Coffee Compound Trigonelline Eases Parkinson’s Symptoms in Mice by Calming Brain Inflammation and Reshaping Gut Microbes

A humble alkaloid found in every roasted coffee bean and fenugreek seed is emerging as one of the more intriguing candidates in the search for disease-modifying therapies for Parkinson’s disease. In a new preclinical study published in Food Science & Nutrition, researchers report that trigonelline, a pyridine alkaloid long overshadowed by caffeine, eased motor deficits, reduced the burden of pathological alpha-synuclein, and dampened neuroinflammation in a mouse model of Parkinson’s disease. Remarkably, the compound’s effects extended well beyond the brain: it also restored gut barrier integrity, remodeled the intestinal microbiome, and shifted the colonic metabolic landscape, suggesting a compound that acts on both ends of the gut-brain axis at once.

Parkinson’s disease is classically defined by the loss of dopaminergic neurons in the substantia nigra pars compacta and the accumulation of misfolded alpha-synuclein into Lewy bodies. But the clinical picture is increasingly understood as systemic. Misfolded alpha-synuclein acts as a damage-associated molecular pattern, provoking microglia, the brain’s resident immune cells, into a chronically activated, pro-inflammatory state. These activated microglia release tumor necrosis factor-alpha, interleukin-1 beta, and reactive oxygen species through pathways such as NF-kappaB, which in turn kills more neurons and further impairs the microglia’s ability to clear alpha-synuclein aggregates. The result is a self-perpetuating neurotoxic loop that current therapies, chiefly levodopa-based dopaminergic replacement, do nothing to interrupt.

The research team, led by investigators at Wenzhou Medical University, began not in the animal facility but in genetic epidemiology. Using two-sample Mendelian randomization with summary data from large genome-wide association studies, they asked whether genetically predicted serum trigonelline levels are causally linked to alpha-synuclein levels. The inverse variance weighted analysis revealed a significant protective association, with an odds ratio of 0.858, and the weighted median method corroborated the result. Sensitivity analyses found no heterogeneity among the instrumental variables, no evidence of horizontal pleiotropy, and no single genetic variant driving the signal. While Mendelian randomization is only as strong as its assumptions, the analysis provided a genetically grounded rationale for the experiments that followed.

For the in vivo work, the team used an adeno-associated virus to overexpress the A53T mutant form of alpha-synuclein in the substantia nigra of C57BL/6J mice, a model that recapitulates alpha-synucleinopathy, dopaminergic degeneration, neuroinflammation, and motor impairment. Two weeks after viral injection, mice received daily oral trigonelline at 5 milligrams per kilogram for 30 consecutive days, a dose chosen from prior mouse work confirming its efficacy and lack of toxicity. The behavioral results were striking. Trigonelline-treated mice traversed a narrow balance beam faster, stayed longer on an accelerating rotarod, and descended a vertical pole more quickly than their vehicle-treated counterparts, collectively indicating preserved motor coordination and reduced bradykinesia.

Histology and biochemistry told a consistent story beneath the behavior. Immunohistochemistry for tyrosine hydroxylase, the rate-limiting enzyme in dopamine synthesis, revealed substantial preservation of dopaminergic neurons in the substantia nigra of trigonelline-treated mice, and Western blotting confirmed that nigral tyrosine hydroxylase protein levels, which had dropped in the disease model, were significantly restored. Just as importantly, immunofluorescence staining showed that the accumulation of phosphorylated alpha-synuclein at serine 129, the pathological hallmark that drives neurotoxicity, was markedly reduced in treated animals, a finding confirmed at the protein level by Western blot. The compound, in other words, appeared to suppress the very pathology that sets the disease cascade in motion.

The mechanistic core of the study lies in inflammatory signaling. In the substantia nigra of diseased mice, microglia bearing the activation markers Iba-1 and CD11b adopted a hypertrophic, ameboid morphology, and levels of the pro-inflammatory effectors COX-2, TNF-alpha, and IL-1 beta rose sharply. Trigonelline treatment reversed all of these changes. Probing upstream, the researchers found that the disease model exhibited hyperactivation of the PI3K/AKT pathway, reflected in elevated phosphorylation ratios of PI3K and AKT, together with increased phosphorylation of the NF-kappaB p65 subunit, the master transcriptional regulator of inflammation. Trigonelline normalized all three phosphorylation events, and quantitative PCR showed corresponding suppression of Il1b, Il6, and Nfkb1 messenger RNA. The data position trigonelline as a suppressor of the AKT-to-NF-kappaB axis that governs microglial activation, a pathway increasingly recognized as a promising therapeutic target in neurodegeneration.

