Getting spatial sample preparation right is the highest-leverage work in mass spectrometry imaging, and it is also where the field’s most persistent mistake occurs: borrowing a protocol from a colleague working on a different molecular class. There is no universally correct preparation, because several of the key decisions help one analyte class and damage another.
Key Takeaways
|
Why Sample Prep Dominates MSI Quality
Because there is no cleanup stage — none at all. In a conventional workflow, extraction and chromatography stand between the sample and the ion source, correcting for a good deal of what happened earlier. Imaging removes both. What is on the slide is what gets ionised, and whatever was lost, moved, or degraded before that point is simply absent from the result with no indication that it was ever there.
Three consequences follow, and they compound.
- Losses are invisible. An analyte washed away or degraded produces no signal, and no signal is indistinguishable from genuine absence. In an untargeted experiment, this is particularly dangerous, as set out in Untargeted Spatial Metabolomics Workflows.
- Movement becomes biology. Analyte delocalisation during a wet step does not look like an artefact. It looks like a diffuse distribution, which is a perfectly plausible biological result.
- The tissue is the matrix. Ionisation happens in whatever chemical environment the preparation left behind, so preparation choices directly affect ion suppression and therefore quantification, as covered in Quantitative Mass Spectrometry Imaging: Challenges and Approaches.
Is There a Universal Protocol?
No — and the reason is worth setting out explicitly, because it is the single most useful thing to understand before designing a workflow. Several preparation decisions are genuinely in conflict: the choice that improves one analyte class measurably degrades another.
Decision | Metabolites | Lipids | Peptides and Proteins | Morphology |
Aqueous washing | Harmful. Removes water-soluble species | Largely tolerated | Helpful. Removes interfering salts | Neutral |
Haematoxylin staining | Not usually applied | Not usually applied | Harmful. Reduces protein detection | Essential. Enables region selection |
Formalin fixation | Variable by species | Generally tolerated | Requires de-crosslinking | Excellent |
Fresh frozen | Preferred | Preferred | Preferred | Poorer than fixed |
Rapid freezing, seconds | Critical | Less critical | Less critical | Neutral |
Matrix by sublimation | Lower coverage | Good, finest resolution | Workable | Not applicable |
Matrix by spraying | Higher coverage | Risk of delocalisation | Workable | Not applicable |
Thicker sections | More signal, more delocalisation risk | More signal | More material for dissection | Poorer optical clarity |
Table 1. How preparation decisions serve or harm each analyte class. Compiled from the published findings discussed across this cluster. Read the rows for conflicts: several decisions cannot simultaneously suit metabolite and protein work.
Three Conflicts Worth Knowing Before You Start Washing. Aqueous washes remove salts that interfere with peptide and protein detection, which is why protein protocols include them. Those same washes remove water-soluble metabolites. A metabolomics experiment run on a protein protocol will lose exactly the analytes it was designed to find. Staining. Work on microdissection of formalin-fixed and stained lung tissue notes that haematoxylin and eosin staining provides critical morphological characterisation enabling identification of anatomical features, while haematoxylin staining has been shown to reduce protein detection by mass spectrometry. You need the stain to choose the region, and the stain costs you the measurement. Matrix deposition. A 2025 comparison of matrix application methods found that with one common matrix at 10 micron imaging, sublimation gave significantly better resolution than spraying, with structural boundaries not easily discerned in the sprayed image. Separate work found an optimised sprayer detected roughly double the metabolites of sublimation. Resolution and coverage pull in opposite directions at the deposition step. |
Working in analytical science?
Register for a FREE Separation Science account to subscribe to the Separation Science Newsletter.
The practical rule follows directly. Establish which analyte class the experiment is for before anything else, accept that the preparation will be suboptimal for other classes, and if you genuinely need two classes, use adjacent sections with different preparations rather than compromising on one. A compromise protocol usually serves neither class well.
The Decision Sequence
Preparation decisions are not a checklist to work through in any order. They form a dependency chain, and optimising a late decision while an early one remains open wastes effort.
Decide | Because It Determines | What It Forecloses |
1. Analyte class | Nearly every subsequent choice, from fixation through to pixel size | Optimal conditions for the classes you did not choose |
2. Fixation and preservation | Molecular integrity, morphological quality, and whether archival material is usable | Fresh frozen cannot be recovered from a fixed block |
3. Section thickness and mounting | Material available per pixel, delocalisation risk, and slide compatibility | Slide choice constrains microdissection and some ionisation approaches |
4. Washing and staining | Which species survive to be measured, and whether morphology is available | Washed-out analytes and stain-suppressed proteins are unrecoverable |
5. Matrix or solvent | Which chemical classes are extracted and ionised, and achievable resolution | Crystal size caps resolution regardless of instrument settings |
6. Pixel size and area | Coverage, acquisition time, and feasibility | Nothing downstream, which is why it comes last |
Table 2. The preparation decision sequence, with what each stage forecloses. The ordering matters: each decision constrains the options remaining below it.
The first row carries the most weight and receives the least deliberation. Analyte class is often treated as a given rather than a decision, yet as Table 1 shows, it determines whether washing helps or harms, whether staining is available, and which deposition method suits. Programmes that begin by choosing a matrix have already skipped four decisions.
The last row is placed deliberately. Pixel size is frequently decided first because it feels like the important specification, but it forecloses nothing downstream and depends on everything above it. The arithmetic that should govern it is in Spatial Resolution vs. Sensitivity in MS Imaging: The Fundamental Trade-off.
