Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2022 Mar 31;33(5):760–771. doi: 10.1021/jasms.1c00327

Evaluation of 6 MALDI-Matrices for 10 μm Lipid Imaging and On-Tissue MSn with AP-MALDI-Orbitrap

Tina B Angerer †,*, Jerome Bour †, Jean-Luc Biagi †, Eugene Moskovets ‡, Gilles Frache †,*
PMCID: PMC9074099  PMID: 35358390

Abstract

graphic file with name js1c00327_0006.jpg

Mass spectrometry imaging is a technique uniquely suited to localize and identify lipids in a tissue sample. Using an atmospheric pressure (AP-) matrix-assisted laser desorption ionization (MALDI) source coupled to an Orbitrap Elite, numerous lipid locations and structures can be determined in high mass resolution spectra and at cellular spatial resolution, but careful sample preparation is necessary. We tested 11 protocols on serial brain sections for the commonly used MALDI matrices CHCA, norharmane, DHB, DHAP, THAP, and DAN in combination with tissue washing and matrix additives to determine the lipid coverage, signal intensity, and spatial resolution achievable with AP-MALDI. In positive-ion mode, the most lipids could be detected with CHCA and THAP, while THAP and DAN without additional treatment offered the best signal intensities. In negative-ion mode, DAN showed the best lipid coverage and DHAP performed superiorly for gangliosides. DHB produced intense cholesterol signals in the white matter. One hundred fifty-five lipids were assigned in positive-ion mode (THAP) and 137 in negative-ion mode (DAN), and 76 peaks were identified using on-tissue tandem-MS. The spatial resolution achievable with DAN was 10 μm, confirmed with on tissue line-scans. This enabled the association of lipid species to single neurons in AP-MALDI images. The results show that the performance of AP-MALDI is comparable to vacuum MALDI techniques for lipid imaging.

Keywords: AP-MALDI, mass spectrometry imaging, lipids, tandem-MS, sample preparation

Introduction

Mass spectrometry imaging (MSI) is a technique capable of locating and identifying atoms and molecules in a sample. By scanning across a surface and recording individual mass spectra at each location, MSI generates “chemical maps” which display the distribution of all detected species on a sample surface. MSI was first performed in 1949,1 but it has only gained traction in the last 20 years. With an ever-growing number of techniques emerging, MSI is now regularly applied in a variety of fields, ranging from inorganic materials science to biomedical research. Especially for biologists, MSI is of great interest because it is now possible to analyze intact macromolecules (e.g., proteins, lipids, and neurotransmitter)2 at cellular resolutions3 while small molecules and atoms can be detected in cell organelles.4 Multimodal experiments connect proteomic and lipidomic data and deepen our understanding of biological processes.5−7 The capabilities, applications, and drawbacks of MSI techniques are described in several reviews.8−12

Lipidomics is of ever-growing importance in the medical field.13 The changes in lipid compositions due to the onset of disease have been recognized, but traditional techniques, such as liquid chromatography mass spectrometry (LC–MS), fail to capture the complexity of heterogeneous samples (e.g., tumors). In contrast to proteins, single lipids cannot be labeled easily. MSI is uniquely suited to localize individual lipid species in a sample, as it distinguishes lipids based on their accurate mass. This improves upon unspecific dyes and techniques requiring sample homogenization (e.g., LC) but is insufficient to determine the exact species, since a number of different lipids can be present within a narrow mass range and lipids can have isomers, even from different classes. For this reason, an increasing number of MSI devices offer high-resolution mass analyzers and tandem-MS capability for structural identification to enhance specificity.14 With these capabilities, MSI has significantly contributed to the field of lipidomics by revealing lipid alterations in various diseases.15−18

Matrix-assisted laser desorption ionization (MALDI)-MSI is a technique where a matrix is applied to a sample surface and molecules of interest are extracted, embedded in matrix crystals, desorbed with a laser beam, ionized, and finally detected with a mass analyzer. The detectable species largely depend on the applied matrix, of which there are numerous options. In general, a MALDI matrix must be able to absorb the laser light and transfer charges to/from the target molecules, usually be vacuum stable, and in the case of MSI, form (sub)micron-sized crystals to enable highly localized detection. The achievable spatial resolution depends on the target molecules, the matrix properties, and the laser beam focus. For lipids, 5–10 μm has been demonstrated, and proteins are usually imaged at 50–100 μm (better spatial resolutions for lipids and proteins have been demonstrated with, e.g., t-MALDI-2).15,19−22

MALDI performed under ambient conditions (atmospheric pressure, AP-MALDI) removes the requirement of a vacuum stable matrix and enables the use of more volatile substances, such as 2′,6′-dihydroxyacetophenone (DHAP) and 2,4,6-trihydroxyacetophenone (THAP). Woods et al. state that the addition of heptafluorobutyric acid (HFBA) stabilizes DHAP in vacuo, but recent reports show its sublimation, even with added HFBA.23,24 An added benefit of AP-MALDI is that the tissue sample is not subjected to the harsh vacuum conditions leading to drying and cracking, which makes subsequent procedures such as histological tissue staining more likely to succeed.25−29 The downside is the shortened mean free path for the generated ions, possibly leading to their neutralization and diminishing sensitivity. This can reduce the useful spatial resolution and the ability to perform tandem-MS. Previous reports investigated AP-MALDI capabilities30 and have compared the performance of several matrices for AP-MALDI imaging,24,31−35 but to our knowledge, there is currently no comprehensive report comparing the numerous matrices available and their lipid coverage on the same sample type and device in positive- and negative-ion mode.

Therefore, in this study we tested 11 sample preparation protocols with six different matrices, washing steps, and additives that have previously been reported to yield good results for high spatial resolution lipid imaging with vacuum and AP-MALDI techniques. Protocols were adapted for the matrices: α-cyano-4-hydroxycinnamic acid (CHCA),36,37 norharmane (Nor),36,38 1,5-diaminonapthalene (DAN),24,39 DHAP,23,24 THAP,40 and 2′,5′-dihydroxybenzoic acid (DHB),38,41 and their performance was evaluated, in terms of signal intensity/ability to perform tandem-MS, lipid coverage and achievable, useful spatial resolution, in positive and negative ion mode. Additionally, 76 peaks were identified using on tissue tandem-MS.

Materials and Methods

Chemicals

Chemicals and solvents (analytical grade) were purchased from the following sources: α-cyano-4-hydroxycinnamic acid 98% (CHCA) (Sigma-Aldrich), 2,5-dihydroxybenzoic acid 98% (DHB) (Sigma-Aldrich), norharmane 98% (Nor) (Acros Organics), 1,5-diaminonaphthalene 97% (DAN) (Sigma-Aldrich), 2′,4′,6′-trihydroxyacetophenone 99.5% (THAP) (Sigma-Aldrich), 2,6-dihydroxyacetophenone 99.5% (DHAP) (Sigma-Aldrich), acetonitrile (ACN) (Honeywell), chloroform (Acros Organics), methanol (Carl Roth), ammonium acetate (AmAc) (VWR). ammonium sulfate (AmS) (Sigma-Aldrich), heptafluorobutyric acid (HFBA) (Sigma-Aldrich), and trifluoroacetic acid (TFA) (Sigma-Aldrich). All chemicals used in this study were stored, handled, and disposed of according to good laboratory practices (GLP).

Sample Preparation

Ten micrometer sagittal mouse brain sections on indium tin oxide (ITO)-coated glass slides (Diamond Coatings, UK) were prepared at Swansea University, as stated in a recent publication.43 Sections were kept at −80 °C until analysis and dried in a vacuum desiccator for 30 min prior to matrix application. Optical images of the tissue sections were taken using an Olympus BX51 microscope (Olympus, Belgium). On some sections, tissue washing was performed with AmAc at 50 mM concentration, chilled to 4 °C, for 3 × 5 s, as described previously.42 Serial sections on separate ITO glass slides were coated with the various matrices using an HTX TM sprayer (HTX Technologies LLC, USA), flow rate 0.12 mL/min, velocity 1200 mm/min, drying time 2 s, line spacing 2.5 mm. Matrix composition and additives, sprayer temperature, number of matrix layers/passes, and laser settings are listed in Table 1. The analyses showing the best performance (most lipids detected, best signal) were repeated on different days in positive-ion mode for CHCA, THAP, and DAN70 and in negative-ion mode for DAN70.

