Self-aspiration sampling extractive electrospray ionization mass spectrometry (EESI-MS) for high-throughput analysis of liquid samples

Transcription

1 Supplement Article Published online in Wiley Online Library Rapid Commun. Mass Spectrom. 2016, 30 (Suppl. 1), (wileyonlinelibrary.com) DOI: /rcm.7616 Self-aspiration sampling extractive electrospray ionization mass spectrometry (EESI-MS) for high-throughput analysis of liquid samples Bin Jia 1, Shuai Zhang 2, Ling Yan 1, Shoubo He 3, Baohong Liu 1, Huali Shen 1 * and Pengyuan Yang 1 * 1 Department of Chemistry and Institutes of Biomedical Sciences, Fudan University, Shanghai , China 2 Department of Anatomy, Histology and Embryology, School of Basic Medical Sciences, Fudan University, Shanghai , China 3 Institute of Modern Physics, Chinese Academy of Sciences, Lanzhou , China RATIONALE: Extractive electrospray ionization mass spectrometry (EESI-MS) was invented as a typical ambient mass spectrometry method (AMS) and has been used for analyzing complex liquid samples. Here, we designed a Venturi effect-based self-aspiration sampling device and applied it to the EESI-MS for high-throughput analysis of liquid sample. METHODS: A special concentric nebulizer was designed and employed to produce a suction force for the direct aspiration of liquid samples, followed by ionization and detection. This sample aspiration process was explained and optimized using computational fluid dynamics (CFD) analysis. Experiment data were recorded to exhibit the sensitivity, memory effect, inter-day reproducibility, throughput, and applicability of the self-aspiration sampling EESI-MS. RESULTS: The limit of detection (LOD) of this method was determined as g/ml (S/N = 3) for caffeine, and the sample throughput and relative standard deviation (RSD) for full scan mode can reach 0.67 samples/s and 4.76%, respectively. Even for MS/MS mode, the frequency can still be kept at 0.4 samples/s (RSD = 4.71%). Inter-day RSD examined in 1 week was below 10% for the signal of characteristic fragment ions of reserpine. Moreover, based on this method, the amount of caffeine in instant coffee was determined as 4.7%. This device was also proven to be suitable for the protein/peptide analysis. CONCLUSIONS: These experiment results, especially the amazing results on sample throughput and inter-day RSD, suggest that we provide a valuable device which can be used for the direct high-throughput qualitative/quantitative mass spectrometry analysis of real liquid samples in ambient. Copyright 2016 John Wiley & Sons, Ltd. 56 Ambient mass spectrometry (AMS) [1 3] has promoted the development of mass spectrometry (MS)-based high-throughput screening technology, [4 6] for its ability in rapid and direct analysis of various samples. Desorption electrospray ionization (DESI), [7 10] the first reported ambient ionization method, has been demonstrated in the high-throughput analysis of pharmaceutical samples, [11] immediately after its invention. During the same time period, direct analysis in real time (DART) [12 15] was invented, and several sample injection devices based on DART have been developed to increase the method s detection speed and throughput. [16,17] Both DESI and DART are mainly used to detect volatile, semi-volatile, or easily desorbed compounds from a solid surface, but they are inconvenient for liquid samples which need to be smeared on special surfaces. Extractive electrospray ionization (EESI) was invented originally for the direct detection of complex liquid samples [18,19] and has recently been successfully used in the * Correspondence to: P. Yang and H. Shen, Department of Chemistry and Institutes of Biomedical Sciences of Shanghai Medical School, Fudan University, Shanghai , China. pyyang@fudan.edu.cn; shenhuali@fudan.edu.cn analysis of exhaled gasses, [20,21] viscous samples, [22 24] solid surfaces, [25 28] etc. With an improved nanoeesi-ms method, Hu et al. [29] achieved the rapid analysis of a single sample within only 1.2 s. However, in nanoeesi-ms experiments, the solution must be poured into a spray bottle before MS analysis. This time-consuming operation brings concern about sample contamination, and the stability of results strongly depends on the proficiency of the operator. Furthermore, the sample volume is too small to make a continuous analysis. The famous Venturi effect [30] has already been used in a few self-pumping devices for conventional ESI-MS [31] and was proposed for AMS by Santos and coworkers [32,33] based on sonic-spray ionization (SSI). In this work, we designed a Venturi self-pumping-based device and combined it with EESI-MS. This can then be used for aspirating and analyzing liquid samples directly from their own normal containers, realizing amazing sample throughput and inter-day reproducibility in the high-throughput qualitative/quantitative AMS analysis. For these advances, the improved EESI-MS method shows promise for practical application in analysis of liquid samples for rapid screening and online monitoring. [34,35] Rapid Commun. Mass Spectrom. 2016, 30 (Suppl. 1), Copyright 2016 John Wiley & Sons, Ltd.

