Comprehensive Spatial Profiling of Species-agnostic Transcriptomes via Stereo-seq

Spatial-omics technologies have revolutionized the study of tissue cellular heterogeneity by enabling the simultaneous quantification of multiple targets, DNA (genomics), RNA (transcriptomics), metabolites (metabolomics), or protein (proteomics)¹. The resolution of these different assays often varies, providing different limitations when measuring analytes, either with increased sparsity when measuring single cell^2,3 or native spatial context⁴. In contrast, bulk methods such as RNA-sequencing only provide average gene expression across entire samples and obscure cellular heterogeneity², or single-cell RNA-sequencing (scRNA-seq)^2,3, which requires tissue dissociation for single-cell sequencing, often requires filtering. Therefore, these methods exclude large cells and aggregates, while bioinformatic methods try to remove fragments and doublets, all of which induce biases^5,6,7. Spatial transcriptomics retains tissue architecture providing critical insights into cellular organization and interactions. Most of the non-targeted spatial sequencing methods do not have single-cell resolution, thus requiring deconvolution of spatial capture areas that containing 10 to 100 cells^8,9,10. However, in the last decade, spatial transcriptomics has outpaced the needs for these methods, as there have been advancements in resolution, larger areas of capture, and the integration of multi-omics approaches, either using experimental strategy or novel techniques^1,11,12.

Spatial Enhanced Resolution Omics-sequencing (Stereo-seq) provides nanoscale resolution (500 nm) and a centimeter-scale capture area, with species-agnostic whole transcriptomic capture allowing for the capture of coding, non-coding, and non-host RNA¹³. This method is currently the highest resolution for spatial transcriptomics and has options to provide the largest fields of view. Yet, this method has some notable limitations, including time and labor requirements of around 15 h hands-on time, spanning 4 days to complete the sequencing-ready libraries. In comparison, other sequencing-based spatial transcriptomics technologies, such as probe-based spatial transcriptomics, polyA-capture array-based spatial transcriptomics, Slide-seq, and deterministic barcoding in tissue for spatial omics sequencing (DBiT-seq) typically require 5 to 8 h of hands-on time over 2 to 3 days (Table 1). For many researchers, these advances outweigh the technical challenges and are ideal for constructing full transcriptomics spatial atlases to study disease progression at cellular and subcellular levels, particularly if layered with other technologies¹.

The Stereo-seq protocol has possibilities beyond traditional mRNA transcriptomics, which relies on poly-T probes for capturing polyadenylated mRNA transcripts. These methods can be modified by adding polyadenylation to fresh tissues¹⁴ or species-specific probes for target predefined gene sets¹⁵ for spatial expression analysis. Instead, Stereo-seq uses random nucleotide probes to select diverse RNAs regardless of species, including microbial RNAs, fungal RNAs, viral RNAs, and non-polyadenylated RNAs such as enhancer RNAs, circular RNAs, and long non-coding RNAs (lncRNAs)¹⁶. Theoretically, it is possible to capture smaller RNAs or reparative elements such as small interfering RNAs (siRNAs), microRNAs (miRNAs), and transposons¹⁷. However, currently the bioinformatics and mapping of these RNA remain challenging and are often filtered out during the initial alignment , which typically uses 30 bp cutoff.

Outlined here is a detailed protocol, with our modifications for a manufacturer's random nucleotide capture on DNA nanoballs (DNB) chips to ensure better understanding and handling of a Stereo-seq protocol illustrated in Figure 1. This protocol includes strategies for molecular biologists for testing new FFPE tissues on Stereo-seq with random nucleotide capture to produce more consistent, higher yields from lower quality archived samples and fatty tissues. The outlined protocol is optimized for FFPE samples from most organs, regardless of species of origin. However, this protocol does not include current modifications needed to handle difficult tissues such as bone marrow, which require decalcification.

No clinical data are provided in this methods paper for the treatment-naive high-grade serous ovarian carcinoma patient specimen. However, associated clinical data were collected following protocols approved by the Institutional Review Board (IRB) at The University of Texas MD Anderson Cancer Center, Department of Gynecologic Oncology and Reproductive Medicine, from participants who provided written informed consent.

NOTE: It is recommended to wear a surgical mask and gloves during the entirety of the protocol to prevent RNase contamination. Researchers with long hair are recommended to pull it back or even wear a hairnet. Always use sterile nuclease free water (NFW) and nuclease free, low-binding DNA/RNA tubes to maximize yield. Non-mammalian tissues, such as plant tissues, may require adherence optimization, involving poly-L-lysine (PLL) and extended permeabilization times due to cell walls. While molecular capture remains theoretically feasible, experimental validation is needed as plant tissues remain untested in our current workflow.

1. Assess RNA quality and RNA fragment distribution value 200 scores

1. Assess RNA quality using RNA Fragment Distribution Value 200 (DV200), which measures the percentage of RNA that is greater than 200 bp.
2. Utilize two 10 µm sections from the same tissue block intended for Stereo-seq processing.
   NOTE: Immediate processing is optimal, but samples can be stored at 4 °C for up to 1 week, -20 °C for 4 weeks, or -80 °C for 6 months.
3. Extract RNA using a non-size selective RNA extraction protocol as described by Patel et al.¹⁸ to allow accurate evaluation of the distribution of RNA fragment sizes.
4. Evaluate RNA quality with one of the following using a fragment analyzer to determine the DV200 score, or the percentage of area under the curve that represents RNA fragments longer than 200 bp. Ideally, tissue should have a DV200 score (> 50%) and concentration (> 10 ng/µL).

