Boosting the Signal Intensity of Nanoelectrospray Ionization by Using

Jun 2, 2017 - Analytical Chemistry. Poole, Grandy, Yu, Boyaci, Gómez-Ríos, Reyes-Garcés, Bojko, Heide, and Pawliszyn. 2017 89 (15), pp 8021–8026...
0 downloads 0 Views 744KB Size
Subscriber access provided by Binghamton University | Libraries

Article

Boosting the Signal Intensity of Nano-Electrospray Ionization by Using a Polarity-Reversing High Voltage Strategy Xiaoyun Gong, Xingchuang Xiong, Yingchen Zhao, Sijian Ye, and Xiang Fang Anal. Chem., Just Accepted Manuscript • Publication Date (Web): 02 Jun 2017 Downloaded from http://pubs.acs.org on June 3, 2017

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 free 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 accessible to all readers and 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 10

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

Boosting the Signal Intensity of Nano-Electrospray Ionization by Using a Polarity-Reversing High Voltage Strategy Xiaoyun Gong,† Xingchuang Xiong,† Yingchen Zhao,‡ Sijian Ye,† Xiang Fang*,† *Email: [email protected], fax: +86-10-6452 6387 † National institute of Metrology, Beijing, China ‡ Henan Institute of Metrology, Zhengzhou, Henan Province, China ABSTRACT: Continuous efforts have been made to further improve the performance of nano-ESI. In this work, we made use of a polarity-reversing high voltage strategy for the generation of nano-ESI (PR-nESI). Typically, a negative high voltage of -3.0 kV was first applied to the electrode and maintained for 6 s. Then the polarity was reversed and a positive high voltage of +1.75 kV was applied for the generation of electrospray. Compared with conventional nano-ESI, PR-nESI significantly enhanced the signal intensity of protonated protein ions. The signal to noise ratio (S/N) of protonated protein ions was increased by 1−2 orders of magnitude. The increase of S/N was even more remarkable at lower concentrations. Furthermore, PR-nESI also had a desalting effect. Metal ion adducts of proteins were effectively removed. No metal ion adducts were identified from the spectra, even if the concentration of salt was increase to mM level. The performance of PR-nESI was confirmed in the detection of different molecules including proteins, peptides, amino acids and other small molecule compounds. The intact folded structure of proteins was preserved during PR-nESI. No unfolding was observed in the spectra. PRnESI was further applied to the analysis of noncovalent protein-ligand complexes and protein digest. At last, investigations into the mechanism of PR-nESI were carried out. The enhancement of signal intensity and desalting effect were related to the electromigration of the solutes in solution. With all the advantages above, PR-nESI would be a promising method in the future analytical and bio-analytical applications.

INTRODUCTION

been found to be highly tolerant of the suppression effects caused by salts and matrices.13-15 With these key advantages, nano-ESI is now regarded as one of the most favorable choices for omics studies such as proteomics, metabolomics and glycomics.16,17 Recently, several variants have been reported that further improved the detection capability of nano-ESI. A linear array of nano-ESI emitters was developed by Smith’s group, which achieved better sensitivity and quantitation performance.18,19 Compared with single emitter, the multi-emitters enabled higher flow rates. This benefited the separation performance of liquid chromatography, and was more convenient in operation. A modification of the multi-emitters from the linear array into a circular array was also carried out. The circular pattern enabled a more uniform electric field experienced by the constituent emitters.20 The appearance of induced nano-ESI offered an alternative way for the generation of nano-ESI.21 The electrode was no longer in direct contact with sample solutions. Instead, it was placed several millimeters away from the solution. A pulsed high voltage was applied to the electrode to induce electrospray. Induced nano-ESI facilitated array operation, thus increasing the throughput of nano-ESI. The sensitivity was also improved, as well as sample economy, due to the even lower flow rate.22 Besides pulsed high voltage, charge pulse was also proven to be capable of inducing electrospray from nano-tips.23 This technique was specially used for the analy-

The development of electrospray ionization (ESI) has extended the applications of mass spectrometry (MS) from gaseous small molecules and ions into far more abundant species of ions and molecules in solutions.1 ESI provides an effective way for ions to be ejected from solution into gas phase for MS detection.2,3 Macromolecules such as proteins, polymers and nucleic acids can easily be ionized by ESI. Applications of ESI in different fields have grown explosively during the last decades including proteomics, metabolomics, glycomics, and as a compound identification tool for synthetic chemists. Extensive scientific and engineering efforts have been made to better understand the mechanism of ESI and further improve its performance. Among these efforts, one of the most successful steps forward is the proposal of nano-electrospray ionization (nano-ESI) from both experimental demonstration and theoretical simulation.4-6 Compared with conventional ESI, the principle advantage of nano-ESI is its improved ionization efficiency and ion transmission. This is attributed to the much smaller initial droplet size emanated from nano-ESI emitter under a reduced flow rate.7,8 The flow rate of nano-ESI is normally at nL/min level. The low flow rate ensures better sample economy and extended time for more comprehensive analysis. Sensitive bottom-up analysis of nanogram, or even sub-nanogram, amounts of complex protein digests has already been achieved by nano-ESI.9-12 In addition, nano-ESI has also

