rTMS ameliorates depressive‐like behaviors and regulates the gut microbiome and medium‐ and long‐chain fatty acids in mice exposed to chronic unpredictable mild stress

Abstract Introduction Repetitive transcranial magnetic stimulation (rTMS) is a clinically useful therapy for depression. However, the effects of rTMS on the metabolism of fatty acids (FAs) and the composition of gut microbiota in depression are not well established. Methods Mice received rTMS (15 Hz, 1.26 T) for seven consecutive days after exposure to chronic unpredictable mild stress (CUMS). The subsequent depressive‐like behaviors, the composition of gut microbiota of stool samples, as well as medium‐ and long‐chain fatty acids (MLCFAs) in the plasma, prefrontal cortex (PFC), and hippocampus (HPC) were evaluated. Results CUMS induced remarkable changes in gut microbiotas and fatty acids, specifically in community diversity of gut microbiotas and PUFAs in the brain. 15 Hz rTMS treatment alleviates depressive‐like behaviors and partially normalized CUMS induced alterations of microbiotas and MLCFAs, especially the abundance of Cyanobacteria, Actinobacteriota, and levels of polyunsaturated fatty acids (PUFAs) in the hippocampus and PFC. Conclusion These findings revealed that the modulation of gut microbiotas and PUFAs metabolism might partly contribute to the antidepressant effect of rTMS.

6][7] There are multiple compositional differences in gut microbiota between patients with MDD and healthy controls, 8,9 and an altered gut metabolome contributes to depressive-like behavior in rats. 10Moreover, prebiotics and probiotics have antidepressive effects partly through regulation of gut microecology 11,12 and fecal transplantation from patients with MDD, or mouse models of depression replicate depressive-like behaviors in recipient germ-free mice. 13,14Besides, the vagus nerve system has been repeatedly identified as the most direct link between the brain and the gut microbiota. 15,16For one thing, the vagus nerve is responsible for regulating metabolic homeostasis and feeding behavior, such as gastrointestinal motility and secretion functions.For another, much of the intestineand microbiota-related processes may influence the activity of the brain through the vagus nerve. 17Corresponding, a previous study found that subdiaphragmatic vagotomy significantly blocked the development of depressive-like behaviors in mice induced by fecal microbiota transplantation (FMT) received from mice subjected to chronic social defeat stress, 18 and vagus nerve stimulation (VNS) is a noninvasive alternative treatment for MDD, which was approved by Food and Drug Administration. 19,20Nevertheless, the metabolites of gut microbiota also play causal roles in the development of MDD via the microbiota-gut-brain axis. 21,22For instance, fatty acids (FAs) are key building blocks of lipids and essential components of the central nervous system (CNS). 23Beyond energy metabolism, FAs and their metabolites play essential roles such as neuroprotective molecules and anti-inflammatory in MDD 24 and the link between MDD and FAs metabolism in the plasma/serum has been also recognized. 25,26portantly, FAs have shown benefits to the brain as part of the direct or indirect link between "gut-health" and "brain-health" because they are naturally fermented or regulated by the gut microbiota. 24evious studies found that gut microbiota can influence fatty acid production and metabolism, 27,28 and altered gut microbiota in depression were associated with disturbed peripheral and central lipids metabolism. 29,30Accordingly, fecal microbial transplants from depression rat model could dysregulate fatty acids metabolism in recipient rats. 31Taken together, the regulation of gut microbiota and fatty acids metabolism might have translational applications in the treatment of depression. 32,33anscranial magnetic stimulation (TMS) is noninvasive neuromodulation technique.It has been successfully used as an important alternative therapy for depression. 34TMS could improve depressive symptoms and prevent the relapse of depressive episodes effectively in depression patients who have not responded to a full course of antidepressants. 35,368][39] A recent clinical study found that deep TMS treatment was revealed to be effective in modulating gut microbiota composition and food cravings in subjects with obesity. 40Another preclinical study also found that 10 Hz low-intensity (13 mT) repetitive TMS (rTMS) treatment showed anti-inflammatory and protective effects on the gut microbiome in chronic restraint stress-treated rats. 41Although our recent study also found that rTMS with higher intensity (1.26 T) regulates brain lipid metabolism in both chronic unpredictable stress (CUS)-treated rats and cuprizone-treated mice, 42,43 little is known about the effect of rTMS on the metabolism of FAs as well as the influence of high-intensity rTMS on the composition of gut microbiota in depression.
Considering the above, the present study used chronic unpredictable mild stress (CUMS)-exposed mice, a well-established mice model of depression 44,45 to determine the influence of rTMS with high-frequency and high-intensity (15 Hz, 1.26 T) on the composition of medium-and long-chain fatty acids (MLCFAs) in the plasma, prefrontal cortex (PFC), and hippocampus (HPC) as well as the composition of gut microbiota.Furthermore, we also explored the relationship between changed gut microbiota, fatty acids, and depressive-like behaviors.

