Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2026 Mar 19;98(12):8839–8850. doi: 10.1021/acs.analchem.5c02567

Spatial Multiomics Combining Lipids and Gene Expression Using MALDI ISH MSI

Kyle A Vanderschoot †, Jacob P Padilla †, Kelli A Steineman †, Christopher M De Caro †, Yongheng Wang ‡, Aijun Wang §,∥, Marie C Heffern †, Elizabeth K Neumann †,*
PMCID: PMC13044879  PMID: 41856123

Abstract

Current spatial gene expression techniques rely upon DNA microarrays, next-generation sequencing, and fluorescence microscopy, which are often costly, time-consuming, and/or restricted in the sampling area. Matrix-assisted laser desorption/ionization in situ hybridization (MALDI ISH) mass spectrometry imaging (MSI) combines the spatiotemporal capabilities of in situ hybridization with MSI using photocleavable peptide mass tags and nucleic acid probes to enable serial detection of lipidomic and gene expression data. We synthesized a copper-catalyzed azide–alkyne cycloaddition (CuAAC) o-nitrobenzylic azide linker that is compatible with solid phase peptide synthesis (SPPS) to enable a flexible and modular conjugation platform between DNA sequences and peptide mass tags. After conjugation of the mass tag and hybridization to the native RNA target, the triazole-functionalized o-nitrobenzyl linker can undergo photolytic cleavage to separate the hybridized nucleic acid sequence from its corresponding peptide mass tag. Using this approach, we synthesized 33 unique photocleavable mRNA probes and successfully validated 12 distinct gene expression patterns across sagittal sections of fresh-frozen murine brain of both wild-type and a Canavan’s disease model. This method enables codetection of native RNA and lipidomic signals from the same tissue section, providing an alternative strategy for spatial multiomic analysis. While this study demonstrates a 12-plex implementation, the platform is scalable and may support expanded multiplexing with continued optimization of peptide design and instrumentation. This approach is broadly applicable across disease models and offers new insights into transcriptome–metabolome interactions.

graphic file with name ac5c02567_0006.jpg

graphic file with name ac5c02567_0005.jpg

Introduction

Historically, transcriptomic-based measurements have been the pinnacle for understanding healthy and diseased cellular function. Largely, this is because mRNA is the functional intermediate between DNA and expressed proteins, providing insight into both ends of the central dogma of biology. Additionally, mRNA can be sequenced and amplified, enabling a range of endogenous expression levels and different splicing variants associated with a singular gene to be concurrently measured. , Amplification is a critical aspect that allows lowly expressed mRNA to be detected regardless of instrumental limits of detection and dynamic range, enabling high-throughput quantitative analysis. As a result of these features, transcriptomics-based measurements have been at the forefront of biological scientific discovery. While bulk or single-cell measurements are informative, biological systems are composed of a unique assortment and arrangement of cells that coordinate together to enable higher order functions, such as memory and cognition within the brain. As such, there has been significant interest and development in spatially resolved transcriptomic approaches, which aim to contextualize the function of an individual cell within its cellular neighborhoods and their anatomical arrangements. To date, the standard method for spatially probing gene expression has been derived from in situ hybridization (ISH). ISH-based approaches leverage complementary base pairing between a known DNA probe and a native mRNA transcript of interest. The location of the DNA probe, and therefore mRNA, is generally visualized using single molecule fluorescence in situ hybridization (smFISH) assay and has been standard practice for preforming spatiotemporal analysis of transcription products. Recent improvements to fluorescent microscopes has enabled subcellular resolution imaging of both live and dead cells. While incredibly powerful, fluorescence microscopy is limited to visualizing a few (usually <7) fluorescent fluorescent channels at a time, due to spectral crowding and spectral bleeding/overlap. Technologies, such as MERFISH (Vizgen), RNAScope (ACDBio), and others, circumvent this issue by introducing adaptor oligonucleotide constructs that allow for additional specificity and sensitivity, making it possible to cycle fluorescently labeled oligonucleotide barcodes on and off of the designed constructs. However, these approaches require a cyclic process through fluorescent channels because of the spectral bleeding/overlap, to enable discrete and unambiguous detection of the applied detection probes. Similar methodologies enable other spatial transcriptomics-based technologies, such as the Xenium platform (10x Genomics) and the GeoMx Digital Spatial Profiler (Nanostring). , In brief, the Xenium platform performs in situ transcriptomic profiling by hybridizing fluorescently labeled detection probes to indirectly target amplified mRNA moleculesgenerated through rolling circle amplificationand utilizes sequential imaging to decode gene-specific barcodes at subcellular resolution. Similarly, the GeoMx Digital Spatial Profiler employs barcoded oligonucleotide probes that bind to RNA or protein targets in tissue sections, with spatial segmentation achieved via UV-directed barcode release from defined regions of interest, followed by quantification using next-generation sequencing or NanoString’s nCounter analysis. To date, these systems have been widely used and demonstrate the importance of spatial analysis within biomedical research. While important, these platforms are expensive, time-consuming, and often limited in either imaging area or sample size. To circumvent these limitations, we have developed a method of performing in situ hybridization with mRNA-correlated mass tags, enabling the use of modern mass spectrometry imaging approaches for exploring gene expression that we have termed MALDI ISH MSI.

MSI is capable of measuring hundreds to thousands of molecules within a section of tissue and can be done through a variety of ionization sources. While each of these approaches have their unique benefits, matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI MSI) is well suited to address the limitations of traditional fluorescence-based ISH or immunohistochemical (IHC) approaches. MALDI MSI benefits from high dynamic range, high sensitivity (attomole), speed (30 pixels/sec), mass accuracy (<5 ppm), multiplexed capacity (concurrent detection of thousands of features) and spatial resolution (<10 μm − commercially; < 5 μm in semicommercial instruments , ), without disturbing their spatial context. Typically in a MALDI MSI experiment, fresh frozen tissue is cryosectioned into 10–20 μm thick sections, thawed mounted onto a conductive surface, and coated in a UV absorbing chemical matrix to enhance and facilitate desorption and ionization of endogenous biomolecules. By rastering the tissue under a UV laser, it is possible to acquire spectra at each raster position for the subsequent construction of ion images. These ion images contain information about the abundance and distribution of each detected molecule. In our case, detected mass tags are correlated to mRNA that is definitive of a cell type state or histological/pathological features. Multimodal methods have been developed to correlate MALDI MSI data with existing spatial transcriptomics platforms, such as Visium (10x Genomics), allowing for untargeted analysis between the metabolome and the transcriptome across serial sections. Furthermore, correlative approaches using the same biological sample, as demonstrated in metaFISH, integrate FISH and MSI, demonstrating multimodal approaches combining the transcriptome and metabolome.

Indirect detection of large biomolecules has previously been used in MSI experiments, such as in Tag-Mass, MALDI IHC (AmberGen) and many others. , Tag-Mass was developed in 2004 and used a photocleavable peptide mass tagging strategy to label oligonucleotides and antibodies for ISH and IHC in MALDI MSI experiments. MALDI IHC follows a similar photocleavable mass tag design for antibody detection of proteins, and has been used to examine antibodies in a highly multiplexed, multiomic manner. , Finally, mass tagging of biomolecules has been used in a variety of ionization sources, including time-of-flight secondary ion mass spectrometry (ToF-SIMS), , desorption electrospray ionization (DESI), laser desorption ionization (LDI), and laser ablation inductively coupled plasma-mass spectrometry (LA-ICP-MS). , AmberGen has released its own version of MALDI ISH, using a photocleavable peptide mass tag conjugated to an oligonucleotide through strain-promoted azide–alkyne cycloaddition (SPAAC). The constructed probe is then used as the indirect detection probe used in an RNAScope assay for multiomic analysis on the same tissue section.

Considering the strengths and limitations of each of these approaches, we have established a method capable of probing mRNA and native metabolites, such as lipids, within a single section of tissue in a cost-efficient and high-throughput manner, with spatial resolution down to 10 μm pixel size using MALDI MSI. Specifically, the key differences described herein include a method for enzymatic signal amplification using a Cu­(I)-catalyzed azide–alkyne cycloaddition (CuAAC)-compatible deoxynucleotide triphosphate (dNTP). This CuAAC-compatible dNTP is then labeled using a flexible and efficient CuAAC-based linker that is easily incorporated for the scalable and customizable synthesis of mass tag labels. Indeed, many of the aforementioned approaches can adopt this enzymatic step to decrease the cost and increase the efficiency of the mass tagging process.

Methods

Chemicals

All chemicals were purchased from Sigma-Aldrich (St. Louis, MO) without additional purification unless otherwise specified.

