Low-melting point agarose as embedding medium for maldi mass spectrometry imaging and laser-capture microdissection-based proteomics

ABSTRACT The combination of MALDI mass spectrometry imaging, laser-capture microdissection, and quantitative proteomics allows the identification and characterization of molecularly distinct

tissue compartments. Such workflows are typically performed using consecutive tissue sections, and so reliable sectioning and mounting of high-quality tissue sections is a prerequisite of

such investigations. Embedding media facilitate the sectioning process but can introduce contaminants which may adversely affect either the mass spectrometry imaging or proteomics analyses.

Seven low-temperature embedding media were tested in terms of embedding temperature and cutting performance. The two media that provided the best results (5% gelatin and 2% low-melting point

agarose) were compared with non-embedded tissue by both MALDI mass spectrometry imaging of lipids and laser-capture microdissection followed by bottom-up proteomics. Two out of the seven

tested media (5% gelatin and 2% low-melting point agarose) provided the best performances on terms of mechanical properties. These media allowed for low-temperature embedding and for the

collection of high-quality consecutive sections. Comparisons with non-embedded tissues revealed that both embedding media had no discernable effect on proteomics analysis; 5% gelatin showed

a light ion suppression effect in the MALDI mass spectrometry imaging experiments, 2% agarose performed similarly to the non-embedded tissue. 2% low-melting point agarose is proposed for

tissue embedding in experiments involving MALDI mass spectrometry imaging of lipids and laser-capture microdissection, proteomics of consecutive tissue sections. SIMILAR CONTENT BEING VIEWED

BY OTHERS MORPHOLOGICAL AND MOLECULAR PRESERVATION THROUGH UNIVERSAL PREPARATION OF FRESH-FROZEN TISSUE SAMPLES FOR MULTIMODAL IMAGING WORKFLOWS Article 28 May 2024 SPATIAL PROTEOMICS OF

SINGLE CELLS AND ORGANELLES ON TISSUE SLIDES USING FILTER-AIDED EXPANSION PROTEOMICS Article Open access 30 October 2024 SPATIALLY RESOLVED ANALYSIS OF FFPE TISSUE PROTEOMES BY QUANTITATIVE

MASS SPECTROMETRY Article 31 July 2020 INTRODUCTION MALDI mass spectrometry imaging (MSI) is a technique that allows for spatial mapping of various analytes including lipids1, peptides2,

metabolites3 and glycans4. The technique is often applied to biological tissues to investigate the molecular composition of heterogeneous specimens5 or to perform a segmentation of the

tissue to highlight characteristics that are not detectable by histology alone6. In order to increase proteome coverage tissue regions can be excised and isolated by laser-capture

microdissection (LCM) and then analyzed with proteomics workflows7. In the context of the MALDI MSI-guided microdissection, highly reproducible sectioning across consecutive sections is

essential to ensure that regions identified by MALDI MSI can be reliably isolated from subsequent tissue sections (i.e. the sectioning and mounting preserves the longitudinal make-up of the

sample). Formalin-fixed paraffin-embedded (FFPE) tissue samples are the preferred format for histopathological analysis8 because of the high quality sections that can be routinely obtained

(essential for many established diagnostic pathways), as well as their compatibility with immunohistochemical molecular assays and the fact that the tissues may be stored at room

temperature. FFPE tissue blocks are routinely used for MALDI–MSI of proteolytic peptides and enzyme-released N-linked glycans9,10, but are less suited to the analysis of lipids and

metabolites because the tissue processing leads to the loss of some lipids/metabolites, and the long processing times lead to post-mortem changes in the molecular composition of the tissue

metabolome. Note that some lipids and metabolites have been detected from FFPE tissue sections, and which may also have diagnostic potential11,12, but the loss of metabolites and post-mortem

changes make it less straightforward to link the metabolic signatures to the tissue’s underlying biology. For these reasons fresh-frozen tissue samples are recommended for MALDI-MSI of

metabolites and lipids, but the fragility of fresh-frozen tissue sections makes the sectioning more challenging, especially for high fat content tissues. The embedding of fresh frozen tissue

within an inert matrix provides structural support and thereby aids sectioning. The most common embedding medium is optimal cutting temperature (OCT) compound, a polymer that has excellent

thermal properties and produces a dense embedding that allows for easy sectioning over a wide range of temperatures. OCT is widely used in immunohistochemistry, histology, and

immunofluorescence. However, the high ionization efficiency of its polymeric components cause heavy ion suppression in MALDI MSI13,14. Protocols have been reported for the removal of OCT

prior to protein15 and lipid16 MALDI-MSI analysis, and alternative embedding media have been developed. The most desirable characteristic of an embedding medium is that it should provide

structural stability to the sample without interfering with the analysis, i.e. providing the same result as non-embedded samples. Nelson et al_._17 conducted an extensive investigation of

the compatibility of many embedding media with MALDI MSI, to embed fragile zebrafish samples, selecting the one that provided the best cutting performance and gave no interference in MSI

analysis. However, the comparison was limited to only MALDI MSI. Dannhorn et al_._18 investigated the properties of several embedding media and their compatibility with MSI techniques