| Animals
Male C57BL/6 mice aged at 8 weeks (18-22 g) used in this study were purchased from the Fourth Military Medical University Animal Center (Xi'an, China) and maintained at 20-25°C in a 12 h alternating light and dark cycle (lights on from 8:00 a.m. to 8:00 p.m.) with food and water available ad libitum.All experiments were approved by the Animal Use and Protection Committee of the Fourth Military Medical University and conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

| Experimental design
As shown in Figure 1A, mice were randomly distributed to the following four groups after 7 days of acclimatization: Sham (n = 18), rTMS (n = 14), CUMS (n = 12), and CUMS + rTMS (n = 12).Mice in the Sham and rTMS group were maintained in home cages for 4 weeks before subjected to sham rTMS or real rTMS (15 Hz, 1.26 Tesla) treatment for 7 days.Mice in the CUMS and CUMS + rTMS groups were subjected to CUMS for 28 days (day 8 to day35) and then received sham rTMS or real rTMS for 7 days (day 36 to day 42).The behavioral tests were conducted 24 h after the final intervention (The training phase for sucrose preference test was initiated after the 6th rTMS intervention (day 41), testing was performed on day 43).Feces were collected and stored in liquid nitrogen prior to behavioral testing.The plasma and brain tissue were collected and frozen in liquid nitrogen 24 h after the last behavioral test.

| CUMS administration
The CUMS was performed as previously described. 46Briefly, mice were singly housed and subjected to various and repeated unpre- Types of stimulation were randomly selected and applied daily, and each of the stressors was equally administered 3 or 4 times during the molding process.

| rTMS treatment
rTMS was delivered at 15 Hz with a 3 s inter-train interval and intensity of 1.26 Tesla using a circular coil (diameter, 23 mm; custom-made YIRD, China) similar to our previous work. 47The daily stimulation consisted of 100 trains of 10 pulses and the trains were administered daily for 7 days at 1000 pulses per day.In the real stimulation groups, the coil was placed over the skull's vertex with the handle paralleling to the line of the animal's vertebral column.In the sham stimulation groups, the coil was held 10 cm above the head to ensure that the animal felt the vibrations produced by coil without any brain stimulation.Mouse was held to restrict movement by hand during the stimulation.Therefore, to exclude putative effects of nonspecific stress, all animals were habituated to the rTMS artifact noise sham stimulation procedure for 4 min every day for 7 days.There were no notable seizures or any behavioral changes throughout the treatment period in either real or sham stimulation.

| Sucrose preference test
The sucrose preference test consisted of two sessions. 48During the training phase (Started on day 41), mice were allowed to consume water and 1% sucrose placed simultaneously freely for 24 h and then were deprived of water and food for 24 h.In the testing phase, mice were allowed to consume water and 1% sucrose, which were placed in pre-weighed bottles freely for 2 h.The sucrose preference rate (%) = sucrose consumption/(sucrose consumption + water consumption).

| Open-field test (OFT)
The open-field test was performed according to a previous study. 49iefly, animals were placed in the center of an open-field box  S1.
(40 cm × 40 cm × 40 cm), and activity was recorded for a period of 5 min, using the open-field activity system (Open-field Ji Liang Co. Ltd., Shanghai, China) and activity software (Top Scan, Clever Sys Inc., USA).
The floor of the open field box was divided equally into 36 squares, including 16 central and 20 peripheral squares.The time spent and distance traveled in the central zone were recorded and measured.

