In the study of degenerative brain diseases, changes in lipids, the main component of neurons, are particularly important because they are used as indicators of pathological changes. One method for the sensitive measurement of biomolecules, especially lipids, is time-of-flight secondary ion mass spectrometry (ToF-SIMS) using pulsed argon cluster ions. In this study, biomolecules including various lipids present in normal mouse brain tissue were measured using ToF-SIMS equipped with pulsed argon cluster primary ions. Based on the ToF-SIMS measurement results, hybrid SIMS (OrbiSIMS), which is a ToF-SIMS system with the addition of an orbitrap mass analyzer, was used to directly identify the biomolecules by the region in the real tissue samples. For this, the results of ToF-SIMS, which measured the tissue samples from a single mouse brain within static limits, were compared with those from OrbiSIMS measured beyond the static limits in terms of the differences in molecular profiling. From this analysis, two types of positive and negative ions were selected for identification, with the OrbiSIMS MS/MS results indicating that the positive ions were glycerophosphocholine and the negative ions were glycerophosphoinositol and sulfatide, a sphingolipid. Then, to confirm the identification of the molecular candidates, lipids were extracted from mirror image tissue samples, and LC-MS/MS also using an orbitrap mass analyzer was performed. As a result, the direct identification of molecular candidate groups distributed in particular regions of the tissue samples via OrbiSIMS was found to be consistent with the identification results by LC-MS/MS for extracted samples.

ToF-SIMS, or time-of-flight secondary ion mass spectrometry, is a mass spectrometry method where secondary ions emitted from the sample surface as a result of sputtering by incident pulsed primary ions are measured. This method, which is carried out within static limits, has the advantage of being able to analyze organic materials including biomolecules.1,2 A static limit refers to using extremely low amounts of primary ions (1013 ions/cm2 or less) so that less than 1% of the top surface layer of an atom or molecule is ion bombarded. In particular, the use of argon cluster ions as primary ions, as an alternative to the more common bismuth cluster ions, can provide measurements with higher intensities at higher mass ranges because of less fragmentation and a higher efficiency of the secondary ions.3,4

While argon cluster ion beams do not have good lateral resolution due to their large beam size, they can obtain full-size images of tissue samples, such as mouse brain tissue.5,6 ToF-SIMS is highly sensitive to lipid-related biomolecules among those present in tissue surfaces, so it has been used in the study of various degenerative brain diseases such as Alzheimer’s disease.7,8 However, for disease studies using tissue mass imaging via pulsed argon cluster ions, higher lateral resolution, mass resolution, and mass accuracy are required. When pulsed argon cluster ions are used, the isotope peaks can be separated by applying a delayed extraction mode to improve mass resolution.6 Despite these improvements, the identification of candidate molecules remains a challenge in secondary ion mass spectrometry. Recently, to support the identification of candidate molecules, MS/MS has been introduced into SIMS equipment. Equipment capable of MS/MS in ToF-SIMS analysis includes the J105 3D Chemical Imager (Ionoptika, UK),9–11 nano TOF III (Physical Electronics, USA),12 and hybrid SIMS (IONTOF, Germany).13 Candidate molecules are selectively introduced into a collision-induced dissociation cell for fragmentation, and the corresponding MS/MS spectra are obtained using an orbitrap mass analyzer13 or a second ToF mass analyzer.9–12 The former SIMS system utilizing an orbitrap is referred to as hybrid SIMS or OrbiSIMS. However, after the introduction of OrbiSIMS, no studies to date have compared the mass spectra obtained from the static mode of ToF-SIMS with the spectra from the dynamic mode of OrbiSIMS.

