Protocol for spatial metabolomics and isotope tracing in the mouse brain

Summary

Mass spectrometry imaging enables high-resolution spatial chemical mapping, yet its application for dynamic analysis with tracers poses challenges. Here, we present a protocol for spatial metabolomics and isotope tracing in the mouse brain. We describe steps for tracer administration, tissue collection, and cryosectioning. We then detail procedures for matrix application, ion identification, and data analysis. This protocol delivers high-quality spatial metabolomics data and is well suited for region-specific tracing analysis in the brain.

Graphical abstract

Highlights

• •Step-by-step workflow for spatial metabolomics and isotope tracing in the mouse brain
• •Guidelines for tracer administration, tissue collection, and cryosectioning
• •Optimized matrix deposition and MSI acquisition for high-resolution imaging
• •Standardized data processing for metabolite annotation and ¹³C enrichment analysis

Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics.

Mass spectrometry imaging enables high-resolution spatial chemical mapping, yet its application for dynamic analysis with tracers poses challenges. Here, we present a protocol for spatial metabolomics and isotope tracing in the mouse brain. We describe steps for tracer administration, tissue collection, and cryosectioning. We then detail procedures for matrix application, ion identification, and data analysis. This protocol delivers high-quality spatial metabolomics data and is well suited for region-specific tracing analysis in the brain.

Before you begin

Profiling brain metabolism presents significant challenges due to the brain’s complex and heterogeneous architecture. Each region has a distinct metabolic profile that standard extraction methods struggle to isolate without cross-contamination. Recent studies highlight the importance of spatially resolved metabolomics in overcoming these limitations.¹ This protocol addresses these challenges by using mass spectrometry imaging (MSI) to map the chemical architecture of the mouse brain. By integrating stable-isotope tracing, it enables simultaneous analysis of static and dynamic metabolite states while preserving tissue anatomy and chemical integrity, avoiding the need for labor-intensive dissection of specific brain regions.

The following steps will guide you through preparing your workspace and reagents for tracer administration, dissecting and sectioning brain tissue, performing chemical matrix deposition, and preparing standards for ion identification. Tissue collection and sectioning must be conducted on a cold surface to maintain anatomical and metabolite integrity. Proper setup, along with meticulous handling of samples, is essential to ensure the success and reproducibility of this protocol.

Institutional permissions

The experiments detailed in this protocol were approved by the Institutional Animal Care and Use Committee (IACUC) at the Icahn School of Medicine at Mount Sinai. Researchers must verify that all experimental procedures adhere to their local and national ethical guidelines for animal research. Before replicating any experiments described in this protocol, appropriate institutional approvals should be obtained to warrant compliance with ethical standards.

Prepare for isotope tracing administration

Timing: 30 min

• 1.Ice cold, 0.22 μm filtered 1x phosphate buffer solution (PBS).
• 2.[U-¹³C]glucose solution in filtered PBS at a concentration of 66.6 mg/mL.
• 3.Blood glucometer with strips.
• 4.A scale.

Prepare for dissection

Timing: 30 min

• 5.Clean the workspace with isopropanol alcohol for disinfection.
• 6.Chill all buffers and equipment on ice.
• 7.Tissue collection should be performed in a previously cleaned laminar flow cabinet.
• 8.A stereo microscope should be ready and clean inside the laminar flow cabinet (Figure 1A).
• 9.Standard dissection tools (i.e., forceps, iris scissors, scalpel, spatula, razor blade).
• 10.Isoflurane anesthesia gas chamber.
• 11.Labeled embedding mold.
• 12.Icebox containing dry ice pellets.

Figure 1.

Tissue collection and preparation for sectioning

(A) Dissection setup.

(B) Measuring blood glucose before ¹³C-glucose injection.

(C) Intraperitoneal injection of the ¹³C tracer.

(D) Dissected brain illustration.

(E) Separating brain hemispheres with a blade.

(F) Midline region of the hemisphere after separation.

(G) Hemisphere placement in the molding block.

(H) Tissue freezing on dry ice.

Prepare for sectioning

Timing: 30 min

• 13.Use a voltmeter to determine the conductivity side of the indium titanium oxide coated (ITO) slide (Figure 2).
• 14.Label on the back of the slides to avoid interference with the coating. Examples of labels include sample ID, date, slide number, section name.
• 15.Label 50 mL tubes for slide storage.
• 16.(Optional) Add silica beads and Kim wipes inside each tube to help absorb moisture during storage in −80°C or shipping to another facility.
• 17.Standard section tools (i.e., forceps, razor blades, brushes, spatula, OCT, and a mirror).
• 18.Icebox containing dry ice pellets.
• 19.Carefully transfer the embedding mold containing brain tissue on dry ice to the cryostat room.

