Electrophoretic Desalting To Improve Performance ... - ACS Publications

Feb 13, 2018 - (MS), ionization is vulnerable to deleterious effects of solution- phase ions such as inorganic salts and ... interfering salt cluster ...
1 downloads 9 Views 2MB Size
Subscriber access provided by UNIV OF DURHAM

Article

Electrophoretic Desalting to Improve Performance in Electrospray Ionization Mass Spectrometry Zezhen Zhang, Christopher J Pulliam, Tawnya G. Flick, and R. Graham Cooks Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b04529 • Publication Date (Web): 13 Feb 2018 Downloaded from http://pubs.acs.org on February 15, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Electrophoretic Desalting to Improve Performance in Electrospray Ionization Mass Spectrometry Zezhen Zhang1, Christopher J Pulliam1, Tawnya Flick2* and R. Graham Cooks1* 1

Department of Chemistry, Purdue University, 560 Oval Drive, West Lafayette, Indiana 47907, United States Department of Analytical Research & Development, Amgen Inc., 1 Amgen Center Drive, Thousand Oaks, California 91320, United States 2

Keywords: Electrophoresis• Separation • Nano ESI • Protein analysis • Phosphate buffer • Matrix effects

Abstract Mass spectrometers are sensitive tools used to identify and quantify both small and large analytes using to the mass-to-charge ratios (m/z) of ions generated by electrospray ionization or other methods. Ionization typical generates protonated or deprotonated forms of the analytes or adducts with adventitious metal ions derived from the spray solvent. The formation of a variety of ionized forms of the analyte as well as the presence of cluster ions complicates the data and can have deleterious effects on the performance of the mass spectrometer, especially under high salt or buffer conditions. To address this, a method involving a dual-electrode nano electrospray source has been implemented to rapidly and temporarily desalt the spray solution of interfering cationic and anionic species using electrophoretic transport from the spray tip. Peptides, proteins, and pharmaceutical drugs all showed improved results after the desalting process as measured by the quality of the mass spectra and the limits of detection achieved. Importantly ordinary phosphate buffers could be used to record protein mass spectra by nano ESI. Introduction Biological samples often exist in complex matrices and despite the performance advantages of mass spectrometers (MS), ionization is vulnerable to deleterious effects of solution-phase ions such as inorganic salts and buffers. These components can suppress analyte signal, cause it to be obscured by interfering salt cluster ions and/or cause the signal for a single component to be spread across multiple ion channels1-5 (e.g. as [M+H]+, [M+Na]+ and [M – H + 2K]+) so both lowering sensitivity and complicating the spectrum. This is especially true for high salt matrices. The fundamentals of these processes have been elucidated in classic papers by Enke6 and by Tang, Page and Smith7 and a recent critical review has appeared8. Off-line desalting methods such has solid phase extraction and gel-filtration have been developed to remove interferents from the matrix9-12. An off-line electrokinetic removal method has also been described13 for small samples. Modification of the buffer14-16 is another widely used approach in which the non-volatile buffer is replaced by a highly volatile MS compatible buffer such as ammonium acetate. This method can be performed on-line or off-line but it requires that the sample be stable enough to survive the matrix change. Size exclusion chromatography (SEC) can be used as an effective alternative to aid in desalting chemical matrices12,17,18. On-line variants of this method incorporate gel filtration cartridges into the SEC to reduce sample consumption while also maximizing the desalting effect12,19. Numerous commercial and research liquid chromatographic (LC) methods have also been developed to desalt matrices for protein analysis prior to MS20,21 but they are time consuming and inconvenient and an 1

