Single cell and subcellular spatial transcriptomics in tissue samples

Abstract

Microarrays and methods of using the microarrays in spatially resolving gene expression are provided. The microarrays and methods provide high cellular resolution and enable the acquisition and sequencing of a wide array of RNA types, including small RNAs (sRNA) and other non-mRNA entities.

Description

PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent CenterSINGLE CELL AND SUBCELLULAR SPATIAL TRANSCRIPTOM ICS IN TISSUE SAMPLESGOVERNMENTAL RIGHTS

[0001] This invention was made with government support under Award No. 1945854 awarded by the National Science Foundation. The government has certain rights in the invention.CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority from Provisional Application number 63 / 635,248, filed April 17, 2024, the entire contents of which are hereby incorporated by reference.FIELD OF THE INVENTION

[0003] The present disclosure provides oligonucleotide microarrays, systems, methods, and kits for spatially resolving gene expression in a tissue sample.BACKGROUND OF THE INVENTION

[0004] The characterization of RNA abundance patterns in specialized tissue and cellular domains in multicellular organisms is critical to understand gene networks and biological systems. This is the focus of recent work to develop spatially resolved gene expression profiling methods that allow for high throughput, in situ transcript profiling. However, current implementations have limitations that prevent their broad adoption, including (1 ) low cellular resolution, yielding inadequate single-cell data, (2) poor capture of transcripts, (3) high cost in part due to prevalent commercial methods, and (4) high complexity. Single-cell RNA-seq (scRNA-seq) sequencing approaches provide high depth and quality data, but lack spatial context and definition.

[0005] Accordingly, there is a need for cost effective and simple methods of spatially resolving gene expression that provide high cellular resolution and that enable the66360232.1PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center acquisition and sequencing of a wide array of RNA types, including small RNAs (sRNA) and other non-mRNA entities.SUMMARY OF THE INVENTION

[0006] One aspect of the present disclosure encompasses a method of spatially mapping gene expression in a tissue sample, wherein the method achieves single-cell resolution. The method comprises the steps of (a) providing or having provided a high- density, high-resolution nucleic acid microarray; (b) contacting cells of a tissue sample with the surface of the nucleic acid microarray; (c) diffusing the free spatially tagged oligonucleotides into cells in contact with the oligonucleotides on the nucleic acid microarray to thereby generate tagged cells; (d) dissociating the tissue sample to form a plurality of cells; (e) performing RNA sequencing in the dissociated tissue sample to generate an expression profile for each tagged cell, wherein the RNA sequencing further sequences the spatially tagged oligonucleotide in the tagged cell; and (f) spatially mapping the expression profile for each tagged cell across the tissue sample based on the identity of the nucleic acid tag sequenced in (e) to thereby generate a gene expression map for the tissue sample. The microarray comprises a solid support comprising a flat surface, the flat surface comprising a two-dimensional (2D) grid of a plurality of polygons, wherein each polygon is assigned a specific coordinate in the grid; and a plurality of spatially tagged oligonucleotides releasably affixed to the flat surface within each polygon, wherein all spatially tagged oligonucleotides comprise a spatial barcode that associates with a coordinate of the polygon. In some embodiments, the tissue sample is an animal or plant tissue sample. In some embodiments, the tissue sample is a whole Wolffia sp. plant.

[0007] The plurality of spatially tagged oligonucleotides can be releasably affixed to the flat surface within each polygon via a cleavable moiety and diffusing the free spatially tagged oligonucleotides into cells can comprise cleaving the cleavable moiety in each spatially tagged oligonucleotide to form free spatially tagged oligonucleotides. The cleavable moiety can be a chemically- or photo-labile cleavable moiety.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0008] In some embodiments, RNA sequencing is single cell RNA sequencing (scRNA- seq). Further, each polygon of the microarray comprises a surface area equal to or smaller than a surface area of a single cell in a tissue sample. In some embodiments, each polygon comprises a surface area of about 10 pm2or smaller. In some embodiments, the polygons are delineated by spaces not exceeding about 0.8 micron in width. Accordingly, the microarray can comprise a grid of 10 pm2polygons, wherein the polygons are distributed such that the center-to-center distance between adjacent polygons is no more than about 10.8 pm. In some embodiments, the polygons is a square.

[0009] The microarray can comprise a grid of polygons, wherein each polygons comprises a surface area of about 10 pm2 or smaller. In some embodiments, the microarray comprises a grid of 10 pm2 polygons, wherein the polygons are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm. In some embodiments, the microarray comprises a grid of about 2,075,600 10 pm2 polygons, wherein the polygons are distributed such that the center-to-center distance between adjacent polygons is no more than about 10.8 pm.

[0010] All spatially tagged oligonucleotides in each polygon can comprise the same spatial barcode. In some embodiments, all spatially tagged oligonucleotides within a polygon further comprise a unique molecular identifier (UM I). In some embodiments, each spatially tagged oligonucleotide consists essentially of a spatial barcode, a UM I, and a cleavable moiety. Further, the spatially tagged oligonucleotides can be RNA oligonucleotides. In some embodiments, each spatially tagged oligonucleotide is an RNA oligonucleotide, wherein each RNA oligonucleotide consists essentially of: (a) a spatial barcode about 12 nucleotides (nt) in length; (b) a UMI about 6-8 nt in length, and (c) a cleavable moiety. Each spatially tagged oligonucleotide can be an RNA oligonucleotide comprising about 26 nt, in the range captured by sRNA library protocols, but not a biologically relevant length.

[0011] The oligonucleotides of the microarray can be synthesized using micromirrors for maskless photolithography (maskless array synthesizer (MAS) arrays). In somePCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center embodiments, the microarray further comprises positional markers. In some embodiments, only one cell in the tissue sample contacts each polygon. In some embodiments, each cell contacted by the nucleic acid microarray comprises at least one spatially tagged oligonucleotide. In some embodiments, step (c) comprises diffusing the spatially tagged oligonucleotides into the tissue sample using electrophoresis.

[0012] The method of any one of the preceding claims, wherein the RNA sequencing in (e) comprises sequencing coding and noncoding RNAs in each cell. The noncoding RNAs can be selected from small RNAs (sRNA), circular RNAs (circRNAs), cleaved or uncapped mRNAs, and non-polyadenylated RNAs, long-non-coding RNAs cleaved or uncapped mRNAs, non-polyadenylated RNAs, mixed host / pathogen (eukaryote / prokaryote), and any combination thereof.

[0013] In some embodiments, the tissue sample is expanded before contacting cells of the tissue sample with the nucleic acid microarray. The tissue sample can be expanded using methods used in expansion microscopy. In some embodiments, the tissue sample is expanded via injection of an expandable polyelectrolyte gel matrix into the tissue sample. In some embodiments, nucleic acids and proteins in the tissue sample are cross-linked to the expandable polyelectrolyte gel matrix before expansion. In some embodiments, the tissue sample is linearly expanded 4 to 5-fold.

