Microbiota–gut–brain axis and its therapeutic applications in neurodegenerative diseases
微生物群-肠-脑轴及其在神经退行性疾病中的治疗应用

Alzheimer’s disease

Accumulating evidence has demonstrated the interaction between gut microbiota and microglia in AD. In the triple transgenic AD (3xTg-AD) mouse model, the development of AD pathologies, including Aβ plaque, hyperphosphorylated tau, synaptic dysfunction, and microglial activation appears to be influenced by the gut microbiome. This is evident as SPF 3xTg-AD mice exhibit greater AD pathologies compared to GF 3xTg-AD mice. Importantly, FMT from AD patients to GF 3xTg-AD mice restored the main AD pathologies and microglial activation.¹⁹⁶

Similar findings have been reported in GF and antibiotic-treated amyloidogenic APP/PS1 mice.^54,197–199 It was reported that GF condition confers protection against Aβ pathology and microglial activation in APP_SWE/PS1_L166P (APP/PS1-21) mice. However, this protection was diminished following FMT from 12-month-old conventionally raised APP/PS1-21 mice to 4-month-old GF APP/PS1-21 mice.¹⁹⁸ Similar trends were also observed in APP/PS1-21 following gut microbiota depletion using long-term (5-week) and short-term (7-day) antibiotic treatment.^54,197 Long-term perturbation of gut microbiome using antibiotic cocktail resulted in reduced Aβ deposition, reduced plaque-localized microglia and altered transcriptional profile of microglia (increased homeostatic microglial genes and decreased MGnD genes) in 7-week-old male APP/PS1-21 mice. Interestingly, these effects were absent in female mice, suggesting potential sexual dimorphism in their responses to gut microbiome manipulation. Importantly, the AD pathologies in antibiotic-treated male mice were partially restored after 3-week FMT from age-matched male APP/PS1-21 mice.¹⁹⁷

Recently, growing studies have illustrated the presence of critical windows of microbial development, during which early-life modulation of the gut microbiome has a long-lasting impact on different aspects of physiology.^{186,200–202} To further validate the time-specific role of the gut microbiome in AD, Dodiya et al. repeated the experiment using short-term 7-day antibiotic treatment administered from postnatal day 14 to day 21, and sacrificed the mice at 9 weeks of age. Consistent with their previous findings,¹⁹⁷ short-term antibiotic treatments effectively reduced Aβ pathology, reduced DAM population, and expression of microglial sensome genes in male APP/PS1-21 mice, which were reversed by FMT from age-matched mice. Interestingly, microglia depletion using colony-stimulating factor 1 receptor inhibitor, PLX5622, mitigated the protective effect of antibiotic treatment against amyloidosis. These results suggest that microglia are essential mediators of the microbiota–gut–brain axis in Aβ pathology.⁵⁴

Similarly, the gut microbiota is required for microglial activation in 5xFAD mice.^203–205 5xFAD mouse model expresses five familial AD mutations and develops Aβ accumulation and gliosis as early as 2 months of age, synaptic degeneration by 4 months of age, and cognitive impairment as early as 4–5 months of age.^206,207 It was found that gut microbiota ablation using 5-month antibiotics treatment prevented microglial activation in 7-month-old 5xFAD mice by reducing immune cell infiltration.²⁰³ Moreover, 5-month antibiotics treatment alleviated AD pathologies and microglial activation in 6.5-month-old 5xFAD mice by inhibiting CCAAT/enhancer binding protein β/asparagine endopeptidase (C/EBPβ/AEP) signaling.²⁰⁴

The involvement of the gut microbiome in facilitating microglial activation is also evident in tau-mediated neurodegeneration. A recent study demonstrated the interplay between gut microbiota, tau and APOE in AD.⁵⁶ Seo and colleagues genetically engineered P301S tau transgenic mice to express different isoforms of human APOE (APOE3 and APOE4) and raised them in conventional or GF environments. Compared to conventionally raised P301S mice expressing human APOE4 (TE4 mice), their GF counterparts showed reduced signs of neurodegeneration (brain atrophy) and neuroinflammation (microglial and astrocytic activation). However, FMT from sex-matched conventionally raised TE4 mice mitigated the neuroprotective effects of GF conditions, indicating that the gut microbiota is responsible for the emergence of tau-mediated neurodegeneration. To further investigate the role of gut microbiota, the researchers induced early-life gut microbiota perturbation through a short-term antibiotic treatment administered from postnatal day 16 to day 22, and sacrificed the mice at 40 weeks of age. Interestingly, they observed sex-dependent and APOE isoform-dependent neuroprotection as the neuroprotective effects of antibiotic treatment were more pronounced in male mice expressing human APOE3 (TE3 mice).⁵⁶ The sex-dependent neuroprotective effects of microbiome perturbations in tau pathology are reminiscent of that observed in amyloid pathology,¹⁹⁷ underscoring the need for future research to consider the gender effects.

Aging is the predominant risk factor for neurodegenerative diseases as a result of the lifetime accumulation of neuropathologies.²⁰⁸ Notably, the gut microbiome of centenarians is associated with “youth-like” signatures (depletion of inflammatory pathobionts and enrichment of beneficial commensals), showing high similarity to those of young individuals.²⁰⁹ To investigate whether the acquisition of “youth-like” microbiota could restore aging-induced neurocognitive and immune impairments, Boehme et al. compared the effects of FMT from young (3–4 months; yFMT) or old (19–20 months; oFMT) mice into old recipient mice. Hippocampal transcriptomic analysis revealed that yFMT reversed the aging-induced alterations in the expression of six microglial sensome genes in old mice. These genes included Trem2, Dap12, C1qb, Fcgr2b, Gpr84, and Tlr13.²¹⁰ Of note, DAP12, also known as TYROBP, is an immunoreceptor tyrosine-based activation motifs (ITAM)-containing transmembrane adapter that associates with TREM2.⁹⁸ Dap12, along with Apoe, were among the most robustly upregulated genes during the microglial transition from homeostatic state to DAM phenotype.⁹⁹ On the other hand, complement component 1q (C1q) is the initiating protein of the classical complement pathway predominantly produced by microglia in the brain.²¹¹ Studies on P301S tau transgenic mice and plaque-bearing mouse models have shown that C1q binds to synapses and facilitates microglial phagocytosis of synapses.^103,212 Recent evidence also revealed that the complement-dependent synapse elimination in P301S mice involves coordinated action between microglia and astrocytes.²¹³

Altogether, the evidence from different mouse models consistently underscores the significance of the microbiota–gut–brain axis in AD pathologies.

Parkinson’s disease

Although GI symptoms and gut microbiota alterations are common in PD patients during the disease course,^214,215 the underlying mechanisms linking the gut microbiome and PD have only been unveiled recently. The first corroboration arises from the study by Sampson et al., which demonstrated that the development of α-synuclein pathology, microglial activation, and motor deficits in α-synuclein-overexpressing (ASO) mice appear to be influenced by the gut microbiome. This is evident as SPF ASO mice exhibit greater PD pathologies compared to their GF and antibiotic-treated counterparts. Importantly, FMT from PD patients to GF ASO mice restored the main disease features, including α-synuclein-mediated motor dysfunction.⁵⁵ In another study, transgenic rats overexpressing α-synuclein displayed progressive gut dysbiosis with aging, whereas a short-term antibiotic treatment mitigated α-synuclein expression in the forebrain.²¹⁶ Furthermore, FMT from 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-treated mice induced motor impairments and neurotransmitter loss in healthy mice. Conversely, FMT from healthy mice ameliorated gut dysbiosis and PD pathologies in MPTP-induced mice, including gut inflammation, glial activation, neurotransmitter abnormalities, and motor dysfunction.²¹⁷ In addition, the development of GI dysfunction and motor symptoms following chronic rotenone administration occurs only in conventionally raised mice, but not in GF mice.²¹⁸ These studies substantiated the significance of the microbiota–gut–brain axis in the pathogenesis of PD.

