Imaging Laser-Induced Fluorescence Detection at the Taylor Cone of

May 28, 2019 - (15) In early work, Zare and co-workers, using a helium–cadmium laser beam (325 .... (1) Outlet tip of the separation capillary; (2) ...
0 downloads 0 Views 3MB Size
Article Cite This: Anal. Chem. 2019, 91, 7738−7743

pubs.acs.org/ac

Imaging Laser-Induced Fluorescence Detection at the Taylor Cone of Electrospray Ionization Mass Spectrometry Mat́ e Szarka,† Maŕ ton Szigeti,†,‡ and Andraś Guttman*,†,‡ †

Downloaded via UNIV OF SOUTHERN INDIANA on July 23, 2019 at 12:24:40 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Horváth Csaba Memorial Laboratory of Bioseparation Sciences, Research Center for Molecular Medicine, Faculty of Medicine, University of Debrecen, 98 Nagyerdei krt, Debrecen 4032, Hungary ‡ Translational Glycomics Laboratory, Research Institute of Biomolecular and Chemical Engineering, University of Pannonia, 10 Egyetem Street, Veszprem 8200, Hungary S Supporting Information *

ABSTRACT: Laser-induced fluorescence detection (LIF) is a powerful tool for the quantitative analysis of fluorescent molecules, widely used in glycan analysis with fluorophore labeled carbohydrates where each species has a common response factor. Electrospray ionization mass spectrometry (ESI-MS), on the other hand, while revealing important structural information about individual analytes, generally can have different response factors for different species. For simpler and improved quantitation with ESI-MS, laserinduced fluorescent images were collected at the Taylor cone of the electrospray interface, enabling simultaneous and robust optical (quantitative) and MS (qualitative) detection of fluorophore labeled sugars. The performance of this universally applicable, interface design independent imaging laserinduced fluorescent (iLIF) system was demonstrated using capillary electrophoresis (CE)-ESI-MS in the analysis of aminopyrene-trisulfonate labeled linear maltooligosaccharides and branched glycans from human immunoglobulin. The limit of detection (LOD) of the iLIF system was in this case 40 attomole. The intra- and interday quantitative (peak area) reproducibilities of the system (RSD) were 4.15% and 6.79%, respectively.

R

rearrangements during fragmentation (e.g., fucose migration of complex carbohydrates),12 resulting in possibly compromised structural information and also affecting quantitation accuracy. Therefore, UV/vis and fluorescent detection is often used prior to ESI in LC-MS systems. Combination of capillary electrophoresis with electrospray ionization mass spectrometry (CEESI-MS) is an important technique for biomolecular analysis with the ability to reduce ion suppression using low nanoliter per minute flow rates.13 Coupling fluorescence detection with liquid phase separation systems offers high sensitivity for fluorophore labeled and/ or autofluorescent molecules passing through the interrogation zone.14 Considering the specific absorption (excitation) and emission properties of the fluorescent molecules, with carefully chosen excitation and emission wavelengths, it can be one of the most sensitive optical detection techniques. Capillary electrophoresis (CE) coupled to MS (CE-MS) is a rapidly growing bioanalytical method for a wide variety of analytical problems. Laser-induced fluorescence detection has been well established for many years.15 In early work, Zare and coworkers, using a helium−cadmium laser beam (325 nm)

ecently developed high sensitivity mass spectrometers are extensively used in many sectors of analytical chemistry, including the multiple omics fields, clinical and regulatory laboratories, and the large and very active biopharmaceutical industry.1−4 The advent of electrospray ionization (ESI)5,6 has enabled a significant increase in the detectable analyte mass range up to a few million Daltons (e.g., viruses), thus making analysis of intact biopolymers possible.7 During a liquid phase ESI process, a Taylor cone is initially formed, leading to nebulized highly charged droplets, which rapidly shrink as they approach the inlet to the mass spectrometer. When the droplets become sufficiently small and ultracharged, they undergo Rayleigh explosion with the release of sample component ions in the gas phase (electrospray plume) that enter the MS instrument. Under appropriate buffer conditions, electrosprayed large biomolecules may closely maintain their in-solution structure in the gas phase, thereby allowing analysis in their native or close to native forms.8 Mass spectrometer analyzer designs have been developed for high resolution in both MS1 and MSn modes for analysis of diverse types of analytes.9 Ion suppression and nonuniform ionization, however, create significant challenges for ESI-MS based quantification.10 Moreover, analyte molecules can lose certain functional residues during the electrospray ionization process11 or be subject to gas-phase structural © 2019 American Chemical Society

