Microbial Metabolomes in Alzheimer’s Disease: From Pathogenesis to Therapeutic Potential

Acknowledging the connection between the GM and AD, recent studies have focused on characterizing this relationship [25,34]. The GM profiles of individuals with AD differ markedly from those of healthy controls, particularly in terms of microbial diversity and the relative abundance of specific bacterial taxa, which vary according to geographic region and the clinical stage of the disease. The current literature consistently reports a reduction in bacterial diversity in AD, as well as notable alterations at the phylum, family, and genus levels [32,35,36,37,38,39,40,41,42,43,44,45]. While the majority of studies report converging findings, discrepancies have been observed in certain bacterial families, such as Bacteroidaceae, Rikenellaceae, and Ruminococcaceae, which yielded conflicting results across some investigations [38,43,45]. Similar inconsistencies are noted at the genus level for Alistipes, Alloprevotella, Barnesiella, Bifidobacterium, Blautia, Coprococcus, Eubacterium, Faecalibacterium, and Ruminococcus [32,38,39,41,43]. Table 1 presents significant differences in the prevalence of GM bacterial taxa between AD patients and cognitively healthy individuals.

Despite considerable inter- and intra-individual variability influenced by factors such as sex, diet, and geographic region, studies have shown that approximately 85% of AD patients exhibit a distinct GM profile compared to age-matched healthy controls [32,46]. It has been stated that GM dysbiosis observed in AD patients may critically impact the MGB axis and contribute actively to AD pathogenesis [47]. Several clinical investigations have explored associations between specific gut microbial taxa and AD biomarkers or disease progression. For instance, Cattaneo et al. [35] reported an increased abundance of the pro-inflammatory genera Escherichia/Shigella and a decreased abundance of the anti-inflammatory species Eubacterium rectale, findings that correlated with elevated peripheral inflammation markers. They found significant positive correlations between Escherichia/Shigella abundance and pro-inflammatory cytokines IL-1β, NLRP3, and CXCL2, whereas E. rectale exhibited negative correlations with these cytokines. Similarly, Vogt et al. [43] documented reduced microbial diversity and altered GM composition in AD patients compared with controls, including decreases in the phyla Actinomycetota and Bacillota, as well as lower abundances of families such as Clostridiaceae, Mogibacteriaceae, Peptostreptococcaceae, Ruminococcaceae, and Turicibacteraceae, and genera including Adlercreutzia, Bifidobacterium, Bilophila, Clostridium, Dialister, and Turicibacter. Conversely, they observed increased abundance of taxa within Bacillota phylum and in families and genera such as Bacteroidaceae, Gemellaceae, Rikenellaceae, Alistipes, Bacteroides, Blautia, Gemella, and Phascolarctobacterium. Zhuang et al. [45] reported altered fecal microbiota composition in AD characterized by decreased Actinomycetota and increased Bacteroidota phyla. Increases were also noted in families Enterococcaceae, Lactobacillaceae, and Ruminococcaceae, as well as in genera Bacteroides, Ruminococcus, and Subdoligranulum, while the abundance of Bacteroidaceae, Lachnoclostridium, Lachnospiraceae, and Veillonellaceae decreased. Ling et al. [40] observed a marked reduction in bacterial diversity in Chinese AD patients, with decreased abundance of butyrate-producing genera such as Butyricicoccus, Coprococcus, Faecalibacterium, Gemmiger, and Roseburia, alongside increased levels of the propionate-producing Akkermansia and of the lactate-producing Bifidobacterium. In addition, genera including Clostridium, Dialister, and Romboutsia were diminished in AD. Hung et al. [38], in a systematic review encompassing 11 studies, reported that patients with AD, but not those with mild cognitive impairment (MCI), exhibited significantly lower GM diversity compared with healthy controls, and noted geographical variation in microbial profiles. The Pseudomonadota phylum and genera Bifidobacterium and Phascolarctobacterium were relatively predominant, whereas Bacillota phylum members and families such as Clostridiaceae, Lachnospiraceae, and Rikenellaceae were relatively less abundant. More recently, Heravi et al. [37] highlighted altered GM signatures in AD patients compared with healthy controls, characterized by elevated proportions of phyla Acidobacteriota and Actinomycetota, as well as family Ruminococcaceae and genus Bacteroides. Conversely, reductions were observed in Bacillota phylum, families Acidaminococcaceae and Lachnospiraceae, and genus Ruminiclostridium.