(MALDI-MSI, desorption-electrospray ionization (DESI), secondary-ion mass spectrometry (SIMS), and imaging mass cytometry (IMC)), as well as with immunohistochemistry, RNAseq, and genomics

analysis, providing an excellent starting point for multimodal studies. Nonetheless, the evaluation of the media compatibility with proteomics analysis was not performed. Quanico et al.19

focused on the integrated MALDI MSI, lipidomics and proteomics analysis from the same tissue section. They were able to perform both MALDI MSI and proteomics from 2% gelatin and 2%

carboxymethyl cellulose (CMC) embedded samples, but they did not focus on the comparison of the two media nor performed a comparison with non-embedded samples. The ideal embedding medium is

a material that does not affect the analysis, providing similar results to that obtained from non-embedded samples. All the previous examples, despite providing extensive characterization of

embedding media, did not benchmark the results to non-embedded tissues. Such benchmarking is essential in order to perform a cost-vs-benefit assessment, and thereby select the experimental

procedure best suited to the goals of the experiment. Here we report an investigation of low-temperature embedding media in terms of their mechanical properties and embedding temperature,

with the goal of determining a reliable embedding medium for the combination of MALDI MSI of lipids, laser-capture microdissection of tissue regions, and quantitative proteomics. We compared

embedded and non-embedded tissues to ensure that the embedding media did not interfere with the analysis. 2% low-melting point agarose and 5% gelatin-embedded were selected as optimal

embedding media, and enabled the collection of high-quality consecutive sections. While 5% gelatin gave a light ion-suppression effect in MALDI-MSI experiment, 2% low-melting point agarose

did not interfere with neither MALDI MSI nor the quantitative proteomics experiments, producing results similar to the non-embedded samples. 2% low-melting point agarose is thus proposed as

embedding medium for MALDI MSI of lipids and proteomics of microdissected tissue samples from consecutive tissue sections. RESULTS AND DISCUSSION An optimal embedding medium should enable

the production of high-quality consecutive sections, allowing for the artifact-free spatially resolved analysis of fragile samples and the isolation of these regions from consecutive

sections. Previously reported embedding media for MALDI MSI include carboxymethyl cellulose (CMC)19,26, gelatin17,19, agarose17, HPMC-PVP (hydroxypropyl methylcellulose and

polyvinylpyrrolidone mixture)18, and poly[N-(2-hydroxypropyl)methacrylamide] (pHPMA)27. Here we focused on CMC, gelatin and low- melting point agarose since these media are relatively

inexpensive and have previously been reported to provide excellent characteristics (structural support and low chemical interference) for MALDI MSI investigations. For instance, a

combination of 2% CMC and 10% gelatin was selected by Nelson et al_._ as the optimal embedding medium for the sectioning of fragile zebrafish embryo samples17. Low melting point agarose is a

type of agarose used for nucleic acid electrophoresis, with a gel temperature a few degrees above room temperature28,29. It was included in the study as a protein-free embedding medium.

Here we evaluated seven combinations of the embedding media in terms of embedding temperature, cutting performance, compatibility with MALDI MSI of lipids, LCM, and quantitative proteomics.

EMBEDDING TEMPERATURE AND CUTTING PERFORMANCE EVALUATION It is known that freeze–thaw cycles can adversely affect MALDI-MSI and proteomics analysis of tissue samples30. To minimize the

number of freeze–thaw cycles experienced by the tissue samples it would be better to perform the sample embedding-and-freezing directly after tissue collection, however such a process would

extend the post-mortem time at room temperature and is not always logistically practical or possible. In these cases, an embedding medium with low embedding temperature is recommended to

minimize sample thawing, especially for small samples. The embedding temperature was evaluated for the embedding media listed below, together with their cutting performance: * 2% CMC * 2%

CMC + 5% gelatin * 2% CMC + 10% gelatin * 5% CMC + 1% low melting point agarose * 5% gelatin * 1% low melting point agarose * 2% low melting point agarose These tests were performed by

embedding chicken breast tissue in the selected media. The embedding temperature was measured with a digital thermometer and defined as the temperature of medium gelatinization. Embedded

samples were then sectioned to test their cutting properties. The embedded blocks were equilibrated at – 20 °C in the cryostat and 5, 8 and 12 μm thick sections were collected. All embedding

media tested allowed the collection of consecutive tissue sections, with little difference between the conditions. Denser media (e.g. 2% CMC + 5% gelatin and 2% CMC + 10% gelatin) performed

slightly better during cutting, producing more solid sections that were easier to manipulate and to mount on the glass microscope slides. A summary of these results is provided in Table 1.

