The gut microbiota and PMD

The intestinal microecosystem is composed of intestinal epithelial cells, the gut microbiota and the intestinal mucosal immune system. Among these, the gut microbiota plays a key role in protecting the intestinal mucosal barrier. The human intestine contains trillions of microorganisms, including bacteria, viruses, archaea and fungi, collectively forming a microbial genome approximately 100–150 times larger than the human genome. The intestinal microbiome plays an important physiological role in food digestion, metabolism, intestinal barrier maintenance and immune system regulation (Ref. Reference Yang, Wang, Chen, Guo and Dong31). The gut microbiota is mainly composed of Firmicutes, Bacteroidetes, Actinobacteria and Proteobacteria, with Firmicutes and Bacteroidetes accounting for 90% of intestinal microbes (Refs Reference Yang, Wang, Chen, Guo and Dong31, Reference Turnbaugh, Ley, Mahowald, Magrini, Mardis and Gordon32).

At the phylum level, patients with depression usually show decreased abundance of Firmicutes and increased abundance of Bacteroidetes, Proteobacteria and Actinobacteria. The relative ratio of Firmicutes to Bacteroidetes (F/B) has been used as a measure of gut microbiota health. Clinical studies have shown that the F/B ratio decreases in MDD patients (Refs Reference Liu, Wang, Chen, Zhang, Zhang and Xie33, Reference Liu, Rowan-Nash, Sheehan, Walsh, Sanzari, Korry and Belenky34). A cross-sectional study found significantly reduced Firmicutes abundance in MDD patients, which may lead to diminished SCFA production, potentially contributing to low-level inflammation in depression (Ref. Reference Huang, Shi, Li, Shen, Shi, Wang, Li, Yuan, Wang, Zhang, Zhao, Zhang, Kang and Liang35). However, this study had limitations, including a small sample size and insufficient assessment of dietary factors (Ref. Reference Huang, Shi, Li, Shen, Shi, Wang, Li, Yuan, Wang, Zhang, Zhao, Zhang, Kang and Liang35). Zhao et al. observed that postmenopausal women exhibited lower F/B ratios and reduced relative abundance of Lachnospira and Roseburia compared to premenopausal women (Ref. Reference Santos-Marcos, Rangel-Zuñiga, Jimenez-Lucena, Quintana-Navarro, Garcia-Carpintero, Malagon, Landa, Tena-Sempere, Perez-Martinez, Lopez-Miranda, Perez-Jimenez and Camargo36). Similarly, a Chinese metagenome-wide association study revealed decreased Firmicutes and Roseburia spp., alongside increased Bacteroidetes and toluene-producing Tolumonas, in postmenopausal women (Ref. Reference Zhao, Chen, Li, Sun, Qin and Wang37). However, some clinical studies have reported that the F/B ratio was increased in patients with MDD (Refs Reference Lai, Deng, Xu, Zhao, Xu, Liu, Guo, Wang, He, Ye, Yang, Liu, Zhang, Wang, Li, Yang, Xie and Rong38, Reference Rong, Xie, Zhao, Lai, Wang, Xu, Liu, Guo, Xu, Deng, Yang, Xiao, Zhang, He, Wang and Liu39, Reference Lin, Ding, Feng, Yin, Zhang, Qi, Lv, Guo, Dong, Zhu and Li40, Reference Arbabi41). A Korean animal study also observed that ovariectomized (OVX) rats had increased F/B ratios and Lachnospiraceae and Ruminococcaceae family abundance while Muribaculaceae family abundance was decreased (Ref. Reference Lim, Song, Kim, Jung, Lee, Shin, Nam and Kim42). In addition, a meta-analysis showed that there was no difference in the abundance of Bacteroidetes and Firmicutes in patients with depression (Ref. Reference Gao, Wang, Liu, Tu, Zhang, Zhang, Sun and Zhang43). Meanwhile, in the subgroup meta-analysis, it was found that in patients with depression who had not taken psychotropic medications, the abundance of Firmicutes decreased while that of Bacteroidetes increased (Ref. Reference Gao, Wang, Liu, Tu, Zhang, Zhang, Sun and Zhang43). It was also found that dietary and regional differences could affect the composition of the gut microbiota (Ref. Reference Gao, Wang, Liu, Tu, Zhang, Zhang, Sun and Zhang43). Therefore, these differences may be attributed to several factors, such as clinical and demographic characteristics, sample differences, dietary patterns, regional differences, as well as medication use and other influencing factors (Table 2). In future research, firstly, strict experimental controls should be implemented, e.g., controlling diet (e.g., standardized dietary records), regional contexts (multicentre studies) and medication use (e.g., the exclusion of recent antibiotic/hormone therapy). Secondly, to verify mechanisms, animal models could be used to simulate different diet/hormone environments to clarify causal relationships between the F/B ratio and depressive phenotypes (e.g., germ-free mice transplanted with specific bacteria). Finally, standardized reporting should also be adopted; population characteristics and experimental methods should be better annotated, and confounding factors in the literature could be explored to help reduce data/results heterogeneity. Lim et al. reported that the relative abundance of beneficial intestinal bacteria (Lactobacillus, Clostridium and Eubacterium) in OVX rats was significantly reduced (Ref. Reference Lim, Song, Kim, Jung, Lee, Shin, Nam and Kim42). Li et al. also showed that the relative abundance of toxin-related Cyanobacteria in OVX rat intestines was increased when compared with a control group (Ref. Reference Li, Liu, Wang, Gong, Yao, Li, Gao and Wei44). Huang et al. observed that PMD model mice had a reduced abundance of beneficial intestinal bacteria, such as Lactobacillus, Alloprevotella, Akkermansia and Allobaculum, as well as an increased abundance of harmful bacteria, such as Muribaculaceae (Ref. Reference Huang, Liu, Xu, Hu, Wang, Shi, Wang, Wang, Lin, Li, Zhao, Jin, Wang, Chen and Wang45). Therefore, PMD occurrence may be related to decreased beneficial bacteria and increased harmful bacteria in the gut. Interestingly, two meta-analyses reported increased beneficial bacteria (Lactobacillus) abundance in patients with MDD (Refs Reference Liu, Wang, Zhang, Chen, Zhang, Wu, Zhao, Wang, Pu, Ji and Xie20, Reference Nikolova, Smith, Hall, Cleare, Stone and Young21), indicating differential roles for different species in this genus, and suggesting no clear boundaries between beneficial and pathogenic bacteria; therefore, a gut microbiota balance appears to be more important than regulating a single bacterial class. These studies suggest that PMD occurrence may be related to a disordered gut microbiota, but specific mechanisms remain unclear. Additionally, a cluster analysis indicated that intestinal microflora composition in patients with depression in the US and China differed, with some microbial changes unique to patients in China, such as increased Eggerthella and Acidaminococcus, and decreased Coprococcus, Fusicatenibacter and Prevotellaceae. Such gut microbiota differences suggest that countries or geographical regions need to develop microbiome databases that are tailored to specific patient characteristics to guide future gut microbiome investigations (Ref. Reference Liu, Wang, Zhang, Chen, Zhang, Wu, Zhao, Wang, Pu, Ji and Xie20). However, few studies have examined gut microbiota differences in patients with PMD across different countries.