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 13

effective on-line non-chromatographic desalting method remains highly desirable. Capillary electrophoresis (CE) methods22 including commercial CE-ESI microfluidic devices are being used to prepare, separate, and electrospray biological and chemical samples into high resolution mass spectrometers23-26 through on-chip desalting of complex matrices. An on-line ion exchange method has also been demonstrated to remove salts27. A very different solution to the problem of on-line desalting was reported Susa et al., who simply reduced the tip size of the ESI glass emitter to minimize salt ion suppression of protein ions28 in order to minimize formation of droplets containing both analyte and salt. An alternative approach, demonstrated by Li29, Zhang30, Wei31 and Gong32 and their coauthors, involves modifying the potential inside the spray source. Li and Huang29 applied an ac potential to a sonic spray source, while the other groups used nano-electrospray (nESI) and minimized matrix effects through electrophoretic manipulation of the sample. Herein, we present an electrophoretic, multi-electrode nESI technique to rapidly (and reversibly) desalt matrices without the addition of chemical modifiers or changes in tip diameter. Differences between the present study and the earlier work are as follows: Zhang’s study30 used a capillary that was split into two sections in order to manipulate the separations and no desalting effects were reported. The strategies of Wei31 and Gong32 used only one electrode and applied different voltages in a temporal sequence so that salts could be removed during the potential change. Note also that a principal difference between the device used in this study and those used by Zhang, Wei and Gong is that our electrophoretic spray can be operated continuously since the sample is directly injected by syringe and additional sample can be added as needed. Also the desalted mass spectrum persists for remarkably long periods in the cases of high salt small or large molecule samples, as shown here for the first time. Experimental All mass spectra were recorded using a linear ion trap mass spectrometer (LTQ, Thermo Scientific). Normally, the spray voltage HV1 (see Fig. 1) was +1.5kV in the positive mode and -1.5kV in the negative ion mode. The separation voltage HV2 was set in the range 3kV - 5kV. The glass capillary used as nanospray emitter has o.d.:1.5 mm, i.d.:0.86 mm with a nominal tip diameter of 10 µm. Capillaries were pulled using P-2000 (Sutter Instrument Co.). A fused silica capillary coated with UV transparent fluoropolymer, bought from Polymicro Technologies, served as the sample injection tube (i.d.: 100 µm, 50 cm length) when connected to a 1 mL syringe. A 0.2mm i.d. enameled wire was used as the separation electrode. Capillary temperature was 200℃ , injection time 100 ms and three microscans were averaged to create a mass spectrum. Other details can be found in Supporting Information (SI). Carfilzomib and Asp-Leu were provided by Amgen Inc., bradykinin, [D-Ala2]-Leucine encephalin, (Val5)-angiotensin I, caffeine solution, reserpine, perfluoroctanoic acid (PFOA); perfluoronananoic acid (PFNA); perfluoroheptanoic acid (PFHA), Cytochrome C, sodium chloride, calcium chloride, potassium chloride, tetramethyl ammonium bromide, didodecyldimethylammonium bromide, tetra-n-butyl ammonium Iodide, methanol and formic acid were purchased from Sigma-Aldrich. Tris and hydrochloride were purchased from J.T. Baker. Purified water came from the Purdue DI water system. All samples were diluted to the desired concentration by using menthanol/water, 1:1. The KCl/Tris-HCl buffer system was 150 mM KCl and 25 mM Tris-HCl (pH 7).

2

ACS Paragon Plus Environment

Page 3 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Results and Discussion The setup, shown in Figure 1, uses a small capillary which contains ca. 6 µL of water (blue). The analyte, shown in purple, is injected for two minutes (4 µL/min) into the tip of the capillary via a syringe pump connected to an insulated fused silica line. A high voltage connector, HV1, provides a potential of 1.5 kV at the front of capillary through conduction via the insulated line. Note that HV1 is set to a typical spray voltage used in spray based ionization sources. The second electrode, HV2, is located inside the glass capillary, but not in contact with the solvent from which it is separated by a ca. 5mm air gap.

Figure 1. Schematic of the apparatus used in electrophoretic spray experiments: solvent (e.g. H2O, shown in blue) is first loaded into the nano-ESI emitter and then sample (red) present in the syringe is loaded into the solvent via the injection tube. Spray voltage HV1 is applied to the needle of the syringe and separation voltage HV2 is applied to a coated electrode, which is inserted into the nano-ESI emitter such that it approaches but does not touch the liquid surface.