Table 1. Overview of Matrix Recipes, Sprayer/Laser Settings, and Experiments Performeda.

matrix (mg/mL) solvents add/ sample prep temp (°C) Z laser (40/20 μm) laser (10 μm) Hip. Crb. Str. adapted from
CHCA (5) CHCl3:MeOH 1:1 0.2%TFA 40 16 3000 Hz 10% 3000 Hz 2.5% × (+) × (±) × (+) Barré et al.36
Hochart et al.37
Nor (7) CHCl3:MeOH 2:1   30 12 500 Hz 10%     × (±) × (+) Barré et al.36
DHB (10) MeOH:H2O 7:3 0.1%TFA 50 8 3000 Hz 10%     × (+) × (+) McMillen et al.38
Leopold et al.41
DHAP (10) ACN:AmSb 7:3 0.05% HFBA 50 8 3000 Hz 7.5%   × (−)c × (−) × (+) Jackson et al.24
Colsch et al.23
THAP (10) ACN:AmSb 7:3 0.05% HFBA 50 8 3000 Hz 15% 3000 Hz 10% × (+) × (−)c × (±) × (+) Pieles et al.40
THAPnS (10) ACN:H2O 7:3 0.05%HFBA 50 8 3000 Hz 15%     × (+)   Pieles et al.40
DAN70 (10) ACN:H2O 7:3   30 8 3000 Hz 5% 2000 Hz 1.5% × (±) × (±) × (+) Sun et al.39
DAN90 (10) ACN:H2O 9:1   30 8 3000 Hz 5%     × (+)   Jackson et al.24
DANtfa (10) ACN:H2O 7:3 0.1%TFA 30 8 3000 Hz 5%     × (+)   Sun et al.39
DANhfba (10) ACN:H2O 7:3 0.05%HFBA 30 8 3000 Hz 5%     × (+)   Sun et al.39
Colsch et al.23
DANw (10) ACN:H2O 9:1 AmAc wash 30 8 3000 Hz 5% 2000 Hz 1.5% × (+) × (±)   Sun et al.39
Angel et al.42
Open in a new tab
a

Listed are matrices + matrix concentration (mg/mL), matrix solvents, matrix additives or sample preparation, HTX-TM sprayer temperature (°C), matrix layers/passes (Z), laser frequency (Hz), and intensity (%) for 40/20 and adjusted settings for 10 μm imaging, ion-mode (±) included for each analyzed brain area (Hip. = hippocampus, spatial resolution 10/20c μm; Crb. = cerebellum, 40 μm; Str. = striatum, 40 μm), and matrix recipe references (adapted from).

b

125 mM (NH4)2SO4 in H2O.

c

Hippocampus imaging with THAP/DHAP in negative ion mode performed at 20 μm spatial resolution.

AP-MALDI-MSI

MALDI analysis on brain sections was performed using an AP-MALDI UHR ion source (Masstech, Inc., USA), which has been described in detail elsewhere,25,26 coupled to an LTQ/Orbitrap Elite high-resolution mass spectrometer (Thermo-Fisher Scientific, USA) in positive- and negative-ion mode. For imaging, the AP-MALDI source was operated in “Constant Speed Raster” motion mode. Whole, sagittal brain sections were analyzed in positive and negative ion mode at 40 μm stepping size (matrix: Norhamane), as well as the cerebellum (Crb., 40 μm), the hippocampus (Hip., 10/20 μm), and a small area in the striatum (Str., 40 μm). Matrices, ion modes, and laser settings used for each analysis are summarized in Table 1. The laser spot size was <10 μm for the hippocampus analysis and <40 μm for all other analyses (20 μm max.). A laser focus of 8.38 μm was determined with scanning electron microscopy (SEM, Figure S3), and using the camera in the source and comparing line to line signal intensities, settings were adjusted for each measurement to ablate as much matrix as possible without oversampling. Spectrum acquisition: 800 ms maximum injection time; mass range: 500–2000 Da (250–1000 Da for striatum (cholesterol) imaging); mass resolution: 120k at m/z 400. Tandem-MS was performed on 76 peaks (Table 2): 1 Da isolation window, and collision-induced dissociation/higher-energy collision dissociation (CID/HCD) was performed with collision energies of 20–55%, adjusted for each lipid species individually. Tandem-MS scans were summed up over 30–120 s. Details on matrix used, collision mode, and energy can be found in the scan header of each analysis in Supporting Information 2. Data analysis and visualization were performed with Thermo Xcalibur 2.2 and Thermo ImageQuest (Thermo-Fisher Scientific, USA), METASPACE,44 MSiReader 1.2 (NC State University, USA),45 LipostarMSI (Molecular Horizons Srl, Italy),46 and OriginPro 2019b (OriginLab Corp., USA). All images are normalized to total ion count (TIC). Lipid identification was performed with METASPACE for hippocampus images (database: LIPIDMAPS, FDR: 20%) and LipostarMSI for cerebellum images (database: LIPIDMAPS, mass accuracy: 2 ppm; mass and isotopic pattern score: 80%+).

Table 2. Lipids Identified with Tandem MS in Positive- and Negative-Ion Modea,b.
Positive-Ion Mode
assignment ion species formula m/z (detected) m/z (exact) Δppm
PC(16:0/9:0(OH)) [M + H]+ C33H65O9NP 650.4372 650.4392 3.00
PC(16:0/9:0(COOH)) [M + H]+ C33H65O10NP 666.4321 666.4341 2.94
PC(18:0/9:0(OH)) [M + H]+ C35H69O9NP 678.4698 678.4705 0.96
PC(16:0/16:0)-TMA [M + K]+ C37H71O8PK 713.4497 713.4518 2.95
PDME(16:0/16:0) [M + H]+ C39H79O8NP 720.5543 720.5538 –0.72
SM(d36:1) [M + H]+ C41H84O6N2P 731.606 731.6062 0.21
PC(16:0/16:0) [M + H]+ C40H81O8NP 734.5689 734.5694 0.72
PC(16:0/16:0) [M + Na]+ C40H80O8NPNa 756.553 756.5514 –2.14
PC(16:0/18:1) [M + H]+ C42H83O8NP 760.5826 760.5851 3.26
PC(16:0/16:0) [M + K]+ C40H80O8NPK 772.5254 772.5253 –0.12
PC(16:0/18:1) [M + Na]+ C42H82O8NPNa 782.5685 782.5670 –1.88
HexCer(d18:1/22:1) [M + H]+ C46H88O8N 782.6489 782.6505 1.98
SM(d40:2) [M + H]+ C45H90O6N2P 785.6534 785.6531 –0.38
PC(18:1/18:1) [M + H]+ C44H85O8NP 786.6004 786.6007 0.42
PC(18:1/18:0) [M + H]+ C44H87O8NP 788.6161 788.6164 0.36
PE(40:7)* [M + H]+ C45H77O8NP 790.5357 790.5381 3.07
PE(18:0/22:6) [M + H]+ C45H79O8NP 792.5555 792.5538 –2.17
PC(16:0/18:1) [M + K]+ C42H82O8NPK 798.5403 798.5410 0.83
PC(16:0/22:6) [M + H]+ C46H81O8NP 806.5691 806.5694 0.41
PE(38:4) [M + K]+ C43H78NO8PK 806.5067 806.5097 3.67
PC(18:1/20:4) [M + H]+ C46H83O8NP 808.5839 808.5851 1.45
HexCer(d18:1/24:2) [M + H]+ C48H90O8N 808.6642 808.6661 2.34
SM(d42:2) [M + H]+ C47H94O6N2P 813.6846 813.6844 –0.25
PC(18:1/18:0)* [M + K]+ C44H86O8NPK 826.5739 826.5723 –1.98
PC(18:0/22:6) [M + H]+ C48H85O8NP 834.6012 834.6007 –0.56
PC(38:6) [M + K]+ C46H80NO8PK 844.5228 844.5253 2.97
PC(38:4) [M + K]+ C46H84O8NPK 848.5556 848.5566 1.19
HexCer(d18:1/24:1(2OH)) [M + Na]+ C48H91O9NNa 848.6564 848.6586 2.59
PC(40:6) [M + K]+ C48H85O8NPK 872.5565 872.5566 0.13
PC(16:0/18:1)+PC(16:0/16:0) [M+M+H]+ C82H163O16N2P2 1494.1488 1494.1470 –1.04
PC(16:0/18:1) [2M+H]+ C84H165O16N2P2 1520.1604 1520.1630 1.64
PC(18:1/18:0)+PC(16:0/18:1) [M+M+H]+ C86H169O16N2P2 1548.1868 1548.1940 4.64
Negative-Ion Mode
assignment ion species formula m/z(measured) m/z(exact) ppm
Cer(d18:0/18:1) [M – H]− C36H70O3N 564.535 564.5350 0.04
CerP(d18:1/18:0) [M – H]− C36H71NO6P 644.5019 644.5025 0.85
PA(16:0/16:1) [M – H]− C35H66O8P 645.4503 645.4501 –0.34
PA(16:0/16:0) [M – H]− C35H68O8P 647.4645 647.4657 1.9
CerP(d38:2) [M – H]− C38H73O6NP 670.5173 670.5181 1.19
PA(16:0/18:1) [M – H]− C37H70O8P 673.4795 673.4814 2.79
SM(d34:1) [M – CH3]− C38H76O6N2P 687.5426 687.5447 2.98
SM(d18:1/18:0) [M – CH3]− C40H80O6N2P 715.575 715.5760 1.33
PE(16:0/18:0) [M – H]− C39H77O8NP 718.5377 718.5392 2.13
PE(P-18:1/18:1) [M – H]− C41H77O7NP 726.5427 726.5443 2.22
PE(P-18:0/18:1) [M – H]− C41H79O7NP 728.5584 728.5600 2.14
PE(18:1/18:1) [M – H]− C41H77O8NP 742.5391 742.5392 0.17
PC(16:0/18:1) [M – CH3]− C41H79O8NP 744.5535 744.5549 1.85
PA(18:0/22:6) [M – H]− C43H72O8P 747.4959 747.4970 1.51
PE(P-18:1/20:1) [M – H]− C43H81O7NP 754.5749 754.5756 0.94
PE(18:0/20:4) [M – H]− C43H77O8NP 766.5373 766.5392 2.52
SM(d18:1/22:0) [M – CH3]− C44H88O6N2P 771.637 771.6386 2.01
PE(P-18:0/22:6) [M – H]− C45H77O7NP 774.5427 774.5443 2.08
PE(18:0/20:4(OH) [M – H]− C43H77O9NP 782.5347 782.5341 –0.72
PC(16:1/22:6) [M – CH3]− C45H75O8NP 788.5244 788.5236 –1.04
PS(18:1/18:0) [M – H]− C42H79O10NP 788.5447 788.5447 0.01
PE(18:0/22:6) [M – H]− C45H77O8NP 790.5374 790.5392 2.31
HexCer(d18:1/22:0(2OH)) [M – H]− C46H88O9N- 798.6443 798.6465 2.70
C18:1-Sulf [M – H]− C42H80O11NS 806.5436 806.5458 2.68
C18(OH)-Sulf [M – H]− C42H80O12NS 822.5411 822.5407 –0.52
HexCer(d18:1/24:0(2OH)) [M – H]− C48H92O9N 826.6757 826.6778 2.49
PS(18:0/22:6) [M – H]− C46H77O10NP 834.5274 834.5291 1.99
PS(18:1/22:0) [M – H]− C46H87O10NP 844.6081 844.6073 –0.93
PI(16:0/20:4) [M – H]− C45H78O13P 857.518 857.5186 0.64
C22-Sulf. [M – H]− C46H88O11NS 862.6062 862.6084 2.50
PI(18:0/18:1) [M – H]− C45H84O13P 863.5636 863.5655 2.20
PS(18:1/24:0) [M – H]− C48H91O10NP 872.6341 872.6386 5.12
C22(OH)-Sulf [M – H]− C46H88O12NS 878.6008 878.6033 2.81
PI(18:1/20:4) [M – H]− C47H80O13P 883.5321 883.5342 2.38
PI(18:0/20:4) [M – H]− C47H82O13P 885.5493 885.5499 0.62
C24:1-Sulf. [M – H]− C48H90O11NS 888.6228 888.6240 1.36
C24-Sulf. [M – H]− C48H92O11NS 890.6376 890.6397 2.31
PI(18:0/20:4(OH)) [M – H]− C47H82O14P 901.5425 901.5448 2.52
C24(OH)-Sulf. [M − H]− C48H92O12NS 906.6333 906.6346 1.40
GA1(d36:1) [M – H]− C6113CH113O23N2 1254.777 1254.7770 0.26
GM1(d36:1) [M – H]− C73H130O31N3 1544.8632 1544.8690 4.00
GM1(d38:1) [M – H]− C7413C H134O31N3 1573.899 1573.9040 3.19
GD1(d36:1) [M – H]− C84H147O39N4 1835.957 1835.9648 4.24
GD1(d36:1) [M – 2H + K]− C84H146O39N4K 1873.916 1873.9210 2.71
Open in a new tab
a