What elevates the study above a straightforward neuroprotection paper is its attention to the periphery. Consistent with the Braak hypothesis, which proposes that Parkinson’s pathology may originate in the enteric nervous system and propagate to the brain via the vagus nerve, the team examined the colons of their mice. Diseased animals showed mucosal atrophy, depleted goblet cells, reduced glandular depth, and fragmented tight junctions, with diminished and discontinuous staining for the barrier proteins ZO-1 and Occludin. Trigonelline restored the continuous, belt-like organization of both proteins and preserved crypt architecture. The compound also reduced the infiltration of Iba-1-positive and CD11b-positive myeloid cells in the colonic lamina propria, suppressed colonic COX-2 and TNF-alpha, and visibly reduced enteric phosphorylated alpha-synuclein, suggesting it lightens the peripheral pathological burden as well as the central one.

Colon content sequencing with 16S rRNA amplicon analysis revealed a parallel microbial transformation. Principal coordinate analysis based on Bray-Curtis distances separated the trigonelline and disease groups cleanly, with a PERMANOVA statistic of 0.7971 and a p-value of 0.001. LEfSe biomarker analysis showed that diseased mice were enriched with pro-inflammatory and opportunistic taxa, including Desulfovibrionaceae, Helicobacteraceae, and the genera Dubosiella and Desulfovibrio, while trigonelline treatment enriched beneficial families such as Erysipelotrichaceae and Clostridiaceae and genera including Ileibacterium, Bilophila, and Gordonibacter. The correlations with behavior were particularly compelling: Desulfovibrio abundance tracked positively with the time needed to complete beam and pole tests, whereas Ileibacterium, expanded by trigonelline, correlated negatively with those latencies. This matters because patient-derived Desulfovibrio strains have previously been shown to enhance alpha-synuclein aggregation in vivo, making their suppression by a dietary compound especially noteworthy.

Untargeted metabolomics of colon contents added a third layer of evidence. The analysis identified 309 differential metabolites, and Procrustes analysis demonstrated clear concordance between the metabolomic configuration and the microbial structure, indicating coordinated remodeling. KEGG topology pointed to enrichment in glycerophospholipid metabolism, taurine and hypotaurine metabolism, and amino acid metabolism. Among the metabolites correlating with better motor performance, one feature tentatively annotated as ganolucidic acid B stood out: it increased with trigonelline treatment, correlated positively with trigonelline-enriched bacterial taxa, and negatively with disease-associated ones. An exploratory network pharmacology analysis of its predicted targets intersected with Parkinson’s genes to yield 146 overlapping targets and ten consensus hub genes, prominently including PIK3CA, PIK3CD, PIK3R1, NFKB1, STAT3, and TNF, with the PI3K-Akt signaling pathway emerging as the most enriched KEGG pathway. The authors are careful to stress that this metabolite’s identity was not confirmed with an authentic standard, and that the network predictions are hypothesis-generating rather than proof of target engagement.

The researchers are equally candid about the study’s broader limits. The AAV-SynA53T model, while faithful to several Parkinson’s-like features, cannot reproduce the chronic, heterogeneous nature of human disease; the signaling changes remain correlative rather than demonstrably causal; and the gut findings reflect system-level associations, leaving open whether trigonelline’s neuroprotection depends on the gut-brain axis or occurs in parallel with direct central effects, particularly since prior work has detected orally administered trigonelline in the brain. Still, the translational arithmetic is tantalizing: the 5 milligram per kilogram mouse dose corresponds to roughly 24.3 milligrams per day for a 60-kilogram adult, a range that overlaps with the trigonelline content of some coffee preparations. That does not mean drinking more coffee will prevent Parkinson’s, since intake varies enormously with bean, roast, and brewing method. But it does mean that a dietary-exposure-range compound has now been shown, in a rigorous multi-omics preclinical framework, to simultaneously quiet neuroinflammation, clear pathological protein, and reshape the gut ecosystem. As the search for disease-modifying Parkinson’s therapies intensifies, trigonelline has earned a place on the shortlist of nutritional neuroprotective candidates worth serious follow-up.

Subject of Research: Neuroprotective and gut microbiome-modulating effects of the coffee alkaloid trigonelline in a mouse model of Parkinson's disease

Article Title: Trigonelline Alleviates Neuroinflammation in AAV‐SynA53T Parkinson's Disease Model Mice: Involvement of PI3K/AKT–NF‐κB Signaling and Gut Microbial‐Metabolic Remodeling

Article References: Chen, M., Dong, Y., Zheng, J., Lin, Y., Li, Q., Wei, C., Zhuo, L., He, Y., Sun, J., Fu, F., & Liu, J. (2026). Trigonelline Alleviates Neuroinflammation in AAV‐SynA53T Parkinson's Disease Model Mice: Involvement of PI3K/AKT–NF‐κB Signaling and Gut Microbial‐Metabolic Remodeling. Food Science & Nutrition, 14(10), Article e72451. https://doi.org/10.1002/fsn3.72451

Image Credits: AI Generated

DOI: 10.1002/fsn3.72451

Keywords: trigonelline, Parkinson's disease, alpha-synuclein, neuroinflammation, gut-brain axis, microglia, PI3K/AKT, NF-kappaB, gut microbiota, Mendelian randomization, metabolomics, coffee