Spatial and molecular characteristics determine tissue function, yet high-resolution methods to capture both concurrently are lacking. Here, we developed high-definition spatial transcriptomics, which captures RNA from histological tissue sections on a dense, spatially barcoded bead array. Each experiment recovers several hundred thousand transcript-coupled spatial barcodes at 2-μm resolution, as demonstrated in mouse brain and primary breast cancer. This opens the way to high-resolution spatial analysis of cells and tissues.
Access optionsAccess options
Subscribe to Journal
Get full journal access for 1 year
only $20.17 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
The raw mouse data have been deposited to NCBI’s GEO archive GSE130682. Raw files for the breast cancer sample are available through an MTA with Å. Borg (email@example.com). All processed data is available at the Single Cell Portal (https://portals.broadinstitute.org/single_cell/study/SCP420).
All code has been deposited on GitHub at https://github.com/klarman-cell-observatory/hdst.
Oh, S. W. et al. Nature 508, 207–214 (2014).
Lein, E., Borm, L. E. & Linnarsson, S. Science 358, 64–69 (2017).
Macosko, E. Z. et al. Cell 161, 1202–1214 (2015).
Zheng, G. X. Y. et al. Nat. Commun. 8, 14049 (2017).
van den Brink, S. C. et al. Nat. Methods 14, 935–936 (2017).
Chen, K. H., Boettiger, A. N., Moffitt, J. R., Wang, S. & Zhuang, X. Science 348, aaa6090 (2015).
Wang, X. et al. Science 361, eaat5691 (2018).
Coskun, A. F. & Cai, L. Nat. Methods 13, 657–660 (2016).
Beliveau, B. J. et al. Nat. Commun. 6, 7147 (2015).
Lee, J. H. et al. Science 343, 1360–1363 (2014).
Ke, R. et al. Nat. Methods 10, 857–860 (2013).
Ståhl, P. L. et al. Science 353, 78–82 (2016).
Michael, K. L., Taylor, L. C., Schultz, S. L. & Walt, D. R. Anal. Chem. 70, 1242–1248 (1998).
Gunderson, K. L. et al. Genome Res. 14, 870–877 (2004).
Nagayama, S., Homma, R. & Imamura, F. Front. Neural Circuits 8, 98 (2014).
Lein, E. S. et al. Nature 445, 168–176 (2007).
Zeisel, A. et al. Cell 174, 999–1014 (2018).
Karaayvaz, M. et al. Nat. Commun. 9, 3588 (2018).
Zhang, X. et al. Breast Cancer Res. 19, 15 (2017).
Rodriques, S. G. et al. Science 363, 1463–1467 (2019).
Salmén, F. et al. Nat. Protoc. 13, 2501–2534 (2018).
Jemt, A. et al. Sci. Rep. 6, 37137 (2016).
Navarro, J. F., Sjöstrand, J., Salmén, F., Lundeberg, J. & Ståhl, P. L. Bioinformatics 33, 2591–2593 (2017).
Dobin, A. et al. Bioinformatics 29, 15–21 (2012).
Anders, S., Pyl, P. T. & Huber, W. Bioinformatics 31, 166–169 (2015).
Costea, P. I., Lundeberg, J. & Akan, P. PLoS ONE 8, e57521 (2013).
Gautheret, D., Poirot, O., Lopez, F., Audic, S. & Claverie, J. M. Genome Res. 8, 524–530 (1998).
Wong, K., Navarro, J. F., Bergenstråhle, L., Ståhl, P. L. & Lundeberg, J. Bioinformatics 34, 1966–1968 (2018).
Sommer, C., Straehle, C., Kothe, U. & Hamprecht, F. A. In IEEE International Symposium on Biomedical Imaging: From Nano to Macro 230–233 (IEEE, 2011).
Lamprecht, M. R., Sabatini, D. M. & Carpenter, A. E. Biotechniques 42, 71–75 (2007).
Lyubimova, A. et al. Nat. Protoc. 8, 1743–1758 (2013).
Schindelin, J. et al. Nat. Methods 9, 676–682 (2012).
Wolf, F. A., Angerer, P. & Theis, F. J. Genome Biol. 19, 15 (2018).
Svensson, V., Teichmann, S. A. & Stegle, O. Nat. Methods 15, 343–346 (2018).