Amyotrophic lateral sclerosis

The motor neuron injury in ALS is thought to arise from multiple interacting pathophysiological mechanisms, including glial dysfunction and neuroinflammation.^89,219,220 Compared to AD and PD, the roles of gut microbiota in ALS remain relatively understudied. Nevertheless, growing evidence has indicated alterations in the gut microbiome and impaired intestinal barrier integrity in both ALS patients and animal models.^221–225

Notably, a gradual reduction in the abundance of Akkermansia muciniphila and reduced bacterial production of nicotinamide (NAM) were observed during the course of disease in SOD1^G93A mice. Conversely, supplementation of A. muciniphila was associated with increased motor neurons in the spinal cord, improved motor function, reduced brain atrophy, and prolonged lifespan.²²³ The gut microbiome is also closely linked to C9orf72 function.⁵⁷ The hexanucleotide (GGGGCC) repeat expansion in the C9orf72 gene has been identified as the leading genetic cause of ALS and FTD.^226–228 It was found that the gut microbiota is a potent modifier of disease severity and microglial activation in C9orf72-mutant mice raised in different environments. Specifically, C9orf72^−/− mice reared at pro-inflammatory environment (Harvard Institute) exhibit significantly different gut microbiota composition, greater autoimmune and inflammatory phenotypes, microglial activation, and shorter lifespan as compared to the C9orf72^−/− mice reared at pro-survival environment (Broad Institute). Importantly, FMT from C9orf72(Broad)^−/− mice significantly ameliorated the autoimmune and inflammatory phenotypes in C9orf72(Harvard)^−/− mice.⁵⁷

Alterations in the gut microbiome have been observed in SOD1^G93A mice prior to the onset of motor dysfunction, muscle atrophy, and immune cell activation in the spinal cord.²²⁴ In addition, gut dysbiosis precedes the aggregation of human-SOD1^G93A protein in the colon and intestine, along with the development of ENS dysfunction in SOD1^G93A mice.²²² These studies indicate an early microbial contribution to ALS disease pathology.

Interestingly, conflicting results have been reported regarding the consequences of gut microbiota depletion in ALS mouse models. Studies have demonstrated the beneficial effects of antibiotic treatment in ameliorating ALS pathologies. Antibiotic treatment has been shown to inhibit SOD1^G93A aggregation in the intestines of SOD1^G93A mice, coupled with improved enteric neuromuscular function.²²² Similar beneficial effects were observed in C9orf72-mutant mice. The administration of antibiotics, either prior to the onset or after the establishment of inflammation, effectively suppressed the emergence of inflammatory and autoimmune phenotypes in C9orf72^−/− mice. Moreover, lifelong antibiotics administration prevented the accumulation of infiltrating myeloid cells within the spinal cord and microglial activation.⁵⁷

However, detrimental effects of antibiotic treatment have also been reported in SOD1^G93A mice. Long-term antibiotic treatment induces motor neuron death in the spinal cord and brain atrophy, thereby exacerbating motor abnormalities in SOD1^G93A mice.²²³ In addition, antibiotic-induced dysbiosis markedly worsened disease progression in SOD1^G93A mice by downregulating homeostatic microglial genes and upregulating MGnD signatures in the spinal cord.²²⁹ Presently, there is a scarcity of research utilizing GF SOD1 mice due to the considerable challenge of rederivation, which is associated with high mortality rates.²²³ Further investigations are warranted to understand the roles of the gut microbiome and the implications of gut microbiota depletion in ALS. With the growing characterization of ALS-associated genes and the development of ALS animal models (extensively reviewed in ref. ²³⁰), we recommend future studies to move beyond SOD1 mice. This is relevant considering that SOD1-ALS accounts for only about 12% of familial and less than 2% of sporadic ALS cases.²³¹

Huntington’s disease

HD is an autosomal dominant neurodegenerative disease caused by a CAG repeat expansion in the huntingtin (HTT) gene. This results in misfolding and accumulation of mutant huntingtin protein in brain cells, including neurons, microglia, and astrocytes.²³² In addition to motor, cognitive, and psychiatric abnormalities, HD patients experience a range of GI disturbances, including nutrient deficiency, diarrhea, and unintended weight loss.^233,234 However, it was only recently that gut dysbiosis has been revealed in preclinical HD models and HD patients, and studies examining the interaction between gut microbiota and microglia remain absent.

The initial evidence of gut dysbiosis emerged from the R6/1 transgenic mouse model of HD. A notable difference in gut microbiota composition was observed between R6/1 mice and wild-type (WT) mice at 12 weeks of age (early disease stage), which coincided with the manifestation of motor deficits and weight loss.²³⁵ Gut dysbiosis and intestinal barrier impairment were also detected in R6/2 mice. It was found that R6/2 mice exhibited increased intestinal permeability and reduced colon length at 16 weeks of age (early-mid disease stage), as compared to age-matched WT mice.²³⁶ Moreover, a higher relative abundance of Bacteroidetes and a lower relative abundance of Firmicutes were reported in both R6/1 and R6/2 mice, as compared to WT mice.^235,236 However, a notable limitation of the current studies is the lack of metabolomic analyses, which greatly hinders our understanding of the metabolites that regulate HD pathogenesis.

A recent study on HD gene expansion carriers (HDGECs) has identified an altered gut microbiome compared to age-matched and gender-matched healthy controls. Moreover, a reduced abundance of Eubacterium hallii was associated with increased severity of motor deficits.²³⁷ E. hallii is a major butyrate-producing species with important health implications, and its depletion has been linked to several diseases.^238–240 In addition, E. hallii has also been shown to influence BA metabolism.^241,242 However, the study did not observe an increase in Bacteroidetes and a reduction in Firmicutes, as reported in both R6/1 and R6/2 mice.^235,236

Interestingly, the degree of gut dysbiosis appears to be influenced by the gender of the mice. This difference is evident in male R6/1 mice, which develops greater gut microbiota alterations than female R6/1 mice at 8 weeks of age. Moreover, the plasma levels of acetate were elevated only in male R6/1 mice at 14 weeks of age.²⁴³ Similar sexual dimorphism was observed in the application of FMT. A recent study performed FMT from WT mice into R6/1 mice and found that male R6/1 mice exhibited greater resistance to FMT engraftment when compared to female R6/1 mice. Consequently, the cognitive function of male R6/1 mice showed no discernible improvement as compared to their female counterparts. However, FMT is ineffective in ameliorating gut dysfunction and motor functions of R6/1 mice.²⁴⁴

Alzheimer’s disease

The effects of gut microbiota manipulation on astrocytes in AD have only emerged recently. Gut microbiota perturbation has also been shown to reduce reactive astrogliosis, promote a shift in astrocytes towards a more homeostatic-like state, and protect against amyloidosis and tau-mediated neurodegeneration. Interestingly, these effects appear to be more prominent in male mice.^56,61 The sexual dimorphism in astrocytic responses to gut microbiome perturbation is reminiscent of that in microglia, further underscoring the importance of future research to account for gender effects. However, the neuroprotective effects of gut microbiota depletion were diminished following microbiota restoration and SCFA supplementation. In particular, FMT from age-matched control mice restored astrogliosis in antibiotic-treated APP/PS1-21 mice, while SCFAs supplementation restored the gliosis and tau pathology in GF TE4 mice.^56,61 Although preliminary, these findings in animals demonstrated the involvement of gut microbiome in facilitating the development and progression of AD pathologies, including the modulation of astrocytic responses.