Table 2. F/B ratio differences and potential influencing factors

Abbreviations: BMI: body mass index; MDD: major depressive disorder; SSRIs: selective serotonin reuptake inhibitors.

When compared to healthy individuals, a gut microbiota disorder occurs in patients with depression (Figure 2). Although data from recent studies are not very consistent, one common feature is increased proinflammatory and decreased anti-inflammatory bacteria (Refs Reference Liu, Wang, Zhang, Chen, Zhang, Wu, Zhao, Wang, Pu, Ji and Xie20, Reference Nikolova, Smith, Hall, Cleare, Stone and Young21, Reference Simpson, Diaz-Arteche, Eliby, Schwartz, Simmons and Cowan22). Among these, proinflammatory bacteria include Alistipes, Eggerthella, Flavonifractor, etc., while anti-inflammatory bacteria include Bifidobacterium spp., Coprococcus, Eucbacterium, Eubacterium rectale, Fecalibacterium, Fecalibacterium prausnitzii, Lactobacillus spp., Prevotella and Roseburia, amongst others (Ref. Reference Riehl19). A disturbed gut microbiota can lead to microglial activation and cause neuroinflammation in the central nervous system (CNS). Microglias are resident immune cells in the brain and have key roles in different neurodevelopmental processes for normal brain maturation and function, such as neurogenesis, synapse shaping and defences against infection. Resting microglia maintain homeostasis via regulated cytokines, which are secreted by astrocytes and other cells. Critically, microglia and astrocyte structure and function changes have been implicated in depression (Refs Reference Zheng, Zhang, Zhang and Duan46, Reference Li, Fang, Zhang, Song, Zhang, Ding, Yao, Chen, Sun, Ding, Wang, Lu, Wu and Hu47). Studies have shown that germ-free (GF) mice exhibited global microglial defects, with altered cell proportions (Refs Reference Erny, Hrabě de Angelis, Jaitin, Wieghofer, Staszewski, David, Keren-Shaul, Mahlakoiv, Jakobshagen, Buch, Schwierzeck, Utermöhlen, Chun, Garrett, McCoy, Diefenbach, Staeheli, Stecher, Amit and Prinz48, Reference Thion, Low, Silvin, Chen, Grisel, Schulte-Schrepping, Blecher, Ulas, Squarzoni, Hoeffel, Coulpier, Siopi, David, Scholz, Shihui, Lum, Amoyo, Larbi, Poidinger, Buttgereit, Lledo, Greter, Chan, Amit, Beyer, Schultze, Schlitzer, Pettersson, Ginhoux and Garel49) and immature phenotypes, leading to impaired innate immune responses (Ref. Reference Erny, Hrabě de Angelis, Jaitin, Wieghofer, Staszewski, David, Keren-Shaul, Mahlakoiv, Jakobshagen, Buch, Schwierzeck, Utermöhlen, Chun, Garrett, McCoy, Diefenbach, Staeheli, Stecher, Amit and Prinz48). Wei et al. observed that butyrate (a gut microbiota metabolite) supplementation improved chronic alcoholic CNS injury by inhibiting microglia-mediated neuroinflammation via GPR109A/peroxisome proliferator-activated receptor gamma (PPAR-γ)/toll-like receptor (TLR) 4-nuclear factor-κB (NF-κB) signalling (Ref. Reference Wei, Yu, Zhang, Ren, Guo, Wang, Chen, Li, Zhang, Wang and Liu50). Furthermore, GF mouse microglia engraftment of change in the form of time and gender specificity. In adult mice, Thion et al. observed that GF female microglia exhibited dysregulated genes related to cell morphogenesis, transcriptional regulation, adaptive immune responses, cell migration and chemotaxis, whereas GF male microglia showed no major changes (Ref. Reference Thion, Low, Silvin, Chen, Grisel, Schulte-Schrepping, Blecher, Ulas, Squarzoni, Hoeffel, Coulpier, Siopi, David, Scholz, Shihui, Lum, Amoyo, Larbi, Poidinger, Buttgereit, Lledo, Greter, Chan, Amit, Beyer, Schultze, Schlitzer, Pettersson, Ginhoux and Garel49). Thus, PMD caused by a disordered gut microbiota may be related to inflammation.

Figure 2. Associations between gut microbiota dysbiosis and pathological changes in the central nervous system (CNS) during depression. An imbalanced gut microbiota increases abundance in the Bacteroidetes phyla, the Enterobacteriaceae family, and the Eggerthella genus, and also decreases abundance of the Firmicutes phyla, and Prevotellaceae and Akkermansiaceae families. A disordered gut microbiota also leads to microglial activation and proinflammatory cytokine secretion (interleukin (IL)-1β, IL-6, etc.), which promotes astrocyte activation, causes neuronal damage (Refs Reference Bassett, Subramaniyam, Fan, Varney, Pan, Carneiro and Chung51,Reference Yang, Liu, Li, Ma, Huo, Ozathaley, Ren, Yuan, Ni, Li, Zhang and Liu52) and neuroinflammation in the CNS, and eventually depression.