Synthesis of Azide-Modified Photocleavable Linker

4-Bromomethyl-3-nitrobenzoic acid (2.5 g, 9.61 mmol) was dissolved in 50 mL of methanol, and the mixture was stirred on ice. Sodium azide (0.75 g, 11.5 mmol, 1.2 equiv) was added as a 0.38 M methanol solution. The mixture turned a dark yellow upon addition and after stirring on ice for 1 h was removed to stir at 22 °C overnight. The resulting mixture was evaporated dry and resuspended in a minimal amount of dichloromethane (DCM) (∼300 mL). The organic layer was washed with 1 M HCl (50 mL), dH2O (50 mL) and a brine solution (50 mL) before being dried over Na2SO4. The DCM was evaporated under a rotary evaporator and further dried under a high vacuum overnight. The resulting solid, a light-yellow powder, (1.86 g, 87%) was weighed and examined using electrospray ionization in negative mode (SI Figure 1) (Exact Mass = 222.0389; [M-H]− = 221.0279) and 1H NMR (400 MHz, DMSO) δ 13.71 (s, 1H), 8.52 (d, J = 1.7 Hz, 1H), 8.30 (dd, J = 8.0, 1.8 Hz, 1H), 7.87 (d, J = 8.0 Hz, 1H), 4.97 (s, 2H) (SI Figure 2, SI Figure 3).

Synthesis of Azide-Modified Photocleavable Peptide Mass Tags

A CEM Liberty Blue 2.0 microwave peptide synthesizer was used to synthesize 0.1 mmol of variable static sequences prior to combinatorial manual synthesis. The first seven residues of the peptide were coupled to an Fmoc-Arg­(Pbf)-Wang resin (0.3 mmol/g, Aapptec Peptides, Louisville, KY). The resin was swollen in anhydrous N,N-dimethylformamide (DMF) for 5 min before deprotecting with 4-methylpiperidine (10% in DMF). Amino acid reagents (0.2 M, Sigma-Aldrich, St. Louis, MO; Advanced ChemTech, Louisville, KY; TCI America, Tokyo, Japan; Thermo Scientific, Waltham, MA) were coupled using 0.1 M diisopropylcarbodiimide (DIC) and 0.1 M OxymaPure for 2 min under microwave radiation at 90 °C. After synthesizing the static region of the peptide, the resin was removed from the synthesizer and combinatorically exposed to manual solid phase synthesis using 5 mL disposable columns (Thermo Scientific Pierce, Waltham, MA). Briefly, peptides were swollen in DMF for 2 h and washed with DMF, followed by deprotection of the Fmoc protecting groups by addition of 4-methylpiperidine (25% in DMF) for 10 min at 22 °C. Amino acid coupling was performed by adding 5 mol equiv of the designated amino acid and 4.9 mol equiv of DIC and OxymaPure to the column and heating at 95 °C for 20 min. The 4-azidomethyl-3-nitrobenzoic acid photocleavable linker was then coupled to the N-terminus before the final deprotection and cleavage of the peptide. Cleavage and deprotection was carried out by using trifluoroacetic acid (TFA), phenol, triisopropylsilane (TIPS), 88:5:5:2 for 2 h at 22 °C. The peptide was then precipitated in cold diethyl ether and centrifuged to remove unwanted deprotection and cleavage products. After washing the solid pellet three times, the ether was evaporated, and the peptide was dried under vacuum. The peptide was solubilized to a 5 mM concentration in 50% acetonitrile and stored at −20 °C until conjugation.

5-Ethynyl-2′-deoxyuridine Triphosphate (5-EdUTP) Tailing of RNA Binding Probes

In brief, 3′OH oligonucleotide sequences (6 nmol, 60 μM, Integrated DNA Technologies, Coralville, IA) were incubated with terminal transferase (100 U, 2 U/μL, New England Biolabs, Ipswich, MA) and 1X TdT buffer (50 mM potassium acetate, 20 mM tris-acetate, 10 mM magnesium acetate, New England Biolabs, Ipswich, MA) cobalt chloride (0.25 mM), and 5-ethynyl-2′-deoxyuracil triphosphate (5-EdUTP, 20 nmol, 200 μM, Abcam, Fremont, CA) for 90 min at 37 °C in a 100 μL reaction volume. Following incubation, the enzyme was heat inactivated at 75 °C for 10 min. The reaction mixture was cooled to 4 °C before proceeding to the mass tag labeling.

Copper Click Mass Tag Labeling of Ethynyl Tailed RNA Binding Probes

The alkyne-modified oligonucleotide (6 nmol, 47.6 μM) was mixed with the azide-modified peptide mass tag (50 nmol, 666.7 μM) and Mili-Q water. CuSO4 (50 nmol, 666.7 μM) and tris­(benzyltriazolylmethyl)­amine (THPTA, 50 nmol, 666.7 μM) are premixed and allowed to incubate for 10 min before adding the azide/alkyne. Sodium-L-ascorbate (200 nmol, 1.7 mM) was then added and mixed vigorously to a final volume of 126 μL. The mixture was degassed under N2, before allowing to incubate at 37 °C for 2 h. The resulting mixture was purified by precipitation, using 0.3 M sodium acetate at pH 5.2 and 0.7x the final reaction volume of isopropanol.

Animal Experiments

All animal procedures were approved by The University of California, Davis (UCD) institutional animal care and use committee (IACUC). C57BL/6J mice (strain #:000664) were purchased from Jackson Laboratory. Mice with Canavan disease (Aspanur7/J, strain #:008607) were purchased from Jackson Laboratory. Day 90 male mice were euthanized by a CO2 overdose. Mouse brains were collected and frozen over an isopentane (J.T Baker, Q22308) and dry ice slurry before being stored at −80 °C.

Matrix-Assisted Laser Desorption/Ionization In Situ Hybridization Assay

Fresh frozen tissue was sectioned to 10 μm and mounted on an indium tin oxide (ITO) coated glass slide (Delta Technologies, Loveland, CO). The mounted tissue was thawed in a vacuum desiccator for 15 min prior to being subjected to fixation with cold 4% paraformaldehyde (PFA, Thermo Scientific, Waltham, MA) in neutral PBS. The tissue was fixed for 15 min at 22 °C, followed by two washes with 1X PBS for 5 min each. The slide was submerged in 70% ethanol, 95% ethanol, 100% ethanol for 3 min each followed by a 5 min incubation 100% CHCl3. The tissue was then rehydrated in 100% ethanol, 95% ethanol, and 70% ethanol for 3 min each. Finally, the slide was placed in 2X saline-sodium citrate (SSC) buffer for 5 min and allowed to incubate at 37 °C for 2 h in 20% formamide (Ambion Inc. Austin, TX)/2X SSC. The mass-tagged oligonucleotide probes were heated briefly at 95 °C in 100% formamide to denature any secondary structure and diluted in hybridization buffer (100 mg/mL dextran sulfate, 20% formamide, 2X SSC, 1 mg/mL bovine serum albumin (BSA), 1 mg/mL yeast tRNA (Thermo Scientific, Waltham, MA), 200 μM ribonucleoside vanadyl complex (New England Biolabs, Ipswich, MA) to 5 ng/μL). The tissue was incubated in a hybridization mixture at 37 °C overnight. Following the incubation, the sample was washed with 20% formamide/2X SSC, 10% formamide/1X SSC, 10% formamide/0.5X SSC for 5 min each at 37 °C, and 150 mM ammonium acetate pH 7.0 two times for 5 min each at 22 °C. The sample was then air-dried for 15 min under darkness. Photocleavage was preformed using the UV illuminator (Phrozen Tech, Hsinchu City, Taiwan), exposing the UV-A light for 30 min at full power.

Mass Spectrometry Imaging Parameters/Image Processing

The sample was then coated in 2′,5′-dihydroxyacetophenone (DHA, 10 mg/mL in 70% ACN) on an M3+ automated aerosolized sprayer (HTX Imaging, Chapel Hill, NC). The sample was sprayed over 8 passes using N2 heated to 50 °C, a flow rate of 120 μL/min, a nozzle velocity of 1200 mm/min, tack spacing of 2.5 mm, and drying time of 2 s. The sample was then run on the timsTOF fleX (Bruker Scientific, Billerica, MA). For untargeted lipid acquisition, MALDI MSI analysis was performed in negative ion mode, although positive ion mode could also be performed. The average lipid spectra were internally recalibrated using known lipid peaks ([PA(32:0)-H]−, [PA(36:1)-H]−, [PE(36:4)-H]−, [PS(36:1)-H]−, [PI(34:1)-H]−, [PI(38:4)-H]−, [PI(40:7)-H]−) and detected m/z values were assigned putative identities from the LipidMAPS database (SI Table 1). To confirm molecular identity, lipid species were analyzed using tandem MS (MS/MS, SI Figure 4.1–4.15). For RNA detection mass tag acquisitions, MALDI MSI was performed in positive ion mode with a mass range from m/z 1,100 to 1,800. More detailed instrumental method parameters can be found in the Supporting Information (SI Table 2, SI Table 3).

