the study have corroborated that MFGM increases neurotransmitters levels, decreases neuroendocrine hormones levels, up-regulates the expression of occludin and ZO-1, and reverses the disordered intestinal flora and SCFAs in CUMS rats. Importantly, fecal microbiota transplantation with CUMS-MFGM donor rats increased the concentrations of 5-HT, DA, occludin, and ZO-1 in recipient rats. Meanwhile, the alterations in the microbiota and SCFAs in the recipient rats were, in part, the same as those in the donor rats. These results illustrated that MFGM could mitigate depression-like behaviors in CUMS rats, possibly via the MGB axis.

Gut microbiome dysbiosis has been reported to correlate with depressive-like behaviors. For example, altered microbiota were found in depression patients and animals [32, 33]. Furthermore, FMT, as a microbiota-targeted technique, has been shown to aggravate depressive-like behaviors in recipient rats via the delivery of dysbiotic microbiota from donor rats with depression [34]. In contrast, our study found that the MFGM intervention significantly increased the richness and diversity of the intestinal flora in CUMS rats, as assessed by the Shannon, Chao1, ACE, and Simpson index. Importantly, recipient rats receiving FMT from CUMS-MFGM donor rats exhibited similarly restored levels of microbial richness and evenness, which demonstrated that MFGM restored the microbial diversity. These results indicated that MFGM could regulate intestinal flora, consistent with the findings of previous studies [35, 36]. At the phylum level, the dominant bacteria, Firmicutes and Bacteroidetes, were associated with the pathogenesis of depression by impacting the homeostasis of energy metabolism [37, 38]. Firmicutes promoted the metabolism of carbohydrates to produce various SCFAs, such as butyrate. This ameliorated the intestinal barrier integrity and restricted translocation of bacterial metabolites and endotoxins into the periphery or brain, thus alleviating inflammation and improving depression-like behavior [39]. Therefore, improvement of the abundances of Firmicutes and Bacteroidetes may be a potential indicator for the improvement of depression. Our results showed that the abundances of Firmicutes and Bacteroidetes were reversed in both the donor and recipient rats that were treated with MFGM. Glycosylation proteins and phospholipids in MFGM could promote the metabolism of the intestinal flora and stimulate the release of mucins, thus impacting the growth of microbiota [40]. Therefore, these glycosylation proteins and phospholipids in the MFGM diet may be the key component in regulating the intestinal flora.

SCFAs, key metabolites produced by the intestinal flora in the colon, might directly modulate depression by crossing the blood-brain barrier (BBB), thereby affecting the levels of neurotransmission, neurotrophic factors, and serotonin [41, 42]. Some SCFAs, such as acetic acid, propanoic acid, and butyric acid, are important neuro-mediators that can influence mood [43]. Consistent with previous studies [44], the levels of acetic acid, propanoic acid, butyric acid, valeric acid, and total SCFAs were decreased in the CUMS rats, while the MFGM treatment reversed the alterations. Emerging evidence indicates that changes in microbiota diversity may lead to variations in SCFA levels [45] and the abundances of Firmicutes [46], Lachnospiraceae [47], Ruminococcaceae [48], and Blautia [49], a member of the Lachnospiraceae family, which can produce SCFAs, thereby impacting intestinal function and resulting in depression-like behavior. In this study, these results demonstrated that the MFGM intervention increased the abundances of Firmicutes, Ruminococcaceae, Lachnospiraceae, and Blautia, supporting a microbiota-mediated mechanism for its antidepressant efficacy. These results were also observed in FMT. In addition, Prevotella, a new dominant flora in the recipient rats, could decrease the levels of SCFAs (especially acetic acid) that were augmented in the MFGM-FMT group [50]. Therefore, these results suggested that MFGM supplementation could regulate the levels of SCFAs by modulating the relative abundances of the SCFAs produced by the intestinal flora.

It has been reported that there is an obvious synergetic effect between the gut microbiota and the barrier in the gut [25]. Disturbance of the intestinal flora induced by chronic stress could compromise the intestinal mucosal barrier and elevate intestinal permeability. This would allow the intestinal flora (especially gram-negative flora) and SCFAs to pass through the gut barrier and enter the periphery or brain, thus affecting brain functions [51, 52]. As revealed in previous studies, the integrity of the gut barrier is impaired in CUMS rats [53], and other neuropsychiatric disorders, such as Parkinson’s disease [54] and Alzheimer disease [55], are associated with destruction of the intestinal barrier. TJ proteins (occludin and ZO-1) play an essential role in maintaining the permeability and integrity of the intestinal mucosal barrier, serving as physical barrier to prevent the diffusion of substances [56]. In this study, histopathological damage in the colon and a decreased level of TJ proteins were observed in CUMS rats. However, the MFGM intervention significantly alleviated the colon damage and up-regulated the expression of TJ protein in brain and colon tissues, indicating that MFGM could maintain the barrier function [57,58,59]. Additionally, FMT of MFGM from donor rats markedly up-regulated the expression of occludin and ZO-1 in the recipient rats. Therefore, these results suggest that MFGM could improve the integrity and reduce the permeability of the intestinal barrier and BBB, which was partially related to the regulation of the intestinal flora.

According to traditional theory, depression is mainly caused by a deficiency in monoamine neurotransmitters, including 5-HT, DA, and NE, and activation of the hypothalamic-pituitary-adrenal (HPA) axis [60, 61]. Our previous study found that the decreased level of monoamine neurotransmitters induced by CUMS stimulation was associated with specific intestinal flora [62]. In this study, both increased monoamine neurotransmitters and modulation of the microbiota related to monoamine neurotransmitters were observed in donor and recipient rats treated with MFGM. These findings corroborated that MFGM ameliorated depression via the regulation of microbiota, thus increasing the levels of monoamine neurotransmitters. It was found that SCFAs directly influence the hypothalamic mechanisms in the brain, which in turn affected the HPA axis [62].

In conclusion, the findings of the present study demonstrated that MFGM could alleviate depressive-like behaviors induced by CUMS and that this was possibly associated with alterations in the microbiota and up-regulation of the levels of SCFAs and monoamine neurotransmitters. However, the precise mechanism through which the MFGM intervention improved depression-like behavior by regulating the microbiota and SCFAs needs to be demonstrated by detecting the key pathways. Additionally, the lack of an FMT group receiving CUMS-MFGM microbiota without antibiotic depletion prevented us from disentangling microbiota effects from antibiotic-induced baseline alterations. Finally, the limitations of the current study included a small sample size and focus on only a short-phase MFGM intervention, whereas the therapeutic implications of chronic MFGM supplementation remain to be elucidated. Therefore, it is important to conduct further experiments with lager sample sizes to investigate the potential of MFGM for depression.

DATA AVAILABILITY

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

FUNDING

This work was supported by the Nutritional Science Research Foundation of BY-HEALTH (grant number TY202002002).

CONFLICT OF INTEREST

The authors report there are no competing interests to declare.