The MGB axis

In recent years, many studies have reported bidirectional communication signal pathways between the gut microbiota and the brain, also known as the MGB axis (Refs Reference Simpson, Diaz-Arteche, Eliby, Schwartz, Simmons and Cowan22, Reference Socała, Doboszewska, Szopa, Serefko, Włodarczyk, Zielińska, Poleszak, Fichna and Wlaź69). This axis comprises the gut, the gut microbiota and the nervous system, collectively referred to as the ‘second brain’ in the human body. The communication network between the gut and the CNS is very complex, including an intestinal nervous system branch, the sympathetic and parasympathetic nerve vegetative nervous system, and neural immune and neuroendocrine signal pathways (Refs Reference Han, Gu, Li, Qian, Wang and Huang3,Reference Foster, Rinaman and Cryan70,Reference Cryan, O’Riordan, Cowan, Sandhu, Bastiaanssen, Boehme, Codagnone, Cussotto, Fulling, Golubeva, Guzzetta, Jaggar, Long-Smith, Lyte, Martin, Molinero-Perez, Moloney, Morelli, Morillas, O’Connor, Cruz-Pereira, Peterson, Rea, Ritz, Sherwin, Spichak, Teichman, van de Wouw, Ventura-Silva, Wallace-Fitzsimons, Hyland, Clarke and Dinan71). In terms of the gut microbiota’s impact on the brain, researchers have used different methods, including antibiotics, probiotic therapy, faecal microbiota transplantation (FMT), gastrointestinal infections and GF studies, to demonstrate its significant influence on mental health (Ref. Reference Gu, Wu, Liu, Dou, Jiang and Liang72). However, underlying mechanisms have not been fully elucidated. The gut microbiota is a key regulator of the MGB axis, which impacts on host physiological functions, emotional changes and social behaviours by regulating neural-, metabolic-, immune- and hormonal- (e.g., oestrogen) mediated mechanisms (Refs Reference Foster, Rinaman and Cryan70,Reference Dinan and Cryan73). In animal models and clinical studies, gut microbiota compositional changes were associated with brain function, e.g., FMT from patients with MDD induced depression-like behaviour in GF mice (Refs Reference Guida, Turco, Iannotta, de Gregorio, Palumbo, Sarnelli, Furiano, Napolitano, Boccella, Luongo, Mazzitelli, Usiello, de Filippis, Iannotti, Piscitelli, Ercolini, de Novellis, di Marzo, Cuomo and Maione74,Reference Marin, Goertz, Ren, Rich, Onengut-Gumuscu, Farber, Wu, Overall, Kipnis and Gaultier75), indicating that intestinal microecological dysbiosis may occur before depression. Sanada et al. reported that abundance of the Prevotellaceae family and Coprococcus and Fecalibacterium genera was lower in patients with MDD when compared with non-depressed controls (Ref. Reference Sanada, Nakajima, Kurokawa, Barceló-Soler, Ikuse, Hirata, Yoshizawa, Tomizawa, Salas-Valero, Noda, Mimura, Iwanami and Kishimoto57). Additionally, Sovijit et al. found that progesterone in a PMD mouse model increased Lactobacillus spp. in the intestines, which improved depression- and anxiety-like behaviours (Ref. Reference Sovijit, Sovijit, Pu, Usuda, Inoue, Watanabe, Yamaguchi and Nagaoka76). These studies suggest that the MGB axis may have important roles in PMD, with axis imbalance triggering PMD mechanisms, potentially related to inflammation. Next, we elaborate on and summarize MGB axis mechanisms that mediate PMD occurrence and development from an inflammation perspective.

Gut microbiota metabolite effects on PMD

SCFAs effects on PMD

SCFAs

SCFAs are one of the most important end-products of intestinal microbial metabolism and are produced by the glycolysis and fermentation of indigestible carbohydrates (Ref. Reference Cheng, Hu, Ju, Liu, Wang, Liu and Zhang77). Mainly found in the cecum and colon of animals and humans, SCFAs are widely distributed in enteroendocrine, immune and nerve cells (Refs Reference Tian, Xu, Pan, Zheng, Yang, Jia, Liu, Garstka, Gao and Yu78,Reference Kimura, Ichimura, Ohue-Kitano and Igarashi79). SCFAs are mainly generated by Bifidobacterium spp., Blautia hydrogentrophica, Prevotella spp., Streptococcus spp., Akkermansia muciniphilia, Bacteroides spp., Anaerostipes spp. and others (Ref. Reference Markowiak-Kopeć and Śliżewska80). Therefore, SCFAs content changes can indirectly reflect changes in intestinal microorganisms. SCFAs are mainly contain acetic acid (Reference Kaliannan, Robertson, Murphy, Stanton, Kang, Wang, Hao, Bhan and Kang60%), propionic acid (Reference Liu, Wang, Zhang, Chen, Zhang, Wu, Zhao, Wang, Pu, Ji and Xie20%), butyric acid (Reference Liu, Wang, Zhang, Chen, Zhang, Wu, Zhao, Wang, Pu, Ji and Xie20%), etc. (Ref. Reference Park, Pilla, Panta, Pandey, Sarawichitr, Suchodolski and Sohrabji81). Mechanistically, SCFAs mainly act by binding to G protein-coupled receptors (GPCRs), mainly GPR43, GPR41 and GPR109A. Receptors are distributed in most human tissues; GPR41 is mainly distributed in adipose tissue, and GPR43 is highly expressed in immune cells, both of which bind to acetate, butyrate and propionate, while GPR109A is mainly expressed in adipose tissue and immune cells, and is only be activated by butyrate (Ref. Reference Liu82).

SCFAs mechanisms in PMD

SCFAs are important mediators in the MGB axis, with crucial roles in the neurobiological mechanisms underlying depression. They directly or indirectly participate in vagal, immune, neuroendocrine and metabolic pathway regulation in the MGB axis (Ref. Reference Cheng, Hu, Ju, Liu, Wang, Liu and Zhang77) (Figure 3). GPCRs in intestinal epithelial and immune cells modulate inflammatory responses induced by SCFAs activation (Ref. Reference Yang, Wang, Chen, Guo and Dong31). Studies have indicated that SCFAs activate GPR43 (i.e., free fatty acid receptor 2, FFAR2) and GPR41 (FFAR3), promote peripheral macrophage, dendritic and T cell activation to exert immune effects, affect immune regulatory T (Treg) cell proliferation and development and inflammatory mediator recruitment, and increase anti-inflammatory factor expression. Thus, SCFAs reduce peripheral inflammatory factor damage to the brain and exert antidepressant roles (Refs Reference Carlessi, Borba, Zugno, Quevedo and Réus83,Reference Makris, Karianaki, Tsamis and Paschou84,Reference Ratajczak, Rył, Mizerski, Walczakiewicz, Sipak and Laszczyńska85). Butyrate reduces lipopolysaccharide (LPS)-induced NF-κB activation via GPR109A (Ref. Reference Chen, Ran, Li, Li, He, Huang, Fu, Liu and Wang86). Acetate binding to GPR43 activates NLRP3 inflammasome activation in the colon (Ref. Reference Macia, Tan, Vieira, Leach, Stanley, Luong, Maruya, Ian McKenzie, Hijikata, Wong, Binge, Thorburn, Chevalier, Ang, Marino, Robert, Offermanns, Teixeira, Moore, Flavell, Fagarasan and Mackay87). Additionally, intestinal 5-HT levels decrease upon SCFAs depletion, while acetate, butyrate and isobutyrate indirectly promote 5-HT production.

Figure 3. Possible SCFAs-mediated mechanisms in perimenopausal depression. In the vagus nerve pathway, SCFAs mainly bind to GPR41, which is highly expressed in the sympathetic nervous system, thus inhibiting the vagus nerve, which is widely distributed in the intestine. In the immune pathway, SCFAs inhibit HDACs, inhibit NF-κB and TLR4 activation, reduce proinflammatory factor levels (IL-6, TNF-α, etc.), and increase anti-inflammatory factor levels (IL-10, etc.), thus exerting antidepressant roles (Refs Reference Shao, Peng, Dong, Gao, Li, Li, Li and Liu98,Reference Liu, Li, Liu, Xiao, Suo, Xie, Yang and Wu99). In the neuroendocrine pathway, SCFAs affect the central nervous system by promoting neurotransmitter secretion, such as 5-HT and DA. In the metabolic pathway, SCFAs exert neuroactive effects by binding to three GPCRs (90). Thus, in a dysregulated gut microbiota, SCFAs content decreases, which inhibits the aforementioned effects and eventually leads to perimenopausal depression. SCFAs: short-chain fatty acids; NF-κB: nuclear transcription factor-κB; 5-HT: 5-hydroxytryptamine; HDACs: histone deacetylase; TNF-α: tumor necrosis factor alpha; GPCR: G protein-coupled receptor; TLR4: toll-like receptor 4.