Received: February 26, 2019 Accepted: May 28, 2019 Published: May 28, 2019 7738

DOI: 10.1021/acs.analchem.9b01028 Anal. Chem. 2019, 91, 7738−7743

Article

Analytical Chemistry

connected to the mass spectrometer. CE separation of the APTS labeled linear maltooligosaccharides, maltose, and released branched hIgG1 glycans were conducted using an MS compatible background electrolyte (BGE) containing 10 mM of ammonium acetate (pH 4.5) and 20% of isopropanol. A two stage sample injection method was applied: starting with a water plug by applying 3.0 psi pressure for 5.0 s, followed by electrokinetic injection of the sample using 10 kV for 20 s. The separation was achieved in 30 min, applying 30 kV electric potential (reversed polarity mode, cathode at the injection side) at 30 °C with constant 2.0 psi pressurization from the inlet end. LOD experiments utilized continuous infusion of APTS labeled maltose solutions from 0.2−20 μM concentration. Mass Spectrometry. A 6500+ QTRAP (Sciex) mass spectrometer was used for all CE-MS measurements in negative ionization mode using Q1 multiple ion scan (Q1MI) targeting the doubly charged APTS labeled carbohydrates. The electrospray voltage was set to −1400 V, the orifice plate temperature was 80 °C, and the curtain gas (N2) pressure was 5.0 psi. For data acquisition and processing, the Analyst version 1.6.3 software (Sciex) was used. Imaging Laser-Induced Fluorescence (iLIF) Detection. The fluorescent imaging system was specifically designed as an easy to apply off-column detection system, built on a modified version of our recently published work.38 For this image acquisition-based detection, a “spyglass” design was assembled using a monocular lens ∼3 cm away from the detection zone with the target (the Taylor cone) in the center of the field of view. Considering the wide Stokes shift of the fluorophore (APTS) labeled molecules (excitationmax: 425 nm; emissionmax: 503 nm), the use of a single optical bandpass filter provided sufficient signal. A 405 nm diode laser (5.0 mW, Laserland, Wuhan, PRC) was used to illuminate the detection target zone, as depicted in Figure 2B. The monocular lens collected the emitted light (image) through a 12.5 mm diameter EO520/10 emission filter (Yulong Optics Co, Ltd., Kunming, PRC). It should be noted that only the Taylor cone was imaged for fluorescence detection, excluding the jet and plume regions. The collected and bandpass filtered light reached an 8megapixel CCD camera (Pi NoIR Sony IMX219, Tokyo, Japan) through an attached eye lens (shown in Figure 2). The core of the detector was an ARM cortex Raspberry Pi-3b minicomputer serving as an image processor unit, running on a Raspbian (Raspberry Pi, Cambridge, U.K.) operating system and executing commands from a graphical user interface. Image processing-based detection was carried out by a custom written python script utilizing time-lapse mode (Raspistill library, Raspberry Pi). Images were produced in jpeg file formats. Trigger signals for image-processing were initiated by the controlling CESI 8000 unit. The electropherogram display and analysis scripts were written in ImageJ/Fiji (Wayne Rasband, NIH, Bethesda, U.S.A.) macro languages.