In-source desalting is achieved using a three step process in either the positive or the negative ion mode. For positive mode analysis, HV1 is set to +1.5 kV and the syringe pump is operated as described above. The data gathered during step 1 is a typical mass spectrum that contains [M+H]+, [M+Na]+ and [M+K]+ species, where M indicates the intact analyte molecule. Next, a second potential, HV2, is applied inductively to the rear of the solvent meniscus, at a potential ranging from -3 kV to -5 kV while HV1 continues to be applied. When on, the positive ion signal is depleted of adduct ions of the type [M+H]+ and [M+Na]+ due to the high negative electric field which sequesters mobile positive ions migrating through the solvent towards the rear of the capillary. After 1 minute, HV2 is removed and solution-phase ions proceed to diffuse to the depleted region in the front of the emitter. Salt adducts are absent from the mass spectrum after removing HV2. We ascribe this phenomenon is to differences in ion mobilities of protons and salt cations; protons have higher mobility and reach the tip of the emitter before Na+ and K+ resulting in a mass spectrum containing primarily protonated ions. Typical data illustrating these points is shown in Figure S1. The simple method just described was applied to bradykinin solutions prepared at concentrations of 5.0 and 0.5 µg/mL in 104 µg/mL aqueous NaCl. Figure 2 a, b illustrates the challenges of analyzing compounds in high salt matrices. Bradykinin at 5 µg/mL is barely detectable and at 0.5 µg/mL is below the LOD for the experiment. Even the addition of 0.1% formic acid (FA) does not significantly improve 3

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 13

the signal-to-noise ratio for bradykinin, Figure 2 c, d. These solutions were then analyzed using the electrophoretic desalting source, Figure 2 e, f. In both spectra, the S/N is improved substantially due to the (temporary) removal of the interfering cations from the tip of the capillary.

Figure 2. (a) Traditional nano-ESI mass spectra recorded at 1.5 kV and averaged for 1min for 5 µg/ml bradykinin in 104 µg/mL aqueous NaCl; (b) as in (a) but using 0.5 µg/mL bradykinin; (c) as in (a) but with 0.1% formic acid; (d) as in (b) but with 0.1 % formic acid; (e) electrophoretic spray of the same solution as in (a) using the same spray voltage (HV1) of 1.5 kV after application for 1 min and then removal of - 5 kV separation voltage (HV2); (f) electrophoretic spray of the same solution as in (b) with spray voltage (HV1) of 1.5 kV after application for 1 min and then removal of - 5 kV separation voltage (HV2). Ion abundances are given in instrument counts.

The desalting process can be repeated multiple times to remove cationic adducts that may gradually reappear during long analysis times. Figure 3 shows data for an experiment in which the electrophoretic potential was applied three times in succession in order to reduce the appearance of unwanted metal adducts while maintaining the desired protonated analyte. In Figure 3a, each period represents data recorded after one desalting cycle. To make sure that salt peaks reappear in the mass spectrum, we decreased the application time of the separation voltage (HV2) from 1 minute to 10 seconds. Moreover, the applied potentials (1.5 kV and – 5 kV) and the distance between the HV2 electrode and the liquid surface (5 mm) were not adjusted from the original values. The mass spectrum after the first cycle is as 4

ACS Paragon Plus Environment

Page 5 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

‘clean’ as it is in the 1 minute mass spectrum and the salt peaks reappear after about 10 minutes. The second spectrum was recorded 15 – 19 min after the first desalting step and it shows the beginning of the reappearance of the salt adducts. Although the signal during the third period is not as stable as that during the first two cycles, the mass spectra are comparable to that observed after the first cycle, which might indicate that repeatedly applying the separation voltage might improve the desalting effect. For clarity, the total ion chronograms (TICs) for protonated and sodiated bradykinin are shown also.