2% CMC showed good cutting performance, allowing the production of consecutive tissue sections. When the embedding temperature was tested, the medium appeared to freeze instead of

gelatinizing at temperatures lower than 3 °C, remaining liquid at 4 °C. A similar behavior was previously reported by Quanico et al., who noted that 2% CMC liquefies after thaw mounting

sections of CMC-embedded tissues on parafilm-coated glass slides19. This property is not ideal for the preparation of precast gels31, or for the fabrication of two-pieces gel molds for the

embedding of small samples32. Despite good cutting performance, the embedding temperature of the other CMC-containing media was in general higher than the other tested media. The higher gel

temperature increased the embedding time and increased the risk of tissue degradation. Moreover, CMC was more difficult to solubilize and significantly increased the preparation time. In

contrast 5% gelatin and 2% low melting point agarose were easy to solubilize and had low gelatinization temperatures, ensuring rapid sample freezing and minimal tissue degradation. The

low-melting/freezing temperature of agarose meant that it remained liquid at room temperature, therefore allowing lower embedding temperatures. The cutting performance of 5% gelatin and

agarose-based media were good and allowed for the collection of consecutive tissue sections. The 5% gelatin and 2% low-melting point agarose were selected for further investigation, with

regards to their compatibility with mass spectrometry analysis. MALDI MSI OF LIPIDS OF EMBEDDED TISSUES An optimal embedding medium for MALDI MSI should not interfere with the molecular

signals obtained from the tissue, both in terms of distribution and detection sensitivity. To test media compatibility with MALDI MSI of lipids, mouse liver samples were embedded with 5%

gelatin and 2% low-melting point agarose, analyzed, and compared with non-embedded mouse liver tissue. Fractional mass plots were obtained for the three conditions to highlight the presence

of background signals arising from the embedded tissues. The comparison between MALDI MSI data of non-embedded samples, 5% gelatin, and 2% low-melting point agarose embedding media is

summarized in Fig. 1. Both media provided good quality tissue blocks that allowed the collection of consecutive sections and proved to be MALDI-MSI compatible, with no traces of

contamination in the MALDI mass spectra. The fractional mass plot of the average mass spectrum for both embedding media (Fig. 1A) showed a large number of molecular ions with masses between

700 and 800 _m/z_ on the integer axis and 0.3–0.7 _m/z_ on the fractional mass axis, corresponding to lipid signals33. The comparison of fractional mass plots showed that samples embedded

with 5% gelatin produced fewer lipid peaks compared to both agarose and the non-embedded samples. Molecular ion images obtained from the agarose-embedded tissue samples exhibited higher

background signals outside the tissue section compared to the gelatin-embedded and non-embedded tissue (Fig. 1B,C) but the fine lobular structure of the liver observable in the non-embedded

tissue was also still recognizable for both agarose-based embedding media (Fig. 1C). Signal arising from the background is likely to come from the embedding, which was not removed prior to

MSI analysis. The retention of the fine structure in the MSI images indicates that the background signal is not due to delocalization of the lipid signals into the embedding media. Both 5%

gelatin and 2% agarose embedding media allowed high quality sections to be obtained, from which good quality MALDI MSI datasets could be recorded. Despite the higher background outside the

tissue obtained with the 2% agarose embedding media, the MALDI-MSI data better matched that obtained from the non-embedded tissue in terms of the appearance of the fractional mass plot, and

the lobular structure of the liver in the MALDI-MSI images. EVALUATION OF LCM-BOTTOM-UP MICROPROTEOMICS PERFORMANCE OF EMBEDDED TISSUES The compatibility of the gelatin and agarose embedding

media with proteomics of microdissected tissue samples was tested by processing small 2 mm2 microdissected tissue samples with the SP321 proteomics protocol. Figure 2 shows a comparison of

the number of identified protein groups, identified proteins, identified peptide groups, peptide-spectrum matches (PSMs), and acquired mass spectra acquired, as well as the identification

rate for the three embedding conditions. PSMs are the MS2 spectra that were assigned to a peptide, including redundancies. Identification rate was calculated as the ratio between PSMs and

total number of MS2 spectra. The percentage of PSMs with zero, one and two missed cleavages sites and the charge distribution of PSMs was also compared to evaluate the digestion efficiency

(Supplementary Information S1). A one-way ANOVA analysis of the proteomics digestion quality parameters showed no significant difference between the three embedding conditions. The number of

proteins groups, proteins, peptides and PSMs identified from 5% gelatin and 2% agarose embedded tissue did not significantly differ from non-embedded tissue. The number of acquired MS/MS

spectra and the identification rate was also not significantly different, demonstrating that the embedding did not introduce any significant difference in the number of detected peptides,

even for the 5% gelatin embedding medium. Type I and type II porcine collagen, which are expected to be major components of porcine skin gelatin, were not identified in any of the samples,

confirming the absence of gelatin contamination from the embedding medium. The percentage of PSMs with zero, one and two missed cleavage sites and the charge state distribution were

identical for the three different conditions, suggesting that the embedding had no effect on digestion efficiency. The effect of the embedding medium on protein quantification was also

investigated. A table containing the quantification of all proteins in the samples is provided in Supplementary Information S2. Figure 3 shows multiscatter plots depicting the pairwise

comparison of the three embedding conditions. The upper right half of the correlation matrix shows the average Pearson correlation coefficient. The average Pearson correlation coefficient of

the protein intensities revealed that the protein intensities obtained from the 2% agarose embedded tissue were a closer match to the non-embedded tissue than the gelatin embedded tissue