Parkinson’s disease

Our understanding of the roles of astrocytic dysfunction and pro-inflammatory glial responses in PD pathogenesis is expanding.²⁴⁶ The amelioration of gut dysbiosis in MPTP-induced mice via FMT, dietary intervention, and probiotic administration has been shown to alleviate neuroinflammation and dopaminergic neuron loss. These effects are attributed to the reduction of glial activation, as well as the restoration of metabolite and neurotransmitter abnormalities.^{217,259–261} In addition, it was found that FMT from healthy human controls mitigated MPTP-induced dysbiosis and neurotoxicity in MPTP-induced mice, whereas FMT from PD patients exacerbated these pathologies.²⁶¹ Taken together, these findings corroborate the contributory roles of gut microbiota in driving glial activation and neuropathologies.

Alzheimer’s disease

Several studies have reported reduced SCFAs-producing species and SCFA levels in individuals with mild cognitive impairment (MCI) and AD patients.^292–295 Notably, reduced fecal levels of SCFAs were negatively associated with Aβ deposition in patients with MCI.²⁹³ In addition, increased levels of HDAC2 and HDAC6 were detected in AD mouse models and AD patients.^296,297 Thus, HDAC inhibition represents a promising approach for the treatment of AD. This is exemplified by the notable findings that the genetic deletion of microglial Hdac1 and Hdac2 substantially ameliorated the cognitive deficits of 5xFAD mice by enhancing microglial phagocytosis of Aβ.²⁹⁸ Despite numerous studies supporting the pivotal roles of SCFAs in mediating gut microbiota-microglia communication, mechanistic studies elucidating the underlying mechanisms of SCFAs in AD remain limited and yield conflicting results.

Studies have demonstrated the neuroprotective effects of sodium butyrate in 5xFAD mice by inhibiting microglial activation and promoting synaptic plasticity (Fig. 3a).^299,300 Moreover, probiotic and prebiotic interventions aimed at elevating SCFA levels have demonstrated neuroprotective effects in AD mouse models by inhibiting glial activation and Aβ deposition.^301–306 Elevating butyrate through probiotic intervention (Clostridium butyricum) has been shown to inhibit microglial activation and reduce the levels of levels pro-inflammatory cytokines in APP/PS1 mice. Furthermore, butyrate exerts anti-inflammatory effects by downregulating the levels of cyclooxygenase-2 (COX-2) and CD11b, and suppressing NF-κB signaling in Aβ-induced BV-2 cells.³⁰³ Notably, an oral combination therapy (AMX0035) comprising sodium phenylbutyrate and tauroursodeoxycholic acid (TUDCA) is currently undergoing a phase II clinical trial to evaluate its safety and biological activity in AD patients [NCT03533257].

However, conflicting findings regarding the roles of SCFAs in AD have also been reported. A recent population-based study has revealed a positive association between serum propionic acid and cognitive decline in older adults, with potential mediation by hypercholesterolemia and diabetes.³⁰⁷ Interestingly, it has been demonstrated that SCFAs induce microglial activation and worsen Aβ pathology in both SPF and GF APP/PS1 mice.¹⁹⁹ Furthermore, SCFA supplementation has been found to trigger C/EBPβ/AEP signaling activation and induce cognitive impairment in GF 3xTg-AD mice.¹⁹⁶ The C/EBPβ is an inflammation-regulated transcription factor that regulates the expression of pro-inflammatory genes, thereby contributing to the pathogenesis of AD.^204,308,309 The AEP is a lysosomal cysteine protease that cleaves tau at N255 and N368 residues, and amyloid precursor protein (APP) at N373 and N585 residues, resulting in amyloidogenic fragmentation and tau hyperphosphorylation.^310,311 In GF 5xFAD mice, the administration of acetate aggravates hippocampal Aβ deposition by disrupting microglial phagocytosis of Aβ.¹⁹⁰ The detrimental effects of SCFAs are similarly evident in a tauopathy mouse model. A recent study demonstrated that SCFA supplementation in GF TE4 mice mitigated the neuroprotective effects of GF rearing, resulting in increased gliosis and tau pathology. Conversely, the depletion of SCFAs-producing bacteria using antibiotic treatment conferred protection against tau-mediated neurodegeneration and neuroinflammation in TE3 mice.⁵⁶ A recent study reported that intermittent fasting is effective in alleviating reactive microgliosis and astrogliosis, Aβ deposition, and cognitive impairment of 5xFAD mice by remodeling the microbiota–gut–brain axis. However, metabolomic analysis of cecal contents found that butyric acid was significantly downregulated in response to intermittent fasting, as compared to mice that were fed ad libitum.³¹²

The considerable heterogeneity observed across distinct mouse models of AD, each characterized by distinct pathological pathways, poses a significant challenge in anticipating the implications of SCFAs. Moreover, the majority of studies that reported the detrimental effects of SCFAs were conducted using GF mice,^{56,190,196,199} which are functionally and structurally abnormal across various physiological functions.³¹³ Thus, caution is warranted when extrapolating these findings to human diseases. In conclusion, it is evident that further investigation is warranted to elucidate the multifaceted nature and contextual significance of SCFAs in AD.

Parkinson’s disease

As with AD described previously, the roles of SCFAs in PD appear to be context-dependent and remain incompletely comprehended. Nevertheless, the majority of studies support the beneficial effects of SCFAs on microglial functions in the context of PD (Fig. 4a). Emerging evidence has revealed that epigenetic perturbation is an important contributor to PD, positioning it as a promising target for potential therapeutic interventions.^286,314 For example, HDAC5 inhibition is effective in attenuating microglial activation and PD-related pathologies in 6-hydroxydopamine (6-OHDA)-lesioned rats.³¹⁵ On the other hand, the activation of GPCRs has been demonstrated to exhibit neuroprotective properties in PD mouse models. Notably, the administration of probiotic Clostridium butyricum triggers the activation of colonic GPR41/43, resulting in the inhibition of microglial activation and mitigation of PD-related pathologies in MPTP-treated mice.³¹⁶ Moreover, the activation of GPR41 in enteric neurons using AR420626 has been shown to mimic the neuroprotective effects of propionate in improving motor functions and preventing dopaminergic neuronal loss in 6-OHDA-induced PD mice.³¹⁷

Boosting the levels of SCFAs with prebiotic intervention upregulates the neuroprotective phenotype of microglia in ASO mice, coupled with reduced motor deficits and α-synuclein aggregation in the SN.³¹⁸ Similarly, the administration of probiotic Bifidobacterium breve CCFM1067 restored the levels of SCFAs in MPTP-induced PD mice, resulting in reduced glial activation, oxidative stress, and motor impairments.²⁶⁰ Additionally, studies have demonstrated the neuroprotective effects of sodium butyrate in MPTP-induced PD mice by attenuating microglial reactivity via the inhibition of TLR4/MyD88/NF-κB and MAPK signaling pathway.^319–321 These findings are in concordance with a substantial body of literature that consistently reports diminished SCFAs-producing species and SCFA levels in prodromal stage of PD and PD patients compared to healthy controls, which may be correlated with the clinical severity of PD.^52,322–331 The observed reduction in SCFAs in PD has inspired a proof-of-concept study to investigate the potential of prebiotic fibers in PD patients [NCT04512599]. Importantly, the trial reported an increased abundance of SCFAs-producing species and increased plasma SCFA levels, along with improved intestinal barrier integrity and reduced intestinal inflammation.³³²

Nonetheless, discrepant findings on the influence of SCFAs on microglial functions in PD have been documented. Of note, SCFA supplementation in GF ASO mice induced microglial activation, α-synuclein accumulation and motor dysfunction.⁵⁵ Conversely, the reduction of SCFAs in MPTP-induced mice yielded beneficial effects, including alleviation of motor dysfunction, microglial and astrocytic activation in the SN.^217,333 The intriguing duality of SCFAs in PD warrants further exploration. We speculate that the activation of C/EBPβ/AEP signaling might serve as a plausible mechanism underlying the detrimental effects of SCFAs, as demonstrated in GF 3xTg-AD mice.¹⁹⁶ A recent study found that the C/EBPβ/AEP signaling is age-dependently activated in human α-synuclein transgenic mice and PD patients, which is responsible for mediating microglial activation and PD pathologies.³³⁴