The purpose of this study is to compare the mass spectra obtained from ToF-SIMS and OrbiSIMS in order to investigate the potential of identifying biomolecules in tissue samples. First, a full image of a coronal section of mouse brain tissue was obtained using ToF-SIMS, and similar reconstructed images with sufficient intensity were grouped with the assumption that similar regions would have similar constituent biomolecules. The ToF-SIMS spectra measured within static limits were compared to the mass spectra from the outer layer of the cortex and corpus callosum areas via OrbiSIMS measurement beyond the static limits. Candidate molecules with a high intensity in the ToF-SIMS spectra were selected, and MS/MS was performed with OrbiSIMS to identify the candidate molecules. Finally, the identification results of the candidate molecules were cross-validated through a comparison with the gold standard LC-MS/MS analysis. The mass profiling was similarly obtained by comparing the mass spectra of ToF-SIMS obtained within static limits and OrbiSIMS obtained in a dynamic mode, but the OrbiSIMS results were more useful due to superior mass resolution and mass accuracy. OrbiSIMS and LC-MS/MS showed similar results in identifying biomolecules. The use of mass peaks obtained based on mass imaging will help to directly identify disease-related biomolecules in biological tissues.

A normal mouse (three-month-old male C57BL/6J) brain was obtained from a laboratory animal resource center, the Korea Research Institute of Bioscience and Biotechnology (KRIBB), and all experimental procedures were performed under policies of the Institutional Animal Care and Use Committee of KRIBB. The dissected mouse brain was immediately immersed in 2-methylbutane for 30 min to start the freezing process and then moved into an −80 °C deep freezer. Before sample preparation, the frozen mouse brain was transferred to a cryostat (Leica CM 3050S, Leica Microsystems Inc., IL) and left for 30 min to reach thermal equilibrium. The brain was fixed on a holder using optimal cutting temperature (OCT) compound and then cut into slices of 12 μm thickness. The slices were then placed on a prechilled stainless steel substrate and stored in the −80 °C deep freezer until analysis. The sample was prepared in a freeze-dried state by drying in a vacuum chamber for 3h on a stainless steel block at −80 °C before being introduced to the instrument.

The ToF-SIMS analysis was performed using a ToF-SIMS V (IONTOF GmbH) equipment. Mass spectra and images were obtained by using an argon cluster ion beam (Ar 3000+, 20 keV) with a 0.15 pA beam current. Delayed extraction modes were used in all analyses to improve mass resolution.6 Data were acquired in positive and negative modes from the same area within static limits using a primary ion dose density of approximately 2.5 × 1011 ions/cm2 for each polarity. Irganox 1010 was used for external mass calibration, where Irganox 1010 dissolved in acetone was dropped near the sample and dried. The mouse brain tissue sample was measured over a wide area using a stage scan, including the Irganox 1010 spot. To obtain images of the whole mouse brain tissue, a patch size of 30 μm and a stage raster size of 8610 × 6210 μm2 were used. For external mass calibration, using the ROI (region of interest) function, mass spectra from the Irganox 1010 region were obtained and applied to the mass spectrum of the mouse brain tissue.6 

The OrbiSIMS analysis was performed using the Hybrid SIMS (IONTOF GmbH) equipment. Data were obtained using the orbitrap mass analyzer with a long pulsed argon cluster ion beam (Ar2000 +, 20 keV) with a 113 pA beam current. Prior to the analysis, mass calibration of the orbitrap mass analyzer was performed using silver cluster secondary ions obtained from a silver plate using Bi1 + ions with 30 keV energy. The mass resolution was 240 000 at m/z 200 and the injection time was 500 ms. Electron flooding with an energy of 20 eV was used to compensate for charging. The surface potential voltage was adjusted in the range of −400 to +400 V to optimize the transfer of secondary ions from the sample to the orbitrap. The mass ranges for the full spectra and MS/MS were m/z 110–1650 and m/z 75–1125, respectively, and the isolation window for MS/MS was 2.5 m/z including isotopes. During MS/MS measurement, the collision energy of the high-energy collision dissociation (HCD) cell was varied between 2 and 150 eV. Long pulse argon cluster ion beams have a stronger beam current compared to the ToF-SIMS argon cluster ion beam, and thus, it is easy to exceed the static limit in a short time. For a mouse tissue sample, 30 s of irradiation in a 300 × 300 μm2 area results in an approximate primary ion dose density of 3.7 × 1013 ions/cm2. After measuring the tissue sample in the dynamic mode, tissue damage revealed the substrate.