[0014] The tissue sample can be a plurality of tissue sections sectioned from the tissue sample. In some embodiments, the tissue sample contacting the microarray has a depth of 1 to 3 cells.

[0015] The tissue sample can be fixed and sectioned into a plurality of tissue sections prior to contacting the array. In some embodiments, the tissue or tissue section is stained prior to (pre-infiltrated) or after contacting the array. The tissue sample can be fixed, cryo-sectioned into a plurality of tissue sections prior to contacting the array, then stained.

[0016] In some embodiments, the method can further comprise imaging the tissue sample before step (c). The method can comprise imaging the tissue sample on the array. In some embodiments, the method comprises fixing the tissue sample, cryo-PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center sectioning into a plurality of tissue sections, staining, transferring onto the array, and imaging. In some embodiments, the imaging correlates each cell on the array with a coordinate polygon. Spatially mapping RNA sequence data in the tissue sample can comprise matching each cell in the tissue sample to the RNA sequence data based on the coordinate of the cell and the sequence of the oligonucleotide spatial tag in the sequenced cell. In some embodiments, spatially mapping RNA sequence data in the tissue sample comprises matching each cell in the tissue sample to the RNA sequence data based on the coordinate of the cell and the sequence of the oligonucleotide spatial tag in the sequenced cell by: (a) converting the stained tissue sample into a map of cells that are overlaid by the array polygons, the positional markers, or both; and (b) correlating the RNA sequence data to the microscopy image by positions relative to the array polygons, the positional markers, or both.

[0017] In some embodiments, the method further comprises integrating the method with array-based barcoding of tissue sections. The tissue sample can be a plurality of tissue sections sectioned from the tissue sample and spatially mapped RNA sequence data in each section of the plurality of tissue sections is assembled into a Z-stack of RNA sequence data to thereby generate a reconstructed 3-dimentional (3D) RNA sequence map and optionally each of the plurality of tissue sections is imaged and the imaging data is combined with each gene expression map for each section to form a three- dimensional imaging and gene expression map for the tissue.

[0018] The method can further comprise building a whole-tissue 3D volumetric model of the tissue sample at a single-cell level before step (b) or of a different sample of the tissue of interest, wherein the 3D model serves as a ground truth and specific reference volume for each corresponding tissue array dataset for subsequent integration. In some embodiments, building a whole-tissue 3D volumetric model comprises imaging via X- Ray microscopy (XRM). In some embodiments, building a whole-tissue 3D model at a single-cell and sub-cellular level comprises combining imaging via XRM with volumetric electron microscopy (vEM) to capture sub-cellular resolution in the 3D model of the tissue. In some embodiments, vEM is serial block-face scanning electron microscopyPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center(SBF-SEM). In some embodiments, the method, further comprises integrating the spatially mapped expression profile for each tagged cell in the 3D RNA sequence map across the tissue with the whole-tissue 3D model.

[0019] In some embodiments, the method further comprises the steps of: (a) building a whole-tissue 3D model at a single-cell and sub-cellular level of the tissue sample or of a different sample of the tissue by combining imaging via XRM with volumetric electron microscopy (vEM); (b) fixing the tissue sample before step (b) and before or after building the whole-tissue 3D model of the tissue sample, cryo-section ing the tissue sample into a plurality of tissue sections, staining the tissue sections, and transferring the tissue sections onto the array; (c) imaging the tissue sections on the microarray before step (c); (d) spatially mapping the expression profile for each tagged cell across each tissue section by matching each cell in the tissue section to the sequence data based on the coordinate of the cell and the sequence of the oligonucleotide spatial tag in the sequenced cell; (e) assembling the sequence data of the plurality of tissue sections into a Z-stack of RNA sequence data to thereby generate a reconstructed 3D RNA sequence map; and (f) integrating the spatially mapped expression profile for each tagged cell in the 3D RNA sequence map across the tissue with the whole-tissue 3D model. Spatially mapping the expression profile for each tagged cell across each tissue section comprises: (a) converting the stained tissue sections into a map of cells that are overlaid by the array spots, the positional markers or both; and (b) correlating the RNA- seq data to the microscopy image by positions relative to the positional markers.

[0020] Another aspect of the instant disclosure encompasses a method of spatially mapping gene expression across three dimensions in a tissue sample. The method comprises the steps of: (a) fixing and cryo-sectioning the tissue sample into a plurality of tissue sections that are 1 -3 cells deep; (b) spatially mapping RNA sequence data in each section according to the method of any one of claims 1 to 43 to generate a set of gene expression maps for each section; and (c) assembling the gene expression maps for each section into a Z-stack of RNA sequence data to thereby generate a reconstructed 3D RNA sequence map. In some embodiments, the method furtherPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center comprises imaging each tissue section and combining the imaging data with each gene expression map for each section to form a three-dimensional imaging and gene expression map for the tissue. In some embodiments, the method further comprises building a whole-tissue 3D volumetric model at a single-cell level of the tissue sample before step (b) or of a different sample of the tissue of interest according to the method described herein above, wherein the 3D model serves as a ground truth and specific reference volume for each corresponding tissue array dataset for subsequent integration. The tissue can be an animal or plant tissue. In some embodiments, the tissue is a plant tissue. In some embodiments, the tissue sample is a whole Wolffia sp. plant.

[0021] An additional aspect of the instant disclosure encompasses a high-density, high- resolution nucleic acid microarray for positionally mapping gene expression in a tissue sample. The microarray comprises: (a) a solid support comprising a flat surface, the flat surface comprising a two-dimensional (2D) grid of a plurality of polygons, wherein each polygon is assigned a specific coordinate in the grid; and (b) a plurality of spatially tagged oligonucleotides affixed to the flat surface within each polygon, wherein all spatially tagged oligonucleotides comprise a spatial barcode that associates with a coordinate of the polygon. The plurality of spatially tagged oligonucleotides can be affixed to the flat surface within each polygon via a cleavable moiety and diffusing the free spatially tagged oligonucleotides into cells can comprise cleaving the cleavable moiety in each spatially tagged oligonucleotide to form free spatially tagged oligonucleotides. In some embodiments, the microarray further comprises positional markers. In some embodiments, each polygon comprises a surface area equal to or smaller than a surface area of a single cell in a tissue. Each polygon can comprise a surface area of about 10 pm2 or smaller. Further, the polygons can be delineated by spaces not exceeding about 0.8 micron in width. In some embodiments, the microarray comprises a grid of 10 pm2 polygons, wherein the polygons are distributed such that the center-to-center distance between adjacent polygons is no more than about 10.8 pm. In some embodiments, the polygons is a square shaped.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0022] In some embodiments, the microarray comprises a grid of polygons, wherein each square comprises a surface area of about 10 pm2 or smaller. In some embodiments, the microarray comprises a grid of 10 pm2 squares, wherein the squares are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm. In some embodiments, the microarray comprises a grid of about 2,075,600 10 pm2 square polygons comprising, wherein the squares are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm. All spatially tagged oligonucleotides in each polygon can comprise the same spatial barcode. In some embodiments, each spatially tagged oligonucleotide consists essentially of a spatial barcode, a unique molecular ID, and a cleavable moiety.