Considerable evidence now suggests that SCFAs are implicated in depression development. In Polish women (aged approximately 50 years old) with depression, acetic acid and isocaproic acid levels were lower in depressed women compared to healthy control women, while acetic acid and propionic acid concentrations were negatively correlated with depression scores (Ref. Reference Skonieczna-Żydecka, Grochans, Maciejewska, Szkup, Schneider-Matyka, Jurczak, Łoniewski, Kaczmarczyk, Marlicz, Czerwińska-Rogowska, Pełka-Wysiecka, Dec and Stachowska88), thereby indicating that circulating SCFAs levels could be used as depression severity indicators. Animal studies reported that increasing the intestinal SCFAs content, particularly propionate and butyrate, effectively improved menopausal symptoms in OVX rats. Notably, propionate stimulated Bifidobacterium, thereby maintaining intestinal barrier permeability, while decreased levels putatively dysregulated neurotransmitter signalling, inflammation and oxidative stress (Ref. Reference Skonieczna-Żydecka, Grochans, Maciejewska, Szkup, Schneider-Matyka, Jurczak, Łoniewski, Kaczmarczyk, Marlicz, Czerwińska-Rogowska, Pełka-Wysiecka, Dec and Stachowska88). Li et al. reported that intrarectal sodium propionate administration for 1 week in chronic unpredictable mild stress (CUMS) rats upregulated 5-HT, norepinephrine (NE) and dopamine (DA) levels, and improved intestinal microecological balance to alleviate depression-like behaviour (Ref. Reference Li, Hou, Wang, Jia, Qin and Wu89). Additionally, butyrate also exerted anti-inflammatory and protective intestinal barrier effects, with potential effects in the immune system and toward ischemic injury. For example, in a dynamic alteration study of the gut microbiota (Ref. Reference Zhao, Wang, Hu, Xing, Gong, Liu, Qin, Xu, du, Ai, Peng and Li90), the butyrate-producing gut microbiota Eubacterium was found to be significantly reduced in a group of PMD model mice, while the relative abundance of gut microbial species, such as Escherichia coli and Veillonella, increased over time after surgery in OVX mice. Significant differences in gut microbial species abundance were recorded between OVX and control groups at different time points. Thus, Veillonella and E. coli may represent intestinal pathogens that affect the immune system and cause inflammatory responses. Notably, Veillonella is capable of producing acetate and propionate, thereby maintaining homeostasis in humans (Ref. Reference Zhang and Huang91). However, it has also been found that Veillonella can activate macrophages through the LPS/TLR4 pathway, triggering intestinal inflammation and further aggravating intestinal flora imbalance (Ref. Reference Zhan, Liu, Pan, Bao, Yan and Hong92). These findings suggest that oestrogen could reduce intestinal inflammation by regulating the abundance of butyrate-producing gut microbiota species. Thus, SCFAs may play important roles in PMD via the MGB axis.

Current studies report that ketamine exerts anti-inflammatory and neurotransmitter effects, and effectively improves treatment-resistant depression (Ref. Reference Wei, Chang and Hashimoto93). Ketamine is an N-methyl-D-aspartate receptor (NMDAR) antagonist, which alters intestinal microbiota composition (Ref. Reference Hashimoto94). It was reported that ketamine significantly increased Lactobacillus johnsonii levels in LPS-induced depressed mice (Ref. Reference Hua, Huang, Liu, Xu, Xu, Wu, Liu, Wang and Yang95). Additionally, ketamine and its metabolites improved SCFAs-producing microbiota, such as Butyricimonas, Turicibacter, Clostridiales, etc., thereby improving the depression state (Ref. Reference Hua, Huang, Liu, Xu, Xu, Wu, Liu, Wang and Yang95). However, while the impact of ketamine on gut microbiota composition has been documented to some degree, how ketamine regulates the intestinal microbiota and its metabolites, such as SCFAs metabolism, remains unclear. Therefore, further studies should investigate the mechanisms whereby ketamine, as a promising antidepressant, regulates gut microbiota metabolism and associated metabolites. The α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor-brain-derived neurotrophic factor (BDNF) mechanistic target of rapamycin (mTOR) signalling pathway is postulated to enhance synaptic function in the medial prefrontal cortex and contribute to rapid antidepressant ketamine effects (Ref. Reference Liang96). Additionally, studies have also shown that women are more sensitive to ketamine than men (Ref. Reference Ponton, Turecki and Nagy97). However, the role of ketamine in PMD via the MGB axis remains unclear; therefore, further studies are required to clarify this.

Interactions between oestrogen and the gut microbiota

The gut microbiota has important roles in the female reproductive endocrine system by interacting with oestrogen, androgens, insulin and other hormones (Ref. Reference Qi, Yun, Pang and Qiao118). Of these, oestrogen deficiency is an important influencing factor in PMD. Oestrogen is a steroid hormone that is mainly secreted by the ovaries and the placenta during pregnancy (Ref. Reference Hu, Ding, Zhang, Kang, Ma and Zhao119). E₂ is the most biologically active oestrogen in women, and the main oestrogen that affects the main brain functions (Ref. Reference Hu, Ding, Zhang, Kang, Ma and Zhao119). Oestrogen in the blood has both free and conjugated forms, of which the latter is the main form, but the biologically active the former (Ref. Reference Hu, Ding, Zhang, Kang, Ma and Zhao119). Oestrogen is a major gut microbiome regulator, with the gut microbiome gene pool (‘oestrogen-ome’) capable of metabolizing oestrogen (Ref. Reference Plottel and Blaser120). The hormone not only prevents beneficial bacterial loss and promotes their growth and reproduction, but also reduces pathogenic bacterial numbers and reduces lipopolysaccharide (LPS)-induced inflammation, thus exerting anti-inflammatory roles (Ref. Reference Park, Pilla, Panta, Pandey, Sarawichitr, Suchodolski and Sohrabji81). Gut microbiota dysbiosis in middle-aged women may potentially reduce free oestrogen levels and trigger oestrogen-related pathology, leading to depressive disorder moods. In terms of oestrogen interactions with gut microbes, the vast majority of studies have been conducted in animal models (Refs Reference Menon, Watson, Thomas, Allred, Dabney, Azcarate-Peril and Sturino121,Reference Khosravi, Bunte, Chiow, Tan, Wong, Poh, Doli Sentosa, Seow, Amoyo, Pettersson, Loke and Vadivelu122). For example, Li et al. showed that 3β-hydroxysteroid dehydrogenase, expressed by Klebsiella aerogenes, effectively degraded E₂, thereby reducing levels in the mouse brain and blood, and eventually leading to depression-like behaviours (Ref. Reference Li, Sun, Tong, le, Yao, Tao, Liu, Jiao, Mei, Chen, Liu, Wang and Li123). Another study observed that Proteobacteria and LPS biosynthesis were reduced in male mice and also in a PMD mouse model treated with E₂, indicating that oestrogen reduced intestinal permeability and LPS-induced inflammation, thereby reducing metabolic endotoxemia (Ref. Reference Kaliannan, Robertson, Murphy, Stanton, Kang, Wang, Hao, Bhan and Kang60).