For LC-MS/MS analysis, mouse brain tissue on the stainless steel substrate was scratched using a pipette tip and placed in a 1.5 ml tube. The sample was resuspended with mobile phase A and then injected to the LC-MS/MS system. Lipids were separated and analyzed using a nanoAcquity (Waters) and LTQ-Orbitrap Elite (Thermo Fischer Scientific). The spray voltages were set to 1.9 kV in the positive mode and 1.5 kV in the negative mode. The column was an AtlantisTM dC18 (3 μm diameter, 300 μm × 100 mm P/N186003505) set to 50 °C. Mobile phase A was 60:40 acetonitrile/water and mobile phase B was 90:10 isopropyl alcohol/acetonitrile, and both A and B contained 10 mM ammonium formate and 0.1% formic acid. The column gradient was developed using a step run from 32% B to 97% B in 21 min at a flow rate of 4 μl/min.

The survey scan settings were as follows: resolution = 120 000, max IT = 200 ms, AGC target = 1E6, and mass range m/z 300–2000. The selected precursor was fragmented by higher-energy collisional dissociation (HCD) and analyzed by the orbitrap mass analyzer. Other parameters for the MS/MS scan were as follows: Top 10 double play, resolution = 30 000, max IT = 500 ms, automatic gain control (AGC) target = 5E4, threshold 25 000, stepped collision energy = 46%, 50%, and 54%, isolation width = 1.0 m/z, dynamic exclusion parameter repeat count = 1, repeat duration time = 10 s, exclusion list size = 500, and exclusion duration time = 30 s.

To clarify what we are trying to determine between the different methods, a graphical representation of what was done in each experimental step is included as . ToF-SIMS measurements were performed on the entire mouse brain tissue sample using pulsed argon cluster ions. OrbiSIMS measurements were carried out on a specific area in the mouse brain tissue, such as the outer layer of the cortex or the corpus callosum, using long pulsed argon cluster ions. Comparison of the ToF-SIMS and OrbiSIMS measurement spectra from the same region of the brain reveals differences and similarities in static and dynamic modes. In addition, four lipid candidates from both regions were analyzed by MS/MS with OrbiSIMS. For validation, LC-MS/MS was performed using extracted samples from the same regions.

Scheme 1.

(a) ToF-SIMS measurements are performed on the entire tissue and reconstructed area (ROI), which is the same area as in the OrbiSIMS analysis. (b) OrbiSIMS measurements are carried out on specific areas. (c) LC-MS/MS analysis is conducted for the same area as in OrbiSIMS, where the sample is collected using a scraper and introduced after lipid extraction.

Scheme 1.

(a) ToF-SIMS measurements are performed on the entire tissue and reconstructed area (ROI), which is the same area as in the OrbiSIMS analysis. (b) OrbiSIMS measurements are carried out on specific areas. (c) LC-MS/MS analysis is conducted for the same area as in OrbiSIMS, where the sample is collected using a scraper and introduced after lipid extraction.

Close modal

We obtained mass spectra and images from the entire coronal region of the mouse brain tissue within static limits using a pulsed argon cluster ion beam. Figure 1 shows the positive and negative spectra obtained from ToF-SIMS with pulsed argon cluster ions from the mouse brain tissue on the stainless steel substrate. The peaks measured in the positive ion spectrum are expected to be cholesterol, triacylglycerols, glycerophospholipids, and glycerolipids. On the other hand, the peaks measured in the negative ion spectrum are expected to be fatty acids, sterol lipids, and sphingolipids. These lipids are biomaterials that can be measured in mammalian tissue, including mouse brain sections.2,14,15

FIG. 1.

Mass spectra of (a) positive and (b) negative ions from ToF-SIMS for normal mouse brain tissue using a pulsed argon cluster ion beam. The types of lipids commonly found in mouse brain tissue samples, namely, cholesterol, fatty acids, sterol lipids, triacylglycerols, glycerophospholipids, glycerolipids, and sphingolipids, are labeled in the positive and negative ion spectra.

FIG. 1.

Mass spectra of (a) positive and (b) negative ions from ToF-SIMS for normal mouse brain tissue using a pulsed argon cluster ion beam. The types of lipids commonly found in mouse brain tissue samples, namely, cholesterol, fatty acids, sterol lipids, triacylglycerols, glycerophospholipids, glycerolipids, and sphingolipids, are labeled in the positive and negative ion spectra.