[0023] In some embodiments, the spatially tagged oligonucleotides are RNA oligonucleotides, DNA oligonucleotides, or RNA / DNA oligonucleotides. In some embodiments, spatially tagged oligonucleotides are RNA oligonucleotides. Each spatially tagged oligonucleotide can be an RNA oligonucleotide, wherein each RNA oligonucleotide consists essentially of: (a) a spatial barcode about 12 nucleotides (nt) in length; (b) a UMI about 6-8 nt in length, and (c) a cleavable moiety.

[0024] In some embodiments, the cleavable moiety is a chemically- or photo-labile cleavable moiety. In some embodiments, the oligonucleotides of the microarray are synthesized using micromirrors for maskless photolithography (maskless array synthesizer (MAS) arrays). Yet another aspect of the instant disclosure encompasses a kit for spatially mapping gene expression in a tissue sample. The kit comprises one or more high-resolution high-density nucleic acid microarray described herein above; and optionally, tissue preparation reagents, microscopy reagents, cell growth media, selection media, or any combination thereof.BRIEF DESCRIPTION OF THE FIGURES

[0025] The following drawings form part of the present specification and are included to further demonstrate certain embodiments of the present disclosure. Certain embodiments can be better understood by reference to one or more of these drawingsPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center in combination with the detailed description of specific embodiments presented herein. The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawing(s) will be provided by the Office upon request and payment of the necessary fee.

[0026] FIG. 1. MAS arrays with 10 pm spots. A schematic (upper portion) and nine “spots” from an actual MAS array with 1 pm spacer; note near complete surface coverage of oligos (light).

[0027] FIG. 2 is a schematic of an embodiment of methods of the instant disclosure, (i) Barcodes from the array are released and migrated up into cells of the fixed tissue section. Imaging performed at this stage, (ii) Cells from the section, now containing positional barcodes, are separated and sorted by the Namocell for scRNA-seq in individual tubes, (iii) mRNA and sRNA libraries (scRNA-seq) made from single cells, capturing a small number of spatial RNA barcodes, (iv) After sequencing, tissue is virtually reconstructed from single cell data; a small number of cells may share barcodes (i.e. , cell #5); others may be fragmented and not captured (light grey).

[0028] FIG. 3 is a diagram of an embodiment of a workflow of a method of the instant disclosure.DETAILED DESCRIPTION

[0029] The present disclosure encompasses methods, oligonucleotide arrays, systems, and kits, for spatially mapping gene expression in a tissue sample. The inventors devised methods, oligonucleotide arrays, and processes capable of positionally resolving gene expression in a tissue sample at the single cell or even sub-cellular level. The microarrays have a high-density and high-resolution and can be manufactured using cost-effective methods. The oligonucleotides and configuration of the oligonucleotides in microarrays of the instant disclosure facilitates a broader and more inclusive capture strategy, enabling the acquisition and sequencing of a wide array of RNA types, including small RNAs (sRNA) and other non-mRNA entities. Methods of thePCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center instant disclosure are beneficial for 3D transcriptom ics in plants that require an enzymatic process to remove the cell wall to isolate cellular contents.I. High-resolution high-density nucleic acid microarray

[0030] One embodiment of the present disclosure encompasses nucleic acid microarray (microarray for short) for positionally mapping gene expression in a tissue sample. The microarray comprises a solid support comprising a flat surface. The flat surface comprises a grid composed of a multitude of polygons and a plurality of spatially tagged oligonucleotides affixed to the flat surface within each polygon. The microarray can be manufactured at a low cost and can be adapted for use in scRNA-seq library construction.

[0031] Importantly, the microarray is characterized by its high-density and high- resolution that enable precise positional resolution of gene expression at a single cell or sub-cellular level. As used herein, the term "high-density" refers to the compact arrangement of numerous distinct features (polygons) within a given area on the microarray surface. A high-density microarray design implies that each polygon, despite its small size, is part of a closely packed configuration, minimizing unused surface area and ensuring that a maximal amount of data can be gathered from the tissue sample. This dense arrangement allows for the detailed mapping of gene expression at a granular level, which is not achievable with lower density arrays. "High- resolution" in the context of microarrays of the instant disclosure refers to the microarray's ability to distinguish and accurately map gene expressions at a very detailed spatial level, down to cellular or even subcellular scales. This resolution is directly influenced by the small size of the polygons on the microarray's surface and the tight packing of these polygons. The minimal spacing between polygons ensures that there are no significant gaps that could result in missing information when analyzing tissues. By minimizing the distance between adjacent polygons, the microarray can more effectively capture the spatial distribution of gene expressions across the tissuePCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center sample, providing insights into the heterogeneity and specific gene expression patterns in a tissue.

[0032] Further, oligonucleotides of the microarray can be releasably affixed to the surface of the microarray in a manner that allows for their subsequent release and diffusion into the cells of a tissue sample. This design contrasts sharply with the limitations inherent in currently available spatial transcriptom ic methods, which predominantly rely on the capture of transcripts directly on the microarray surface. Such surface capture techniques often lead to suboptimal sequence data due to inefficient transcript capture. This inefficiency not only hampers the quality of data obtained but also significantly limits the scope of transcriptom ic analysis by restricting it primarily to mRNA. The configuration of the oligonucleotides in microarrays of the instant disclosure facilitates a broader and more inclusive capture strategy, enabling the acquisition and sequencing of a wide array of RNA types, including small RNAs (sRNA) and other non-mRNA entities. This comprehensive approach enhances the depth and diversity of transcriptom ic profiling, offering a more complete and nuanced understanding of the gene expression landscape within tissue samples.(a) Solid support and grid configuration

[0033] The microarray comprises a solid support comprising a plurality of spatially tagged oligonucleotides affixed to a surface of the microarray. The microarray comprises a solid support comprising a flat surface comprising a two-dimensional (2D) grid composed of a multitude of polygons within which the oligonucleotides are affixed. Each polygon can be assigned a specific coordinate in the grid, facilitating precise spatial localization of gene expression data. It is important to note that the microarray comprises a flat surface with minimal spacing between polygons, thereby providing a more uniform and comprehensive coverage of the tissue surface when compared, e.g., to other technologies that utilize oligonucleotides affixed to beads or spotting methods. The inherent design of microarrays of the instant disclosure eliminates the gaps between probes that are unavoidable with bead- or spot- based methods, ensuring thatPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center no area of the tissue sample is left unanalyzed and enabling more precise spatial resolution of gene expression across the entire sample.