Note: Ensure the tissue is not exposed to ambient temperature for more than 2 min at any time to prevent potential degradation of metabolites.

• 20.Verify that the cryostat is functioning properly and set the temperature to −20°C.

Figure 2.

Sectioning the brain tissue

(A) Tissue placement in the cryostat chamber.

(B) Application of OCT to the chuck.

(C) Positioning the brain tissue laterally with the midline facing up.

(D) Checking ITO slide conductivity.

(E) Tissue sectioning.

(F) Mounting tissue on the conductive side of the ITO slide.

(G) Warming the tissue slightly with a finger to aid attachment.

(H) Storing tissue in a 50 mL tube on dry ice before freezing at −80°C.

Prepare for MSI identification

Timing: 30 min

• 21.Prepare glass vials with the name of each chemical standard.
• 22.Weigh out 1–3 mg of each chemical standard.
• 23.Add in 50% acetonitrile with 0.1% trifluoroacetic acid (TFA) to each standard to make a solution that is at a final concentration of 10 μg/μL.

Reagent	Calculated m/z	Detected m/z [M + H]	Final concentration	Amount
50% Acetonitrile with 0.1% TFA				1500 μL
CHCA			5 mg/mL	500 μL
GABA	104.0703	104.0701	10 μg/μL	200 μL
Glutamic acid	148.0602	148.0603	10 μg/μL	200 μL
Glutamine	147.0761	147.0764	10 μg/μL	200 μL

• 24.
  To confirm m/z information for standards:

  • a.
    Spot 0.5 μL of prepared α-cyano-4-hydroxycinnamic acid (CHCA) matrix with 0.5 μL of desired standard on a stainless-steel target or ITO coated glass slide.
  • b.
    Allow the sample to dry fully before analysis.
• 25.Analyze dried standard with the same parameters listed in part 5 (MSI operation and data acquisition).

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Chemicals, peptides, and recombinant proteins
α-Cyano-4-hydroxycinnamic acid (CHCA)	Sigma-Aldrich	Cat #C2020
y-Aminobutyric acid (GABA)	Sigma-Aldrich	Cat # A2129
Glutamic acid	Sigma-Aldrich	Cat #G1251
Glutamine	Sigma-Aldrich	Cat #G3126
Acetonitrile LCMS grade	Honeywell B&J	Cat # LC015-4
Trifluoroacetic acid (TFA)	Sigma-Aldrich	Cat #T6508
PBS - phosphate-buffered saline (10X) pH 7.4, RNase-free	Thermo Fisher Scientific	Cat # AM9624
D-glucose (U-13C6, 99%)	Cambridge Isotope Laboratories, Inc.	Cat # CLM-1396-PK
Isoflurane, USP	Piramal Pharma Ltd.	Cat # 66794-013-25
Tissue-Tek O.C.T. compound	Sakura	Cat # 4583
Pure white silica gel beads	Dry & Dry
BD Lo-Dose U-100 Insulin syringes	Fisher	Cat# 14-826-79
Experimental models: Organisms/strains
C57BL/6 strain mice, males/females, P21	Jackson Laboratory	000664
Software and algorithms
IMAGEREVEAL MS	Shimadzu Corporation	https://www.shimadzu.com/an/products/life-science-lab-instruments/imaging/imagereveal-ms/index.html
BioRender	BioRender	https://www.biorender.com
R version 4.1.1	Posit	https://posit.co/

Step-by-step method details

Part 1. Injection of tracers

Timing: 6 h

This section details tracer preparation and administration. Proper fasting, accurate dosing, and blood glucose monitoring help achieve effective tracer uptake and data reproducibility.

Note: This protocol uses [U-¹³C]glucose, the brain’s primary energy substrate, because its fully label form allows tracking of key metabolic pathways, including glycolysis, the Krebs cycle, and anaplerosis, as most glucose-derived metabolites incorporate multiple ¹³C atoms at specific positions. Additionally, [U-¹³C]glucose minimizes interference from naturally abundant ¹³C, as most downstream metabolites contain two or more labeled carbons.

• 1.Fast the mice for 5 h.