focused onto a separation capillary column, achieved fmol level detection limits for dansyl-amino acids.16 Laser-induced fluorescence (LIF) detection was further optimized by Dovichi and co-workers, who applied a liquid filled region at the outlet end of the capillary in a form of a sheath flow cuvette, leading to subattomole level sensitivity, while eliminating postcolumn band broadening.17 Based on these encouraging results, commercially available systems became widespread for CELIF analyses18 and automated high-resolution DNA sequencing.19−21 Built on its high sensitivity and selectivity, CE-LIF has become a crucial detection method applied in the pharmaceutical industry22−26 and in the biotherapeutics field.27 However, only a few attempts have been made so far to couple CE with MS through fluorescence or UV/vis absorbance detectors.28,29 Albeit some authors have reported progress in the field,30−32 the application of the technology is still lagging due to the lack of an easily applicable system. One of the handicaps of online optical detection for CE-MS has been the long capillary distance between the detection point and the electrospray,33 representing a risk to adequately align the optical and MS traces, particularly for complex mixtures. In the past decade, attempts have been made to utilize the electrospray plume to fluorescently profile solvent fractionation,34 combining fluorescence spectroscopy with electrospray ionization to examine its effects on protein denaturation35 and for laser ablation electrospray ionization.36 However, up to now, no reports have been published on the utilization of the electrospray Taylor cone for quantitative fluorescence detection of analyte molecules separated by liquid phase separation techniques, such as CE, LC, or microchips prior to mass spectrometry. Here, a novel and universally applicable system is introduced, capable to support imaging laser-induced fluorescence (iLIF) detection at the Taylor cone of the electrospray, i.e., immediately prior to the sample components entering the MS orifice. The system allows the essentially simultaneous optical (quantitative) and mass (structural) detection. Imaging the fluorescence at the Taylor cone is shown to provide subnanomolar detection sensitivity in the negative ionization MS mode for fluorescently labeled carbohydrates. The approach is universal, i.e., independent of the ESI interface design.



EXPERIMENTAL SECTION Chemicals. Acetic acid (glacial), ammonium acetate, isopropanol, sodium-cyanoborohydride (1 M in THF), maltose, acetonitrile, and HPLC grade water were obtained from Sigma-Aldrich (St. Louis, MO, U.S.A.). Human immunoglobulin G1 was from Molecular Innovations (Novi, MI, U.S.A.). The Fast Glycan Labeling and Analysis Kit, including the tagging dye of aminopyrenetrisulfonic acid (APTS), the maltooligosaccharide ladder, and the magnetic beads for excess dye removal were from Sciex (Brea, CA, U.S.A.). The PNGase F enzyme was purchased from Asparia Glycomics (San Sebastian, Spain). Sample Preparation. A 1.0 mg of maltooligosaccharide mixture, 1.0 mg of maltose, and PNGase F released N-glycans from 1.5 mg of human Immunoglobulin G1 (hIgG1) were APTS labeled following the procedure published earlier by Szigeti et al.37 Capillary Electrophoresis. A CESI 8000 Plus High Performance Separation - ESI Module was used for all separations with bare fused silica surface OptiMS capillaries (both from Sciex), employing a porous sprayer that was



RESULTS AND DISCUSSION In this work, imaging laser-induced fluorescent detection with a capillary electrophoresis−mass spectrometry electrospray ionization interface (CE-ESI-MS) was accomplished by focusing the laser on the liquid at the outlet end of the capillary, where a Taylor cone was formed with the application of high voltage. As illustrated in Figure 1, the Taylor cone (2) was fluorescently imaged (5−6), with simultaneous collection of mass data information (4). 7739

DOI: 10.1021/acs.analchem.9b01028 Anal. Chem. 2019, 91, 7738−7743

Article

Analytical Chemistry

the electrospray at the outlet end of the separation capillary. Figure 2A shows, in the light path, the objective lens target (4) with a bandpass filter (3), an eye lens (2), and a CCD camera (1). Fluorescent excitation at the Taylor cone was achieved via illumination by a 405 nm laser (2 in Figure 2B), and the emitted light was transmitted to the smart imaging system via the objective lens shown in Figure 2A. During detection, the fluorescent emission was imaged at the Taylor cone by the CCD camera through the detection tube assembly (1 in Figure 2B). The detection tube was placed at approximately 3 cm above the NanoSpray source (4 in Figure 2B) on an adjustable 3D stage holder and carefully positioned to focus on the Taylor cone at the outlet end of the separation capillary (3 in Figure 2B), illuminated by the laser source (2 in Figure 2B). Image collection was accomplished by an 8-megapixel noncooled CCD camera (1 in Figure 2A) and analyzed by ImageJ/Fiji software, running on Raspberry Pi-3b as a standalone system. Figure 2C depicts the fluorescence image of the electromigrating APTS labeled maltose in the separation capillary (1 in Figure 2C) and at the Taylor cone (2 in Figure 2C), with the detection zones shown as dotted squares. First, an APTS labeled maltooligosaccharide mixture was separated and detected by the CE-iLIF-ESI-MS system. Figure 3 compares the CE separation with fluorescent imaging in the capillary approximately 0.15 mm before the end of the tip/ sprayer as control (trace A; sampling area 1 in Figure 2C), at the Taylor cone (trace B; sampling area 2 in Figure 2C), and

Figure 1. Schematic representation of the imaging laser-induced fluorescence (iLIF) detection at the Taylor cone of CE-ESI-MS. (1) Outlet tip of the separation capillary; (2) Taylor cone; (3) jet and electrospray plume; (4) MS orifice; (5) imaging laser; (6) CCD camera.