Close modal

The high intensity peaks from the spectra in Fig. 1 were reconstructed into images that were then grouped by similarity, as shown in Fig. 2. Figure 2(a) labels the various regions in a total ion image of the coronal section of the mouse brain tissue. In the positive ion images in Fig. 2(b), it can be confirmed that m/z 369.39 and 386.36 expected to be cholesterol peaks are clearly apparent in the corpus callosum and anterior commissure areas.16,17 This is thought to be because these areas contain a high density of nerve fibers, the surfaces of which are coated with various lipids, including cholesterol, to efficiently transmit electrical signals to the nerve fiber bundles. These results are in good agreement with those of a previous study conducted with bismuth cluster ions.17  Figure 2(c) shows reversed images from those in Fig. 2(b). The images appear similar but slightly differ in detail. The m/z 184.11 ion is well known as the phosphocholine headgroup of phosphatidylcholine (PC), and since PC is a component of the phospholipid double membrane on the surface of cells, it is expected to appear in regions containing neuron cells.2 In the images, m/z 753.61, 769.61, and 1463.12 show a high intensity, especially in the septum area, while m/z 478.39, 734.58, 1463.12, 1468.12, 1491.11, and 1494.11 present a high intensity in the piriform area. In Fig. 2(d), the images appear bright over the entire tissue area including the corpus callosum, anterior commissure, cortex, and caudoputamen. Similar to the positive ion image classification, images can also be grouped by similar patterns using negative ion images. In particular, m/z 722.55 in Fig. 2(h) shows a high intensity in the region presumed to be the substantia innominate, and the images in Fig. 2(i) show a significant intensity in the cortex and hypothalamus areas. Based on the observed high intensity areas in similar images, MS/MS analysis was conducted using OrbiSIMS to determine whether images that appear similar contain the same biomolecules or not.

FIG. 2.

(a) Regions of the mouse brain highlighted in a total ion image. Using a pulsed argon cluster ion beam within static limits, (b)–(d) positive ion images and (e)–(i) negative ion images were classified by similarity.

FIG. 2.

(a) Regions of the mouse brain highlighted in a total ion image. Using a pulsed argon cluster ion beam within static limits, (b)–(d) positive ion images and (e)–(i) negative ion images were classified by similarity.

Close modal

We wanted to compare lipids in the outer layer of the cortex, gray matter made up of nerve cells, and in the corpus callosum, a bundle of nerve fibers, in the ToF-SIMS and OrbiSIMS results.2 OrbiSIMS obtains mass spectra in the dynamic mode from the use of long pulsed argon cluster ions similar to continuous ions. In Fig. 3, we compare the spectra obtained from ToF-SIMS within static limits with the spectra obtained from OrbiSIMS in the dynamic mode. Figures 3(a) and 3(b) plot the positive and negative spectra from the approaches measured in the outer layer of the cortex, respectively, and Figs. 3(c) and 3(d) plot the positive and negative spectra obtained from the corpus callosum, respectively. The mass spectra of ToF-SIMS shown in Fig. 3 were obtained using the ROI function, which selects only the outer layer of the cortex and corpus callosum areas from the full image. Lipid groups measured in ToF-SIMS, such as triacylglycerols, glycerophospholipids, glycerolipids, fatty acids, sterol lipids, and sphingolipids, were also measured in OrbiSIMS. The mass spectra of triacylglycerols, glycerophospholipids, and sphingolipids obtained in the two areas show different patterns, and these pattern differences were measured equally by both ToF-SIMS and OrbiSIMS. However, the spectrum obtained with OrbiSIMS showed a much better mass resolution. The mass resolution of PC (m/z 760.59) was approximately 4000 for ToF-SIMS and approximately 120 000 for OrbiSIMS (Orbitrap). The mass accuracy of PC (m/z 760.59) was approximately 8.94 ppm for ToF-SIMS and 0.26 ppm for OrbiSIMS. These results demonstrate the advantages of using OrbiSIMS to analyze biomolecules containing various lipids in the mouse brain.

FIG. 3.