Note: Avoid <4 h because animals could still be post-prandial.

Note: To carry out this protocol, we used three mice (one male and two females) aged P21. This sample size was chosen based on the protocol's high reproducibility among biological replicates, as discussed in the Expected outcome section, such that it ensures reliable metabolic measurements with minimal variability.

• 2.Weigh the animals to calculate tracer solution volume to be injected (2 mg/g). The required volume is calculated as follows:

Volume (mL) = weight (g) / 33.3 (g/mL)

• 3.Put the total amount of tracer solution in a 1.5 mL microcentrifuge tube and take it from there with a low-dosage syringe for animal injection.

Note: Avoid the formation of bubbles.

• 4.Check blood glucose in the tail (Figure 1B).
• 5.Inject 2 mg/g of weight of [U-¹³C]glucose solution intraperitoneally (Figure 1C).
• 6.Wait for 30 min after the injection.
• 7.Measure blood glucose in the tail and quickly proceed with tissue collection.

Part 2. Tissue collection

Timing: 5 min

This section outlines rapid brain extraction and storage to maintain tissue integrity for metabolomic analysis. Proper handling and quick-freezing help minimize metabolite and tissue degradation and support high-quality MSI data.

• 8.Put the animal in a gas chamber with an adequate dosage of isoflurane 5% at a flow rate of 5% v/v mixed with 25–50% O₂, for about 2 min.
• 9.Carefully decapitate the animal and drop the head of the animal in ice cold water to reduce brain’s temperature.
• 10.(Optional) Collect blood from the cervical stump into a labeled microcentrifuge tube containing 25 μL of heparin.

Note: We recommend measuring [U-¹³C]glucose enrichment in blood using a standard protocol, as we described recently.² The goal is to check that the percentage of [U-¹³C]glucose in plasma exceeds 30%, which, in our experience, is optimal for detecting sufficient ¹³C signal from [U-¹³C]glucose-derived metabolites in the brain.

• 11.In a dissection plate, quickly open the skull and remove the brain (Figure 1D).
• 12.Separate the two hemispheres with a blade (Figures 1E and 1F).
• 13.Carefully place them in the embedding molds (Figure 1G).
• 14.Place the embedding molds in dry ice pellets to allow them to rapidly freeze (Figure 1H).
• 15.Wrap the molds with parafilm paper and store them at −80°C until further use.

Pause Point: Brain hemispheres can be stored at −80°C for up to four months. For longer storage, sample quality should be evaluated by comparing the data with previously obtained MSI data from fresh tissue to assess potential analyte intensity loss.

CRITICAL: Handle the brain tissue with extreme care, especially during extraction and placement in the embedding mold, to preserve its anatomy, as any disruption may compromise the quality and accuracy of the MSI data.

Part 3. Sectioning

Timing: 1.5 h

This section describes the cryosectioning method to obtain high-quality brain tissue slices. Proper tissue orientation, precise cutting at 10–15 μm, and careful handling minimize artifacts and preserve tissue anatomical features.

• 16.
  Tissue mounting and positioning:

  • a.
    Transfer the embedding mold containing brain tissue samples and labeled slides into the cryostat chamber, then carefully remove the parafilm from the embedding mold (Figure 2A).
  • b.
    Allow the tissue and slides to acclimate to the cryostat temperature for at least 10 min.

    Note: Avoid leaving the tissue in the chamber overnight, as most cryostats have a defrost cycle that may cause the temperature to briefly rise to room temperature before returning to −20°C.

  • c.
    Place a labeled, empty 50 mL tube on dry ice for later slide collection.
  • d.
    Drop a small amount of OCT compound on the center of the chuck to create a stable base for the tissue (Figure 2B).
  • e.
    Using forceps, place the brain hemisphere onto the OCT compound with the cutting surface facing up (Figure 2C).
  • f.
    Ensure the brain is centered and properly oriented to achieve even sectioning.
  • g.
    Verify the orientation of the conductive side of the ITO slide to avoid mishaps (Figure 2D).
• 17.
  Tissue sectioning:

  • a.
    Begin slicing by rotating the handle clockwise until a smooth, even surface is achieved (Figure 2E).

    Note: The recommended thickness for imaging is between 10-15 μm.