2. Workspace cleaning

1. Clean all surfaces with 30% bleach. Allow surfaces to dry for at least 30 min to prevent the formation of chloroform from the combination of bleach and ethanol.
2. Wipe surfaces with DNA decontamination wipes (DDW). Spray down all surfaces with RNase decontamination solution (RDS) and then 70% ethanol prepared with NFW.
   NOTE: Never mix bleach or any other alcohol. This combination produced chloroform, which is highly toxic and can cause unconsciousness, organ damage, or death even at low exposures.

3. Stereo-seq random oligonucleotide capture chip (N-chip) preparation

1. Clean the workspace as described in steps 2.1 and 2.2, including all systems used in this step.
2. Remove the Stereo-seq N-chip from its packaging, record the chip ID, and clean the edges of the Stereo-seq N-chip and glass slide with 100% ethanol without touching the chip. Allow to airdry completely.
3. Resin application to secure the chip to the slide (Optional for users using Xylenes).
   NOTE: Although non-xylene clearing agents are often recommended by commercial vendors, it has been observed that non-xylene clearing agents can be less effective depending on the paraffin mixture. They are often not compatible with certain plastics, synthetic resins, or bee wax-embedded tissues, and do not clear heavily cross-linked or high-melting-point paraffins efficiently. If uncertain of the paraffin composition used for embedding the sample, it is recommended to use xylenes for optimal clearing. Furthermore, xylenes are better able to clear fatty tissues, such as breast, ovary, ovum, or brain tissue.
   1. Shake or mix resin well to make sure it is homogeneously white. Using a micropipette, carefully apply 10 µL of UV-curable ceramic resin around the edge of the Stereo-seq N-chip (Figure 2A).
      NOTE: Resin can cause skin and eye irritation and may lead to allergic skin reactions with repeated exposure. When pipetting or handling this resin, always wear nitrile gloves, safety goggles, and a lab coat. Work in a well-ventilated area and wear a mask when using large volumes or if heating solution and avoid direct contact with skin or eyes.
   2. Use a sealed foam swab, remove excess resin, leaving only a thin line around the chip edge. Place the slide into the UV curing device, ensuring not to touch the chip surface.
   3. Place the resin-applied chip on top of an opaque mask on the 405 nm UV light source to illuminate the back of the slide and cure for 5 min. This is illustrated in Figure 2B.
      NOTE: Do not overexpose the Stereo-seq N-chip to UV light and do not touch resin on the chip surface, as it may cause detrimental effects to the nanoballs.
   4. After curing, use sealed foam swabs soaked in each of the following: 100% molecular grade ethanol; 70% ethanol; and finally NFW to clean around the edge where resin was placed. If white specks come off, clean the edge again. Never touch the chip surface (Figure 2C).
   5. Allow the slide to airdry completely. After inspection, store the chips overnight at 4 °C with a desiccant if needed. However, it is recommended to continue with chip preparation and sectioning immediately when possible.
      NOTE: Always wear appropriate personal protective equipment, including UV-protective eyewear, when working with UV light. Handle ethanol and resin in a well-ventilated area, preferably a fume hood. Dispose of waste materials according to institutional guidelines.
4. Chip Preparation with Poly-L-lysine (Optional for fatty tissues or tissues with adherence issues.)
   1. Clean the workspace as described in steps 2.1 and 2.2, including all systems used in this step.
      NOTE: Do not spray chemicals directly into the PCR machine. For all pre-PCR steps, work in a PCR hood, if possible, to prevent contamination including previous amplification of cDNA.
   2. Submerge the chip three times in fresh NFW using a 50 mL conical tube.
   3. Dry the chip using compressed air, using the full force of compressed air at approximately 60° to 80° angle going diagonally (45°) across the surface of the chip from about 5 cm away from the surface of the chip (Figure 2E). If the surface appears cloudy, repeat the wash with fresh NFW and dry again (the cloud surface usually means the ceramic was not sufficiently washed off or cured in the previous steps).
   4. Apply 150 µL of PLL solution to cover the entire chip surface. Incubate for 10 min at room temperature (22 °C to 25 °C). This is illustrated in Figure 2D.
   5. Remove the PLL solution and wash the chip twice with nuclease free water for 10-15 s.
   6. Dry the edges of the slide carefully after the second wash (without touching the chip).
   7. Firmly dry the chip using compressed air, using the full force of compressed air at approximately 60° to 45° angle going diagonally across the surface of the chip from about 5 cm away from the surface of the chip (Figure 2E).
      NOTE: The chip needs to be completely dry within 30 s or less. If it takes longer than this, the DNB will be disturbed, which will cause artifacts in the capture.
   8. Proceed to mounting within 1 h and keep the chip dry.
      NOTE: If the chip surface appears cloudy after drying, wash with fresh NFW and repeat the drying process. If there are artifacts on the surface of the chip after drying, this can cause capture issues in those areas. Avoid using PLL to adhere if the tissue does not require it. Touching the chip can cause scratches and loss of capture areas before cDNA release, or uneven capture after release. Resin on the chip surface leads to speckled loss of capture areas. Placing the chip too close to the compressed air nozzle may create crystals and a spray of damaged DBN marks across the chip surface; however, this is rare and usually occurs only if the nozzle is almost touching the chip and the compressed air bottle is too cold. It is recommended to practice this procedure several times with assessment chips, as this step is technically challenging.