(0.704 vs 0.680 respectively). DISCUSSION The central tenet behind multimodal imaging _using consecutive tissue sections_ is the conservation of the biologically relevant spatio-molecular

information between consecutive tissue sections. Nonetheless, the sectioning of fragile and heterogeneous tissues can complicate such multimodal approaches, since the collection of

superimposable sections can be challenging. Embedding media help to overcome this disadvantage by facilitating the collection of reproducible consecutive sections. OCT is the most widely

used embedding medium for fresh-frozen samples. However, when samples are to be analysed by mass spectrometry, OCT is not recommended, since it is acts as a contaminant causing ion

suppression13,14. This is particularly apt for the SP3 methodology, which is based on the adsorption of proteins and peptides onto bead surfaces (OCT is based on polydiethylene glycol, a

surfactant, which can interfere with this adsorption). Tissue washes have been developed to reduce this contamination15,16, and alternative embedding media have been developed to enable

MALDI-MSI based multimodal approaches. Nonetheless, a complete characterization of embedded media which included a comparison with non-embedded tissue was missing. Here we focused on the use

of combinations of CMC, gelatin and low-melting point agarose as mass spectrometry-compatible media. The embedding temperature, the cutting performance, and the compatibility with MALDI MSI

of lipids and LCM-proteomics were evaluated. Among all the embedding media tested, 5% gelatin and 2% low melting point agarose allowed for low-temperature embedding, optimal cutting

performances and mass spectrometry compatibility compared to non-embedded sample. In fact, no contamination was observed in the MALDI MSI of lipids experiment. The comparison of the

fractional mass plots of MSI data of embedded samples with non-embedded tissue suggested a potential ion-suppression effect of gelatin embedding. Instead, the number of lipids identified in

the 2% agarose-embedded sample was comparable with the non-embedded tissue. Another potential advantage of the agarose-based embedding medium over the gelatin-based medium is the possibility

of performing on-tissue tryptic digestion for MALDI MSI of peptides. The 2% agarose medium produced a higher background outside the tissue but this phenomenon had no effect on the ion

distribution recorded inside the tissue. In fact, the absence of ion delocalization is confirmed by the retention of the spatial features of the sample (in our case, the fine lobular

structure of the liver). The signal surrounding the tissue is likely to be produced by the ionization of the agarose embedding, which was not removed before MSI data acquisition. Hematoxylin

and eosin H&E staining of consecutive tissue sections showed that the agarose-based embedding remained attached to the slide during staining. Nevertheless, the background of the 2%

agarose embedding in the H&E staining was negligible (Supplemental Fig. S1). Both gelatin and agarose were found to be compatible with laser-capture microdissection and proteomics and

the embedding medium did not influence the number of identified proteins nor the quality of protein digestion. The protein intensities obtained from tissues embedded in 2% low melting point

agarose had a slightly higher correlation coefficient with those obtained from not embedded tissue, compared to 5% gelatin. Based on these results, both 5% gelatin and 2% low-melting point

agarose were suitable embedding media for LCM-proteomics experiments. Nonetheless, low-melting point agarose outperformed gelatin in MALDI MSI analysis of lipids. 2% low-melting point

agarose showed good performances for the embedding of frozen biological samples for multimodal experiment involving MALDI MSI of lipids and proteomics of excised regions from consecutive

tissue sections. CONCLUSIONS Mass spectrometry-based multimodal spatial experiments are based on the assumption of the conservation of the spatial details between consecutive sections. This

assumption is no longer valid if the reproducibility across the section collection is not maintained, due to the introduction of distortion during tissue sectioning and mounting. Embedding

media are crucial to ensure proper section collection and to collect reproducible sections. An ideal embedding medium should provide structural stability to the tissue and should not

interfere with the molecular analysis i.e. producing the same results as the non-embedded tissue. Thus benchmarking with non-embedded tissue is crucial for the evaluation of embedding media.

While several MALDI MSI-compatible embedding media have been proposed17,18, only 2% CMC and 2% gelatin were also tested for proteomics analysis19, but which were not benchmarked against

non-embedded tissue. Here we reported the use of 5% gelatin and 2% low-melting point agarose for the embedding of fresh-frozen samples, MALDI MSI of lipids, microdissection, and quantitative

proteomics. The comparison of the results obtained from the embedded tissues with the non-embedded samples demonstrated little adverse effects of the embedding process, limited to just

light ion-suppression effect of gelatin in MALDI-MSI experiments. 2% low-melting point agarose is thus proposed as embedding medium for the acquisition of MALDI MSI of lipids and proteomics

of microdissected tissue samples from consecutive tissue sections. MATERIALS AND METHODS MATERIALS CMC (C9481-500G, carboxymethylcellulose sodium, medium viscosity) and gelatin

(48723-500G-F) were purchased from Sigma-Aldrich (St. Gallen, Switzerland); QA-Agarose low melting point was obtained from MP Biomedical, Santa Ana, CA, USA. Disposable PVC plastic embedding

moulds (15 × 15 × 5 mm) were obtained from Polysciences (Warrington, PA, USA). HPLC grade water and DMSO (dimethyl sulfoxide) were purchased from Thermo Fisher Scientific (Dreieich, Hessen,

Germany). Histology grade ethanol was purchased from DiaPath S.p.A. (Martinengo, Italy). Harris hematoxylin was purchased from VWR International LLC (Milan, Italy), cOmplete Protease

Inhibitor Cocktail was purchased from La Roche (Basel, Switzerland), and Trypsin/Lys-C mix Mass Spec grade was obtained from Promega (Milan, Italy). ITO (indium tin oxide) coated glass

slides for MALDI were purchased from Bruker Daltonics (Billerica, MA, USA), and the polyethylene naphthalate (PEN) slides and Adhesive Cap 200 were purchased from Zeiss (Oberkochen,