We thank the NGI Stockholm, SciLifeLab, Flatiron Institute and Simons Foundation for providing infrastructure support. We thank L. Gaffney, A. Hupalowska and J. Rood for help with manuscript preparation. Work was supported by the Knut and Alice Wallenberg Foundation, Swedish Foundation for Strategic Research, Swedish Cancer Society and the Swedish Research Council (to S.V., J.L. and P.L.S.), the EMBO long-term fellowship (no. ALTF 738-2017) to J.K., Early Postdoc Mobility fellowship (no. P2ZHP3_181475) to D.S., the NIH HuBMAP HIVE project and the Klarman Cell Observatory and HHMI (to A.R.). S.V is supported as a Wallenberg Fellow at the Broad Institute of MIT and Harvard.
F.S., J.F., J.L. and P.L.S. are authors on patents PCT/EP2012/056823 (WO2012/140224), PCT/EP2013/071645 (WO2014/060483) and PCT/EP2016/057355 applied for by Spatial Transcriptomics AB (10x Genomics) covering the described technology. M.R. is employed by Illumina Inc. A.R. is a founder and equity holder of Celsius Therapeutics and an SAB member of Syros Pharmaceuticals and ThermoFisher Scientific.
Peer review information: Nicole Rusk was the primary editor on this article and managed its editorial process and peer review in collaboration with the rest of the editorial team.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Integrated supplementary information
(a) Split-pool bead synthesis overview. Three rounds of split-and-pool were performed to produce a bead pool with 2,893,865 different oligonucleotide combinations. In the first round (top), 65 different oligonucleotides were attached covalently to the bead surface. All 65 different bead moieties were pooled together and distributed into 211 wells. These 211 wells in round two (middle) carried 211 new individual oligonucleotide sequences. These were ligated to the beads produced in Round1 (Methods). Next, 65x211 different bead moieties were pooled together and redistributed into 211 new wells carrying new oligonucleotide sequences in Round three (bottom). Oligonucleotides were again ligated and all the beads pooled for a final bead pool of 65x211x211 different bead moieties. (b) The complete bead pool was made using 65 unique barcodes in Pool1, 211 in Pool2 and 211 in Pool3 for a total of 2,893,865 uniquely barcoded beads as shown in (a). The decoding process is split into 3 sequential hybridizations. First, Pool1 in decoded. For decoding 65 barcodes, a total of 4 cycles is needed (although 4 cycles could in theory decode 81 barcodes). In each decoding cycle, a decoder pool is added that consists of a unique combination of 65 complementary oligonucleotides. In cycle#1, complementary oligos 1-27 are labeled with a green fluorophore, oligos 28-54 with a red one and 55-65 are unlabeled (depicted as black). After hybridization, the fluorescence is recorded and generates a read-out as shown in the zoomed-in view. In cycle#2, three colors are again used but attached to different oligos: 1-9; 28-36 and 55-63 are labeled green, 10-18; 37-45 and 64-65 are labeled red and 19-17; 46-54 are not labeled. In cycle#3, the colors are again attached to different oligos as well as in cycle#4. The readout of all color combinations per decoding cycle is listed under “Unique color code” and can be spatially tracked on the array (example; (x,y) positions in the zoomed in view). The 4 decoded unique beads are identified as: bead#1, beads#2, bead#3 and bead#60. Now, each spatial (x,y) position (as detected by fluorescence) is connected to a unique ATGC barcode sequence.
(a) Barcode usage. Mean percent of spatial barcodes (y axis) that are retained in each of the following steps: after decoding (n=30 slides), barcode mapping after sequencing (n=3) and removal of “clashing” (redundant) barcodes (n=30 slides). (b) Sequencing depth (y axis, left) and library saturation (y axis, right) for three replicate MOB HDST libraries (n=3 sections; Replicate 1 is derived from one animal and Replicates 2-3 from a second animal). (c) Complexity of barcodes, UMIs and genes for three replicate MOB HDST libraries. Average number (y axis) of barcodes and UMI counts (left, x axis) and genes (right, x axis) recovered under the tissue boundaries after sequencing and filtering. Statistics derived same as for (c). (d) Recovered number of UMIs (x axis) per spatial barcode for each replicate library. (e) HDST specificity to the tissue section. Total UMIs for protein-coding genes per spatial barcode for all three MOB HDST replicates. Number of UMIs (log2(raw UMI counts), color bar) mapped at each position on the HDST array. Error bars: s.e.m. in all respective panels.