4.Tissue sectioning and mounting

1. Ensure that the workspace is clean as described in steps 2.1 and 2.2. Ensure new blades and NFW are utilized for all water baths and ice during tissue sectioning and mounting. Do not touch the chip directly, and only cover 80% of the chip with tissue.
   NOTE: It is recommended to use a thickness of 5 µm for standard tissue or 4 µm for high-fat tissue. Use a ribbon of serial sections to match modalities across samples, if applicable.
2. Dry slides at 42 °C for 3 h either in an oven or on a PCR machine with the lid open. Then transfer slides or change temperature on oven or PCR machine to 37 °C and bake overnight for 12-48 h. It is recommended to use around 18 h.
3. Optional stopping point: After drying, seal the slides in aluminum bags with desiccant at 4 °C for up to 1 week before proceeding.
   NOTE: Longer drying times may reduce RNA integrity. Too short drying will result in lower tissue adhesion to the slide. Baking at a higher temperature may reduce RNA integrity. Slides can be shipped for external processing (e.g., host analysis). However, shipping is not recommended for microbiome analysis due to increased potential contamination during transit.

5. Deparaffinize slides

1. Ensure that the workspace is clean as described in steps 2.1 and 2.2.
2. Prepare reagents for deparaffinization and Stereo-seq protocol.
   1. Aliquot the required reagents: Two 30 mL of Xylenes in 50 mL conical tubes, two 30 mL of 100% ethanol in 50 mL conical tubes.
   2. Prepare the following diluted reagents with NFW: 96% ethanol per chip (2 conical or 1mL required), 90% ethanol per chip, 80% ethanol per chip, 70% ethanol per chip, 50% ethanol per chip, 30% ethanol per chip, 2.5 mL of 5x saline-sodium citrate (SSC) from a 20x solution, 5 mL of 0.01N HCl (dilute 0.1N HCl with NFW to pH 2). For ethanol dilutions, either prepare one 30 mL solution in 50 mL conical tubes or a 500 µL solution in a 1.5 mL tube if using silicone gaskets.
   3. Prepare the following solutions and keep them on ice: i) 32.5 mL of 0.1x SSC from a 5x solution (keep 30 mL of solution in a conical tube and put the other 2.5 mL into a 5mL tube for future use), ii) 400 µL of 0.1x SSC with 5% RI, keep on ice.
   4. Prepare the following diluted reagents with 0.1 N HCl.
      1. Prepare 10x permeabilization reagent (PR) solution by adding 1 mL of 0.1 N HCl to the commercially provided lyophilized reagent.
         NOTE: Only keep on ice for 1 h, make sure to aliquot for future use, and freeze at -20 °C.
      2. Prepare 500 µL of 1x PR solution (dilute a 10x PR solution aliquot with 0.01 N HCl, keep on ice for up to 6 h).
         NOTE: It is critical that the pH of the 0.01 N HCl is within 10% of pH 2.0, as pepsin-based enzymes perform optimally at pH 2.
3. Deparaffinize slides by soaking slides in Xylenes (recommended) for 2 rounds at 20 min each.
4. Immediately, rehydrate the sample with ethanol (100%, 96%, for 2 rounds, 5 min each, followed by 90%, 70%, 50%, and 30% for 1 round, 2 min). Finally, rehydrate with NFW for 1 min. For ethanol dilutions, use 30 mL of solution in a 50 mL conical tube for each slide if preparing samples for microbiome studies, or use 30 mL for every two slides if preparing non-microbiome samples. For all subsequent dilutions, prepare 30 mL of solution in a 50 mL conical tube, or, if using silicone gaskets, use 500 µL of solution per 1 cm² chip to fully cover the chip area.
   OPTIONAL: It is suggested to place a silicone chamber from a 3-chamber removal slide around the chip to create a seal after the 100% ethanol step, as illustrated in Figure 2F-H. This allows the use of smaller volumes of reagents and reduces the risk of damaging the chip. Alternatively, 50 mL conical tubes can be used to fully submerge the Stereo-seq Slide throughout the procedure.

6. (Optional) Nuclear imaging and ssDNA staining

NOTE: Allow the FFPE decrosslinking reagent to reach room temperature ( ~25 °C) before use in 6.1. Other staining solutions are possible, but fluorescence nuclear staining is currently required for single-cell segmentation and cell-binning of transcriptomic data.

1. Fluorescence staining using single-stranded DNA (ssDNA) dye
   1. Prepare 2 µL of ssDNA staining solution in SSC buffer by diluting the Qubit ssDNA reagent with 5x SSC buffer at a 1:1 dilution per 1 to 2 chips. 
      NOTE: The ssDNA dilution can change based on the microscope used when imaging.  Therefore new microscopes should be tested with an ssDNA dye dilutions set using assessment chip and test tissue to determine best dynamic range.
   2. Create a master mix of the staining solution (200 µL per chip) using a 1:20 dilution of RNase Inhibitor (RI), 1:200 dilution of the ssDNA dilution, and the remaining 5x SSC.
   3. Add staining solution (150 µL per chip) and incubate in the dark for 5 min at room temperature. Gently remove the staining solution from the corner of the chip using a pipette.
      NOTE: If using the removable silicone chamber, remove it before staining the chip with ssDNA dye mix.
   4. Wash the chip twice with 0.1x SSC (150 µL per chip each time) for 10-15 s. Dry the edges of the slide carefully with lens paper after the second wash (without touching the chip).
   5. Add ~3-5 µL of glycerol (use less for an inverted microscope) to mount the cloverslip (be careful of bubbles).
      NOTE: Ensure that the correct volume of glycerol is applied to the slide area, typically about 4 µL for a 1 cm² chip. Using too much glycerol can cause the coverslip to shift when the slide is held vertically. Using too little glycerol increases the risk of tissue damage when the coverslip is removed and may introduce bubbles, which can negatively affect imaging and single-cell segmentation results.
   6. Drop the coverslip from about 1 cm height onto the chip.If using an intervened microscope, make sure the coverslip does not slide off when held vertically. If it does remount without adding more glycerol.
   7. Capture fluorescent images of ssDNA-stained tissue with a widefield microscope with 10% overlap between image tiles.
      NOTE: It is recommended to do a manual focus map of 9-13 points with a widefield microscope using a 10x objective. A confocal microscope can be used, but it is much more sensitive to focal changes across the tissue and chip.
   8. Stitch in image processing software such as ImageJ, utilizing theoretical overlap to generate the approximate positioning of each tile and subsequently perform pixel-matching computational overlap to fine-tune image tiling
   9. Process images in StereoMap software and ensure Image QC passes before proceeding.
   10. Submerge the Stereo-seq slide in 30 mL of 0.1x SSC until the coverslip detaches.

