Supporting information
Coupling of Ultrafast LC with Mass Spectrometry by DESI
Yi Cai1, Yong Liu2, Roy Helmy2 and Hao Chen1,3*
1Center for Intelligent Chemical Instrumentation, Department of Chemistry and Biochemistry;3Edison Biotechnology Institute,Ohio University, Athens, OH, USA 45701
2Merck & Co., Inc. One Merck Drive, Whitehouse Station, New Jersey, USA 08889
1. LC/DESI-MS experimental details
Chemicals
HPLC-grade methanol and acetonitrile were purchased from Fisher Chemicals (Pittsburgh, PA) and from EMD Chemicals Inc. (Billerica, MA), respectively. Dopamine hydrochloride, 3-methoxytramine hydrochloride (≥95.5%), L-kynurenine, L-tryptophan (≥98%), codeine, cocaine hydrochloride, flunitrazepam, ketoprofen (≥98%), ibuprofen (≥98%) and fenoprofen calcium salt hydrate were all purchased from Sigma-Aldrich (St. Louis, MO). Cola was Pepsi purchased from Walmart.
LC/DESI-MS apparatus and LC separation conditions
In all monolithic column LC experiments, the mobile phase elution flow rate was kept at 3.0 mL/minand a 20 μL injection loop was used for sample loading.Withsuch a flow rate and 100 μmi.d. orifice being used, the splitting ratio was measured to beca. 4:96, which is subject to small change depending on the composition of mobile phase solvent (probably due to viscosity change from varying solvent composition). For example, elution using 5% and 95% MeOHat the flow rate of3.0 mL/min led to splitting ratios of 4.5:95.5 and 3.9:96.1, respectively.
Phenomenex Onyx Monolithic C18 column (100×4.6 mm) was employed for the separation of dopamine, 3-methoxytyramine, L-tryptophan and L-kynurenine in urine. The urine sample doped with these compounds was diluted 1:1 (v/v) with water to have the final concentration of 60 μM for each species, and then filtered with 0.2 μm Nylon Membrane for removing possible particulates before LC injection.Solvent A was 0.1% FA in H2O, and solvent B was 0.1% FA in ACN with elution program as such:0-1min, 1% B; 1-3min, 1% B was ramped to 50%.
For analysis of acidic anti-inflammatory drug mixture consisting of ketoprofen, fenoprofen and iburopfen(100 μM each), Phenomenex Onyx Monolithic C18 column (100×4.6 mm) was used again. The isocratic elution program was usedH2O:MeOH:FA(30:70:0.05, by volume).
For the UPLC experiments, ACQUITY UPLC® BEH C18 column (50×2.1 mm) was employed for the separation. A cola sample spiked with codeine, cocaine and flunitrazepam(150 μM each) was diluted 10 times by water before injection for UPLC/DESI-MS analysis. The mobile phase consisted of A: 0.1% FAin H2O and B: 0.1% FA in ACN. A linear gradient program ran from 23% to 90% solvent B in 3 min. With the mobile phase elution flow rate kept at 0.3 mL/min, the splitting ratio using the PEEK capillary tube carrying the 100 μmi.d. orifice was 1:1.For the high temperature (80oC) separation of the drug mixture in cola, a gradient program held23% B for 0.3 min then ramped to90% B in 0.7 min with the flow rate of 1.0 mL/min. The splitting ratio was measured to be ca. 3:7.
2. Additional supporting data
Figure 1S.Apparatus of LC/DESI-MS using an orifice (i.d. 50 or 100 μm) drilled on PEEK capillary tubing for splitting
Figure 2S.EICs of a) dopamine (m/z 154), b) 3-methoxytyramine (m/z 168), c) L-kynurenine (m/z 209) and d) L-tryptophan (m/z204) acquired by LC/ESI-MS
Table S1. Data showing LC/DESI-MS sensitivity for the urine sampleanalysis
Compounds / Injection concentrations / ng to MS / Signal/Noise (S/N)*dopamine / 450 ng/mL / 0.36 / 5
3-methoxytyramine / 250 ng/mL / 0.20 / 7
L-tryptophan / 620 ng/mL / 0.50 / 10
L-kynureine / 600 ng/mL / 0.48 / 14
*The S/N was measured based on the ratio of the signal height (SIM signal in our experiment) to the standard deviation of blank signal height (noise).