Nevertheless, some studies have reported conflicting findings, showing no significant alterations in the GM of AD patients compared to controls. For instance, Cirstea et al. [48] found that although the GM diversity was reduced in AD patients, the overall microbial composition did not differ significantly from that of the control group. Furthermore, elevated levels of the phylum Bacteroidota have been observed both in cognitively normal controls with Aβ-positive plasma [49], and in AD patients [36,43,45]. Additionally, as highlighted by Jemimah et al. [50], the abundance of genera Bacteroides and Phascolarctobacterium in AD patients appears to be modulated by geographic factors, with diet and lifestyle playing a pivotal role in these variations.

Several clinical studies have investigated the association between specific gut microbial taxa and clinical biomarkers of AD progression. For instance, Cattaneo et al. [35] identified a positive correlation between increased levels of Escherichia/Shigella and markers of peripheral inflammation as well as brain amyloidosis. Vogt et al. [43] reported associations between certain gut genera and cerebrospinal fluid (CSF) biomarkers of AD, including Aβ42/Aβ40 ratios, chitinase-3-like protein 1, phosphorylated tau (p-tau), and the p-tau/Aβ ratio. Specifically, they found that elevated levels of Bacteroides and Blautia correlated positively with p-tau and the p-tau/Aβ42 ratio. On the other hand, Sheng et al. [49] observed a negative correlation between plasma Aβ42 and Aβ42/Aβ40 ratios and the abundance of the family Desulfovibrionaceae and the genera Bilophila and Faecalibacterium, suggesting that higher levels of these taxa may be linked to reduced amyloid deposition in the brain. In addition, postmortem analyses of AD brain tissue have revealed co-localization of lipopolysaccharides (LPS) and Escherichia coli with amyloid plaques, implying a potential mechanistic link between GM components and amyloid pathogenesis [39]. This evidence supports the hypothesis that amyloid pathology in AD may be initiated during MCI as a consequence of shifts in the GM. Nevertheless, the causal involvement of these bacterial components in AD development, as well as the extent of their presence, remains a matter of debate within the scientific community. Several factors contribute to this controversy: (i) a causal relationship has not been definitively established; (ii) LPS-induced animal models of AD have inherent limitations, as they may not fully capture the complex pathophysiological mechanisms of human neurodegenerative diseases; and (iii) the co-localization of bacterial components with plaques demonstrates a correlation, but does not provide conclusive evidence of direct causation. Additional hypothesis might be considered. In this context, antibodies may be used to detect Aβ deposits in AD. However, antibody specificity is crucial for accurately identifying plaques and distinguishing them from artifacts. Misfolded proteins, for instance, may be erroneously interpreted as plaques, while low specificity can result in false positives or negatives. Consequently, the use of highly specific antibodies targeting modified or deposited forms of Aβ, rather than general Aβ antibodies, is essential for precise diagnosis and for elucidating the role of Aβ in AD pathogenesis.