| Tail suspension test (TST)
Tail suspension test was performed following a previous study. 50ce were hung from a horizontal bar 60 cm above the floor by its tail for 6 minutes.Total immobility time was recorded and calculated without the first 2 minutes of the test session via automated tracking software (Freeze Scan, Clever Sys, Inc.).Criteria Immobility was set as being lack of skeletal movement for at least 1 s and mouse that crawled back up its tail was removed from the analysis.

| Blood and tissue collection
Mice were anesthetized with 2,2, 2-tribromoethanol 24 h after the behavioral testing, and their blood and brain tissues were collected.
Blood was collected from the mice eye pits into a 1.5 mL centrifuge tube pretreated with 20 μL of 100 U heparin sodium (Changzhou Qianhong Biopharma Co., Ltd.China), and then centrifuged at 1500 g for 15 min at 4°C.Then the plasma (supernatant) was collected and stored in liquid nitrogen prior to use.Subsequently, mice brains were removed and rinsed in ice-cold phosphate buffered saline.The mPFC was isolated in a brain tank (68,713; RWD, Shenzhen, China), and the hippocampus was isolated under an anatomical microscope.
All tissues were weighed and cut into pieces on ice and then were frozen and stored in liquid nitrogen.Finally, 8 samples of each group were used for fatty acid detection by Gas Chromatography-Mass Spectrometry (GC-MS).

| Total fatty acids extraction
For brain tissue, samples were mixed with 2 mL of acidified methanol and incubated at 80°C for 30 min to achieve fatty-acid methyl esterification.For plasm, 50 μL of each sample was mixed with 1 mL chloroform methanol (2:1 v/v) in a 2 mL glass centrifuge tube and ultrasonicated for 30 min.To achieve fatty-acid esterification, 2 mL of 1% sulfuric acid in methanol was added to the supernatant, and then the mixture was incubated in water bath at 80°C for 30 min.
Subsequently, both brain and plasm samples were extracted with 1 mL of hexane and washed with 5 mL of ddH 2 O. Next, 25 μL of internal standard (NU-CHEK-PREP methyl esterified fatty acids mixture added with methyl Salicylate) was mixed with 500 μL supernatant of the extract before GS-MS. 52

| Medium-and long-chain fatty acid measurements
Each 1 μL of extracted samples from brain tissue and plasm was analyzed by GC-MS in a single ion monitoring mode.Briefly, samples were separated and detected by Agilent 7890A/5975C system with Agilent DB-WAX column (Agilent, 0.25 μm, 0.25 mm × 30 m).
The initial column temperature remained at 50°C for 3 min, and then increased to 220°C at 10°C/min, and remained at 220°C for 20 min.Helium was used as the carrier gas, and mass spectrometry assay was performed using the Electron impact ion source (EI).
The temperatures of the injection port, transmission line, and ion source were 280, 250, and 230°C, respectively.The stability and repeatability of the system were tested and evaluated through QC (quality-control) sample in SIM (Selected ion Monitor) scanning mode throughout the experiment.Finally, MSD ChemStation was applied to analyze the mass data to determine the concentration of each compound. 52

| Statistical analysis
Behavioral and fatty acids data are presented as mean ± SD.Statistical analyses were conducted using GraphPad v.8.0 (GraphPad software) and R software package (http://www.R-proje ct.org/).The normal distribution of continuous data was detected by Shapiro-Wilk test.Data that did not satisfy normal distribution or homogeneity of variance were analyzed by nonparametric test (Kruskal-Wallis) and others were subjected to two-way analysis of variance (ANOVA) followed by a Bonferroni post-hoc test for pairwise comparisons.Pearson correlation was applied to analysis of the associations between fatty acids levels and behaviors, and the Spearman's rank correlation coefficient was applied to analysis of the correlations between behaviors and the top abundance 50 species at the genus level.All tests for significance were two-tailed, and p < 0.05 was considered significant.

