Can Ganoderma lucidum polysaccharides combat cognitive aging through the Blautia coccoides-butyrate-NLRP3 axis?

From dysbiosis correlation to a named microbial-metabolite-inflammasome axis

Human evidence has established that gut microbiome composition relates to cognitive aging: in a cross-sectional cohort of 288 older adults with normal cognition, subjective cognitive decline, MCI, Alzheimer's disease, or subcortical vascular dementia, a data-driven microbiome cluster characterized by higher Bacteroidota, Proteobacteria, and Verrucomicrobiota was associated with greater brain age, and brain age significantly mediated the relationship between cluster membership and MMSE score (indirect effect ß = -0.352, p = 0.041) [2]. That study was explicitly cross-sectional and could not establish temporal precedence or causality [2]. Separately, reviews of Ganoderma lucidum have catalogued neuroprotective polysaccharide and triterpenoid activities across neurodegenerative models, but noted that most studies used impure extracts and that the specific bacterial or metabolite intermediaries were undefined [7]. The new paper's contribution is to fill that gap by naming Blautia coccoides as the enriched taxon, butyrate as the effector metabolite, and NLRP3 as the central target, then testing each link with FMT, bacterial reconstitution, butyrate supplementation, and NLRP3-knockout mice [1].

What the mouse experiments actually show, and how strong the causal chain is

In D-galactose-treated male C57BL/6J mice, GLPs (150 mg/kg) improved novel object recognition, Morris water maze, and Y-maze performance, reduced hippocampal p53, p21, and p16, restored synaptic proteins MAP2, PSD95, synaptophysin, BDNF, and MBP, and lowered IL-6, TNF-α, and IL-1β [1]. 16S profiling identified Blautia as the core genus enriched by GLPs, with B. coccoides showing the greatest fold-change increase, and fecal butyrate rose in parallel [1]. Fecal microbiota transplantation from GLP-treated donors reproduced the cognitive and anti-inflammatory effects in recipients, and direct B. coccoides gavage or sodium butyrate supplementation (150 mg/kg/day) each replicated the phenotype [1]. In NLRP3-knockout mice, the benefits were absent, and in BV2 microglial cells, butyrate and B. coccoides-derived supernatant suppressed NLRP3 activation in vitro [1]. The authors themselves note that the causal chain would be strengthened by germ-free mice colonized with B. coccoides or by specific butyrate receptor agonists and inhibitors, which were not performed [1].

How this compares with other gut-brain and neuroimmune interventions

The GLP study is not the only attempt to link a peripheral intervention to central neuroinflammation. A 2026 study of porcine plasma-derived small extracellular vesicles in 5×FAD mice reported that intravenous delivery crossed the blood-brain barrier, reduced amyloid burden, attenuated neuroinflammation, and rescued cognitive deficits, with the authors framing the approach as a scalable, multi-target biologic rather than a single-axis intervention [8]. A 2027 paper on Alzheimer's disease proposes a microbiota-butyrate-BHB axis and explicitly invokes HDAC inhibition and NLRP3 suppression as mechanisms, which overlaps with the GLP paper's butyrate-NLRP3 logic but positions BHB, not Blautia coccoides, as the key metabolite [4]. A 2025 study of low-molecular-weight Ganoderma lucidum polysaccharide peptide in spinal cord injury rats reported reduced M1 microglia, increased M2 microglia, improved BBB scores, and reduced colon inflammation, indicating that Ganoderma polysaccharides can modulate microglial phenotype and gut barrier in a different injury model [3]. A 2025 human ex vivo SHIME study of Ganoderma lucidum and Hericium erinaceus found that microbiome-derived cell-free supernatants upregulated BDNF, CDNF, and MANF and modulated CREB/BDNF signaling, providing human-relevant evidence that Ganoderma preparations can engage gut-brain signaling, though it did not measure cognition or NLRP3 [6]. These comparisons matter because they show the GLP paper's axis is one plausible route among several, not the only one.

Where the conclusion stops: mouse models, D-galactose aging, and unmeasured human outcomes

The most important boundary is that all causal claims come from mice. The aging model was D-galactose-induced premature aging in 7–8-week-old male C57BL/6J mice, not natural chronological aging, and the authors acknowledge that accelerated or genetically modified aging models may not recapitulate the chronic pathophysiology of natural aging [1]. The B. coccoides reconstitution was performed in antibiotic-pretreated SPF mice, not germ-free or gnotobiotic animals, so residual microbiota could contribute [1]. The study did not identify which structural features of GLPs enrich B. coccoides, and the authors call for advanced glycomics to isolate active fractions [1]. Human evidence remains indirect: the cohort study linking microbiome clusters to brain age and MMSE was cross-sectional and could not infer causality [2], and the human ex vivo mushroom study measured neurotrophic signaling, not cognitive outcomes [6]. Other limitation evidence reinforces that cognitive phenotyping in mice is method-sensitive: a 2026 Barnes maze protocol paper shows that odor cues and protocol design can disrupt interpretation of learning trials, and that different dementia models produce different behavioral profiles [9]. A 2026 tauopathy model paper further shows that even within mouse models, genetic background and transgene design strongly shape pathology and memory deficits [5]. Finally, a 2025 study of noise-induced hearing loss in 17-month-old mice and a 2025 theta-shaking study in senescence-accelerated mice both illustrate that aging-related cognitive and sensory decline involves multiple, region-specific mechanisms beyond a single gut-inflammasome axis [10][11].

About These Sources

This research page is built on 11 peer-reviewed studies — published from 2024 to 2027, 11 from 2024 or later — selected as the most relevant from 13 studies that passed quality screening, drawn from 83 papers retrieved from a database of over 500 million.

Sources used in this answer