Additional research has focused on delineating differences in GM profiles between MCI and AD, in order to clarify their relationship with disease progression. In a U.S. longitudinal cohort study of 108 nursing home elders (ages 65–94) with a 5-month follow-up, where shotgun metagenomics (NextSeq 500) was used, Haran et al. [36] found that GM composition varies by dementia subtype. AD patients showed increased abundance of Alistipes, Bacteroides, Barnesiella, and Odoribacter, as well as decreased abundance of Lachnoclostridium. Other dementia types were characterized by increased Barnesiella and Odoribacter, but decreased Collinsella, Eubacterium, Lachnoclostridium, and Roseburia. In China, a prospective cross-sectional study including 97 subjects (33 with AD, 32 with amnestic MCI, and 32 healthy controls, aged 50–85 years), where 16S rRNA sequencing was used, Liu et al. [41] reported a reduction in fecal microbial diversity in AD patients compared to both MCI and control groups, with significant compositional differences. The families Clostridiaceae and Lachnospiraceae were markedly decreased in both AD and MCI groups relative to controls, while Blautia, Ruminococcus, and the family Ruminococcaceae showed further decline in AD versus MCI. Moreover, a progressive increase in Enterobacteriaceae abundance was noted from controls to AD patients, whereas Veillonellaceae abundance increased from MCI to controls. Li et al. [39], in a cross-sectional study conducted in China including 90 subjects (30 with AD, 30 with MCI, and 30 healthy controls, with a mean age of 65.5 years), using 16S rRNA gene sequencing, also demonstrated that microbial diversity was diminished in both AD and MCI patients, with no significant differences between these two groups. However, they identified 11 genera that differed between AD patients and healthy controls across fecal and blood samples. Genera such as Bifidobacterium, Blautia, Dorea, Escherichia, Lactobacillus, and Streptococcus were increased in AD, while genera such as Alistipes, Bacteroides, Parabacteroides, Paraprevotella, and Sutterella were decreased.

Complementary studies have explored the relationship between variations in GM profiles and factors such as disease severity [51,52], the presence of neuropsychiatric symptoms [53], and an increased likelihood of positive amyloid and p-tau status [54]. For instance, Guo et al. [51] performed a comparative analysis of fecal samples from individuals with AD, MCI, and healthy controls. Their results showed no significant differences in microbial α-diversity among the groups, but β-diversity was elevated in both AD and MCI patients. Compared to healthy controls, AD patients exhibited decreased abundances of Bacteroides, Lachnospira, and Ruminiclostridium, as well as an increase in Prevotella. MCI patients displayed a similar pattern of alteration when compared to AD patients. However, Lachnospira was the only genus significantly reduced in MCI compared to controls. Notably, the negative correlation between Prevotella abundance and cognitive function persisted in the MCI group. These findings suggest that while the overall richness and evenness of microbial communities (α-diversity) may remain relatively stable in early to moderate stages of cognitive decline, β-diversity often differs because it reflects changes in community composition and structure associated with AD-related neurodegeneration and metabolic alterations.

Yıldırım et al. [52] reported that individuals with AD or MCI exhibit GM dysbiosis characterized by a reduction in beneficial bacteria such as Bacteroides, in addition to an increase in pro-inflammatory genera including Prevotella. The genera Faecalibacterium, Fusicatenibacter, Lactobacillus, and Roseburia were found to be decreased in abundance, whereas Akkermansia and Escherichia/Shigella were more prevalent in AD samples. Verhaar et al. [54] conducted a comparative analysis of GM compositions among patients with AD, MCI, and subjective cognitive decline, identifying only two genera, Phascolarctobacterium and Subdoligranulum, with significantly different abundances across groups, while α- and β-diversity metrics showed no notable differences. Their findings highlighted that the most prominent predictors of amyloid and p-tau status were members of the Lachnospiraceae family, including Lachnoclostridium, Marvinbryantia, Monoglobus, Roseburia, and Ruminococcus. In contrast, elevated levels of Alistipes and Odoribacter correlated with biomarker profiles indicative of more typical AD pathology, such as increased amyloid and decreased p-tau in CSF. Furthermore, key predictors of amyloid levels included Anaerostipes, Eubacterium, and Subdoligranulum, while p-tau predictors encompassed Lachnospiraceae family members Blautia and Lachnoclostridium. Wanapaisan et al. [55] observed no significant differences in α- and β-diversity between AD, MCI, and control groups in a Thai cohort. However, bacterial abundance varied, as Lachnospiraceae family members and Clostridium genus were decreased in both AD and MCI compared to controls. In addition, AD patients showed increased levels of Bacteroides, Escherichia/Shigella, Holdemanella, Megamonas, and Romboutsia, whereas MCI patients exhibited higher abundances of Agathobacter, Faecalibacterium, and Fusicatenibacter. The authors suggested that reductions in Agathobacter, Clostridium, and Faecalibacterium in patients with AD may correlate positively with amygdala and hippocampal brain volumes, regions known to undergo atrophy during early cognitive decline [56].