In addition, growing evidence is shedding light on the distinct effects of different SCFAs on microglial functions. In particular, acetate has been found to restore microglial homeostasis in GF mice, while propionate and butyrate do not have the same restorative effect.¹⁹⁰ This is further complicated by a recent study which reported that both reduced fecal propionic acid and butyric acid following low-dose maslinic acid treatment, as well as increased fecal acetic acid following high-dose maslinic acid treatment, demonstrated neuroprotective effects against PD pathologies in MPTP-treated mice. However, only high-dose maslinic acid treatment reduced microglial activation and neuroinflammation, and the effects are speculated to be mediated by acetic acid.³³⁵ On the other hand, a comparison between acetate, propionate, and butyrate in primary microglia has reported an abundant overlap between butyrate and propionate in microglial transcriptomic profile. However, individual SCFA failed to achieve comparable effects as the combined SCFA treatment.¹⁹⁹ Thus, we recommend future studies to examine the effects of individual SCFA supplementation and combined SCFA supplementation on microglial function in PD mouse models. The beneficial effects of butyrate supplementation on PD pathologies might be attributed to its potent HDAC inhibitory activity and its ability to promote intestinal barrier and BBB integrity.^336–339

Amyotrophic lateral sclerosis

Growing evidence has linked epigenetic dysregulations to ALS.³⁴⁰ Notably, genetic knockdown and pharmacological inhibition of HDACs have been shown to ameliorate ALS pathology in different ALS models.^341–344 Furthermore, studies have identified reduced butyrate-producing species in ALS patients and SOD1^G93A mice.^{221,222,345,346}

Butyrate supplementation has demonstrated neuroprotective effects in SOD1^G93A mice.^222,347 The administration of butyrate has been shown to increase the abundance of butyrate-producing Butyrivibrio fibrisolvens, restore Paneth cell homeostasis and enhance the intestinal barrier integrity of SOD1^G93A mice. This is accompanied by reduced SOD1^G93A aggregation in the intestine, thereby slowing disease progression and prolonging the lifespan of SOD1^G93A mice.³⁴⁷ Moreover, butyrate treatment reduces SOD1^G93A aggregation and GFAP expression in the colon and lumbar spine of SOD1^G93A mice, resulting in improved enteric neuromuscular function.²²² Using motor neuron-like NSC34 cells with overexpression of hSOD1^G93A, it was found that butyrate improved mitochondrial bioenergetics by improving mitochondrial network and upregulating the transcription of peroxisome proliferator-activated receptor-gamma coactivator-1α (PGC1α).³⁴⁸ The PGC1α signaling is a master regulator of mitochondrial biogenesis and energy metabolism with significant therapeutic relevance in neurodegenerative diseases, including ALS.³⁴⁹

Alzheimer’s disease

Several studies have linked an altered BA profile to AD. Although studies in AD patients have yielded slightly heterogeneous results in BA composition, they have reported a consistent pattern of reduced levels of primary BAs and elevated levels of secondary BAs. Secondary BAs associated with AD and cognitive impairments include DCA, LCA, glycodeoxycholic acid, taurodeoxycholic acid, glycolithocholic acid, and taurolithocholic acid.^355–358 Moreover, an elevated level of secondary BA taurohyodeoxycholic acid was reported in the serum of GF 3xTg-AD mice receiving FMT from AD patients.¹⁹⁶ In another study, the administration of a traditional Chinese medicine decoction effectively ameliorated Aβ plaque pathology, neuroinflammation, and cognitive impairment in APP/PS1 mice. These beneficial effects were attributed to the suppression of gut dysbiosis and a reduction in the serum levels of secondary BAs, namely DCA and taurohyodeoxycholic acid.³⁵⁹

Nevertheless, it is noteworthy that certain secondary BAs serve significant biological functions, including the regulation of host metabolism, immunity, and resistance against intestinal pathogen expansion.^21,360–364 More importantly, several secondary BAs have demonstrated pronounced neuroprotective and anti-inflammatory activities, particularly ursodeoxycholic acid (UDCA) and its amidated conjugate, TUDCA (Fig. 3c). They exert anti-inflammatory effects by activating the nuclear receptor Farnesoid X receptor (FXR) and membrane receptor Takeda G-protein-coupled receptor 5 (TGR5), both of which are found in microglia and neurons.^365,366

The anti-inflammatory activity of TUDCA has been demonstrated in LPS-stimulated primary microglial cells and BV-2 microglial cells.^367–369 In addition, TUDCA is effective in attenuating microglial reactivity in LPS-treated mice.^367,370,371 Notably, TUDCA treatment has demonstrated the ability to inhibit Aβ deposition and glial activation in APP/PS1 mice and HFD-fed A7-Tg mice.^372–375

Moreover, BAs have the potential to modulate the gut–brain axis by maintaining intestinal homeostasis, as TGR5 and FXR signaling are pivotal regulators of intestinal immune response and barrier function.^365,376 Peripheral BAs provide signals to the CNS indirectly via TGR5-dependent glucagon-like peptide-1 (GLP-1) pathway and the FXR-dependent fibroblast growth factor 15 or 19 (FGF15/19) pathway.³⁵⁰ Currently, a phase II clinical trial is underway to examine the safety and biological activity of AMX0035, an oral combination of TUDCA and sodium phenylbutyrate, in the treatment of AD [NCT03533257].

Parkinson’s disease

The dysregulation of BA homeostasis has been implicated as a pivotal factor in the pathogenesis of PD. Studies have revealed elevated levels of secondary BAs in PD patients, while findings regarding primary BAs have yielded varying results.^377–380 An elevated level of primary BA (cholic acid) was identified in the plasma of PD patients.^379,380 Conversely, a reduction in glycine-conjugated primary BAs (glycocholic acid and glycochenodeoxycholic acid) was reported in post-mortem frontal cortex samples of PD patients, which was associated with the duration of PD diagnosis.³⁷⁷ However, another study reported increased levels of glycine- and taurine-conjugated primary BAs in plasma samples of PD patients.³⁷⁹ The discrepancy in BA profiles between plasma samples and brain samples warrants further investigation to understand the potential implications for PD pathogenesis. Notably, elevated levels of BAs were detected in the plasma samples of pre-PD patients, indicating that alterations in BA profile manifest many years before the onset of PD.³⁸¹ Moreover, PD patients exhibited reduced plasma levels of neuroprotective BAs, namely UDCA and TUDCA.³⁷⁹ This observation aligns with the findings obtained in prodromal PD mice.³⁸² Interestingly, a significant elevation in the risk of PD was observed among individuals who underwent cholecystectomy (removal of the gallbladder), a surgical procedure that has been associated with detrimental effects on both gut microbiota and BA composition.^383–385 It was reported that mice that received FMT from patients with post-cholecystectomy diarrhea exhibited increased levels of secondary BAs, specifically DCA, LCA, and hyodeoxycholic acid.³⁸³

Similar to AD, the administration of UDCA and TUDCA represents a promising therapeutic approach to counteract microglial activation and neuroinflammation in PD (Fig. 4b). Multiple in vivo studies have demonstrated the neuroprotective effects of UDCA and TUDCA in counteracting mitochondrial dysfunction, oxidative stress, and neuroinflammation within PD mouse models.^386–392 Moreover, the administration of a low-protein high-carbohydrate diet effectively mitigated PD pathologies in MPTP-treated mice by increasing the serum concentrations of TUDCA.³⁹³ These encouraging results have prompted two clinical trials, NCT03840005 (phase II) and NCT02967250 (phase I), which investigate the application of UDCA in PD patients. UDCA was found to be safe and well-tolerated in PD patients at a daily dose of 30 mg/kg.³⁹⁴

