Highlights
-
Deterministic barcoding in tissue enables NGS-based spatial multi-omics mapping
-
DBiT-seq identified spatial patterning of major tissue types in mouse embryos
-
Revealed retinal pigmented epithelium and microvascular endothelium at cellular level
-
Direct integration with scRNA-seq data allows for rapid cell type identification
Summary
We present deterministic barcoding in tissue for spatial omics sequencing (DBiT-seq) for co-mapping of mRNAs and proteins in a formaldehyde-fixed tissue slide via next-generation sequencing (NGS). Parallel microfluidic channels were used to deliver DNA barcodes to the surface of a tissue slide, and crossflow of two sets of barcodes, A1-50 and B1-50, followed by ligation in situ, yielded a 2D mosaic of tissue pixels, each containing a unique full barcode AB. Application to mouse embryos revealed major tissue types in early organogenesis as well as fine features like microvasculature in a brain and pigmented epithelium in an eye field. Gene expression profiles in 10-μm pixels conformed into the clusters of single-cell transcriptomes, allowing for rapid identification of cell types and spatial distributions. DBiT-seq can be adopted by researchers with no experience in microfluidics and may find applications in a range of fields including developmental biology, cancer biology, neuroscience, and clinical pathology.
Introduction
In multicellular systems, cells do not function in isolation but are strongly influenced by spatial location and surroundings (
;
;
). Spatial gene expression heterogeneity plays an essential role in a range of biological, physiological, and pathological processes (
;
;
). For example, how stem cells differentiate and give rise to diverse tissue types is a spatially regulated process that controls the development of different tissue types and organs (
;
). Mouse embryonic organogenesis begins at the end of the first week, follows gastrulation, and continues through birth (
). When, and how exactly, different organs emerge in an early embryo is still inadequately understood due to highly dynamic spatial organization of tissues and cells at this stage. An embryonic organ could differ substantially in anatomical and molecular definitions as compared to the adult counterpart. In order to dissect the initiation of early organogenesis at the whole embryo scale, it is desirable to not only measure genome-wide molecular profiles for cell type identification but also interrogate spatial organization in the tissue context with high spatial resolution.
Despite the latest advent of massively parallel single-cell RNA-sequencing (scRNA-seq) (
;
) that revealed astonishing cellular heterogeneity in many tissue types, including the dissection of all major cell types in developing mouse embryos from E9 to E14 (
;
), the spatial information in the tissue context is missing in scRNA-seq. Spatial transcriptomics emerged to address this problem (
). Early attempts were all based on multiplexed single-molecule fluorescent
in situhybridization (smFISH) via spectral barcoding and/or sequential imaging (
;
). It evolved rapidly over the past years from detecting a handful of genes to hundreds or thousands (e.g., seqFISH, MERFISH) (
;
), and recently to the whole transcriptome level (e.g., seqFISH+) (
). However, these methods were technically demanding, requiring high-sensitivity single-molecule fluorescence imaging systems, sophisticated image analysis processes, and a lengthy repeated imaging workflow to achieve high multiplexing (
). Moreover, they were all based upon a finite panel of probes that hybridize to known mRNA sequences, limiting their potential to discover new sequences and variants. Fluorescent
in situsequencing methods (e.g., FISSEQ, STARmap) (
;
) were additionally reported, but the number of detectable genes was limited, and their workflow resembled sequential FISH, again requiring a lengthy, repeated, and technically demanding imaging process.
It is highly desirable to develop new methods for high-spatial-resolution, unbiased, genome-scale molecular mapping in intact tissues at cellular level, which does not require sophisticated imaging but capitalizes on the power of next generation sequencing (NGS) to achieve higher sample throughput and cost efficiency. Spatial transcriptome mapping at cellular level (spot size ∼10 μm) was demonstrated with Slide-seq that utilized a self-assembled monolayer of DNA-barcoded beads on a glass slide to capture mRNAs released from a frozen tissue section placed on top (
). A similar method, called high-definition spatial transcriptome (HDST), used 2 μm beads in a microwell array chip to further reduce the spot size (
). However, these emergent methods are limited by low number of detected genes (∼150 genes per pixel in Slide-seq), incompatibility with fixed tissues, potential lateral diffusion of release mRNAs, and sophisticated process for bead decoding. Moreover, they are all limited to spatial transcriptomes and have yet to realize multi-omics spatial sequencing.