As previously stated, two primary hypotheses have been traditionally proposed to elucidate the etiology and pathogenic mechanisms of AD. The amyloid cascade hypothesis [8,57] suggests that the accumulation of Aβ plaques within the nervous system constitutes a central risk factor for AD. Specifically, the aggregation of neurotoxic Aβ42 peptides in the brain is recognized as a critical event leading to amyloid plaque formation [58]. This Aβ aggregation initiates a neurotoxic cascade, which triggers cytoskeletal disruptions, neuronal dysfunction, and ultimately neuronal death [59]. Studies have identified the Toll-like receptor 2 (TLR2)-myeloid differentiation primary response 88 (MyD88) signaling pathway as playing a pivotal role in Aβ generation [60]. Experimental evidence in murine models shows that deficiency of MyD88 enhances Aβ clearance via low-density lipoprotein receptor-related protein 1 (LRP1) [61], thereby reducing pro-inflammatory mediators and brain Aβ accumulation [62]. This reduction has been correlated with improved cognitive performance in rat models. Nevertheless, some studies have questioned the essentiality of MyD88 signaling, suggesting that MyD88-dependent pathways may not be strictly required for neuroglial activation or the development of Aβ pathology in the brain [63].

Another crucial hypothesis concerning AD pathogenesis centers on the hyperphosphorylation of tau protein, which leads to the formation of neurofibrillary tangles [64]. Tau protein plays an important role in promoting microtubule assembly and maintaining their stability [65]. Its hyperphosphorylation disrupts synaptic plasticity, which is essential for coordinating memory-related pathways [66]. Thus, tau hyperphosphorylation is directly implicated in triggering the neurodegenerative processes characteristic of AD [67]. Pathogenic tau also induces the release of mitochondrial DNA from microglia, which activates the antiviral cGAS–IFN signaling pathway, resulting in sustained interferon release. This persistent immune activation impairs synaptic integrity and plasticity, ultimately contributing to ongoing cognitive deficits [68]. Recent evidence highlights phosphorylated tau protein 217 as a promising biomarker for AD diagnosis and progression monitoring [69]. Building on these insights, the "dual prion disorder" hypothesis has been proposed, which posits that the pathological interplay between Aβ and tau proteins represents a viable therapeutic target [70,71]. In addition to the well-established amyloid cascade and tau phosphorylation hypotheses, other significant theories explore the roles of neuroinflammation [72], mitochondrial dysfunction [73], and cholinergic deficits [74] in AD pathogenesis. Collectively, these diverse perspectives reflect the ongoing efforts within the scientific community to reveal the complex mechanisms underlying AD.