| rTMS treatment ameliorates depressive-like behaviors in CUMS mice
As shown in Figure 1 and Table S1, no identified differences were observed in the total distance (F = 0.189, p = 0.666; and F = 0.010, p = 0.921; respectively; data is not shown in the figure) or center distance (F = 2.788, p = 0.101; and F = 3.797, p = 0.057; respectively; Figure 1C) traveled in OFT either for the CUMS or for the rTMS treatment factors.Significant differences were observed in time spent in center (F = 5.083, p = 0.028; Figure 1D) of OFT and the sucrose preferences rate in the SPT (F = 6.189, p = 0.016; Figure 1E) for the CUMS factor, as well as immobility times displayed in the TST (Figure 1F) for both the CUMS factor (F = 5.879, p = 0.019) and rTMS factor (F = 5.982, p = 0.018).We also observed significant differences for the interaction factor in the percentage of center distance (F = 6.782, p = 0.012), time spent in center (F = 6.595, p = 0.013) as well as sucrose preference rate (F = 7.105, p = 0.010).Post hoc comparisons further showed that CUMS reduced the distance traveled in the central area and time spent in center of OFT and sucrose preferences rate in the SPT, but increased the immobility times in the TST significantly (CUMS vs. Sham, p < 0.05).
rTMS treatment ameliorated the depressive-like behavior of CUMS mice effectively, as evidenced by increasing central distance in the OFT, sucrose preferences rate in the SPT, and decreasing immobility time in the TST observed in rTMS + CUMS group (CUMS vs. CUMS + rTMS, p < 0.05), which are consistent with our previous study. 47

| rTMS changes the composition of gut microbiota in mice
We obtained a total of 1,117,341,014 bases and 2,582,388 highquality 16S rRNA gene sequences from 56 fecal samples.After downstream analysis, 924,726 sequences and 724 species-level OTUs were obtained from the Sham group, 597,869 sequences and 622 OTUs from the rTMS group, 522,695 sequences and 585 species-level OTUs from the CUMS group, and 537,098 sequences and 680 OTUs from the CUMS + rTMS group (Figure 2A).There was a significant difference in OTUs numbers between each group We validated differences in taxonomic composition between the four groups through linear discriminant analysis (LDA) and effect size (LEfSe) analysis (Figure 3A).We found that the relative abun-

| Microbiota signatures specific for CUMS and rTMS treatment
In order to observe the effect of CUMS on the composition of  Moreover, the immobility time in TST was negatively correlated with the abundance of 6 bacteria, but positively correlated with the abundance of Dubosiella and Ileibacterium.(p < 0.05) (Figure 4A and Table S2).

| Predicting the gene function of gut microbiota using KEGG
To further understand the functional information related to changes in the gut microbiota, functional prediction of important bacterial taxa among the four groups was achieved using PICRUSt2.A total of 245 pathways of level 3 were found to differ in functional abundance among the four groups through the Kyoto encyclopedia of genes and genomes (KEGG) database (p < 0.05), including most lipid metabolism, amino acid metabolism and Carbohydrate metabolism.
Interestingly, significant differences were observed in the functional abundance levels of fatty acid metabolism (F = 5.217, p = 0.026) and fatty acid biosynthesis (F = 5.922, p = 0.018) for the CUMS factor (Table S3).We also observed significant differences for the interaction factor in the functional abundance levels of carbon metab-  S3).
In addition, we observed significant differences for the rTMS treatment factor in the concentrations of the following: 4 PUFAs such as C22:5N6, C20:5N3 and C22:2N6, and 2 MUFAs (C14:1N5 and C18:1TN9), and 3 SFAs (C10:0, C20:0 and C22:0).Consistently, significant differences were also observed for the interact factor in the concentrations of the following: total MLCFAs, total PUFAs F I G U R E 4 Correlation between depressive-like behaviors and the differential gut microbiota and differences in gene function abundance between the four groups (sham group: n = 18; rTMS group: n = 14; CUMS group: n = 12; CUMS + rTMS group: n = 12).(A)The correlation between depressive-like behaviors and the differential gut microbiota was analyzed at the genus level by Spearman's rank correlation coefficients, *p < 0.05; **p < 0.01, detailed statistical information is provided in Table S2.(B-G) the relative abundance of 6 KEGG pathways, (B) carbon metabolism, (C) citrate cycle (TCA cycle), (D) fatty acid metabolism, (E) fatty acid biosynthesis, (F) glycerophospholipid metabolism, and (G) phenylalanine, tyrosine, and tryptophan biosynthesis.The circle represents one value from individual mice.KEGG, Kyoto encyclopedia of genes and genomes; rTMS, repetitive transcranial magnetic stimulation; CUMS, chronic unpredicted mild stress; SP, sucrose preference.Two-way analysis of variance (ANOVA) followed by a Bonferroni post-hoc test for pairwise comparisons was used in date from B-G, detailed statistical information is provided in Table S3.S4; Pearson correlation was used in (E), *p < 0.05; **p < 0.01, detailed statistical information is provided in Table S5.