[0034] As explained above, microarrays of the instant disclosure are characterized by their high-density and high-resolution. Each polygon comprises a surface area equal to or smaller than a surface area of a single cell in a tissue, thereby allowing for resolution of positional gene expression at the cellular or even subcellular level in some tissues. Further, the polygons can be tightly packed with minimal spacing between polygons. These features allow for a broad application in studying cellular heterogeneity, structure, and function, contributing to a deeper understanding of the complex interplay between gene expression and cellular physiology in diverse biological systems.

[0035] Cell sizes can and will vary depending on the organism, the tissue, and the stage of development or growth of the cells, among other variables. Accordingly, the size of each polygon in a microarray can and will vary according to the intended tissue or organism to be studied. The variability in cell sizes across different organisms, tissues, and even within a single organism illustrates the significance of this feature. For instance, typical human cells can range in diameter from about 10 to 30 micrometers (pm), with neurons being among the largest, extending up to several hundred micrometers in length. Conversely, cells in plants can vary more broadly in size, from the relatively small cells of algae, which can be as little as a few micrometers across, to the large cells found in some plant tissues, such as the parenchyma of certain aquatic plants, which can reach sizes of up to several centimeters.

[0036] By adjusting the polygon size to be on par with or smaller than the size of the cells of interest, the microarray can generate high-resolution data capable of distinguishing between cells within a tissue and different cell types within a heterogeneous tissue sample. This capability is particularly valuable when considering the vast differences in cell sizes and shapes found across tissues. For example, epithelial cells, which are typically polygonal and closely packed, may require a different polygon size for optimal resolution compared to elongated and sparsely distributed neuronal cells.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0037] In some embodiments, a microarray of the instant disclosure comprises polygons comprising a surface area of about 10, 11 , 12, or 13 pm2or smaller. In some embodiments, a microarray of the instant disclosure comprises polygons comprising a surface area of about 10 pm2or smaller.

[0038] In some embodiments, the polygons are delineated by spaces as minimal as microarray-manufacturing technology. See Section l(b) for methods of manufacturing such high density and high resolution microarrays. In some embodiments, the polygons are delineated by spaces not exceeding about 0.8 micron in width. Accordingly, when the microarray comprises a grid of 10 pm2polygons, and the polygons are delineated by spaces not exceeding about 0.8 micron in width, the polygons are distributed such that the center-to-center distance between adjacent polygons is no more than about 10.8 pm.

[0039] The number of polygons in a grid on a surface of a microarray of the instant disclosure can and will vary depending on the dimensions of the tissue sample under investigation, the size of the polygons of the grid, and limitations imposed by the microarray's manufacturing process and the capabilities of data-reading devices. The methods for preparing the 2D grid systems can be as detailed herein below. For instance, the total number of polygons within a grid can span a broad range, with potential quantities between about 1 million and about 3 million polygons. In some embodiments, counts can fall within a narrower range of about 1 .5 million to about 2.5 million polygons, or can number about 2,075,600 polygons, depending on the specific requirements of the microarray application and the aforementioned factors. In some embodiments, the number of polygons in a grid can range from about 1 million to about 3 million polygons, from about 1 .5 million to about 2.5 million polygons, or can be about 2,075,600 polygons.

[0040] A 2D grid of the microarray of the instant disclosure can be any of several known 2D grid systems, each offering advantages depending on the analysis requirements and the nature of the data. The selection of a specific grid system can optimize the spatial resolution and analytical efficiency of the microarray, taking into account the tissue ofPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center interest along with the sizes and shapes of cells within that tissue. Non-limiting examples of some commonly used 2D grid systems include a cartesian grid system, hexagonal grid system, triangular grid system, polar grid system, isometric grid system, logarithmic grid system, quadtrees, and Voronoi diagrams. When selecting a grid system for a microarray, it's important to consider the tissue of interest and the cell size(s) and shape(s) of cells in the tissue.

[0041] A cartesian grid system utilizes two perpendicular axes (X and Y) to define square or rectangular grid cells. It is straightforward and widely used, facilitating easy mapping and indexing of spatial data. This system is particularly useful for microarrays analyzing tissues with relatively uniform cell sizes and shapes, allowing for straightforward alignment and comparison of gene expression data across different regions.

[0042] Hexagonal grids can be made of tessellated hexagons that cover a plane without gaps or overlaps. This grid system can offer closer packing of cells, potentially reducing the distance between points within the grid compared to other grid systems. Hexagonal grids can be advantageous for microarrays by providing higher spatial resolution and more efficient coverage, especially useful in tissues with varying cell densities.

[0043] Triangular grid systems comprising equilateral triangles allow for a high degree of connectivity between grid points, which can be beneficial for analyzing complex tissue structures. The triangular grid can offer advantages in microarray designs that require detailed mapping of highly heterogeneous tissues, facilitating the study of intricate cellular interactions.

[0044] Polar grid systems can be based on circles divided into sectors. The polar grid system is centered around a central point and is useful for data that naturally aligns in circular patterns. In microarray applications, polar grids can be suitable for analyzing circular or radially symmetric tissue samples, allowing for an intuitive representation of gene expression data in such contexts.

[0045] An isometric grid system creates a 3D effect on a 2D plane using equilateral triangles, facilitating the visualization of data with pseudo-three-dimensionalPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center perspectives. Isometric grids can be useful in microarrays for providing a more nuanced view of tissue structures that have significant depth or layering.

[0046] A logarithmic grid system employs logarithmic scaling on its axes, making it suitable for representing data that spans several orders of magnitude. In microarray contexts, logarithmic grids can be particularly useful for analyzing gene expression levels that vary widely across different regions of a tissue sample.

[0047] Quadtrees divide the space into four quadrants recursively based on the presence of data, allowing for efficient data representation and querying. This adaptive grid system can be highly beneficial for microarrays analyzing tissues with uneven distribution of cells, as it allows focusing resolution where it is most needed.

[0048] Voronoi diagrams partition the plane into regions based on distances to a specific set of points, creating cells that are closer to each respective point than to any other. Voronoi diagrams can be exceptionally useful for microarrays in modeling cellular territories or domains within tissues, providing insights into spatial relationships and dependencies among cells.