Only a few clinically relevant studies in this area have been published; Shin et al. divided subjects based on low, medium and high sex hormone levels, and showed that females in the high-dose group had increased Bacteroidetes and decreased Firmicutes abundance when compared with females in the low-dose group, while Slackia and Butyricimonas were significantly negatively correlated with serum estradiol levels (Ref. Reference Shin, Park, Sim, Kim, Joung and Shin124). In a paired premenopausal and postmenopausal female study, Santos-Marcos et al. reported that serum estradiol levels were positively correlated with the Gammaproteobacteria class and an unknown genus from Myxococcales, which was negatively correlated with Prevotellaceae (Ref. Reference Santos-Marcos, Rangel-Zuñiga, Jimenez-Lucena, Quintana-Navarro, Garcia-Carpintero, Malagon, Landa, Tena-Sempere, Perez-Martinez, Lopez-Miranda, Perez-Jimenez and Camargo36). Zhu et al. identified a weak positive correlation between estradiol levels and Shewanella putrefaciens and Erwinia amylovora (Ref. Reference Zhu, Liao, Yao, Liang, Li, Liu, Yang, Ji, Wei, Tan, Liang, Chen, Lin, Zhu, Huang, Tian, Tang, Wang and Mo125). These results suggest that interactions between the gut microbiota and oestrogen have key roles in PMD development via immune-mediated inflammatory pathways (Figure 4).

Figure 4. Potential interaction mechanisms between gut microbiota and oestrogen in perimenopausal depression (PMD). Gut microbiota dysbiosis reduces β-glucuronidase production, impairing the deconjugation of conjugated oestrogen into bioactive free oestrogen. Oestrogen deficiency promotes proinflammatory cytokine secretion (e.g., TNF-α, IL-1β, IL-6) by T cells and macrophages, leading to peripheral immune activation and chronic inflammation. These peripheral cytokines traverse the gut-brain axis, activate microglia and amplify neuroinflammation through sustained release of proinflammatory mediators. Concurrently, diminished free oestrogen entry into the brain suppresses oestrogen receptor β (ERβ) signalling, exacerbating depressive phenotypes in PMD. Furthermore, oestrogen deficiency perpetuates gut dysbiosis, establishing a vicious cycle between neuroendocrine dysfunction and microbial imbalance.

Estrogen and inflammatory responses

Many studies have shown that depression is a neuroinflammatory disease, with neuroinflammation implicated in depression occurrence and development (Ref. Reference Köhler, Freitas, Maes, de Andrade, Liu, Fernandes, Stubbs, Solmi, Veronese, Herrmann, Raison, Miller, Lanctôt and Carvalho126). Studies have also shown that decreased ovarian function in women during menopause is associated with increased proinflammatory cytokine levels (Ref. Reference Pfeilschifter, Köditz, Pfohl and Schatz127). E₂ exerts regulatory effects on NF-κB, which is a central regulator of inflammatory pathways and regulates multiple target gene expression. Oestrogen deficiency affects oestrogen target gene expression, leading to increased IL-7 levels and promoting T cell activation, which further induces proinflammatory cytokine secretion, such as IL-1, IL-6 and TNF (Ref. Reference Cheng128). Currently, considerable evidence suggests that oestrogen exerts anti-inflammatory effects by binding to the ER. The oestrogen receptors ERα, ERβ and GRP30 are widely expressed in microglia, astrocytes and neurons, and exert anti-inflammatory effects (Refs Reference Kata, Gróf, Hoyk, Ducza, Deli, Zupkó and Földesi129,Reference Qu, Li, Xu, Zhang, Xie and Wang130,Reference Baez-Jurado, Rincón-Benavides, Hidalgo-Lanussa, Guio-Vega, Ashraf, Sahebkar, Echeverria, Garcia-Segura and Barreto131,Reference Hwang, Lee, Kim and Kwon132,Reference Smith, das, Butler, Ray and Banik133,Reference Correa, Ronchetti, Labombarda, de Nicola and Pietranera134). Upon binding to the ERα in microglia, E₂ activates phosphatidylinositol 3-kinase (PI3K)/protein kinase B (AKT) signalling, which in turn inhibits NF-κB activation, thereby reducing proinflammatory cytokine transcription (Refs Reference Murphy, Guyre and Pioli135,Reference Sabbatini and Kararigas136). E₂ also increases GABA levels in the hippocampus and frontal cortex by activating ERβ and/or GPR30, and up-regulates GABA-related genes in the amygdala and hippocampus (Ref. Reference Tongta, Daendee and Kalandakanond-Thongsong137). Studies have shown that increasing ERβ expression alleviates depression-like behaviours during perimenopausal periods in OVX-CUMS mice (Ref. Reference Zhang, Wang, Pan, Yang and Wu110), whereas ERβ knockout mice have significantly increased anxiety-like behaviours (Ref. Reference Walf, Koonce and Frye138). Furukawa et al. found that E₂ may affect social behaviour in OVX mice by regulating TPH expression in the raphe nucleus and serotonin release in the amygdala via GPR30 (Ref. Reference Furukawa, Izumo, Aoki, Nagashima, Ishibashi and Matsuzaki139). Upon activation, microglia released proinflammatory cytokines (TNF-α, IL-1β and IL-6), which initiated neuroinflammation (Ref. Reference Jia, Gao and Hu140). Also, neurons triggered ATP release by activating NMDARs, which induced microglial processes (Ref. Reference Dissing-Olesen, LeDue, Rungta, Hefendehl, Choi and MacVicar141) and led to depression (Refs Reference Zhang, He, Qiao, Zhou, He, Yi, Zhang, Mo, Li, Jiang and You142,Reference Wang, He, Sun, Ren, Liu, Wang and Yang143). Clinical studies also reported that peripheral estradiol serum levels in perimenopausal women were inversely correlated to serum IL-8 and TNF-α levels, and also microglial and astrocyte reactivity (Ref. Reference Liang96). Thus, oestrogen may exert anti-inflammatory effects by activating the ER on glial cells, while its deficiency exerts inflammatory responses in glial cells, which impacts on brain function. Notably, there is a ‘critical period’ for oestrogen neuroprotective effects. It was reported that oestrogen therapy should be given immediately after brain injury, as administration at 122 weeks after oophorectomy results in treatment ineffectiveness, and long-term oestrogen deprivation decreases hippocampal ERα receptors (Ref. Reference Hwang, Lee, Kim and Kwon132). Additionally, oestrogen, above physiological doses, was shown to activate inhibitory ERβ levels, thereby aggravating depression-like behaviours (Refs Reference Sun, Li, Zhao, Dong, Xie, Liu, Yang and Cui144,Reference Li, Zhang, Chen, Liu, Liu, Ning, Cao and Zhao145). Furthermore, several studies reported individual efficacious differences for oestrogen alone in treating PMD, e.g., transdermal oestrogen (0.1 mg/day for 4 or 12 weeks) or oral oestrogen (1.25 mg/day for 4 weeks) improved symptoms in patients with PMD (Ref. Reference Xiang146). Joffe et al. observed that perimenopausal women with depression had improved symptoms after transdermal estradiol treatment (0.05 mg/day for 8 weeks) (Ref. Reference Joffe, Petrillo, Koukopoulos, Viguera, Hirschberg, Nonacs, Somley, Pasciullo, White, Hall and Cohen147). However, other studies reported no significant differences from placebo (Ref. Reference Schmidt, Wei, Martinez, Dor, Guerrieri, Palladino, Harsh, Li, Wakim, Nieman and Rubinow148). Therefore, oestrogen therapy for PMD has limitations. In the future, multi-centre, large-sample, long-term randomized controlled trials must systematically evaluate dose–response relationships, intervention times and biomarker safety to clarify the clinical applicability of oestrogen therapy.