2 High-throughput anlaysis by improved EESI-MS EXPERIMENTAL Reagents and solutions Methanol (A.R. grade) was purchased from the Chinese Chemical Reagent Co. Ltd. (Shanghai, China). Cytochrome C from equine heart and caffeine were purchased from Sigma-Aldrich, Inc. (Milwaukee, WI, USA). Reserpine was purchased from AB Sciex Pte. Ltd. (Foster City, CA, USA), and the TP4 tryptic peptide test mixture was purchased from Michrom Bioresources Inc. (Auburn, CA, USA). All chemicals were directly used without any pretreatment except for the dissolution and dilution with deionized water (18.1 MΩ cm) when needed. Design of self-aspiration sampler As depicted in Supplementary Fig. SI-1 (Supporting Information), the spray nozzle of the self-aspiration sampler was inner cone shaped with an angle of 90 (the γ in Supplementary Fig. SI-1B, Supporting Information) to facilitate the formation of a low-pressure region. The selfaspiration sampler used a simple Swagelok T-element with appropriate ferrules and stainless steel needle for the gas flow (i.d. = 0.5 mm, o.d. = 1.0 mm) and a fused-silica capillary (i.d. = 0.2 mm, o.d. = 0.32 mm) at the nebulizing exit for liquid flow. This fused-silica capillary was 0.4 m long, and the other end of it was inserted into the sample container (placed behind the sampling device, about 0.2 m lower than the spray nozzle). Briefly, when the N 2 gas (0.3 MPa, measured at the outlet of the nitrogen tank) was blown out, the static pressure in the region around the outlet of the silica capillary decreased below the atmospheric pressure, as explained by the Venturi effect, [30,32] and the analyte or spray solution was sucked out and nebulized. To further explain and optimize the sampling device, full 3D numerical simulations were performed to model the gas flow velocity and static pressure in the spray nozzle of the sampler using computational fluid dynamics (CFD) software (Ansys Workbench 2.0 Framework, version ). The details for establishing and operating the numerical model and CFD software are described in the Supporting Information. In our experiments, the axial distance ( Δd shown in Supplementary Fig. SI-1B) between the two ports of the silica capillary and the steel capillary and the shape of the sample spray nozzle were optimized by CFD analysis. Depending on the optimization results (see Supplementary Figs. SI-2 and SI-3, Supporting Information), we employed the inner cone-type nozzle for the self-aspiration sampling device, and the Δd value was kept at about 200 μm. EESI-MS The experiments were carried out using a commercial linear ion-trap mass spectrometer (LTQ-XL, Thermo Fisher Scientific, San Jose, CA, USA) equipped with the selfaspiration sampling EESI source. As shown in Fig. 1, methanol/water solution (1/1, v/v) was injected into the ESI emitter to generate the primary charged droplets. The designed self-aspiration sampler was used as the sample channel for EESI. The distance (a) between the spray tips of the EESI source and the MS inlet was 5 mm, and the distance (b) between the end-tips of the two sprays was 2 mm. The angle (α) formed between the ESI source and the inlet capillary of the LTQ-MS instrument was 150, and the angle (β) formed between the ESI spray and the neutral sample spray was 60. In the ESI emitter, the voltage was +3.5 kv, and the flow rate of solvent was 5 μl/min. The temperature of the heated capillary of the LTQ-MS was maintained at 200 C. A full scan mass spectrum was recorded with an average of ~30 scans, and the background was subtracted. In the tandem mass spectrometry experiments, collision-induced dissociation (CID) was performed with a 1.4 Da mass-to-charge window width and a 15 37% collision energy (CE). Figure 1. The integral schematic of the self-aspiration sampling EESI source. 57 Rapid Commun. Mass Spectrom. 2016, 30 (Suppl. 1), Copyright 2016 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/rcm