| Effect of rTMS on MLCFAs in the hippocampus and prefrontal cortex
To investigate the effect of rTMS on brain fatty acid metabolism, medium-and long-chain fatty acid changes in the hippocampus (Figure 6) and prefrontal cortex (PFC) (Figure 7) were also detected.
In terms of PFC (Figure 7A-D and Table S8), there are significant differences for the CUMS factor in concentrations of following: total PUFAs and 7 PUFAs such as C22:6N3, C22:4N6 and C20:2N6, total SFAs and 3 SFAs (C6:0, C17:0 and C21:0).We observed significant differences for the rTMS treatment factor in the concentrations of C20:5N3 and C12:0.Moreover, we also observed significant differences for interaction factor in the concentrations of following: total The immobility time in TST was negatively correlated with levels of C20:3N6 and C20:5N3 in the PFC (Figure 7E and Table S9).These results suggested that CUMS remarkably reduces the concentration of MUFAs and PUFAs in the hippocampus and PUFAs in the PFC, which was also partially restored by rTMS treatment.

| DISCUSS ION
In the current study, we performed integrative analysis to assess the effect of high-frequency rTMS treatment on the composition of gut microbiota and medium-and long-chain fatty acids  S6.Pearson correlation was used in (E), *p < 0.05; **p < 0.01, detailed statistical information is provided in Table S7.
Bidirectional gut-brain axis communication has been widely accepted. 53,54Brain can modulate the gastrointestinal tract and enteric nervous system via the parasympathetic and sympathetic branches of the autonomic nervous system and the HPA axis directly, 55 and can influence the enteric microbiota indirectly by altering its microenvironment. 56As a result, neuromodulation technology could affect the composition and function of gut microbiota. 57Conversely, gut microbiota can regulate many aspects of host physiology thus influencing brain development and function. 58Although the involvement of gut microbiota in the pathogenesis of depression has been largely reported, including the relationship between the characteristic gut microbiota, diversity of gut microbiota and the severity of depressive symptoms in patients with depression as well as the effect of antidepressants on the composition of gut microbiota, 59,60 the results of characteristic microbiota and diversity identified in different studies are not completely consistent.For instance, there is no consensus regarding whether the gut microbiota richness and diversity changes in depression models.A recent work found that the Shannon and Simpson indices showed no significant difference whereas the Chao1 and ACE indices was decreased, 61 while another work showed that the diversity of the fecal microbiome but not the richness was reduced in CUMS treated rats. 62Similarly, the community diversity and richness estimators were not changed in CUMS-treated mice in a previous study, 63 but these estimators were reduced in another study. 64Inconsistent with the previous study, 41 the present study showed that the alpha diversity in the CUMS model was significantly reduced and rTMS treatment reversed these changes effectively (including Ace, Chao1, and Shannon).In addition, the beta diversity analysis showed a different community composition of the gut microbiome between the CUMS-and rTMS-treated group, indicating the potential effect of rTMS on the diversity of intestinal flora.
Moreover, the anti-anxiety effect of rTMS has been extensively reported, 72 and the add-on rTMS treatment can improve anxiety symptoms in patients with depression comorbid with anxiety. 73On the other hand, both patients with anxiety disorder and anxiety-like animal models exhibit gut microbiota imbalance. 74,75Together, intestinal dysbacteriosis might be a common pathogenesis both in anxiety and depression.However, the co-pathogenic bacteria of anxiety and depression, as well as the regulatory effect of TMS on the microbiota of anxiety animal models, still need further exploration.
Accumulating evidence indicated that depression is linked to abnormal lipid metabolism, especially fatty acids and their metabolites, which are the key messengers in the bidirectional communication between the gut microbiota and the brain. 10,76In consistent, the present study found that 245 Kyoto encyclopedia of genes and genomes (KEGG) pathways varied significantly in relative abundance among the groups and the relative abundance of energy metabolism (carbon metabolism and TCA cycle), phenylalanine, tyrosine and tryptophan biosynthesis, glycerophospholipid metabolism, and fatty acid metabolism and biosynthesis pathways were decreased in CUMS-treated mice, all of which were normalized after rTMS treatment.It suggested that the regulation of metabolic function of bacteria may be related to the antidepressant effect of rTMS.
Fatty acids (FAs) play an important role in regulating energy homeostasis, neurotransmission and signaling pathways, and ultimately affects emotional behavior. 24,77FAs are classified as SFAs, MUFAs, and PUFAs.Previous studies found that intake of SFAs induces depressive-like behavior in rodents, 78,79 whereas intake of MUFAs might be benefit to brain function, such as the facilitation of neurotransmitter signal transduction. 80,81In the present study, there is no significant difference in the total SFAs in the plasma, hippocampus, and PFC between each group.Meanwhile, total MUFAs were decreased in plasma and hippocampus in CUMS group, which could not be normalized after rTMS treatment.It suggests that SFAs may not be closely related to CUMS-induced depressive-like behaviors.
