Tissue Preparation for Spatial Transcriptomics: A Complete Guide

Tissue preparation is the step in spatial transcriptomics that can determine whether an experiment succeeds or fails before the sample ever reaches the instrument. Getting fixation, sectioning, and RNA quality right up front matters more than any single downstream analysis choice. This guide walks through fixation methods, sectioning practices, RNA integrity assessment, and the pitfalls that most often derail a spatial transcriptomics run.

Why tissue quality determines spatial transcriptomics success

The quality of the tissue entering a spatial transcriptomics workflow sets an upper limit on the quality of the resulting data, and no amount of computational correction afterward can fully recover RNA that degraded before sectioning. Both sequencing-based and imaging-based platforms depend on RNA molecules being intact and accessible at the moment of capture or hybridization, which makes tissue preparation the first and most consequential step in any spatial transcriptomics project (Figure 1).

Figure 1: A linear flowchart of the tissue preparation workflow for spatial transcriptomics, from collection and preservation through sectioning and RNA quality checkpoints. Credit: AI-generated image created using Google Gemini (2026).

Formalin fixation, storage conditions, and handling between collection and sectioning each introduce risks to RNA integrity, and their effects compound. Research using data from the Genotype-Tissue Expression project found that gene expression changes measurably within hours of death. Additionally, the degree of change varies by tissue and by how long a sample sits before it is preserved after death. That finding underscores why the interval between tissue collection and stabilization deserves as much attention as the fixation protocol itself.

Because tissue preparation sits upstream of every platform choice, it deserves the same planning attention researchers typically reserve for choosing a transcriptomics platform. A platform selected for its resolution or gene coverage will still underperform on poorly prepared tissue, regardless of how capable the underlying chemistry is.

Fixation methods: FFPE versus fresh frozen tissue

The choice between FFPE tissue and fresh frozen tissue shapes what a spatial transcriptomics experiment can achieve. FFPE preserves tissue morphology exceptionally well and allows long-term storage at room temperature, but the formalin cross-linking process fragments RNA and chemically modifies bases in ways that fresh frozen tissue avoids entirely.

Fresh frozen tissue generally yields higher-quality, less fragmented RNA and is compatible with a wider range of spatial platforms without additional recovery steps. The trade-off is that fresh frozen samples require continuous cold-chain handling, are more prone to morphological artifacts such as tissue tearing during sectioning, and cannot be archived indefinitely at room temperature like FFPE blocks.

Fixation duration matters independent of which method is chosen. A study using RNAscope in situ hybridization on human brain tissue found that RNA detection varied and background autofluorescence increased significantly as fixation time lengthened. This is why standardized, minimal fixation windows generally outperform prolonged fixation done for scheduling convenience. Neither fixation approach is universally superior; the right choice depends on whether a project prioritizes morphological detail and archival stability or maximum RNA yield and simplicity (Table 1).

Table 1: A comparison of FFPE and fresh frozen tissue for spatial transcriptomics.

Sectioning and placement in spatial transcriptomics

Section thickness and placement accuracy directly affect how well a spatial transcriptomics platform can resolve individual cells and capture representative RNA. Most platforms specify a narrow thickness range, and sections that fall outside it can introduce inconsistent tissue folding, incomplete permeabilization, or reduced signal across the capture area.

Section thickness also interacts with cell biology in ways that are easy to overlook. A study of human cerebral cortex tissue found that neuronal nuclei, at roughly 20 micrometers (µm) in diameter, are often larger than the 10 to 12 µm sections typically used on spot-based spatial platforms, so a single section can capture only part of an affected nucleus and lose signal as a result.

Placement onto the capture surface or slide also matters more in spatial transcriptomics than in standard histology, since misalignment or bubbles introduced during mounting can create gaps in spatial coverage that no amount of downstream analysis can fill in. Working in an RNase-free environment throughout sectioning is essential regardless of fixation method, since even trace RNase contamination can degrade RNA in the minutes between cutting and fixation or freezing of the section.