Listed are assigned lipid, ion species, chemical formular, m/z value from tandem-MS spectra (detected) and calculated (exact) from the formula, and mass deviation (Δppm). Abbreviations for all lipid species listed at the bottom of the table. Tandem-MS spectra for each lipid shown in Supporting Information 2. Spatial distribution in MS-images can be viewed in METASPACE.

b

Phosphatidylcholine (PC), phosphatidyl-dimethylethanolamine (PDME), sphingomyelin (SM), hexosylceramide (HexCer), ceramide (Cer), ceramide 1-phosphate (CerP), phosphatidylethanolamine (PE), phosphatidylserine (PS); phosphatidic acid (PA), sulfatide (Sulf), phosphatidylinositol (PI), asialo-/monosialo-/disialo-ganglioside (GA/GM/GD); “P-“ plasmanyl-ether lipid; “d” 1,3-dihydroxy, long-chain base in sphingolipids.

ToF-SIMS-MSI

ToF-SIMS analysis was performed using an TOF.SIMS 5 (IONTOF GmbH, Germany) with a 25 kV Bi3+ primary analysis beam. Dried brain sections with DAN70 matrix were analyzed in burst alignment, delayed extraction, positive-ion mode with a total primary ion dose of 7 × 1011 ions/cm2, cycle time 105 us, random raster mode, 1 frame/patch, 1 shot/frame/pixel, 25 scans, mass range: 1–1000 Da, mass resolution: 5000 at m/z 300, image size: 256 × 256 μm and 512 × 512 pixels, spatial resolution: 0.5 μm/pixel. Data analysis and visualization was performed using SurfaceLab 7 (IONTOF GmbH, Germany).

SEM

SEM analysis was performed using a Quanta 200 field emission gun scanning electron microscope (FEG SEM, Philips-FEI, USA). DAN70 matrix analysis on brain was performed in a low-vacuum environment (60 Pa) with a “Large Field Detector” (LFD) for a topographical image. CHCA matrix analysis was performed with a Genesis XM 4i energy dispersive spectrometer (from EDAX) system for elemental mapping and line-scans, with a back-scattered, composition mode detector called BSED (for high vacuum), generating a grayscale, chemical composition SEM image, with “heavier” elements areas corresponding to brighter areas.

Results

The aim of this study was to evaluate the performance of various matrices/sample preparation protocols for AP-MALDI in terms of lipid coverage, signal intensity, and high-resolution imaging capability. For this purpose, several images were taken on sagittal mouse brain sections. Brain sections are often used in comparison studies, as they are rich in a great variety of lipids and the results can be compared to previously published material and entries in databases like, e.g., METASPACE. Figure 1 shows an overview of the data sets included in this study (summarized in Table 1): Full brain sections (Figure 1a, spatial resolution: 40 μm) were imaged to obtain a detailed overview of the lipid distributions in the brain and to determine the ideal location to perform tandem-MS on specific lipids. The location of six lipid species and representative spectra in positive- and negative-ion mode are shown in Figure 1g. Evidently, for species like SM(d16:1/24:1) is it important to know the location prior to on-tissue tandem-MS analysis as it is only present in the ventricles (Figure 1g). The full brain data sets are available in METASPACE: AP_MALDI_Full_Brain. The ability to detect cholesterol was tested on a small area in the striatum, around the fiber tracts (Figure 1c), with a resolution of 40 μm, in positive-ion mode, and a mass range of 250–1000 Da. To test instrument and matrix performance for higher spatial resolution imaging, the hippocampus with its intricate structures was imaged at 10 μm (Figure 1d, Figure S2) in positive and negative ion mode (for ganglioside detection, DHAP and THAP analysis was performed at 20 μm to shorten analysis time). Hippocampus data sets in METASPACE: AP_MALDI_Hippocampus. To determine lipid coverage (Figure 2), the cerebellum was imaged in positive ion mode (Figure 1e, Figure S1) and negative ion mode (Figure 1f, Figure S1) at 40 μm spatial resolution. DAN matrix performed well in positive and negative ion mode and attempts were made to improve its performance further, including tissue washing (DANw) and acidic additives (DANtfa, DANhfba). Initially HFBA was reported to stabilize DHAP in vacuum which would make it unnecessary in atmospheric conditions.23 Since HFBA is also a strong acid, that similar to TFA is used as ion-pairing agent in liquid chromatography, we investigated its protonating and potentially ion enhancing effects.47 The biggest improvement to DAN performance for lipid detection was achieved by changing the solvent ratios of acetonitrile and water (ACN:H2O) from 9:1 (DAN90, 90% ACN) to 7:3 (DAN70, 70% ACN) without decreasing the achievable spatial resolution (demonstrated in Figure 3 and Figure 4). Cerebellum data sets in METASPACE: positive ion mode AP_MALDI_MATRIX_pos; negative ion mode AP_MALDI_MATRIX_neg. Images for all cerebellum and hippocampus analyses are displayed in the Supporting Information 1 (Figure S1 and S2 respectively). Brain tissue not imaged was used for on-tissue tandem-MS.

Figure 1.

Figure 1

Open in a new tab

AP-MALDI imaging on sagittal brain section. (a) Full brain AP-MALDI-image, 40 μm spatial resolution, (red: PS(40:6) green: PI(38:5), blue: C24:1-Sulf), (b) H&E stained sagittal mouse brain section (Allen Developing Mouse Brain Atlas, data set P56, sagittal), (c–f) small area AP-MALDI-images of (c) fiber tracts/striatum images for cholesterol analysis (40 μm, red: PC(32:0) green: cholesterol), (d) hippocampus imaged at 10 μm, (red: PS(40:6) green: PI(38:5), blue: C24:1-Sulf), (e) cerebellum in positive ion mode (red: SM(d36:1), green: HexCer(d40:2), blue: PC(34:1)), and (f) negative-ion mode (40 μm, red: SM(d36:1), green: C24:1-Sulf, blue: PI(38:4)). Matrix and ion mode stated in each image, scalebar: 500 μm. (g) Six representative lipid images and spectra in positive-ion and (mirroring) in negative-ion mode.

Figure 2.