Importantly, levels of oleic acid (18:1N9) were decreased after CUMS both in the plasma and hippocampus, and rTMS treatment only restored it in the hippocampus.Due to oleic acid can increase neural stem cell mitotic activity and drive hippocampal neurogenesis in mice 82 and prolonged intake of oleic acid enriched diet in human reduces the risk of depression, 83,84 the increased oleic acid in the hippocampus might be involved in the antidepressant effect of rTMS.
Nevertheless, more studies focus on the involvement of PUFAs  86 Previous studies reported that decreased PUFAs in patients with depression 87 and lower N3 PUFAs levels in baseline showed reduced response to standard antidepressants. 88Meanwhile, nutritional intervention with EPA or DHA provides beneficial anti-inflammatory and antidepressant effects both in clinical and basic research. 89,90Here, we found that levels of ALA and LA were decreased in the plasma and hippocampus, and the levels of EPA and AA were decreased in the plasma, hippocampus, and PFC, whereas level of DHA was decreased only in hippocampus and PFC in CUMS-induced mice.Importantly, rTMS treatment restored levels of total PUFAs, including EPA and DHA in the hippocampus and PFC.Meanwhile, rTMS treatment also increased EPA in the plasma of CUMS-treated mice effectively, suggesting that these N3 PUFAs were involved in the protective effect of rTMS.Furthermore, AA, a precursor for two main endocannabinoids-anandamide (AEA) and 2-arachidonoylglycerol (2-AG), plays a key role in modulating synaptic plasticity and neurotransmitter release in the brain. 91,92Previous work found that serum AA was decreased in childhood and adolescent patients with depression, 93 and a dysfunctional endocannabinoid system (ECS) in the CNS has also been increasingly implicated in the pathophysiology of depression. 94The present study found that rTMS increased level of AA in PFC of CUMS-induced mice, indicating that rTMS might play a neuroprotective and antidepressant role by regulating AA metabolism to enhance endocannabinoid signaling.Moreover, our results also found that rTMS treatment reversed the level of eicosatrienoic acid (20:3N3) in the plasma and hippocampus of CUMS-induced mice, which was negatively associated with the severity of depression.As eicosatrienoic acid is produced during arachidonic acid metabolism and plays a role in neuroprotective effects on the CNS, 95 this PUFA may also involve in the antidepressant effects of rTMS, and more details remain to be determined.It cannot be ignored that the soluble epoxide hydrolase (sEH) is a key enzyme in the metabolism of PUFA and plays a key role in the inflammation that is involved in depression. 96,97Meanwhile, epoxy fatty acids such as epoxyeicosatrienoic acids (EETs) and epoxyeicosapentaenoic acids (EDPs) have been found to exert neuroprotective effects through potent anti-inflammatory actions. 98Unfortunately, the present study did not observe the content of sEH and epoxy fatty acids, which may further explain the antidepressant effect of TMS.
On the contrary, recent studies also found N3 supplementation is not enough to make recommendations in preventing depression symptoms 99 and even nonclinically beneficial effect of N3 on depressive symptomology when compared to placebo. 100 On the other hand, AA is also a substrate for the synthesis of the inflammatory mediators, such as prostaglandins (PGs) and leukotrienes (LTs).PGE2 mediates depression-like behavior induced by repeated social defeat stress and PGD2 has been implicated in depression-like behaviors induced by chronic stress. 101Moreover, a previous clinical study also found depressive symptoms were negatively correlated with serum AA levels. 102Furthermore, multiple studies suggested the measurement of the N6/N3 ratio rather than the content of N6 or N3 might be a useful indication for depressive symptoms. 103,104This disparity might be related to the high individual variability in FAs composition together with the presence of confounders.The ability to control for potential effects of environment, diet, and lifestyle on the FAs and gut microbiome might be an important advantage of performing animal studies.However, little is known about this controversial result in animal research, which needs to be verified in future studies.
Finally, there are some limitations that should be noted.First, the underlying mechanisms of rTMS alter intestinal flora and fatty acid metabolism remains unclear.We speculate that the regulation of the brain-gut axis may be one of the underlying mechanisms.Due to the subdiaphragmatic vagus nerve plays a crucial role in the crosstalk between the brain and gut microbiota, 105,106 and subdiaphragmatic vagotomy (SDV) blocks depression-related behaviors in the antibiotic cocktail-treated mice after ingestion of specific microbe. 18,107It is necessary to observe the influence of SDV on the antidepressant effects as well as the regulation of gut microbiome and fatty acid metabolism of rTMS.Meanwhile, the effects of different patterns of rTMS on gut microbiota and fatty acids metabolism still need to clarify.Although the current study has confirmed the correlation analysis between changes either in gut microbiota or in fatty acids content and behavioral outcomes, the correlation between the gut microflora and fatty acids content in brain still needs to be investigated.