7. Decrosslinking process

NOTE: Avoid touching or tilting the chip during this step, and pipette gently.

1. Make sure the FFPE decrosslinking reagent is at warm room temperature  (~25 °C) and inspect to make sure there are no particles that make the solution cloudy. If particles are visible, heat to 55 °C and cool to 30 °C before use.
2. Turn on the thermocycler with the following settings: 30 °C for equilibration (a heated lid of 85 °C); 95 °C for a 30 min incubation; 4 °C infinite hold time.
3. Assemble the cassette gasket, then place the Stereo-seq chip slide in the cassette.
4. Add FFPE decrosslinking reagent (400 µL per chip) into the well of the cassette containing the chip. Apply sealing tape to the cassette and confirm it is tightly sealed.
5. Start the 30 min incubation at 95 °C from step 7.2. Once 10 min have passed, place 10 mL of methanol in a 15 mL container (about 1 mL per slide will be needed) in a -20 °C freezer.
6. Carefully remove the chip from the thermocycler, transfer the cassette to the bench, and peel off the sealing tape.
7. Remove and discard the FFPE decrosslinking reagent using a pipette. Once all of the decrosslinking reagents have been removed, detach and discard the cassette and gasket.
8. Equilibrate the Stereo-seq slide to room temperature ~23 °C. Under a sterile fume hood, dry off the edge of the slide and apply a new silicone chamber (if applicable). Add 500 µL of chilled methanol from the -20 °C freezer, making sure that the entire section is completely submerged for 20 min.
   NOTE: Alternatively, the entire Stereo-seq slide can be submerged in a 50 mL conical tube filled with 30 mL of chilled methanol.
9. Prewarm the 1x PR solution on a 37 °C dry block for 10 min before use  around 5 min before the end of methanol fixation (step 7.8).
10. After fixation is completed, move the slide to a sterile fume hood. Wipe off any excess methanol on slide and wait for evaporation of remaining methanol on chip (~5 min). Assemble a new cassette and gasket. Once the methanol has fully evaporated, place the Stereo-seq chip slide in the cassette.

8. Permeabilization and reverse transcription

NOTE: For steps 8.1-8.11, pipette slowly and do not touch the chip. Excess pressure will cause diffusion or marks in the DNBs.

1. Add 200 µL of permeabilization reagent to the chip. Apply sealing tape to the Stereo-seq slide cassette and ensure it is tightly sealed.
2. Incubate at 37 °C for 30 min.
3. Prepare the wash buffer by adding 380 µL of 0.1x SSC and 20 µL of RI (RNA Inhibitor) for each slide. Make 400 µL per slide and keep on ice until used.
4. Pull out the reagents for FFPE reverse transcription (FFPE RT), then spin down all of the following, and keep the enzyme mix, oligos, and dimers on ice. Warm the FFPE RT buffer to room temperature (~23 °C).
5. 5 min before min before the end of the permeabilization, prepare 200 µL per chip of FFPE RT mix. Add FFPE RT buffer (158 µL), FFPE RT enzyme mix (30 µL), FFPE RT oligo (10 µL), and FFPE dimer (2 µL). Mix this gently and keep it on ice.
6. Keeping the chip on the thermocycler, change incubation temperature to room temperature (~23 °C). Remove the seal and gently discard the 1x PR solution.
7. Wash the chip very gently with 200 µL of 0.1x SSC (with 5% RI) per 1 cm² chip.
8. Remove 0.1x SSC (with 5% RI) wash from the corner of the chip.
9. Add 200 µL of FFPE RT solution to the corner of the chip.
10. Apply sealing tape to the Stereo-seq slide cassette and ensure it is tightly sealed; otherwise, evaporation reduce yield.
11. Start the protocol FFPE RT incubation isothermal reaction at 42 °C with 45 °C heated lid for (5 to 24 h).
    NOTE: It is recommended to incubate for 12 to 16 h and be consistent across experiments.

9. cDNA release and purification

NOTE: For steps 9.6-9.7, pipette slowly and do not touch the chip. Excess pressure will cause diffusion or marks in the DNBs.