Aside from the direct effects of brain BAs, peripheral BAs provide signals to the CNS indirectly via the TGR5-dependent GLP-1 pathway and FXR-dependent FGF15/19 pathway.³⁵⁰ Notably, the modification of gut microbiota, BA metabolism and activation of intestinal TGR5 by dioscin treatment led to enhanced secretion of GLP-1 in both the intestine and brain of MPTP-treated mice. This is coupled with reduced glial activation and amelioration of motor deficits. Importantly, the co-administration of UDCA and dioscin further enhanced the neuroprotective effects of dioscin.³⁹⁰ GLP-1 is an incretin hormone that plays a pivotal role in stimulating insulin secretion and lowering blood glucose levels, making it highly relevant in the context of neurodegenerative diseases.³⁹⁵ The administration of GLP-1 receptor agonist inhibited microglial activation, prevented microglial-mediated conversion of astrocytes into the neurotoxic A1 phenotype, and effectively protected against α-synucleinopathy-induced neurodegeneration.³⁹⁶ Furthermore, the administration of probiotic Clostridium butyricum restored the levels of colonic GLP-1 and expression of cerebral GLP-1 receptor in MPTP-treated mice, leading to reduced microglial activation and alleviated motor deficits.³¹⁶

Amyotrophic lateral sclerosis

Alterations in BA profiles have been documented in ALS patients and SOD1^G93A mice, emphasizing the potential relevance of BAs to ALS pathology.^397,398 A significant reduction in primary BAs and neuroprotective BA (TUDCA) was recently reported in the fecal samples of ALS patients with cognitive impairment compared to those with normal cognition.³⁹⁸ The existing evidence strongly supports the neuroprotective effects of TUDCA in ALS. TUDCA has been found to confer protection against cyclopiazonic acid-induced degeneration in both mouse and human stem cell-derived hSOD1^G93A motor neurons. Furthermore, TUDCA enhances neuromuscular junction innervation in the tibialis anterior muscle of early-stage hSOD1^G93A mice.³⁹⁹

Notably, an oral combination of TUDCA and sodium phenylbutyrate (AMX0035) is effective in slowing functional decline and prolonging the overall survival of ALS patients.^400–402 The promising data from the phase II CENTAUR trial has recently led to its approval by the US FDA for the treatment of ALS.²¹⁹ The phase III PHOENIX trial of this combination is ongoing [NCT05021536]. In addition, a phase III clinical trial is underway to evaluate the safety and efficacy of TUDCA as an add-on treatment to riluzole in ALS patients [NCT03800524].

Alzheimer’s disease

Two recent metabolomic studies have reported elevated TMAO levels in the brains of aged mice.^406,407 In addition, the levels of TMAO in the plasma and brain of 18-month-old 3xTg-AD mice are markedly higher than those in 8-month-old mice.⁴⁰⁸ Importantly, elevated levels of TMAO have been observed in the plasma and cerebrospinal fluid (CSF) of individuals with MCI and AD.^409,410 Several studies have shown that TMAO has the ability to traverse the BBB and contribute to neurodegeneration by inducing microglial and astrocytic activation.^411–414

TMAO supplementation has been shown to induce cognitive impairment in APP/PS1 mice by promoting neuroinflammation, Aβ and tau pathology (Fig. 3b). Interestingly, the study also found that TMAO supplementation disrupted the integrity of the intestinal barrier and BBB.⁴¹⁴ Moreover, TMAO induces inflammation and aggravates Aβ and tau pathology in D-galactose/AlCl₃-induced AD mice by activating the PI3K/AKT/mTOR signaling pathway.⁴¹⁵ Conversely, the reduction of TMAO following the administration of 3,3-dimethyl-1-butanol alleviated cognitive impairment of APP/PS1 mice by attenuating Aβ pathology and neuroinflammation.⁴¹⁶ The administration of probiotic Lactobacillus plantarum is effective in remodeling the gut microbiota and reducing TMAO levels of APP/PS1 mice, resulting in amelioration of neuroinflammation and neurodegeneration.⁴¹⁷ In addition, physical exercise has been shown to reduce the serum concentrations of TMAO, TMA, and betaine in APP/PS1 mice, conferring protection against neuroinflammation and AD pathologies.⁴¹⁴

Parkinson’s disease

An evident increase in TMAO synthesis was detected in the post-mortem frontal cortex of PD patients with dementia, in comparison to patients with normal cognitive function and MCI.³⁷⁷ Moreover, elevated plasma TMAO levels in PD patients are correlated with disease severity and motor symptom progression. Patients with higher baseline TMAO levels are at higher risk of experiencing deterioration in motor symptoms and cognitive function.⁴¹⁸ However, contradictory findings have been reported regarding the roles of TMAO in PD. A study on drug-naïve early-stage PD patients found that the plasma TMAO levels were lower in PD patients compared to healthy controls. Moreover, patients with lower plasma TMAO levels are associated with a higher rate of increase in levodopa-equivalent dose, as well as a higher risk of dementia conversion.⁴¹⁹ Further research is needed to clarify the contribution of TMAO in the pathogenesis of PD.

Nevertheless, the existing evidence from PD models has indicated potential detrimental effects of TMAO (Fig. 4c). TMAO pre-treatment has been shown to worsen MPTP-induced microglial and astrocytic activation in the striatum, SN, and hippocampus of MPTP-treated mice.^412,413 Using midbrain organoid models, it was found that TMAO induced pathological changes similar to PD, including loss of dopaminergic neurons, astrocytic activation, and phosphorylation of α-synuclein.⁴²⁰

Alzheimer’s disease

An altered composition of indole-producing bacteria was observed in APP/PS1 mice, concomitant with compromised intestinal barrier integrity and cognitive impairment. The administration of an indole mixture (indole, indole-3-acetic acid, and indole-3-propionic acid) attenuated the microglial reactivity and neuroinflammation by activating microglial AHR signaling and inhibiting NLRP3 inflammasomes. Consequently, the treatment led to reduced Aβ deposition, diminished tau hyperphosphorylation and improved cognitive function (Fig. 3d).⁴²⁶ Moreover, the administration of high-tryptophan diet proved effective in alleviating gut dysbiosis and inhibiting microglial reactivity in APP/PS1 mice by activating central AHR signaling and inhibiting NF-κB signaling.⁴²⁷

On the other hand, a significantly higher concentration of indoxyl-3-sulfate was reported in the brain and serum of GF 3xTg-AD mice receiving FMT from AD patients, when compared to their counterparts receiving healthy control microbiota.¹⁹⁶ Indoxyl-3-sulfate is a uremic toxin adversely related to several mental and neurological conditions, including anxiety, anorexia nervosa, AD, and PD.^{377,428–431} An in vitro study revealed that it induces oxidative stress and inflammation in primary astrocytes and mixed glial cell cultures.⁴³² Further studies are needed to comprehensively understand the roles of different indole derivatives in the pathogenesis of AD.

