J. Will Thompson (Presenter)
Bio: I received my Ph.D. in Analytical Chemistry from the University of North Carolina Chapel Hill under Prof. Jim Jorgenson, studying extensions of UHPLC theory and practice. In 2007, I assisted Professor Arthur Moseley in founding the Duke Proteomics Core Facility, which is now the Duke Proteomics and Metabolomics Shared Resource. Currently, I am Assistant Director of the shared resource and Assistant Research Professor in the Department Pharmacology and Cancer Biology.
Authorship: J. Will Thompson(1), Lisa St. John-Williams(1), David M. Brass(1), Daniel W. Morgan(2), M. Arthur Moseley(1), Scott M. Palmer(1), and Matthew W. Foster(1)
(1) Duke University School of Medicine; (2) National Institutes of Environmental Health Sciences
Occupational exposure to diacetyl (butanedione) vapor, utilized as artificial butter flavoring, has been shown to induce fibrotic lung disease called bronchiolitis obliterans (BO), yet there are no methods to monitor chronic occupational exposure. In an effort to develop a group of circulating biomarker of occupational exposure, two separate quantitative metabolite methods were utilized to profile plasma from rats exposed to 125 mg/Kg diacetyl or sterile H2O as a control via intratracheal instillation. A number of putative biomarkers of DA exposure were apparent in the plasma, including increases in long chain acylcarnitines, differentiation in a marker of oxidative stress (Met:Met-SO ratio), and decrease in proline:lysine ratio.
Occupational exposure to 2,3-butanedione (diacetyl) vapor, utilized as artificial butter flavoring in popcorn and e-cigarettes, has been shown to induce fibrotic lung disease called bronchiolitis obliterans (BO), yet there are no methods to monitor chronic occupational exposure. In an effort to measure changes in serum metabolites and develop a group of circulating biomarkers of exposure or progression to BO, two separate quantitative metabolite methods were utilized to profile plasma from a rat model exposed to intrathecal instillation of 2,3-butanedione (diacetyl, DA) versus water as control.
Rats were exposed to 125 mg/Kg diacetyl or sterile H2O as a control via intratracheal instillation. On day 7 post-exposure we collected blood, bronchoalveolar lavage (BAL) and lung tissue. Lungs were harvested, fixed in formalin, and H&E stained. Blood was collected from each animal by cardiac puncture into tubes containing EDTA, and spun to collect plasma from each animal. Plasma from 11 animals (n=5 control, n= 6 DA) was analyzed using the AbsoluteIDQ p180 quantitative metabolomics platform (Biocrates, Inc) according to manufacturer protocol, with data collection performed on an Acquity UPLC coupled to Xevo TQ-S mass spectrometer. LC-MS/MS methodology was utilized to measure levels of amino acids and biogenic amines, while flow-injection (FIA) – MS/MS was used to quantify acylcarnitines, glycerophospholipids and sphingolipids. To provide a more detailed lipid analysis, 50 uL plasma was extracted with MeOH/MTBE, and high-resolution relative quantification of lipids was performed by ESI+ and ESI- LC-MS/MS on an Acquity UPLC coupled to Synapt G2 high resolution mass spectrometer. These methods allowed independent analysis of lipid classes including glycerophospholipids, sphingolipids, mono-, di- and triacylglycerols, cholesteryl esters, and free fatty acids. Data analysis was performed in several packages, including Targetlynx (Waters), Progenesis QI (Nonlinear Dynamics), MetIDQ (Biocrates), Skyline, and JMP Pro (SAS Institute).
Data QC was performed in part by measuring a Study Pool QC (SPQC) in triplicate at beginning, middle, and end of each analytical platform. For the AbsoluteIDQ p180 analysis, 150 analytes were reliably measured with %CV < 25%, with a mean %CV of 2.5% and median %CV of 1.5%. For high resolution lipidomics profiling, the ESI+ data measured approximately 7,000 lipid features with %CV < 25% (mean = 9.5%), and the ESI- data measured an additional 2,450 lipid features with %CV < 25% (mean = 11.3%).
To validate the model based on known pathology of BO, we observed loss of clara cell secretory protein (CCSP) in airway lining fluid, and the presence of BO lesions were confirmed in rats treated with DA vapor versus controls by hemotaxin and eosin (H&E) staining. There was no change observed in total plasma protein. Plasma long-chain acylcarnitines (C14-C18) were increased in DA exposed animals, and palmitoylcarnitine has recently been shown to decrease lung function by inhibiting pulmonary surfactant. No change was observed in free carnitine (Otsubo et al, J. Biol. Chem. 2015). Circulating proline levels decreased (p = 2e-3), consistent with proline being utilized for generation of fibrotic collagen. Lysine levels increased (p = 4e-3), potentially due to decreased activity of the lysine degradation pathway, which utilizes proline as a substrate. Consequently, the Proline/Lysine ratio was highly differentiated between control and DA-exposed animals (2.8-fold, p = 8e-4) and may represent a sensitive and easily translatable biomarker. Methionine sulfoxide (Met-SO), a molecule considered a marker of oxidative stress as a reaction product of methionine (Met) with reactive oxygen species (ROS) was found to decrease with DA exposure (p = 4e-3). Correcting for initial levels of Met by calculating the Met-SO/Met ratio led to a modest improvement in p-value (p = 1e-3). Decreases in unsaturated chain (e.g. 16:1, 18:1, 18:2) lysophsphatidylcholines (LPC) and diacyl phosphatidylcholines (PC aa) were consistently observed in both FIA-MS/MS and LC-MS/MS data in DA treated animals (p~2e-3 to 2e-4). The levels of each of the free fatty acids was inversely correlated with the corresponding LPC, showing significant increases with DA. This suggests that the decrease in LPC content is not due to limitation in FFA substrates but instead caused by either decreased choline or dysfunction of LPC synthesis with DA exposure.
Plasma metabolites are able to differentiate animals with diacetyl-induced BO lesions from corresponding controls. Targeted assays for these plasma metabolites can easily be deployed for validation of flavoring-induced BO in animal model and human samples.
References & Acknowledgements:
IP Royalty: no
|Planning to mention or discuss specific products or technology of the company(ies) listed above:||