1. Prepare 400 µL of cDNA release buffer by warming cDNA release buffer at 55 °C to dissolve any precipitate that may have formed.
2. Equilibrate cDNA release buffer to a warm room temperature (25 °C). Do not put on ice.
3. Spin down the cDNA release buffer and the cDNA release enzyme at 100 x g in a tabletop centrifuge before use.
4. Prepare cDNA buffer by adding 20 µL of release enzyme to 380 µL of cDNA release buffer.
5. Keeping the chip on the thermocycler, change incubation temperature to room temperature (~23 °C). Remove the seal and gently remove and discard the FFPE RT solution.
6. Wash the chip once, incredibly gently, with 0.1x SSC per chip (200 µL).
7. Remove 0.1x SSC from the corner of the chip, pipetting very cautiously.
8. Add 400 µL of the prepared cDNA release solution to each chip.
9. Seal the Stereo-seqs slide cassette by applying sealing tape (or PCR cover) to the cassette. Ensure the cassette is sealed tightly to prevent any evaporation, leakage or contamination.
10. Incubate at 55 °C for ≥ 5 h (minimum of 5 h up to 24 h) with lid heated to 60 °C.
    NOTE: It is recommended to incubate for 5 to 7 h or to be consistence across experiments with 2 h.

10. Collect the cDNA release mix

1. Preheat NFW to 37 °C (at least 500 µL per chip of 1 cm²).
2. Remove the seal and pipette vigorously in each corner and center of the chip over the tissue, but do not scrape or touch the tissue. Gather all the cDNA Release Mix from the chip and transfer to a 1.5 or 2.0 mL centrifuge tube. Record chip identification (chip ID) numbers on each tube.
3. Add 350 µL of pre-heated NFW directly onto the chip surface. And once again pipette vigorously with a smaller pipette (P100 or P200), tilting the tip to apply more shear force, but do not touch the tissue or scrap the chip.
4. Combine the wash from step 10.3 (350 µL) with cDNA Release Mix from step 10.2 (400 µL). Make sure to combine only the samples from a single chip.
5. Place 100 µL of NFW on the chip, seal the Stereo-seq chip slide in the cassette, and store it in the fridge until the end of the protocol.
   NOTE: In cases of low yield, consider re-releasing the cDNA. The presence of DNA on the chip can be assessed using an ssDNA dye and reimaging with or without a coverslip.

11. Collect the cDNA bead clean up

1. Resuspend cDNA from step 10.4 with an equal volume of Solid Phase Reversible Immobilization (SPRI) beads. Mix until the beads are evenly distributed throughout the solution to select DNA fragments over ~100 bp.
2. Incubate for 10 min at a temperature from 27 °C to 37 °C (30° C recommended).
   1. During incubation. Prepare 5 mL of 80% fresh ethanol in NFW for every 2 chips processed.
3. Take the tubes with the incubated bead mix and briefly spin down at 100 x g on a tabletop centrifuge.
4. Place the tubes onto a magnetic separation rack and incubate for at least 5 min (until the liquid becomes clear).
5. Carefully remove the supernatant without disturbing the beads and save it in a clean labeled tube. Label cDNA release with chip ID and date.
6. Add 1.5 mL of 80% ethanol to the tubes without disturbing the beads and incubate for 30 s before removing the ethanol.
7. Repeat step 11.6 one additional time, this time removing as much ethanol as possible using a small pipette tip without disturbing the beads.
8. Air dry the beads in the tubes until they are no longer reflective. The time will vary, so check regularly to prevent overdrying (cracks), which will result in low yield.
   NOTE: Measure the humidity and temperature of the room. At high temperatures (27 °C) and low humidity (less than 20%), beads will dry very quickly (3 to 5 min). At low temperatures (23 °C), and high humidity (more than 40%), beads will dry much more slowly (7 to 15 min).
9. Remove the bead tubes from the magnet rack, add 22 µL of NFW, and mix by vortexing.
10. Incubate in a dry bath or PCR machine at 27 °C for 5 to 15 min.
11. Briefly centrifuge and place the tube back onto the magnetic separation rack until the liquid becomes clear (2 to 5 min).
12. Transfer 21 µL of supernatant containing cDNA into the sample-labeled 0.2 mL PCR tube without disturbing the beads.
13. Repeat steps 11.9 to 11.12 for each tube, collecting a total of 42 µL of cDNA. Continue directly with the amplification of cDNA (steps 12.1 to 12.8).
    NOTE: Delay in starting amplification will result in decreased yield, increased short reads, and degradation of single-stranded DNA template.

12. cDNA amplification

1. Remove the cDNA amplification mix and FFPE cDNA primers from the freezer and place them on wet ice to thaw.
2. Spin down the cDNA amplification mix and FFPE cDNA primers on a tabletop centrifuge at ~100 x g.
3. Add 50 µL of cDNA amplification solution and 8 µL of FFPE cDNA primers to each cDNA tube from 42 µL from (step 11.13) and place in a thermocycler with the flat cap adapter to prevent tube deformation.
4. Run the thermocycler program for the cDNA amplification: initial denaturation: 95 °C for 5 min; cycling: [98 °C for 20 s, 58 °C for 20 s, 72 °C for 3 min] × 15 cycles; final extension: 72 °C for 5 min; holding temperature: 12 °C (if processing immediately) / 4° C (if held overnight).
5. Measure the DNA concentration of 1 µL cDNA pre- and post-bead clean up.
   NOTE: The amount of cDNA product before cleanup should be calculated to exceed 1 µg for each 1 cm² chip. The post-cleanup concentration should be at least 10 ng/µL. Lower yields are typically associated with reduced library diversity. Significant sample loss at this stage is correlated with DV200 scores; however, DV200 is not a perfect predictor of results.
6. Repeat cDNA bead cleanup from steps 11 (bead clean up) utilizing the PCR product from step 12.4, with three modifications: use smaller PCR magnetic rack, add only 200 µL of 80% ethanol in step 11.6, and resuspend in 22 µL of Tris-EDTA (TE) buffer (pH 8) in step 11.9.
7. Store beads in a tube with 20 µL of NFW, labeled with sample names, barcode, date, and chip ID.
8. Measure the cDNA fragment distribution using a fragment analyzer. The fragment size range should be between 250-400 bp. If the fragments are too small, repeat bead cleaning but use room temperature (~25 °C) instead of 27 °C.
   OPTIONAL STOPPING POINT: After measuring the cDNA, it can be stored in -20 °C for up to 1 month and -80 °C for up to 1 year. Beads can be stored in NFW for 1 week at 4 °C. Do not freeze beads.