The gut microbiota and inflammatory responses

Recent studies have shown that the gut microbiota has an increasingly close relationship with inflammatory responses in host physiological and pathological processes. Intestinal microbes maintain mucosal barrier integrity by regulating intestinal epithelial cell growth and differentiation, tight junction protein expression and intestinal mucosal permeability. Changes in intestinal microbial composition can damage the intestinal mucosal barrier, destroying connection functions between cells, increasing permeability and increasing the transport of inflammatory mediators (Ref. Reference Yang, Wang, Chen, Guo and Dong31). Inflammatory responses associated with gut microbes are not only present in gastrointestinal diseases, such as irritable bowel syndrome and colorectal cancer, but also affect cardiovascular, reproductive, metabolic, autoimmune and neurodegenerative diseases (Ref. Reference Yang, Wang, Chen, Guo and Dong31). A potential link between the gut microbiota and MDD is low-grade chronic inflammation (Ref. Reference Park, Pilla, Panta, Pandey, Sarawichitr, Suchodolski and Sohrabji81). Studies have shown that gut microbiota metabolites and microbial cell components (e.g., LPS) can pass through the damaged intestinal barrier (‘intestinal leakage’), leading to increased inflammatory factor levels, such as IL-6, TNF-α and IL-1β, thereby exacerbating systemic inflammatory responses (Refs Reference Liu, Wang, Chen, Zhang, Zhang and Xie33,Reference Kiecolt-Glaser, Wilson, Bailey, Andridge, Peng, Jaremka, Fagundes, Malarkey, Laskowski and Belury149,Reference Slyepchenko, Maes, Jacka, Köhler, Barichello, McIntyre, Berk, Grande, Foster, Vieta and Carvalho150). These cytokines can reach the brain via neuroanatomical and neuroendocrine pathways and influence mental health and behaviour (Ref. Reference Guo, Wang, Su, Xu, Ji, Zhang, Zhao, Ma, Deng and Ma151). Oestrogen fluctuations in perimenopausal women can aggravate gut microbiota imbalance, which then triggers the activation of the following key inflammatory signalling pathways.

The NOD-like receptor protein 3 (NLRP3) inflammasome pathway

Inflammasomes play a key role in the activation of the innate immune system and the maturation of inflammatory cytokines. The dysregulation of inflammasomes may be related to MDD (Ref. Reference Wong, Inserra, Lewis, Mastronardi, Leong, Choo, Kentish, Xie, Morrison, Wesselingh, Rogers and Licinio156). Clinical studies have shown that treatment failures in patients with major depression are associated with elevated inflammatory mediator serum levels (Ref. Reference Haroon, Daguanno, Woolwine, Goldsmith, Baer, Wommack, Felger and Miller111). In recent years, researchers have proposed the ‘microbiota-gut-inflammasome-brain axis,’ which suggests that interactions between the intestinal flora and inflammasomes can affect the intestinal microecological balance and physiological functions in the brain (Ref. Reference Pellegrini, Antonioli, Calderone, Colucci, Fornai and Blandizzi23). In particular, NLRP3 inflammasomes have important roles in depression occurrence and development. Studies have shown that Gram-negative bacteria, such as Francisella novicida, Salmonella typhimurium, Citrobacter and E. coli, can stimulate bone marrow-derived macrophages to produce IL-1β and IL-18 in an NLRP3-dependent manner (Ref. Reference Man, Karki, Sasai, Place, Kesavardhana, Temirov, Frase, Zhu, Malireddi, Kuriakose, Peters, Neale, Brown, Yamamoto and Kanneganti157). An Aspergillus chimaera was shown to stimulate newly recruited monocytes to induce NLRP3-dependent IL-1β release and exacerbate intestinal inflammatory injury (Ref. Reference Seo, Kamada, Muñoz-Planillo, Kim, Kim, Koizumi, Hasegawa, Himpsl, Browne, Lawley, Mobley, Inohara and Núñez158). Recent evidence has also shown that the gut microbiota promotes proinflammatory cytokine production such as IL-1β via NLRP3 inflammasome activation, leading to acute pancreatitis, colitis and depression (Ref. Reference Huang, Ma, Ze, Zhao, Zhang, Lv, Zheng and Liu159). Together, these findings suggest that gut microbes activate the inflammasome in immune cells in an NLRP3-dependent manner.

The biological characteristics and activation mechanisms of the NLRP3 inflammasome