A growing body of human research has identified several bacterial species, including Borrelia burgdorferi, Chlamydia pneumonia, and Helicobacter pylori, as factors that may increase susceptibility to AD by promoting tau protein hyperphosphorylation, elevating pro-inflammatory bacteria such as Escherichia/Shigella, and reducing anti-inflammatory gut microorganisms such as Ruminococcus [75,76]. Bostanciklioğlu [15] proposed three primary mechanisms linking the GM to AD pathogenesis: (i) central nervous system (CNS) inflammation and cerebrovascular degeneration triggered by bacterial metabolites and amyloids; (ii) impaired autophagy-mediated protein clearance caused by GM dysbiosis; and (iii) modulation of brain neurotransmitter levels via vagal afferents influenced by the GM. Accumulating evidence indicates that disturbances in human GM homeostasis may contribute to the deposition of tau and Aβ peptides [43,77,78]. Multiple gut microbial species, including Bacillus subtilis, E. coli, Salmonella enterica, and S. enterica serotype Typhimurium, have been shown to produce amyloid fibers [79,80,81,82,83]. While microbial amyloids share only tertiary structural similarities with human CNS amyloids, they act as prion-like agents through molecular mimicry, inducing pathogenic β-sheet conformations in host amyloidogenic proteins [16,84,85,86,87]. In elderly individuals, potential mechanisms of amyloid propagation include neuron-to-neuron or distal neuronal spread, as well as direct crossing of the BBB or transport via other cells, such as astrocytes, fibroblasts, microglia, and immune cells [16]. Furthermore, Aβ, tau, and α-synuclein may contribute to neurodegeneration by inducing or exacerbating BBB disruption. These pathological proteins can compromise BBB integrity either directly, by affecting key components such as pericytes and endothelial cells, or indirectly via the activation and dysfunction of brain macrophages [88].

Systemic inflammatory responses triggered by bacterial secretions or structural components have been implicated in the disruption and increased permeability of the BBB. This process involves activation of the TLR4/IRF-3 signaling pathway within endothelial cells lining cerebral blood vessels. Similarly, systemic inflammation induced by bacterial products has been linked to impairment of the intestinal epithelial barrier integrity [89]. Moreover, activation of the lipopolysaccharide (LPS)/TLR4 pathway in glial cells within the CNS has been shown to promote neuroinflammation [90]. Bacterial LPS presence in the brain, facilitated by translocation across both the intestinal barrier and the BBB, has been implicated in AD pathogenesis [91]. LPS stimulates leukocyte and microglial receptors, including TLR4-CD14 and TLR2, resulting in elevated cytokine production and increased Aβ accumulation. In addition, Aβ1–42 acts as an agonist for TLR4 receptors, activating NF-κB pathways that further amplify cytokine release, exacerbating Aβ accumulation, damaging oligodendrocytes, and contributing to the myelin degeneration observed in AD brains [92]. In a recent hypothesis proposed by Brown and Heneka [93], LPS is suggested to contribute to AD pathophysiology through peripheral infections or GM dysbiosis, which elevate LPS concentrations in circulation and the brain, thereby promoting amyloid and tau pathologies along with microglial activation and subsequent neurodegeneration characteristic of AD. Supporting evidence for this hypothesis includes: (i) elevated LPS levels detected in the blood and brains of AD patients; (ii) AD risk factors associated with increased LPS levels or heightened responses; (iii) LPS-induced upregulation, aggregation, inflammation, and neurotoxicity of Aβ; (iv) LPS-driven tau phosphorylation, aggregation, and propagation; (v) LPS-mediated microglial priming, activation, and neurotoxicity; and (vi) LPS-induced synaptic and neuronal loss and memory impairment in AD mouse models, as well as cognitive deficits observed in humans.

The MGB axis plays a pivotal role in the pathophysiology of AD, with its underlying mechanisms extensively characterized [94]. Alterations in GM diversity and their metabolic products in AD disrupt the homeostatic balance of the GM, leading to impaired communication between microorganisms, the gut epithelium, and the immune system, ultimately resulting in developmental and functional deficits of the gut-immune barrier [95]. Metabolites derived from the GM exacerbate damage to the gut mucosal immune system, increase intestinal permeability (commonly referred to as "leaky gut"), and promote abnormal bacterial translocation [96]. These changes intensify systemic inflammatory responses, characterized by a dysregulated Th17/Treg cell ratio and elevated circulating levels of pro-inflammatory cytokines including IL-6, IL-1β, IFN-γ, and TNF-α. Collectively, these immune alterations contribute significantly to the pathological processes associated with AD [95,97,98].