13. Library preparation (Day 4)

1. Amplification Reaction Mix Preparation
   1. Remove the library PCR amplification solution, library barcode sets, cDNA product, and NFW. Thaw the cDNA and reagents on ice and water at room temperature and spin down at 100 x g for 10 s.
2. Determine the library barcode strategy for each 25 µL of PCR barcode primer mix.
   NOTE: For this sequencer, all bases must be represented equally at each position in a barcode set. Most commercial kits compatible with this are therefore designed with predefined barcode tetrads (sets of four). These tetrads can be applied to individual samples when splitting across a sequencing lane, or they can be pooled together for a single sample when sequencing one sample per lane. Maintaining balanced barcodes is imperative, as unbalanced barcodes will reduce the amount of usable data generated by the sequencer.
3. Determine the volume of cDNA for 20 ng (e.g., 10 ng/µL would need 2 µL). Using this make a NFW (25 μL) solution of diluted cDNA (dilute to ~1 ng/μL).
4. Combine library amplification mix (50 µL) with diluted cDNA in NFW (25 µL) and PCR barcode primer mix (25 µL) to the 0.2 mL PCR tube.
5. Mix the well by pipetting and spin down at 100 x g for 10 s. (Do not vortex)
6. Put the 0.2 mL PCR tubes into the thermocycler and make sure to have an insert to prevent tubes from being crushed.
7. Run the thermocycler program for the cDNA amplification: initial denaturation: 95 °C for 5 min; cycling: [98 °C for 20 s, 58 °C for 20 s, 72 °C for 3 min] × 8 cycles; final extension: 72 °C for 5 min; holding temperature: 12 °C (if processing immediately) / 4 °C if held overnight.
   NOTE: It is possible to use varying amounts of input cDNA to prepare the library. However, the amount used will affect the number of PCR cycles required for amplification. For example, the vendor recommends 20 ng of input DNA and the use of 8 cycles. It is also possible to use 40 ng with 7 cycles or 80 ng with 6 cycles. Using less than 20 ng or more than 80 ng of input cDNA is not recommended.

14. Library amplification bead clean up

1. Quantify the Raw Library Product as done in step 12.5.
2. Repeat bead cleanup from step 11 (bead clean up) utilizing the raw library product from step 13.7 with three modifications: use 0.8:1 ratio bead to product, use smaller PCR magnetic rack, use only 200 µL of 80% ethanol compared to step 11.7, and resuspend in 22 µL of TE buffer (pH 8) in step 11.9. Prepare around 1mL of fresh 80% ethanol fresh per two 1cm² chips.
3. OPTIONAL STOPPING POINT: After measuring the library, it can be stored at -20 °C for up to 1 month and -80 °C for up to 1 year. Beads can be stored at NFW for 1 weekat 4 °C. Do not freeze the beads.

15. DNBSEQ-T7RS system for sequencing Stereo-Seq Random Oligonucleotide Primed DNBs

1. Using a PE75 kit on the DNBSEQ-T7RS system, allocate the cycles as 25 for Read 1 and 62 for Read 2 (PE25+62), followed by a 10 bp index:
   1. The first read (read 1 of 25 cycles) captures the spatial Chip ID (CID) or location-specific information from the chip-based barcodes.
   2. The second read (read 2 of 62 cycles) captures the Molecular Identifier (MID), a random 6 bp sequence, directly attached to the gene inserts with 3 additional dark cycles used for phasing correction and barcode offset calibration. This is followed by a gene insert of 30-59 bp from read 2, which is identified by mapping to known genomes or transcriptomes.
      NOTE: Longer gene inserts are possible, but would require different paired end kits, and modification to the mapping workflow.
   3. The index cycle for demultiplexing covers 10 bps based on the sequencing barcodes pooled in step 13.2.
      NOTE: Typical sequencing time of approximately 10-12 h for a 75 PE flow cell. Sequencing is most often done at a sequencing core, or commercial facility, and not an in-house process for most labs, as it requires a DNBSEQ-T7RS or similar sequencing instrument.

The data below demonstrates the ability to perform single-cell segmentation using the SAW pipeline and mapping to non-poly adenylated RNAs such as tRNA (TRDMT1) from an FFPE section. The unbiased data from Stereo-seq allow one to examine the importance of not just mRNAs but also tRNAs, rRNAs, and non-host RNAs (if pertinent). Using the direct output from the SAW pipeline16 and the stereopy for quality assessment of the Stereo-seq output in Figure 3A-G, we generated standard spatial overviews, gene expression, as well as clustering and QC from the cell bins (single nucleus overlays created via the saw nuclei image). The number of transcripts detected varies depending on tissue type: immune-dense tissues typically exhibit lower transcript counts (approximately 100 genes per cellbin), whereas aneuploid tumor cells can have as many as 1,000 transcripts per cellbin. Please note that the clustering of this sample would still require bioinformatics processing to provide cluster identities, as the SAW default resolution is very high (0.8 to 1), and filtering the data based on read counts, poorly defined cells/nuclei are recommended using Python packages such as stereopy4. Cellbins containing greater than 100 genes are considered acceptable for most analysis, as this data is similar to single-cell nuclei data. During filtering, classical single cell metrics can be used, including gene per cellbin of < 100 (removing low cellbins), cellbin area to remove undefined cellbin, and duplicates or hot spots cellbin. Other spatial analysis or filtering can include tissue edge effect, space to nearest cells. Other cell analysis, such as a high percentage of mitochondria per cellbin is not suggested (Figure 3H).