The NLRP3 inflammasome is a complex composed of NLRP3 receptor protein, adaptor protein apoptosis-associated speck-like protein containing a CARD (ASC) and Caspase-1, which is widely distributed in the immune system and non-immune systems, such as macrophages, B cells, T cells, neurons, astrocytes and microglia (Refs Reference Yang, Wang, Chen, Guo and Dong31,Reference Carranza-Aguilar, Hernández-Mendoza, Mejias-Aponte, Rice, Morales, González-Espinosa and Cruz160). The mechanism of action of NLRP3 inflammasome involves dual signalling activation: The first priming signal is the release of LPS due to dysbiosis of gut microbiota (such as a decreased F/B ratio) (Ref. Reference Xia, Guo, Lian, Yan, Ma, Cheng, Xu, He and Zhang4). Pattern recognition receptors (PRRs), such as TLR4, are activated upon recognizing LPS, subsequently inducing activation of the NF-κB signalling pathway (Ref. Reference Xia, Guo, Lian, Yan, Ma, Cheng, Xu, He and Zhang4). NF-κB translocates to the nucleus, upregulating the transcription and expression of precursor proteins such as NLRP3, pro-IL-1β and pro-IL-18 (Ref. Reference Xia, Guo, Lian, Yan, Ma, Cheng, Xu, He and Zhang4). The second activation signal occurs when gut microbiota metabolites (such as ATP and SCFAs) or pathogen-associated molecular patterns (PAMPs) trigger the binding of NLRP3 with ASC and Caspase-1, assembling into the inflammasome complex (Refs Reference Xia, Guo, Lian, Yan, Ma, Cheng, Xu, He and Zhang4,Reference Wong, Inserra, Lewis, Mastronardi, Leong, Choo, Kentish, Xie, Morrison, Wesselingh, Rogers and Licinio156,Reference Lee, Suh, Ryter and Choi161,Reference Lin and Mei162). Activated Caspase-1 cleaves pro-IL-1β and pro-IL-18 into mature IL-1β and IL-18, and induces Gasdermin d-mediated pyroptosis, releasing large amounts of IL-1β and IL-18 to further amplify the inflammatory response (Refs Reference Xia, Guo, Lian, Yan, Ma, Cheng, Xu, He and Zhang4,Reference Wong, Inserra, Lewis, Mastronardi, Leong, Choo, Kentish, Xie, Morrison, Wesselingh, Rogers and Licinio156,Reference Lee, Suh, Ryter and Choi161,Reference Lin and Mei162). IL-1β and IL-18 then activate microglia via an impaired BBB or vagal afferent signals. IL-1β activates microglia via IL-1R1 to release TNF-α and IL-6, and inhibits hippocampal neurogenesis (Figure 5).

Figure 5. NLRP3 inflammasome mechanisms in PMD mediated by the MGB axis. Oestrogen deficiency disrupts the gut microbiota, which disrupts the gut barrier and leads to harmful substance release (e.g., LPS). TLR4 in intestinal epithelial and immune cells recognizes LPS and activates NF-κB and the NLRP3 inflammasome via MyD88 to promote caspase-1 maturation and inflammatory cytokine (IL-1β and IL-18) release, thereby destroying the intestinal barrier and increasing intestinal permeability, triggering intestinal and peripheral inflammation. Inflammation increases BBB permeability and enhances microglial activity via the MGB axis, leading to neuroinflammation and NLRP3 inflammasome activation, which ultimately causes PMD. TLR4: toll-like receptor 4; LPS: lipopolysaccharide; MyD88: myeloid differentiation primary response protein 88: NLRP3: NOD- like receptor protein 3; NF-κB: nuclear transcription factor-κB; IL-1β: interleukin-1β; and IL-18: interleukin-18; PMD: perimenopausal depression.

NLRP3 activation and neuroinflammatory injury in the hippocampus

The hippocampus is an important part of the limbic system. The NLRP3 inflammasome and its activation in the hippocampus were related to depression-like behaviour induced by oestrogen deficiency in animals. Several animal studies reported that depression-like behaviour in OVX animals was associated with NLRP3 inflammasome activation and enhanced IL-1β, IL-18, TLR4 and NF-κB expression in the hippocampus (Refs Reference Xu, Sheng, Bao, Wang, Lu and Ni153,Reference Wang, Xu, Sheng, Ni and Lu163,Reference Menze164,Reference Hou165). However, IL-1β, IL-18, caspase-1 and NLRP3 inflammasome gene knockouts improved depression-like behaviour in mice (Refs Reference Xia, Guo, Lian, Yan, Ma, Cheng, Xu, He and Zhang4,Reference Yamanishi, Miyauchi, Mukai, Hashimoto, Uwa, Seino, Li, Gamachi, Hata, Kuwahara-Otani, Maeda, Watanabe, Yamanishi, Yagi, Okamura and Matsunaga166). Clinical studies also reported that serum IL-1β, IL-18, NLRP3 and caspase-1 expression in MDD patients was significantly increased, but after antidepressant treatment, indicators were significantly decreased, thereby effectively improving depressive states in patients (Refs Reference Alcocer-Gómez, de Miguel, Casas-Barquero, Núñez-Vasco, Sánchez-Alcazar, Fernández-Rodríguez and Cordero167,Reference Harsanyi, Kupcova, Danisovic and Klein168). Therefore, these observations reinforce the notion that the NLRP3 inflammasome may be at the centre of inflammatory cascades following oestrogen deficiency, ultimately leading to depression. Wong et al. showed that caspase-1 knockout reduced anxiety- and depression-like behaviours in mice with chronic restraint stress (CRS), and found that the antibiotic minocycline regulated the gut microbiota and inhibited caspase-1, thereby improving depression-like behaviour and gut microbiota disorder in CRS mice (Ref. Reference Wong, Inserra, Lewis, Mastronardi, Leong, Choo, Kentish, Xie, Morrison, Wesselingh, Rogers and Licinio156). Zhang et al. observed that behavioural changes were improved in NLRP3 gene-deficient mice when compared to CUMS mice, and when faecal microbes from NLRP3 gene-deficient mice were transplanted into recipient CUMS mice, mice exhibited significantly improved depression-like behaviour (Ref. Reference Zhang, Huang, Cheng, Wang, Chao, Li, Zheng, Xie, Zhang and Yao169). Another study confirmed that NLRP3 inflammasome activity in CUMS mice was inhibited by a TLR4 inhibitor, thus effectively relieving a depressive state in animals (Ref. Reference Fu, Wang, Hao, Dang and Jiang170). Studies have shown that chronic ethanol exposure (CEE) induces depression-like behaviour in mice via FMT. Severe hippocampal neuroinflammation and NLRP3 inflammasome activation are also observed in recipient mice, indicating that depression-like behaviour is regulated by the gut microbiota (Ref. Reference Yao, Zhang, Yu, Yuan, Shen, Lan, Liu, Chen, Meng, Wu, Zhang and Wang171). Inflammatory factors are critical for NLRP3 inflammasome activation in the hippocampus. A causal mediation analysis in mice receiving the NLRP3-shRNA group FMT showed increased Firmicutes, Actinobacteria, Erysipelotrichi and Allobaculum abundance, and reduced Bacteroidetes, Bacteroidia, Verrucomicrobiae and Bacteroidales abundance, while mice showed increased depression-like behaviour risks, which were mediated by LPS, IL-1β, TNF-α, interferon (IFN)-γ and IL-12p70 (Ref. Reference Yao, Zhang, Yu, Yuan, Shen, Lan, Liu, Chen, Meng, Wu, Zhang and Wang171). These observations suggest that the gut microbiota activates the NLRP3 inflammasome in the hippocampus via peripheral inflammatory factors (Ref. Reference Yao, Zhang, Yu, Yuan, Shen, Lan, Liu, Chen, Meng, Wu, Zhang and Wang171). These aforementioned studies suggest that the MGB axis-mediated NLRP3 inflammasome in the hippocampus may be an important treatment target for PMD. However, the interplay between the gut microbiota and the NLRP3 inflammasome is poorly understood. Notably, most current studies have focused on animal models, while humans and rodents exhibit significant differences in NLRP3 expression profiles: the activation threshold of NLRP3 in human microglia is lower than that in rodents, and human microglia are more sensitive to oestrogen fluctuations (Ref. Reference Koller, Nguyen, Snouwaert, Gabel and Ting172), suggesting that species-specific mechanisms require further validation.