  • b.
    After collecting each section, use a brush to remove any excess tissue debris.
  • c.
    Close the collecting mirror before slicing for collection.
  • d.
    Use a brush to gently unfold or straighten the tissue on the slide if necessary.
  • e.
    Gently tap a spare slide on the tissue (Figure 2F).
  • f.
    Warm the back of the slide with your hand to help the tissue adhere to the slide surface (Figure 2G).
  • g.
    Evaluate the quality of the section.

    Note: Ensure all target anatomical features are clearly visible, for example hippocampus or cerebellum, with no folds, tears, or significant distortions.

• 18.
  Collect sections:

  • a.
    Once quality is confirmed, begin collecting sections.
  • b.
    (Optional) Alternate collection between frosty slides and ITO-slides to ensure adjacent tissue is available for MSI and histology.
  • c.
    When collecting more than one brain on a single slide, complete one brain before starting the next to minimize setup time and reduce potential cross-contamination.
  • d.
    Clean any surface touched by the tissue with pure ethanol.
  • e.
    Place the collected slides in the 50 mL tubes on dry ice (Figure 2H).

    Note: Store up to two slides per tube, making sure that the tissue surface faces away from each other to prevent damage.

  • f.
    Transfer the tube to dry ice for transportation and store them in a −80°C freezer.
• 19.
  Tissue storage and cleaning up:

  • a.
    After finishing the sectioning, remove the remaining brain tissue from the chuck and return it to its embedding mold.

    Note: Based on our experience, it is preferable to collect all sections at once rather than at different time points, as this minimizes tissue damage caused by temperature fluctuations during transfers between the −80°C freezer and the −20°C cryostat chamber.

  • b.
    Quickly wrap the block in parafilm, making sure the label remains visible, and place it on dry ice for transfer to −80°C freezer.
  • c.
    Safely dispose of the razor blade in a sharp container.
  • d.
    Remove the collecting mirror, chuck, and all tools from the cryostat chamber.
  • e.
    Clean the chamber thoroughly with pure ethanol.
  • f.
    Transfer samples and collected tubes via dry ice to the −80°C freezer.

    Pause Point: Sectioning can immediately follow tissue collection or start from tissues stored at −80°C. Completed slides can be kept at −80°C for up to four months, as indicated earlier.

    CRITICAL: Check that sections are of high quality, with target anatomical features clearly visible and free from folds, tears, tissue distortions, or OCT residue on top of the section, as these issues can negatively impact the quality of the MSI data.

Part 4. Matrix deposition

Timing: 1.5 h

This section details the application of a CHCA matrix to tissue sections, which is essential for optimal ionization during MSI. Proper sample handling, uniform deposition, and condensation prevention help preserve metabolite integrity and maintain signal consistency.

Note: Matrix deposition should be performed right after sectioning or after samples have been properly stored in the −80°C freezer.

• 20.Remove the slides from −80°C freezer and place them in a vacuum desiccator for 30 min to allow them to warm to room temperature.
• 21.Make sure there is no condensation on the slide before proceeding to matrix application (Figure 3A).
• 22.
  (Optional) Prepare a spare sample for ion identification.

  • a.
    Spot 0.3 μL of each standard on the tissue or on a slide using subsequent spotting of the standard and a solution of 5 mg/mL of CHCA in 50% acetonitrile with 0.1% TFA.
  • b.
    Allow the solution to fully dry.
• 23.Weigh out 300 mg (± 10 mg) of CHCA and place it in the center of the iMLayer matrix deposition boat (Figure 3B).
• 24.Spread the matrix evenly into a uniform layer (Figure 3C).
• 25.Attach the top cover for the matrix boat and secure it by tightening the screw at the top.
• 26.Place the matrix boat and slide into the iMLayer with the tissue side facing the matrix boat (Figure 3D).
• 27.Use the green LED feature to confirm that the light is targeting the glass slide and not the tissue.
• 28.Close the door of the iMLayer and select the method for CHCA, 0.7 μm thickness.
• 29.Click “Start” on the front display (Figure 3E).
• 30.After 30 min, wait for the instrument to return to room temperature (Figure 3F).
• 31.Remove the slide from the instrument (Figures 3G and 3H).

CRITICAL: Make sure that the tissue slide remains free of condensation before matrix deposition, as this can damage the tissue and result in the loss of soluble metabolites.

Figure 3.

Matrix deposition

(A) ITO slide before processing.

(B and C) Loading the matrix into the matrix deposition boat for even distribution.

(D) Placing the matrix and ITO slide side by side in the IMLayer.

(E) Starting the matrix deposition process after preparing settings.