(a) Positive and (b) negative ion mass spectra measured from the outer layer of the cortex using ToF-SIMS and OrbiSIMS. (c) Positive and (d) negative ion mass spectra measured from the corpus callosum using ToF-SIMS and OrbiSIMS. The mass spectra obtained from the outer layer of cortex and corpus callosum areas show differences in triacylglycerols, glycerophospholipids, and sphingolipids.

FIG. 3.

(a) Positive and (b) negative ion mass spectra measured from the outer layer of the cortex using ToF-SIMS and OrbiSIMS. (c) Positive and (d) negative ion mass spectra measured from the corpus callosum using ToF-SIMS and OrbiSIMS. The mass spectra obtained from the outer layer of cortex and corpus callosum areas show differences in triacylglycerols, glycerophospholipids, and sphingolipids.

Close modal

Next, we compared the MS/MS spectra of biomolecules in the mouse brain tissue sputtered by argon cluster ions with the MS/MS spectra measured via LC-MS/MS. Both methods adopted fragmentation using CID (chemical induced dissociation) and used an orbitrap as a mass analyzer. Figure 4 shows the full mass spectrum of positive ions acquired using OrbiSIMS and the MS/MS spectra obtained using OrbiSIMS and LC-MS/MS. MS/MS was performed for positive m/z 734.57 and 760.59 ions among the many peaks measured in the mass spectrum. In the MS/MS spectra of m/z 734.57 and 760.59 measured using OrbiSIMS, in addition to the precursor molecule, the m/z 184.08 peak, which is well known as a phosphocholine (C5H15NPO4+) peak2, was measured at a collision energy (CE) of 15 eV [Figs. 4(b) and 4(d)]. Therefore, the two positive ion peaks are expected to be PC. The MS/MS spectra were obtained from LC-MS/MS by extracting the lipids from the mouse brain tissue. LC-MS/MS spectra show the precursor and phosphocholine peaks [Figs. 4(c) and 4(e)], matching the MS/MS results obtained using OrbiSIMS. Figure 5(a) presents the full mass spectrum of negative ions obtained using OrbiSIMS. MS/MS analysis in OrbiSIMS was performed on m/z 885.55 and 888.63 ions. For the negative m/z 885.55, two MS/MS spectra were obtained under the conditions of 20 and 40 eV CE, showing slight differences [Fig. 5(b)]. The MS/MS spectrum measured under the condition of 40 eV CE showed various fragment peaks including m/z 78.96 (PO3 −, phosphate), m/z 153.00 (C3H6O5P, glycerophosphate), m/z 214.01 (C6H10O8P, inositol head group), and m/z 303.23 (C20H31O2, arachidonic acid), but the precursor molecule peak was not measured. Under the 20 eV CE condition, the precursor molecule, m/z 419.26 (C21H40O6P), and m/z 581.31 (C27H50O11P) were measured, which were absent at CE 40 eV. It is expected that the precursor ion lost arachidonic acid in this condition, giving the m/z 581.31 peak, and also lost arachidonic acid and inositol, giving the m/z 419.26 peak. From these results, the precursor ion is expected to be glycerophosphoinositol (PI). Nuffel et al. obtained similar results from human lung tissue by measuring all the same peaks we obtain here at CE 20 and 40 eV in the MS/MS spectrum at CE 56.7 eV.18 At m/z 885.55, various peaks measured by OrbiSIMS were observed in the LC-MS/MS spectra, including peaks related to the precursor ion, glycerophosphate, inositol head group, and arachidonic acid [Fig. 5(c)]. In the OrbiSIMS MS/MS spectrum of the other negative ion m/z 888.63, the precursor ion, the sulfo-galactose-related peak m/z 241.00 (C6H9SO8 −), and m/z 96.96 (SO4H) were measured. Accordingly, this peak was estimated to be sulfatide [Fig. 5(d)].19 In the LC-MS/MS spectrum of m/z 888.63, the peaks related to sulfate and sulfo-galactose were measured [Fig. 5(e)], confirming the same results as those in OrbiSIMS MS/MS.

FIG. 4.