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

sis of samples with extremely small volume. In addition, direct-current high voltage was used to generate induced nano-ESI as well, which was assisted by the addition of certain solvent to the tip to dissolve the pre-dried sample.24 This method was advantageous to the analysis of samples with buffers. Efforts have also been paid to integrate separation procedures with nano-ESI. The integration of capillary electrophoresis with nano-ESI was realized by capillary zone electrophoresis (CZE), which involved an electrokinetically pumped sheath-flow interface within the nano-tip.25 This method enabled fast and sensitive bottom-up analysis of complex proteome digests.9 Direct separation of analytes from matrices in the tip zone was achieved by step-voltage nano-ESI.26 This technique made use of the electromigration speed difference between analytes and matrices. It was further applied in the removal of matrices in small-volume samples. Although nano-ESI has now been widely used in analytical and biological researches, there are still great demands for methods to further improve the sensitivity and detection capability of nano-ESI. In this work, we made use of a polarityreversing high voltage strategy for the generation of nanoESI (PR-nESI). Compared with conventional nano-ESI, PRnESI significantly enhanced the signal intensity of protonated protein ions. The signal to noise ratio (S/N) of protonated protein ions was increased by 1−2 orders of magnitude. The increase of S/N was even more remarkable at lower concentrations. Furthermore, PR-nESI also had a desalting effect. Metal ion adducts of proteins were effectively removed, thus simplifying the spectrum. In addition, PR-nESI was fast and simple in operation. The enhancement of signal intensity was achieved without adding any pretreatment or additives in the samples. The intact folded structure of proteins was preserved during PR-nESI. No unfolding was observed. With all these advantages, PR-nESI would be a promising method in the future analytical and biological applications.

Page 2 of 10

were adjustable. Detailed discussions were made in the results and discussion section. Typically, a negative high voltage of -3.0 kV was applied to the electrode for 6 s. Afterwards, the polarity of the high voltage was reversed, and a positive high voltage of +1.75 kV was applied to the electrode to generate electrospray. Except for the high voltage strategy, the other settings of PR-nESI were the same as nano-ESI. Other experimental details were described in the Supporting Information.

RESULTS AND DISCUSSION Enhancement of Signal Intensity. PR-nESI significantly enhanced the signal intensity of protonated protein ions. To further investigate this issue, an aqueous solution of cyt c was used as the sample. Native cyt c has a helical globular conformation with one covalently attached heme group.27 It is a highly water soluble protein and has an isoelectric point (pI) of 10.4. Generally, cyt c exists in a native folded structure between pH = 3 and 7.28,29 It undergoes a highly cooperative, acid-induced unfolding transition between pH = 2 and 3.30 Cyt c was dissolved in ultrapure water in a concentration of 10 µM for direct detection. A representative nano-ESI mass spectrum of cyt c was shown in Figure 2a. A unimodal charge state distribution (CSD) was observed. The CSD mainly ranged from 7+ to 11+ and centered at 8+, corresponding to a native folded structure. Obvious metal ion adducts could be identified from the spectrum. The amplified spectrum of the 8+ charged peak showed that Na+ adducts were remarkably more abundant than protonated cyt c ions (Figure 2b). Multiple Na+ ions were bonded to cyt c molecules. Adduct with three Na+ ions was the most abundant species.

EXPERIMENTAL SECTION

Figure 1. Schematic setup of PR-nESI. (a) The strategy of high voltage. Typically, a negative high voltage of -3.0 kV was applied to the electrode for 6 s. Then the polarity of the high voltage was reversed, and a positive high voltage of +1.75 kV was applied to the electrode for the generation of electrospray. (b) The setup of PR-nESI. Except for the high voltage strategy, the other settings of PR-nESI were the same as nano-ESI.

PR-nESI was generated by using a polarity-reversing high voltage strategy (Figure 1). All of the analytes were detected in positive polarity mode. A negative high voltage was first applied to the electrode for a certain period of time (Figure 1a). The value and duration time of the negative high voltage

Figure 2. Detection results of cyt c obtained by nano-ESI and PRnESI. Cyt c was dissolved in ultrapure water in a concentration of 10 µM. (a) Mass spectrum of cyt c obtained by Nano-ESI. The spray voltage was +1.75 kV. (b) Amplified mass spectrum of the 8+ charged peak in panel (a). The S/N represented the signal to noise ratio of the protonated ion peak. (c) Mass spectrum of cyt c obtained by PR-nESI. A negative high voltage of -3.0 kV was first applied to the electrode and maintained for 6 s, then a positive high voltage of +1.75 kV was applied to generate electrospray. (d) Amplified mass spectrum of the 8+ charged peak in panel (c).

2

ACS Paragon Plus Environment

Page 3 of 10

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

The results were improved when PR-nESI was used for the detection of cyt c (Figure 2c). A unimodal CSD was also observed. The CSD mainly ranged from 7+ to 10+ and centered at 8+, showing that no unfolding of cyt c occurred. The intact folded structure of cyt c was maintained during PR-nESI. More importantly, the signal intensity of protonated cyt c ions was significantly enhanced. Protonated cyt c ions became the most abundant species. In comparison, the signal intensity of Na+ adducts was effectively reduced to a very low level (Figure 2d). The reduction of Na+ adducts simplified the spectrum, making it more convenient to identify the analytes. To quantify the enhancement of signal intensity, the S/N of protonated cyt c ion was calculated and compared between nanoESI and PR-nESI. Taken into consideration of the fact that the 8+ charged peak (m/z = 1529) was the strongest peak in the spectra of both nano-ESI and PR-nESI, the S/N of this peak was calculated for the comparison. The signal intensity of the noise within the mass range of m/z = 1480−1500 (The left side noise of the 8+ charged peak) and m/z = 1690−1710 (The right side noise of the 8+ charged peak) was averaged to give a final mean noise. The signal intensity of the 8+ charged peak was compared to the mean noise to give the S/N. The S/N of the protonated 8+ charged peak obtained by nano-ESI was 111 (Figure 2b). In comparison, the S/N obtained by PR-nESI was 2180 (Figure 2d). The S/N was enhanced by 20 folds.