(F) Checking the chamber temperature before opening to avoid burns.

(G) ITO slide after matrix deposition.

(H) Placing the slide in the appropriate cassette and (I) inserting it into the IMScope.

Part 5. MSI operation and data acquisition

Timing: 3–5 h

This section describes the acquisition of MSI data for spatial metabolite detection. Proper sample alignment, optimized laser settings, and precisely defined scanning regions are essential for high data resolution and reproducibility.

Note: This step should be performed within 0.5–1.5 h after matrix application.

• 32.Insert the matrix coated sample into the iMScope QT sample holder and insert it into the iMScope QT instrument (Figure 3I).
• 33.Perform the tilt correction and sample stage alignment to ensure proper positioning.
• 34.Acquire a wide field microscope image of the tissue, including the standards spotted on it.
• 35.Focus the microscope at 40x at the center of the tissue for optimal clarity.
• 36.Navigate to the acquisition tab and use the free-form shape icon to annotate the tissue region where data acquisition is desired. For example, outline the entire tissue region to ensure it is fully included.
• 37.Define a rectangular scanning area that fully covers the entire tissue, so the annotated region is completely included in the scan.
• 38.Double click inside the area and adjust the rectangle size in μm unit.
• 39.Set X-axis and Y-axis pitch to 25.0 μm.
• 40.Set method parameters for scanning MSI in positive mode as follows:

Laser firing parameters
Number of Laser Shots	100 shots
Repetition Rate	2000 Hz
Laser Diameter	2
Laser Intensity	55.0
MS Acquisition Parameters
Polarity	Positive
Event Type	MS
Measurement Mode	Normal
Mass Range m/z	70–190

• 41.Select start to begin data acquisition.
• 42.Once the data acquisition is completed, the mass spectrometry image file (IMDX) will be automatically created.
• 43.Click “Reset” to return the sample plate to its original position and remove the sample from the instrument. Place it back in the tube and store it in −80°C.

Part 6. Data analysis and visualization

Timing: 1–3 h

This step outlines the process of metabolite identification, tissue region segmentation, and ¹³C enrichment analysis. Accurate data normalization and ion annotation are key for reliable metabolite detection and precise spatial mapping.

• 44.Launch IMAGEREVEAL software and select the “Image Analysis” option.
• 45.Load the MSI data by using the taskbar on the left side of the screen to select “Add IMDX file”. The mass spectrometry image will show up (Figure 4A, Left).
• 46.Normalize the signal by selecting “Pre-processing setting”.
• 47.Choose TIC for normalization, click “OK”, and execute the “Pre-processing” step.
• 48.Set up the data matrix for ion annotation.
• 49.Under “Data Matrix Setting”, select “Target” as the analysis method and click “Create list”.
• 50.Select “Endogenous metabolites” from the internal library and configure matrix, polarity, and tolerance to CHCA, positive, 0.0001 Da, respectively.
• 51.Click “OK” to accept the setting.
• 52.
  (Optional) Create new data matrix from spotted standards:

  • a.
    Confirm the identity of compounds by comparing the calculated and detected m/z values in the [M + H] form, obtained from the spotted experiment.

    Note: We recommend an acceptable threshold of below 20 ppm to ensure reliable identification.

  • b.
    Navigate to “Settings” and select “Compound Template”.
  • c.
    Click the green + on the left side to create a new compound template and name it.
  • d.
    Click the green + on the right side to add details for each compound.
  • e.
    Enter m/z values detected from previous experiments with standards. Do not check the “Calculate Adduct Ion” box since the m/z is already in [M + H] form.
  • f.
    Click “OK” to save the compound template. The new compound template can now be used to generate a new data matrix.
  • g.
    This approach enables the identification of ion annotations.
• 53.After setting up the data matrix, execute “Data Matrix Calculation”. The raw data matrix can be exported in CSV format.
• 54.
  For segmentation and Uniform Manifold Approximation and Projection (UMAP) analysis (in IMAGEREVEAL MS):

  • a.
    Navigate to “Segmentation Calculation”, select UMAP and click “Execute”.
  • b.
    The UMAP image (Figure 4A, Mid) and UMAP 3D plot (Figure 4A, Right) are generated.
  • c.
    Region of interest (ROI) can be localized using the UMAP of the whole brain.
  • d.
    Navigate to “ROI setting”, import the UMAP image by selecting “Reference Image” and then “Import”.
  • e.
    Use the polygon icon to annotate ROIs (Figure 4B, Left).
  • f.
    Assign “Group Attribute” to label each ROI and click “OK”.