5

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 13

Figure 3. a) Total ion chronogram and average electrophoretic spray mass spectra during multiple cycles of desalination. b) Average mass spectrum recorded during 5-15 min of the first cycle, showing only protonated peptide, c) 18 - 22 min showing appearance of Na+ and K+ adducts, d) 45 – 55 min, showing removal of Na+ and K+ adducts by the third cycle of cleaning. Separation voltage HV2 was turned on for just 10 seconds at the beginning of each cycle. The three dashed lines indicate the time at which the Na+ and K+ adducts begin to reappear in the mass spectra. Sample: 1µg/mL carfilzomib in 10µg/mL NaCl and 20µg/mL KCl. e) TIC ion chronograms for [M+H]+ and [M+Na]+

To explore how the voltages and separation time affect the mass spectrum and to determine conditions which result in the highest S/N, HV1 and HV2 were varied, see Table S1. The situation involves ion mobility in solution as well as charged droplet emission and is, not surprisingly complex, but the simple result is that as the offset potential (HV1 + HV2) was increased from 0 to 3.5 kV, the time during which desalted spectra could be observed also increased. When HV1 + HV2 was equal to zero, there was no desalting, and there was no difference in the mass spectra recorded before and after the separation voltage was applied. The duration for which HV2 was applied was also observed to have a significant effect on the spectrum. When HV2 was on for 10 sec, the desalted spectrum persisted for 10 minutes before adducts began to form. In the case where HV2 was on for 60 sec, and other conditions were kept constant, the desalted spectrum persisted for over an hour. Full data and experimental conditions are given in the Supporting Information, Table S2. Hereafter, unless otherwise specified, a 1 minute separation time and 3.5 kV offset voltage (HV1 1.5 kV and HV2 -5 kV) were used. Under these conditions, 1 µg/mL carfilzomib in 10 µg/mL of NaCl and 20 µg/mL of KCl were sprayed for over an hour with no observable cationic adducts (Figure S1). A three-peptide mixture consisting of bradykinin at 5µg/mL, (Val5)angiotensin at 10 µg/mL, and [D-Ala2]-Leucine encephalin at 10 µg/mL was also successfully sprayed for more than an hour (Figure S2). 6

ACS Paragon Plus Environment

Page 7 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Interferents in MS analysis not only arise from inorganic salts but may also arise in the form of precharged organic ions such as quaternary ammonium salts. To investigate this matter, didodecyldimethyl ammonium bromide 1 µg/mL, tetra-n-butyl ammonium iodide 1 µg/mL, and [D-Ala2]-leucine encephalin 5µg/mL were analyzed using both nESI and electrophoretic spray (Figure S3). In the nESI spectrum, sodium adducts and interferences from the quaternary ammonium appeared in the spectrum. After a separation time of just 1 min, the sodium adduct was completely removed; however, the ammonium peak remained present. Nevertheless, after another 1 minute cycle, the ammonium adduct was also removed. Some analytes are best measured as negatively charged ions using negative spray voltages but these are also vulnerable to competition with and interference by adducts such as chloride and bromide (Figure 4a, b). Removal of these species can be performed in the same manner as in the positive mode with reversed potentials, viz., using the rear electrode is supplied with a high positive voltage and the front (spray) electrode is set to -1.5 kV. Using these conditions, 1 µg/mL of carfilzomib prepared in 15 µg/mL NaCl, 15 µg/mL CaCl2, and 60 µg/mL (CH3)4NBr, showed peaks corresponding to these anionic adducts which were greatly reduced after exposure to the desalting potential (Figure 4c, d). However, there is an anionic adduct at [M+45]- which is believed to arise from adventitious formate.

7

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 13

Figure 4. Mass spectra in negative ion mode for 1µg/mL carfilzomib M = 719.91 in 15µg/mL NaCl, 15µg/mL CaCl2 and 60µg/mL (CH3)4NBr. (a) Traditional nano-ESI mass spectrum recorded at -1.5 kV with signal averaging for 1min; (b) as in (a) but showing low mass range; c) electrophoretic spray of the same solution as in (a) with spray voltage (HV1) of -1.5 kV but after application for 1 min then removal of 5 kV separation voltage (HV2); (d) as in (b) but showing low mass range. Ion abundances are given in instrument counts.