obtained by nano-ESI was 3. The value obtained by PR-nESI was 156, which was 52 folds higher than that of nano-ESI. Further reducing the concentration to 10 nM resulted in a failure for nanoESI to detect cyt c. No signal of cyt c could be identified from the spectrum. In comparison, PR-nESI successfully achieved the detection of cyt c and gave a S/N of 58. When the concentration was decreased to 1 nM, PR-nESI still achieved the detection cyt c and gave a S/N of 5. No signal of cyt c could be identified from the spectrum obtained by nano-ESI. Duration Time of the Negative High Voltage. The duration time of the negative high voltage was of great importance to the final signals. Thus it was specially optimized. Cyt c was used as the analyte. It was dissolved in ultrapure water in a concentration of 10 µM for direct analysis. Different length of duration time was investigated, including 1 s, 3 s, 6 s, and 9 s (Figure 4). The value of the negative high voltage was fixed at -3.0 kV. The negative high voltage was maintained for a certain period of time and then turned into a positive high voltage of +1.75 kV to generate electrospray.

Figure 4. Optimization of the duration time of negative high voltage. An aqueous solution of 10 µM cyt c dissolved in ultrapure water was used as the sample. A negative high voltage of -3.0 kV was first applied to the electrode and maintained for different periods of time, and then a positive high voltage of +1.75 kV was applied to generate electrospray. (a) A representative EIC of the protonated 8+ charged cyt c ion under different duration time. (b) The ratio of RPR-nESI/RnESI under different duration time. RPR-nESI represented the S/N ratio of protonated 8+ charged cyt c ion obtained by PR-nESI, while RnESI represented that obtained by nano-ESI. Each test was repeated for five times.

Figure 3. Detection results of cyt c in lower concentrations. Cyt c was dissolved in ultrapure water in different concentrations. Mass spectra of cyt c obtained by nano-ESI in the concentration of 100 nM (a), 10 nM (c) and 1 nM (e). The spray voltage was +1.75 kV. Mass spectra of cyt c obtained by PR-nESI in the concentration of 100 nM (b), 10 nM (d) and 1 nM (f). A negative high voltage of -3.0 kV was first applied to the electrode and maintained for 6 s. Then a positive high voltage of +1.75 kV was applied to generate electrospray.

The enhancement of signal intensity was even more remarkable at lower concentrations (Figure 3). For cyt c sample with a concentration of 100 nM, the S/N of the protonated 8+ charged peak

A representative extracted ion chromatograph (EIC) of the protonated 8+ charged cyt c ion under different duration time was shown in Figure 4a. During the test, a positive high voltage of +1.75 kV was first applied to generate conventional nano-ESI. A relatively low signal intensity was obtained. Afterwards, PR-nESI was carried out. A negative high voltage of -3.0 kV was applied and maintained for 1 s, and then turned back into the positive high voltage of +1.75 kV. A sharp increase of the signal intensity was observed. The signal intensity soon reached a top and then went down to the level of conventional nano-ESI. This process was repeated with longer duration time of 3 s, 6 s and 9 s. With the increase of duration time, the signal intensity reached a higher top. At the same time, it took a longer time to reach the top. To quantify the enhancement of signal intensity under different duration time, the S/N ratio of the protonated 8+ charged cyt c ion obtained by PR-nESI (RPR-nESI) was compared with that obtained by nano-ESI (RnESI) to give the ratio of RPR-nESI/RnESI. As was shown in Figure 4b, the RPR-nESI/RnESI gradually increased from 13 to 20 with extended duration time. However, the increase became gentle from 6 s to 9 s. Taking into consideration of the fact that it would take rather a long time for PR-nESI to reach the top of the

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 10

showed that Na+ adducts were significantly more abundant than protonated holo-Mb ions (Figure 6b). Multiple Na+ ions were bonded to holo-Mb molecules. Adducts with three and four Na+ ions were the most abundant species. The results were improved when using PR-nESI for the detection of holo-Mb (Figure 6c). A unimodal CSD was also observed. The CSD mainly ranged from 9+ to 14+ and centered at 10+, corresponding to a folded structure of holo-Mb. Apo-Mb could hardly be identified from the spectrum. The non-covalently bounded heme group was preserved during PR-nESI. Besides, the the signal intensity of protonated holo-Mb ions was significantly enhanced. Protonated holo-Mb ions became the most abundant species, while the signal intensity of Na+ adducts was effectively reduced to a very low level (Figure 6d). The reduction of Na+ adducts simplified the spectrum and made it more convenient for the identification of the analytes.

signal intensity under the duration time of 9 s (Figure 4a), the duration time of 6 s was chosen for subsequent experiments. The Value of the Negative High Voltage. The value of the negative high voltage also had considerable impact on the final signals. In order to find out the optimal value, different values were investigated including -2.5 kV, -2.75 kV, -3.0 kV, -3.5 kV, -4.0 kV and -5.0 kV (Figure 5). The duration time was fixed at 6 s. The negative high voltage was applied to the electrode and maintained for 6 s. Then a positive high voltage of +1.75 kV was applied to generate electrospray. The ratio of RPR-nESI/RnESI was also used to quantify the changes of signal intensity under different values. According to the results, the RPR-nESI/RnESI increased from 12 to 15 with the rise of the negative high voltage from -2.5 kV to -3.0 kV. Further increasing the negative high voltage from -3.0 kV to -5.0 kV resulted in a decrease of the RPR-nESI/RnESI from 15 back to 12. As the highest ratio of RPR-nESI/RnESI was obtained under the value of -3.0 kV, this value was chosen for subsequent experiments.

