Desorption ionization using through-hole alumina membrane offers higher reproducibility than 2,5-dihydroxybenzoic acid, a widely used matrix in Fourier transform ion cyclotron resonance mass spectrometry imaging analysis

lower signal intensity in standard PC analysis. Interestingly, it offered higher signal intensities for all the endogenous lipid ions. Lower fluctuations of both mass accuracies and signal intensities were observed in DIUTHAME-MS. Conclusions: Our results demonstrated that DIUTHAME-MS offers higher reproducibility for mass accuracies and intensities than MALDI-MS in both standard lipid and mouse brain tissue analyses. It can potentially be used instead of conventional MALDI-MS and mass spectrometry imaging analyses to achieve highly reproducible data for mass accuracy and intensity.

lower signal intensity in standard PC analysis. Interestingly, it offered higher signal intensities for all the endogenous lipid ions. Lower fluctuations of both mass accuracies and signal intensities were observed in DIUTHAME-MS.
Conclusions: Our results demonstrated that DIUTHAME-MS offers higher reproducibility for mass accuracies and intensities than MALDI-MS in both standard lipid and mouse brain tissue analyses. It can potentially be used instead of conventional MALDI-MS and mass spectrometry imaging analyses to achieve highly reproducible data for mass accuracy and intensity.

| INTRODUCTION
Mass spectrometry imaging (MSI) is a label-free molecular imaging technique that allows the simultaneous mapping of hundreds of molecules in tissues or cells. Various ionization techniques such as matrix-assisted laser desorption ionization (MALDI), 1 desorption electrospray ionization, 2 and secondary ion 3

have been used in MSI
for the visualization of proteins, 4 fatty acids, 5 phospholipids, 6,7 nucleotide, 8 neurotransmitters, 9 small metabolites, and exogenous compounds 10,11 in a wide range of biological samples. MALDI-MSI is currently the most widely used MSI technique where thin tissue sections are generally thaw-mounted on an indium tin oxide (ITO)coated glass slide and subsequently coated with a matrix solution, resulting in the formation of co-crystals between the matrix molecules and the analytes. The sample surface is then irradiated by laser energy that causes the rapid heating of the matrix molecules. Consequently, the heated molecules, as well as the analytes, undergo desorption and ionization. Ionization of the analytes is thought to occur through charge transfer between the excited matrix molecules and the neutral analytes. 12,13 MALDI-MSI has been shown to have the capability of detecting molecules with higher sensitivity in a wide mass range than the other MSI techniques. However, the quality and the reproducibility of MALDI data are directly affected by the sample preparation method. 14 The results often differ from person to person, even from experiment to experiment, due to the difficulty in maintaining identical conditions for matrix application. 15,16 Homogenous spraying has been achieved by the emergence of automatic matrix sprayers, such as the TM-Sprayer™ (HTX Technologies, Carrboro, NC, USA) or the iMLayer™ (Shimadzu Co. Ltd, Kyoto, Japan), although these are costly and the procedure is time-consuming. 13 To combat matrix crystal heterogeneity, ionic liquid matrices (ILMs) have been reported to offer higher signal intensity, enhanced spot homogeneity, increased signal reproducibility, and similar or better detection limits compared with the widely used crystalline matrix 2,5-dihydroxybenzoic acid (DHB). Unfortunately, the use of ILMs provides a stronger tendency for alkali metal-ion adduct generation. 17 In addition, the optimization of ILMs is a trial-and-error process, and the base-to-acid ratio can alter the performance of these materials. 18 Thus, sample preparation is still a bottleneck in achieving reproducible data in MALDI-MS and MSI analyses.
High mass accuracy acts as a powerful "filter" that significantly limits the number of possible molecular formulae. 19,20 Therefore, high reproducibility of accurate mass measurements is essential for characterizing small molecules. 21,22 Fourier transform MS or Fourier transform ion cyclotron resonance (FTICR) MS facilitates the highest mass accuracy and resolving power among all the most frequently used MS techniques over a broad m/z range. [23][24][25] Even for a small number of ions, it outperforms many other frequently applied MS methods. 26 Our group has previously exploited this advantage of FTICR to explore active chemical components in a traditional beverage. 27 However, the space charge effects caused by a high number of ions derived from the DHB matrix have been reported to reduce the reproducibility of accurate mass measurements. 28 Surface-assisted laser desorption ionization (SALDI) is an alternative technique, which uses the optophysical properties of solid compounds rather than the spectral properties of organic matrix compounds. It could be a useful technique for detecting low-mass molecules from the sample without matrix-derived peaks. 29 Previously, matrix-free laser desorption ionization methods have been described that are based on nanostructure surfaces such as nanomaterial-assisted laser desorption ionization, 30 36 In addition, the preparations of those inorganic materials can be complex and costly. 29 The regularity and the diameters of the nanoposts are highly critical for achieving optimized results. 45 Recently, our group has developed novel ionization-assisting membrane chips based on SALDI procedures that use a porous alumina membrane called desorption ionization using throughhole alumina membrane (DIUTHAME). The capillary action of DIUTHAME has a very large aspect ratio across the throughhole alumina membrane that enables many loading protocols, including sample impregnation from the opposite side to the laser exposure surface. It is assumed that the size and shape of the through-holes lead to the conversion and transfer of photon energies. Previously, our group reported that the ionic yield depends to some degree on the inner diameter of the throughhole alumina membrane. 46 Successful feasibility studies in which a DIUTHAME chip was applied for the MS study of liquid samples 46 and MSI analysis of brain tissue sections of mice 47 were previously reported. In contrast to MALDI, DIUTHAME is a matrix-free ionization-assisting method; thus, interference induced by matrix-derived ion peaks in the lower mass range, generally found in MALDI-MS, is not observed in DIUTHAME-MS analyses. [46][47][48] In addition, DIUTHAME significantly reduces the sample preparation time and does not require a qualified technician or specialized instrument for matrix application. 46,47 In the study presented here, we used a DIUTHAME chip in the preparation of standard lipid and mouse brain tissue samples followed by FTICR MS analysis to investigate the reproducibility of mass accuracies and signal intensities of the measured ions. The results were compared with those obtained using MALDI-MS, where DHB was used as a matrix. Hamamatsu Photonics. The shape of the through-holes, the developmental procedures, and characteristics have been previously described in detail. 46 All materials used in this study were of the highest purity. Two C57BL/6J male mice (age: 90 days) purchased from SLC Inc.