The desalted signal in the negative ion mode lasted for approximately 10 min, which is sufficient for many MS experiments. Negative mode analysis was successfully extended to common environmental contaminants such as perfluorinated organic acids: perfluoroctanoic acid (PFOA, 1 µg/mL), perfluoronananoic acid (PFNA, 1 µg/mL), and perfluoroheptanoic acid (PFHA, 1 µg/mL) in 10 mg/mL NaCl, see Figure S4. The electrophoretic spray protocol was applied to the protein cytochrome C prepared in 1:1 methanol:water without addition of salt. When analyzed by nESI, the 7+ and 8+ charge states were present in high abundance in the mass spectrum (Figure 5a) but when exposed to the electrophoretic voltage, the charge state was increased significantly. Figure 5b and 5c were recorded with a rear electrode potential of -4 kV and -5 kV (offset = 2.5 kV and 3.5 kV) respectively. The increase in the charge state of cytochrome C is attributed to denaturing as a result of the higher relative proton abundance near the tip of the emitter at the time of analysis. To test this, formic acid was added to a fresh solution of cytochrome C and analyzed via nESI. The data shown in Figure S5 indicate that the effect of adjusting bulk pH and of applying the desalting potentials were similar. Figure 5 also illustrates decreased tailing of the peaks, consistent with removal of adventitious cationic adducts. The enlarged image of the 8+ charged peak in Fig. 5 a and b, show that Na+ adducts were remarkably more abundant than were protonated cytochrome C ions using traditional nESI; multiple Na+ ions are bound to cytochrome C molecules. In comparison, after removal of the separation voltage, the signal intensity of protonated 8+ charged ions was significantly enhanced and protonated ions became the most abundant species which indicates that this method effectively reduced the signal intensity of salt adducts to a very low level improving S/N ratios by more than an order of magnitude. We also operated the electrophoretic spray ionization system under conditions typically used in protein solution chemistry, i.e. using 150 mM KCl and 25mM Tris-HCl buffer at pH 7, a buffer which mimics the intracellular environment but is normally inimical to successful mass spectrometry. The same KCl/TrisHCl buffer system has been used to compare our method against the reduced capillary size method of ref.28. As seen by comparing the data in Fig. 6 with and without electrophoretic desalting, a large improvement in signal/noise is evident for cytochrome C and myoglobin. A comparable result was reported for yeast alcohol dehydrogenase (ADH), bovine serum albumin (BSA) and other proteins when using a quadrupole-time-of-flight (Q-TOF) instrument and using small emitter tips28. In this case the protein ion signal occurred in the m/z 4000 to 8000 range and was enhanced by the small initial droplet size. In the m/z 1000 to 2000 range effective electrophoretic desalting was seen. This region appeared to be more effectively desalted in these experiments than in those of Susa and coworkers28.

8

ACS Paragon Plus Environment

Page 9 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 5. (a) Traditional nano-ESI mass spectrum recorded at 1.5 kV and averaged for 1 min for 50µg/mL Cytochrome C; (b) Electrophoretic spray of the same solution as in (a) but with spray voltage (HV1) of 1.5 kV and after application for 1 min and then removal of – 4 kV separation voltage (HV2); (c) Electrophoretic spray of the same solution as in (a) but with spray voltage (HV1) of 1.5 kV and after application for 1 min and then removal of – 5 kV separation voltage (HV2). Ion abundances are given in instrument counts. Explanded versions of the 8+ charge peak are shown.