Figure 2 presents the details of the imaging laser-induced fluorescence detection system focusing on the Taylor cone of

Figure 3. Capillary electrophoresis separation of APTS labeled maltooligosaccharides, detected by fluorescence imaging in the separation capillary approximately 0.15 mm from the tip/sprayer as control (A), at the Taylor cone (B), and by mass spectrometry (C) in negative ionization mode. The inserts above the electropherograms in Panels A and B show the actual fluorescent signal in the capillary by xz orthogonal rendering of the collected image sequences. The insets above Trace B and C show the blow up portion of the separation emphasizing the peak distribution change from DP9−DP13. Conditions: BGE: 10 mM ammonium acetate (pH 4.5) and 20% isopropanol, capillary: 90 cm of 30 μm ID (150 μm OD) bare fused silica; injection order: water plug: 3.0 psi for 5.0 s, sample: 10 kV for 20 s; applied voltage: 30 kV (reversed polarity, cathode at the injection side); separation temperature: 30 °C. A small positive pressure (2.0 psi) was applied from the injection end during the entire analysis to stabilize electrospray. The electrospray voltage was set to −1400 V, the orifice plate temperature was 80 °C, and the curtain gas (N2) pressure was 5.0 psi.

Figure 2. Imaging laser-induced fluorescence (iLIF) detection system for capillary electrophoresis − ESI mass spectrometry. Panel (A) detection tube assembly: 1, CCD camera; 2, eye lens; 3, bandpass filter; 4, objective lens. Panel (B) CE-iLIF-ESI-MS system: 1, detection tube assembly; 2, laser; 3, capillary tip/sprayer; 4, nanospray source. Panel (C) Fluorescence images for detection within the capillary and Taylor cone: 1, separation capillary; 2, Taylor cone (the yellow dotted squares represent the actual sampling areas for the capillary and Taylor cone detection. The red dotted line depicts the outlet end of the separation capillary). 7740

DOI: 10.1021/acs.analchem.9b01028 Anal. Chem. 2019, 91, 7738−7743

Article

Analytical Chemistry

Figure 4. Capillary electrophoresis separation of APTS labeled human IgG1 branched N-glycans detected by fluorescence imaging in the separation capillary approximately 0.15 mm from the tip/sprayer as control (A), at the Taylor cone (B), and by mass spectrometry (C) in negative ionization mode. The inserts above the electropherograms in Panels A and B show the actual fluorescent signal in the capillary, and at the Taylor cone, respectively, by xz orthogonal rendering of the collected image sequences. Panel D depicts the extracted ion electropherograms with mass based identification of nine glycans along with their structural interpretation. Conditions were the same as in Figure 3.

by mass spectrometry detection (trace C). In the system used here, it was possible to focus the laser on the fused silica capillary very close to the outlet of the column and used as control for the off capillary detection; however, in other designs, this may not be possible. The insets above the electropherograms in panels A and B show the actual fluorescent images in the capillary, and at the Taylor cone, respectively. The negligible differences in the fluorescent detection profiles between the in-capillary control (A) and at Taylor cone (B) detection traces (Pearson Product Correlation was 0.996) emphasize the applicability of this detection option. The approximately 1.5× higher detection sensitivity in trace B was probably due to the signal collection through the air in comparison to through the fused silica capillary wall in trace A. The differences, on the other hand, in the peak distribution profiles between the fluorescent images and the MS data is apparent from the comparison of the fluorescent signals (RFU, both in trace A and B) and total ion count data (TIC, trace C). While the fluorescent detection efficiency was considered the same for all labeled maltooligosaccharide species (one fluorophore per oligosaccharide, i.e., constant response factor), representing their true quantitative distribution, a steeper decrease was observed in the TIC signal intensity for the longer oligosaccharides, i.e., after the maltodecaose (DP10) peak compared to the RFU signals as depicted by the insets above trace B and C. The short distance between the Taylor