(a) Correlation of average protein-coding gene expression profiles for each HDST replicate (n=3) and for bulk RNA-Seq. Histogram of UMIs for each sample. Denoted are the Spearman correlation coefficients between groups. Red line represents fitted density distributions. (b) Numbers of protein-coding genes detected in each of three HDST replicates and bulk RNA-Seq and their intersections.
(a) Differentially expressed genes between morphological layers in HDST (Replicate1). Shown is the scaled expression level (color bar, genes expression is scaled between 0 and 1 by subtracting the minimum across layers and dividing by the maximum) of genes (columns) that are recovered as differentially expressed between layers (rows) based on lightly binned and smoothed data. (b) Labeling of morphological layers. HDST H&E image of a main olfactory bulb (left; scale bar 300 µm) and matching HDST (x,y) barcodes annotated into 9 different morphological areas (right, color code). (c-d) Shown is the spatial organization of barcodes, colored by summed gene expression for all genes differentially expressed in a specific morphological layer (label). (c) Genes up regulated in specific layers. (d) Genes down regulated in specific layers.
Supplementary Figure 5 Spatial expression patterns from HDST are validated in the Allen Brain Atlas.
(a) Layer specific signatures agree between HDST and the Allen Brain Atlas (ABA) for the major layers: granular (GL), glomerular (GCL) and mitral layers (M/T). -log10(P value) of Fisher’s test (color scale) for the enrichment of genes associated with each layer in lightly binned HDST (columns) and in the Allen Brain Atlas (rows). (b) Consistency in gene patterns. Shown are illustrative examples of summed gene expression (columns) whose spatial expression is found only as HDST specific, matches between HDST and the Allen Brain Atlas data or is found only in the Allen Brain Atlas (rows). Top illustration denotes layer-specific annotations (color code) in HDST where the grey color annotates all barcode positions not present in that layer. (c) Spatial expression of individual genes plotted using HDST data (black). Allen Brain Atlas ISH images shown side by side to each of the HDST expressed genes.
(a) H&E image of a main olfactory bulb (left), matching ST barcodes annotated into 9 different morphological areas (right, color code). (b) Nuclei segmentation and light binning enhance assignment specificity. Frequency distribution of posterior probabilities of spatial barcode assignments to two of the selected cell types (OBINH2 and OBNBL1) in data following nuclear segmentation (“sn-like”, blue), light “sc-like” (5X, cyan) and large (38X, yellow) “ST-like” binning, and HDST (1X; tan). Only posterior probabilities passing the cutoff used in cell type assignment (Methods) were plotted. (c) Spatial assignment of OBNBL1 cells. Spatial plots of posterior probabilities with assigned (color bar) or not assigned (light grey; only in 38X) barcode positions to the OBNBL1 cell type in all bin sizes (HDST, sn-like, 5X and 38X). (d) Cell type assignment frequency. Percentage of barcode positions with a single cell type assignment out of the total number of barcode positions (y axis) in each of the binning approaches and HDST (x axis). (e) Differentially expressed genes between all 63 cell types used in the prediction task on 5X “sc-like” bins. Shown is the scaled expression level (color bar) of most significant two genes (columns; bottom annotation) per cell type (columns; top annotation). (f) Number of different RNA species detected as specific in the HDST nuclear segments. (g) Numbers of genes detected in each of the HDST nuclear segments, outside HDST nuclear segments, single nuclei data and single cell data created in suspension and their intersections.
Supplementary Figure 7 Differentially expressed genes and layer-specific enrichment in the HDST profiled breast cancer resection.
(a,b) Differentially expressed genes between morphological layers with a single annotation region associated with a lightly binned (x,y) position. Scaled expression (color bar, genes expression is scaled between 0 and 1 by subtracting the minimum across layers and dividing by the maximum) of the top 10 genes (columns; ranked according to t-test statistics) that are differentially expressed in each of four major regions (b, rows) based on lightly binned and smoothed data, or all 13 annotation regions (c, rows). (c) Enrichment of segmented spatial barcodes with assigned cell types (columns) to any of the morphological layers (rows) as annotated in (b). -log10(P value) (Fisher’s exact test, Bonferroni adjusted, P<0.01) represents the color bar. Grey tiles represent non-significant values.
Supplementary Figs. 1–7.