Altered Gut Microbial Metabolites in Amnestic Mild Cognitive Impairment and Alzheimer’s Disease: Signals in Host–Microbe Interplay

Intimate metabolic host–microbiome crosstalk regulates immune, metabolic, and neuronal response in health and disease, yet remains untapped for biomarkers or intervention for disease. Our recent study identified an altered microbiome in patients with pre-onset amnestic mild cognitive impairment (aMCI) and dementia Alzheimer’s disease (AD). Thus, we aimed to characterize the gut microbial metabolites among AD, aMCI, and healthy controls (HC). Here, a cohort of 77 individuals (22 aMCI, 27 AD, and 28 HC) was recruited. With the use of liquid-chromatography/gas chromatography mass spectrometry metabolomics profiling, we identified significant differences between AD and HC for tryptophan metabolites, short-chain fatty acids (SCFAs), and lithocholic acid, the majority of which correlated with altered microbiota and cognitive impairment. Notably, tryptophan disorders presented in aMCI and SCFAs decreased progressively from aMCI to AD. Importantly, indole-3-pyruvic acid, a metabolite from tryptophan, was identified as a signature for discrimination and prediction of AD, and five SCFAs for pre-onset and progression of AD. This study showed fecal-based gut microbial signatures were associated with the presence and progression of AD, providing a potential target for microbiota or dietary intervention in AD prevention and support for the host–microbe crosstalk signals in AD pathophysiology.


High-throughput fecal untargeted metabolomics profiling
Chemical Reagents. HPLC grade methanol and ultrahigh quality water were purchased from Thermo Fisher Scientific (USA).
Sample Preparation. The fresh fecal samples were collected and stored at -80 °C until analysis. The samples were extracted by methanol at a ratio of 3 mL/g [11], and ceramic beads (1 mm, OMNI, USA) were added for homogenizing (Omni International, USA).
Then, the mixtures were centrifuged twice (12,000 rpm, 10 min, Eppendorf, Germany) and the supernatant was collected and filtered using 0.22 μm syringe filters (Millipore, USA). An equal volumes of all extracted samples were pooled as the QC sample.
Metabolomics profiling analysis. In this study, the Dionex UltiMate 3000 RS system coupled with Q Exactive HF-X mass spectrometry (MS) (Thermo Fisher Scientific, USA) was applied for metabolite separation and MS detection [12]. The liquid chromatographic separation was performed by a Hypersil Gold C-18 column  -40% B during 0.5-8 min, 40%-98% B during 8-12 min, 98% B during 12-14 min, and 2% of B at last 2.5 min. MS detection was conducted by Q Exactive HF-X MS with heated-ESI-II (HESI-II) ion source (Thermo Fisher Scientific, USA) as previously described [12]. The acquisition mode was a full MS with a m/z range 70-1050 followed by data-dependent MS 2 (dd-MS 2 ). The resolution was set at 60,000 and 15,000 for full MS and dd-MS 2 , respectively. The MS 2 spectrometry data were acquired with the collision energy of 20, 40 and 60 eV.
In order to delete the confounding factors during the experiment, the blank sample (100% HPLC-grade water) was run with the sample simultaneously. Ten QC samples were run before sample analyzing to equilibrate the detection system, and one QC sample was run every ten samples during sample processing to monitor the stability of the acquisition system [13].
Data processing. The data processing was performed in Compound Discoverer 3.1 software (Thermo Fisher Scientific, USA) according to the manufacturer's user guideline. The multivariate statistical analysis including principle component analysis (PCA) and partial least-squares-latent structure discriminate analysis (PLS-DA) were performed using SIMCA-P 13.0 (Umetrics AB, Sweden). And the univariate analysis was conducted by one-way ANOVA or Kruskal-Wallis test in SPSS software (version 16.0, SPSS Inc., USA) and GraphPad Prism 6 (GraphPad Inc., USA). To identify potential biomarkers for AD, the relative abundance of metabolites were calculated by assigning the total peak area of the metabolic profiles from one sample to 10 7 , and receiver operating characteristic (ROC) analysis based on the relative abundance of differential metabolites were performed. Amethyst Chemicals (China), respectively.
Sample Preparation. The fecal sample from each subject were collected and derivatized as previously described [14,15]. Briefly, samples were prepared by mixing with 10% isobutanol, then ceramic beads were added for homogenizing (50 Hz, 30 s). After centrifugation (12000rpm, 5min), the supernatant was obtained and chloroform were added to remove lipophilic compounds. For chloroformate derivatization, the esterification was performed using isobutanol and isobutyl chloroformate, and NaOH and pyridine were added as base and catalyst, respectively. One boiling stone (Acros Organics, USA) was added to avoid bumping. Finally, hexane was added and centrifuged, and the upper hexane-isobutanol phase was transferred into a gas chromatography (GC) vial for detection. Sample Preparation. Fecal samples were prepared as previously described [16].
The mobile phase consisted of A (water with 0.005% formic acid, v/v) and B (acetonitrile with 0.005% formic acid, v/v). The eluent gradients were set as follows: 7 23%-33% of B during 0-2 min, 33%-34% B during 2-6 min, 34%-70% B during .01 min, 70%-95% B during 11.01-15 min, and 95% of B at last 5 min. And the acquisition mode of MS detection was multiple reactions monitoring (MRM). Data were collected using Mass Hunter software (Agilent Technologies, USA). The BAs were identified by referring to the retention time and ion pairs of standard chemicals (Table S5) and quantified by the internal standard calibration curves.

Measurement of circulating lipopolysaccharide level
The serum lipopolysaccharide (LPS) was measured using limulus amoebocyte lysate (LAL) chromogenic endpoint assay (Hycult Biotech, Uden, Netherlands) in 48 subjects (AD, n = 12; aMCI, n = 27; HC, n = 9) [18]. The serum samples were diluted 1:3 with endotoxin-free water and heated at 75°C for 5 minutes in a water bath to neutralize the endotoxin inhibiting compounds. Then, the LAL reagent was added and incubated with sample for 20 min at 25°C. Finally, the reaction was terminated by adding the stop solution and measured by spectrophotometer (Biotek, Vermont, USA). 8 Tables   Supplementary Table S1. Mass list library of twenty tryptophan metabolites.