Toward Omics-Scale Quantitative Mass Spectrometry Imaging of Lipids in Brain Tissue Using a Multiclass Internal Standard Mixture

Mass spectrometry imaging (MSI) has accelerated our understanding of lipid metabolism and spatial distribution in tissues and cells. However, few MSI studies have approached lipid imaging quantitatively and those that have focused on a single lipid class. We overcome this limitation by using a multiclass internal standard (IS) mixture sprayed homogeneously over the tissue surface with concentrations that reflect those of endogenous lipids. This enabled quantitative MSI (Q-MSI) of 13 lipid classes and subclasses representing almost 200 sum-composition lipid species using both MALDI (negative ion mode) and MALDI-2 (positive ion mode) and pixel-wise normalization of each lipid species in a manner analogous to that widely used in shotgun lipidomics. The Q-MSI approach covered 3 orders of magnitude in dynamic range (lipid concentrations reported in pmol/mm2) and revealed subtle changes in distribution compared to data without normalization. The robustness of the method was evaluated by repeating experiments in two laboratories using both timsTOF and Orbitrap mass spectrometers with an ∼4-fold difference in mass resolution power. There was a strong overall correlation in the Q-MSI results obtained by using the two approaches. Outliers were mostly rationalized by isobaric interferences or the higher sensitivity of one instrument for a particular lipid species. These data provide insight into how the mass resolving power can affect Q-MSI data. This approach opens up the possibility of performing large-scale Q-MSI studies across numerous lipid classes and subclasses and revealing how absolute lipid concentrations vary throughout and between biological tissues.


Method for haematoxylin and eosin staining
At UOW, post-MALDI imaged tissue sections were stained for histological and anatomical features by the haematoxylin and eosin (H&E) staining method.MALDI matrix was removed by immersion of slides in 100% methanol for 30 sec.Tissues were rehydrated by a series of graded ethanol washes of 95% ethanol (aq.) and 70% ethanol (aq.) and deionized water for 2 min each.Slides were stained by haematoxylin for 3 min, blued by rinsing in running tap water until clear and a 1 min wash in distilled water.Tissues were then stained in eosin for 30 sec, washed in two changes of 95% ethanol and once in 100% ethanol for 1 min each and transferred to xylene.Glass coverslip were placed onto samples by Quick hardening mounting medium.
Digital optical scans of H&E-stained section were acquired at 10x magnification on a Falcon SP8 Confocal microscope (Leica Systems, Germany).
At UM, H&E staining was performed on the same 12-µm sections used for MALDI-MSI experiments.The residual matrix was removed by submerging the slides in 70% ethanol for 3 minutes.After a brief wash in Mill-Q water, the slide was re-submerged in 70% ethanol for 3 minutes, followed by Milli-Q water for 3 minutes.The slides were stained in hematoxylin for 3 minutes, followed by rinsing under running tap water for 3 minutes, then placed in eosin for 10 seconds.After placing the slides under running tap water for 1 minute, they were left in 100% ethanol for 1 minute followed by xylene for 30 seconds.The stained sections were mounted with Entellan and covered with a glass coverslip, then dried at room temperature.The stained sections were scanned with a digital scanner (Aperio CS2) at 20x magnification.
For both UM and UOW, high-resolution digital optical H&E images of tissue sections were uploaded onto a custom annotation portal for the annotation of key brain features namely prefrontal cortex, midbrain, hindbrain, basal ganglia and cerebellum.These features were co-registered to MSI data (.imzML) for extraction of regions of interest (ROI) for downstream analyses of lipid profile distributions.