The electrophoretic method of sample cleanup works well for removal of other solution-phase ions, besides the salt peaks; for example, some matrix peaks are also removed from the mass spectra. However, based on the result of small molecules mixtures, non-charged matrix constituents will continue to appear in the MS (c.d. Figure S6). To further investigate the mechanism of desalting we enquired into what happens in the system when applying the separation voltage (the stage during which we have no observable signal). To obtain this information we switched to the opposite detection mode during this 1 minute period. The results in Figures S7 and Figure S8 prove that during the application of the separation voltage, the spray is still occurring and that the separation voltage does indeed attract positive ions to the rear of the glass capillary and away from the spray tip, which is consistent with our expectation that the higher mobility of the protons contributes to the results reported in the paper. In fact, in several other experiments (data not shown), salts were discovered to be deposited at the rear of the capillary or on the tip of the HV2 wire when spray times were long enough.

9

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 13

Figure 6. (a) Traditional nano-ESI mass spectrum recorded at 1.5 kV and averaged for 30 seconds for 5 µM Cytochrome C in 150 mM KCl and 25 mM Tris-HCl buffer showing only salt and buffer cluster ions (b) Electrophoretic spray mass spectrum of the same solution as in (a) with 2 min application of -5 kV separation voltage (HV2) data recorded after HV2 removal. (c) Traditional nano-ESI mass spectrum recorded at 1.5 kV and averaged for 30 seconds for 5 µM Myoglobin in 150 mM KCl and 25 mM Tris-HCl buffer showing only salt and buffer cluster ions (d) Electrophoretic spray mass spectrum of the same solution as in (c) with 2 min -5 kV separation voltage (HV2) data recorded after HV2 removal and showing the characteristic myoglobin charge state distribution.

Conclusions Electrophoretic nESI can be used to reduce the deleterious effects of adduct formation of salts by sequestering them in the rear of the emitter prior to electrospray analysis. This technique requires two electrodes to create a potential gradient which induces ion migration. Both the offset potential and the duration of the separation voltage were varied and optimized. Perhaps the most striking feature of the methodology is the difference between the time for electrophoresis (10 sec to 1 min) and the time for reappearance of salt adduct (10 min to more than 1 hour). The coulombic barrier presented by the mobile protons is the underlying cause, as will be discussed separately. Importantly, this method is suitable for removal of both organic and inorganic cations and anions and is compatible with mass spectrometers capable of nano-electrospray ionization. It is also successful in recording mass spectra from phosphate buffers, a notoriously difficult process in mass spectrometry.

10

ACS Paragon Plus Environment

Page 11 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Author information Corresponding Authors *E-mail: [email protected] ‡E-mail: [email protected] Supporting Information The Supporting Information is available free of charge on the ACS Publications website Instrumental setup plus results and discussion. Optimization and duration of applied voltages (Tables S1 and S2). Desalinization performance in positive mode with stages of process (Fig. S1) performance with peptide mixture (Fig. S2). Comparison of organic and inorganic salt removal (Fig. S3). Negative ion mode performance (Fig. S4), pH effects (Fig. S5) and mixture analysis (Fig. S6). Exploration of mechanism, Fig. S7 and S8.

Acknowledgements The authors acknowledge financial support for this work by Amgen Inc. and the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, Separations and Analysis Program (DEFG02-06ER15807).appreciate valuable discussions with Stephen. T. Ayrton, Ryan. M. Bain, Zhenwei Wei, Da Ren, Michael Achmatowicz and Elcin Icten. References (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21)