Figure 2

Open in a new tab

Lipid coverage for various matrices with AP-MALDI. (a) Number of identified lipids by class, detected in the cerebellum using five matrices (10 sample preparation protocols) in positive ion mode and (b) five matrices (six sample preparation protocols) in negative-ion mode listed in e). Displayed in the graph is the relative number of lipids normalized to the highest number of detected lipids for each class; the actual number is stated above CHCA (pos) and DAN70 (neg). Scores plots (PC1 vs PC2) of PCA analysis for cerebellum mass spectra in (c) positive- and (d) negative-ion mode (loadings plots: Figure S7). (e) Sum of assigned lipids for each matrix in positive- and negative-ion mode. Σ+ (x) DHAP was not analyzed in positive ion mode, Σ– (x) all matrices were analyzed in negative-ion mode but excluded if insufficient signal was detected (below 1 × 103 counts). Lipids were assigned using LipostarMSI.

Figure 3.

Figure 3

Open in a new tab

Capabilities of the AP-MALDI-Orbitrap system demonstrated on the hippocampus, (a) MS1 scan on brain tissue (matrix: DAN70, negative ion mode, showing m/z 788.5447 and m/z 788.5236. (b) Distribution of PS(18:1/18:0) (green) and PC(16:1/22:6)–CH3 (red) in a 10 μm AP-MALDI image, white arrows show the positions of linescans in (c) scale bar: 500 μm. (c) Two linescans, normalized to their individual maximum intensity (100%). (d) On-tissue tandem-MS analysis of m/z 788.5 ± 0.5 Da, containing PS (green) and PC (red) fragments. Fragmentation mechanism shown for both lipids, referring to the letters above each mass peak.

Figure 4.

Figure 4

Open in a new tab

Single-cell imaging in the hippocampus with AP-MALDI. (a) 10 μm Hippocampus analysis (matrix: DAN70, positive-ion mode), RGB overlay of: SM(d36:1) red, gray matter; PC(18:0/22:6) green, single cells; PC(18:1/20:4) blue, pyramidal layer. (b) Microscope image of the hippocampus pre matrix application, features corresponding to the distribution of PC(18:0/22:6) highlighted in green. (c) Single-ion image of PC(18:0/22:6) with the same features highlighted as in b), color scale: hot. (d) Microscope image of the cerebellum pre matrix application. (e) Single-ion image of PC(18:0/22:6) in the cerebellum, spatial resolution: 40 μm, color scale: hot. All scale bars: 500 μm.

Lipid coverage with various matrices

Figure 2 shows the number of lipid species detected in the cerebellum (Figure 1d/e) using all 11 sample preparation protocols for positive (Figure 2a) and negative (Figure 2b) ion mode, grouped into lipid classes (data was excluded if the most intense signals were below 1 × 103 counts). Spectra for all included data sets in positive (Figure S4) and negative ion mode (Figure S5) and signal-to-noise ratios for several signals (Figure S6) can be found in the Supporting Information 1. Using LipostarMSI and the LIPIDMAPS database, lipids were putatively assigned (Figure 2a/e), with strict selection criteria (mass accuracy: 2 ppm; mass and isotopic pattern score: 80%+). The most species detected using one matrix were 155 in positive (THAP), 136 in negative ion mode (DAN70), and 224 for positive and negative ion mode combined (DAN70). Repeat measurements on different days for THAP (155 lipids 2020–12–01, 144 lipids 2020–10–19) and CHCA (136 lipids 2021–01–19, 124 lipids 2020–10–15) in positive ion mode, and DAN70 in positive (88 lipids 2020–10–07, 83 lipids 2021–01–25) and negative ion mode (136 lipids 2020–09–24, 120 lipids 2021–01–26) showed similar results (data sets included in the cerebellum METASPACE projects). Other parameters and/or databases (e.g., HMBD, SwissLipids) could lead to different results in terms of peak identity and the number of species, but this consistent approach was suitable to determine performance of each matrix/protocol.

In negative ion mode, DAN70 was the only matrix that enabled the detection of a broad range of lipid species while DHAP and THAP mainly produced sulfatide and ganglioside signals (Figure 2b/d). Tissue washing (DANw) did increase signal intensities about 1.5-fold and double the S/N ratio for certain peaks (Figure S6) compared to DAN70. Jackson et al. noted that DHAP is superior to DAN for detecting gangliosides, especially for intact GD1(d36:1) at m/z 1835.965. Similarly, in this study DHAP was the only matrix producing sufficient GD1 signal to perform on tissue tandem-MS (Supporting Information 2).

PCA analysis of the spectra (average sum spectra for the whole cerebellum image, as shown in Figure S1) shows that, in positive ion mode (Figure 2c, loadings plots in Figure S7), matrices CHCA, Nor and DHB produce higher intensity, sodiated/potassiated [M+Na/K]+ species while THAP and DAN (especially with AmAc wash, DANw) favor protonated [M + H]+ species. Also, DAN70 shows higher intensities for long chain fatty acid PC species. It has been reported that the addition of ammonium salts to a matrix, aids in suppressing sodium and potassium, especially for oligonucleotide analysis with THAP.40,50 Similarly, we observed that THAPnS (no salt added) produces almost exclusively [M+Na/K]+ species. We did not observe a drastic signal increase after tissue washing (DANw), as was previously reported,42,51 and no improvement in the number of detected species (Figure 2e). The slight decrease in the number of detected species can be explained by the lack of salt adducts which often cause a single lipid species to be detected 3 times in positive ion mode [M+H/Na/K]+. The addition of acids, used in the CHCA and THAP protocols, to DAN (DANtfa, DANhfba) did not improve performance either. Based on those results, we chose the matrices providing the best lipid coverage and the highest signal (CHCA, THAP and DAN70) to test the ultimate capabilities of our instrumental setup.

Instrumentation capabilities: Spatial resolution, mass resolution and tandem-MS

The MassTech AP-MALDI source in combination with an Orbitrap mass analyzer and an HTX TM sprayer for sample preparation, allows to attain MS-imaging with 10 μm spatial resolution while collecting high resolution mass spectra (up to 240k FWHM at m/z = 400), and on-tissue tandem-MS spectra. Figure 3 shows the analysis of the hippocampus from a sagittal mouse brain section (also shown in Figure 1d) and lipid species PS(18:1/18:0)–H at m/z 788.5447 and PC(16:1/22:6)–CH3 at m/z 788.5236. To distinguish those lipids, a resolving power of about 40k is necessary, which is well within the capabilities of the Orbitrap (Figure 3a). PC(16:1/22:6) is exclusively located in the hippocampus while PS(18:1/18:0) is mainly in the surrounding fiber tracts (Figure 3b). Linescans across the hippocampus/fiber tract interface demonstrate that we can monitor chemical changes with a spatial resolution of 10 μm (Figure 3c, position of the linescans is shown in 2b, indicated with white arrows).

SEM images show that DAN70 matrix applied with an HTX-TM sprayer produces ∼1 μm crystals and that the AP-MALDI laser can be focused below 10 μm (Figure S3). At 10 μm spatial resolution sufficient signal is produced to assign 52 species with DAN70 in the hippocampus (72 with CHCA, 134 with THAP) in positive ion mode, and 121 species in negative ion mode (METASPACE, database: LIPIDMAPS, FDR:20%). Significantly more species are assigned in the THAP data set due to the slightly different analysis area which included the lateral ventricle with unique lipids. (The cerebellum data sets used to compare lipid coverage do not have this issue). In terms of spatial resolution, CHCA and DAN70 produce sharp images and perform better than THAP although all data sets were acquired with the same analysis settings (line spacing and scanning speed) and laser focus was <10 μm (Figure S2). The reason could be that laser settings were adjusted for maximum signal and were higher for THAP than for other matrices. Less laser energy would have provided less signal but could have resulted in a sharper image.

Figure 3d shows a tandem-MS spectrum of m/z 788.5, and it demonstrates that AP-MALDI produces sufficient lipid signal on tissue to perform tandem-MS analysis. PS(18:1/18:0) and PC(16:1/22:6) are fairly well spatially separated in the hippocampus but can overlap in other brain areas and are therefore both present in the tandem-MS spectrum (Figure 3d). Due to the complexity of biological samples, it is to be expected that a tandem-MS spectrum will contain more than one lipid species, albeit with different intensities. Due to the high mass resolution and mass accuracy provided by the Orbitrap this is not an issue, as one still can identify both species and assign their fragments accordingly.