Fixation, Sectioning, and Mounting
These three settle the physical substrate everything else acts on. Each is treated fully in the spokes below; what matters at this level is how they interact.
Fixation is the decision most often inherited rather than made, because the tissue already exists. Formalin-fixed paraffin-embedded material offers excellent morphology, room-temperature stability, and access to biobank collections, at the cost of crosslinking that must be reversed for protein work and that affects some analytes more than others. Fresh frozen preserves molecular integrity better and gives poorer morphology, while demanding an unbroken cold chain. Both are workable for imaging, which is worth stating because FFPE is still sometimes assumed unsuitable. Readers approaching the same decision from a laboratory operations rather than a method development perspective may find our sister publication’s treatment useful, in FFPE vs. Fresh Frozen for Spatial Biology: Sample Handling Decisions, which covers fixation chemistry and platform quality thresholds for a lab management audience.
One aspect of preservation is genuinely analyte-specific rather than a general trade-off, and it is the one most often underestimated. For labile metabolites, the interval between excision and freezing matters on a timescale of seconds. A study of harvest-induced hypoxia in mouse tissue found 31 metabolites significantly increased within 30 seconds of dissection and 128 within 10 minutes, with the broad metabolome reaching half its total measured change in around 3.58 minutes. For lipid and protein work the same delays are far less consequential, as discussed in Spatial Lipidomics in Tissue.
Sectioning and mounting then set thickness, which determines material per pixel, and slide format, which constrains what can follow. Conductive slides are required for some ionisation geometries; specialised membrane or glass slides are required for microdissection, as covered in Laser Capture Microdissection + LC-MS Proteomics. Choosing a slide before knowing the downstream workflow is a common and easily avoided error.
What Does Washing Actually Remove?
More than intended, and this is the conflict worth understanding in detail because it is the one most often applied unthinkingly from a borrowed protocol.
The purpose of washing in peptide and protein imaging is to remove salts and lipids that suppress ionisation and crowd the spectrum. Ammonium acetate and graded alcohol washes are established for this and they work. The difficulty is that solubility does not discriminate by research intent: an aqueous wash that removes sodium and potassium salts also removes small polar metabolites, and a lipid-removing organic wash removes the lipids a lipidomics experiment was designed to map.
Three practical positions follow, depending on what you are measuring.
- Peptide and protein work. Wash as the protocol specifies. Salt removal materially improves spectra and the analytes of interest are not solubilised by the washes used.
- Metabolite work. Do not adopt a washing step from a protein protocol. If salt suppression is a problem, address it through ionisation approach and mass resolving power instead, as discussed in MALDI Imaging Mass Spectrometry: How It Works.
- Lipid work. Aqueous washes are generally tolerated and can improve spectra by removing salts, but organic washes will remove the analytes. The distinction between wash chemistries matters more here than the presence or absence of a wash.
The broader lesson generalises beyond washing. Every preparation step that removes something is removing it by chemical property rather than by relevance, and the properties that define an interferent in one experiment define the analyte in another. That is the mechanism behind most of Table 1.
How Do You Build a Reproducible Prep SOP?
By recording the decisions rather than only the steps, since a protocol that lists actions without stating what they were chosen for cannot be adapted when the analyte changes.
Record | Why It Belongs in the Document |
The analyte class this protocol is for | Prevents the borrowing error that Table 1 exists to warn against. Should be the first line |
Time from excision to freezing | Critical for metabolites and worth recording for everything, since it cannot be reconstructed later |
Fixation details, or that tissue was fresh frozen | Determines what de-crosslinking or retrieval is required downstream |
Section thickness and slide type | Constrains ionisation approach, microdissection compatibility, and material per pixel |
Washing steps, with chemistry and duration, or explicitly none | The most consequential omission in most protocols. Recording that no wash was performed is as important as recording one |
Matrix or solvent, and deposition method with parameters | Sets both coverage and the resolution ceiling |
Staining, if any, and where it sits in the sequence | Affects protein recovery and must be visible to anyone interpreting a low yield |
Anything deliberately not done, and why | The rationale is what allows the protocol to be adapted rather than only repeated |
Table 3. A preparation record that supports adaptation rather than only repetition. The first and last rows are the ones that make the difference between a protocol and a recipe.
That final row is worth emphasising. A protocol recording only what was done can be repeated exactly and cannot be modified intelligently, because the next person does not know which steps were essential and which were inherited. Recording why a wash was omitted, or why a particular deposition method was chosen, converts a recipe into something a colleague working on a different analyte class can reason about.
This section covers each decision in depth. Fixation and preservation, including which molecular classes survive each route, are in FFPE vs. Fresh Frozen for Mass Spectrometry Imaging. Thickness, slide selection, and the handling practices that preserve spatial fidelity are in Tissue Sectioning and Mounting for Spatial MS. Matrix chemistry, deposition method, and the crystal size that caps resolution are in Matrix Selection and Application in MALDI Imaging. And registering ion images to stained sections and other modalities, which is what makes a molecular map interpretable alongside histology, is in Coregistration With Histology and Multimodal Imaging.
For how these preparation choices interact with ionisation approach and instrument selection, see Mass Spectrometry Imaging: Principles, Techniques, and Applications and, for the ambient alternative that avoids matrix application altogether, DESI and Ambient Ionization Imaging. For the wider spatial landscape, see Spatial Analysis in Analytical Science: Mass Spectrometry Imaging and Spatial Omics.
This article was produced under Separation Science's AI Editorial Guidelines.