Method for LC-MRM analyses of PE and PC lipid species
Orthogonal analyses of PE and PC species in crude brain extracts was performed in a separate cohort of wild type (WT) C57BL/6N mice aged to 24 weeks at Merck & Co., Inc following inlife conditions described above for housing conditions and isolation of brain specimens.
Immediately prior to LC-MS/MS analyses, fresh frozen hemi brain sections were thawed on ice and homogenized in MeOH:H2O (1:1, v/v) buffer with volumes normalized across samples by wet tissue weight to a final value of 0.1 g/mL.Automated addition of stable isotopically labeled standards (PC 15:0_18:1d7 and PE 15:0_18:1d7, final concentrations of 10 g/mL; Avanti Polar Lipids, Alabaster, AL) and lipid extraction was performed in 96 well plate format using a Microlab Nimbus workstation (Hamilton, Reno, Nevada) following methods described in Zhang et al, J. Lipid.Res, 2022.Briefly, lipid extraction of 100 uL of brain homogenate was performed using chloroform/MeOH + 0.1% butylated hydroxytoluene at 2:1.The plate was mixed vigorously for 30 min, and phase separation was performed by centrifugation (Sorvall Legend centrifuge, Kendro Laboratory, Germany) at 2500 rpm for 10 min at 15°C.The lower phase was transferred to a separate plate and dried under nitrogen gas at room temperature.To the remaining upper aqueous phase, 360 μl of chloroform was added and repeated as above.
After centrifugation, the lower organic phase was pooled with the previous organic fraction.
Targeted quantification of endogenous PE and PC species by LC-MRM analyses was performed as described previously by Zhang and colleagues.Briefly, analyses utilized a duo channel UPLC system (Thermo Scientific Waltham, MA) coupled to a Sciex (Framingham, MA) 6500 triple quadrupole using electrospray ionization operating in negative mode. 1 Separation was achieved using a HALO hydrophilic interaction chromatography (HILIC) (90Å, 2.7 µm, 4.6 mm X 150 mm, Advanced Materials Technology) with mobile phases A: water containing 15 mM ammonium acetate, B: 98.5/1.5 ACN/water containing 15 mM ammonium acetate.Targeted peak area integration was performed manually using MultiQuant (Version 3.03, Sciex).                .Each data point is the average of n=3 biological replicates measured on each system.The majority of outliers can be explained by isobaric overlap encountered in the lower resolution Q-TOF data which adds additional peak intensity in the extracted mass windows (see methods), or by the increased sensitivity of one system for a given lipid class, particularly for species that appear at low intensity that is close to the noise level in one or both systems.

Figure S1 .
Figure S1.Extracted lipid species from analysis of mouse brain tissue on an Orbitrap Elite system....

Figure S2 .
Figure S2.Extracted lipid species from analysis of mouse brain tissue on a timsTOF Flex system......

Figure S1 .
Figure S1.Extracted lipid species from analysis of mouse brain tissue on an Orbitrap Elite system.Reference IS peak shown in red and endogenous lipid species shown in blue; (a) MALDI-MSI in negative mode: [M-H] -ions for (i) PE-O and (ii) PA.(b) Laser post ionization MALDI-2 MSI in positive mode: [M+H] + ions for (i) PE-O, (ii) PC-O, (iii) LPC and [M+Na] + ions for PC.

Figure S3 .
Figure S3.Orbitrap Elite unprocessed averaged on-tissue spectra.(a) Negative mode: Left panel; offtissue showing norharmane matrix ion peak and internal standard peaks for LPE, PA, PE, SHexCer, PG, PS and PI sub-classes detected as [M-H] -ions; Right panel, on-tissue averaged spectra from a whole brain sagittal section with a variety of endogenous lipid species labelled.(b) Positive mode: Left panel, off-tissue showing DHB matrix ion peak and internal standard peaks for LPE, LPC, PE, HexCer, SM, PC and Hex2Cer as [M+H] + ions with PC also shown as [M+Na] + ion; Right panel on-tissue averaged spectra from a whole brain sagittal section with a variety of endogenous lipid species labelled.

Figure S4 .
Figure S4.Histogram showing the ratio of lipid signal intensities to their corresponding internal standard in the averaged on-tissue spectrum from the Orbitrap Elite data.Note the phosphatidylethanolamine (PE) species; PE, PE-O, LPE and LPE-O species correspond to those detected as [M-H] -ions.