The identity of 76 peaks (listed in Table 2) was confirmed with on tissue tandem-MS (MS2 and MS3 for GD1(d36:1)) of which 60 were unique lipid species and 16 were repeat species detected as [M+H/Na/K]+, were detected in both ion modes (e.g., PE(18:0/22:6) as [M ± H]± at m/z 790.539 and 792.554), or peaks consisting of lipid dimers (e.g., m/z 1548.194 = PC(18:1/18:0)+PC(16:0/18:1), ≠ CL(78:1)). The mass accuracy (ppm) in full scan (MS1) and tandem-MS data are within ±1 ppm (with lock mass) and ±3 ppm (without lock mass). The m/z values in Table 2 stem from the tandem-MS spectra, where the lock-mass often was not included. All tandem-MS spectra, with analysis conditions in the scan header plus: identified fragments, mass accuracy and molecular formulas of precursor and fragment ions, are listed in the Supporting Information 2. For simplicity, phospholipid fragments resulting from a headgroup loss (e.g., PS -serine) are referred to as PA(x/x) and fatty acid fragments as FA(x/x). Lipid species and fragments were identified using a combination of the LipostarMSI lipid catalogue with rule-based fragmentation entries, LIPIDMAPS and published literature.52−60

Imaging Single Cells in Tissues with AP-MALDI

Figure 4 shows that AP-MALDI imaging with DAN70 matrix has the potential of imaging individual neuron cells in tissues. PC(18:0/22:6) (Figure 4a,c) shows a unique distribution, localized to small areas in the hippocampus. The associated microscopic image (Figure 4b), taken before matrix application, shows small features corresponding to this distribution, that appear to be single cells. Their location, mainly in the stratum oriens, surrounding pyramidal cells (PC(18:1:20:4) blue, Figure 4a), suggests that those are inhibitory neurons/basket cells. This is supported by the resemblance of the PC(18:0/22:6) hippocampus images (Figure 4a,c) to immunohistochemistry images of basket cells48 and the PC(18:0/22:6) distribution in the cerebellum, most intense in the region of the Purkinje cells (Figure 4d,e) that are surrounded by basket cells.49 This association will have to be confirmed in future experiments.

Cholesterol Imaging with AP-MALDI

Overall, DAN matrix performed best or well in all experiments we conducted, apart from cholesterol imaging. Figure 5 shows images (Figure 5a) and spectra (Figure 5b) of the striatum/fiber tracts with cholesterol as [M – H2O + H]+ species (m/z 369.3516). Cholesterol species [M – H]+ at m/z 385.3464, reported in the literature,61−63 was detected at ∼10% of the intensity of m/z 369.3516. Cholesterol is not detectable with DAN70 and Nor, which produces low intensity signals in general. DHAP and THAP produced relatively low (below 1 × 103 counts) and CHCA and DHB high cholesterol signals (1 × 104 counts and above). While cholesterol detection was vastly different, other lipids were detected at similar levels in the range of 1–4 × 104 (Figure S8, lower for Nor). For DHB, cholesterol-related peaks were the most intense signals in the white matter.

Figure 5.

Figure 5

Open in a new tab

Cholesterol imaging with AP-MALDI, SEM, and ToF-SIMS. (a) AP-MALDI fiber tract/striatum images with various matrices showing cholesterol at m/z 369.3516 (green) and PC(16:0/16:0) at m/z 734.5688 (red). Scalebar: 500 μm. (b) Associated mass spectra showing the cholesterol peak and detection levels for each matrix. (c) SEM and (d) ToF-SIMS image of a brain section covered in DAN matrix, in the fiber tract region, cholesterol, m/z 369.35 (green); PC-headgroup, m/z 184.07 (red); m/z 196.87 (blue).

The tendency of cholesterol to migrate to the surface and crystallize during tissue drying has been reported previously.64,65 During matrix application, DAN70 does not seem to alter the cholesterol crystals. SEM images of DAN70 on brain sections show regular DAN crystals in the gray matter (Figure S3). In the fiber tract region, additional larger spikelike crystals are visible (Figure 5c). Those crystals resemble cholesterol crystals on dried tissue sections reported in the literature.65 Indeed, ToF-SIMS imaging on brain tissue covered with DAN70 matrix confirmed them to be cholesterol crystals (Figure 5d). The addition of 0.1% TFA to DAN70 matrix did not alleviate the issue. No such crystals were found on the surfaces of tissue slices covered with CHCA or DHB, suggesting that these matrices can dissolve and incorporate cholesterol, therefore enabling its detection with MALDI.

Discussion

We analyzed 11 sample preparation protocols with six different, vacuum stable, and unstable matrices to evaluate their performance for lipid imaging on brain tissue with AP-MALDI-Orbitrap-MSI and on-tissue tandem-MS.

Using strict selection criteria, we were able to assign 155/136 lipid species in AP-MALDI images in positive-/negative-ion mode. These results are comparable to vacuum MALDI-MSI approaches.66 AP-MALDI detected lipids with sufficient S/N to perform on-tissue tandem-MS and determine the structure of 76 peaks (60 unique lipids, Table 2) and with the added benefit of less pronounced tissue drying and cracking due to the harsh vacuum environment.29,32 Of the studied matrices, DAN was the most versatile matrix (high spatial resolution, high signal intensity and numerous lipids detected in positive+negative ion mode, similar results found for vacuum MALDI),67 but other matrices were superior for studying specific molecules (e.g., DHAP for gangliosides, DHB for cholesterol, more species detected in positive ion with THAP/CHCA). DAN protocol alterations such as tissue washing (DANw) and the addition of acids (DANhbfa, DANtfa) did increase performance but not as drastically as changing the solvent proportions, increasing water and decreasing acetonitrile. Using DAN dissolved in 70% acetonitrile and 30% water (DAN70) increased the number of detected lipids more than 2-fold (compared to 90% acetonitrile in DAN90) without decreasing spatial resolution. Further increasing the water content might yield even better results but could make it difficult to dissolve DAN completely. Norharmane could be used in both ion modes as well, but due to low signal intensity (even taking the low noise levels in the spectrum into account), many isotopical peaks fell below the signal-to-noise threshold and therefore fewer peaks were assigned with our selection criteria. Nevertheless, many lipids were present and the full brain analysis with norharmane could be used to guide tandem-MS analysis.

The majority of MALDI-MSI publications state their imaging resolution/pixel size in terms of laser crater size or stage stepping size.32 This can be misleading, as it does not accurately reflect the spatial resolution for detecting chemical changes in the sample. In this study, we demonstrated our spatial resolution of 10 μm with on-tissue linescans of lipid species that combine laser focus, stage stepping size (or in our case, raster speed), and crystal size into one metric. A better laser focus is possible, but this would significantly reduce the signal. A value of 10 μm corresponds to the size of one cell and was sufficient to correlate lipid species with single cells/neurons in the hippocampus (Figure 4). It should be noted that even though it is possible to image an entire brain section at 10 μm spatial resolution and at high mass resolution, the resulting data files would be too large to process on most computers. Therefore, reducing the spatial resolution for the analysis of larger areas is advised.

DHB used to be the gold-standard for MALDI analysis, and it is still used in many matrix comparison studies.38,42 Here, DHB was outperformed in all aspects by other matrices, other than for cholesterol detection. Cholesterol is of interest to many scientists because of its abundance and its involvement in metabolism and disease.68,69 It can be easily detected with ToF-SIMS,65 but MALDI usually requires additional steps to enhance cholesterol ionization.43,70 The main issue seems to be that cholesterol forms large, solid crystals on the sample surface. Even though they are destroyed by the laser, cholesterol is not ionized sufficiently without being integrated into (and cocrystallized with) the matrix. Therefore, only matrix protocols that can dissolve cholesterol crystals enable its detection with (AP-)MALDI-MSI. The DAN-matrix protocols used here left the cholesterol crystals intact; however, other lipids were still detected in the fiber tracts.

For tandem-MS analysis, THAP and DAN70 worked comparatively well due to their high signal intensities and lipid coverage. For most lipids assigned with LipostarMSI and METASPACE, tandem-MS confirmed their identity in accordance with their possible assignments. Only the assigned cardiolipins detected in positive-ion mode were discovered to be lipid dimers instead. This highlights the importance of tandem-MS analysis, not only for the structural elucidation of the detected species but also for confident assignments.

All data sets included in this study contained hundreds of assigned species with different distributions. This data would have been too vast to include in this manuscript. Therefore, for transparency all data sets were uploaded to METASPACE, where their lipid distributions can be viewed. Additionally, the METASPCE projects, created for this publication, can be expanded upon, as further analyses with novel matrix compounds are performed. Links to all data set: full brain positive/negative: AP_MALDI_Full_Brain; hippocampus positive/negative: AP_MALDI_Hippocampus; cerebellum positive: AP_MALDI_MATRIX_pos; cerebellum negative: AP_MALDI_MATRIX_neg.

Conclusion

For a long time, lipids have taken a backseat to proteins concerning their importance in disease mechanisms. This was partially due to the inability to track the changes in lipid distributions in heterogeneous samples, changes which can be very subtle in homogenized sample extracts. MSI techniques like AP-MALDI-imaging can capture those changes, but data quality can depend strongly on sample preparation. We tested 11 sample preparation protocols for six matrices and the instrument capabilities of the AP-MALDI-Orbitrap system for lipidomics studies. We defined their characteristics in terms of lipid coverage, signal intensities, high spatial resolution imaging capability and usability for positive and negative ion mode. Every matrix had its advantages and disadvantages and knowing their characteristics is crucial for deciding which one is best suited for the scientific needs of a study. For now, we recommend THAP for on tissue tandem-MS, CHCA or DAN for high spatial resolution imaging in positive-ion mode and DAN for high spatial resolution imaging and tandem-MS in negative-ion mode. For multiple experiments on one sample (positive- and negative-ion mode imaging plus tandem-MS) DAN matrix in a solution with higher water content (as demonstrated here with DAN70 containing 30% H2O) is the best suited matrix. The matrix recipes were adapted from previous MALDI/AP-MALDI publications and had all been optimized by their respective users, but they ultimately represented only a small number of options. Therefore, we intend to keep testing matrices and expanding the publicly available data sets in METASPACE. In conclusion, AP-MALDI has shown to be comparable to vacuum MALDI for lipid detection, with the added benefit of less pronounced tissue drying and no requirement for vacuum stable matrices. In addition, the AP-MALDI source in combination with Orbitrap-MS allows for a direct transposition of conventional LC–MS fragmentation parameters for lipids. Therefore, AP-MALDI can be considered cost-effective addition to widely available LC/HRMS instruments and a valuable asset in applied, biomedical research.