    • i.
      To analyze all ROI at once, keep all ROI selected.
    • ii.
      To analyze each ROI individually unselect all other ROI before continuing to UMAP segmentation.
  • g.
    Re-execute “Data Matrix Calculation” and “Segmentation Calculation”.
  • h.
    The updated UMAP image (Figure 4B, Mid) and UMAP 3D plot (Figure 4B, Right) are generated.
  • i.
    To export the cluster results, go to “Cluster Average Spectrum” tab.
  • j.
    Export each cluster average spectrum result in text format.
• 55.
  For cluster analysis, import matrix data and cluster average spectrum from IMAGEREVEAL MS into RStudio.

  • a.
    Use the RefMet library³ to classify the compounds listed in the matrix data.

    Note: For compounds that cannot be classified using the R script, manually add classification information using the RefMet online tool.

  • b.
    Plot chemical composition of regions such as the entire brain (Figure 4C, Left), cortex (Figure 4C, Mid) and cerebellum (Figure 4C, Right) using the ggplot2 package in RStudio.
• 56.
  For ¹³C enrichment analysis, select the desired metabolites and their corresponding isotopologues (Figure 4D).

  Note: We chose glutamate, glutamine, and GABA because their masses were confirmed using the previously mentioned standards and because they are highly enriched from [U-¹³C]glucose in the brain.

  • a.
    Use “Add MS image” to extract MSI data.
  • b.
    Extract images for each isotopologue (M0, M1, etc.) using the annotation library from Shimadzu IMAGEREVEAL MS (endogenous metabolites) (Figure 4E).
  • c.
    ¹³C natural abundance was corrected using the IsocorrectoR R package.⁴
  • d.
    Use Cardinal R library⁵ to process the peaks and R programming software to analyze and visualize the data (Figure 4E and 4F).
  • e.
    A full description of the data analysis pipeline in R can be found in this GitHub repository (https://github.com/Neurometabolomics/Protocol_Spatial_Neurometabolomics).

    Note: This analysis compares ¹³C enrichment data from MSI with data from mouse brain tissue extracts obtained using the same [U-¹³C]glucose injection protocol and analyzed by GC-MS, as we previously described² (illustrated in Figure 4F).

Figure 4.

Analysis of MSI data

(A) Metabolite distribution by signal intensity (left panel, red: high intensity, blue: low intensity), UMAP analysis (middle panel), segmentation distribution (right panel).

(B) Region of interest of the cortex and cerebellum (left panel), UMAP analysis of cortex (top-middle panel) and cerebellum (bottom-middle panel) region, segmentation distribution of cortex (top-right panel) and cerebellum (bottom-right panel) region.

(C) Chemical composition for each cluster of the entire brain (left panel), cortex (middle panel), and cerebellum (right panel).

(D) Schematic of [U-¹³C]glucose metabolism and ¹³C tracer distribution across downstream metabolites.

(E) Anatomical distribution of M0 (unlabeled) and M2 (labeled with two ¹³C carbons) of key metabolites derived from [U-¹³C]glucose, with enrichment analysis shown in the heatmap.

(F) M0 and M2 isotopologue enrichment analysis of glutamate, glutamine, and GABA from MSI (N = 3 mice), as well as cerebellum and cortex extracts (N = 8) analyzed by GC-MS, from P21 mice administered the same [U-¹³C]glucose injection protocol. The Pearson correlation analysis (top-left panel, R = 0.86) between MSI and GC-MS is also illustrated, demonstrating the strong agreement between both methods. An unpaired Student’s t-test was used to perform statistical comparisons between groups. ∗p < 0.05, ∗∗p < 0.01.

Expected outcomes

Using this protocol, we are able to produce high-resolution images of the mouse brain’s chemical composition, measure absolute abundances of metabolites within specific regions of interest, and analyze the ¹³C enrichment of targeted metabolites, as shown in the data analysis and visualization section (Figures 4A–4F). The UMAP image displays well-defined clusters that represent the chemical profiles of distinct brain structures—such as the cortex and cerebellum—while highlighting heterogeneous and region-specific chemical distributions (Figures 4B and 4C). This MSI protocol preserves the brain’s chemical integrity and clearly delineates its morphology, with anatomical features appearing in sharp contrast in the UMAP images.