| Sample preparation for MS analysis
(Hamamatsu, Japan) were used in this study. The mouse brains were excised after euthanasia by cervical dislocation and subsequently frozen in powdered dry ice and stored at −80 C in a refrigerator until analysis. The frozen tissues were cryosectioned (coronal section; 10 μm for MALDI-MS and 30 μm for DIUTHAME-MS) at −20 C using a CM1950 cryostat (Leica Biosystems, Buffalo Grove, IL, USA). The DIUTHAME chip was then applied on the tissue section of the mouse brain, as described previously. 47 Briefly, inside the cryostat chamber, the tissue section of the mouse brain (30 μm thick) was mounted on a precooled conductive ITO-coated glass slide (100 Ω; Matsunami, Japan), and a precooled DIUTHAME chip was placed very gently onto the tissue section ( Figure 1). A finger was pressed (2-3 min) straight down onto the tissue slice on the ITO-coated glass slide such that the molecules in the tissue sample could thaw and be raised off the upper surface of the DIUTHAME by capillary action and then allowed to dry for 5-10 min.

| Mass spectrometer calibration
Before analysis, the calibration of the FTICR mass spectrometer was carried out with sodium formate (1 mg/mL in 50% MeOH) cluster ions. The sodium formate solution was externally injected directly into the ESI source, and the data were acquired in the mass range m/z 500-1200. The standard deviations (SDs) were calculated based on five measurements under identical experimental conditions where the resultant mass errors were less than 1 ppm ( Figure S1 and Table S1, The applied function of the calibration used for converting the ion cyclotron frequencies to m/z values is represented as Equation 1, where A and B are the calibration constants and f is the calculated frequency: This calibration equation was initially derived by Ledford et al. 49 According to this derivation, the second term, B/f 2 , denotes the field of direct current trapping and the effect of space charge.

| Data acquisition and analysis
Mass spectra were acquired in positive ion mode using a SolariX XR  Table S2 (supporting information). The laser power for DIUTHAME-MS was

| Endogenous PC(36:4) analysis in brain tissue
To verify the reliability of the highly reproducible data for DIUTHAME-MS, we examined the mass accuracy and intensity of an endogenous PC(36:4), which was used against the PC(18:2/18:2) standard in mouse brain tissue sections at m/z 782.569. The mass errors of DIUTHAME-MS and MALDI-MS were −3.961 ppm ( Figure 3A) and −4.472 ppm ( Figure 3B), respectively. The mass accuracy and intensity values for endogenous PC(36:4) were significantly (p < 0.05) higher with a smaller SD for DIUTHAME-MS than for MALDI-MS ( Figures 3C and 3D). These results showed that DIUTHAME-MS also provides higher reproducibility in mass accuracy and intensity measurements in mouse brain tissue sections than MALDI-MS. and 798.541) showed significantly (p < 0.05) higher intensities in DIUTHAME-MS than in MALDI-MS ( Figure 4D). Here, DIUTHAME-MS also showed a smaller SD in intensity measurements for all selected ions than MALDI-MS. These results indicated that DIUTHAME-MS offers higher reproducibility in mass accuracy and intensity measurements than the DHB matrix for mouse brain tissue lipids.

| Analyses of some abundant lipids in brain tissue
Next, we showed the distribution of mass accuracy and intensity values for the selected lipid ions in the mouse brain section using dot plots. Interestingly, DIUTHAME-MS showed smaller fluctuations in This is consistent with our previous report. 48  In DIUTHAME experiments, the purpose of using thick sections (30 μm) is to ensure sufficient capillary action and surface adhesion between the tissue section and the porous alumina membrane. If the thickness is less than 20 μm, the gap between the tissue section and the DIUTHAME membrane might remain after thawing. A thick section is less likely to suffer from incomplete surface adhesion than a thinner section. 47 The DIUTHAME chip is currently being developed to minimize the gap between tissue section and membrane after mounting, ensuring proper adhesion for a section thickness of 10 μm. DIUTHAME-MS showed higher intensity signals in the analysis of mouse brain tissue sections than MALDI-MS ( Figures 3D and 4D, respectively). However, the liquid sample analysis showed the opposite results ( Figure 2D). This could be due to there being insufficient sample retention on the top surface of the DIUTHAME membrane as precipitation occurs under the membrane. A high volume or concentrated liquid sample addition onto the DIUTHAME chip may overcome this problem.

| CONCLUSIONS
Our results demonstrated that DIUTHAME offers high reproducibility in terms of mass accuracy and intensity measurements in both liquid and mouse brain section analyses compared with MALDI-MS. It can potentially be used instead of the conventional MALDI-MS and MSI analyses to achieve highly reproducible data. Due to its high mass accuracy and intensity with high reproducibility features, DIUTHAME can be used to identify unknown compounds in non-targeted MS and MSI analyses by limiting the number of possible formulae for a given ion.

PEER REVIEW
The peer review history for this article is available at https://publons. com/publon/10.1002/rcm.9076.