Next, the cellbin using the proprietary software (StereoMap4) and stereopy (coding-based) provides examples of spatial visualization. Figure 4A demonstrates the red box using nuclei ssDNA image. The interactive software shows the spatial distribution of putative cell types, such as Figure 4B and the single gene interactively Figure 4C, such as a non-poly adenylated RNA tRNA aspartic acid methyltransferase 1 (TRDMT1). This transcript is also visualized across the whole tissue using stereopy Figure 4D. However, a more detailed analysis of the spatial relationships of the single cell and transcriptomics is required to provide insight into the importance of any transcripts observed.

Figure 1: Workflow diagram for the protocol. (A) Sample selection and capture chip setting. (B) Day 1: capture chip preparation, sectioning, and overnight baking (during this time, the chip should not be touched). (C) Day 2: Short bake, deparaffinization, staining, and imaging (during this time, touching of the chip should be limited to prevent scratch artifacts). (D) Day 2: decrosslinking, methanol fixation, permeabilization, FFPE reverse transcription should be done with slow even pipetting to prevent diffusion artifacts, (E) Day 3: cDNA release prerelease slow pipetting, post release vigorous pipetting to collect all cDNA, cDNA cleaned up, cDNA amplification and a second clean up, (F) Day 4 is the QC for the cDNA, library barcode selection, NGS library preparation and submission. During Day 1-2, it is very necessary to be very careful with the slide and chip and not scratch the surface. However, in the cDNA release step (when removing the cDNA), it is necessary to be very vigorous when pipetting to get decent yields. Please click here to view a larger version of this figure.

Figure 2: User innovations. First, the application of resin to be used when clearing with xylenes (A) adding resin, (B) curing resin, (C) cleaning resin after curing. Second, the application of PLL (D) adds PLL to the coat slide. The third is the slide drying technique (E) angle and closeness of compressed air during drying. The last adjustment of a silicone chamber for the rehydration process, (F) placing a cut silicone chamber round chip area, (G) lift test to check silicone chamber adherence, (H) adding or removing liquid from the silicone chamber without touching the chip. Please click here to view a larger version of this figure.

Figure 3: Quality control assessment. This figure covers basic outputs from stereopy and saw to assess the quality of the sample. First (A) ssDNA nuclear staining image required to create single-cell level bins (cellbins) across the tissue. It is a confocal microscope stitched single-channel ssDNA staining image for the nuclear location and track lines. The stereopy (B) spatial gene expression capture map with the number of genes captured per cellbin. SAW output showing pre filtered pipeline for cellbin output including (C) single cell level spatial lieden clustering, (D) lieden cluster UMAP, and (E-G) violin plots of (E) molecular identifier count, (F) gene type count and (G) cell area distribution per cellbin. Lastly, a stereopy quality control metrics standard for single cell sequencing (H) percentage mitochondrial by total count by cellbin. Please click here to view a larger version of this figure.

Figure 4: Simple spatial visualizations. This figure provides a (A) ssDNA nuclear staining image for orientation, with the red box highlighting the regions zoomed in regions in B and C. Images (B-C) provide a StereoMap 4 software, which provides interactive user visualization, (B) the cellbin with liaden cluster labels, (C) with expression levels of tRNA aspartic acid methyltransferase 1 (TRDMT1) with cellbins. Lastly, (D) shows a stereopy spatial expression of TRDMT1 across the whole tissue. Please click here to view a larger version of this figure.

Table 1: Method comparisons of Sequencing-based Spatial Transcriptomics methods. This table summarizes key features of major sequencing-based spatial transcriptomics platforms currently available. Methods are compared by tissue fixation compatibility, field of view (FOV, in mm), spatial resolution (in µm), RNA species captured, estimated protocol hands-on time (in h), total protocol duration (in days), and relevant citations or commercial protocol1,13,20,21,22,23,24.

The Stereo-seq protocol is highly detailed and involves several critical steps that require precision, timing, and a clean environment to ensure success especially with microbiome or non-host profiling that may require specific precautions.

The use of nuclease free water during all steps is required, and the capture chip must not be touched with anything other than the sample and reagents. Exercise extreme caution when using the cassette assembly, taking time to utilize practice chips. A scratch on the tissue area will ruin tissue spatial resolution, and a scratch on the track lines can likely prevent the spatial resolution or the sample from being automatically aligned. Further, for high RNase/DNase tissues such as pancreatic or liver, changes of water during sectioning are recommended. Additionally, to maintain RNA quality it is suggested to store FFPE tissue blocks with a small amount of wax-seal on the face, at 4 °C, and in a desiccator to prevent degradation. If stored this way, the samples will still need to be rehydrated on the surface and equilibrated to cutting room temperature before using^25,26,27. The cutting room should be at 22 to 25 °C. Furthermore, the samples best suited for Stereo-seq are ones with a DV200 score of higher than 50, but scores with DV200 over 30 may also be sequenced with limited results. If microbiome is of interest, it is best practice to not reuse any materials and only use fresh reagents, wear masks, and keep areas clean if not sterile for the whole process.