F
I G U R E 1 rTMS ameliorates depressive-like behaviors in CUMS treated mice.(A) The experimental design.After 1 week of acclimatization, mice were subjected to CUMS or maintained in their home cages for 4 weeks, then rTMS or Sham stimulation was administered for 7 days.Behavioral testing and fecal collection were performed after the last rTMS intervention (the training phase for sucrose preference test was initiated on day 41), then the peripheral blood and brain tissues were collected for medium-and long-chain fatty acids measurements.(B) Representative real-time movement traces in the OFT for each group.(C) Percentage of distance traveled in central zone in the OFT.(D) Quantification of the time spent in center of the OFT.(E) The percentage of sucrose consumption in the SPT.(F) Immobility time measured in the TST.The dot represents one value from individual mice (sham group: n = 18; rTMS group: n = 14; CUMS group: n = 12; CUMS + rTMS group: n = 12); rTMS, repetitive transcranial magnetic stimulation; CUMS, chronic unpredicted mild stress; SPT, sucrose preference test; OFT, the open field test; TST, tail suspension test.Data was analyzed with two-way analysis of variance (ANOVA) followed by a Bonferroni post-hoc test for pairwise comparisons and detailed statistical information is provided in Table

F I G U R E 2
Differential gut microbial characteristics in mice of each group (sham group: n = 18; rTMS group: n = 14; CUMS group: n = 12; CUMS + rTMS group: n = 12).(A) The number of common and unique OUTs among the four groups is displayed by the Venn diagram and histogram.(B-F) Alpha diversity analysis index, including the (B) observed species (H = 40.268,p < 0.001), (C) Ace index (H = 34.421,p < 0.001), (D) chao1 index (H = 38.750,p < 0.001), (E) Shannon index (H = 27.362,p < 0.001), and (F) Simpson index (H = 18.144, p < 0.001).(G-I) PCoA plots of bacterial beta-diversity on the basis of (G) Bray curtis, (H) Unweighted UniFrac distance and (I) Weighted UniFrac distance.The circle represents one value from individual mice (B-F).rTMS, repetitive transcranial magnetic stimulation; CUMS, chronic unpredicted mild stress; OUT, Operational taxonomic unit; PCoA, Principal coordinates analysis.Nonparametric test (Kruskal-Wallis) was used in B-F.Proteobacteria, Deferribacterota, and Cyanobacteria were enriched in the CUMS + rTMS group.At the family level, Lactobacillaceae, Erysipelotrichaceae, and Bifidobacteriaceae were enriched in the CUMS group, while Lachnospiraceae, Oscillospiraceae, Desulfovibrionaceae, Ruminococcaceae, Tannerellaceae, Sutterellaceae, Deferribacteraceae, Bacteroidaceae, norank_o_Gastranaerophilales, norank_o_Clostridia_ vadinBB60_group, and Rs-E47_termite_group were consistently higher in the