3 B. Jia et al. Sample analysis In the quantitative analysis experiment, a series of caffeine solutions (0.1, 0.5, 1, 5, 10, and 50 ng/ml) were used to establish the quantitative standard curve. For actual sample analysis, a standards addition method based on the selfaspiration sampling EESI-MS was used to quantify the concentration of caffeine in instant coffee. The instant coffee was dissolved in water and diluted to 50 ng/ml. Then, the standard caffeine was spiked to obtain coffee samples spiked with 0.5, 1, 2.5, 5, 10, 25, and 50 ppb caffeine. The memory effect was evaluated by alternately introducing a row of standard solutions (caffeine and reserpine were used respectively, 100 ng/ml) and blank solutions to the self-aspiration sampling EESI-MS. Both full mass scan mode and MS/MS mode were used to evaluate the memory effect. Inter-day reproducibility of the device was assessed with 100 ng/ml reserpine. High-throughput experiment The high-throughput analysis experiments were carried out by successively introducing samples of reserpine solutions (50 ng/ml). To improve the data collection efficiency, the maximum injection time of the LTQ was set to 1 ms. Upon CID, the maximum injection time was set to 50 ms for recording stable MS/MS spectra. RESULTS AND DISCUSSION Qualitative/quantitative analysis The mass spectrum of the caffeine solution was recorded in positive ion detection mode, as shown in Fig. 2(A). The peak at m/z 195 corresponds to the quasi-molecular ions [M + H] + of caffeine. MS/MS experiments were further performed, and the MS 2 spectrum of m/z 195 is shown in Fig. 2(B). The major fragment ions were detected at m/z 138 and m/z 110 and were assigned as the characteristic fragments of protonated caffeine generated by the loss of CH 3 NCO and CO, respectively. This CID spectrum agrees with the previous experimental results from tandem linear ion traps, [29] supporting the confirmation of caffeine. The intensity of the characteristic fragment ion at m/z 138 vs the concentrations shows a good linear relationship with a linear fit correlation coefficient of As an example by using this reported method, a linear dynamic range from 0.1 to 50 ppb was illustrated for quantitative analysis of caffeine. The limit of detection (LOD) of caffeine was calculated to be 0.45 ng/ml, corresponding to a concentration with a threefold increase in abundance (at m/z 138) compared with the blank measurements, as depicted in Supplementary Fig. SI-4 (Supporting Information). Furthermore, the self-aspiration sampling EESI method was applied to determine the caffeine concentration of instant coffee sample by the standards addition method. 58 Figure 2. Direct analysis of a caffeine standard solution using the self-aspiration sampling EESI-MS. (A) MS spectrum of caffeine (50 ppb); (B) MS/MS spectrum of protonated caffeine (m/z 195); and (C) linear range for caffeine determination using the signal of the characteristic fragment ions at m/z 138; curve I, recorded from caffeine standard solution; curve II, recorded from diluted coffee spiked with caffeine. wileyonlinelibrary.com/journal/rcm Copyright 2016 John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2016, 30 (Suppl. 1), 56 61

4 High-throughput anlaysis by improved EESI-MS The plot of the abundance of the characteristic fragment ion (m/z 138) vs the concentrations of spiked caffeine was recorded as curve II in Fig. 2(C). Based on this curve, the original concentration of caffeine was calculated to be 2.37 ng/ml in the diluted coffee solution and approximately 4.7% (w/w) in the instant coffee powder. Memory effect Memory effects are of great concern in high-throughput analysis. As shown in Supplementary Fig. SI-5 (Supporting Information), the selected ion current (SIC) of m/z 195 indicated that the residual looks very serious. However, in CID mode, the characteristic fragments of caffeine were produced in lower abundance from