Wang, G.; Cole, R. B. Anal. Chem. 1994, 66, 3702-3708. Cavanagh, J.; Benson, L. M.; Thompson, R.; Naylor, S. Anal. Chem. 2003, 75, 3281-3286. Iavarone, A. T.; Udekwu, O. A.; Williams, E. R. Anal. Chem. 2004, 76, 3944-3950. Pan, P.; McLuckey, S. A. Anal. Chem. 2003, 75, 5468-5474. Wilson, D. J.; Konermann, L. Anal. Chem. 2005, 77, 6887-6894. Enke, C. G. Anal. Chem. 1997, 69, 4885-4893. Tang, K.; Page, J. S.; Smith, R. D. J. Am. Soc. Mass. Spectr. 2004, 15, 1416-1423. Furey, A.; Moriarty, M.; Bane, V.; Kinsella, B.; Lehane, M. Talanta 2013, 115, 104-122. Jiang, Y.; Hofstadler, S. A. Anal. Biochem. 2003, 316, 50-57. Fornea, D. S.; Wu, Y.; Marcus, R. K. Anal. Chem. 2006, 78, 5617-5621. Manduzio, H.; Martelet, A.; Ezan, E.; Fenaille, F. Anal. Biochem. 2010, 398, 272-274. Tubaon, R. M.; Haddad, P. R.; Quirino, J. P. Proteomics 2017, 17, Article no.1700011. Tubaon, R. M.; Haddad, P. R.; Quirino, J. P. J. Chromatogr. A 2017, 1488, 134-139. Clarke, D. J.; Campopiano, D. J. Analyst 2015, 140, 2679-2686. Flick, T. G.; Cassou, C. A.; Chang, T. M.; Williams, E. R. Anal. Chem. 2012, 84, 7511-7517. DeMuth, J. C.; McLuckey, S. A. Anal. Chem. 2014, 87, 1210-1218. Liu, C.; Wu, Q.; Harms, A. C.; Smith, R. D. Anal. Chem. 1996, 68, 3295-3299. Lazar, A. C.; Wang, L.; Blättler, W. A.; Amphlett, G.; Lambert, J. M.; Zhang, W. Rapid Commun. Mass. Sp. 2005, 19, 1806-1814. Mathur, S.; Park, J. D.; Kim, D. H.; Hartmann, R. W. J. Biomol. Screen. 2005, 10, 30-35. Fountain, K. J.; Gilar, M.; Gebler, J. C. Rapid Commun. Mass. Sp. 2004, 18, 1295-1302. Pendela, M.; Hoogmartens, J.; Van Schepdael, A.; Adams, E. J. Sep. Sci. 2009, 32, 3418-3424. 11

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 13

(22) (23)

Tomlinson, A.; Guzman, N.; Naylor, S. J. Cap. Electroph. 1995, 2, 247-266. Li, J.; Thibault, P.; Bings, N. H.; Skinner, C. D.; Wang, C.; Colyer, C.; Harrison, J. Anal. Chem. 1999, 71, 3036-3045. (24) Deng, Y.; Henion, J.; Li, J.; Thibault, P.; Wang, C.; Harrison, D. J. Anal. Chem. 2001, 73, 639-646. (25) Vrouwe, E. X.; Gysler, J.; Tjaden, U. R.; van der Greef, J. Rapid Commun. Mass. Sp. 2000, 14, 1682- 1688. (26) Wang, Y.; Feng, P.; Sosic, Z.; Zang, L. J. Anal. Bioanal. Tech. 2017, 8:359. doi: 10.4172/21559872.1000359

(27) 1021. (28) (29) (30) (31) 52, (32)

Chen, Y.; Mori, M.; Pastusek, A. C.; Schug, K. A.; Dasgupta, P. K. Anal. Chem. 2011, 83, 1015Susa, A. C.; Xia, Z.; Williams, E. R. Anal. Chem. 2017, 89, 3116-3122. Li, G.; Huang, G. J. Mass. Spectrom. 2014, 49, 639-645. Zhang, Z.; He, M.; Liu, L.; Xiong, X.; Fang, X.; Xu, W. Talanta 2017, 164, 45-51. Wei, Z.; Han, S.; Gong, X.; Zhao, Y.; Yang, C.; Zhang, S.; Zhang, X. Angew. Chem. Int. Edit. 2013, 11025-11028. Gong, X.; Xiong, X.; Zhao, Y.; Ye, S.; Fang, X. Anal. Chem. 2017, 89, 7009-7016.

12

ACS Paragon Plus Environment

Page 13 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Analysis of myoglobin from Tris buffer by nESI without (left) and with (right) electrophoretic desalting

ACS Paragon Plus Environment