Acknowledgments

We thank Roberto Angelini, Eylan Yutuc, and William J. Griffiths from the Medical School, Swansea University, UK, for providing the brain tissue sections, MassTech, Inc. for their financial and material support of the AP-MALDI demo lab at LIST, and Venkat Panchagnula for proofreading the manuscript. The project was supported by the Luxembourg National Research Fund (FNR) (SKIMAS, Project number: 14292830).

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jasms.1c00327.

  • Additional AP-MALDI images for all data sets included in this study, SEM images of DAN70 and CHCA matrices, mass spectra for cerebellum analysis with all matrices in positive-/negative-ion mode and signal-to-noise levels for selected species, PCA loadings plots, and images and spectra of the small area-striatum analyses (PDF)

  • Tandem-MS data for 76 peaks with assigned fragments (PDF)

Author Contributions

T.B.A. and G.F. designed the experiments. J.B. performed the ToF-SIMS analysis. J.L.B. performed the SEM analysis. T.B.A. performed the AP-MALDI analyses and wrote the manuscript. All authors commented on the manuscript and discussed the results.

The authors declare no competing financial interest.

Supplementary Material

js1c00327_si_001.pdf (1.5MB, pdf)
js1c00327_si_002.pdf (4.3MB, pdf)

References

  1. Herzog R. F. K.; Viehböck F. P. Ion Source for Mass Spectrography. Phys. Rev. 1949, 76 (6), 855–856. 10.1103/PhysRev.76.855. [DOI] [Google Scholar]
  2. Shariatgorji M.; Nilsson A.; Fridjonsdottir E.; Vallianatou T.; Källback P.; Katan L.; Sävmarker J.; Mantas I.; Zhang X.; Bezard E.; Svenningsson P.; Odell L. R.; Andrén P. E. Comprehensive mapping of neurotransmitter networks by MALDI-MS imaging. Nat. Methods 2019, 16 (10), 1021–1028. 10.1038/s41592-019-0551-3. [DOI] [PubMed] [Google Scholar]
  3. Tian H.; Sheraz née Rabbani S.; Vickerman J. C.; Winograd N. Multiomics Imaging Using High-Energy Water Gas Cluster Ion Beam Secondary Ion Mass Spectrometry [(H2O)n-GCIB-SIMS] of Frozen-Hydrated Cells and Tissue. Anal. Chem. 2021, 93 (22), 7808–7814. 10.1021/acs.analchem.0c05210. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Thomen A.; Najafinobar N.; Penen F.; Kay E.; Upadhyay P. P.; Li X.; Phan N. T. N.; Malmberg P.; Klarqvist M.; Andersson S.; Kurczy M. E.; Ewing A. G. Subcellular Mass Spectrometry Imaging and Absolute Quantitative Analysis across Organelles. ACS Nano 2020, 14 (4), 4316–4325. 10.1021/acsnano.9b09804. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Kaya I.; Jennische E.; Lange S.; Malmberg P. Multimodal chemical imaging of a single brain tissue section using ToF-SIMS, MALDI-ToF and immuno/histochemical staining. Analyst 2021, 146 (4), 1169–1177. 10.1039/D0AN02172E. [DOI] [PubMed] [Google Scholar]
  6. Kaya I.; Sämfors S.; Levin M.; Borén J.; Fletcher J. S. Multimodal MALDI Imaging Mass Spectrometry Reveals Spatially Correlated Lipid and Protein Changes in Mouse Heart with Acute Myocardial Infarction. J. Am. Soc. Mass Spectrom. 2020, 31 (10), 2133–2142. 10.1021/jasms.0c00245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Van Nuffel S.; Quatredeniers M.; Pirkl A.; Zakel J.; Le Caer J.-P.; Elie N.; Vanbellingen Q. P.; Dumas S. J.; Nakhleh M. K.; Ghigna M.-R. Multimodal imaging mass spectrometry to identify markers of pulmonary arterial hypertension in human lung tissue using MALDI-ToF, ToF-SIMS, and hybrid SIMS. Anal. Chem. 2020, 92 (17), 12079–12087. 10.1021/acs.analchem.0c02815. [DOI] [PubMed] [Google Scholar]
  8. Buchberger A. R.; DeLaney K.; Johnson J.; Li L. Mass Spectrometry Imaging: A Review of Emerging Advancements and Future Insights. Anal. Chem. 2018, 90 (1), 240–265. 10.1021/acs.analchem.7b04733. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Goodwin R. J. A.; Takats Z.; Bunch J. A Critical and Concise Review of Mass Spectrometry Applied to Imaging in Drug Discovery. SLAS DISCOVERY: Advancing the Science of Drug Discovery 2020, 25 (9), 963–976. 10.1177/2472555220941843. [DOI] [PubMed] [Google Scholar]
  10. Amstalden van Hove E. R.; Smith D. F.; Heeren R. M. A. A concise review of mass spectrometry imaging. Journal of Chromatography A 2010, 1217 (25), 3946–3954. 10.1016/j.chroma.2010.01.033. [DOI] [PubMed] [Google Scholar]
  11. Pól J.; Strohalm M.; Havlíček V.; Volný M. Molecular mass spectrometry imaging in biomedical and life science research. Histochemistry and Cell Biology 2010, 134 (5), 423–443. 10.1007/s00418-010-0753-3. [DOI] [PubMed] [Google Scholar]
  12. Römpp A.; Both J.-P.; Brunelle A.; Heeren R. M. A.; Laprévote O.; Prideaux B.; Seyer A.; Spengler B.; Stoeckli M.; Smith D. F. Mass spectrometry imaging of biological tissue: an approach for multicenter studies. Anal. Bioanal. Chem. 2015, 407 (8), 2329–2335. 10.1007/s00216-014-8410-7. [DOI] [PubMed] [Google Scholar]
  13. O’Donnell V. B.; Ekroos K.; Liebisch G.; Wakelam M. Lipidomics: Current state of the art in a fast moving field. WIREs Systems Biology and Medicine 2020, 12 (1), e1466 10.1002/wsbm.1466. [DOI] [PubMed] [Google Scholar]
  14. Ellis S. R.; Paine M. R. L.; Eijkel G. B.; Pauling J. K.; Husen P.; Jervelund M. W.; Hermansson M.; Ejsing C. S.; Heeren R. M. A. Automated, parallel mass spectrometry imaging and structural identification of lipids. Nat. Methods 2018, 15 (7), 515–518. 10.1038/s41592-018-0010-6. [DOI] [PubMed] [Google Scholar]
  15. Kaya I.; Brinet D.; Michno W.; Başkurt M.; Zetterberg H.; Blenow K.; Hanrieder J. Novel Trimodal MALDI Imaging Mass Spectrometry (IMS3) at 10 μm Reveals Spatial Lipid and Peptide Correlates Implicated in Aβ Plaque Pathology in Alzheimer’s Disease. ACS Chem. Neurosci. 2017, 8 (12), 2778–2790. 10.1021/acschemneuro.7b00314. [DOI] [PubMed] [Google Scholar]
  16. Eberlin L. S.; Ferreira C. R.; Dill A. L.; Ifa D. R.; Cooks R. G. Desorption electrospray ionization mass spectrometry for lipid characterization and biological tissue imaging. Biochimica et Biophysica Acta (BBA) - Molecular and Cell Biology of Lipids 2011, 1811 (11), 946–960. 10.1016/j.bbalip.2011.05.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Dimovska Nilsson K.; Neittaanmäki N.; Zaar O.; Angerer T. B.; Paoli J.; Fletcher J. S. TOF-SIMS imaging reveals tumor heterogeneity and inflammatory response markers in the microenvironment of basal cell carcinoma. Biointerphases 2020, 15 (4), 041012. 10.1116/6.0000340. [DOI] [PubMed] [Google Scholar]
  18. Tian H.; Sparvero L. J.; Anthonymuthu T. S.; Sun W.-Y.; Amoscato A. A.; He R.-R.; Bayır H.; Kagan V. E.; Winograd N. Successive High-Resolution (H2O)n-GCIB and C60-SIMS Imaging Integrates Multi-Omics in Different Cell Types in Breast Cancer Tissue. Anal. Chem. 2021, 93 (23), 8143–8151. 10.1021/acs.analchem.0c05311. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Zavalin A.; Yang J.; Hayden K.; Vestal M.; Caprioli R. M. Tissue protein imaging at 1 μm laser spot diameter for high spatial resolution and high imaging speed using transmission geometry MALDI TOF MS. Anal Bioanal Chem. 2015, 407 (8), 2337–42. 10.1007/s00216-015-8532-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Niehaus M.; Soltwisch J.; Belov M. E.; Dreisewerd K. Transmission-mode MALDI-2 mass spectrometry imaging of cells and tissues at subcellular resolution. Nat. Methods 2019, 16 (9), 925–931. 10.1038/s41592-019-0536-2. [DOI] [PubMed] [Google Scholar]