the ion at m/z 195 in blank solutions, indicating that the response at m/z 195 in the blank sample is a false positive instead of a residual. Additionally, reserpine was used as another standard sample to evaluate the memory effect. The results are shown in Supplementary Fig SI-6 (Supporting Information); an intense peak at m/z 609, corresponding to the quasi-molecular ions [M + H] + of reserpine, was detected. Also, the signal of reserpine was not observed in the blank sample. Actually, the memory effect of the self-aspiration sampling EESI device was faint and can be easily eliminated by introducing a blank solution. High-throughput analysis of samples As depicted in Fig. 3(A), a total ion current plot with 40 injection peaks was recorded for the full mass spectra of the reserpine solutions in a time range of min. With a speed of 0.67 samples/s, the relative standard deviation (RSD) of the signal at m/z 609 for all 40 liquid samples was calculated to be 4.76%. More details are shown and discussed in Supplementary Fig. SI-7 (Supporting Information). Upon CID, the major fragment ions were detected at m/z 397, which are assigned as the characteristic fragments of protonated reserpine that are generated by the loss of C 10 H 11 O 5. [36,37] Thus, the sampling frequency was sacrificed to achieve reliable MS/MS spectra in the high-throughput analysis. Twelve reserpine MS/MS spectra were successfully recorded at a speed of 0.4 samples/s, as shown in Fig. 3(B), with an RSD of 4.71% for the characteristic fragment signal at m/z 397. Even when considering the alternatively added blank to reduce possible memory effect in real analysis, the halved frequency of the sampling throughput is still quite amazing. A group of different samples, such as caffeine, reserpine, and blank water, were also used to test the throughput, as shown in Supplementary Fig. SI-8 (Supporting Information). The throughput for different samplesis0.56samples/sor0.28samples/s(takingwater as blank). Figure 3. The high-throughput experiments. (A) The total ion current of the continuous detection of 40 reserpine solutions in 1 min and (B) the highthroughput quantitative analysis results, including the MS/MS spectra of m/z 609 and selected ion current of characteristic fragment ions at m/z 397 produced from m/z Rapid Commun. Mass Spectrom. 2016, 30 (Suppl. 1), Copyright 2016 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/rcm

5 B. Jia et al. 60 Moreover, the inter-day reproducibility of the selfaspiration sampling EESI-MS method was examined by repeating the experiment for 7 days. The inter-day RSD in 1 week was below 10% (9.6%), as shown in Supplementary Fig. SI-9 (Supporting Information). In addition to small molecules, peptides and proteins can also be analyzed by self-aspiration sampling EESI. Here, the TP4 tryptic peptide test mixture (TP4) and cytochrome c (Cyt C) were tested on the device. As shown in Supplementary Fig. SI-10 (Supporting Information), the doubly charged ions [M + 2H] 2+ for all four peptides in this mixed solution were detected, and the multiply charged ions of Cyt C were also recorded in a full mass spectrum with charge states from +7 to +16. All these results indicate that the self-aspiration sampling EESI device has high sample throughput, high repeatability, wide applicability, faint memory effect, and has no significant effect on the ionization efficiency and sensitivity of EESI-MS. Thus, the self-aspiration sampling EESI device is very suitable for the direct high-throughput qualitative/quantitative mass spectrometry analysis of real liquid samples in ambient. CONCLUSIONS According to CFD optimization calculation results, we designed a Venturi effect-based self-aspiration sampling device and combined it with EESI-MS experimentally. It was proved that this self-aspiration sampling EESI-MS method has relative low RSD and LOD, faint memory effect, and amazing sample