Sphingolipids, classified as bioactive lipids, are essential to cellular signal transduction pathways [141]. These molecules regulate numerous cellular functions, including neuronal stress responses, cell proliferation, differentiation, and maturation [142]. Host sphingolipids promote the release of pro-inflammatory mediators that impact neuronal viability and mitochondrial apoptosis, thereby contributing to neuroinflammation processes implicated in various brain disorders [104]. Emerging evidence suggests that bacterial-derived sphingolipids can stimulate the gut immune system and metabolic homeostasis. However, evidence regarding their direct impact on the human CNS remains limited. Metabolomic profiling of brain tissue and blood samples during preclinical and prodromal phases of AD reveals mixed patterns, with some ceramides increased and certain sphingomyelin levels decreased, which is associated with the progression of AD pathology [143,144]. Moreover, sphingolipid concentrations have been correlated with CSF Aβ levels, brain atrophy, and cognitive deterioration, underscoring their potential utility as early biomarkers for AD [143].

Phospholipids form the fundamental components of the cellular membrane lipid bilayer, serving as a crucial protective barrier for both cellular and subcellular structures. Beyond this structural role, they are involved in maintaining cellular homeostasis, regulating immune responses, and modulating oxidative stress and neuroinflammation within the brain [28]. The predominant phospholipids in the brain are phosphatidylcholine and phosphatidylethanolamine, which relative abundances are altered in the context of AD pathology [145]. The catabolism of these phospholipids is considered a significant metabolic disturbance in AD, with decreased levels strongly correlated with the extent of amyloid deposition and neurofibrillary tangle formation [146].

A recent investigation identified the presence of LPS, which constitute endotoxins produced by Gram-negative bacteria such as Bacteroides fragilis, E. coli, and S. enterica, within amyloid plaques in postmortem brain tissues from individuals with AD [147]. This discovery suggests a possible co-occurrence of these LPS-producing bacteria alongside E. coli in the pathogenesis of AD. In addition, elevated LPS levels have been detected in the blood of AD patients, as well as in the neocortex and hippocampus, where concentrations exceed those found in control subjects [148]. These findings implicate LPS as a potential risk factor contributing to cognitive decline and AD progression [149]. LPS has been shown to compromise both the intestinal barrier and the BBB, triggering a potent neuroinflammatory response, promoting Aβ deposition, and inducing abnormal tau hyperphosphorylation [93]. Such effects lead to microglial activation and neuroinflammation [150], as well as oxidative stress and mitochondrial dysfunction mediated by NADPH oxidase 2 (NOX2), culminating in neuronal injury and synaptic degeneration [149]. Furthermore, LPS acts as an agonist for Toll-like receptor 4 (TLR4), a critical receptor in microglial innate immunity [151], thereby facilitating the release of pro-inflammatory cytokines and further enhancing Aβ accumulation [28].

Trp is an essential aromatic amino acid obtained through dietary intake that is subject to regulation by the GM [185]. Recent metabolomics analyses have revealed significant differences in Trp-derived catabolites (TRYCATs) between individuals with AD and healthy controls, suggesting that dysregulated Trp metabolism, which is driven by microbial imbalance, may contribute to cognitive impairment in AD [47]. A key pathway implicated in this dysregulation is the kynurenine pathway (KP), which appears to play a critical role in AD pathophysiology through the exacerbation of neuroinflammation [186]. The GM has been shown to influence circulating Trp levels, directing its metabolism through the KP and resulting in the production of several neuroactive metabolites. Among these, there are neuroprotective compounds such as KYN and the antioxidant 3-hydroxyanthranilic acid (3-HAA), as well as neurotoxic metabolites including 3-hydroxykynurenine (3-HK) and quinolinic acid (QUIN) [187]. In the context of AD, KP dysregulation is characterized by: (i) reduced plasma concentrations of Trp and 3-HAA; (ii) elevated KYN/Trp ratios and increased serum levels of 3-HK; and (iii) accumulation of QUIN in the hippocampus [186]. Importantly, these KP-derived metabolites exhibit strong correlations with key pathological markers of AD, including amyloid-β42 (Aβ42) and phosphorylated tau (p-Tau181) [188].