In addition, this protocol enables the spatial analysis of ¹³C-enriched intermediates derived from [U-¹³C]glucose metabolism in the mouse brain. While ¹³C enrichment is traditionally assessed in tissue extracts,^2,6 Figure 4E illustrates the spatial distribution of M+0 (unlabeled) and M+2 (labeled with two ¹³C atoms) isotopologues of key metabolites derived from [U-¹³C]glucose, such as glutamate, glutamine, and GABA, along with their enrichment profiles across a sagittal section of the brain. Moreover, the protocol is highly reproducible among biological replicates, with a coefficient of variation of 21.1% in the cerebellum and 22.3% in the cortex (see full enrichment analysis in the GitHub repository: (https://github.com/Neurometabolomics/Protocol_Spatial_Neurometabolomics/blob/main/Code/Enrichment_Analysis.Rmd). Comparing this ¹³C enrichment analysis with data from cortical and cerebellar extracts of P21 wild-type mice injected with the same [U-¹³C]glucose protocol and analyzed via GC-MS,² we observed a strong correlation (R = 0.86) between MSI and GC-MS measurements. Across all metabolites, MSI and GC-MS showed highly comparable enrichment patterns, particularly for glutamate. However, some differences were observed in glutamine, where MSI detected a 9.5% higher M2 enrichment (MSI/GC-MS ratio: 1.45) in the cortex, along with a corresponding 16.4% decrease (ratio: 0.72) in M0 enrichment. In the cerebellum, MSI reported a 13.1% higher M2 enrichment in glutamine (ratio: 1.81) compared to GC-MS, with no differences in the corresponding M0. For GABA, cerebellar M2 enrichment was 2.8% higher in MSI (ratio: 1.12) relative to GC-MS, while M0 and cortical isotopologues showed no significant differences. These findings indicate that while MSI effectively captures ¹³C enrichment trends, moderate differences may exist when compared to conventional tissue extract-based mass spectrometry, particularly in glutamine and GABA. Nonetheless, MSI provides ¹³C enrichment data that closely align with GC-MS data, making it a powerful tool for flux analyses with the added advantage of spatial resolution.

Altogether, these data provide critical insights into region-specific glucose metabolism in the mouse brain and support metabolic flux analysis to estimate the activity of pathways, such as the Krebs cycle, at a regional level. Beyond central carbon metabolism, the protocol can be adapted to investigate other pathways using specific tracers. For instance, ¹³C-labeled fatty acids could be used to analyze fatty acid oxidation in the brain, while deuterated water could enable the study of lipid synthesis at a regional level, as we previously described in tissue extracts.^6,7 Further validation will be needed to fully establish these approaches. Overall, the ability to administer various substrates with different labeling patterns enhances the protocol’s versatility, such that it offers a foundation for future studies on spatial brain metabolism under both physiological and pathological conditions.

Quantification and statistical analysis

Statistical analysis and data visualization were performed using R statistical software version 4.11. An unpaired, two-tailed Student’s t-test was used to compare MSI and GC-MS measurements, with the significance level (α) set at 0.05. Pearson correlation analysis was applied to assess the linear relationship between MSI and GC-MS data.

Limitations

This protocol provides a robust framework for high-resolution spatial metabolomics in mouse brain tissue, yet it has some limitations. The accuracy and reliability of MSI data can be affected by environmental factors such as humidity and temperature fluctuations, which may alter ionization efficiency during sample preparation. Additionally, tracer stability and matrix deposition require freshly prepared solutions to minimize degradation, maintain consistency, and avoid contamination; any inconsistencies in how the tissue samples are handled can compromise spatial resolution, and metabolite and anatomical integrity. Further, this protocol is optimized for glucose tracing but may need adjustments for other tracers, especially those with distinct chemical properties such as lipids or amino acids. Consistent sample and tracer handling throughout the experiment and adherence to standardized conditions are essential to minimize these limitations. Further validation of detected metabolites, their quantification, and ¹³C enrichment can be performed using other methodologies, such as GC-MS or LC-MS, from tissue extracts.

Troubleshooting

Problem 1

Insufficient isotope tracer uptake (related to injection of tracers, part 1).

Potential solution

Confirm that tracer solutions are prepared at the correct concentration and are properly filtered. Monitor animal blood glucose post-injection to verify tracer uptake and adjust fasting time if necessary.

Problem 2

Loss of metabolite integrity during tissue preparation (related to tissue collection and sectioning, parts 2 and 3).

