|Stepwise HPLC Extraction and Purification of Endogenous Urinary Metabolites for Their Annotation and Identification in Large Scale Phenotyping Studies
|Exploiting Separation Science
|Dr Luke Whiley
|Dr Elena Chekmeneva
Dr David Berry
Prof Zoltan Takats
Prof Elaine Holmes
Prof Jeremy Nicholson
|MRC-NIHR National Phenome Centre, Imperial College London
Abstract Information :
As large scale metabolite phenotyping studies evolve, greater emphasis is being placed on biological interpretation of results. In order to understand biomarkers and link them to specific pathways accurate annotation of features is key. To date structural elucidation is a complex task for synthetic chemistry, drug discovery and natural products research, however it is now considered a significant bottleneck in the metabolic phenotyping workflow. Low concentration metabolites can be notoriously difficult to annotate, whilst many biological standards are unavailable for purchase or can be prohibitively expensive. Here, we present a urine based pipeline for metabolite extraction and increased confidence annotation using traditional and widely available analytical scale HPLC-MS technology coupled with a simple off the shelf fractionation system. The resulting workflow enables the purification and concentration of small molecule metabolites in urine that could be easily implemented in analytical laboratories without the need for specialist purification LC systems.
In order to perform the primary HPLC fractionation a composite urine pool (from 8 healthy volunteers) underwent a simple desalting procedure using ion-exchange. Desalted urine was then pre-concentrated (10X) by evaporation under a stream of nitrogen. One hundred 1mL injections of the pooled concentrate then underwent primary fractionation using a traditional reversed-phase (RP) C18 HPLC (4.6mm ID) column. This process lead to the formation of the fraction bank - 120 sub-fractions that are stored at -80°C for downstream analysis. Repeatability was monitored by use of a 1/20 split to MS (Waters TQS) operating on full scan acquisition. The process of desalting urine provided an increase in chromatographic resolution and precision between injections during the primary fractionation. This led to prolonged column lifetime, enabling an increase in the number of repeat fractions, and therefore enabling the purification of more bulk material - a key requirement to reach a concentration suitable for NMR structural elucidation.
Primary fractions then underwent a secondary and tertiary isocratic separation to further purify target metabolites, taking advantage of subtle differences in column chemistry, mobile phase selectivity, and varied LC conditions to achieve required purity for downstream NMR structural elucidation. All effluent throughout the fractionation process was re-collected, dried under N2 and returned to the -80°C fraction bank.
For features that experienced co-elution of contaminant features at all steps of the workflow, statistical heterospectroscopy (SHY) was employed to correlate mass spectrometry data with NMR spectra, enabling isolation of relevant spectral peaks from contaminant noise. Following a stepwise LC purification workflow unknown urine metabolites were successfully extracted at NMR acceptable concentrations acceptable for further NMR analysis and annotated using a combination of MS/MS fragmentation (Waters G2 QTOF) and NMR (Bruker 600MHz and 800 MHz). The workflow is now in place to enable faster annotation of urinary metabolites from large-scale phenotyping projects.