Germany). Speedbeads A (GE65152105050250), Speedbeads B (GE65152105050250), 1M HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) solution, EDTA (ethylenediaminetetraacetic acid),

EGTA (ethylene glycol-bis(β-aminoethyl ether)-_N,N,N′,N′-_tetraacetic acid), sodium dodecyl sulphate (SDS), trifluoroethanol (TFE), dithiothreitol (DTT), iodoacetamide (IAA), formic acid,

acetonitrile (ACN) and HPLC grade ethanol were purchased from Sigma-Aldrich (Milan, Italy). METHODS SAMPLE EMBEDDING Fresh frozen mouse liver (male C57BL/6, adult) was kindly provided by M.

Caleo at the Institute of Neuroscience of the CNR in Pisa and stored at – 80 °C until use. Chicken breast was purchased from a local supermarket, cut into 1 cm × 1 cm × 0.5 cm pieces and

stored at – 80 °C until use. Embedding solutions were prepared by dissolving the powdered compound in HPLC grade water and heating the resulting solution to 95 °C. A dry-ice ethanol bath was

prepared, and the lower part of the embedding mold was submerged into the bath. Embedding media were cooled down to a temperature just above their hardening temperature. A small piece of

frozen tissue was then placed in the mold and the embedding medium carefully added with a plastic pipette, avoiding the formation of bubbles. After the medium had solidified the mold was

transferred on dry ice, wrapped in aluminum foil and stored at -80 °C. Tissue blocks were sectioned using a Leica CM1950 cryostat (Leica, Wetzlar, Germany). Tissue sections of 12 μm

thickness were mounted onto indium-tin-oxide (ITO) conductive glass slides for MALDI MSI experiments. 12 μm thick sections were also collected on polyethylene naphthalate (PEN) membrane

glass slides for microdissection. Non embedded mouse liver tissue was sectioned alongside the embedded samples, to benchmark the results obtained from the embedded samples. The cryostat

blade was cleaned with ethanol after the collection of each tissue section to avoid cross contamination. MALDI MSI Tissue sections mounted on ITO glass slides were thawed under vacuum for 15

min prior to matrix application. One replicate was processed for each condition. The MALDI matrix norharmane (7 mg/mL in CHCl3:MeOH 70:30) was sprayed onto the tissue sections using a

SunCollect automated spraying system (SunChrom, Friedrichsdorf, Germany). X and Y deposition speed were set to Low 3 and Medium 1 respectively while the height of the nozzle was set to 50

mm. The matrix solution was applied in 8 layers (2 layers at 5 µL/min followed by 6 layers at 10 µL/min). The matrix coated tissue section was then dried under vacuum for 15 min prior to

data acquisition. MALDI MSI was performed using an elevated pressure (EP) MALDI source20 (Spectroglyph, LLC., Kennewick, WA, USA) equipped with a 349 nm laser (Spectra-Physics, Santa Clara,

CA, USA), coupled to an Orbitrap Q-Exactive Plus mass spectrometer (Thermo Scientific). The laser was operated at 1.65 A and 500 Hz, the source pressure was 7.2 Torr and the pixel size was

30 × 30 µm. Mass spectra were acquired in the range 150–2000 _m/z_ at 70 000 resolving power at _m/z_ 200. The position file was aligned to the raw MS files using ImageInsight (v.

0.1.0.11550, Spectroglyph, LLC). LASER-CAPTURE MICRODISSECTION AND QUANTITATIVE PROTEOMICS Tissue sections were stained with hematoxylin at 4°C and all solutions used for staining contained

a tablet of protease inhibitor. The LCM system used was an Apotome2 Axio Observer Z 1 microscope equipped with a PALM MicroBeam (Zeiss, Oberkochen, Germany). PEN slides were mounted on the

LCM system and the laser used to isolate around 2 mm2 of homogeneous liver tissue. The dissected tissue pieces were collected using RoboLPC (Robo laser pulse catapulting) mode, which works

by cutting the selected region border and catapulting the dissected tissue piece into the cap of an Adhesive Cap 200 tube with a defocused laser pulse. Each tissue sample was processed in

triplicate using an SP3 proteomics workflow21,22. In brief, rectangular tissue regions were transferred from the adhesive cap to 0.2 mL tubes containing 10 μL of SP3 lysis buffer and 10 μL

TFE. Samples were lysed by sonication and quantified using a modified BCA assay suited for small sample amounts23. Aliquots of 1.95 μg of proteins were prepared and the sample volume was

equalized to 40 μL with 1:1 SP3 lysis buffer:TFE. 2 μL of carboxylate coated paramagnetic beads (100 mg/mL suspension of 50% Speedbeads A and 50% Speedbeads B) were added to each sample.