throughput. This proposed self-aspiration sampling EESI-MS method shows promise for practical application in high-throughput analysis of liquid samples for rapid screening and online monitoring. Acknowledgements This work was supported by the MOST 863 (2014AA and 2014AA020902) and the NSF of China ( ). The precision-regulating device for the EESI was supplied by Prof. Huanwen Chen of East China Institute of Technology, Jiangxi, China. REFERENCES [1] R. G. Cooks, Z. Ouyang, Z. Takats, J. M. Wiseman. Ambient mass spectrometry. Science 2006, 311, [2] J. S. Wiley, J. T. Shelley, R. G. Cooks. Handheld lowtemperature plasma probe for portable point-and-shoot ambient mass spectrometry. Anal. Chem. 2013, 85, [3] R. M. Alberici, R. C. Simas, G. B. Sanvido, W. Romão, P. M. Lalli, M. Benassi, I. B. Cunha, M. N. Eberlin. Ambient mass spectrometry: bringing MS into the real world. Anal. Bioanal. Chem. 2010, 398, 265. [4] L. Li, B. Feng, J. Yang, C. Chang, Y. Bai, H. Liu. Applications of ambient mass spectrometry in high-throughput screening. Analyst 2013, 138, [5] S. C. Cheng, M. Huang, L. C. Wu, C. C. Chou, C. N. Cheng, S. S. Jhang, J. Shiea. Building blocks for the development of an interface for high-throughput thin layer chromatography/ ambient mass spectrometric analysis: a green methodology. Anal. Chem. 2012, 84, [6] L. Shen, J. Zhang, Q. Yang, N. E. Manicke, Z. Ouyang. High throughput paper spray mass spectrometry analysis. Clin. Chim. Acta 2012, 420, 28. [7] Z. Takats, J. M. Wiseman, B. Gologan, R. G. Cooks. Mass spectrometry sampling under ambient conditions with desorption electrospray ionization. Science 2004, 306, 471. [8] D. R. Ifa, N. E. Manicke, A. L. Dill, R. G. Cooks. Latent fingerprint chemical imaging by mass spectrometry. Science 2008, 321, 805. [9] C. R. Ferreira, L. Wu, F. G. Vogt, E. R. Bornancini, R. G. Cooks. Fiducial markers for distribution of drug and excipient on tablet surfaces by multimodal desorption electrospray ionization mass spectrometry (DESI-MS) imaging. Anal. Lett. 2014, 41, 91. [10] Z. Takats, J. M. Wiseman, R. G. Cooks. Ambient mass spectrometry using desorption electrospray ionization (DESI): instrumentation, mechanisms and applications in forensics, chemistry, and biology. J. Mass Spectrom. 2005, 40, [11] H. Chen, N. N. Talaty, Z. Takats, R. G. Cooks. Desorption electrospray ionization mass spectrometry for highthroughput analysis of pharmaceutical samples in the ambient environment. Anal. Chem. 2005, 77, [12] R. B. Cody, J. A. Laramee, H. D. Durst. Versatile new ion source for the analysis of materials in open air under ambient conditions. Anal. Chem. 2005, 77, [13] R. B. Cody. Observation of molecular ions and analysis of nonpolar compounds with the direct analysis in real time ion source. Anal. Chem. 2008, 81, [14] R. B. Cody, A. J. Dane. Soft ionization of saturated hydrocarbons, alcohols and nonpolar compounds by negative-ion direct analysis in real-time mass spectrometry. J. Am. Soc. Mass Spectrom. 2013, 24, 1. [15] A. Martínez-Villalba, L. Vaclavik, E. Moyano, M. T. Galceran, J. Hajslova. Direct analysis in real time high-resolution mass spectrometry for high-throughput analysis of antiparasitic veterinary drugs in feed and food. Rapid Commun. Mass Spectrom. 2013, 27, 467. [16] C. M. Jones, F. M. Fernandez. Transmission mode direct analysis in real time mass spectrometry for fast untargeted metabolic fingerprinting. Rapid Commun. Mass Spectrom. 2013, 27, [17] A. H. Grange. Semi-quantitative analysis of contaminants in soils by direct analysis in real time (dart) mass spectrometry. Rapid Commun. Mass Spectrom. 2013, 27, 305. [18] H. Chen, A. Venter, R. G. Cooks. Extractive electrospray ionization for direct analysis of undiluted urine, milk and other complex mixtures without sample preparation. Chem. Commun. 