Figure S5 .
Figure S5.Influence of IS spraying method on MSI data.Different lipids species detected in (a) negative ion mode and (b) positive ion mode using regular MALDI.For each lipid species the sample prepared by spraying only matrix is on the left, the ion image after spraying with methanol followed by matrix is shown in the centre and the ion image sprayed with IS mixture follows by matrix shown on the right.Intensities for each lipid species were selected using an m/z window of ±12.0 ppm compared to the theoretical m/z of the lipid species.(a) Upper panel timsTOF MALDI-MSI negative ion mode detected as [M-H] -; PI 38:4, PE 38:4, LPI 18:0 and SHexCer 42:2;O2.(b) Lower panel timsTOF MALDI MSI positive ion mode detected as [M+H] + , PC 32:0.SM 36:1;O2,, PE 38:4, and LPC 16:0.All images are TIC normalised.

Figure S6 .
Figure S6.Influence of IS spraying method on acquired mass spectra.Averaged mass spectra acquired using the acquired using the timsTOF in (a) negative ion mode and (b) positive ion mode following (top) matrix application, (middle) spraying with methanol followed by matrix application and (bottom) spraying with IS mix followed by matrix.

Figure S9 .
Figure S9.Reproducibility of quantitative mass spectrometry imaging.Figure shows PE 40:6 and myelin-rich HexCer 42:2;O2 species measured in positive-ion mode using MALDI-2.Data is generated from 3 biological replicates.IS normalized m/z images have significantly less stripes or line streaks in comparison to the original m/z images due to the correction of MALDI-2 artefacts (refer to main text).The right hand panel shows the 3 ppm selection window used for each species.

Figure S10 .
Figure S10.Representative internal standard normalized ion images acquired using the timsTOF.Figure shows different lipids species detected in (a) negative ion mode using MALDI and (b) positive ion mode using MALDI-2.For each lipid species the original ion image is shown on the left, the classspecific internal standard ion image is shown in the centre and the IS normalized ion image is shown on the right.Intensities for each lipid species were selected using an m/z window of ±12.0 ppm compared to the theoretical m/z of the lipid species (a) Left panel timsTOF MALDI-MSI negative ion mode detected as [M-H] -; PE 38:4, PG 44:12, PI 38:4, PS 36:2 and SHexCer 42:2;O2.(b) Right panel timsTOF MALDI-2 MSI positive ion mode detected as [M+H] + , PE 38:4, SM 36:1;O2, HexCer 42:2;O2, HexCer 42:2;O2 and PC 32:0.

Figure S14 .
Figure S14.Violin plot comparing PC to PE ratio obtained by MALDI-MSI vs LC-MS/MS.Total PC and PE were obtained by summing up all class-specific species from respective analyses.PC to PE ratio was determined by dividing the total class levels.For MALDI-MSI, individual points represent a single measurement from a particular region, and for LC-MS/MS, measurements from extracted brainhomogenates of ten individual wild-type mice aged 24 weeks.

Figure S15 .
Figure S15.Correlation of Q-MSI data acquired using the Orbitrap and timsTOF following averaging of all on-tissue pixels for each section.Data is provided for (a) PC-O, [M+H] + , (b) SM, [M+H] + (c) Cer, [M+H] + and (d) PE-O, [M-H] -, (e) LPC, [M+H] + , (f) LPC, [M-H] -and (g) Hex2Cer, [M+H] +.Each data point is the average of n=3 biological replicates measured on each system.The majority of outliers can be explained by isobaric overlap encountered in the lower resolution Q-TOF data which adds additional peak intensity in the extracted mass windows (see methods), or by the increased sensitivity of one system for a given lipid class, particularly for species that appear at low intensity that is close to the noise level in one or both systems.

Table S2 .
Ions used for mass recalibration EN-Endogenous Lipid, IS-Internal Standard

Table S3 .
List of adducts used for Q-MSI of PC species.Adducts chosen to reduce isobaric interferences.The corresponding normalization IS mass is shown for each adduct type.ppmparts per million.