[0049] In some embodiments, the microarray comprises a solid support comprising a flat surface, the flat surface comprising a 2D cartesian grid of a plurality of polygons, wherein each polygon is assigned a specific coordinate in the grid. In some embodiments, a microarray of the instant disclosure comprises polygons comprising a surface area of about 10, 11 , 12, or 13 pm2or smaller. In some embodiments, a microarray of the instant disclosure comprises polygons comprising a surface area of about 10 pm2or smaller. In some embodiments, the polygons are delineated by spaces not exceeding about 0.8 micron in width. Accordingly, when the microarray comprises a grid of 10 pm2polygons, and the polygons are delineated by spaces not exceeding about 0.8 micron in width, the polygons are distributed such that the center-to-center distance between adjacent polygons is no more than about 10.8 pm. In some embodiments, the microarray comprises a grid of about 2,075,600 10 pm2polygons, wherein the squares are distributed such that the center-to-center distance between adjacent polygons is no more than about 10.8 pm.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0050] In some embodiments, the polygons are square shaped. In some embodiments, when the polygons are square shaped, the microarray comprises a solid support comprising a flat surface, the flat surface comprising a 2D cartesian grid of a plurality of squares, wherein each square is assigned a specific coordinate in the grid. In some embodiments, a microarray of the instant disclosure comprises squares comprising a surface area of about 10, 11 , 12, or 13 pm2or smaller. In some embodiments, a microarray of the instant disclosure comprises squares comprising a surface area of about 10 pm2or smaller. In some embodiments, the squares are delineated by spaces not exceeding about 0.8 micron in width. Accordingly, when the microarray comprises a grid of 10 pm2squares, and the squares are delineated by spaces not exceeding about 0.8 micron in width, the squares are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm. In some embodiments, the microarray comprises a grid of about 2,075,600 10 pm2squares, wherein the squares are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm. In some embodiments, the microarray can be as depicted in FIG. 1

[0051] In some embodiments, the microarray further comprises positional markers that facilitate the precise localization of gene expression data. Positional markers can facilitate precise localization of gene expression data by precisely localizing each single cell, mapping the cells back to the array and to the tissue section that is imaged after it is applied to the array, relative to the positional markers on the array for positional alignment in both the image and sequencing data. Spatially mapping gene expression in a tissue sample can be as described in Section II herein below.

[0052] In some embodiments, a microarray of the instant disclosure comprises a solid support comprising a flat surface, the flat surface comprising a 2D grid of a plurality of polygons and positional markers, wherein each polygon is assigned a specific coordinate in the grid. In some embodiments, each polygon comprises a surface area equal to or smaller than a surface area of a single cell in a tissue. In some embodiments, each polygon comprises a surface area of about 10 pm2or smaller. InPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center some embodiments, the polygons are delineated by spaces not exceeding about 0.8 micron in width. In some embodiments, the microarray comprises a grid of 10 pm2polygons, wherein the polygons are distributed such that the center-to-center distance between adjacent polygons is no more than about 10.8 pm.

[0053] In some embodiments, the polygons are square shaped. When the polygons are square shaped, each square can comprise a surface area of about 10 pm2or smaller. In some embodiments, the microarray comprises a grid of 10 pm2squares, wherein the squares are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm. In some embodiments, the microarray comprises a grid of about 2,075,600 10 pm2square polygons comprising, wherein the squares are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm.(b) Photolithography

[0054] As explained above, microarrays of the instant disclosure are characterized by their high-density and high-resolution. The small size of the polygons ensures that the microarray can cover a wide range of gene expressions with high precision, while the tight packing (high-density) of these polygons allows for an extensive dataset to be collected from a minimal sample area. This combination is important for applications requiring detailed mapping of gene expressions in tissue, such as in the study of complex biological processes, disease mechanisms, and the development of targeted therapies.

[0055] Accordingly, microarrays of the instant disclosure can be fabricated using any method capable of fabricating microarrays comprising the high density and high resolution of microarrays of the instant disclosure. However, it is important to note that the polygon densities necessary for the functionality of this invention exceed the capabilities of conventional microarray manufacturing techniques. Such densities represent a recent advancement in the field, thus distinguishing the present disclosure.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0056] The evolution of microarray technology from its inception in the late 1980s to the present day has been marked by continuous advancements aimed at increasing both the diversity and density of features that can be arrayed on a support surface. The manufacturing of microarrays capable of hosting large, complex sets of heteropolymers such as nucleic acids and peptides with high feature density is critical for their utility in a wide range of applications in medicine, biotechnology, and materials science. The absolute number of different heteropolymers represented on a microarray (feature diversity), the number of different heteropolymers per unit area (feature density), and the quality of these heteropolymers are key technical characteristics that correlate directly with the microarray's value.

[0057] The method chosen for manufacturing a microarray significantly impacts not only its intrinsic technical characteristics but also its economic value. This includes factors such as customizability of content, manufacturing flexibility, cost structure, and turnaround time from content specification to manufacture completion. Despite significant commercial success over the past four decades with various manufacturing approaches, the quest for further improvements in microarray performance, particularly in terms of feature density and manufacturing efficiency, remains a high priority.

[0058] Conventional manufacturing technologies, divided into spotting and in situ synthesis methods, each come with their limitations. Spotting methods, though customizable and cost-effective, generally produce larger features, significant spacing between features, and thus lower feature densities. In situ synthesis methods, including ink-jet printing, electrochemistry, and photolithography, offer finer control over feature size and density but face fundamental physical and technical constraints that challenge further improvements. Notably, the resolution and fidelity of microarrays produced by these methods are limited by issues such as cross-talk during synthesis and sequence errors, which are exacerbated as feature density increases.

[0059] Maskless photolithography (MAS arrays) emerges as a superior method for fabricating the high-density microarrays of the instant disclosure. Unlike traditional photolithography, which relies on static masks to pattern light, masklessPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center photolithography uses a programmable digital micromirror device (DMD) to dynamically direct UV light for polymer synthesis on the microarray surface. This approach eliminates the need for costly and inflexible photomasks, significantly reducing manufacturing time and expenses. Moreover, maskless photolithography offers unparalleled precision in controlling feature size and spacing, enabling the production of microarrays with the small polygons and narrow spaces between them as required for this invention. These arrays are flexible, as printing by maskless photolithography means that array content is designed on a computer and each synthesis run can be different, with a cost of tens of dollars per array. They can be synthesized with diverse chemistries including chemically labile nucleotides to allow high efficiency cleavage from the surface, without disturbing their localization.

[0060] The high-density, small polygon sizes, and minimal spaces between polygons achievable through maskless photolithography are not only technically feasible but also critically important for the intended application of spatially resolving gene expression at cellular or subcellular levels. The precise control over polymer synthesis afforded by this technology allows for the creation of microarrays with extremely high feature densities — essential for the detailed analysis and understanding of complex biological samples. This technology aligns with the requirements for manufacturing the advanced microarrays envisioned in this disclosure, providing the necessary infrastructure to support the novel capabilities introduced by these high-density arrays.

[0061] Accordingly, in some embodiments, microarrays of the instant disclosure are manufactured using maskless photolithography. In some embodiments, microarrays of the instant disclosure are manufactured using maskless photolithography capable of generating polygons delineated by spaces not exceeding about 0.5 micron to about 1 .0 micron in width. In some embodiments, microarrays of the instant disclosure are manufactured using maskless photolithography capable of generating polygons delineated by spaces not exceeding about 0.5 micron, about 0.6 micron, about 0.7 micron, about 0.8 micron, about 0.9 micron, about 0.10 micron, or about 1 .0 micron, in width. In some embodiments, microarrays of the instant disclosure are manufacturedPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center using maskless photolithography capable of generating polygons delineated by spaces not exceeding about 0.8 micron in width. In some embodiments, microarrays of the instant disclosure are manufactured using maskless photolithography capable of generating polygons comprising a surface area of about 15 pm2to about 5 pm2or smaller. In some embodiments, microarrays of the instant disclosure are manufactured using maskless photolithography capable of generating polygons comprising a surface area of about 15 pm2, about 14 pm2, about 13 pm2, about 12 pm2, about 11 pm2, about 10 pm2, about 9 pm2, about 8 pm2, about 7 pm2, about 6 pm2, or about 5 pm2, or smaller. In some embodiments, microarrays of the instant disclosure are manufactured using maskless photolithography capable of generating polygons comprising a surface area of about 10 pm2or smaller.