Figure 5. The ratio of RPR-nESI/RnESI obtained under different negative high voltages. RPR-nESI represented the S/N ratio of protonated 8+ charged cyt c ion obtained by PR-nESI, while RnESI represented that obtained by nano-ESI. An aqueous solution of 10 µM cyt c dissolved in ultrapure water was used as the sample. Different values of negative high voltage were investigated including -2.5 kV, -2.75 kV, -3.0 kV, -3.5 kV, -4.0 kV and -5.0 kV. The negative high voltage was applied to the electrode and maintained for 6 s. Then a positive high voltage of +1.75 kV was applied to generate electrospray. Each test was repeated for five times.

Detection of Holo-Mb. PR-nESI was further applied to the detection of holo-Mb so as to confirm its detection capability for proteins. Holo-Mb is another commonly featured protein for mass spectrometry studies. Native holo-Mb has a tightly folded conformation with a non-covalently bounded heme group in a hydrophobic pocket.31 The non-covalent complex (holo-Mb) could be preserved under native electrospray conditions.32-34 When exposed to acidic environments, holo-Mb rapidly denatures via a short-lived unfolded intermediate before it loses the heme group and yields apo-Mb.35 HoloMb was dissolved in ultrapure water in a concentration of 10 µM for direct detection. A representative nano-ESI mass spectrum of holo-Mb was shown in Figure 6a. Both holo-Mb and apo-Mb were identified. The CSD of holo-Mb mainly ranged from 10+ to 17+ and centered at 12+. Besides, the 9+ charged peak represented a more folded structure of holo-Mb molecules. The CSD of apo-Mb ranged from 10+ to 16+ and centered at 11+. Severe Na+ ion adduction was observed. The amplified spectrum of the 12+ charged peak of holo-Mb

Figure 6. Detection results of holo-Mb obtained by nano-ESI and PRnESI. Holo-Mb was dissolved in ultrapure water in a concentration of 10 µM. (a) Mass spectrum of holo-Mb obtained by Nano-ESI. The spray voltage was +1.75 kV. (b) Amplified mass spectrum of the 12+ charged peak in panel (a). The S/N represented the signal to noise ratio of the protonated 10+ charged ion peak. (c) Mass spectrum of holo-Mb obtained by PR-nESI. A negative high voltage of -3.0 kV was first applied to the electrode and maintained for 6 s, then a positive high voltage of +1.75 kV was applied to generate electrospray. (d) Amplified mass spectrum of the 10+ charged peak in panel (c).

Quantification of the enhancement of signal intensity was also carried out. The S/N of protonated 10+ charged holo-Mb ion was calculated and compared between nano-ESI and PR-nESI. The signal intensity of the noise within the mass range of m/z = 1730−1740 (the left side noise of the 10+ charged peak) and m/z = 1800−1810 (the right side noise of the 10+ charged peak) was averaged to give a final mean noise. The signal intensity of the 10+ charged peak was compared to the mean noise to give the S/N. The S/N of the protonated 10+ charged peak obtained by nano-ESI was 13 (Figure 6b). In comparison, the S/N obtained by PR-nESI was 5789 (Figure 6d). The S/N was enhanced by 445 folds. Detection of Insulin. Besides holo-Mb, insulin was also used to confirm the detection capalibity of PR-nESI. Insulin was a relatively small protein. It was dissolved in ultrapure water in a concentration of 10 µM for direct detection. A representa-

4

ACS Paragon Plus Environment

Page 5 of 10

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

tive nano-ESI mass spectrum of insulin was shown in Figure S1a. A unimodal CSD was observed. The CSD ranged from 3+ to 5+ and centered at 5+. Compared with cyt c and holoMb, the overall signal intensity of insulin was much weaker and the adduction of Na+ ions with insulin molecules was more serious. The amplified spectrum of the 5+ charged peak showed that multiple Na+ ions were bonded to insulin molecules and adduct with five Na+ ions was the most abundant species (Figure S1b). The protonated 5+ charged ion peak could hardly be identified from the spectrum. When PR-nESI was applied to the detection of insulin, a better result was obtained (Figure S1c). A unimodal CSD was observed. It ranged from 3+ to 5+ and centered at 5+. The signal intensity of Na+ adducts were effectively reduced (Figure S1d). Mainly protonated peaks were observed in the spectrum. The signal intensity of protonated peaks was remarkably enhanced. Considering that no protonated 5+ charged ion peak was identified in the spectrum obtained by nano-ESI (Figure S1b), the S/N of protonated 4+ charged insulin ion was calculated and compared between nanoESI and PR-nESI. The signal intensity of the noise within the mass range of m/z = 1400−1420 (the left side noise of the 4+ charged peak) and m/z = 1560−1580 (the right side noise of the 4+ charged peak) was averaged to give a final mean noise. The signal intensity of the protonated 4+ charged peak was compared to the mean noise to give the S/N. The S/N of the protonated 4+ charged peak obtained by nano-ESI was 6 (Figure S1a). In comparison, the S/N obtained by PR-nESI was 2252 (Figure S1c). The S/N was enhanced by 375 folds. Desalting Effect. Metal ion adducts of proteins were easily formed during ionization when metal ions were present in the solution. This had been proven by the results of nano-ESI above. Even though the proteins were dissolved in ultrapure water, severe metal ion adduction was observed. Among the three investigated proteins, insulin formed Na+ adducts more seriously than the other two proteins (Figure S1a and b). No protonated 5+ charged peak was identified in the spectra, though the 5+ charged ion was the most abundant species of insulin. The results were improved when PR-nESI was employed for the detection. The signal intensity of protonated insulin ions were significantly enhanced, while the signal intensity of Na+ adducts were effectively reduced (Figure S1c and d). A desalting effect was involved in the process of PR-nESI. The desalting effect had also been proven by the detection results of cyt c and holo-Mb.