Tryptophan catabolites (TRYCATs) have been shown to promote neurogenesis in murine models via mechanisms dependent on activation of the aryl hydrocarbon receptor (AhR) [189]. Notably, AhR has also been implicated in the regulation of sphingolipid metabolism, where it enhances sphingolipid and sphingosine-1-phosphate (S1P) levels, thereby contributing to the maintenance of axonal myelination and neural integrity [190]. Among Trp-derived microbial metabolites, indole plays a pivotal neuroprotective role in the context of AD. It modulates neuroinflammation, supports intestinal barrier function, and contributes to immune homeostasis [191]. Importantly, a marked reduction in the abundance of indole-producing bacteria, particularly within the phyla Actinomycetota, Bacillota, and Bacteroidota, has been reported in patients with AD, paralleled by a significant decline in circulating indole levels [47]. Indole-3-propionic acid (IPA), a specific indole derivative synthesized by Clostridium sporogenes, is capable of intestinal absorption and subsequent penetration of the BBB. Within the CNS, IPA exerts antioxidant effects by neutralizing hydroxyl radicals, mitigating DNA damage, and inhibiting amyloid fibril formation [191]. The neuroprotective potential of indole compounds is attributed to their multifaceted biological activities, including antioxidative, anti-inflammatory, immunomodulatory, and anti-amyloidogenic actions.

Amino acids serve as essential precursors in the biosynthesis of various neurotransmitters. Increasing evidence indicates that the GM not only modulates the availability of these amino acids but also significantly influences neurotransmitter synthesis and activity [192]. Evidence has shown that gut microorganisms possess the capacity to synthesize and regulate a wide range of neuroactive compounds within the host, including GABA, serotonin, dopamine, and norepinephrine [193]. These microbiota-derived neuromodulators exert systemic effects through the MGB axis, thereby modulating neuronal metabolism and CNS function.

Probiotics are defined as live microbial supplements and have demonstrated efficacy in restoring GM balance and providing therapeutic benefits for patients with AD [20]. Nevertheless, evidence supporting their clinical outcomes is still limited. Recent research highlights the promising effects of a novel probiotic formulation, SLAB51, which comprises strains such as Streptococcus thermophilus DSM 32245, Bifidobacterium animalis subsp. lactis DSM 32246 and DSM 32247, Lactobacillus acidophilus DSM 32241, Lactobacillus helveticus DSM 32242, Lacticaseibacillus paracasei DSM 32243, L. plantarum DSM 32244, and Levilactobacillus brevis DSM 27961. This probiotic cocktail has been shown to improve glucose homeostasis, reduce tau phosphorylation [251], and increase SCFAs such as acetate, butyrate, and propionate, while simultaneously decreasing inflammation and Aβ deposition [252]. In addition, SLAB51 activates a SIRT1-dependent pathway that mitigates oxidative stress in the brains of AD mouse models [253]. More recently, SLAB51 has been reported to inhibit cholesterol biosynthesis, reduce the ω-6/ω-3 fatty acid ratio, and attenuate neuroinflammation and oxidative stress, collectively leading to decreased aggregation of Aβ and tau proteins and slowing AD progression [254].

As demonstrated by Song et al. [255], the probiotic L. plantarum strain DP189 effectively regulates GM dysbiosis and inhibits tau hyperphosphorylation. Furthermore, this strain has been shown to suppress the production of trimethylamine (TMA) and its hepatic metabolite trimethylamine-N-oxide (TMAO), leading to reduced expression of the clu gene (clusterin). These effects collectively alleviate neuroinflammation and neuropathological defects in APP/PS1 transgenic mice [256].