For single-cell resolution and cellular binning (cellbin), imaging of the cell nuclei with ssDNA dye is required. However, the concentration of this dye is camera-specific, and dependent on the dynamic range that can be imaged by the system. This requires titration and testing of dye concentrations for each new device, typically a 1:2 to 1:10 dilution of the ssDNA dye from manufacturer recommendations. During imaging, it is recommended to manually focus on 3-5 areas without tissue and 4-6 areas with tissue; however, most consistent results occurred using a 13-point setup to ensure quality control with clear cellular focus, as both the track lines and tissue must be found in the focal depth.

Certain reagents are more sensitive to timing than others. Specifically, the permeabilization reagent must be resuspended in a 0.01 N HCl solution (pH 2.0 +/- 0.1). Store the aliquoted 10x PR solution to avoid degradation from freeze-thaw cycles. Timing varies by tissue type, with thicker sections requiring between 15 and 30 min when embedded in soft paraffin. Handle chips carefully post-permeabilization to avoid damage, diffusion, and artifacts from disturbing the chip/tissue surface. The sequencing strategy for the library barcode mix is a very important pooling strategy for sequencing and multiplexing. If sequencing a single sample on a single lane, it is easiest to just use a barcode set (BC1-4, BC5-8, BC9-12, or BC13-16). Other strategies are detailed in the manufacturer's handbook.

The addition of a ceramic resin sealant around the chip can help prevent xylene's penetration of the adhesive that seals the chip to the slide (Figure 2A-C). In addition, the use of poly-L-Lysine is recommended for high fatty tissues and fragile tissues, but may lead to issues with capture due to residues from the PLL drying processes (Figure 2D-E). As such, it is not recommended for all tissues. To reduce reagents, utilize a silicone chamber (Figure 2F-H), which decreases the volume from 30 mL to 500 µL of reagent needed, specifically for microbiome work, which requires fresh reagents for each slide. It is recommended to do additional filtration of reagents when trying to capture microbiome data. Further, utilizing sealed ethanol that is only accessed via a needle/syringe will provide cleaner reactions than ethanol that is open to the air. SPRI beads concentration can be modified by decreasing the concentration to select longer fragments, but this will reduce the overall yield. This protocol was modified on the SPRI bead steps to have a more consistent yield by warming the beads to (~30 °C) and using double elution to increase yield consistency.

Stereo-seq is currently one of the highest resolutions with the largest capture areas for sequencing-based spatial transcriptomics (SST) methods. It offers nanoscale resolution (200 nm nanoball spot size with center-to-center distance of 500 nm), species-agnostic capture, and a large field of view (up to 13 cm x 13 cm). By comparison, Slide-seq V2 achieves ~10 µm resolution with high sensitivity (~50% RNA capture efficiency) but is limited by smaller capture areas with bead-based barcoding²⁴. Similarly, probe based whole transcriptome capture methods and spatial transcriptomics A'tailling capture method provided robust whole mRNA transcriptome profiling using A-tailing capture, but at a much lower resolution (55 µm spot size with center-to-center distance of 100 µm), requiring computational methods to deconvolute the cell mixtures per spot. The newer whole genome probe-based capture methods have improved resolution of 2 µm tiles (recommending 8 µm x 8 µm) and rely on species-specific probe panels mostly focused on protein coding genes; however, the method tends to be more robust and less finicky. Another SST method, DBiT-seq, uses microfluidics-based spatial barcoding to achieve ~10 µm near single-cell resolution with flexibility in experimental design, but lacks the nanoscale precision and scalability of Stereo-seq²⁸. Other cellular-level resolution SST platforms with cellular dissociation are the easiest to combine with single-cell sequencing data, but this method has many of the disadvantages of single-cell sequencing due to its methodology of disaggregation, filtering, and profiling using single-cell RNA methods. Lastly, Slide-tag (10 µm resolution) is more cost-effective but currently does not match Stereo-seq's nanoscale capabilities²⁸.

Aside from SST methods, probe-based methods excel in targeted transcript detection with high sensitivity but are limited to predefined probes, although some have expanded to up to 5,000 probes panels^29,30,31. Still, Stereo-seq provides unbiased whole-transcriptome analysis limited to SST methods^28,32. Imaging-based methods^33,34 achieve single-molecule or single-cell resolution but are constrained by multiplexing limits and imaging throughput. Overall, Stereo-seq is a labor-intensive process, including multiple time-sensitive steps that are extremely influenced by the hands of the operator for the duration of the multiday protocol. The protocol capture rate for each individual space will not be as sensitive as target probe-based methods or imaging methods, but offers unparalleled resolution and unbiased capture in untargeted spatial transcriptomics, as highlighted by its species-agnostic capture ability. The protocol involves a multi-day workflow with numerous steps critical steps, including tissue preparation for sectioning and adhesion, deparaffination, rehydration and staining of the tissue before time-sensitive fluorescent imaging, followed time and pH sensitive by tissue permeabilization, RNA capture, in situ RNA amplification, cDNA release, clean up and amplification, and lastly by library preparation. All of which require meticulous attention, and unique technical skills often best performed by different people, each with detail and precise timing to ensure high-quality results. Many of these steps are highly time-sensitive, particularly tissue permeabilization, RNA capture and enzymatic reactions, where delays or deviations can lead to RNA degradation or loss of spatial resolution. In addition, after permeabilization, there is an increased chance of diffusion the more the slide is moved. Lastly, while library preparation itself does not demand extensive or specialized equipment, relying primarily on standard laboratory tools such as thermocyclers, pipettes, and magnetic racks for bead purification, it still requires an experienced operator to prevent RNA loss, PCR bubbles, or drying during amplification, and understanding of the issues such as uneven amplification, tissue sectioning artifacts, or contamination. This reliance on skilled personnel adds to the complexity of the protocol.