Fig 1.
Experimental scheme showing potential cell fates.
Upon exposure to cocaine in a microfluidic bioreactor, naïve and cocaine-experienced cells present different exometabolomic profiles, demonstrated here as color change. Our experiments were designed to determine whether cocaine-experienced cells went to a conditioned state A that was different from state B reached by naïve cells.
Fig 2.
Cocaine exposure scheme for both in-culture and online cell experiments.
The time course of cocaine administration to naïve (blue) and experienced (green) T-cell populations is shown. For the in-culture experiments, experienced samples 1E-4E and naïve samples 1N-4N were withdrawn for analysis at the times shown.
Fig 3.
Solid phase extraction desalter.
Setup starting from initial sample effluent flow incorporates two sample loops, three valves, and two C18 columns. During (A), sample effluent fills sample loop 2 for 9 minutes, while the aqueous solvent flows through sample loop 1, over column 1, and to waste. The organic solvent flows over column 2 and to the mass spectrometer. (B) Upon switching of the valves, the sample effluent fills sample loop 1 for 9 minutes. Aqueous solvent forces the 1.5 μL head of aqueous solvent, the 4.5 μL of sample effluent, and an additional 2.1 μL of aqueous solvent over column 2 to equilibrate the column, load the sample, and rinse away the salts. Organic solvent runs through column 1 and to the detector. (C) The next valve switch again exchanges the sample loop filled by effluent, while column 1 is equilibrated, loaded, and rinsed. The analytes captured on column 2 are eluted by the organic solvent and sent to the detector. (D) When the valves switch again, the sample effluent fills the opposite loop, column 2 is equilibrated, loaded, and rinsed, and column 1 is eluted with organic solvents and those analytes are sent to the detector. The pattern repeats until the experiment is completed, with each cycle requiring 9 minutes.
Fig 4.
Multitrap Nanophysiometer (MTNP)-solid phase extraction (SPE)-nanoelectrospray ionization (nESI)-ion mobility (IM)-mass spectrometry (MS) platform.
The platform includes Harvard Apparatus syringe pumps, a Nikon Eclipse Ti-e inverted fluorescence microscope with stage incubator, a solid phase extraction desalter with Eksigent and nanoAcquity pumps for solvent flow control, and a nanoelectrospray ionization source for continuous flow sample analysis with the Waters Synapt G2 ion mobility-mass spectrometer. Beyond the initial experimental setup, all components are fully automated and capable of running multi-hour experiments limited by the lifetime of cells in the MTNP.
Fig 5.
Exometabolomic profiles of naïve and cocaine-experienced Jurkat T cells from online cellular analysis.
Walking principal component analyses of exometabolomic profiles of naïve (shades of yellow) and cocaine-experienced (shades of turquoise) Jurkat T cells were constructed following online cellular analysis. Each numerically labeled data point represents a 9-minute column elution, with six data points collected per step (all marked with the step number and connected in order of collection with the dotted line). In steps 1 and 3, both cell populations received plain RPMI. In steps 2 and 4, both received cocaine in RPMI at 60 μg/mL. As profiles switch from RPMI to cocaine exposure, the data points move towards the right and vice versa, with the exception of naïve cell step 4, which stays closer to step 1. Further analysis of the data reveals this inconsistency may be explained by the death of the cells. Data were grouped not only based on the experimental step, but also by the experience level of the cells, as the cells receiving a 4-hour pre-incubation in cocaine group separately from those that did not receive this dose.
Fig 6.
Benzoylecgonine (BE) time course and fragmentation data.
Top: BE time course data for experienced (blue) and naïve (purple) cells. While data were gathered sequentially, plots are overlaid to highlight the increased abundance of BE in experienced cells. The absence of the expected increase in BE corresponding to step 4 (the last step of cocaine exposure) suggests a decrease in cocaine metabolism, possibly due to cell death. Bottom: The fragmentation spectra of BE are shown with parent ion of m/z 290.
Fig 7.
Metabolism of cocaine showing molecular weight for each metabolite.
The linkages in this network were adapted from Xia et al. [30]. Putatively identified species are shown in bold.
Fig 8.
Additional metabolite time course data compared to benzoylecgonine (BE).
Experimental conditions for each group of cells are shown above the graph with solid black lines indicating exposure to cocaine media. Anhydroecgonine (AHE) and hydroxybenzoylecgonine (HOBE), two additional metabolites of cocaine, provide examples of both variation between naïve and experienced cell groups in the case of AHE and consistency between these two groups in the case of HOBE. The increases in BE and AHE from naïve to experienced groups are statistically significant with respective p values of 5.7 x 10–4 and 1.12 x 10–3. Three unidentified metabolites (m/z 645, m/z 478, and m/z 330) that contribute to the separation between media exposure groups are also shown. m/z 330 and m/z 478 show no statistical significance between naïve and experienced cell groups while the increase in m/z 645 is statistically significant (p = 1.6 x 10–3).
Fig 9.
Exometabolomic profiles of naïve and cocaine-experienced Jurkat T cells from in-culture analysis.
Walking principal component analysis of exometabolomic profiles of naïve (shades of yellow) and cocaine-experienced (shades of turquoise) Jurkat T cells using in-culture analysis. Each numerically labeled data point represents the end point of a 54-minute exposure to the indicated media. As with the online experiment, during steps 1 and 3 both cell populations received plain RPMI, while during steps 2 and 4 both received cocaine in RPMI at 60 μg/mL. Variance between steps 1 and 3 and steps 2 and 4 results in unsupervised grouping according to media exposure type. Separation between naïve and experienced cell groups, particularly when under cocaine exposure, does not occur to the same degree, as yellow and turquoise points corresponding to steps 2 and 4 are heavily intermixed.