Figure 7. Detection results of insulin samples added with NaCl. Insulin was dissolved in ultrapure water in a concentration of 10 µM. 1 mM NaCl was added to the sample before detection. (a) Mass spectrum obtained by nano-ESI. The spray voltage was +1.75 kV. (b) Mass spectrum obtained by PR-nESI. A negative high voltage of -3.0

kV was first applied to the electrode and maintained for 6 s, then a positive high voltage of +1.75 kV was applied to generate electrospray.

Further investigations were made into the desalting capability of PR-nESI. An aqueous solution of 10 µM insulin dissolved in ultrapure water was prepared as the sample. 1 mM NaCl was added to the sample before detection. According to the results obtained by nano-ESI, the spectrum was dominated by the signals of NaCl clusters (Figure 7a). The signal of insulin was barely identified with serious Na+ adduction. In comparison, PR-nESI successfully avoided the interference caused by NaCl and achieved detection of insulin (Figure 7b). No NaCl cluster was identified in the spectrum. Only the signal of insulin was observed. Still, a unimodal CSD was observed in the spectrum. The CSD ranged from 3+ to 5+ and centered at 5+. The peaks observed in the spectrum were all protonated peaks. No Na+ adduct was identified. The results above showed that PR-nESI not only enhanced the signal intensity of proteins, but also achieved effective desalting of the sample. Investigations were also made to see if PR-nESI was tolerant to commercial PBS buffers. Insulin was dissolved in commercial 0.1×PBS buffer in a concentration of 1 µM. No signal of insulin was identified from the spectrum obtained by nano-ESI (Figure S2a). Only salt clusters were observed. PBS buffer was a mixture of several nonvolatile salts with high concentrations. The presence of PBS buffer in samples could cause severe suppression effect. When PR-nESI was used, the result was significantly improved (Figure S2b). Salt clusters were removed. A clear unimodal CSD of insulin was observed. Applications. PR-nESI was applied to the analysis of the noncovalent protein-ligand system of lysozyme and NAG3. Lysozyme was dissolved in ultrapure water in a concentration of 1 µM. NAG3 was added to the sample in a concentration of 4 µM. Direct analysis of the sample by nano-ESI achieved the detection of the noncovalent complexes of lysozyme and NAG3 (Figure S3a). However, serious Na+ adduction was observed. The presence of Na+ adducts affected the identification of the peaks. When the sample was buffered with 10 mM NH4Ac, the CSD of lysozyme and the noncovalent complexes shifted to a lower charge state (Figure S3b). The presence of NH4Ac inhibited the formation of Na+ adducts, making it more convenient for the identification of the peaks. However, the abundance of the noncovalent complexes was reduced compared with the result obtained without NH4Ac. The signal of the lysozyme-(NAG3)2 complex became rather weak. When PR-nESI was used for the analysis of the sample, an improved result was obtained (Figure S3c). Na+ adducts were eliminated. High abundances of lysozyme-NAG3 and lysozyme(NAG3)2 complexes were identified. Three unimodal CSDs were observed in the spectrum, representing lysozyme, lysozymeNAG3 and lysozyme-(NAG3)2 complexes, respectively. The system of lysozyme and maltohexaose was also investigated (Figure S4). Quite Similar results were obtained. PR-nESI was also applied to the analysis of the tryptic digest of Mb. As was shown in Figure S5a, a very complicated mass spectrum was obtained by nano-ESI. High abundances of Na+ adducts were observed, which severely affected the identification of the product peptides. Although 13 peptides were identified from the spectrum (Table S1), many of them had quite low signal intensi-

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

ties. The amino acid sequence of Mb was shown in Table S2. When PR-nESI was used for the analysis of the digest, an improved mass spectrum was obtained (Figure S5b). Na+ adducts were successfully removed. Only protonated peptide ions were observed, which remarkably simplified the spectrum. Furthermore, the signals of many peptides were enhanced at the same time. This was advantageous to the identification of the product peptides (Table S3). Two more peptides were identified from the spectrum obtained by PR-nESI with high abundances. Finally, the detection capability of PR-nESI for small molecule compounds was confirmed. Different kinds of compounds were investigated, including drugs (Figure S6), amino acids (Figure S7), peptides (Figure S8) and some other organic compounds (Figure S9). PR-nESI remarkably enhanced the signals of these compounds and removed the Na+ adducts at the same time. Investigation of the Mechanism. The enhancement of signal intensity and desalting effect were supposed to be related to the electromigration of the ions in samples. The electromigration of ions driven by the applied high voltage had been proven to play an important role during ESI.36-38 When conventional nano-ESI was carried out in positive polarity, all the positive charged ions were forced to migrate towards the tip. Metal ions had a relatively higher mobility than proteins, thus they were better accumulated to the tip than proteins. As a result, the adduction of metal ions with protein molecules was aggravated. To solve this problem, a positive step voltage was employed by Zhang’s group.26 Typically, a high voltage of +5.2 kV was first applied to the electrode and maintained for 30 s. This high voltage significantly accelerated the migration of ions, resulting in the separation of metal ions and analytes. Then a lower high voltage of +2.4 kV was applied to the electrode to generate electrospray. The signals of analytes and matrices were obtained at different times due to the separation achieved in the previous step. In this work, we made use of a polarity-reversing high voltage strategy. Typically, a negative high voltage of -3.0 kV was first applied to the electrode and maintained for 6 s. The difference in migration speed led to the separation of protein and metal ions. Then a positive high voltage of +1.75 kV was applied to generate electrospray. The signals of analytes and metal ions were obtained at different times as a result of the previous separation step. The mechanism of PR-nESI was similar to that of SV-nano-ESI developed by Zhang’s group. But the polarity-reversing high voltage strategy used in PR-nESI significantly improved the perfomance. A much lower high voltage of -3.0 kV was employed in PR-nESI for the separation. And this negative high voltage was maintained for a much shorter time of only 6 s. A remarkable enhancement of S/N was obtained for the investigated proteins using PR-nESI. No unfolding of the proteins was observed during detection. For acidic proteins, such as insulin (pI = 5.4), they were negatively charged when dissolved in pure water. When the negative high voltage of -3.0 kV was applied, Na+ and protein ions moved in different directions (Figure S10a). Na+ ions moved towards the electrode while protein ions moved towards the tip, resulting in the separation of the two ion species. When the negative high voltage was switched into the positive high voltage of +1.75 kV,