cone and the MS orifice (a few mm) and the transmission efficiency of the spray process resulted in practically simultaneous fluorescent image acquisition of the electropherogram along with obtaining the structural data information for the fluorophore labeled analyte molecules. Next, a mixture of PNGase F released and APTS labeled human immunoglobulin G1 branched N-glycans were analyzed by capillary electrophoresis with imaging laser-induced fluorescence and MS detection. Figure 4, trace A shows the iLIF signal, collected approximately 0.15 mm from the tip/ sprayer end of the capillary (control). Trace B shows the iLIF signal at the Taylor cone, while trace C shows the total ion current of mass spectrometry detection. Trace D displays the extracted ion electropherogram with mass based identification of the higher abundant carbohydrate structures identified in this study. Due to background electrolyte composition limitations for CE-MS, neither polymeric additives nor nonvolatile buffer components were used, resulting in decreased separation efficiency and comigration of the Man6, A2G1, and FA2 structures as well as the FA2BG1 and FA2[3]G1 glycans, respectively. In addition, only partial separation of the positional isomers of the core-fucosylatedmonogalactosylated-biantennary glycans with 1−6 (FA2[6]G1) and 1−3 (FA2[3]G1) linkages was obtained. Similar to Figure 3, here the N-glycan profiles detected by iLIF (both in the capillary and at the Taylor cone) and by MS also 7741

DOI: 10.1021/acs.analchem.9b01028 Anal. Chem. 2019, 91, 7738−7743

Article

Analytical Chemistry

itself, making reliable fluorescent light collection challenging. Using the direct imaging based detection system introduced in this paper, one can digitally compensate for any movement occurring in the field of view (i.e., sprayer tip wavering). Consequently, this robust setup supports simultaneous quantitative (fluorescent image based) and qualitative (mass information based) data acquisition for any liquid phase separation systems with sprayers connected to ESI-MS. In addition to the quantitation challenges caused by ion suppression and ionization efficiency differences during electrospray ionization, unpredictable analyte fragmentation, and rearrangement might also occur, making structural interpretation ambiguous by using the MS data only. Thus, the accompanying fluorescent detection information on the separation can be crucial not only for precise quantification, but to avert any data misinterpretation. The MS signal, on the other hand, revealed comigration of several carbohydrate structures in the human IgG1 CE-ESI-MS N-glycan profile, that was otherwise not apparent from the iLIF trace, due to the limited resolving power of the MS friendly buffer system used. The detection limit of the fluorescent imaging system was comparable to that of the detection limit of our mass spectrometer in negative mode operation, thus, represented a good match for simultaneous fluorescent image and mass data acquisition. The iLIF detection at the Taylor cone is universal (i.e., not system dependent); thus, it could be applied to other configurations, such as nano liquid chromatography or microchip electrophoresis with the use of a variety of sprayer designs. If greater detection sensitivity or linear range required, a high-end cooled CCD camera could be used. It is also important to note that label-free detection of tryptophan containing proteins and peptides could also be implemented by employing UV lasers and corresponding UV sensitive CCD cameras using the above-described fluorescent imaging system.

exhibited differences, probably due to structure based ionization efficiency variances39 compromising exact quantification by solely relying on the MS data. Table 1 depicts the Table 1. Area Percent Distribution of the Five Major Peaks in Figure 4 Detected by iLIF at the Taylor Cone and by MS % area structures