CUMS + rTMS group.Finally, we identified 7 genera that were abundant in the CUMS group and 16 genera were abundant in the CUMS + rTMS group (Figure 3C).Notably, 14 genera altered in CUMS-treated mice (9 increased and 5 decreased) were restored after rTMS treatment.The correlation between depressive-like behaviors and the differential gut microbiota at the genus level showed that the percentage of distance traveled in the center area was negatively correlated with the abundance of Enterorhabdus and Lactobacillus, but positively correlated with the abundance of 16 bacteria such as unclassified_f_ prevotellaceae, Colidextribacter and Lachnospiraceae_NK4A136_group (p < 0.05).The percentage of sucrose preference rate was negatively correlated with the abundance of 7 bacteria such as Lactobacillus, Enterorhabdus, and Dubosiella, but positively correlated with the abundance of 21 bacteria such as Lachnospiraceae_NK4136_group, Eubacterium_xylanophilum_group, and Lachnoclostridium (p < 0.05).
olism, citrate cycle (TCA cycle), fatty acid metabolism, fatty acid biosynthesis, glycerophospholipid metabolism and phenylalanine, tyrosine and tryptophan biosynthesis.Post hoc comparisons further revealed that functional abundance levels of carbon metabolism, citrate cycle (TCA cycle), fatty acid metabolism, fatty acid biosynthesis, glycerophospholipid metabolism and phenylalanine, tyrosine F I G U R E 3 Differential taxonomic composition of gut microbiota in mice of each group (sham group: n = 18; rTMS group: n = 14; CUMS group: n = 12; CUMS + rTMS group: n = 12).(A) LDA score showed significant bacterial differences between these four groups based on the LEfSe and LDA analyses.(B) LDA score showed significant bacterial differences between the Sham and CUMS group.(C) LDA score showed significant bacterial differences between the CUMS and CUMS + rTMS groups.Only taxa with an LDA significance threshold >3.5 was presented.rTMS, repetitive transcranial magnetic stimulation; CUMS, chronic unpredicted mild stress; LDA, Linear discriminant analysis; LEfSe, Linear discriminant analysis effect size. and tryptophan biosynthesis were significantly reduced in CUMStreated mice (CUMS vs. Sham, p < 0.01), which were effectively normalized after rTMS treatment (CUMS + rTMS vs. CUMS, p < 0.01) (Figure 4B-G and Table

in depression. 85
According to the position of the first double bond, PUFAs are distinguished into N3 and N6 PUFAs.The two essential PUFAs linoleic acid (18:2N6, LA) and alpha-linolenic acid (18:3N3, ALA), which are mainly provided by diet and act as precursors of arachidonic acid (20:4N6, AA), eicosapentaenoic acid (20:5N3, EPA), and docosahexaenoic acid (22:6N3, DHA), regulate both the structure and the function of cells in the brain.