MALDI ISH Data Analysis

MALDI ISH data were analyzed by using SCiLS Lab (Bruker Scientific, version 2025b). RNA targets were detected via peptide mass tags released upon UV-induced photocleavage of an o-nitrobenzyl-based linker. Photocleavage produces a defined and chemically related series of linker-derived photoproducts (“linker states”) originating from a common parent peptide mass tag. As multiple linker states may be detected for a single mass tag, ion image generation requires the selection of a representative m/z value.

For each peptide mass tag, predefined analytical criteria were used for selection of ion image selection: (i) reproducible detection across experiments consistent spatial alignment with chemical identities that originate from the o-nitrobenzyl photocleavage mechanism (SI Scheme 4), (ii) dominance of signal intensity relative to other linker-derived species associated with the same mass tag, and (iii) absence of isobaric overlap with species derived from the other mass tags applied in a multiplexed setting. The selected linker state satisfying these criteria is selected as the representative species for the gene and denoted with [cAZL-X] where X represents the linker state corresponding to the structures defined in Figure . While the selected m/z value may vary depending on the composition of the multiplexed tag set, the selected m/z represents the most intense and analytically robust photoproduct for each gene.

2.

2

Open in a new tab

Proposed structures of photocleavable linker after UV photolysis reaction and separation from the RNA binding probe from the triazole. Top: Proposed structures of photocleavable linker products corresponding to the m/z values associated with each of the products. A) Aryl nitroso radical species generated in-source laser exposure. B) o-nitrosobenzaldehyde generated from elimination of benzylic triazole from benzisoxazolidine intermediate. C) Nitroxyl radical generated from an aryl nitroso radical (A). D) Hydroxybenzisoxazole generated from hydration, cyclization, and dehydration of o-nitrosobenzaldehyde (B). E) 13C isotope of the hydroxybenisoxazole (D). F) 13C isotope of the hydroxybenzisoxazole (D). G) Benzoxazoline created from the in-source dehydration of the hydroxybenzisoxazole photoproduct (D). H) 13C isotope of benzisoxazole. I) o-aminobenzaldehyde made from the reduction and dehydration of the nitroxyl radical (C). J) 13C isotope of o-aminobenzaldehyde (I). Bottom: Spatial distribution of detected ions (A-J) coming from the same tissue used in the full MALDI ISH assay.

Ion images were generated using identical preprocessing parameters and root-mean-square (RMS) normalization using SCiLs Lab to calculate the intensity within each experiment. Custom Python scripts were developed and implemented to enumerate and calculate the m/z values for all chemically plausible peptide-derived ions, including photocleavage products, synthetic truncations, and common MALDI adducts. A second script was used to cross reference the library of theoretically plausible m/z values with the experimentally observed m/z values, using a 5 ppm mass error tolerance and selecting a single representative ion per gene using the criteria described above.

For multiomic and higher multiplexed experiments, RNA-associated mass tag ion images were coregistered and overlaid with untargeted lipid ion images acquired using consistent spatial coordinates from the same tissue. Feature finding was preformed using SCiLS Lab T-REX2 algorithm, and spatial segmentation was carried out using a bisecting k-means algorithm and applied to the overlaid layers to define regions of interest for receiver operating characteristic (ROC) analysis and intensity comparison.

Results and Discussion

Here, we describe a customizable method for the multiplexed spatial detection of RNA and endogenous metabolites using a shared MALDI MSI workflow. This approach builds upon a prior mass tagging strategy to enable multiomic assessment of gene expression and metabolomic signatures within a single, fresh frozen tissue section (Figure ). Our approach involves the incorporation of peptide-encoded mass tags to the nucleic acid hybridization moieties capable of being analyzed by MALDI MSI. By incorporating a photocleavable linker between the nucleic acid sequence and the correlated synthetic peptide tags, mass tags can be temporally released after being delivered to their appropriate spatial location. In brief, our method for mass labeling mRNA probes is designed in a flexible, modular format, allowing tags and probes to be mixed and matched in a cost-effective, scalable synthesis and application. We synthesized an o-nitrobenzylic azide photocleavable linker (SI Scheme 1) that is compatible with standard Fmoc protected peptide synthesis (SI Scheme 2). With the goal of having a flexible, modular, and expandable design. An azide-modified linker allows for copper-catalyzed azide–alkyne cycloaddition (CuAAC) click chemistry, which enables orthogonal and efficient labeling. This approach ensures that each mass tag has a unique m/z value and contains amino acid residues that are highly ionizable for easy MALDI MSI detection (SI Table 4, SI Table 5).

1.

1

Open in a new tab

General MALDI ISH MSI workflow. Fresh frozen tissue is sectioned to 10 μm and mounted to an ITO glass slide. The native tissue is coated in 2′,5′-dihydroxyacetophenone (DHA) for untargeted lipid analysis via MALDI MSI. Only after lipid imaging is completed, the matrix is removed, and the same section undergoes MALDI ISH analysis, using photocleavable mass-tagged DNA probes to detect mRNA targets. Following photolytic cleavage of the mass tags separated from hybridized probes, the slide is recoated with DHA and reanalyzed by MALDI MSI (e.g., using a Bruker timsTOF fleX) to localize gene expression via the corresponding mass tag m/z signal.

MALDI ISH Probe Design

Careful parameters were considered when designing the peptide mass tags to avoid potential side reactions, increase the signal intensity, and prevent on-bead aggregation during synthesis. One such parameter is the choice of amino acids. Proline (P) was avoided due to difficulty associated with synthesis. Cysteine (C) is prone to the formation of disulfide bridges and was excluded. Lysine (K) and histidine (H) at the N-terminus or middle position have been reported to decrease signal and were therefore not used. Tryptophan (W) and methionine (M) are sensitive to oxidation and were not used. Asparagine (N) and glutamine (Q) are susceptible to loss of ammonia during ionization and were used sparingly. Isoleucine (I) is isobaric with leucine (L) and, therefore, was not used. Of the remaining ten amino acids, alanine (A), leucine (L), valine (V), phenylalanine (F), tyrosine (Y), aspartic acid (D), glutamic acid (E), serine (S), and threonine (T) were used, resulting in easier synthesis and increased purity during diethyl ether precipitation. Additionally, the amino acid position within the mass tag is a key consideration. Arginine (R) on the C-term has been shown to promote signal intensity; therefore, all mass tags contain arginine in this position to enhance analysis. Repeating units of either charged residues or hydrophobic residues was avoided to prevent aggregation of peptides during and after synthesis.

Mass tag diversity was achieved by using a combinatorial-like approach. A 7-mer static peptide was made with a CEM Liberty Blue 2.0 microwave peptide synthesizer and fractioned into 12 different vessels. Three additional amino acid residues were then manually coupled via standard Fmoc-SPPS. Finally, the azide-modified photocleavable linker was added to the N-terminus of each peptide in a similar manner before acidic cleavage from the resin, collection, precipitation and analysis. Additionally, mass tag ionization efficiency is critical to achieving a sensitive and accurate assay result. High ionization efficiency allows for the detection of mass tags (and thus RNA indirectly) in low abundance, with acceptable signal-to-noise (S/N) ratios (>3 S/N). To ensure this, positively charged amino acid (e.g., arginine) are included within the designed static sequence such that every synthesized mass tag will have a permanent positive charge, even without matrix protonation. This design enhances the assay’s sensitivity for the use of only three mRNA probe sequences with a to detect endogenous RNA. This simplifies bioinformatic design in contrast to traditional smFISH assays that requires 25–48 probes labeled in a single fluorescence channel to achieve comparable sensitivity.

An advantage of our approach compared to traditional FISH approaches is the tailing step, which permits multiple mass tags to label the same tail, amplifying the signal further. By using three probe sequences per gene target, we significantly reduce material cost without sacrificing signal intensity, while still maintaining the ability to probe for potential isoforms and splicing variants. The all-in-one hybridization protocol also minimizes tissue degradation compared to fluorescent assays that use a cyclic hybridize/strip strategy. Lastly, while we use a peptide mass tag here, it should be noted that anything that can be attached to the carboxylic acid of the azide-modified photocleavable linker can be used as a mass tag, including lipids, small molecules, etc.