Parkinson’s disease

A recent metagenomic analysis of the gut microbiome of PD patients has revealed a reduction in the tryptophan biosynthesis pathway.³³¹ Dietary supplementation of tryptophan has been shown to alleviate inflammation and motor deficits of rotenone‐induced PD rats by inhibiting NF‐κB activation, but the neuroprotective effects were blocked by an AHR inhibitor.⁴³³ In addition, studies have demonstrated the neuroprotective effects of indole-3 carbinol in rotenone-induced PD rats and LPS-treated rats (Fig. 4d).^434,435 Notably, indole-3-carbinol effectively prevents rotenone-induced α-synuclein accumulation, astrocytic activation, neuroinflammation, and motor dysfunction by activating the SIRT1/AMP-activated protein kinase (AMPK) signaling pathway.⁴³⁴ Moreover, indole-3-carbinol is effective in counteracting LPS-induced oxidative stress and neuroinflammation.⁴³⁵

On the other hand, the uremic toxin indoxyl sulfate is increased in PD patients.^377,429 An elevated CSF/plasma ratio of indoxyl sulfate was detected in PD patients, with notably higher concentrations observed in patients experiencing motor fluctuations compared to those without motor fluctuations.⁴²⁹ Furthermore, a marked upregulation of indoxyl sulfate was identified in the post-mortem frontal cortex samples of PD patients with dementia, compared to patients with normal cognitive function and MCI.³⁷⁷

Alzheimer’s disease

A recent lipidomic analysis of the brains of AD patients has revealed a significant reduction in the ratio of omega-3 to omega-6 fatty acids compared to age-matched cognitively normal individuals. In addition, the study found that omega-3 fatty acids are positively correlated with cognitive function and negatively correlated with Aβ, neurofibrillary tangles burden, and Braak stage. Conversely, the pro-inflammatory metabolites of omega-6 fatty acids are positively correlated with AD pathologies (Fig. 3e).⁴³⁸ Indeed, a higher dietary intake of omega-3 fatty acids is associated with a reduced risk of dementia and cognitive decline.⁴³⁹ The beneficial effects of omega-3 fatty acids may be attributed to their ability to stimulate microglial phagocytosis of Aβ₄₂ and promote the transition of microglia to an anti-inflammatory phenotype, as demonstrated in human CHME-3 microglial cells.⁴⁴⁰

A recent study reported that GF 3xTg-AD mice exhibited reduced genes involved in AA metabolism, along with decreased levels of AA-associated inflammatory enzymes, in comparison to their SPF counterparts. However, FMT from AD patients to GF 3xTg-AD mice remarkably increased the relative abundance of Bacteroides strains involved in PUFA metabolism, resulting in increased concentrations of AA and its metabolites (leukotriene B4, prostaglandin E2, and 12-hydroxyheptadecatrienoic acid). Consequently, the elevated levels of AA-associated metabolites induced microglial activation and the development of AD pathologies in AD-humanized ex-GF 3xTg-AD mice by activating the C/EBPβ/AEP signaling pathway.¹⁹⁶ Similarly in another study, FMT from AD patients to Thy1-C/EBPβ transgenic mice led to increased levels of prostaglandin E2 and 12-hydroxyheptadecatrienoic acid, resulting in C/EBPβ/AEP pathway activation, microglial activation, and AD pathologies.⁴⁴¹ Indeed, elevated levels of leukotriene B4 and prostaglandin E2 were recently detected in the post-mortem brains of AD patients.⁴³⁸ Interestingly, the co-administration of glyceryl-conjugated prostaglandin E2 and SCFAs additively induced microglial activation in GF 3xTg-AD mice.¹⁹⁶

Amyotrophic lateral sclerosis

PUFAs and their bioactive derivatives are vital regulators of neuronal function and neuroinflammation,⁴³⁶ making them highly relevant in the context of ALS. This is evident in two recent metabolomic analyses which revealed significant dysregulation of lipid metabolism, including PUFAs, among ALS patients compared to healthy controls.^442,443 Notably, a higher plasma level of omega-3 fatty acid α-linolenic acid (ALA) was associated with slower functional decline and longer survival. In addition, higher plasma levels of omega-3 fatty acid eicosapentaenoic acid (EPA) and omega-6 fatty acid linoleic acid (LA) were associated with a lower risk of death, although they did not impact the rate of functional decline.⁴⁴⁴ The protective effects of ALA on ALS are similarly supported by another study, wherein males with higher pre-diagnostic plasma ALA levels are at a lower subsequent risk of developing ALS.⁴⁴⁵ These results are supported by a prospective study that observed a markedly reduced risk of ALS in individuals with higher dietary intakes of omega-3 fatty acids, including ALA and marine omega-3 fatty acids.⁴⁴⁶ On the other hand, an increased risk of ALS is observed in males with higher pre-diagnostic plasma levels of omega-3 fatty acid DHA, and in females with higher pre-diagnostic plasma levels of omega-6 fatty acid AA.⁴⁴⁵

Perturbation in lipid-related metabolism pathways, including AA metabolism, has been identified as a common metabolic signature in a multi-omics analysis of spinal motor neurons (sMNs) derived from four ALS hiPSC lines of various genetic background (SOD1^A4V, C9ORF72, TDP-43^Q343R, and sporadic).⁴⁴⁷ Importantly, the reduction of AA levels using 5-lipoxygenase inhibitor (caffeic acid) is effective in rescuing the reduced percentage of HB9::GFP⁺ cells (a marker of motor neurons) and increased percentage of 7AAD⁺ cells (a marker of cell death) in C9ORF72 and SOD1^A4V ALS lines. Moreover, caffeic acid treatment is effective in delaying disease onset and prolonging the lifespan of SOD1^G93A mice. These beneficial outcomes are attributed to the attenuation of microglial and astrocytic activation, and an increase in the number of motor neurons in the spinal cord.⁴⁴⁷

Parkinson’s disease

Reduced plasma levels of BCAAs and aromatic amino acids were detected in PD patients, which were significantly correlated with disease severity.^452,453 This may be related to the reduction in Prevotella copri in PD patients,^331,454 which is the species responsible for BCAA biosynthesis.⁴⁴⁸ A recent study observed reduced serum BCAA levels in rotenone-induced PD mice, accompanied by intestinal dysfunctions, motor deficits, α-synuclein accumulation, and loss of dopaminergic neurons in SN. Conversely, the administration of a high BCAA diet reversed the pathological changes induced by rotenone, along with reduced levels of pro-inflammatory cytokines in the colon and SN.⁴⁵⁵ Furthermore, administration of probiotic L. plantarum CCFM405 restored the biosynthesis of BCAAs, restored intestinal barrier integrity, and inhibited microglial and astrocytic activation in rotenone-induced PD mice.⁴⁵⁶

However, we recommend caution against excessive elevation of BCAAs as BCAAs imbalance is associated with detrimental health consequences, including hyperphagia, metabolic disorders, cardiovascular diseases, and reduced lifespan.^450,457 Another factor warranting consideration is the potential competition between amino acids and levodopa, a commonly prescribed antiparkinsonian drug, in utilizing the intestinal and BBB transporters and consequent interference with levodopa pharmacokinetics.^458,459

Amyotrophic lateral sclerosis

Neurodegenerative diseases, including ALS, are associated with disrupted NAD⁺ homeostasis and NAD⁺ depletion.⁴⁶¹ A marked decrease in the levels of NAM has been observed in both the serum and CSF of ALS patients compared to healthy controls. The serum levels of NAM are positively associated with the ALS functional status score.²²³

A notable difference in the metagenomic NAM biosynthetic pathway has been reported between SOD1^G93A mice and WT mice. The altered NAM biosynthesis was linked to A. muciniphila, which showed a gradual reduction in disease progression in SOD1^G93A mice, but not in WT mice. Notably, the administration of A. muciniphila significantly improved the motor function and prolonged the lifespan of SOD1^G93A mice. These beneficial effects of A. muciniphila were attributed to the enhanced NAM biosynthesis, which resulted in increased levels of NAM in the serum and CSF of SOD1^G93A mice.²²³ In addition, NAM treatment has been shown to improve the survival of sMNs derived from both sporadic and familial ALS hiPSCs, as well as isogenic iPSC lines harboring SOD1^L144F and TDP-43^G298S mutations, by rescuing the mitochondrial respiration defects.⁴⁶³

