# Mechanistic links between the benefits of pharmacologically high thiamine (vitamin B1) in Alzheimer's disease to Advanced Glycation Endproducts (AGE)

> **NIH NIH R01** · WINIFRED MASTERSON BURKE MED RES INST · 2024 · $1,018,953

## Abstract

Experimental data linking thiamine (vitamin B1) deficiency to Alzheimer’s disease (AD) inspired our
clinical trial, which generated preliminary evidence that pharmacological thiamine produced by the drug
benfotiamine provides clinical benefit. We hypothesize that pharmacological thiamine is protective by diminishing
the formation of advanced glycation endproducts (AGE). AGE are proteins and lipids that become glycated and
harmful following exposure to reducing sugars. AGE cause irreversible damage to biological macromolecules by
altering their structural and functional integrity. Abundant evidence links AGE to AD. In AD animal models,
thiamine deficiency increases AGE and exacerbates plaques and tangle formation, while increased thiamine
diminishes AGE and pathology. In our pilot clinical trial, pharmacological thiamine levels diminished global
plasma AGE levels and improved symptoms in patients with AD. Interestingly, in AD patients, the effects of high
thiamine are diminished in patients carrying the APOE4 genotype, the most significant genetic risk factor for
sporadic AD. We postulate that this is because APOE4 increases unique AGE at earlier stages.
 Optimizing this therapeutic approach requires a better understanding of the mechanism underlying the
action of benfotiamine. All previous related AD and thiamine studies have utilized AGE antibody surveys. This
data is limited to a small range of AGE and provides no data on the proteins and specific sites modified with
glycation. We will provide this critical data by using multiple state-of-the-art mass spectrometric measures of
AGE. Global glycaproteomics will identify glycated proteins and specific sites of AGE modifications and AGE-
omics will identify a broad range of crosslinking and non-crosslinking AGE. A second major gap is the lack of
our understanding on how APOE4 modifies the response to thiamine. Novel APOE3 and APOE4 humanized
APP mouse models will allow us to test these interactions. We will test our hypotheses: (1) In AD autopsy
brains at different stages of the disease, AGE modifications are critical to the pathophysiology of AD in
an APOE-dependent manner. (2) In mouse models of AD, thiamine deficiency drives AD-like pathology
and memory loss by causing specific brain and blood AGE modifications which are modified by APOE
genotype. (3) In mouse models, benfotiamine is beneficial by diminishing specific AGE and the treatment
must be initiated at an earlier stage of disease in APOE4 mice.
 These studies will dramatically improve our understanding of the role of AGE in AD and its link to a treatment
of pharmacological thiamine levels. Defining the interaction of AGE and thiamine in the etiology and progression
of AD will enable the development of specific AGE signatures for targets of engagement for therapeutic trials
and as AD diagnostic and prognostic biomarkers. These studies will define the differential effectiveness of
thiamine by APOE genotype and define the most effective the...

## Key facts

- **NIH application ID:** 10806795
- **Project number:** 1R01AG082228-01A1
- **Recipient organization:** WINIFRED MASTERSON BURKE MED RES INST
- **Principal Investigator:** Sarah Ann Flowers
- **Activity code:** R01 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2024
- **Award amount:** $1,018,953
- **Award type:** 1
- **Project period:** 2024-02-01 → 2029-01-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10806795

## Citation

> US National Institutes of Health, RePORTER application 10806795, Mechanistic links between the benefits of pharmacologically high thiamine (vitamin B1) in Alzheimer's disease to Advanced Glycation Endproducts (AGE) (1R01AG082228-01A1). Retrieved via AI Analytics 2026-05-26 from https://api.ai-analytics.org/grant/nih/10806795. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*