Proteins were denatured at 95 °C for 5 min, then reduced with the addition of 2 μL of 200 mM DTT and incubated at 45°C for 30 min. Proteins were alkylated with the addition of 2 μL of 400 mM

IAA followed by incubation at room temperature for 30 min after which the reaction was quenched by the addition of 2 μL of 200 mM DTT. The solution was then acidified by the addition of 2

μL of formic acid, and ACN added to a final concentration of 50% to promote protein adsorption to the magnetic beads. The tubes were then placed on a magnet to immobilize the magnetic beads

and the bead-bound proteins were washed with EtOH 70% and ACN 100%. The proteins were eluted from the beads using 10 μL 50mM HEPES pH 8, and with sonication (30 s ON, 30 s OFF, 5 min, 18 °C)

to help protein detachment. The proteins were then digested overnight (18 h, trypsin/Lys-C mix Mass Spec grade, 1:25 enzyme/protein). The resulting proteolytic peptides were then purified:

the ACN concentration was raised to 95% to promote peptide adsorption to the beads, which were then washed with 100% ACN and finally eluted using 20 μL of 2% DMSO aqueous solution. Peptide

samples were diluted 1:1 with 10% formic acid and injected into the EASY-nLC 1000 coupled to an Orbitrap Fusion mass spectrometer (Thermo Scientific, Bremen, Germany). Peptide ions were

analyzed using the Top Speed data-dependent method, with a 3 s cycle; MS1 scans were performed in the Orbitrap (m/z 375 to 1500 at 120 K resolution with an AGC Target 5 × 105 and 100 ms

maximum injection time) and MS2 scans were acquired in the ion trap using a 1.6 _m/z_ isolation window, 30% HCD Collision Energy, and an AGC target of 5 × 103. DATA ANALYSIS MALDI MSI Thermo

RAW file were converted into mzXML format using RawConverter24 and the tool ORBIIMAGEmzXML2Tricks (v.0.10, G. Eijkel) was used to convert the mzXML spectra file and the EP-MALDI XML

position file into a .mat data structure. The MALDI MSI dataset was TIC normalized. To evaluate the number and nature of detected peaks, the fractional mass plot of the _m/z_ list of each

MALDI MSI dataset was plotted. The fractional mass plot is a visualization method based on the decomposition of the ion _m/z_ values into their integer and decimal components. The fractional

mass plot was obtained by plotting the integer part as the _x_ coordinate and the decimal part as the _y_ coordinate. LC–MS/MS LC–MS/MS data were analyzed using the Proteome Discoverer

(v.2.1, Thermo Fisher Scientific, Rockford, IL, USA) and searched against the SwissProt Mus Musculus database (Uniprot, 11 June 2019, 17,021 entries). An in-house contaminant database was

added to the search (253 entries), which included type I (alpha-1 and alpha-2 chains) and type III (alpha-1) porcine collagen. Searches were performed with a precursor mass tolerance of 10

ppm using a strict FDR of 0.01. A maximum of two missed cleavages were allowed. Methionine oxidation (+ 15.995 Da) were set as a dynamic modification and carbamidomethylation of cysteine (+ 

57.021 Da) was set as a static modification. Quantification was performed only on proteins identified by at least one unique peptide; proteins identified from the contaminant database were

discarded. Protein intensities were used as a proxy of protein abundance. The proteomics data were analyzed using Perseus 1.5 software25 and GraphPad Prism v5 for Windows (GraphPad Software,

La Jolla, CA, USA, www.graphpad.com). The number of identified protein groups, proteins, peptide groups, peptide-spectrum-matches (PSM), and mass spectra were retrieved from Proteome

Discoverer software. Identification rate was defined as the percentage ratio of PSM to mass spectra. The distribution of peptide charge and the percentage of missed cleavages were retrieved.

Identification parameters (number of identified protein groups, proteins, peptide groups, peptide-spectrum-matches (PSM), mass spectra, identification rate, percentages of missed cleavages

and charge distribution) were compared using a one-way ANOVA performed with using (α = 0.05). Protein intensities were log2 transformed and the median of protein intensity of each sample was

subtracted from each protein intensity to correct for small differences in the number of peptides loaded on the LC–MS/MS instrument. Perseus software was used to produce a multiscatter plot

(based on Pearson’s correlation coefficient) to compare the protein intensities obtained from the technical replicates and from the different embedding media/control. DATA AVAILABILITY The

MALDI MSI and the proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE34 partner repository with the dataset identifier PXD045349.

(https://proteomecentral.proteomexchange.org/cgi/GetDataset?ID=PXD045349). REFERENCES * Moerman, A. M. _et al._ Lipid signature of advanced human carotid atherosclerosis assessed by mass

spectrometry imaging. _J. Lipid Res._ DEC, jlr.RA120000974 (2020). * Mittal, P. _et al._ Lymph node metastasis of primary endometrial cancers: Associated proteins revealed by MALDI imaging.