Alzheimer’s disease

The dysregulation of the serotonergic system represents a complex pathological process in AD due to its involvement in multiple AD pathologies, including APP processing and Aβ deposition.^471–473 A recent study reported lower brain serotonin transporter availability and higher cortical Aβ deposition in individuals with MCI compared to healthy controls.⁴⁷⁴ Furthermore, significantly reduced urine and serum serotonin concentrations have been reported in AD patients compared to controls.⁴⁷⁵ Interestingly, the administration of selective serotonin reuptake inhibitors (SSRIs) is effective in suppressing Aβ levels in human and AD mouse models.^471,476,477

Differences in genes related to the serotonergic system have been reported between GF and SPF 3xTg-AD mice, which suggests higher hippocampal serotonin signaling in the GF mice.¹⁹⁶ This study corroborates the involvement of gut microbiome in regulating serotonergic system in AD. However, there are currently sparse studies investigating the potential of gut microbiota manipulation to restore the serotonergic system in AD, and the available studies did not consistently show improvements in the serotonergic system. The co-administration of prebiotics fructo-oligosaccharides (FOS) and galacto-oligosaccharides (GOS) increased the concentration of serotonin in the brains of APP/PS1 mice.⁴⁷⁸ However, another study did not observe any improvement in the serotonergic system following prebiotic Sparassis crispa-1 polysaccharide administration in d-galactose/AlCl3-induced AD mice.⁴⁷⁹ Nevertheless, the available evidence suggests that probiotics, prebiotics and FMT hold immense potential in AD, as they have shown promising results in modulating the serotonergic system in different neurological disorders.^34,480–483

Parkinson’s disease

Beyond the established dopaminergic dysfunction characterizing PD, growing evidence implicates serotonergic dysfunction in the progression of PD, which is associated with disease burden and may precede motor manifestations or dopaminergic dysfunction.^484,485 Furthermore, the development of neuropsychiatric symptoms in PD patients, including apathy, anxiety, and depression, is associated with serotonergic dysfunction.^486,487 In addition, studies have demonstrated the modulation of the serotonergic system by α-synucleinopathy. For instance, the overexpression of human α-synuclein in serotonin receptors in raphe nuclei in mice has been found to impair forebrain serotonin neurotransmission and trigger a depressive-like phenotype.⁴⁸⁸ Conversely, the serotonergic neurotransmission deficits were alleviated following antisense oligonucleotide-induced α-synuclein knockdown.^488,489

A recent metagenomic analysis of the gut microbiome of PD patients has revealed a dysregulation in the synthesis and metabolism of multiple neurotransmitters, including serotonin. This dysregulation may be attributed to the observed reduction in the tryptophan biosynthesis pathway and sporulation genes.³³¹ Additionally, a recent study found that healthy mice, upon receiving FMT from PD mice, exhibited impaired motor function and reduced striatal dopamine and serotonin levels.²¹⁷ These findings underscore the potential significance of gut microbiota manipulation to restore the serotonergic system in PD. Notably, prebiotic polymannuronic acid administration led to the elevation of serotonin and its metabolite, 5-hydroxyindoleacetic acid (5-HIAA), in the striatum of MPTP-induced mice.⁴⁹⁰ Moreover, FMT from control mice mitigated MPTP-induced gut dysbiosis and the decline of striatal serotonin and 5-HIAA in recipient mice.²¹⁷ Conversely, recent studies investigating the impact of probiotics on the serotonergic system in PD have shown less favorable outcomes. In a recent randomized controlled trial (RCT), no significant changes in serum serotonin levels were observed following a 3-month adjunctive probiotic Bifidobacterium animalis subsp. lactis treatment in PD patients receiving conventional regimen (Benserazide and dopamine agonists).⁴⁹¹ Furthermore, the administration of L. plantarum PS128 was ineffective in restoring MPTP-induced serotonin reduction.⁴⁹² Future studies may explore the potential benefits of multi-strain probiotic formulations for achieving the desired serotonergic effects in PD. These studies should emphasize on the importance of informed strain selection, including the potential synergy or additive effects of individual probiotic strains.⁴⁹³

Alzheimer’s disease

The AD pathologies, including Aβ and tau pathology, have been shown to induce GABAergic dysfunction and contribute to excitatory and inhibitory (E/I) imbalance.^501–503 An abnormal increase in tonic inhibition has been reported in APP/PS1 mice and 5xFAD mice due to aberrant release of GABA by reactive astrocytes, leading to impaired synaptic plasticity and memory.^504,505 However, recent evidence suggests that GABAergic hypoactivation potentiates neurodegeneration and cognitive impairment in AD patients.^502,506 Further research is warranted to elucidate the observed discrepancies in the role of GABAergic system within AD pathogenesis, including the involvement of astrocytes.

A recent study has reported variations in the expression levels of genes related to GABAergic system between GF and SPF 3xTg-AD mice, suggesting an increase in hippocampal GABA production and GABA receptors in the GF counterparts.¹⁹⁶ These findings suggest a potential avenue for manipulating the gut microbiota to address GABAergic dysfunction in AD. In an initial investigation with limited sample size, it was reported that FMT from an AD patient resulted in a reduction of GABA levels in the fecal metabolites of recipient mice.⁵⁰⁷ Conversely, yFMT effectively restored hippocampal GABA levels in aged recipient mice towards the levels observed in young mice, coinciding with improvements in cognitive behavior.²¹⁰ Moreover, the administration of prebiotic Sparassis crispa-1 polysaccharide increased the hippocampal levels of GABA in D-galactose/AlCl₃-induced AD mice, resulting in improved cognitive function.⁴⁷⁹ Future studies employing larger sample sizes and AD mouse models could provide valuable insights into the potential therapeutic implications of modulating the gut microbiome to target the GABAergic system in AD.

Parkinson’s disease

Growing evidence has reported alterations in the GABAergic system in PD, which are associated with motor symptoms, psychomotor symptoms, axial symptoms, cognitive impairment, olfactory dysfunction, and visual hallucinations.^508–510 An alteration in the synthesis and metabolism of GABA was recently reported in a metagenomic analysis of the PD gut microbiome.³³¹ Recent studies have explored the potential therapeutic role of probiotics and prebiotics in addressing these underlying pathologies. The administration of probiotic Pediococcus pentosaceus and prebiotic polymannuronic acid has shown promise in alleviating gut dysbiosis and reducing GABA levels in the brains of MPTP-induced mice, resulting in the amelioration of motor dysfunction and neuronal degeneration.^490,511 A recent RCT found that the co-administration of probiotic and conventional regimen in PD patients increased the abundance of species-level genome bins (SGBs) involved in GABA synthesis and reduced the abundance of SGBs encoding GABA degradation, as compared to PD patients receiving conventional regimen alone.⁴⁹¹

Parkinson’s disease

Aside from the aggregation of misfolded α-synuclein in the intracellular Lewy bodies, a major neuropathological hallmark of PD is the loss of dopaminergic neurons within the SN.⁵¹⁹ Dopaminergic deficiency underlies the cardinal motor features of PD, including bradykinesia, rigidity, resting tremor, and postural instability.⁵²⁰ Thus, the mainstay pharmacological treatments for the management of PD motor symptoms are predominantly dopamine-related interventions, including levodopa preparations, dopamine agonists, and monoamine oxidase-B inhibitors.⁵¹⁹

The gut microbiome influences multiple aspects of dopaminergic system within PD pathophysiology, as well as the efficacy of dopamine-related drugs.^{518,521–523} This influence is exemplified by a study in which FMT from PD mice reduced striatal levels of dopamine and its metabolites in recipient mice, accompanied by impaired motor function.²¹⁷ Moreover, metagenomic analysis of the gut microbiome in PD patients has detected a reduced capacity to generate dopamine precursors, particularly tyrosine.³³¹ Several gut microbes have shown detrimental effects on the dopaminergic system in PD mouse models.²¹⁵ For instance, the administration of Proteus mirabilis during the premotor phase of MPTP-treated mice aggravated striatal dopaminergic neuronal damage and motor symptoms.⁵²⁴