(a) Full mass spectrum of positive ions measured using OrbiSIMS. (b) OrbiSIMS MS/MS of the m/z 734.57 positive ion. (c) LC-MS/MS of the m/z 734.57 positive ion. (d) OrbiSIMS MS/MS of the m/z 760.59 positive ion. (e) LC-MS/MS of the m/z 760.59 positive ion. All spectra were measured from the same mouse brain tissue sample.

FIG. 4.

(a) Full mass spectrum of positive ions measured using OrbiSIMS. (b) OrbiSIMS MS/MS of the m/z 734.57 positive ion. (c) LC-MS/MS of the m/z 734.57 positive ion. (d) OrbiSIMS MS/MS of the m/z 760.59 positive ion. (e) LC-MS/MS of the m/z 760.59 positive ion. All spectra were measured from the same mouse brain tissue sample.

Close modal
FIG. 5.

(a) Full mass spectrum of negative ions measured using OrbiSIMS. (b) OrbiSIMS MS/MS of the m/z 885.55 negative ion. (c) LC-MS/MS of the m/z 885.55 negative ion. (d) OrbiSIMS MS/MS of the m/z 888.63 negative ion. (e) LC-MS/MS of the m/z 888.63 negative ion. All spectra were measured from the same mouse brain tissue sample.

FIG. 5.

(a) Full mass spectrum of negative ions measured using OrbiSIMS. (b) OrbiSIMS MS/MS of the m/z 885.55 negative ion. (c) LC-MS/MS of the m/z 885.55 negative ion. (d) OrbiSIMS MS/MS of the m/z 888.63 negative ion. (e) LC-MS/MS of the m/z 888.63 negative ion. All spectra were measured from the same mouse brain tissue sample.

Close modal

In this work, we performed mass imaging of the coronal section of mouse brain tissue using ToF-SIMS with pulsed argon cluster ions. Full-sized images of the mouse brain were classified according to similar reconstructed images with sufficient intensity. We compared the mass spectra of ToF-SIMS measured in the outer layer of the cortex and corpus callosum regions of the normal mouse brain tissue to those of OrbiSIMS measured in the same regions. We found that the ToF-SIMS mass spectrum measured within static limits and the OrbiSIMS mass spectrum measured in dynamic mode were not significantly different, meaning that the two modes using argon cluster ions give similar results for biological tissue.

We then tried to identify candidate molecules present in specific areas of the mouse brain tissue. With the inclusion of an orbitrap mass analyzer and HCD cell, hybrid SIMS (OrbiSIMS) with MS/MS functionality measured the mass spectra in the dynamic mode of the outer layer of the cortex and corpus callosum. As a result of performing MS/MS using OrbiSIMS, it was confirmed that two candidate molecules with m/z 734.57 and 760.59 positive ion peaks were both PC (phosphatidylcholine), which was consistent with the LC-MS/MS result. Negative ion candidates m/z 885.55 and 888.63 were identified as PI (glycerophosphoinositol) and sulfatide, respectively, which were also consistent with the LC-MS/MS results. The significance of this study is that the candidate molecules measured in specific areas were not compared with compound molecules for identification but were rather directly identified in the mouse brain tissue. There is a definite capability to identify lipid molecules with strong intensity distributed in the tissue, but it is still difficult to identify molecules with weak intensity. However, utilizing OrbiSIMS, which exhibits results similar to the LC-MS/MS analysis results used as the golden standard, can provide an alternative means for the identification of biomolecule candidates present at specific locations in tissue. Identifying biomolecules based on the mass imaging of biological tissues is expected to support disease-related research.

This work was supported by the Development of Measurement Standards and Technology for Biomaterials and Medical Convergence funded by the Korea Research Institute of Standards and Science (No. KRISS-2023-GP2023-0007), the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Korean Government (MSIT) (No. 2021RC1C1011598), and the Nano Material Technology Development Program (Nos. 2016M3A7B6908929 and 2018M3D1A1058814) of the National Research Foundation (NRF) funded by the Ministry of Science and ICT, Republic of Korea.

The authors have no conflicts to disclose.

A normal mouse (three-month-old male C57BL/6J) brain was obtained from a laboratory animal resource center, the Korea Research Institute of Bioscience and Biotechnology (KRIBB), and all experimental procedures were performed under policies of the Institutional Animal Care and Use Committee of KRIBB.