The RNA binding sequences were designed by inputting The National Center for Biotechnology Information Consensus Coding Sequence (NCBI CCDS) , into Stellaris Probe Designer (LGC Biosearch Technologies, Petaluma, CA) to generate a set of 20mer sequences that were then confirmed as exclusive sequences to the RNA target. The proposed probe sequences were then filtered for guanine/cytosine composition, melting temperature (T m), homopolymer regions, and potential secondary structures (SI Table 6). For each gene target, three distinct 20mer probes were selected to tile nonoverlapping regions of the mRNA transcript, following a strategy analogous to smFISH. This design increases the number of independent hybridization events per transcript, thereby improving signal robustness and minimizing false negatives due to local sequence structure or accessibility. Here we refer to “plex” as the complete sets of three oligonucleotide sequences that target the same gene and not the total number of oligonucleotides that are applied to the tissue at one time.

After oligonucleotide synthesis, the MALDI ISH probe is synthesized as depicted in (SI Scheme 3). Oligonucleotides are modified using 5-ethynyl-2′-deoxyuridine triphosphate (5-EdUTP) using terminal transferase (TdT) for template-independent tailing. Because of the structural similarity to thymidine, the TdT enzyme is highly efficient at incorporating the click-compatible, nucleobase. CuAAC was chosen over SPAAC because of the potential decrease in enzymatic incorporation efficiency associated with steric bulk from the dibenzocyclooctyne group. The length of the added tail was optimized to reach lengths greater than 200 bases (SI Figure 5), providing an amplification domain for the probe, allowing for detection by increasing the number of available sites for mass tag labeling. Following CuAAC labeling, the constructed probe was purified via isopropyl alcohol precipitation and used in the ISH portion of the workflow.

MALDI ISH Probe Analysis

Before the MALDI ISH workflow was run, the product of each step in the conjugation chemistry was analyzed. First, the azide-modified peptide mass tags were visualized using a MALDI spot tests to verify successful synthesis and any potential side products from the synthesis (SI Figures 6.1 – 6.32). To further assess photolytic behavior, the azide-modified peptides were exposed to UV light in solution, leveraging the greater reactivity of the azide to generate the expected photocleaved products that would be detected in the MALDI ISH assay (SI Figures 6.33–6.63). In this quality assessment, the expected o-nitrosobenzaldehyde product was observed, but since the reaction was carried out in solution, the hydrated product was observed as the main product. The main byproducts detected arose from incomplete couplings of the N-terminal residues introduced during manual synthesis as well as neutral losses (e.g., N2, CO2). To ensure confident mass tag identification, all detected m/z values were assigned based on expected truncation product mass, accounting for potential adducts within the 5 ppm error. Peaks below 1100 Da were excluded from analysis to reduce background interference.

Because the peptide mass tags were designed using a restricted amino acid alphabet, certain truncation products can result in isobaric m/z values, despite differences in amino acid sequence. While click chemistry eliminates most labeling side products lacking the azide-modified linker, minor truncation products with overlapping m/z values can still occur. Nevertheless, the resolving power of the timsTOF fleX allows effective discrimination of signals within <5 ppm mass error. Furthermore, isobaric and linker-derived variants can be separated using more advanced instrumental techniques such as trapped ion mobility spectrometry (TIMS), which represents a future direction for this work.

Next, the CuAAC product was analyzed using the 5-ethynyl-2′-deoxyuridine, rather than 5-EdUTP, as the alkyne source. This substitution simplified the analysis by minimizing the need for oligonucleotide-specific liquid chromatography (LC) methods while still preserving the generated triazole present in the full construct (SI Table 7). The starting material was visualized using LC-MS, (seen as [M + H]2+ from two R residues), followed by the CuAAC product. The CuAAC product was aliquoted and exposed to UV-A light to monitor the photolytic cleavage products over time (SI Figure 7A). Because triazole leaving groups are less commonly used in photocleavable linkers, we assessed the efficiency of the photolytic separation over time (SI Figure 7B). At 30 min, the cleaved product became the base peak in the mass spectrum and was selected as the minimum UV-A exposure time for subsequent ISH experiments.

Triazoles are not often photocage leaving groups compared to o-nitrobenzylic alcohols or amines, however, triazoles are photoreactive and stable during photolysis. In our design, the triazole does not serve as the photoreactive center but instead serves as a stable chemical linkage between the peptide mass tag and the o-nitrobenzyl moiety. The photocleavage step of the MALDI ISH workflow is driven by a 405 nm light source, which enables the photolysis of the o-nitrobenzyl to proceed. Following the same mechanistic pathway that has been extensively studied for esters, ethers, carbamates, oximes, amines, amides and imidazoles, the photocleavage occurs at the benzylic carbon, releasing the triazole and generating an o-nitrosobenzaldehyde. ,, While local electronic effects can influence cleavage rates and efficiencies, the fundamental photolytic mechanism proceeds via a Norrish Type II-like mechanism, using aci-nitro and benzisoxazolidine intermediates to rearrange yielding a o-nitrosobenzaldehyde [cAZL-B] (SI Scheme 4).

Following photocleavage of the o-nitrobenzyl moiety, the resulting o-nitrosobenzaldehyde on the N-terminus of the peptide can undergo two distinct decomposition reactions. In the first, nucleophilic attack by water yields a hydrated aci-nitro intermediate that can cyclize to form a dihydrobenzisoxazolidine intermediate, consistent with classic o-nitrobenzyl chemistry. , Subsequent dehydration of this intermediate generates a hydroxybenzisoxazole [cAZL-D], which is known to form under mildly acidic conditions. , During the desorption/ionization process, the hydroxybenzisoxazole can be dehydrated in the ionization source to give a benzoxazoline [cAZL-G].

In the second pathway, the nitroso group may engage in photoinduced one-electron processes, forming a nitroso-derived radical intermediate [cAZL-A]. Such radical chemistry is well established for nitrosoarenes and can proceed through proton-coupled reductions to yield aryl nitroxyl radical [cAZL-C] or arylhydroxylamines, which are readily dehydrated under energetic MALDI plume conditions to afford the corresponding o-aminobenzaldehyde [cAZL-I] (SI Scheme 4). These transformations are consistent with the known propensity of nitrosoarenes to undergo reduction to anilines, as described in nitroso literature, and explains the appearance of a loss of 16 Da relative to the o-nitrosobenzaldehyde precursor. Though these intermediates are transient, their existence is strongly supported by the consistent formation across mass tag analogs, characteristic fragmentation behavior (i.e., loss of H2O), and colocalization in tissue, which confirms their shared origin from a single peptide mass tag following photocleavage.

Each linker state variant observed in the full MALDI ISH assay was validated as a peptide by MS/MS analysis, confirming that the detected m/z values contain the expected amino acid composition of the corresponding tag (Figure ; SI Figures 8.1–8.6). Although some of the proposed structures are transient or potentially radical-containing, it is plausible that the MALDI matrix facilitates their stabilization, serving as a protonation sink and enabling detection. , Importantly, the spatial distribution of the different species is highly colocalized, supporting the conclusion that they arise from the same gene-specific hybridization event and are attributable to a single peptide mass tag. Different probes may preferentially use different linker states based on the chemical environment in which they are found and the peptide sequence to which they are attached to. Despite the minor difference in intensity the selection criteria of m/z values follow the criteria outlined in the Data Analysis section in the Methods section.

To assess whether peptide mass tag–labeled probes can be applied simultaneously without ionization efficiency interference, we compared relative signal intensity and spatial localization between single-plex and multiplex MALDI ISH experiments (SI Figure 9). A single-plex experiment was run in parallel to a 3-plex mixture containing the same probe, enabling a direct evaluation of multiplex-associated ion suppression and spatial fidelity. We observe that the signal intensity and localization in the single-plex and multiplexed experiments are preserved under multiplexed conditions, indicating that coapplication of multiple gene-targeted probe sets does not compromise detectability or spatial information. These results support the feasibility of scaling MALDI ISH to higher multiplex levels and establish a foundation for subsequent multigene and multiomic imaging experiments presented in this study.

We also evaluated the linearity and sensitivity of the MALDI ISH assay by titrating the concentration of mass-tagged probes and quantifying the resulting ion signal intensities (SI Figure 10). By doing so, we observed a clear, linear dose–response relationship across a range of hybridization concentrations (R2 = 0.95). This linearity confirms that MALDI ISH enables semiquantitative detection of transcript abundance without sacrificing spatial specificity. While inherent pixel-to-pixel variability arises from tissue heterogeneity and off-tissue regions during MADLI analysis, mean signal intensities remained proportional to probe concentration, demonstrating the assay’s reproducibility. Multiplexed experiments at 50 ng/μL further confirmed that simultaneous hybridization of multiple probes maintains signal intensity and spatial fidelity comparable to single-plex experiments.