Page 6 of 10

the protein ions were directly sprayed out of the tip. The unidirectional flow of the solution towards the tip further promoted this process (Figure S10b). The signal of protein ions was thus obtained in the first place. In comparison, the signal of Na+ ions was delayed because it took time for Na+ ions to move to the tip. Further experimental confirmations were made on the hypothesis. Insulin was dissolved in ultrapure water in a concentration of 10 µM for direct analysis. The EIC of the 5+ charged insulin ion showed a fast increase and soon reached its top value (the blue zone in Figure S11a). Only protonated peaks were observed during this time with remarkably enhanced signal intensity (Figure S11c). It took a little time for the signal to reach its top. This was due to the fact that when the negative high voltage was applied, the tip was negatively charged. When the positive high voltage was just applied, it took time for the tip to get positively charged. The signal of insulin was thus delayed. In comparison, the trend of Na+ ions (represented by the Na+ adducts) was right on the contrary. The signal intensity of Na+ adducts was kept at a rather low level at the beginning. Then, the signal intensity increased gradually to a much higher level (The red zone in Figure S11b). Mass spectrum obtained during this time was dominated by the peaks of Na+ adducts (Figure S11d). For basic proteins, such as cyt c (pI = 10.4), they were positively charged when dissolved in pure water. When the negative high voltage of -3.0 kV was applied, Na+ and protein ions moved in the same direction towards the electrode (Figure S12a). Na+ ions had a much higher migration speed. Thus they moved a further distance, resulting in the separation of Na+ from protein ions. When the positive high voltage of +1.75 kV was applied, Na+ and protein ions moved towards the tip (Figure S12b). As the solution was sprayed out of the tip under both the negative and positive high voltages, it kept flowing towards the tip. This unidirectional flow of the solution made it possible for protein ions to spray out of the tip before they were caught up by Na+ ions. The signal of Na+ and protein ions was thus obtained at different times. As it would take time for protein ions to move to the tip, a gradual increase of the signal intensity was observed. Experimental confirmation was also carried out using cyt c solutions (Figure S13). The result matched well with the hypothesis. Mass spectra obtained during the application of the negative high voltage of -3.0 kV were investigated. A rather weak signal was obtained and no signal of cyt c was observed (Figure S14a). In comparison, a much stronger signal was obtained when conventional nano-ESI was carried out with a lower negative high voltage of -1.6 kV (Figure S14b) for the same sample. Cyt c ion with a charge state of 7- was identified from the spectrum. The results indicated that the negative high voltage of -3.0 kV in PRnESI might be too high to form a stable electrospray. A discharge process was supposed to happen during this time, resulting in a decayed signal. The nano-tips were checked under a microscope to see if the high voltage used in PR-nESI would have an impact on their internal diameter (Figure S15). An increase of the tip internal diameter was observed from 3 µm to 5 µm after PR-nESI. Mass spectra were acquired to ensure that PR-nESI was carried out successfully (Figure S16a). The increase of the internal diameter was supposed to be due to the discharge occurred during the negative high volt-

6

ACS Paragon Plus Environment

Page 7 of 10

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

age. Further investigations were made to see if the improved performance obtained by PR-nESI was related to the increase of tip internal diameter. After PR-nESI, the tip in Figure S15c was used to carry out conventional nano-ESI (Figure S16b). The signal decayed remarkably compared with that in Figure S16a. High abundances of Na+ adducts were observed. This revealed that the improved performance obtained by PR-nESI was not caused by the increase of tip internal diameter. Although the internal diameter of the tips was changed, it had little impact on the performance of PR-nESI. PR-nESI could be repeated on the same tip for multiple times (Figure S17). According to the EIC of the protonated 8+ charged cyt c ion, the signals obtained during the repeated four cycles were about the same. PR-nESI was tolerant to the change of the tip internal diameter. Finally, the detection capability of PR-nESI in negative polarity was investigated. An aqueous solution of 10 µM insulin added with 1 mM NaCl was use as the sample. Conventional nano-ESI failed in the detection of insulin (Figure S18a), due to the presence of NaCl. In comparison, detection of insulin was successfully achieved by PR-nESI (Figure S18b). The interference of NaCl was overcome by PR-nESI. As was revealed by Figure 7 and Figure S18, insulin could be detected in both positive and negative polarities. However, positive polarity was recommended. This was because positive ion mode was more commonly used in the detection of proteins and peptides. Most standard analysis methods and spectral libraries of proteins and peptides were built in positive ion mode. The mechanism of PR-nESI enabled the detection of both basic and acidic proteins and peptides in positive ion mode. Thus positive ion mode was used in most cases in the present work. An alternating-current (AC) high voltage was actually used for the generation of PR-nESI. However, the mechanism of PR-nESI was quite different from that of conventional AC-ESI. The key difference lay in the frequency of the AC voltage. Typically, a high-frequency AC voltage was used for the generation of ACESI.39,40 The frequency of the AC voltage generally ranged from 80−400 kHz,41 resulting in a rather limited time for ion migration. AC-ESI produced a comparable signal to noise ratio (S/N) to that of conventional direct-current electrospray ionization (DC-ESI), but its signal intensity was much lower.41 In PR-nESI, the negative high voltage was maintained for 6 s before it was switched into positive high voltage. This ensured sufficient time for the migration of ions, leading to more complete separation of protein and metal ions. Much higher signal intensity and S/N were obtained by PR-nESI than DC-ESI. In PR-nESI, protein and metal ions were separated and sequentially detected based on their migration speed difference in solutions. This was beneficial for the elimination of the suppression effect caused by metal ions. Separation and sequential detection of the components in samples had also been achieved by other ionization sources, such as probe electrospray (PESI).42,43 But the mechanism was different. In PESI, analytes were detected sequentially in the order of their surface activities. Interactions between analytes and the probe surface were also supposed to have an impact on the detection sequence.44 Electrospray generated on porous substrates allowed for sequential detection of proteins based on their molecular sizes and shapes.45 This technique made use of the separation effect of proteins on porous substrates, fol-