We sought to develop a completely different approach, which was to spatially barcode biomolecules in tissues rather than to capture them on a solid-phase substrate. Previously, we developed microfluidic channel-guided patterning of DNAs or antibodies on a glass slide for multiplexed protein assay (
,
). We speculated that a microfluidics-confined delivery of molecular barcodes to a tissue section could enable high-spatial-resolution barcoding of mRNAs or proteins directly in tissue. Herein, we report on deterministic barcoding in tissue for spatial omics sequencing (DBiT-seq). A microfluidic chip with parallel channels (10, 25, or 50 μm in width) was placed directly against a fixed tissue slide to introduce oligo-dT-tagged DNA barcodes A1–A50 that annealed to mRNAs to initiate
in situreverse transcription. This step resulted in stripes of barcoded cDNAs inside the tissue. Afterward, the first microfluidic chip was removed and another chip was placed on the same tissue slide with the microchannels perpendicular to the first flow direction to introduce a second set of DNA barcodes B1–B50, which were subsequently ligated at the intersections to form a 2D mosaic of tissue pixels, each containing a distinct combination of barcodes Ai and Bj (i = 1–50, j = 1–50). Then, the tissue was digested to recover spatially barcoded cDNAs that were collected to an Eppendorf tube, template-switched, PCR amplified, and tagmentated to prepare a library for NGS sequencing. Proteins could be co-measured by applying a cocktail of antibody-derived DNA tags (ADTs) to the fixed tissue slide prior to flow barcoding, similar to Ab-seq or CITE-seq (
;
). We demonstrated high-spatial-resolution co-mapping of whole transcriptome and a panel of 22 proteins in mouse embryos (E10–E12). DBiT-seq faithfully detected all major tissue types in early organogenesis and identified the fine features such as brain microvascular networks and a single-cell-layer of melanocytes lining an optical vesicle. We found that the gene expression profiles of 10-μm tissue pixels were dominated by single-cell transcriptomes, and an integrated analysis allowed for rapid identification of cell types in relation to spatial distribution. The microfluidic chip was directly clamped onto the tissue slide, and the reagent dispensing was performed by directly pipetting into the inlet holes, requiring no prior experience in microfluidic control. Thus, DBiT-seq could be readily adopted by researchers from a wide range of fields in biological and biomedical research.
ResultsDBiT-Seq Workflow
The workflow of DBiT-seq is described in Figure 1A (see also Figure S1). A tissue section pre-fixed with formaldehyde on a standard aminated glass slide was used. A polydimethylsiloxane (PDMS) microfluidic chip (Figure S1C) containing 50 parallel microchannels (down to 10 μm in width) was placed on the tissue slide to introduce a set of DNA barcode A solutions. To assist the assembly, an acrylic clamp was used to hold the PDMS chip firmly against the tissue slide (Figure 1B). The inlet holes were ∼2 mm in diameter and ∼4 mm in depth allowing for ∼5 μL reagents to be directly pipetted into the inlets. The outlet holes were roofed with a global cover connected to a house vacuum to pull the reagents all the way from the inlets to the outlets through the tissue surface, which took several seconds for a 50 μm chip and up to 3 min for a 10 μm chip. Barcode A is composed of an oligo-dT sequence for binding mRNAs, a distinct spatial barcode Ai (i = 1–50, 8-mer), and a ligation linker (15-mer). Reverse transcription was conducted during the first flow for in situ synthesis of first strand cDNAs that immediately incorporate barcode A. Then, the first PDMS chip was removed and another PDMS chip with the microchannels perpendicular to those in the first flow barcoding was placed on the same tissue to introduce a second set of barcodes Bj (j = 1–50), each containing a ligation linker (15-mer), a distinct spatial barcode Bj (j = 1–50, 8-mer), a unique molecular identifier (UMI), and a PCR handle (22-mer) functionalized with biotin, which was used later to perform cDNA purification with streptavidin-coated magnetic beads. Also added to the barcode B reagents were T4 ligase and a complementary ligation linker to perform in situ ligation at the intersections, resulting in a mosaic of tissue pixels, each containing a distinct combination of barcodes Ai and Bj (i = 1–50, j = 1–50). The tissue slide being processed could be imaged during each flow or afterward such that the tissue morphology can be correlated with spatial omics map. To co-measure proteins and mRNAs, the tissue slide was stained with a cocktail of 22 antibody-derived DNA tags (ADTs) (
) (see Table S1) prior to microfluidic flow barcoding. Each of the ADTs contains a distinct barcode (15-mer) and a polyadenylated tail that allowed for protein detection using a workflow similar to that for mRNAs detection. After forming a spatially barcoded tissue mosaic, cDNAs were collected, template-switched, and PCR amplified to make a sequencing library. Using a paired-end (2 × 100) NGS sequencing, we detected spatial barcodes (AiBj, i = 1–50, j = 1–50) from one end and the corresponding transcripts and protein barcodes from the other end to computationally reconstruct a spatial expression map. It is worth noting that unlike other methods, DBiT-seq permits the same tissue slide being imaged during or after the flow barcoding (Figure S1D) to precisely locate the pixels and perform correlative analysis of tissue morphology and spatial omics maps at high precision. The sequences of ADTs, DNA barcodes, and key reagents are summarized in Tables S1, S2, and S3, respectively.