iLIF detector

MS

Man5 FA2; A2G1; Man6 FA2(6)G1; FA2(3)G1; FA2BG1 FA2G2

8.26 24.65 23.96 21.06 22.07

5.48 38.50 29.06 18.09 8.87

area percent distribution of the five major peaks in Figure 4, showing heavily decreasing MS signal values for the larger, later migrating branched carbohydrates in comparison to the iLIF signal. On the other hand, comigration of several species were revealed by the MS signal, emphasizing the importance of the simultaneous use of both detections systems, especially when full separation is not obtained. The detection limit of the imaging laser-induced fluorescent detection system was evaluated by infusing a dilution series of APTS labeled maltose in the concentration range of 0.2−20 μM. The CCD camera was acquiring images in exhaustive condition (ISO 800; 200 ms exposure time in burst mode for the time-lapse imaging). At such elevated ISO level, high exposure, and low readout rate, the sensitivity could be extremely enhanced; however, linearity and stability started to drop sharply40 since the camera was not cooled, and it was not generating data in the raw image format. For quicker encoding and easier processing, hardware accelerated JPEG encoding was used. The RGB (24 bit) images were split into the 3 channels, and only the green channel was kept for intensity processing leaving an 8-bit dynamic range with 256 steps of intensity scale, controlling the dynamic range of the detection. With this simple, not cooled CCD camera, the limit of detection in the imaging area depicted in Figure 2C at the Taylor cone was approximately 40 attomole. Please note that the detection limit was also influenced by the size of the actual imaging area, as explained earlier by Szarka and Guttman.38 Using the iLIF detection arrangement introduced in this paper, excellent intra- and interday peak area reproducibility values were obtained as depicted for the major peaks of Figure 4 and shown in Supplementary Figure 1, panel A and B with the average RSD of 4.15% and 6.79%, respectively, emphasizing the robustness of the system.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.9b01028.





Supplementary Figure 1 shows the intra- (A) and interday (B) peak area reproducibility values for the major peaks of Figure 4 (PDF)

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected].

CONCLUSIONS Simultaneous fluorescent imaging and mass spectrometry detection of APTS labeled linear and branched carbohydrates has been implemented by using a diode laser based excitation source and a simple, noncooled CCD camera for image acquisition and processing in a hyphenated capillary electrophoresis ESI-MS system. The long distance and, therefore, low magnification lens assembly allowed relatively deep depth of field to collect photons from the subnanoliter volume of the Taylor cone at the outlet end of the separation capillary. Traditionally used PMT setups, while making single photon level detection theoretically possible, would require a fixed position detection zone window, in which case tiny ripples in the electric field could avert the Taylor cone and/or the tip

ORCID

Máte Szarka: 0000-0003-2670-5554 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors gratefully acknowledge the support of the National Research, Development and Innovation Office (NKFIH) (K 116263) grants of the Hungarian Government, the EFOP-3.6.3-VEKOP-16-2017-00009 cofinanced by EU and the European Social Found. This work was also supported by the BIONANO_GINOP-2.3.2-15-2016-00017 project and the V4-Korea Joint Research Program, project National 7742

DOI: 10.1021/acs.analchem.9b01028 Anal. Chem. 2019, 91, 7738−7743

Article

Analytical Chemistry

(35) Rodriguez-Cruz, S. E.; Khoury, J. T.; Parks, J. H. J. Am. Soc. Mass Spectrom. 2001, 12, 716−725. (36) Nemes, P.; Vertes, A. Anal. Chem. 2007, 79, 8098−8106. (37) Szigeti, M.; Guttman, A. Methods Mol. Biol. 2017, 1503, 265− 272. (38) Szarka, M.; Guttman, A. Anal. Chem. 2017, 89, 10673−10678. (39) Schmidt, A.; Karas, M.; Dulcks, T. J. Am. Soc. Mass Spectrom. 2003, 14, 492−500. (40) Pagnutti, M.; Ryan, R. E.; Cazenavette, G.; Gold, M.; Harlan, R.; Leggett, E.; Pagnutti, J. Journal of Electronic Imaging 2017, 26, 013014.

Research, Development and Innovation Office (NKFIH; NN 127062) grants of the Hungarian Government. The stimulating discussions with Dr. Barry L. Karger are also greatly appreciated. This is contribution #150 from the Horváth Csaba Memorial Laboratory of Bioseparation Sciences.