  21. DeLaney K.; Phetsanthad A.; Li L. Advances in high-resolution maldi mass spectrometry for neurobiology. Mass Spectrom. Rev. 2020, 10.1002/mas.21661. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Andersen M. K.; Høiem T. S.; Claes B. S. R.; Balluff B.; Martin-Lorenzo M.; Richardsen E.; Krossa S.; Bertilsson H.; Heeren R. M. A.; Rye M. B.; Giskeødegård G. F.; Bathen T. F.; Tessem M.-B. Spatial differentiation of metabolism in prostate cancer tissue by MALDI-TOF MSI. Cancer & Metabolism 2021, 9 (1), 9. 10.1186/s40170-021-00242-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Colsch B.; Woods A. S. Localization and imaging of sialylated glycosphingolipids in brain tissue sections by MALDI mass spectrometry. Glycobiology 2010, 20 (6), 661–667. 10.1093/glycob/cwq031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Jackson S. N.; Muller L.; Roux A.; Oktem B.; Moskovets E.; Doroshenko V. M.; Woods A. S. AP-MALDI Mass Spectrometry Imaging of Gangliosides Using 2,6-Dihydroxyacetophenone. Journal of The American Society for Mass Spectrometry 2018, 29 (7), 1463–1472. 10.1007/s13361-018-1928-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Doroshenko V. M.; Laiko V. V.; Taranenko N. I.; Berkout V. D.; Lee H. S. Recent developments in atmospheric pressure MALDI mass spectrometry. Int. J. Mass Spectrom. 2002, 221 (1), 39–58. 10.1016/S1387-3806(02)00893-X. [DOI] [Google Scholar]
  26. Moyer S. C.; Marzilli L. A.; Woods A. S.; Laiko V. V.; Doroshenko V. M.; Cotter R. J. Atmospheric pressure matrix-assisted laser desorption/ionization (AP MALDI) on a quadrupole ion trap mass spectrometer. Int. J. Mass Spectrom. 2003, 226 (1), 133–150. 10.1016/S1387-3806(02)00972-7. [DOI] [Google Scholar]
  27. Geier B.; Sogin E. M.; Michellod D.; Janda M.; Kompauer M.; Spengler B.; Dubilier N.; Liebeke M. Spatial metabolomics of in situ host-microbe interactions at the micrometre scale. Nat. Microbiol 2020, 5 (3), 498–510. 10.1038/s41564-019-0664-6. [DOI] [PubMed] [Google Scholar]
  28. Bancroft J. D.; Gamble M.. Theory and practice of histological techniques; Elsevier Health Sciences, 2008. [Google Scholar]
  29. Ostrowski S. G.; Paxon T. L.; Denault L.; McEvoy K. P.; Smentkowski V. S. Preparing Biological Samples for Analysis by High Vacuum Techniques. Microscopy Today 2009, 17 (2), 48–53. 10.1017/S1551929500054511. [DOI] [Google Scholar]
  30. Ryumin P.; Brown J.; Morris M.; Cramer R. Investigation and optimization of parameters affecting the multiply charged ion yield in AP-MALDI MS. Methods 2016, 104, 11–20. 10.1016/j.ymeth.2016.01.015. [DOI] [PubMed] [Google Scholar]
  31. Keller C.; Maeda J.; Jayaraman D.; Chakraborty S.; Sussman M. R.; Harris J. M.; Ané J.-M.; Li L. Comparison of Vacuum MALDI and AP-MALDI Platforms for the Mass Spectrometry Imaging of Metabolites Involved in Salt Stress in Medicago truncatula. Frontiers in Plant Science 2018, 9, 1238. 10.3389/fpls.2018.01238. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kompauer M.; Heiles S.; Spengler B. Atmospheric pressure MALDI mass spectrometry imaging of tissues and cells at 1.4-μm lateral resolution. Nat. Methods 2017, 14 (1), 90–96. 10.1038/nmeth.4071. [DOI] [PubMed] [Google Scholar]
  33. Elia E. A.; Niehaus M.; Steven R. T.; Wolf J.-C.; Bunch J. Atmospheric Pressure MALDI Mass Spectrometry Imaging Using In-Line Plasma Induced Postionization. Anal. Chem. 2020, 92 (23), 15285–15290. 10.1021/acs.analchem.0c03524. [DOI] [PubMed] [Google Scholar]
  34. Niehaus M.; Robinson K. N.; Murta T.; Elia E. A.; Race A. M.; Yan B.; Steven R. T.; Bunch J. MALDI-2 at Atmospheric Pressure—Parameter Optimization and First Imaging Experiments. J. Am. Soc. Mass Spectrom. 2020, 31 (11), 2287–2295. 10.1021/jasms.0c00237. [DOI] [PubMed] [Google Scholar]
  35. Steven R. T.; Shaw M.; Dexter A.; Murta T.; Green F. M.; Robinson K. N.; Gilmore I. S.; Takats Z.; Bunch J. Construction and testing of an atmospheric-pressure transmission-mode matrix assisted laser desorption ionisation mass spectrometry imaging ion source with plasma ionisation enhancement. Anal. Chim. Acta 2019, 1051, 110–119. 10.1016/j.aca.2018.11.003. [DOI] [PubMed] [Google Scholar]
  36. Barré F.; Rocha B.; Dewez F.; Towers M.; Murray P.; Claude E.; Cillero-Pastor B.; Heeren R.; Porta Siegel T. Faster raster matrix-assisted laser desorption/ionization mass spectrometry imaging of lipids at high lateral resolution. Int. J. Mass Spectrom. 2019, 437, 38–48. 10.1016/j.ijms.2018.09.015. [DOI] [Google Scholar]
  37. Hochart G.; Bonnel D.; Stauber J.; Stamatas G. N. Biomarker Mapping on Skin Tape Strips Using MALDI Mass Spectrometry Imaging. J. Am. Soc. Mass Spectrom. 2019, 30 (10), 2082–2091. 10.1007/s13361-019-02277-5. [DOI] [PubMed] [Google Scholar]
  38. McMillen J. C.; Fincher J. A.; Klein D. R.; Spraggins J. M.; Caprioli R. M. Effect of MALDI matrices on lipid analyses of biological tissues using MALDI-2 postionization mass spectrometry. Journal of Mass Spectrometry 2020, 55 (12), e4663 10.1002/jms.4663. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Sun C.; Liu W.; Ma S.; Zhang M.; Geng Y.; Wang X. Development of a high-coverage matrix-assisted laser desorption/ionization mass spectrometry imaging method for visualizing the spatial dynamics of functional metabolites in Salvia miltiorrhiza Bge. Journal of Chromatography A 2020, 1614, 460704. 10.1016/j.chroma.2019.460704. [DOI] [PubMed] [Google Scholar]
  40. Pieles U.; Zürcher W.; Schär M.; Moser H. E. Matrix-assisted laser desorption ionization time-of-flight mass spectrometry: a powerful tool for the mass and sequence analysis of natural and modified oligonucleotides. Nucleic acids research 1993, 21 (14), 3191–6. 10.1093/nar/21.14.3191. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Leopold J.; Popkova Y.; Engel K. M.; Schiller J. Recent Developments of Useful MALDI Matrices for the Mass Spectrometric Characterization of Lipids. Biomolecules 2018, 8 (4), 173. 10.3390/biom8040173. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Angel P. M.; Spraggins J. M.; Baldwin H. S.; Caprioli R. Enhanced sensitivity for high spatial resolution lipid analysis by negative ion mode matrix assisted laser desorption ionization imaging mass spectrometry. Analytical chemistry 2012, 84 (3), 1557–1564. 10.1021/ac202383m. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Angelini R.; Yutuc E.; Wyatt M. F.; Newton J.; Yusuf F. A.; Griffiths L.; Cooze B. J.; El Assad D.; Frache G.; Rao W.; Allen L. B.; Korade Z.; Nguyen T. T. A.; Rathnayake R. A. C.; Cologna S. M.; Howell O. W.; Clench M. R.; Wang Y.; Griffiths W. J. Visualizing Cholesterol in the Brain by On-Tissue Derivatization and Quantitative Mass Spectrometry Imaging. Anal. Chem. 2021, 93 (11), 4932–4943. 10.1021/acs.analchem.0c05399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Palmer A.; Phapale P.; Chernyavsky I.; Lavigne R.; Fay D.; Tarasov A.; Kovalev V.; Fuchser J.; Nikolenko S.; Pineau C.; Becker M.; Alexandrov T. FDR-controlled metabolite annotation for high-resolution imaging mass spectrometry. Nat. Methods 2017, 14 (1), 57–60. 10.1038/nmeth.4072. [DOI] [PubMed] [Google Scholar]
  45. Robichaud G.; Garrard K. P.; Barry J. A.; Muddiman D. C. MSiReader: an open-source interface to view and analyze high resolving power MS imaging files on Matlab platform. J. Am. Soc. Mass Spectrom. 2013, 24 (5), 718–21. 10.1007/s13361-013-0607-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Tortorella S.; Tiberi P.; Bowman A. P.; Claes B. S. R.; Ščupáková K.; Heeren R. M. A.; Ellis S. R.; Cruciani G. LipostarMSI: Comprehensive, Vendor-Neutral Software for Visualization, Data Analysis, and Automated Molecular Identification in Mass Spectrometry Imaging. J. Am. Soc. Mass Spectrom. 2020, 31 (1), 155–163. 10.1021/jasms.9b00034. [DOI] [PubMed] [Google Scholar]