lowed by in situ electrospray of the separated proteins from the tip. Ion migration played an important role in PR-nESI. It was also made use of by analyte migration electrospray ionization (AMESI) for the removal of biological matrix from analytes.46 Though similar, the separation mechanisms of the two methods were different. In PR-nESI, the separation of different ion species was based on their migration speed difference in solution. But in AMESI, the separation of analytes from biological matrix was actually based on their solubility difference in the solvent. Analytes were easily extracted from the biological sample into the solvent due to their high solubility in the solvent, while biological matrix remained in the sample due to its limited solubility in the solvent. Subsequent migration of the analytes in the solvent led to the transfer of the analytes from the back end of the solvent to the tip. The analytes were then sprayed out from the tip. Ion migration played different roles in PR-nESI and AM-ESI. Although the signal intensity of analytes changed over time in PR-nESI, quantitative analysis could be achieved by adding isotopic internal standard into samples. Analytes and their isotopic internal standards had very similar physical and chemical properties. Thus they underwent almost the same migration and ionization process. Their relative abundance in the obtained mass spectra was steady. Quantification could then be achieved by comparing the relative abundance of analytes and their isotopic internal standards in the obtained mass spectra.

CONCLUSIONS In this work, a novel ionization method, namely PR-nESI, was developed. It made used of a polarity-reversing high voltage strategy for the generation of nano-ESI. Significant enhancement of signal intensity was achieved compared with conventional nanoESI. The S/N was increased by 1−2 orders of magnitude. Remarkable desalting effect was also involved in PR-nESI. Only protonated protein ions were identified in the obtained mass spectra, even if the concentration of salt was increased to mM level. Metal ion adducts were effectively removed. Furthermore, the intact folded structure of proteins was preserved during PR-nESI. No unfolding was observed. Investigations into the mechanism of PR-nESI revealed that the enhancement of signal intensity and desalting effect were related to the electromigration of the solutes in solutions. Compared with conventional methods for the enhancement of signal intensity, PR-nESI required no additional additives and no modifications of the experimental setup of nano-ESI. Only a novel high voltage strategy was used, which was simple in manipulation. The desalting effect was also advantageous, with no pretreatment needed. With all these advantages, PR-nESI would be a promising method in future analytical and bio-analytical applications.

ASSOCIATED CONTENT Supporting Information. Experimental details; detection results of insulin samples, noncovalent complexes of lysozyme and saccharides, Mb digest, small molecule compounds including drugs, amino

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

acids, peptides, RB and diphenylamine; schematic graph of the mechanism of PR-nESI.

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected], fax: +86-10-6452 6387.

Author Contributions All authors have given approval to the final version of the manuscript.

Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT This work is supported by the Ministry of Science and Technology of China (No. 2016YFF0102603) and the National Natural Science Foundation of China (Nos. 21575132 and 21605135). We sincerely thank our teammates Chang Li, Shiying Chu and Rui Zhai for their help in the experiments. We also thank Yuxuan Luo for his help in the literatures.

REFERENCES (1) (2) (3) (4) (5)

(6) (7) (8) (9)

(10)

(11) (12) (13) (14) (15) (16)

Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M., Science, 1989, 246, 64-71. Kebarle, P.; Tang, L., Anal. Chem., 1993, 65, A972-A986. Kebarle, P.; Verkerk, U. H., Mass Spectrom. Rev., 2009, 28, 898-917. Wilm, M. S.; Mann, M., Int. J. Mass Spectrom., 1994, 136, 167-180. Wilm, M.; Shevchenko, A.; Houthaeve, T.; Breit, S.; Schweigerer, L.; Fotsis, T.; Mann, M., Nature, 1996, 379, 466-469. Gibson, G. T. T.; Mugo, S. M.; Oleschuk, R. D., Mass Spectrom. Rev., 2009, 28, 918-936. El-Faramawy, A.; Siu, K. W. M.; Thomson, B. A., J. Am. Soc. Mass Spectrom., 2005, 16, 1702-1707. Marginean, I.; Kelly, R. T.; Page, J. S.; Tang, K. Q.; Smith, R. D., Anal. Chem., 2007, 79, 8030-8036. Sun, L. L.; Zhu, G. J.; Zhao, Y. M.; Yan, X. J.; Mou, S.; Dovichi, N. J., Angew. Chem.-Int. Edit., 2013, 52, 1366113664. Waanders, L. F.; Chwalek, K.; Monetti, M.; Kumar, C.; Lammert, E.; Mann, M., Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 18902-18907. Yue, G. H.; Luo, Q. Z.; Zhang, J.; Wu, S. L.; Karger, B. L., Anal. Chem., 2007, 79, 938-946. Luo, Q.; Yue, G.; Valaskovic, G. A.; Gu, Y.; Wu, S. L.; Karger, B. L., Anal. Chem., 2007, 79, 6174-6181. Schmidt, A.; Karas, M.; Dulcks, T., J. Am. Soc. Mass Spectrom., 2003, 14, 492-500. Juraschek, R.; Dulcks, T.; Karas, M., J. Am. Soc. Mass Spectrom., 1999, 10, 300-308. Karas, M.; Bahr, U.; Dulcks, T., Fresenius J. Anal. Chem., 2000, 366, 669-676. Want, E. J.; Nordstrom, A.; Morita, H.; Siuzdak, G., J. Proteome Res., 2007, 6, 459-468.