REFERENCES

(1) Sun, L.; Zhu, G.; Zhang, Z.; Mou, S.; Dovichi, N. J. J. Proteome Res. 2015, 14, 2312−2321. (2) Bush, D. R.; Zang, L.; Belov, A. M.; Ivanov, A. R.; Karger, B. L. Anal. Chem. 2016, 88, 1138−1146. (3) Han, M.; Rock, B. M.; Pearson, J. T.; Rock, D. A. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2016, 1011, 24−32. (4) Lew, C.; Gallegos-Perez, J. L.; Fonslow, B.; Lies, M.; Guttman, A. J. Chromatogr. Sci. 2015, 53, 443−449. (5) Dole, M.; Hines, R. L.; Mobley, R. C.; Ferguson, L. D.; Alice, M. B.; Mack, L. L. J. Chem. Phys. 1968, 49, 2240. (6) Mack, L. L.; Rheude, A.; Dole, M.; Kralik, P. J. Chem. Phys. 1970, 52, 4977. (7) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Science 1989, 246, 64−71. (8) Konermann, L. J. Am. Soc. Mass Spectrom. 2017, 28, 1827−1835. (9) Costello, C. E. Curr. Opin. Biotechnol. 1999, 10, 22−28. (10) Saraswat, S.; Snyder, B.; Isailovic, D. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2012, 902, 70−77. (11) Wang, Y.; Santos, M.; Guttman, A. J. Sep Sci. 2013, 36, 2862− 2867. (12) Mucha, E.; Lettow, M.; Marianski, M.; Thomas, D. A.; Struwe, W. B.; Harvey, D. J.; Meijer, G.; Seeberger, P. H.; von Helden, G.; Pagel, K. Angew. Chem., Int. Ed. 2018, 57, 7440−7443. (13) Jarvas, G.; Fonslow, B.; Yates, J. R., 3rd; Foret, F.; Guttman, A. J. Chromatogr. Sci. 2017, 55, 47−51. (14) Jorgenson, J. W.; Lukacs, K. D. Science 1983, 222, 266−272. (15) Jorgenson, J. W.; Lukacs, K. D. Clin Chem. 1981, 27, 1551− 1553. (16) Gassmann, E.; Kuo, J. E.; Zare, R. N. Science 1985, 230, 813− 814. (17) Cheng, Y. F.; Dovichi, N. J. Science 1988, 242, 562−564. (18) Burolla, V.; Pentoney, S. J.; Zare, R. N. American biotechnology laboratory 1989, 7, 6. (19) Chen, D. Y.; Swerdlow, H. P.; Harke, H. R.; Zhang, J. Z.; Dovichi, N. J. J. Chromatogr. 1991, 559, 237−246. (20) Karger, B. L.; Guttman, A. Electrophoresis 2009, 30, S196−202. (21) Drossman, H.; Luckey, J. A.; Kostichka, A. J.; D’Cunha, J.; Smith, L. M. Anal. Chem. 1990, 62, 900−903. (22) Kist, T. B.; Termignoni, C.; Grieneisen, H. P. Braz. J. Med. Biol. Res. 1994, 27, 11−19. (23) Schultz, N. M.; Huang, L.; Kennedy, R. T. Anal. Chem. 1995, 67, 924−929. (24) Baars, M. J.; Patonay, G. Anal. Chem. 1999, 71, 667−671. (25) Reif, O. W.; Freitag, R. J. Chromatogr A 1995, 716, 363−369. (26) Tagliaro, F.; Turrina, S.; Pisi, P.; Smith, F. P.; Marigo, M. J. Chromatogr., Biomed. Appl. 1998, 713, 27−49. (27) Guttman, A. TrAC, Trends Anal. Chem. 2013, 48, 132−143. (28) Sturm, S.; Stuppner, H. Electrophoresis 1998, 19, 3026−3032. (29) Coenen, G. J.; Kabel, M. A.; Schols, H. A.; Voragen, A. G. Electrophoresis 2008, 29, 2101−2111. (30) Foret, F.; Kirby, D. P.; Vouros, P.; Karger, B. L. Electrophoresis 1996, 17, 1829−1832. (31) Gennaro, L. A.; Salas-Solano, O. Anal. Chem. 2008, 80, 3838− 3845. (32) Huhn, C.; Ruhaak, L. R.; Mannhardt, J.; Wuhrer, M.; Neususs, C.; Deelder, A. M.; Meyer, H. Electrophoresis 2012, 33, 563−566. (33) Khan, S.; Liu, J.; Szabo, Z.; Kunnummal, B.; Han, X.; Ouyang, Y.; Linhardt, R. J.; Xia, Q. Rapid Commun. Mass Spectrom. 2018, 32, 882−888. (34) Zhou, S.; Cook, K. D. Anal. Chem. 2000, 72, 963−969. 7743

DOI: 10.1021/acs.analchem.9b01028 Anal. Chem. 2019, 91, 7738−7743