  47. Shibue M.; Mant C. T.; Hodges R. S. Effect of anionic ion-pairing reagent concentration (1–60 mM) on reversed-phase liquid chromatography elution behaviour of peptides. Journal of chromatography. A 2005, 1080 (1), 58–67. 10.1016/j.chroma.2005.02.047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Udakis M.; Pedrosa V.; Chamberlain S. E. L.; Clopath C.; Mellor J. R. Interneuron-specific plasticity at parvalbumin and somatostatin inhibitory synapses onto CA1 pyramidal neurons shapes hippocampal output. Nat. Commun. 2020, 11 (1), 4395. 10.1038/s41467-020-18074-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Zhou J.; Brown A.; Lackey E.; Arancillo M.; Lin T.; Sillitoe R. Purkinje cell neurotransmission patterns cerebellar basket cells into zonal modules defined by distinct pinceau sizes 2020, e55569. 10.7554/eLife.55569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Sugiyama E.; Masaki N.; Matsushita S.; Setou M. Ammonium Sulfate Improves Detection of Hydrophilic Quaternary Ammonium Compounds through Decreased Ion Suppression in Matrix-Assisted Laser Desorption/Ionization Imaging Mass Spectrometry. Anal. Chem. 2015, 87 (22), 11176–11181. 10.1021/acs.analchem.5b02672. [DOI] [PubMed] [Google Scholar]
  51. Wang H.-Y. J.; Liu C. B.; Wu H.-W. A simple desalting method for direct MALDI mass spectrometry profiling of tissue lipids. J. Lipid Res. 2011, 52 (4), 840–849. 10.1194/jlr.D013060. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Cerruti C. D.; Benabdellah F.; Laprevote O.; Touboul D.; Brunelle A. MALDI imaging and structural analysis of rat brain lipid negative ions with 9-aminoacridine matrix. Anal. Chem. 2012, 84 (5), 2164–71. 10.1021/ac2025317. [DOI] [PubMed] [Google Scholar]
  53. Hsu F. F.; Turk J. Studies on sulfatides by quadrupole ion-trap mass spectrometry with electrospray ionization: structural characterization and the fragmentation processes that include an unusual internal galactose residue loss and the classical charge-remote fragmentation. J. Am. Soc. Mass Spectrom. 2004, 15 (4), 536–46. 10.1016/j.jasms.2003.12.007. [DOI] [PubMed] [Google Scholar]
  54. Jarmusch A. K.; Alfaro C. M.; Pirro V.; Hattab E. M.; Cohen-Gadol A. A.; Cooks R. G. Differential Lipid Profiles of Normal Human Brain Matter and Gliomas by Positive and Negative Mode Desorption Electrospray Ionization - Mass Spectrometry Imaging. PLoS One 2016, 11 (9), e0163180 10.1371/journal.pone.0163180. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Magny R.; Regazzetti A.; Kessal K.; Genta-Jouve G.; Baudouin C.; Melik-Parsadaniantz S.; Brignole-Baudouin F.; Laprevote O.; Auzeil N. Lipid Annotation by Combination of UHPLC-HRMS (MS), Molecular Networking, and Retention Time Prediction: Application to a Lipidomic Study of In Vitro Models of Dry Eye Disease. Metabolites 2020, 10 (6), 225. 10.3390/metabo10060225. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Mi J.; Han Y.; Xu Y.; Kou J.; Li W. J.; Wang J. R.; Jiang Z. H. Deep Profiling of Immunosuppressive Glycosphingolipids and Sphingomyelins in Wild Cordyceps. J. Agric. Food Chem. 2018, 66 (34), 8991–8998. 10.1021/acs.jafc.8b02706. [DOI] [PubMed] [Google Scholar]
  57. Nakanishi H.; Iida Y.; Shimizu T.; Taguchi R. Analysis of oxidized phosphatidylcholines as markers for oxidative stress, using multiple reaction monitoring with theoretically expanded data sets with reversed-phase liquid chromatography/tandem mass spectrometry. J. Chromatogr B Analyt Technol. Biomed Life Sci. 2009, 877 (13), 1366–74. 10.1016/j.jchromb.2008.09.041. [DOI] [PubMed] [Google Scholar]
  58. Qu F.; Zhang H.; Zhang M.; Hu P. Sphingolipidomic Profiling of Rat Serum by UPLC-Q-TOF-MS: Application to Rheumatoid Arthritis Study. Molecules 2018, 23 (6), 1324. 10.3390/molecules23061324. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Spickett C. M.; Reis A.; Pitt A. R. Identification of oxidized phospholipids by electrospray ionization mass spectrometry and LC-MS using a QQLIT instrument. Free Radic Biol. Med. 2011, 51 (12), 2133–49. 10.1016/j.freeradbiomed.2011.09.003. [DOI] [PubMed] [Google Scholar]
  60. Zianni R.; Bianco G.; Lelario F.; Losito I.; Palmisano F.; Cataldi T. R. Fatty acid neutral losses observed in tandem mass spectrometry with collision-induced dissociation allows regiochemical assignment of sulfoquinovosyl-diacylglycerols. J. Mass Spectrom 2013, 48 (2), 205–15. 10.1002/jms.3149. [DOI] [PubMed] [Google Scholar]
  61. Börner K.; Nygren H.; Hagenhoff B.; Malmberg P.; Tallarek E.; Månsson J.-E. Distribution of cholesterol and galactosylceramide in rat cerebellar white matter. Biochimica et Biophysica Acta (BBA) - Molecular and Cell Biology of Lipids 2006, 1761 (3), 335–344. 10.1016/j.bbalip.2006.02.021. [DOI] [PubMed] [Google Scholar]
  62. Careri M.; Ferretti D.; Manini P.; Musci M. Evaluation of particle beam high-performance liquid chromatography-mass spectrometry for analysis of cholesterol oxides. Journal of Chromatography A 1998, 794 (1), 253–262. 10.1016/S0021-9673(97)00764-4. [DOI] [Google Scholar]
  63. Demirev P.; Barlo Daya D. D. N.; Brinkmaim G.; Fenyö D.; Häkansson P.; Sundqvist B. U. R. On molecular ion formation by hydride abstraction in plasma desorption mass spectrometry. International Journal of Mass Spectrometry and Ion Processes 1993, 123 (1), 69–75. 10.1016/0168-1176(93)87055-W. [DOI] [Google Scholar]
  64. Sjövall P.; Lausmaa J.; Johansson B. Mass spectrometric imaging of lipids in brain tissue. Analytical chemistry 2004, 76 (15), 4271–4278. 10.1021/ac049389p. [DOI] [PubMed] [Google Scholar]
  65. Angerer T. B.; Mohammadi A. S.; Fletcher J. S. Optimizing sample preparation for anatomical determination in the hippocampus of rodent brain by ToF-SIMS analysis. Biointerphases 2016, 11 (2), 02A319. 10.1116/1.4941064. [DOI] [PubMed] [Google Scholar]
  66. Bowman A. P.; Bogie J. F. J.; Hendriks J. J. A.; Haidar M.; Belov M.; Heeren R. M. A.; Ellis S. R. Evaluation of lipid coverage and high spatial resolution MALDI-imaging capabilities of oversampling combined with laser post-ionisation. Anal. Bioanal. Chem. 2020, 412 (10), 2277–2289. 10.1007/s00216-019-02290-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Kaya I.; Michno W.; Brinet D.; Iacone Y.; Zanni G.; Blennow K.; Zetterberg H.; Hanrieder J. Histology-Compatible MALDI Mass Spectrometry Based Imaging of Neuronal Lipids for Subsequent Immunofluorescent Staining. Anal. Chem. 2017, 89 (8), 4685–4694. 10.1021/acs.analchem.7b00313. [DOI] [PubMed] [Google Scholar]
  68. Björkhem I.; Leoni V.; Meaney S. Genetic connections between neurological disorders and cholesterol metabolism. J. Lipid Res. 2010, 51 (9), 2489–2503. 10.1194/jlr.R006338. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Griffiths W. J.; Wang Y. Oxysterol research: a brief review. Biochem. Soc. Trans. 2019, 47 (2), 517–526. 10.1042/BST20180135. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Barré F. P. Y.; Paine M. R. L.; Flinders B.; Trevitt A. J.; Kelly P. D.; Ait-Belkacem R.; Garcia J. P.; Creemers L. B.; Stauber J.; Vreeken R. J.; Cillero-Pastor B.; Ellis S. R.; Heeren R. M. A. Enhanced Sensitivity Using MALDI Imaging Coupled with Laser Postionization (MALDI-2) for Pharmaceutical Research. Anal. Chem. 2019, 91 (16), 10840–10848. 10.1021/acs.analchem.9b02495. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

js1c00327_si_001.pdf (1.5MB, pdf)
js1c00327_si_002.pdf (4.3MB, pdf)

Articles from Journal of the American Society for Mass Spectrometry are provided here courtesy of American Chemical Society

RESOURCES