A short checklist helps standardize sectioning across a lab or core facility:

• Confirm section thickness matches the platform's validated range before cutting.
• Use RNase-free blades, slides, and surfaces for every sectioning session.
• Avoid tissue folds, tears, or bubbles when mounting sections onto the capture area.
• Minimize the time between sectioning and the next fixation or permeabilization step.
• Document lot numbers and handling conditions for every tissue block processed.

Assessing RNA integrity with DV200 and RIN

RNA integrity assessment is the quality control step that determines whether a tissue sample is worth carrying forward into an expensive spatial transcriptomics run. The RNA integrity number, developed to standardize RNA quality scoring from electrophoretic traces, remains the standard metric for fresh frozen tissue. However, it performs poorly on FFPE-derived RNA because formalin fixation eliminates the distinct ribosomal peaks that the RIN algorithm depends on.

DV200 was developed specifically to address that gap. A direct comparison of RNA samples from FFPE tissue, fresh frozen tissue, and cell lines found that the DV200 index outperformed the RIN equivalent as a predictor of successful sequencing library preparation, with a DV200 threshold near 66% correlating with reliably successful library outcomes in that study. Other FFPE-focused RNA quality research reports a broader working scale: a DV200 above 70% is generally considered high quality, values between 50% and 70% are usable with adjusted input amounts, and values below 30% are typically too degraded for reliable sequencing.

Uncovering Hidden Genetic Causes of Sudden Cardiac Death With Long-Read Sequencing

It explores how long-read sequencing was used to identify a pathogenic dystrophia myotonica protein kinase (DMPK) repeat expansion associated with inherited cardiac disease and sudden cardiac death after previous testing failed to find a diagnosis.
View Webinar

Advertisement

A consistent workflow for assessing RNA integrity before committing tissue to a spatial run includes the following steps.

1. Extract a small RNA aliquot from an adjacent section or scroll rather than the section intended for the spatial assay.
2. Run the aliquot on a capillary electrophoresis instrument to generate an electropherogram.
3. Calculate RIN for fresh frozen samples or DV200 for FFPE samples, since the two metrics are not interchangeable.
4. Compare the result against the platform manufacturer's validated quality thresholds before proceeding.
5. If quality falls below the threshold, adjust input amounts, reconsider the fixation protocol, or select an alternative tissue block.

Common pitfalls in spatial transcriptomics tissue preparation

Overfixation is one of the most common and most preventable failure modes in spatial transcriptomics tissue preparation. Leaving tissue in formalin longer than necessary, whether due to workflow delays or convenience, increases RNA fragmentation and cross-linking well beyond what standard protocols anticipate, and this effect cannot be reversed later in the workflow.

Inconsistent handling between tissue blocks is a second common pitfall, particularly in multi-site or multi-operator studies. Variation in cold ischemia time before fixation, storage temperature, or time from sectioning to permeabilization introduces batch effects that are difficult to distinguish from genuine biological variation once sequencing or imaging data is in hand. Established biospecimen handling guidance recommends standardized, documented protocols precisely to minimize this kind of preanalytical variability across samples.

Skipping RNA quality assessment entirely is the third and most costly pitfall, since it means quality problems are discovered only after an expensive spatial run has already failed. Building a brief quality control step into the standard workflow, using DV200 or RIN as appropriate, is inexpensive relative to the cost of a failed spatial transcriptomics experiment and should be treated as non-negotiable rather than optional.

Tissue preparation as the foundation for reliable spatial transcriptomics

Tissue preparation is not a preliminary formality before the real spatial transcriptomics experiment begins; it is the step that determines whether that experiment can succeed at all. Fixation method, section quality, and a documented RNA integrity check together account for most of the variability researchers otherwise attribute to platform performance.

Labs that standardize these steps, choosing fixation deliberately, sectioning within validated parameters, and assessing RNA quality with the appropriate metric before committing tissue to an assay, consistently see more reliable spatial transcriptomics data with fewer failed runs. That discipline pays for itself well before a project reaches the stage of weighing single-cell integration approaches or interpreting results within the broader spatial biology landscape.

This content includes text that has been created with the assistance of generative AI and has undergone editorial review before publishing. Technology Networks' AI policy can be found here.