Question: Can I use an autostainer with the Visium CytAssist FFPEv2 assay?
Answer: Please refer to the Demonstrated Protocol for Visium CytAssist FFPE deparaffinization and H&E staining. Modifications to the Demonstrated Protocol steps are unsupported. We cannot guarantee assay performance even when implementing the guidance listed in this article. The information provided should only be considered a starting point for further optimization. 10x Genomics Support does not have additional information beyond what is provided in this article regarding using autostainers and will not be able to provide additional details from internal experiments or guidance.
Outlined below is a summary of our learnings and considerations from limited testing of autostainer with FFPE samples. The observations here are based on our limited testing with one specific autostainer model. The setup with different autostainer models can vary. Our recommendation would be to modify the autostainer protocol to match the 10X protocol as closely as possible, use 10x supported staining reagents, and perform a pilot run before proceeding with any large batches.
FFPE mouse brain and human tonsil samples were deparaffinized and stained using the Epredia™ Gemini™ AS Automated Slide Stainer (Epredia A81500001). Mayer’s hematoxylin (Sigma MHS-16), Dako bluing reagent (Agilent CS70230-2), and alcoholic eosin (Sigma HT110116) were used to stain the samples.
Additional optimization of staining and wash times may be needed to achieve desired staining intensity. In our limited testing, we altered the eosin incubation and wash times to produce a darker stain and observed that sample complexity was not impacted (Fig 1).
For all autostainer protocols on the Gemini, initial agitation was set for all steps. The limit was set to critical for the following steps: water incubation at the end of deparaffinization, hematoxylin incubation, bluing incubation, and eosin incubation. Slides were coverslipped with 85% glycerol and imaged before proceeding with library generation.
Mouse brain samples were run through the CytAssist Spatial Gene Expression assay. The median UMIs per spot were comparable between manually stained and autostainer stained samples (Fig.2a).
Human tonsil samples were run through the CytAssist Spatial Gene and Protein Expression assay. The median UMIs per spot in the gene expression library was comparable between the manually stained and autostainer stained samples (Fig.2b). In the protein expression library, the antibody distribution and antibody clustering were comparable between the manually stained and autostainer stained samples (Fig.2c).
There are some considerations when using an autostainer:
- Coverslipping: We recommend coverslipping with 85% glycerol. Slides mounted with Cytoseal will require additional steps to remove the coverslip and residual mounting media. Please ensure the mounting media is completely removed. Any residue mounting media can impact the downstream assay performance.
- Staining: We recommend using 10x supported staining reagents. Increasing the hematoxylin incubation time is a higher risk modification. Changing the incubation times for bluing buffer and eosin is a lower risk modification.
- Water: Pressure/flow of water source may be suboptimal and require a pump. Installing a water filter may be necessary to filter out contaminants.
- We recommend replacing solutions weekly. Some autostainer models do not come with airtight container lids, causing faster evaporation of the reagents.
Figure 1. Effect of Eosin Optimization on UMI’s per spot. FFPE mouse brain samples processed through the autostainer with different staining protocols and run through the CytAssist Spatial Gene Expression assay using 6.5mm Visium slides. The median UMIs per spot were comparable between the differing staining protocols. Scale bars are 500 um.
Figure 2. Effect of Autostainer Use on Visium CytAssist FFPE Samples. a. FFPE mouse brain samples were processed through the autostainer and run through the CytAssist Spatial Gene Expression assay using 6.5mm Visium slides. The median UMI’s per spot were comparable between manually and autostainer stained samples. b. FFPE human tonsil samples were run through the CytAssist Spatial Gene and Protein Expression assay using 6.5mm Visium slides. The median UMI’s per spot were comparable between manually and autostainer stained samples. c. Antibody distribution and clustering was comparable between manually and autostainer stained samples. Scale bars are 500 um.
Product: Visium CytAssist for FFPE, Visium CytAssist Gene and Protein