_Proteomics_ 16, 1793–1801 (2016). Article  CAS  PubMed  Google Scholar  * Dilillo, M. _et al._ Ultra-high mass resolution MALDI imaging mass spectrometry of proteins and metabolites in a

mouse model of glioblastoma. _Sci. Rep._ 7, 1–11 (2017). Article  CAS  Google Scholar  * Drake, R. R. _et al._ Defining the human kidney N-glycome in normal and cancer tissues using MALDI

imaging mass spectrometry. _J. Mass Spectrom._ 55, e4490 (2020). Article  CAS  PubMed  PubMed Central  Google Scholar  * Greco, F. _et al._ Mass spectrometry imaging as a tool to investigate

region specific lipid alterations in symptomatic human carotid atherosclerotic plaques. _Metabolites_ 11, (2021). * Balluff, B. _et al._ De novo discovery of phenotypic intratumour

heterogeneity using imaging mass spectrometry. _J. Pathol._ 235, 3–13 (2015). Article  CAS  PubMed  Google Scholar  * von Eggeling, F. & Hoffmann, F. Microdissection—an essential

prerequisite for spatial cancer omics. _Proteomics_ 20, 1–10 (2020). Google Scholar  * Bass, B. P., Engel, K. B., Greytak, S. R. & Moore, H. M. A review of preanalytical factors

affecting molecular, protein, and morphological analysis of formalin-fixed, paraffin-embedded (FFPE) tissue: How well do you know your FFPE specimen?. _Arch. Pathol. Lab. Med._ 138,

1520–1530 (2014). Article  PubMed  Google Scholar  * Powers, T. W. _et al._ MALDI imaging mass spectrometry profiling of N-glycans in formalin-fixed paraffin embedded clinical tissue blocks

and tissue microarrays. _PLoS One_ 9, 1–11 (2014). Article  Google Scholar  * Ly, A. _et al._ Site-to-site reproducibility and spatial resolution in MALDI–MSI of peptides from formalin-fixed

paraffin-embedded samples. _Proteomics Clin. Appl._ 13, 1–10 (2019). Article  Google Scholar  * Ly, A. _et al._ High-mass-resolution MALDI mass spectrometry imaging of metabolites from

formalin-fixed paraffin-embedded tissue. _Nat. Protoc._ 11, 1428–1443 (2016). Article  PubMed  Google Scholar  * Denti, V. _et al._ Spatial multiomics of lipids, N-glycans, and tryptic

peptides on a single FFPE tissue section. _J. Proteome Res._ 21, 2798–2809 (2022). Article  CAS  PubMed  PubMed Central  Google Scholar  * Goto-Inoue, N., Hayasaka, T., Zaima, N. &

Setou, M. Imaging mass spectrometry for lipidomics. _Biochim. Biophys. Acta Mol. Cell Biol. Lipids_ 1811, 961–969 (2011). * Goodwin, R. J. A. Sample preparation for mass spectrometry

imaging: Small mistakes can lead to big consequences. _J. Proteomics_ 75, 4893–4911 (2012). Article  CAS  PubMed  Google Scholar  * Weston, L. A. & Hummon, A. B. Comparative LC-MS/MS

analysis of optimal cutting temperature (OCT) compound removal for the study of mammalian proteomes. _Analyst_ 138, 6380–6384 (2013). Article  ADS  CAS  PubMed  Google Scholar  * Truong, J.

X. M. _et al._ Removal of optimal cutting temperature (O.C.T.) compound from embedded tissue for MALDI imaging of lipids. _Anal. Bioanal. Chem._ 413, 2695–2708 (2021). * Nelson, K. A.,

Daniels, G. J., Fournie, J. W. & Hemmer, M. J. Optimization of whole-body zebrafish sectioning methods for mass spectrometry imaging. _J. Biomol. Tech._ 24, 119–127 (2013). Article 

PubMed  PubMed Central  Google Scholar  * Dannhorn, A. _et al._ Universal sample preparation unlocking multimodal molecular tissue imaging. _Anal. Chem._ 92, 11080–11088 (2020). Article  CAS

  PubMed  Google Scholar  * Quanico, J., Franck, J., Wisztorski, M., Salzet, M. & Fournier, I. Integrated mass spectrometry imaging and omics workflows on the same tissue section using

grid-aided, parafilm-assisted microdissection. _Biochim. Biophys. Acta - Gen. Subj._ 1861, 1702–1714 (2017). Article  CAS  PubMed  Google Scholar  * Belov, M. E. _et al._ Design and

performance of a novel interface for combined matrix-assisted laser desorption ionization at elevated pressure and electrospray ionization with orbitrap mass spectrometry. _Anal. Chem._ 89,

7493–7501 (2017). Article  CAS  PubMed  Google Scholar  * Hughes, C. S. _et al._ Ultrasensitive proteome analysis using paramagnetic bead technology. _Mol. Syst. Biol._ 10, 757–757 (2014).

Article  PubMed  Google Scholar  * Pellegrini, D. _et al._ Quantitative microproteomics based characterization of the central and peripheral nervous system of a mouse model of Krabbe

disease. _Mol. Cell. Proteomics_ 29, mcp.RA118.001267 (2019). * De Graaf, E. L., Pellegrini, D. & McDonnell, L. A. Set of novel automated quantitative microproteomics protocols for small

sample amounts and its application to kidney tissue substructures. _J. Proteome Res._ 15, 4722–4730 (2016). Article  PubMed  Google Scholar  * He, L., Diedrich, J., Chu, Y.-Y. & Yates,

J. R. Extracting accurate precursor information for tandem mass spectra by RawConverter. _Anal. Chem._ 87, 11361–11367 (2015). Article  CAS  PubMed  PubMed Central  Google Scholar  *

Tyanova, S. _et al._ The Perseus computational platform for comprehensive analysis of (prote)omics data. _Nat. Methods_ 13, 731–740. https://doi.org/10.1038/nmeth.3901 (2016). Article  CAS 