To further evaluate the reproducibility and spatial consistency of MALDI ISH signal across biological replicates, we analyzed the intensity and spatial consistency of mass tag distribution across two sections of WT mouse brain tissue coming from four different animals (SI Figure 11). Consistent ion intensities were observed for the shared tags and the associated linker variants across all eight sections, with box plots demonstrating low variability in signal intensity. Ion images for each shared tag also revealed conserved spatial localization patterns across replicates, confirming that the peptide mass tag labeling is both technically reproducible and spatially robust across biological specimens. For visualization, the most intense m/z value among the related linker variants derived from each peptide mass tag was selected to represent the spatial distribution, ensuring that the displayed images reflect the predominant and most analytically stable species detected.

MALDI ISH in Canavan’s Disease Mouse Model

The multiplexing and multiomic capabilities of MALDI ISH was demonstrated using a 33-plex gene panel of a sagittal section of wild type (WT) mouse brain, where untargeted lipid analysis and gene expression were analyzed serially from the same tissue section (Figure ). First, the tissue was thawed under a vacuum and coated with a DHA matrix before being analyzed for lipids in negative ion mode at 20 μm. The matrix was subsequently removed during the fixation and ethanol dehydration steps. After hybridization with mass-tagged probes (SI Table 8), washing, and UV cleavage, the tissue was resprayed with DHA for analysis in positive mode. While data in this study were primarily at 20 μm spatial resolution, imaging at 10 μm is also achievable (SI Figure 12). Feature finding of the RNA-associated average mass spectra revealed 2,163 features and cross-referenced against 3,600 theoretical sequence and truncation products. The custom python script then applied the selection criteria for depicting the correct m/z within the 5 ppm error mass tolerance and S/N ratio >3 (SI Table 9).

3.

3

Open in a new tab

Multiomic capability of MALDI ISH demonstrated by combining untargeted lipid profiling and RNA profiling from the same tissue section from a 33-plex experiment (n = 1). A) Multicolor ion image displaying the spatial distribution of the detected mass tags associated with Igfbp5 (green), Chrna2 (blue), and Sla (red). A.1-A.8) Single ion image of gene probe and mass tag pairings detected in the 33-plex experiment in wild type mouse brain (20 μm spatial resolution). Linker states denoted [cAZL-D], where D = hydroxybenzisoxazole. Additional single ion images shown in SI Figure 13. B) Multicolor ion image of RNA and lipid overlaid data acquired from the same tissue section using the full MALDI ISH. B.1-B.8) Representative lipid single-ion images from the MALDI ISH data set. All scale bars are 1 mm.

While each ion image has a visually different localization, we have compared our MALDI ISH images to smFISH (SI Figure 14). Although hybridization between poly­(dU) tails of the probe and endogenous RNA poly­(A) tails is theoretically possible, such duplexes are among the weakest of all nucleic acid interactions, which is further discouraged by the applied hybridization conditions. We do not thoroughly test this here and acknowledge that we do not rule it out, but plan to address this more thoroughly in consecutive, validation studies. Fully synthesized mass tag probe constructs are precipitated and washed with 80% ethanol prior to their use in the assay, minimizing unbound azide-modified peptides that could nonspecifically adsorb to the tissue and generate false-positive signals (SI Figure 15). Further, the probes generated for CuAAC attachment have the theoretical potential to hybridize with the poly­(A) tails of endogenous mRNA; however, the calculated melting temperatures of such poly­(dU):poly­(A) interactions are less than those of the designed binding sequence and do not produce significant alteration to signal-to-noise ratio. Moreover, the high formamide concentration and low-salt wash stringency effectively eliminate these weak, noncomplementary interactions. Additional components in the hybridization buffer, such as yeast tRNA and BSA, further reduce nonspecific probe–protein or probe–nucleic acid binding within the tissue. While the density of labeling within the poly­(U) tail is unknown, the anatomical specificity of the resulting MALDI ISH signals is consistent with that of smFISH controls (SI Figure 14). We would like to note that MALDI ISH data can be viewed as complementary to chromogenic assays, as is done in the Allen Brain Atlas, that will have apparent inconsistencies, especially for lowly and diffusely expressed genes, due to different chemistries and data analysis strategies being employed.

Finally, this multiomic approach was used to analyze differences between WT (n = 1) and a Canavan’s Disease (CD) mouse model (n = 1). The 33-plex ISH panel and spatial lipidomics were performed to assess the molecular differences between these samples. A bisecting k-means machine learning algorithm (SCiLS Lab, Version 2025b) was used to clusters to compare between the two tissues (Figure A.1, Figure B.1). WT and CD tissues were segmented simultaneously using the same list of m/z features, segmenting lipids and RNA as separate overlaid layers to define regions of the aligned imaging data sets because of the differences in intensity. Similar clusters were acquired for each sample, including clusters associated with the corpus callosum, brain stem, cerebellar granular and molecular layers, neocortex, and hippocampus (SI Figure 16.1 and SI Figure 16.2). Gene expression differences were most pronounced in the thalamus, a region that experiences vacuolization and loss of spatial organization in CD. Within this region, Meis2, a transcription factor critical for neurogenesis and neuronal maturation demonstrates the expected neuronal loss associated with the vacuolization in CD (SI Figure 13). The Meis2 cluster in the thalamus was then selected for lipidomic analysis by overlaying ISH-defined boundaries onto the same tissue’s lipidomic data (Figure A.3, Figure B.3). Spatial correlation between the two data sets reveals that the thalamic subregion demonstrates both increases and decreases in sulfatide expression. Receiver operating characteristic (ROC) scores were calculated to determine the most discriminating features found from the segmented regions. In summary, SHexCer­(d40:1), PI(38:2), SHexCer­(t40:1), SHexCer­(t36:1), and SHexCer­(d38:1) were calculated to be discriminating features (ROC > 0.8).

4.

4

Open in a new tab

MALDI ISH multiplexing comparing fresh frozen WT mouse brain (n = 1) to a CD mouse model (n = 1) at 20 μm spatial resolution. A.1, B.1) Bisecting k-means segmentation of RNA-associated peptide mass tags on aligned data sets. Each color represents a dominant RNA-associated feature defining a spatial region. A.2, B.2) Bisecting k-means clustering of lipid ion images acquired from the same tissue sections. Each color represents a dominant lipid species that helps define the spatial region A.3, B.3) Segmented Meis2 expression region from A.1,B.1 overlaid with corresponding lipid data showing SHexCer­(t36:1) in WT (A.3) and CD (B.3). A.4) Average mass spectrum from the integrated data set with selected mass tags from the 33-plex experiment and top-ranking ROC scores within the Meis2-segmented region. More detailed spectra and single-ion images are provided in SI Figure 13. B.4) Ion intensity box plots of the top-ranking ROC score within the Meis2-segmented region. All scale bars = 1 mm.

Sulfatides are enriched in myelin sheaths and synthesized by mature oligodendrocytes, and their dysregulation is linked to white matter disease and neuroinflammation. , In the context of CD, a leukodystrophy caused by ASPA deficiency and subsequent N-acetylaspartate accumulation, the increase in sulfatides within Meis2-depleted thalamic regions may indicate a compensatory response by glial cells, particularly by oligodendrocytes and their progenitor cells, attempting to maintain or repair thalamic tissue in the face of neuronal loss. This is especially evident when comparing WT and CD models, where the CD model shows a high abundance of SHexCer­(t36:1) that is lacking in the WT (Figure A.3, Figure B.3). This midlength hydroxylated fatty acid chain SHexCer species has been observed in isolated pro-oligodendroblasts, supporting the hypothesis of glial compensation. , Conversely, SHexCer­(d40:1) and SHexCer­(t40:1) were less abundant and are produced by more mature oligodendrocytes that are the primary target of autoimmune demyelination in CD.

To visualize the integrated data set, the average mass spectrum of the Meis2-segmented region was plotted (Figure A.4), highlighting several peptide mass tags from the 33-plex experiment and the top-ranking ROC scoring lipids. Corresponding intensity boxplots (Figure B.4) further illustrate the differential abundances of the discriminating lipids between WT and CD animals. More detailed spectra and single-ion images are provided in SI Figure 13.