2006, [19] H. Chen, S. Yang, M. Li, B. Hu, J. Li, J. Wang. Sensitive detection of native proteins using extractive electrospray ionization mass spectrometry. Angew. Chem. 2010, 122, [20] G. Gamez, L. Zhu, A. Disko, H. Chen, V. Azov, K. Chingin, G. Krämer, R. Zenobi. Real-time, in vivo monitoring and pharmacokinetics of valproic acid via a novel biomarker in exhaled breath. Chem. Commun. 2011, 47, [21] M. Li, J. Ding, H. Gu, Y. Zhang, S. Pan, N. Xu, H. Chen, H. Li. Facilitated diffusion of acetonitrile revealed by quantitative breath analysis using extractive electrospray ionization mass spectrometry. Sci. Rep. 2013, 3, [22] X. Li, B. Hu, J. Ding, H. Chen. Rapid characterization of complex viscous samples at molecular levels by neutral desorption extractive electrospray ionization mass spectrometry. Nat. Protocols 2011, 6, [23] W. S. Law, H. Chen, J. Ding, S. Yang, L. Zhu, G. Gamez, K. Chingin, Y. Ren, R. Zenobi. Rapid characterization of complex viscous liquids at the molecular level. Angew. Chem. Int. Ed. 2009, 48, wileyonlinelibrary.com/journal/rcm Copyright 2016 John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2016, 30 (Suppl. 1), 56 61

6 High-throughput anlaysis by improved EESI-MS [24] W. S. Law, H. Chen, R. Balabin, C. Berchtold, L. Meier, R. Zenobi. Rapid fingerprinting and classification of extra virgin olive oil by microjet sampling and extractive electrospray ionization mass spectrometry. Analyst 2010, 135, 773. [25] H. Chen, B. Hu, Y. Hu, Y. Huan, Z. Zhou, X. Qiao. Neutral desorption using a sealed enclosure to sample explosives on human skin for rapid detection by EESI-MS. J. Am. Soc. Mass Spectom. 2009, 20, 719. [26] H. Gu, S. Yang, J. Li, B. Hu, H. Chen, L. Zhang, Q. Fei. Geometry-independent neutral desorption device for the sensitive EESI-MS detection of explosives on various surfaces. Analyst 2010, 135, 779. [27] H. Chen, R. Zenobi. Neutral desorption sampling of biological surfaces for rapid chemical characterization by extractive electrospray ionization mass spectrometry. Nat. Protocols 2008, 3, [28] H. Chen, S. Yang, A. Wortmann, R. Zenobi. Neutral desorption sampling of living objects for rapid analysis by extractive electrospray ionization mass spectrometry. Angew. Chem. 2007, 119, [29] B. Hu, X. Peng, S. Yang, H. Gu, H. Chen, Y. Huan, T. Zhang, X. Qiao. Fast quantitative detection of cocaine in beverages using nanoextractive electrospray ionization tandem mass spectrometry. J. Am. Soc. Mass Spectrom. 2010, 21, 290. [30] C. J. Geankoplis. Transport Processes and Unit Operations, (3rd edn.). Prentice-Hall: Englewood Cliffs, NJ, [31] T. Wachs, J. Henion. Electrospray device for coupling microscale separations and other miniaturized devices with electrospray mass spectrometry. Anal. Chem. 2001, 73, 632 [32] V. G. Santos, T. Regiani, F. F. G. Dias, W. Romão, J. L. P. Jara, C. F. Klitzke, F. Coelho, M. N. Eberlin. Venturi easy ambient sonic-spray ionization. Anal. Chem. 2011, 83,1375. [33] N. Liu, X. Lu, Y. H. Yang, C. X. Yao, B. M. Ning, D. C. He, L. He, J. Ouyang. Monitoring binding affinity between drug and alpha 1-acid glycoprotein in real time by Venturi easy ambient sonic-spray ionization mass spectrometry. Talanta 2015, 143, 240. [34] Y. Liu, B. H. Zheng, M. Wang, Y. X. Xu, Y. W. Qin. Optimization of sampling frequency for routine river water quality monitoring. Sci. Sin. Chim. 2014, 57, 772. [35] Q. Z. Yao, Z. G. Yu, L. L. Li, H. T. Chen, T. Z. Mi. Transformation and source of nutrients in the Changjiang Estuary. Sci. Sin. Chim. 2014, 57, 779. [36] J. W. Hager. A new linear ion trap mass spectrometer. Rapid Commun. Mass Spectrom. 2002, 16, 512. [37] J. W. Hager. Product ion spectral simplification using timedelayed fragment ion capture with tandem linear ion traps. Rapid Commun. Mass Spectrom. 2003, 17, SUPPORTING INFORMATION Additional supporting information may be found in the online version of this article at the publisher s website. 61 Rapid Commun. Mass Spectrom. 2016, 30 (Suppl. 1), Copyright 2016 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/rcm