Page 8 of 10

(17) Xu, J. J.; Gao, J.; Yu, C. L.; He, H.; Yang, Y. M.; Figeys, D.; Zhou, H., Anal. Chem., 2016, 88, 583-591. (18) Kelly, R. T.; Page, J. S.; Tang, K. Q.; Smith, R. D., Anal. Chem., 2007, 79, 4192-4198. (19) Kelly, R. T.; Page, J. S.; Zhao, R.; Qian, W. J.; Mottaz, H. M.; Tang, K. Q.; Smith, R. D., Anal. Chem., 2008, 80, 143149. (20) Kelly, R. T.; Page, J. S.; Marginean, I.; Tang, K. Q.; Smith, R. D., Anal. Chem., 2008, 80, 5660-5665. (21) Huang, G. M.; Li, G. T.; Cooks, R. G., Angew. Chem.-Int. Edit., 2011, 50, 9907-9910. (22) Wei, Z. W.; Xiong, X. C.; Guo, C. A.; Si, X. Y.; Zhao, Y. Y.; He, M. Y.; Yang, C. D.; Xu, W.; Tang, F.; Fang, X.; Zhang, S. C.; Zhang, X. R., Anal. Chem., 2015, 87, 1124211248. (23) Li, A. Y.; Hollerbach, A.; Luo, Q. J.; Cooks, R. G., Angew. Chem.-Int. Edit., 2015, 54, 6893-6895. (24) Gong, X. Y.; Xiong, X. C.; Wang, S.; Li, Y. Y.; Zhang, S. C.; Fang, X.; Zhang, X. R., Anal. Chem., 2015, 87, 97459751. (25) Wojcik, R.; Dada, O. O.; Sadilek, M.; Dovichi, N. J., Rapid Commun. Mass Spectrom., 2010, 24, 2554-2560. (26) Wei, Z. W.; Han, S.; Gong, X. Y.; Zhao, Y. Y.; Yang, C. D.; Zhang, S. C.; Zhang, X. R., Angew. Chem.-Int. Edit., 2013, 52, 11025-11028. (27) Bushnell, G. W.; Louie, G. V.; Brayer, G. D., J. Mol. Biol., 1990, 214, 585-595. (28) Shastry, M. C. R.; Luck, S. D.; Roder, H., Biophys. J., 1998, 74, 2714-2721. (29) Konno, T., Protein Sci., 1998, 7, 975-982. (30) Goto, Y.; Hagihara, Y.; Hamada, D.; Hoshino, M.; Nishii, I., Biochemistry, 1993, 32, 11878-11885. (31) Evans, S. V.; Brayer, G. D., J. Mol. Biol., 1990, 213, 885897. (32) Katta, V.; Chait, B. T., J. Am. Chem. Soc., 1991, 113, 8534-8535. (33) Li, Y. T.; Hsieh, Y. L.; Henion, J. D.; Ganem, B., J. Am. Soc. Mass Spectrom., 1993, 4, 631-637. (34) Loo, J. A., Mass Spectrom. Rev., 1997, 16, 1-23. (35) Konermann, L.; Rosell, F. I.; Mauk, A. G.; Douglas, D. J., Biochemistry, 1997, 36, 6448-6454. (36) Blades, A. T.; Ikonomou, M. G.; Kebarle, P., Anal. Chem., 1991, 63, 2109-2114. (37) Van Berkel, G. J.; Kertesz, V., Anal. Chem., 2007, 79, 5510-5520. (38) Vanberkel, G. J.; Zhou, F. M., Anal. Chem., 1995, 67, 3958-3964. (39) Chetwani, N.; Cassou, C. A.; Go, D. B.; Chang, H. C., J. Am. Soc. Mass Spectrom., 2010, 21, 1852-1856. (40) Chetwani, N.; Cassou, C. A.; Go, D. B.; Chang, H. C., Anal. Chem., 2011, 83, 3017-3023. (41) Sarver, S. A.; Chetwani, N.; Dovichi, N. J.; Go, D. B.; Gartner, C. A., J. Am. Soc. Mass Spectrom., 2014, 25, 524529. (42) Mandal, M. K.; Chen, L. C.; Hiraoka, K., J. Am. Soc. Mass Spectrom., 2011, 22, 1493-1500.

8

ACS Paragon Plus Environment

Page 9 of 10

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

(43) Mandal, M. K.; Chen, L. C.; Yu, Z.; Nonami, H.; ErraBalsells, R.; Hiraoka, K., J. Mass Spectrom., 2011, 46, 967-975. (44) Hu, B.; So, P. K.; Chen, H. W.; Yao, Z. P., Anal. Chem., 2011, 83, 8201-8207. (45) Hu, B.; Yao, Z. P., Anal. Chem., 2016, 88, 5585-5589. (46) Wang, Q.; Zhong, H. B.; Zheng, Y. J.; Zhang, S. C.; Liu, X. N.; Zhang, X. R.; Zhang, X. L.; Zhang, Z. P., Analyst, 2014, 139, 5678-5681.

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

TOC

10 ACS Paragon Plus Environment

Page 10 of 10