While additional biological replicates are needed to validate and quantify these findings, the results presented here demonstrate another framework for integrated spatial multiomics. By combining multiplexed ISH with untargeted MADLI MSI, we demonstrate the ability to simultaneously visualize changes in gene expression and lipid composition. Unlike conventional untargeted lipidomic workflows that rely on anatomical landmarks or morphological features, our approach leverages gene expression to define regions of interest. This enables more accurate lipidomic analysis and a more biologically relevant interpretation to disease-specific changes. This strategy can further be expanded to capture cell-type-specific changes, whether related or unrelated to underlying genetic mutations, offering a robust and flexible method for interrogating complex tissue heterogeneity in health and disease.

Conclusions

Here, we demonstrate a method for performing MALDI ISH MSI on the murine brain, targeting 33 genes through a cost-effective, flexible, and modular synthesis platform. We leverage a CuAAC-compatible linker that is capable of being incorporated with standard Fmoc peptide synthesis and undergoes photolysis efficiently. While we have leveraged the power of this technique using a small set of hybridization sequences correlated to a single peptide mass tag, the indiscriminate labeling chemistry can be applied in a smFISH-like assay using a modified nucleobase tailing strategy. Additionally, this mass tagging approach could be used as a substitute for fluorescent oligonucleotide detection probes in more elaborate nucleic acid architectures, such as those used in MERFISH, Xenium, RNAscope, and others.

Furthermore, we demonstrated the capability to integrate untargeted lipidomics with targeted gene expression data, derived from the same tissue section. The use of a mass spectrometer enables multiomic detection of lipids and mRNA features that allows for cell type information to give additional context to the cell’s lipid profile. This serial approach can easily be applied to incorporate the detection of small polar metabolites, drugs, and secondary metabolites. Additionally, the high resolving power of the mass spectrometer allows for simultaneous multiplexing like never before. While a small sampling has been shown here, the modular design allows for easy expansion to include any gene target of choice. This approach can be further expanded and complemented further to simultaneously detect both genes and proteins in a combined workflow MALDI IHC and MALDI ISH to cover entire biochemical pathways. Finally, it should be noted that MALDI ISH is not limited to MALDI ionization and that different ionization sources may be easily integrated to this kind of photocleavable mass tagging strategy.

Future efforts will be aimed at optimizing the platform to address the extraneous m/z values that were identified due to the photolysis reaction. The simplicity and compatibility of the synthesis were attractive options; however, the additional m/z values generated from the cyclization and loss of water was not anticipated and complicate the design of the peptide mass tags. Additionally, the synthesis and purification of the peptide mass tags can be improved, eliminating truncation products and helping reduce the possible number of overlapping m/z values. We hope the system can further what we are then able to further investigate more complex systems and diseases, helping reveal unseen relationships between gene expression and metabolomic signatures.

Supplementary Material

ac5c02567_si_001.pdf (9.8MB, pdf)

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.analchem.5c02567.

  • Additional experimental data, including MALDI spot tests and MS/MS spectra of peptide mass tags, analysis of photocleavage kinetics, LC-MS characterization of CuAAC products, validation of MALDI ISH probe performance, spatial ion images for individual gene targets, and smFISH comparisons across wild-type and CD mouse brain tissues (PDF)

The authors declare no competing financial interest.