Potential gut microbiota applications for PMD treatment

A growing body of evidence now suggests that different treatments targeting the microbiota can effectively improve depression by re-establishing the correct intestinal microecological balance. Among these, FMT, probiotics and prebiotics have potential clinical applications (Figure 6).

Figure 6. Potential gut microbiota application for PMD: therapeutic approaches targeting the microbiome, include FMT, probiotics (Lactobacillus, Bifidobacterium, etc.) and prebiotics (fructo-oligosaccharides, BOS, etc.). PMD: perimenopausal depression; FMT: faecal microbiota transplantation.

FMT

FMT rapidly reshaped the gut microbiome by transplanting faecal microbiota from healthy donors into the gut of patients. FMT has been shown to relieve depressive state (Table 3), e.g., animal studies have shown that FMT from healthy donors alleviated alcohol-induced depression-like behaviour in mice (Ref. Reference Liang96), and also improved stress-induced depression-like behaviour in rats by improving the gut microbial balance, reducing intestinal barrier damage and inhibiting neuroinflammation (Ref. Reference Ponton, Turecki and Nagy97). In a randomized controlled trial, two female patients with MDD, aged 50–60 years old, showed significant improvement in depressive symptoms after the oral administration of frozen FMT capsules for 4 weeks, and in one patient, effects lasted 8 weeks (Ref. Reference Doll, Vázquez-Castellanos, Schaub, Schweinfurth, Kettelhack, Schneider, Yamanbaeva, Mählmann, Brand, Beglinger, Borgwardt, Raes, Schmidt and Lang181). Furthermore, a recent randomized controlled, double-blind trial validated the feasibility, efficacy and safety of an FMT enema in adults with moderate to severe depression (Ref. Reference Green, Berk, Mohebbi, Loughman, McGuinness, Castle, Chatterton, Perez, Strandwitz, Athan, Hair, Nierenberg, Cryan and Jacka182). At the same time, despite potential side effects and complications, the accessibility and tolerability of enema is high in some patients, suggesting its great therapeutic potential (Ref. Reference Green, Berk, Mohebbi, Loughman, McGuinness, Castle, Chatterton, Perez, Strandwitz, Athan, Hair, Nierenberg, Cryan and Jacka182). Critically, FMT opens up new avenues for PMD treatment; however, there is a lack of clinical research on FMT in PMD, and more studies are required to fully verify its efficacy.

Table 3. FMT action mechanisms in depression

Abbreviations: CEE: conjugated equine oestrogens; CUMS: chronic unpredictable mild stress; FMT: faecal microbiota transplantation; IBS: irritable bowel syndrome; IDO1: indoleamine 2,3-dioxygenase 1; MDD: major depressive disorder; NLRP3: NOD-like receptor protein 3; SCFAs: short-chain fatty acids; SD: Sprague Dawley.

Conclusions and outlook

Studies examining gut microbiota alterations in PMD patients are controversial. However, it cannot be denied that gut microbiota composition and metabolites in patients with depression are significantly different from healthy individuals, with gut microbiota status in patients with depression more often in a proinflammatory state. Many studies have reported that the gut microbiota regulates inflammatory signalling via the NLRP3 inflammasome in the MGB axis and then affects brain homeostasis. However, how the brain regulates intestinal inflammation via efferent pathways remains unclear. Therefore, an in-depth understanding of the relationships between the gut microbiota and the NLRP3 inflammasome, and the identification of related foods or probiotics that regulate intestinal microecological balance, can provide new treatment directions for PMD. To date, research on gut microbiota mechanisms in oestrogen-deficiency-induced depression has mainly focused on animal models, but there is a lack of relevant clinical research on the dynamic monitoring of gut microbiota changes in depression. It will be important to investigate the effects of oestrogen deficiency on the gut microbiota and its associated metabolic and immune diseases in the short term. Additionally, there is a relative lack of clinical research on PMD. In the future, for patients with PMD, we need to comprehensively examine gut microbiota changes, perform microbial-targeted therapies, dynamically monitor gut microbiota changes, and conduct long-term, follow-up prognosis studies.

Animal models of OVX undergo surgical removal of the ovaries, resulting in a dramatic decrease in hormone levels, and this change is significantly different from the course of natural menopause in humans. Natural menopause in humans is a gradual process, during which the body has a series of complex physiological regulatory mechanisms to adapt to the changes in hormone levels. However, the OVX model lacks this natural adaptive regulatory process. Furthermore, microbiota changes and neuroinflammatory responses during perimenopause in humans may be influenced by multiple factors, including lifestyle, dietary habits and genetic background, which are difficult to fully simulate in animal models. This may be an important reason for the difference in the incidence of depression between OVX animal models and the actual situation in humans. Oestrogen may improve female depression through two pathways of action. On the one hand, oestrogen can indirectly inhibit depression by interacting with gut microbiota/intestinal epithelial cells, regulating the composition and function of the intestinal microbial community, affecting the synthesis and metabolism of neurotransmitters and regulation of the immune system. On the other hand, oestrogen can directly act on the nervous system by interacting with oestrogen receptors in the brain to regulate the release of neurotransmitters and neuroendocrine function, thereby improving depressive symptoms. It is not clear whether the interaction pathway between oestrogen and gut microbiota/intestinal epithelial cells is the main pathway. Although these probiotics showed positive effects in OVX animal models, it is not sufficient to conclusively infer the dominance of this pathway in the treatment of depression in humans. Therefore, the interaction between oestrogen and gut microbiota/intestinal epithelial cells cannot be considered as the main way to treat female depression based on the results of animal experiments, and more in-depth research based on humans is needed in the future.

Although several studies have shown that regulating the gut microbiota can improve depressive symptoms, therapeutic effects can significantly vary between individuals, with such effects potentially due to multi-dimensional interactions in host genetics, strain functions and metabolic phenotypes. Many studies have only focused on common probiotics such as Lactobacillus and Bifidobacterium, while other potentially beneficial or harmful microbial communities are poorly studied. This narrow perspective may mean that some important microbial-host interaction mechanisms are inadvertently missed or ignored. Therefore, to overcome the limitations of OVX animal models, to systematically integrate human pathophysiological characteristics and clinical data, to improve the value of translational medicine, and formulate corresponding translational research strategies, ‘precision microbiota medicine’ must be promoted via integrated multi-omics, artificial intelligence predictions and dynamic monitoring technology to address both standardization and ethical challenges (Table 4).

Table 4. Strategies addressing future PMD translational research

Abbreviations: PMD: perimenopausal depression; OVX: ovariectomized; SCFAs: short-chain fatty acids.