[0062] In some embodiments, microarrays of the instant disclosure are manufactured using maskless photolithography capable of generating polygons comprising a surface area of about 15 pm2, about 14 pm2, about 13 pm2, about 12 pm2, about 11 pm2, about 10 pm2, about 9 pm2, about 8 pm2, about 7 pm2, about 6 pm2, or about 5 pm2, or smaller, delineated by spaces not exceeding about 0.5 micron, about 0.6 micron, about 0.7 micron, about 0.8 micron, about 0.9 micron, about 0.10 micron, or about 1 .0 micron, in width. In some embodiments, microarrays of the instant disclosure are manufactured using maskless photolithography capable of generating polygons comprising a surface area of about 10 pm2or smaller, delineated by spaces not exceeding about 0.8 micron in width. Accordingly, when the microarray comprises a grid of about 15 pm2to about 5 pm2or smaller polygons, and the polygons are delineated by spaces not exceeding about 0.5 micron to about 1 .0 micron in width, the polygons are distributed such that the center-to-center distance between adjacent polygons is no more than about 16 pm to about 5.05 micron. In some embodiments, the microarray comprises a grid of polygons distributed such that the center-to-center distance between adjacent polygons is no more than about 10.8 pm.

[0063] In some embodiments, microarrays of the instant disclosure are manufactured using maskless photolithography comprising 1080p resolution. In some embodiments,PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center microarrays of the instant disclosure are manufactured using maskless photolithography comprising 4K resolution.(c) Oligonucleotides

[0064] The nucleic acid microarray of the instant disclosure further comprises a plurality of spatially tagged oligonucleotides releasably affixed to the flat surface within each polygon. In some embodiments, the oligonucleotides are affixed to the flat surface of the microarray using methods described in Section l(b) herein above.

[0065] Oligonucleotides of microarray of the instant disclosure can comprise any nucleic acids or their derivatives, contingent upon the microarray's specific application. This flexibility ensures that the oligonucleotides can be optimally configured for their intended analytical or diagnostic purposes. For example, in applications targeting single-cell RNA sequencing (scRNA-seq), the oligonucleotides can be RNA oligonucleotides to directly complement the RNA transcripts being analyzed, facilitating a more direct and efficient hybridization process. However, the scope of the oligonucleotides extends beyond RNA to include DNA oligonucleotides, as well as hybrids of RNA and DNA and oligonucleotides that include derivatives of RNA and DNA. In some embodiments, oligonucleotides of the instant disclosure are RNA oligonucleotides.

[0066] In some embodiments, spatially tagged oligonucleotides of the instant disclosure are releasably affixed to the flat surface of a microarray within each polygon via a cleavable moiety. These moieties facilitate the controlled release or activation of oligonucleotides upon specific stimuli. Non-limiting examples of cleavable moieties that can be used to facilitate a controlled release include chemically-labile cleavable moieties, photo-labile, cleavable moieties, enzyme-labile cleavable moieties, or any combination thereof.

[0067] In some embodiments, a cleavable moiety of the instant disclosure is a chemically-labile cleavable moiety. Non-limiting examples of chemically-labile cleavable moieties include ester linkages and disulfide bonds. Ester linkages are commonly used due to their stability under physiological conditions and their cleavage in the presence ofPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center esterases or through hydrolysis under basic conditions. For instance, acetate-protected ester linkages are cleavable in mild alkaline conditions, making them suitable for selective release strategies in microarrays. Disulfide bonds are cleavable through reduction, and can be used where changes in redox conditions can be used as a trigger for cleavage. A non-limiting example of a disulfide bond linkage is dimethyl 3,3'- dithiobispropionimidate (DTBP) which is cleavable by reducing agents like dithiothreitol (DTT) or tris(2-carboxyethyl)phosphine (TCEP), offering controlled release of oligonucleotides under reducing conditions.

[0068] In some embodiments, a cleavable moiety of the instant disclosure is a photo- labile cleavable moiety. Non-limiting examples of photo-labile cleavable moieties include ortho-nitrobenzyl groups and coumarin-based linkers. Ortho-nitrobenzyl groups offer UV light-sensitive cleavage options, allowing for spatially controlled release through targeted illumination. Non-limiting examples of ortho-nitrobenzyl groups suitable for use as a cleavable moiety include 6-Nitroveratryloxycarbonyl (NVOC) and 2- nitrobenzyl ether (NBE). Coumarin-based Linkers such as 7-diethylaminocoumarin-4- yl)methyl (DEACM) are also cleavable by UV light.

[0069] In some embodiments, a cleavable moiety of the instant disclosure is an enzyme- labile cleavable moiety. Non-limiting examples of enzyme-labile cleavable moieties include peptide linkers and glycosidic linkers. Peptide linkers such as USER enzyme (Uracil-Specific Excision Reagent) can be designed to be cleaved by specific proteases, making them suitable for environments where enzyme presence can be used as a trigger. Glycosidic linkers such as [3-glycosidase sensitive linkers are cleavable by glycosidase enzymes.

[0070] The spatially tagged oligonucleotides comprise a spatial barcode. Using methods of the instant disclosure, spatial barcodes can assign precise positional information of expression data by indexing the locations of polygons from which the sequencing data is derived. Accordingly, in some embodiments, each polygon comprises a different spatial barcode that associates with a coordinate of the polygon.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent CenterIn some embodiments, all spatially tagged oligonucleotides in each polygon comprise the same spatial barcode.

[0071] The length of spatial barcodes within the oligonucleotides can and will vary depending on the microarray's capacity for spatial resolution and the total number of distinct spatial locations (polygons) that can be uniquely identified. The relationship between the length of spatial barcodes and the number of polygons on the microarray is a function of the need to uniquely identify each polygon's spatial position on the microarray. The high-density embodiment of the microarray, with potentially millions of polygons tightly packed across its surface, necessitates a barcode system with sufficient complexity to distinguish each of these locations. The length of the barcode determines its encoding capacity — the number of unique sequences it can generate — and thus directly impacts the microarray's ability to achieve high-resolution spatial mapping. A longer barcode increases the number of unique identifiers available, which is essential for microarrays designed to cover extensive areas or to provide ultra-high resolution at the cellular or subcellular level. However, increasing barcode length also raises the complexity of synthesis and analysis. Therefore, the chosen length represents a compromise, ensuring that the barcodes can be long enough to offer the necessary encoding capacity without unduly complicating the microarray's manufacture or use.