References

  1. Byron S. A., Van Keuren-Jensen K. R., Engelthaler D. M., Carpten J. D., Craig D. W.. Translating RNA Sequencing into Clinical Diagnostics: Opportunities and Challenges. Nat. Rev. Genet. 2016;17(5):257–271. doi: 10.1038/nrg.2016.10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Harvey S. E., Cheng C.. Methods for Characterization of Alternative RNA Splicing. Methods Mol. Biol. 2016;1402:229–241. doi: 10.1007/978-1-4939-3378-5_18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Douglas A. G. L., Wood M. J. A.. RNA Splicing: Disease and Therapy. Briefings in Functional Genomics. 2011;10(3):151–164. doi: 10.1093/bfgp/elr020. [DOI] [PubMed] [Google Scholar]
  4. Zhang L., Chen D., Song D., Liu X., Zhang Y., Xu X., Wang X.. Clinical and Translational Values of Spatial Transcriptomics. Signal Transduction and Targeted Therapy. 2022;7(1):111. doi: 10.1038/s41392-022-00960-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Jung N., Kim T. K.. Spatial Transcriptomics in Neuroscience. Experimental and Molecular Medicine. 2023;55(10):2105–2115. doi: 10.1038/s12276-023-01093-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Piwecka M., Rajewsky N., Rybak-Wolf A.. Single-Cell and Spatial Transcriptomics: Deciphering Brain Complexity in Health and Disease. Nature Reviews Neurology. 2023;19(6):346–362. doi: 10.1038/s41582-023-00809-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Pinkel D., Landegent J., Collins C., Fuscoe J., Segraves R., Lucas J., Gray J.. Fluorescence in Situ Hybridization with Human Chromosome-Specific Libraries: Detection of Trisomy 21 and Translocations of Chromosome 4. Proc. Nati. Acad. Sci. 1988;85:9138–9142. doi: 10.1073/pnas.85.23.9138. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Enninful A., Baysoy A., Fan R.. Unmixing for Ultra-High-Plex Fluorescence Imaging. Nat. Commun. 2022;13(1):3473. doi: 10.1038/s41467-022-31110-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Xia C., Babcock H. P., Moffitt J. R., Zhuang X.. Multiplexed Detection of RNA Using MERFISH and Branched DNA Amplification. Sci. Rep. 2019;9(1):7721. doi: 10.1038/s41598-019-43943-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Wang F., Flanagan J., Su N., Wang L. C., Bui S., Nielson A., Wu X., Vo H. T., Ma X. J., Luo Y.. RNAscope: A Novel in Situ RNA Analysis Platform for Formalin-Fixed, Paraffin-Embedded Tissues. Journal of Molecular Diagnostics. 2012;14(1):22–29. doi: 10.1016/j.jmoldx.2011.08.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Janesick A., Shelansky R., Gottscho A. D., Wagner F., Williams S. R., Rouault M., Beliakoff G., Morrison C. A., Oliveira M. F., Sicherman J. T., Kohlway A., Abousoud J., Drennon T. Y., Mohabbat S. H., Taylor S. E. B.. High Resolution Mapping of the Tumor Microenvironment Using Integrated Single-Cell, Spatial and in Situ Analysis. Nat. Commun. 2023;14(1):8353. doi: 10.1038/s41467-023-43458-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Walker J. M., Orr M. E., Orr T. C., Thorn E. L., Christie T. D., Yokoda R. T., Vij M., Ehrenberg A. J., Marx G. A., McKenzie A. T., Kauffman J., Selmanovic E., Wisniewski T., Drummond E., White C. L., Crary J. F., Farrell K., Kautz T. F., Daoud E. V., Richardson T. E.. Spatial Proteomics of Hippocampal Subfield-Specific Pathology in Alzheimer’s Disease and Primary Age-Related Tauopathy. Alzheimer's and Dementia. 2024;20:783. doi: 10.1002/alz.13484. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. He S., Bhatt R., Brown C., Brown E. A., Buhr D. L., Chantranuvatana K., Danaher P., Dunaway D., Garrison R. G., Geiss G., Gregory M. T., Hoang M. L., Khafizov R., Killingbeck E. E., Kim D., Kim T. K., Kim Y., Klock A., Korukonda M., Kutchma A., Lewis Z. R., Liang Y., Nelson J. S., Ong G. T., Perillo E. P., Phan J. C., Phan-Everson T., Piazza E., Rane T., Reitz Z., Rhodes M., Rosenbloom A., Ross D., Sato H., Wardhani A. W., Williams-Wietzikoski C. A., Wu L., Beechem J. M.. High-Plex Imaging of RNA and Proteins at Subcellular Resolution in Fixed Tissue by Spatial Molecular Imaging. Nat. Biotechnol. 2022;40(12):1794–1806. doi: 10.1038/s41587-022-01483-z. [DOI] [PubMed] [Google Scholar]
  14. Spraggins J. M., Caprioli R. M.. High-Speed MALDI-TOF Imaging Mass Spectrometry: Rapid Ion Image Acquisition and Considerations for Next Generation Instrumentation. J. Am. Soc. Mass Spectrom. 2011;22(6):1022–1031. doi: 10.1007/s13361-011-0121-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Angerer T. B., Bour J., Biagi J.-L., Moskovets E., Frache G.. Evaluation of 6 MALDI-Matrices for 10 Μm Lipid Imaging and On-Tissue MSn with AP-MALDI-Orbitrap. J. Am. Soc. Mass Spectrom. 2022;33(5):760–771. doi: 10.1021/jasms.1c00327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Müller M. A., Kompauer M., Strupat K., Heiles S., Spengler B.. Implementation of a High-Repetition-Rate Laser in an AP-SMALDI MSI System for Enhanced Measurement Performance. J. Am. Soc. Mass Spectrom. 2021;32(2):465–472. doi: 10.1021/jasms.0c00368. [DOI] [PubMed] [Google Scholar]
  17. 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. doi: 10.1038/s41592-019-0536-2. [DOI] [PubMed] [Google Scholar]
  18. 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. doi: 10.1038/nmeth.4071. [DOI] [PubMed] [Google Scholar]
  19. Ma S., Leng Y., Li X., Meng Y., Yin Z., Hang W.. High Spatial Resolution Mass Spectrometry Imaging for Spatial Metabolomics: Advances, Challenges, and Future Perspectives. TrAC Trends in Analytical Chemistry. 2023;159:116902. doi: 10.1016/j.trac.2022.116902. [DOI] [Google Scholar]
  20. Vicari M., Mirzazadeh R., Nilsson A., Shariatgorji R., Bjärterot P., Larsson L., Lee H., Nilsson M., Foyer J., Ekvall M., Czarnewski P., Zhang X., Svenningsson P., Käll L., Andrén P. E., Lundeberg J.. Spatial Multimodal Analysis of Transcriptomes and Metabolomes in Tissues. Nat. Biotechnol. 2024;42:1046. doi: 10.1038/s41587-023-01937-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. 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. doi: 10.1038/s41564-019-0664-6. [DOI] [PubMed] [Google Scholar]
  22. Lemaire R., Stauber J., Wisztorski M., Van Camp C., Desmons A., Deschamps M., Proess G., Rudlof I., Woods A. S., Day R., Salzet M., Fournier I.. Tag-Mass: Specific Molecular Imaging of Transcriptome and Proteome by Mass Spectrometry Based on Photocleavable Tag. J. Proteome Res. 2007;6(6):2057–2067. doi: 10.1021/pr0700044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Yagnik G., Liu Z., Rothschild K. J., Lim M. J.. Highly Multiplexed Immunohistochemical MALDI-MS Imaging of Biomarkers in Tissues. J. Am. Soc. Mass Spectrom. 2021;32(4):977–988. doi: 10.1021/jasms.0c00473. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Lim M. J., Yagnik G., Henkel C., Frost S. F., Bien T., Rothschild K. J.. MALDI HiPLEX-IHC: Multiomic and Multimodal Imaging of Targeted Intact Proteins in Tissues. Front. Chem. 2023;11:1182404. doi: 10.3389/fchem.2023.1182404. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Claes B. S. R., Krestensen K. K., Yagnik G., Grgic A., Kuik C., Lim M. J., Rothschild K. J., Vandenbosch M., Heeren R. M. A.. MALDI-IHC-Guided In-Depth Spatial Proteomics: Targeted and Untargeted MSI Combined. Anal. Chem. 2023;95(4):2329–2338. doi: 10.1021/acs.analchem.2c04220. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Le M. U. T., Shon H. K., Nguyen H. P., Lee C. H., Kim K. S., Na H. K., Lee T. G.. Simultaneous Multiplexed Imaging of Biomolecules in Transgenic Mouse Brain Tissues Using Mass Spectrometry Imaging: A Multi-Omic Approach. Anal. Chem. 2022;94(26):9297–9305. doi: 10.1021/acs.analchem.2c00676. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Keren L., Bosse M., Thompson S., Risom T., Vijayaragavan K., Mccaffrey E., Marquez D., Angoshtari R., Greenwald N. F., Fienberg H., Wang J., Kambham N., Kirkwood D., Nolan G., Montine T. J., Galli S. J., West R., Bendall S. C., Angelo M.. MIBI-TOF: A Multiplexed Imaging Platform Relates Cellular Phenotypes and Tissue Structure. Sci. Adv. 2019;5:eaax5851. doi: 10.1126/sciadv.aax5851. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Song X., Zang Q., Li C., Zhou T., Zare R. N.. Immuno-Desorption Electrospray Ionization Mass Spectrometry Imaging Identifies Functional Macromolecules by Using Microdroplet-Cleavable Mass Tags. Angew. Chem., Int. Ed. 2023;62(9):e202216969. doi: 10.1002/anie.202216969. [DOI] [PubMed] [Google Scholar]
  29. Thiery G., Shchepinov M. S., Southern E. M., Audebourg A., Audard V., Terris B., Gut I. G.. Multiplex Target Protein Imaging in Tissue Sections by Mass Spectrometry - TAMSIM. Rapid Commun. Mass Spectrom. 2007;21(6):823–829. doi: 10.1002/rcm.2895. [DOI] [PubMed] [Google Scholar]
  30. Seuma J., Bunch J., Cox A., McLeod C., Bell J., Murray C.. Combination of Immunohistochemistry and Laser Ablation ICP Mass Spectrometry for Imaging of Cancer Biomarkers. PROTEOMICS. 2008;8(18):3775–3784. doi: 10.1002/pmic.200800167. [DOI] [PubMed] [Google Scholar]