A substantial body of literature has demonstrated the potential of microbiome-based therapeutics in alleviating dopaminergic damage and motor deficits in PD mouse models.^{217,261,321,393,492,525,526} Adjunctive probiotic administration was found to increase serum dopamine levels and improve motor function in PD patients receiving conventional regimen, compared to patients receiving conventional regimen alone.⁴⁹¹ Moreover, the administration of berberine has been shown to stimulate the production of dopa and dopamine in the gut microbiota by activating tyrosine hydroxylase and dopa decarboxylase. Dopa, in turn, crosses the BBB and is converted into dopamine. Consequently, the increased levels of dopamine in the brain protected against MPTP-induced motor deficits.³⁶

Alzheimer’s disease

The significance of cholinergic signaling in AD was long recognized.⁵³¹ A hallmark of cognitive impairment is the loss of basal forebrain cholinergic neurons, which is driven by APP processing, Aβ deposition, tau hyperphosphorylation and dysregulated immune response.⁵²⁷ At present, there is a limited number of studies examining the potential of microbiome-based therapeutics for modulating the cholinergic system in AD, but the evidence available supports their beneficial effects on the cholinergic system. Notably, the administration of probiotic L plantarum MTCC1325 restored the levels of acetylcholine in the brains of D-galactose-induced AD rats by reducing AChE activity.⁵³² Similarly, another study found that prebiotic FOS from Morinda officinalis was effective in restoring acetylcholine in the brains of d-galactose-induced AD rats by reducing AChE levels.⁵³³ Further investigation into the neuroprotective potential of microbiome-based therapeutics in the cholinergic system in other AD animal models is warranted.

Parkinson’s disease

Aberrant glutamatergic neurotransmission represents a key contributing factor to neurodegeneration and PD.^377,510,539 However, our understanding of the influence of the gut microbiome on glutamate in PD remains limited, and the studies available primarily examine glutamate within the gut microbiome. Although PD is generally associated with glutamatergic hyperactivity, a recent metagenomic analysis of PD gut microbiome revealed a reduction in glutamate/glutamine synthesis genes and pathways, along with an increase in glutamate degradation pathway.³³¹ These findings are in agreement with the results of an earlier meta-analysis of the PD gut microbiome.⁵⁴⁰ However, it is noteworthy that glutamate does not readily cross the BBB, which may limit its direct influence on brain glutamate levels.^541,542 On the contrary, several transporters for glutamine, the direct precursor of glutamate, have been identified.⁵⁴³ Notably, the combined administration of probiotics with conventional regimen led to a reduction in serum glutamine concentrations in PD patients compared to the placebo group receiving the conventional regimen only.⁴⁹¹ Further studies are warranted to examine the potential and underlying mechanisms of microbiome-based therapeutics in modulating glutamatergic signaling in PD.

Alzheimer’s disease

Accumulating evidence has indicated dysfunctional ghrelin signaling in AD. It was found that Aβ binds to and inhibits hippocampal GHSR1α activity in 5xFAD mice, resulting in synaptic deficits and cognitive decline.⁵⁵⁴ In addition, plasma ghrelin levels reduce gradually in 3xTg-AD mice with aging, which is reversed by the administration of multi-strain probiotics formulation (SLAB51), leading to the attenuation of cognitive impairment.⁵⁵⁵ A recent study in the elderly reported an age-dependent reduction in plasma ghrelin and an elevated ratio of plasma liver-expressed antimicrobial peptide 2 (LEAP2)/ghrelin, which is associated with cognitive impairment. Subsequently, the study found that reducing the plasma LEAP2/ghrelin ratio reversed cognitive deficits in aged mice by restoring hippocampal neurogenesis, alleviating synaptic loss, and inhibiting neuroinflammation.⁵⁵¹ Moreover, the activation of ghrelin signaling has been shown to attenuate microglial activation in mouse models of normal aging, accelerated aging and AD.^556,557 The neuroprotective effects of ghrelin might be attributed to its ability to impact age-related pathways, including sirtuin-1 activation, with studies demonstrating its ability to prolong lifespan in different mouse models of aging.^556,558

Parkinson’s disease

The ghrelin receptors are expressed in various brain regions, including the SN. Ghrelin participates in the regulation of motor functions by modulating the dopaminergic neuronal excitability within the SN.^559,560 Additionally, ghrelin promotes the differentiation of midbrain neural stem cells into dopaminergic neurons by activating the Wnt/β-catenin pathway.⁵⁶¹ Conversely, the inhibition of central GHSR1α activation via intra-SN administration of a selective GHSR1α inhibitor resulted in impaired motor coordination.⁵⁶² Evidence from PD patients has identified reduced plasma ghrelin levels in the fasting state,^563,564 as well as an impaired ghrelin response and secretion in the postprandial state.^564,565 These findings have provided a pivotal basis for further investigations into the neuroprotective effects of ghrelin in PD. Indeed, accumulating evidence has demonstrated the neuroprotective effects of ghrelin in multiple rodent models of PD, including MPTP-induced PD mice,^{560,566–568} 6-OHDA-induced PD rats,^569,570 and A53T α-synuclein transgenic mice.⁵⁷¹ These beneficial effects are attributed to its ability to promote autophagy, enhance mitochondrial biogenesis and bioenergetics, mitigate neuroinflammation, and inhibit apoptosis.^{560,566–569}

Alzheimer’s disease

Leptin is generally considered beneficial for cognitive function, as studies have consistently shown that individuals with higher leptin levels are associated with a reduced risk of dementia and AD.^585–587 This is supported by a recent meta-analysis that has reported reduced blood leptin levels in AD patients compared to cognitively normal individuals.⁵⁸⁷ Moreover, leptin is involved in counteracting Aβ pathology and tau phosphorylation to prevent synaptic dysfunction.^588,589 This is evident in a recent study that has revealed a robust association between low plasma leptin and increased CSF Aβ concentration in patients with cognitive impairment, which indicates increased brain amyloid deposition.⁵⁹⁰ Notably, the elevation of plasma leptin levels via chronic administration of multi-strain probiotics has been shown to ameliorate Aβ pathology in 3xTg-AD mice.⁵⁵⁵

Parkinson’s disease

Although leptin primarily acts on neurons in the hypothalamus, leptin signaling also has an important role in preserving the dopaminergic system.^591–593 Leptin receptors are widely expressed in several extra-hypothalamic regions, including dopaminergic neurons in SN.^594,595 However, our understanding of the role of leptin in PD remains incomplete, with limited evidence from clinical studies. Notably, a recent meta-analysis, including only 198 PD patients and 182 controls, found that the serum leptin levels in PD patients were slightly lower compared to those in the control group, although the difference did not reach statistical significance.⁵⁹⁶ These limited findings emphasize the need for more comprehensive research to understand the mechanisms and functions of leptin in the onset, progression, and management of PD.

Alzheimer’s disease

Diabetes is a well-established risk factor for cognitive impairment and dementia, with dementia now among the leading causes of death in the diabetic population.^602–604 Given these connections, there is a growing interest in repurposing the anti-diabetic class of drug GLP-1R agonists for AD.⁶⁰⁵ Substantial preclinical evidence supports the neuroprotective effects of GLP-1R agonists on cognition in AD, which is attributed to the ability to reduce Aβ pathology, tau phosphorylation, glial activation, and neuroinflammation.^598,606 Notably, GLP-1R agonist, NLY01, inhibits Aβ-induced microglial activation via microglial GLP-1R, resulting in reduced reactive astrocyte conversion in both 5xFAD and 3xTg-AD mice.⁶⁰⁷ In addition, the neuroprotective effects of GLP-1R agonists were substantiated by clinical studies that reported a reduction in cognitive impairment and lower dementia incidence among type 2 diabetes patients.^608,609 These encouraging results have prompted two phase III RCTs, NCT04777396 and