Seq-Scope Protocol: Repurposing Illumina Sequencing Flow Cells for High-Resolution Spatial Transcriptomics

Abstract

Spatial transcriptomics (ST) technologies represent a significant advance in gene expression studies, aiming to profile the entire transcriptome from a single histological slide. These techniques are designed to overcome the constraints faced by traditional methods such as immunostaining and RNA in situ hybridization, which are capable of analyzing only a few target genes simultaneously. However, the application of ST in histopathological analysis is also limited by several factors, including low resolution, a limited range of genes, scalability issues, high cost, and the need for sophisticated equipment and complex methodologies. Seq-Scope-a recently developed novel technology-repurposes the Illumina sequencing platform for high-resolution, high-content spatial transcriptome analysis, thereby overcoming these limitations. Here we provide a detailed step-by-step protocol to implement Seq-Scope with an Illumina NovaSeq 6000 sequencing flow cell that allows for the profiling of multiple tissue sections in an area of 7 mm × 7 mm or larger. In addition to detailing how to prepare a frozen tissue section for both histological imaging and sequencing library preparation, we provide comprehensive instructions and a streamlined computational pipeline to integrate histological and transcriptomic data for high-resolution spatial analysis. This includes the use of conventional software tools for single cell and spatial analysis, as well as our recently developed segmentation-free method for analyzing spatial data at submicrometer resolution. Given its adaptability across various biological tissues, Seq-Scope establishes itself as an invaluable tool for researchers in molecular biology and histology.

Fig. 1.

Overview of Experimental Procedures.

Fig. 2.

Example outputs from the Illumina SAV application showing (a) base composition at each cycle, (b) raw (Density) and pass filter (PF) cluster density in each lane, and (c) scatterplot of cluster occupancy (x axis) and % PF in each lane (blue – Lane 1, cyan – Lane 2, yellow – Lane 3, red – Lane 4). Each dot represents an individual tile in one of the four flow cell lanes.

Fig. 3.

Flow cell disassembly and dicing. (a) Disassembly of NovaSeq S4 flow cell using a scalpel. (b) Flow cell disassembled into top, middle, and bottom layers. The top layer is upside down to position the cluster surface upward. (c) Glass dicing procedure. (d) Schematic of the flow cell diced into separate fragments and imaging areas.

Fig. 4.

Liquid Handling in Seq-Scope Chip Surface Treatment. (a) Enzymatic and chemical solutions were first placed on Parafilm in a humidified chamber. (b) The chip is then overlaid onto the droplet solutions to treat the cluster surface.

Fig. 5.

Tissue Freezing Chamber. (a) Schematic instructions for preparing a tissue freezing chamber. An LN₂-chilled isopentane bath provides a medium for rapidly freezing OCT tissue block. (b) Demonstration of the tissue freezing procedure.

Fig. 6.

Tissue Sectioning and Attachment. (a) Cryostat set-up for sectioning. (b) Demonstration of maneuvering tissue onto the Chip surface. (c) Tissue melting for Chip attachment.

Fig. 7.

Liquid and Chip Handling Procedure in STEP 2. Tissue Fixation and Imaging Steps. The photographs are intended to serve as a guide in how to perform the procedures; therefore, actual tissues are not included in the pictures. (a) The diced Chip laid flat on the laboratory bench surface. (b) Chip covered with 1 mL of PBS. All liquid handling steps in tissue fixation and imaging will be performed similarly. The surface tension of all liquids is strong enough to hold them together on the Chip surface. (c-d) Setup of a glycerol-mounted chip/coverslip assembly on the bench (c) or in the Keyence imager (d).

Fig. 8.

Seq-Scope Adapter Frame and Silicone Isolator used in STEP 3. Library Construction Procedures. (a) Schematic diagram showing how the Chip is placed in the Adapter Frame and how Silicone Isolators are attached to the Chip. (b) Empty Adapter Frame. (c) Adapter Frame loaded with a Chip attached to Silicone Isolators. (d) Example of how liquid handling is performed on a Seq-Scope Chip, attached by Silicone Isolators to the Adapter Frame.

Fig. 9.

Overview of Computational Procedures

Fig. 10.

Spatial arrangement between tiles, which is determined by comparing the histology images and the spatial barcodes.

Fig. 11.

An output sbcd image from step 1–4), showing the distribution of spatial barcodes (i.e., HDMIs). The fiducial marks should be aligned horizontally and vertically with correct parameter settings.

Fig. 12.

An output smatch image from Step 2–1), showing the distribution of spatial barcodes (i.e., HDMIs) that matched to the 2^nd-seq FASTQ files. The distribution should be consistent to the tissue area.

Fig. 13.

An output sge image from step 2–2), showing the distribution of spatial barcodes (i.e., HDMIs) that aligns to the reference genome, coloring exon-aligned transcripts as red, unspliced transcripts as green, and mitochondrial transcripts as blue.

Fig. 14.

Comparison between sge (left) and aligned histology (right) images. Boxed areas are sequentially magnified below. The aligned histology image is the output image from step 2–3), a tiff file with histology imaging information that has the same dimension as in smatch (Fig. 12) and sge (Fig. 13) images.

Fig. 15.

An exemplary output image from step 3–2, illustrating distinct zonation of hepatocellular factors. Refer to Fig. 16 for the color-coding scheme. The LDA settings are d18 (10 μm-sided hexagons) and nF6 (number of factors = 6). Boxed area is magnified on the right.

Fig. 16.

Factor color legend for Fig. 15. Factors 0–4 represents metabolically zonated hepatocytes along the portal-central axis. 0: Extreme periportal, 1: Periportal, 2, 4: Central, 3: Extreme Central. 5: Red blood cells.

Fig. 17.

An exemplary multi-dimensional clustering result obtained using Seurat. (A) UMAP manifold visualizing 20 clusters representing different cell types. Notable clusters include 0, 3, 5: Pericentral hepatocytes, 2, 4: Periportal hepatocytes, 1: Midzone hepatocytes, 6, 16: B cells, 7: Cells undergoing antiviral response, 9: Red blood cells, 10: hepatic stellate cells, 11: Myofibroblasts and macrophages, 12: Hepatocytes with damage response, 13: Cholangiocytes/liver progenitors. The clusters were obtained using following parameters: 10 μm-sided hexagons (d18) and FindCluster resolution of 1.00. These results reproduce the liver results that were described in the original Seq-Scope paper.

Fig. 18.

An exemplary pixel-level output image from step 3–4) showing clear zonation of hepatocellular factors. The color-coding scheme is the same as shown in Fig. 16. Boxed area is magnified on the right.

Fig. 19.

Pixel-level projection of Seurat cluster factors learned from the analysis shown in Fig. 17. The color coding scheme for each factor is presented in Fig. 20. Boxed area is magnified on the right.

Fig. 20.

Factor color legend for Fig. 19. Color code for each factor whose number is corresponding to what was presented in Fig. 17.

Fig. 21.

Seq-Scope analysis using shallowly sequenced liver data (~163 million 2^nd-Seq read inputs). (a) sge image produced as in Fig. 13. (b) FICTURE projection image produced as in Fig. 18 with d24 (14 μm-sided hexagons) of shallowly sequenced data. (c) Spatial gene expression plot visualizing each indicated transcript with dots of corresponding color and varying alpha levels. Warm colors represent pericentral expression while cold colors represent periportal expression. Abundantly expressed genes, such as Alb, were plotted with lower alpha levels. (d) H&E histology overlaid with spatial gene expression plots visualizing central area-specific Glul and Oat transcripts.