  31. Hutchinson R. W., Cox A. G., McLeod C. W., Marshall P. S., Harper A., Dawson E. L., Howlett D. R.. Imaging and Spatial Distribution of β-Amyloid Peptide and Metal Ions in Alzheimer’s Plaques by Laser Ablation-Inductively Coupled Plasma-Mass Spectrometry. Anal. Biochem. 2005;346(2):225–233. doi: 10.1016/j.ab.2005.08.024. [DOI] [PubMed] [Google Scholar]
  32. Bell J. M., Yagnik G., Dettori L. G., Carvalho P., Wan Z., Rothschild K. J., Lim M. J.. Photocleavable Mass-Tagged Oligonucleotide Probes for Multiplexed and Multiomic Tissue Imaging of Targeted Transcripts. J. Am. Soc. Mass Spectrom. 2025;36:1621. doi: 10.1021/jasms.5c00057. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Marmur J., Ts’o P. O. P.. Denaturation of Deoxyribonucleic Acid by Formamide. Biochim. Biophys. Acta. 1961;51(1):32–36. doi: 10.1016/0006-3002(61)91013-7. [DOI] [PubMed] [Google Scholar]
  34. Sud M., Fahy E., Cotter D., Brown A., Dennis E. A., Glass C. K., Merrill A. H., Murphy R. C., Raetz C. R. H., Russell D. W., Subramaniam S.. LMSD: LIPID MAPS Structure Database. Nucleic Acids Res. 2007;35:D527–D532. doi: 10.1093/nar/gkl838. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Breci L. A., Tabb D. L., Yates J. R., Wysocki V. H.. Cleavage N-Terminal to Proline: Analysis of a Database of Peptide Tandem Mass Spectra. Anal. Chem. 2003;75(9):1963–1971. doi: 10.1021/ac026359i. [DOI] [PubMed] [Google Scholar]
  36. Tabb D. L., Huang Y., Wysocki V. H., Yates J. R.. Influence of Basic Residue Content on Fragment Ion Peak Intensities in Low-Energy Collision-Induced Dissociation Spectra of Peptides. Anal. Chem. 2004;76(5):1243–1248. doi: 10.1021/ac0351163. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Seidler J., Zinn N., Boehm M. E., Lehmann W. D.. De Novo Sequencing of Peptides by MS/MS. PROTEOMICS. 2010;10(4):634–649. doi: 10.1002/pmic.200900459. [DOI] [PubMed] [Google Scholar]
  38. Moses L., Pachter L.. Museum of Spatial Transcriptomics. Nat. Methods. 2022;19(5):534–546. doi: 10.1038/s41592-022-01409-2. [DOI] [PubMed] [Google Scholar]
  39. Sayers E. W., Beck J., Bolton E. E., Brister J. R., Chan J., Connor R., Feldgarden M., Fine A. M., Funk K., Hoffman J., Kannan S., Kelly C., Klimke W., Kim S., Lathrop S., Marchler-Bauer A., Murphy T. D., O’Sullivan C., Schmieder E., Skripchenko Y., Stine A., Thibaud-Nissen F., Wang J., Ye J., Zellers E., Schneider V. A., Pruitt K. D.. Database Resources of the National Center for Biotechnology Information in 2025. Nucleic Acids Res. 2025;53(D1):D20–D29. doi: 10.1093/nar/gkae979. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Pruitt K. D., Harrow J., Harte R. A., Wallin C., Diekhans M., Maglott D. R., Searle S., Farrell C. M., Loveland J. E., Ruef B. J., Hart E., Suner M.-M., Landrum M. J., Aken B., Ayling S., Baertsch R., Fernandez-Banet J., Cherry J. L., Curwen V., DiCuccio M., Kellis M., Lee J., Lin M. F., Schuster M., Shkeda A., Amid C., Brown G., Dukhanina O., Frankish A., Hart J., Maidak B. L., Mudge J., Murphy M. R., Murphy T., Rajan J., Rajput B., Riddick L. D., Snow C., Steward C., Webb D., Weber J. A., Wilming L., Wu W., Birney E., Haussler D., Hubbard T., Ostell J., Durbin R., Lipman D.. The Consensus Coding Sequence (CCDS) Project: Identifying a Common Protein-Coding Gene Set for the Human and Mouse Genomes. Genome Res. 2009;19(7):1316–1323. doi: 10.1101/gr.080531.108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Rychlik W., Rhoads R. E.. A Computer Program for Choosing Optimal Oligonudeotides for Filter Hybridization, Sequencing and in Vitro Amplification of DNA. Nucleic Acids Res. 1989;17(21):8543–8551. doi: 10.1093/nar/17.21.8543. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Holmes C. P.. Model Studies for New O-Nitrobenzyl Photolabile Linkers: Substituent Effects on the Rates of Photochemical Cleavage. J. Org. Chem. 1997;62(8):2370–2380. doi: 10.1021/jo961602x. [DOI] [PubMed] [Google Scholar]
  43. Qvortrup K., Nielsen T. E.. A Photolabile Linker for the Solid-Phase Synthesis of 4-Substituted NH-1,2,3-Triazoles. Chem. Commun. 2011;47(11):3278–3280. doi: 10.1039/c0cc05274d. [DOI] [PubMed] [Google Scholar]
  44. Okamura H., Iida M., Kaneyama Y., Nagatsugi F.. O-Nitrobenzyl Oxime Ethers Enable Photoinduced Cyclization Reaction to Provide Phenanthridines under Aqueous Conditions. Org. Lett. 2023;25(3):466–470. doi: 10.1021/acs.orglett.2c04015. [DOI] [PubMed] [Google Scholar]
  45. Kalbag S. M., Roeske R. W.. Photolabile Protecting Group for Histidine. J. Am. Chem. Soc. 1975;97(2):440–441. doi: 10.1021/ja00835a046. [DOI] [PubMed] [Google Scholar]
  46. Bochet C. G.. Photolabile Protecting Groups and Linkers. J. Chem. Soc., Perkin Trans. 1. 2002;(2):125–142. doi: 10.1039/b009522m. [DOI] [Google Scholar]
  47. Chen L.-J., Burka L. T.. Formation of o-Nitrosobenzaldehyde from Hydrolysis of o-Nitrobenzyl Tosylate. Evidence of Intramolecular Nucleophilic Interaction. Tetrahedron Lett. 1998;39(30):5351–5354. doi: 10.1016/S0040-4039(98)01068-5. [DOI] [Google Scholar]
  48. Gaplovsky M., Il’ichev Y. V., Kamdzhilov Y., Kombarova S. V., Mac M., Schwörer M. A., Wirz J.. Photochemical Reaction Mechanisms of 2-Nitrobenzyl Compounds: 2-Nitrobenzyl Alcohols Form 2-Nitroso Hydrates by Dual Proton Transfer. Photochem. Photobiol. Sci. 2005;4(1):33–42. doi: 10.1039/b409927c. [DOI] [PubMed] [Google Scholar]
  49. Zuman P., Shah B.. Addition, Reduction, and Oxidation Reactions of Nitrosobenzene. Chem. Rev. 1994;94(6):1621–1641. doi: 10.1021/cr00030a007. [DOI] [Google Scholar]
  50. Asakawa D., Calligaris D., Smargiasso N., De Pauw E.. Ultraviolet Laser Induced Hydrogen Transfer Reaction: Study of the First Step of MALDI In-Source Decay Mass Spectrometry. J. Phys. Chem. B. 2013;117(8):2321–2327. doi: 10.1021/jp311464k. [DOI] [PubMed] [Google Scholar]
  51. Smargiasso N., Quinton L., De Pauw E.. 2-Aminobenzamide and 2-Aminobenzoic Acid as New MALDI Matrices Inducing Radical Mediated In-Source Decay of Peptides and Proteins. J. Am. Soc. Mass Spectrom. 2012;23(3):469–474. doi: 10.1007/s13361-011-0307-5. [DOI] [PubMed] [Google Scholar]
  52. Mathews D. H., Sabina J., Zuker M., Turner D. H.. Expanded Sequence Dependence of Thermodynamic Parameters Improves Prediction of RNA Secondary Structure1. J. Mol. Biol. 1999;288(5):911–940. doi: 10.1006/jmbi.1999.2700. [DOI] [PubMed] [Google Scholar]
  53. Lein E. S., Hawrylycz M. J., Ao N., Ayres M., Bensinger A., Bernard A., Boe A. F., Boguski M. S., Brockway K. S., Byrnes E. J., Chen L., Chen L., Chen T.-M., Chi Chin M., Chong J., Crook B. E., Czaplinska A., Dang C. N., Datta S., Dee N. R., Desaki A. L., Desta T., Diep E., Dolbeare T. A., Donelan M. J., Dong H.-W., Dougherty J. G., Duncan B. J., Ebbert A. J., Eichele G., Estin L. K., Faber C., Facer B. A., Fields R., Fischer S. R., Fliss T. P., Frensley C., Gates S. N., Glattfelder K. J., Halverson K. R., Hart M. R., Hohmann J. G., Howell M. P., Jeung D. P., Johnson R. A., Karr P. T., Kawal R., Kidney J. M., Knapik R. H., Kuan C. L., Lake J. H., Laramee A. R., Larsen K. D., Lau C., Lemon T. A., Liang A. J., Liu Y., Luong L. T., Michaels J., Morgan J. J., Morgan R. J., Mortrud M. T., Mosqueda N. F., Ng L. L., Ng R., Orta G. J., Overly C. C., Pak T. H., Parry S. E., Pathak S. D., Pearson O. C., Puchalski R. B., Riley Z. L., Rockett H. R., Rowland S. A., Royall J. J., Ruiz M. J., Sarno N. R., Schaffnit K., Shapovalova N. V., Sivisay T., Slaughterbeck C. R., Smith S. C., Smith K. A., Smith B. I., Sodt A. J., Stewart N. N., Stumpf K.-R., Sunkin S. M., Sutram M., Tam A., Teemer C. D., Thaller C., Thompson C. L., Varnam L. R., Visel A., Whitlock R. M., Wohnoutka P. E., Wolkey C. K., Wong V. Y., Wood M., Yaylaoglu M. B., Young R. C., Youngstrom B. L., Feng Yuan X., Zhang B., Zwingman T. A., Jones A. R.. Genome-Wide Atlas of Gene Expression in the Adult Mouse Brain. Nature. 2007;445(7124):168–176. doi: 10.1038/nature05453. [DOI] [PubMed] [Google Scholar]
  54. Qiu S., Palavicini J. P., Wang J., Gonzalez N. S., He S., Dustin E., Zou C., Ding L., Bhattacharjee A., Van Skike C. E., Galvan V., Dupree J. L., Han X.. Adult-Onset CNS Myelin Sulfatide Deficiency Is Sufficient to Cause Alzheimer’s Disease-like Neuroinflammation and Cognitive Impairment. Mol. Neurodegener. 2021;16(1):64. doi: 10.1186/s13024-021-00488-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Hannun Y. A., Obeid L. M.. Sphingolipids and Their Metabolism in Physiology and Disease. Nat. Rev. Mol. Cell Biol. 2018;19(3):175–191. doi: 10.1038/nrm.2017.107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Hirahara Y., Wakabayashi T., Mori T., Koike T., Yao I., Tsuda M., Honke K., Gotoh H., Ono K., Yamada H.. Sulfatide Species with Various Fatty Acid Chains in Oligodendrocytes at Different Developmental Stages Determined by Imaging Mass Spectrometry. Journal of Neurochemistry. 2017;140(3):435–450. doi: 10.1111/jnc.13897. [DOI] [PubMed] [Google Scholar]
  57. Ahmed S. S., Gao G.. Making the White Matter Matters: Progress in Understanding Canavan’s Disease and Therapeutic Interventions Through Eight Decades. JIMD Rep. 2014;19:11–22. doi: 10.1007/8904_2014_356. [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

ac5c02567_si_001.pdf (9.8MB, pdf)

Articles from Analytical Chemistry are provided here courtesy of American Chemical Society

RESOURCES