PubMed  Google Scholar  * Tran, T. N., Afanador, N. L., Buydens, L. M. C. & Blanchet, L. Interpretation of variable importance in partial least squares with significance multivariate

correlation (sMC). _Chemom. Intell. Lab. Syst._ 138, 153–160 (2014). Article  CAS  Google Scholar  * Strohalm, M. _et al._ Poly[N -(2-hydroxypropyl)methacrylamide]-based tissue-embedding

medium compatible with MALDI mass spectrometry imaging experiments. _Anal. Chem._ 83, 5458–5462 (2011). Article  CAS  PubMed  Google Scholar  * Zhang, N., Wang, J., Ye, J., Zhao, P. &

Xiao, M. Oxyalkylation modification as a promising method for preparing low-melting-point agarose. _Int. J. Biol. Macromol._ 117, 696–703 (2018). Article  CAS  PubMed  Google Scholar  *

Klepárník, K., Fanali, S. & Boček, P. Selectivity of the separation of DNA fragments by capillary zone electrophoresis in low-melting-point agarose sol. _J. Chromatogr. A_ 638, 283–292

(1993). Article  Google Scholar  * Goodwin, R. J. A., Lang, A. M., Allingham, H., Borén, M. & Pitt, A. R. Stopping the clock on proteomic degradation by heat treatment at the point of

tissue excision. _Proteomics_ 10, 1751–1761 (2010). Article  CAS  PubMed  Google Scholar  * Gill, E. L., Yost, R. A., Vedam-Mai, V. & Garrett, T. J. Precast gelatin-based molds for

tissue embedding compatible with mass spectrometry imaging. _Anal. Chem._ 89, 576–580 (2017). Article  CAS  PubMed  Google Scholar  * Ivanov, D. P. & Grabowska, A. M. Spheroid arrays for

high-throughput single-cell analysis of spatial patterns and biomarker expression in 3D. _Sci. Rep._ 7, 1 (2017). Article  Google Scholar  * Visscher, M. _et al._ Data processing pipeline

for lipid profiling of carotid atherosclerotic plaque with mass spectrometry imaging. _J. Am. Soc. Mass Spectrom._ 30, 1790–1800 (2019). Article  ADS  CAS  PubMed  PubMed Central  Google

Scholar  * Vizcaíno, J. A. _et al._ The proteomics identifications (PRIDE) database and associated tools: Status in 2013. _Nucleic Acids Res._ 41, 1063–1069 (2013). Article  Google Scholar 

Download references AUTHOR INFORMATION AUTHORS AND AFFILIATIONS * Institute of Life Sciences, Sant’Anna School of Advanced Studies, Pisa, Italy Francesco Greco * Fondazione Toscana Gabriele

Monasterio, Pisa, Italy Francesco Greco * Fondazione Pisana per la Scienza ONLUS, San Giuliano Terme (PI), Italy Francesco Greco, Luca Fidia Pardini, Asia Botto & Liam Andrew McDonnell *

Department of Chemistry and Industrial Chemistry, University of Pisa, Pisa, Italy Luca Fidia Pardini & Asia Botto Authors * Francesco Greco View author publications You can also search

for this author inPubMed Google Scholar * Luca Fidia Pardini View author publications You can also search for this author inPubMed Google Scholar * Asia Botto View author publications You

can also search for this author inPubMed Google Scholar * Liam Andrew McDonnell View author publications You can also search for this author inPubMed Google Scholar CONTRIBUTIONS L.A.M. and

F.G. conceived the idea of the present work which was carried out under the supervision of L.A.M. F.G. chose the methodology and developed the software for the data analysis. F.G., L.F.P,

and A.B. conducted the investigation. F.G. wrote the original manuscript draft ad L.A.M. provided review and editing. CORRESPONDING AUTHOR Correspondence to Liam Andrew McDonnell. ETHICS

DECLARATIONS COMPETING INTERESTS The authors declare no competing interests. ADDITIONAL INFORMATION PUBLISHER'S NOTE Springer Nature remains neutral with regard to jurisdictional claims

in published maps and institutional affiliations. SUPPLEMENTARY INFORMATION SUPPLEMENTARY INFORMATION 1. SUPPLEMENTARY INFORMATION 2. SUPPLEMENTARY INFORMATION 3. RIGHTS AND PERMISSIONS

OPEN ACCESS This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or

format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or

other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not

included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission

directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Reprints and permissions ABOUT THIS ARTICLE CITE THIS ARTICLE Greco,

F., Pardini, L.F., Botto, A. _et al._ Low-melting point agarose as embedding medium for MALDI mass spectrometry imaging and laser-capture microdissection-based proteomics. _Sci Rep_ 13,

18678 (2023). https://doi.org/10.1038/s41598-023-45799-5 Download citation * Received: 15 September 2023 * Accepted: 24 October 2023 * Published: 31 October 2023 * DOI:

https://doi.org/10.1038/s41598-023-45799-5 SHARE THIS ARTICLE Anyone you share the following link with will be able to read this content: Get shareable link Sorry, a shareable link is not

currently available for this article. Copy to clipboard Provided by the Springer Nature SharedIt content-sharing initiative