Potential solution

Rapidly freeze the brain tissue on dry ice after collection, keep it frozen during transportation to prevent degradation, and store it at −80°C. Perform sectioning on a pre-cooled cryostat stage and make sure the tissue is not exposed to room temperature before matrix deposition.

Problem 3

Inconsistent metabolite signal intensity (related to matrix deposition, part 4).

Potential solution

Ensure that matrix deposition is uniform across tissue sections, as variations can lead to uneven ionization. Verify that the matrix application equipment is calibrated, and that tissues are equilibrated to room temperature in a vacuum desiccator to prevent condensation before matrix application.

Problem 4

Low spatial resolution in MSI data (related to MSI operation and data acquisition, part 5).

Potential solution

Adjust the laser intensity, diameter settings, and pitch size on the mass spectrometer to optimize spatial resolution. Use fresh matrix solution and ensure it is applied evenly. Avoid excessive drying time between matrix application and analysis to preserve tissue morphology.

Problem 5

Identification of metabolites and their corresponding isotopologue ions (related to data analysis and visualization, part 6).

Potential solution

Given the large number of metabolites detected in each run, individually confirming all of them is impractical. Instead, using a metabolite library is recommended for broad identification. For compounds of particular interest, obtain chemical standards for each metabolite or isotopologue and analyze them using the dried droplet method for MALDI analysis. This approach also allows screening for the optimal matrix and ionization mode for analyte detection. Alternatively, perform MS/MS analysis using the dried droplet method or spare samples to profile fragment patterns of unlabeled species (M0) and identify isotopologues (e.g., M1, M2, etc.) by matching fragment patterns with the corresponding mass shifts.

Problem 6

Poor reproducibility between runs (related to tissue sectioning through data analysis, from parts 2 to 6).

Potential solution

Standardize all sample handling and preparation steps. Calibrate the mass spectrometer regularly and run quality control samples with each batch to ensure consistent instrument performance. Carefully label and organize all slides to avoid cross-contamination. Furthermore, computational tools, such as peak alignment algorithms, can mitigate some effects of instrumental variability by correcting for m/z shifts and improving data consistency.

Alternatives: Reagents: Alternative stable-isotope tracers or matrix compounds may be tested for specific metabolites.

Alternatives: Equipment: If an iMLayer matrix deposition system is unavailable, manual matrix application methods (e.g., airbrush, homebuilt sublimation) may be used, although they require greater operator skill and may result in lower reproducibility. The iMLayer deposits roughly 50 μg/cm² of CHCA on the slide after sublimation.

Alternatives: GC-MS or LC-MS analysis of whole tissue homogenate for analyte confirmation, as shown in Figure 4F.

For further troubleshooting guidance, refer to individual protocol steps as needed to ensure optimal results.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Isaac Marin-Valencia (isaac.marin-valencia@mssm.edu).

Technical contact

Technical questions on executing this protocol should be directed to and will be answered by Watit Sontising (watit.sontising@mssm.edu) and Francine Yanchik-Slade (fryanchikslade@shimadzu.com).

Materials availability

All materials generated in this study are available from the lead contact upon request. No new unique materials were generated.

Data and code availability

The protocol includes all datasets generated or analyzed during this study. New code was created to analyze the data and is available in this GitHub repository: https://github.com/Neurometabolomics/Protocol_Spatial_Neurometabolomics.

Acknowledgments

This work was supported in part by NIH grants 5K08NS110877-04 (I.M.-V.) and 1R21NS133826-01A1 (I.M.-V.) and the March of Dimes Foundation (I.M.-V.).

Author contributions

W.S., C.R.-N., F.Y.-S., M.A.H., M.G.-R., J.V., M.N.B., and I.M.-V. conceptualized the study. W.S., C.R.-N., F.Y.-S., and M.A.H. developed the experimental protocol, which W.S. and F.Y.-S. optimized. M.G.-R. and J.V. collected and sectioned brain tissue samples. S.Y. contributed to data analysis, and W.S. developed the R-based data analysis pipeline. I.M.-V. supervised the project. W.S., F.Y.-S., and I.M.-V. drafted the manuscript, and all authors revised and approved the final version.

Declaration of interests

F.Y.-S., M.A.H., and M.N.B. were employed by Shimadzu Scientific Instruments, Inc., and S.Y. was employed by Shimadzu Corporation during the development of the protocol and/or preparation of the manuscript.