Feasibility of assessment of regulatory lipids in breath condensate as potential presymptomatic harbingers of pulmonary pathobiology

Regulatory lipids from the airway surface readily form aerosols that can be recovered non-invasively by cooling expired breath to form breath condensate (BC). Regulatory lipids have been detected previously utilizing enzyme-linked-immunosorbent serologic assay (ELISA). Here we test the feasibility of assessment of regulatory lipids in BC by mass spectrometry so presently unknown lipid regulatory components can be detected without addition of specific antibodies as in the ELISA procedure. Baseline regulatory lipids were detected in > pg/mL BC in control animals or human lung tissue culture cells. In nearly every case animals exposed to toxins or infectious bacteria showed increases in the BC regulatory components. Lipids were recovered from BC by solid phase extraction. Phosphatidylcholine (PC) based lipids were detected as the progenitor (parent) ions of isomers that fragmented in producing product positive ions at m/z 184 (of phosphocholine) in tandem MS using capillary HPLC and electrospray ionization. BC eicosanoids such as prostaglandins, thromboxane, and isoprostanes require capillary gas chromatography for separation and detection that necessitates methoximation, pentafluorobenzyl (PFB) ester formation, and trimethyl silylation of hydroxyls prior to gas chromatography/ion trap tandem mass spectrometry of negative ions after chemical ionization (NICI). Tetradeuterated internal standards were utilized for quantitation with the GC/NICI/MS. Changes in concentrations of lipids and eicosanoids were observed in piglets, and rats exposed to aerosolized 100 mu g/kg lipopolysaccharide (LPS), or 50 mu g/kg and 150 mu g/kg aerosolized Staphylococcal enterotoxin B (SEB) in BC as well as in human THP-1 cell culture cell supernatants and bronchoalveolar lavage (BAL) samples in rats. Responses of the molecular species of phosphatidylcholines (PCs), platelet activating factors (PAFs) and specific eicosanoids correlated to the toxin and bacterial infections suggesting that patterns of differential responses could be detected with further experimentation. Initial targets included prostaglandins (PGE(2), PGF(2 alpha)), thromboxane (TXB2), and prostacyclin (as 6-Keto PGF(1 alpha)) that show differential responses to inflammation, the leukotriene (LTB4) and PGD(2) for allergic responses, isoprostanes (8-iso-PGF(2 alpha)) for free radical oxidative stress responses, and HETEs for differential lipoxygenase activities. PAFs and lysoPAFs have been shown to increase with inflammation and in the feasibility experiments reported here. Preliminary studies show pulmonary responses of piglets to intrathecal exposure of toxicants (LPS and SEB) or infections with Actinobacillus pleuropneumoniae induce increased levels of lipids and two eicosanoids with the suggestion that differential patterns might be detected with expanded testing. Preliminary experience indicates numerous other eicosanoids were available for assay in BC. This suggests an important potential application of BC to observe a wide array of factors to establish comprehensive profiles for physiological and pathophysiological states. Ultimately this technique could be used as a non-invasive possibly presymptomatic assessment of pulmonary pathobiology. (c) 2005 Elsevier B.V. All rights reserved.