A Metabolite Level View of the Gut-Brain Axis in Dementia and Aging: Quantification of the Kynurenine/Tryptophan Pathway Using Mass Spectrometry Luke Whiley (Presenter) Imperial College London Presenter Bio: Whilst studying for an undergraduate degree in Biochemistry with Toxicology, I initially developed an interest in analytical technologies, metabolite profiling and phenotyping. This lead me to complete a PhD at King’s College London in mass spectrometry based biomarker discovery. Following my PhD, I made the conscious decision to spend time in regulated laboratories to develop key skills in mass spectrometry analysis, including analyst roles within the anti-doping laboratory at the London 2012 Olympic Games, the Glasgow 2014 Commonwealth Games and the UK Defence science and technology laboratory. I made the transition back to academia as a post-doctoral researcher at the MRC-NIHR National Phenome Centre (NPC) and work on developing mass spectrometry based metabolomics workflows including metabolite identification and quantitative metabolite pathway analysis.

Authors: Luke Whiley (1,2), Leanne Nye (3), Nick Andreas (3), Katie Chappell (2), Ian Wilson (3), Matthew Lewis (2,3), Jeremy Nicholson (2,3), Jonathan Swann (2,3), Elaine Holmes (1,2,3)
1- UK Dementia Research Institute, Imperial College London, 2 - MRC-NIHR National Phenome Centre, Imperial College London, 3 - Division of Integrative Systems Medicine and Digestive Disease, Imperial College London

Compositional variation in the gut microbiota has been reported in dementia and aging. The intestinal bacteria are now known to influence biochemical processes in the brain and modify host behaviours and cognitive development. This gut-brain axis arises through a variety of mechanisms that are not completely understood. One mechanism is the tryptophan-kynurenine pathway. Intestinal bacteria can metabolize tryptophan, a precursor for serotonin, reducing its bioavailability. In addition, microbes can interact with the immune system altering the amount tryptophan converted to kynurenine and its neuroactive metabolites. A UPLC-MS method was developed to quantify serum/plasma metabolites in this pathway and assess their variation in large clinical cohorts reflective of neurological disease, cognition and dementia.

Introduction

The influence of the gut microbiome on brain health is becoming more apparent. Differences in our gut microbial composition have been reported in neurological conditions including Alzheimer’s disease [1,2] and Parkinson’s disease [3]. However, the metabolism that underlies this gut brain axis interaction has not yet been elucidated. The tryptophan and kynurenine pathway is of interest as their bioavailability is controlled by microbial colonies [4] and is therefore reflective of gut composition at the metabolite level, whilst downstream metabolites include neuroactive molecules that influence receptors in the central nervous system [5]. To study the kynurenine pathway in clinical cohorts a targeted metabolomics UPLC-MS method was developed to quantify key metabolites and intermediates from the pathway. The method presented several significant challenges in the development phase, including a vast dynamic range of endogenous analytes. Once these were overcome the method was validated in both serum and plasma.

Methods

Calibration points were prepared for 16 compounds of the kynurenine pathway at a concentration range reflective of expected endogenous concentration. This presented a challenging working dynamic range between major metabolites and minor intermediates. Calibration standards, QCs and biological samples underwent identical sample clean-up preparation including protein precipitation with methanol. Samples were then passed through hybrid protein and lipid removal SPE plates prior to drying and re-suspension LC compatible diluent. UPLC-MS analysis was completed on a Waters Acquity UPLC system coupled to a Waters Xevo TQ-S. MRM transitions for 15 compounds were analysed in positive ionisation mode with 1 MRM transition monitored in negative ionisation MRM.

Results

All compounds were quantified with a stable isotope labelled internal standard. Linearity of r2 > 0.99 was achieved for each, with QC acceptance of 15% (20% for the lowest QC). Significant challenges were encountered in the vast differences of dynamic range between major metabolites and minor intermediates. Initially, tryptophan was particularly vulnerable to poor linearity across the desired calibration range, with in-source ionisation saturation effecting analysis. To resolve this, the instrument was de-tuned for sensitivity in negative ionisation mode (capillary voltage 0.15V), taking advantage of the rapid polarity switching on the Waters TQ-S system resulting in linearity across the required range 160 ng/mL to 16000 ng/mL. To preserve column lifetime across multi-plate analysis of large cohorts a hybrid protein and phospholipid removal SPE step was included. This required a large amount of method optimisation to balance sample clean-up whilst achieving a high percentage analyte recovery.

Conclusions & Discussion

A UPLC-MS method was developed to quantify a panel of significant endogenous analytes that are reflective of gut microbiome activity at a metabolic level. 16 metabolites that are associated with the gut microbiome and kynurenine pathway were fully quantified and validated in serum and plasma. The method was employed to investigate metabolic changes in the pathway in large cohorts reflective of neurological disease. The method will be adapted and validated for use in multiple biofluids and tissues, including brain and CSF, with the aim to improve our understanding of the gut-brain axis, and the metabolites that influence it.

[1] Porter RJ et al. - Am J Psychiatry. 2000 Apr;157(4):638-40

[2] Gulaj E et al. - Adv Med Sci. 2010;55(2):204-11

[3] Chang KH et al. – Mol Neurobiol. 2018; [Epub ahead of print]

[4] Kennedy PJ et al - Neuropharmacology. 2017; 112(Pt B):399-412

[5] Reyes-Ocampo J et al -Oxid Med Cell Longev. 2014; 2014:646909