NMR, UV and IR of collected 3-methoxytyramine
With 3-methoxytyramine collected at hand via MS-directed purification after LC separation, 1H NMR, UV and IR spectra of the collected explicitly were acquired, confirming its structure. In the 1H NMR spectrum, the methoxyl hydrogens -OCH3 give rise to a singlet peak at δ=3.861 ppm; CH2-CH2 hydrogens produce two triplet peaks at δ=3.141 ppm and at δ=2.865 pp, respectively. Three peaks at 6.695-6.846 ppm correspond to the three aromatic hydrogens.The maximum UV absorption of 3-methoxytyramineoccurs at 282 nm. In the IR spectrum, major peaks observed include 3141 cm-1(O-H stretch), 1525 cm-1 (aromatic ring C-C stretch), 1156 cm-1(C-O stretch), 1278 cm-1(C-N stretch) and 799 cm-1 (N-H wag).
Improving the recovery yield of the MS-directed purification
In this study, with further reducing the orifice i.d. from 100 μm down to 50 μm, the splitting ratio of 0.9:99.1 (approximately 1:99)was obtained when the mobile phase elution flow rate was kept at 3.0 mL/min. The use of the smaller orifice would improve the sample collection yield up to 99% with only 1% sample being consumed for DESI-MS detection. Experiment was conducted to testify this hypothesisand the mixture of dopamine, 3-methoxytyramine, L-tryptophan and L-kynureninewas again chosen as the test sample. Using the smaller orifice (50 μmi.d.), the four samples were successfully detected online by DESI-MS (data not shown). During the run of LC/DESI-MS, we collected 3-methoxytyramine from the PEEK capillary tube outletand quantified the collection yield using UV absorption spectroscopy.
LC with a UV detector (detection wavelength was selected at 210 nm) was used for the quantification of the DESI-MS-directed collection yield. 3-Methoxytyramine standard stock solutions were diluted into the standard solutions with five different concentrations (1, 2, 5, 10, and 20 μM), which were analyzed three times by LC/UVusing the same LC separation condition described in text using monolithic C18 column to establish the calibration curve (Figure 3S). The collected 3-methoxytyraminesolution was also injected for LC/UV analysis in triplicate measurements. The peak areas of collected 3-methoxytyraminesolutions were brought into the regression equation of the calibration curve to get the concentrations of the collected 3-methoxytyraminesolutions (Table 2S). Then, based on the collected volume, the moles of collected samples can be obtained to compare with the original amount of injected 3-methoxytyraminefor LC separation to obtain the recovery yield of 98.6±0.6% on average (Table 2S). This collection yield is fairly close to the actual splitting ratio value of 99.1%.
Figure 3S.Calibration curves of 3-methoxytyramine(UV absorption peak area vs. 3-methoxytyramine solution concentration)
Table 2S. Quantification results for the collection yield
3-MT stands for “3-methoxytyramine”.
Figure4S.EICs of a) codeine (m/z 300), b) cocaine (m/z304) and c) flunitrazepam (m/z314) acquired by UPLC/DESI-MS
Figure 5S.EICs of a) codeine (m/z 300), b)cocaine (m/z304) and c) flunitrazepam(m/z314) acquired by UPLC/ESI-MS
When elevated column temperature (80 oC) was used, the elution flow rate for UPLC separation was increased to 1.0 mL/min and the separation was completed in 45 s (Figure 3). In addition, the separated drugs were also online collected with the aid of DESI-MS detection and re-analysis of the collected drugs was performed using ESI-MS. The ESI-MS spectra shown in Figure 6S confirm that purified individual drugs were collected.
Figure 6S.ESI-MS spectra of a) collected codeine, b) collected cocaine and c) collected flunitrazepam after high temperatureUPLC separation
1