Hyun Kyong Shon: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Software (equal); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal). Jin Gyeong Son: Conceptualization (supporting); Formal analysis (supporting); Investigation (supporting); Writing – original draft (supporting). Sun Young Lee: Formal analysis (supporting); Methodology (supporting); Validation (supporting). Jeong Hee Moon: Formal analysis (equal); Methodology (supporting); Writing – original draft (supporting). Ga Seul Lee: Formal analysis (supporting); Visualization (supporting). Kyoung-Shim Kim: Formal analysis (supporting); Resources (supporting). Tae Geol Lee: Conceptualization (lead); Funding acquisition (lead); Project administration (lead).

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
J.
Xu
,
S.
Ostrowski
,
C.
Szakal
,
A. G.
Ewing
, and
N.
Winograd
,
Appl. Surf. Sci.
231–232
,
159
(
2004
).
2.
M. K.
Passarelli
and
N.
Winograd
,
Biochim. Biophys. Acta (BBA)
1811
,
976
(
2011
).
3.
H. K.
Shon
,
Y.-L.
Cho
,
C. S.
Lim
,
J. S.
Choi
,
S. J.
Chung
, and
T. G.
Lee
,
Surf. Interface Anal.
46
,
189
(
2014
).
4.
I.
Yamada
,
J.
Matsuo
,
N.
Toyoda
,
T.
Aoki
, and
T.
Seki
,
Curr. Opin. Solid State Mater. Sci.
19
,
12
(
2015
).
5.
A. S.
Mohammadi
,
N. T. N.
Phan
,
J. S.
Fletcher
, and
A. G.
Ewing
,
Anal. Bioanal. Chem.
408
,
6857
(
2016
).
6.
H. K.
Shon
,
S.
Yoon
,
J. H.
Moon
, and
T. G.
Lee
,
Biointerphases
11
,
02A321
(
2016
).
7.
A. N.
Lazar
et al,
Acta Neuropathol.
125
,
133
(
2013
).
8.
M. T.
Le
,
H. K.
Shon
,
H.
Nguyen
,
C.
Lee
,
K.
Kim
,
H.
Na
, and
T. G.
Lee
,
Anal. Chem.
94
,
9297
(
2022
).
9.
J. S.
Fletcher
,
S.
Rabbani
,
A.
Henderson
,
P.
Blenkinsopp
,
S. P.
Thompson
,
N. P.
Lockyer
, and
J. C.
Vickerman
,
Anal. Chem.
80
,
9058
(
2008
).
10.
J. S.
Fletcher
,
H. L.
Kotze
,
E. G.
Armitage
,
N. P.
Lockyer
, and
J. C.
Vickerman
,
Metabolomics
9
,
535
(
2013
).
11.
N. T. N.
Phan
,
M.
Munem
,
A. G.
Ewing
, and
J. S.
Flecture
,
Anal. Bioanal. Chem.
409
,
3923
(
2017
).
12.
G. L.
Fisher
,
A. L.
Bruinen
,
N. O.
Potočnik
,
J. S.
Hammond
,
S. R.
Bryan
,
P. E.
Larson
, and
R. M. A.
Heeren
,
Anal. Chem.
88
,
6433
(
2016
).
13.
M. K.
Passarelli
et al,
Nat. Methods
14
,
1175
(
2017
).
14.
D.
Debois
,
M.-P.
Bralet
,
F. L.
Naour
,
A.
Brunelle
, and
O.
Laprévoe
,
Anal. Chem.
81
,
2823
(
2009
).
15.
Y.
Hirahara
et al,
J. Neurochem.
140
,
435
(
2017
).
16.
P.
Sjövall
,
J.
Lausmaa
, and
B.
Johansson
,
Anal. Chem.
76
,
4271
(
2004
).
17.
D.
Touboul
,
F.
Kollmer
,
E.
Niehuis
,
A.
Brunelle
, and
O.
Laprévoe
,
J. Am. Soc. Mass Spectrom.
16
,
1608
(
2005
).
19.
F.-F.
Hsu
and
J.
Turk
,
Am. Soc. Mass Spectrom.
15
,
536
(
2004
).