The probiotic cocktail VSL#3, comprising eight strains, L. plantarum, Lactobacillus delbrueckii subsp. bulgaricus, L. acidophilus, L. paracasei, Bifidobacterium breve, Bifidobacterium longum, B. longum subsp. infantis, and Streptococcus salivarius subsp. thermophilus, has been shown to effectively reduce serum levels of prostaglandin and deoxycholic acid. This reduction contributes to the amelioration of intestinal inflammation and gut permeability. However, its effects on brain amyloid plaque deposition, pro-inflammatory cytokine levels, and gliosis appear to be limited [257].

In summary, probiotics have been demonstrated to modulate GM composition and lipid metabolism homeostasis, thereby exerting beneficial effects on brain inflammation, oxidative stress, and the pathologies of Aβ and tau. Thus, targeting the microbial community through probiotic interventions offers promising avenues for the development and clinical evaluation of novel preventive and therapeutic strategies for AD.

Prebiotics, comprising human milk oligosaccharides, non-digestible oligosaccharides, and soluble fermentable fibers, represent a promising alternative or complement to probiotic supplementation. These compounds have been shown to enhance the growth of beneficial bacteria, such as bifidobacteria and lactobacilli, thereby improving cognitive impairment in APP/PS1 mice via the MGB axis [258,259]. The neuroprotective properties of prebiotics highlight their potential as oral formulations for the prevention and treatment of AD [260]. Specifically, fructooligosaccharides have demonstrated beneficial effects in APP/PS1 mice [261], while mannanoligosaccharides have been found to reshape the GM, maintain intestinal barrier integrity, increase SCFA production, inhibit neuroinflammation and oxidative stress, and alleviate cognitive and behavioral deficits in 5xFAD mice [262]. In addition, xylooligosaccharides, abundant plant-derived biopolymers resembling oligosaccharides, have been shown to significantly alter GM composition and improve AD symptoms such as anxiety and depression, and may also help preserve or enhance cognitive function in individuals at risk for AD. [262,263]. While prebiotic supplementation holds promise for modulating microbial composition and function to improve AD outcomes, further research is necessary to fully assess their therapeutic applicability.

Fecal microbiota transplantation (FMT) involves transferring the fecal microbiota of a healthy donor into the gut of a patient or diseased animal to restore GM balance and improve disease outcomes [264]. Increasing evidence supports the therapeutic potential of FMT from wild-type (WT) mice in ameliorating AD-related pathologies in transgenic ADLPAPT mouse models. This intervention has been shown to reduce Aβ plaques, neurofibrillary tangles, glial reactivity, and cognitive impairments, indicating that restoring GM homeostasis via FMT may benefit AD treatment [265]. Similarly, FMT from WT mice has been reported to positively reshape the GM in APP/PS1 mice by increasing Bacteroidota abundance while reducing Pseudomonadota and Verrucomicrobiota populations, elevating butyrate levels, and significantly improving Aβ accumulation, synaptic dysfunction, neuroinflammation, and cognitive deficits [266]. Moreover, Qian et al. [267] demonstrated that FMT from WT mice modulates glycerophospholipid metabolism in APP/PS1 mice, further alleviating Aβ pathology and neuroinflammation. In contrast, FMT from AD mice has been found to impair memory and neurogenesis in WT mice [268,269]. Nevertheless, the precise mechanisms by which GM influences these outcomes require further clarification, but these findings collectively suggest a promising role for healthy FMT in modulating AD pathology [270].

A clinical trial has reported that FMT improved clinical manifestations in AD patients, notably cognitive impairments. For instance, a case study described an AD patient who received FMT from a healthy 47-year-old woman and subsequently exhibited significant improvements in cognitive function, as evidenced by enhanced mini-mental state examination and clinical dementia rating scores [271]. Despite the promising efficacy and safety of FMT demonstrated in murine models, several challenges remain for its clinical application. These include acceptance and compatibility between recipients and donors, individual variations in immune responses, and lifestyle factors, such as diet, that influence GM composition. Furthermore, relatively few studies have systematically addressed GM variations following FMT. Table 3 provides an overview of current research on GM-based therapies aimed at mitigating AD-related pathological features.