[0072] In some embodiments, spatial barcodes of microarrays of the instant disclosure can be sufficiently long to uniquely identify or index each polygon of the microarray and, by extension, each cell in a tissue tagged by the oligonucleotide. In some embodiments, spatial barcodes can range from about 9 nucleotides (nt) in length to about 15 nt in length. In some embodiments, spatial barcodes can range from about 12 nucleotides (nt) in length, a dimension that balances complexity with practical synthesis and sequencing considerations. In some embodiments, spatial barcodes are about 12 nucleotides (nt) in length. A 12 nt barcode can theoretically encode 412(approximately 16.8 million) unique combinations, far exceeding the requirements for microarrays with densities in the range of about 1 million to about 3 million polygons. This amplePCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center encoding capacity ensures that each polygon can be associated with a unique spatial barcode, thereby eliminating potential ambiguity in spatial mapping.

[0073] In addition to a cleavable moiety and spatial barcodes, oligonucleotides of the instant disclosure can further comprise additional sequences or elements to expand its functionality and application range. Non-limiting examples of such additional sequences or elements include a unique molecular identifier (UMI), capture sequences such as target-specific probes complementary to specific RNA or DNA targets of interest, enabling the detection and quantification of specific genes or transcripts; adapters such as PCR primer binding sites and sequencing adapters; additional barcodes such as sample barcodes that can allow tracing the origin of each sequence read to its specific sample after pooled sequencing; spacers or linkers that can provide physical separation between functional regions of the oligonucleotide; regulatory elements such as promoters for applications involving in vitro transcription; internal standards that can serve as controls or references within an assay; and conditional sequences such as temperature-sensitive or pH-sensitive sequences that can be useful for controlling reaction kinetics or for environmental sensing applications within a biological sample.

[0074] In some embodiments, spatially tagged oligonucleotides further comprise a unique molecular identifier (UMI). UMIs are short nucleotide sequences incorporated into oligonucleotides used in various high-throughput sequencing applications, including single-cell RNA sequencing (scRNA-seq). UMIs are designed to be distinct for each molecule in a sample, enabling precise quantification of individual nucleic acid molecules, reducing amplification bias, and improving the accuracy of data analysis. UMIs can comprise a series of nucleotides (usually 6-12 bases long) that are randomly incorporated into oligonucleotides. The length and composition of UMIs are carefully designed to ensure a sufficiently large combination of sequences to uniquely tag each molecule in a complex sample. The randomness and uniqueness of each UMI sequence are crucial for their effectiveness in distinguishing between individual molecules, even those that are otherwise identical in sequence. UMIs enable accurate quantification and differentiation of individual nucleic acid molecules within a sample,PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center particularly in applications involving amplification steps, such as PCR, which can introduce bias. During the sequencing process, each molecule is tagged with a UM I before any amplification. After sequencing, reads with the same UMI are grouped together, assuming they originate from the same original molecule. This allows to correction for PCR amplification bias, identifying duplicate reads, and improving detection sensitivity. UMIs are utilized across various sequencing applications, with notable but non-limiting examples including single-cell RNA sequencing (scRNA-seq), targeted sequencing and variant detection, and ChlP-seq and ATAC-seq.

[0075] In some embodiments, a microarray of the instant disclosure comprises a plurality of spatially tagged RNA oligonucleotides affixed to the flat surface within each polygon via a cleavable moiety, wherein all spatially tagged oligonucleotides within a polygon comprise a unique molecular identifier (UMI) and a spatial barcode that associates with a coordinate of the polygon. In some embodiments, all spatially tagged oligonucleotides in each polygon comprise the same spatial barcode. In some embodiments, each spatially tagged oligonucleotide consists essentially of a spatial barcode, a unique molecular ID, and a cleavable moiety.

[0076] In some embodiments, spatially tagged oligonucleotides are RNA oligonucleotides. In some embodiments, each spatially tagged oligonucleotide is an RNA oligonucleotide, wherein each RNA oligonucleotide consists essentially of: a spatial barcode about 12 nucleotides (nt) in length; a UMI about 6-8 nt in length, and a cleavable moiety. In some embodiments, the cleavable moiety is a chemically- or photo-labile cleavable moiety. In some embodiments, each spatially tagged oligonucleotide is an RNA oligonucleotide comprising about 26 nt, in the range captured by sRNA library protocols, but not a biologically relevant length.(d) Embodiments of a high-resolution high-density nucleic acid microarray

[0077] In some embodiments, high-density, high-resolution nucleic acid microarrays of the instant disclosure comprise a solid support comprising a flat surface, the flat surface comprising a two-dimensional (2D) grid of a plurality of polygons and positionalPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center markers, wherein each polygon is assigned a specific coordinate in the grid; and a plurality of spatially tagged RNA oligonucleotides affixed to the flat surface within each polygon via a cleavable moiety, wherein all spatially tagged oligonucleotides within a polygon comprise a unique molecular identifier (UMI) and a spatial barcode that associates with a coordinate of the polygon.

[0078] In some embodiments, a microarray of the instant disclosure comprises squares comprising a surface area of about 10, 11 , 12, or 13 pm2or smaller. In some embodiments, a microarray of the instant disclosure comprises squares comprising a surface area of about 10 pm2or smaller. In some embodiments, the squares are delineated by spaces not exceeding about 0.8 micron in width. Accordingly, when the microarray comprises a grid of 10 pm2squares, and the square polygons are delineated by spaces not exceeding about 0.8 micron in width, the squares are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm. In some embodiments, the microarray comprises a grid of about 2,075,600 10 pm2squares, wherein the squares are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm.

[0079] In some embodiments, the microarray further comprises positional markers that facilitate the precise localization of gene expression data. Positional markers can facilitate precise localization of gene expression data by precisely localizing each single cell, mapping the cells back to the array and to the tissue section that is imaged after it is applied to the array, relative to the positional markers on the array for positional alignment in both the image and sequencing data.

[0080] In some embodiments, microarrays of the instant disclosure are manufactured using maskless photolithography. In some embodiments, microarrays of the instant disclosure are manufactured using maskless photolithography capable of generating polygons comprising a surface area of about 10 pm2or smaller, delineated by spaces not exceeding about 0.8 micron in width. In some embodiments, microarrays of the instant disclosure are manufactured using maskless photolithography comprising 1080pPCT PATENT APP Danforth Ref. DDPSC0134-401 -PCT Via Patent Center resolution. In some embodiments, microarrays of the instant disclosure are manufactured using maskless photolithography comprising 4K resolution.

[0081] In some embodiments, a microarray of the instant disclosure comprises a plurality of spatially tagged RNA oligonucleotides affixed to the flat surface within each polygon via a cleavable moiety, wherein all spatially tagged oligonucleotides within a polygon comprise a unique molecular identifier (UMI) and a spatial barcode that associates with a coordinate of the polygon. In some embodiments, all spatially tagged oligonucleotides in each polygon comprise the same spatial barcode. In some embodiments, each spatially tagged oligonucleotide consists essentially of a spatial barcode, a unique molecular ID, and a cleavable moiety.

[0082] In some embodiments, spatially tagged oligonucleotides are RNA oligonucleotides. In some embodiments, each spatially tagged oligonucleotide is an RNA oligonucleotide, wherein each RNA oligonucleotide consists essentially of: a spatial barcode about 12 nucleotides (nt) in length; a UMI about 6-8 nt in length, and a cleavable moiety. In some embodiments, the cleavable moiety is a chemically- or photo-labile cleavable moiety.

[0083] In some embodiments, a high-density, high-resolution nucleic acid microarray of the instant disclosure comprises (a) a solid support comprising a flat surface, the flat surface comprising a 2D grid of about 2,075,600 10 pm2squares, wherein the squares are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm, wherein each polygon is assigned a specific coordinate in the grid; (b) a plurality of spatially tagged RNA oligonucleotides affixed to the flat surface within each polygon; and (c) positional markers. Each spatially tagged oligonucleotide is an RNA oligonucleotide consisting essentially of: a spatial barcode about 12 nucleotides (nt) in length; a UMI about 6-8 nt in length, and a chemically- or photo-labile cleavable moiety.II. Spatially mapping gene expressionPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0084] One embodiment of the instant disclosure encompasses a method of spatially mapping gene expression (spatial transcriptom ics) in a tissue sample. Importantly, methods of the instant disclosure can achieve single cell RNA sequence resolution of spatial data within or starting from intact tissues. While methods have been previously established in many organisms that allow for transcriptional analysis at the level of individual tissue sections or cell types, methods of the instant disclosure have several advantages. Previously established methods include: (1 ) fluorescence-activated cell sorting (FACS) of fluorescently labeled cells, used in animals and plants, of both transgenic and antibody labeled cell types, most often in animals from cells in a suspension or soft tissues; (2) isolation of nuclei tagged in specific cell types (INTACT), developed in plants, but applied to multiple animal systems; and (3) laser capture microdissection (LCM) of tissue sections, developed in animals, but widely deployed in studies of other kingdoms. More modern approaches, including Drop-seq or the Chromium system (10X Genomics), use isolated single cells. While all of these methods have been useful, they each have inherent limitations. The FACS and INTACT methods can require transgenic organisms - challenging in some species - or antibodies that can produce variable results. Additionally, FACS, Drop-seq, and the Chromium system require separation of cells (often chilled, live cells, introducing a major and unwanted treatment and variable), reducing spatial data from the cells. LCM is laborious and can require many hours to collect just hundreds of cells, and it can be limited in quality by the cell type. Technologies that do offer high spatial resolution for RNA localization lack breadth in the number of transcripts characterized. Until recently, one of the few feasible ways to get precise spatial data was to perform fluorescent in situ hybridization (FISH) microscopy, which presents throughput limitations.

[0085] Spatial transcriptom ics methods of the instant disclosure integrate microarrays with RNA sequencing to make innumerable barcoded libraries, indexed by the X / Y coordinate location on the microarray from which the RNA sequence is derived. Before development of methods of the instant disclosure, spatial transcriptom ics methods included placing tissue sections on microarrays covered with reverse transcriptionPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center oligo(dT) primers, each having unique positional barcodes. After fixing and imaging the sample on the array, the mRNA is captured by the oligo(dT) primers on the surface of the array and reverse transcribed into cDNA before subsequent use in library preparation and RNA-seq. Finally, the gene expression data is mapped to specific locations within the tissue section via the barcodes embedded within the array. The method has been commercialized as the 10X Genomics Visium platform. This approach is a powerful tool, yet it has limitations, including array resolution, which, as the Visium system, consists of 55 pm spots with a 100 pm center-to-center spacing, gridded on a 6.5 mm x 6.5 mm array 44. Animal and plant cells can be similarly sized, and may be as small as 10 pm; thus, with the Visium system, many cells are measured together, and there are broad, unsampled gaps in a tissue section. Resolution is being improved by both the 10X company with the release of the “HD” Visium system (consisting of <10 pm spots - but 10X has provided few technical details). Also problematic is the cost of the Visium system, which is roughly $1200 for one slide including kits. Although higher resolution methods comprising beads have been developed, these methods have several drawbacks: (1 ) the methods are highly complex and not transferable to a typical molecular biology lab; (2) the size of the measured surface is only 3 mm ad circular; (3) their efficiency of bead coverage is only 85% (due to gaps and spaces between the beads); and (4) have an efficiency of deconvoluting the beads of only 70%.

[0086] Conversely, methods of the instant disclosure simplify and improve spatial transcriptom ics methods by using densely-packed arrays that substantially improve resolution, minimize gaps, and substantially lower costs when compared to currently available systems. Methods of the instant disclosure comprise use of the high-density, high-resolution nucleic acid microarrays of spatially tagged oligonucleotides of the instant disclosure to tag cells of a tissue, followed by expression profiling of tagged single cells and spatially mapping the expression profile for each tagged cell across the tissue. The method can further comprise correlating transcriptional data with multiscale 3D features of organs and tissues from high resolution microscopy data obtained fromPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center the tissue to generate unparalleled transcriptional data of a whole tissue or even an entire organism.(a) Spatial transcriptomics workflow

[0087] Methods of the instant disclosure comprise providing or having provided a high- density, high-resolution nucleic acid microarray on a flat surface. The microarray comprises a plurality of spatially tagged oligonucleotides affixed to a flat surface of the microarray via a cleavable moiety. In some embodiments, a high-density, high- resolution nucleic acid microarray can be as described in Section I herein above.

[0088] In some embodiments, the high-density, high-resolution nucleic acid microarray comprises a solid support comprising a flat surface, the flat surface comprising a two- dimensional (2D) grid of a plurality of polygons and positional markers, wherein each polygon is assigned a specific coordinate in the grid; and a plurality of spatially tagged RNA oligonucleotides affixed to the flat surface within each polygon via a cleavable moiety. In some embodiments, all spatially tagged oligonucleotides within a polygon comprise a unique molecular identifier (UMI) and a spatial barcode that associates with a coordinate of the polygon. In some embodiments, the microarray comprises a grid of about 2,075,600 10 pm2squares, wherein the squares are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm. In some embodiments, the microarray further comprises positional markers that facilitate the precise localization of gene expression data. In some embodiments, microarrays of the instant disclosure are manufactured using maskless photolithography comprising 1080p resolution.

[0089] In some embodiments, a microarray of the instant disclosure comprises a plurality of spatially tagged RNA oligonucleotides affixed to the flat surface within each polygon via a cleavable moiety, wherein all spatially tagged oligonucleotides within a polygon comprise a unique molecular identifier (UMI) and a spatial barcode that associates with a coordinate of the polygon. In some embodiments, each spatially tagged oligonucleotide consists essentially of a spatial barcode, a unique molecular ID,PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center and a cleavable moiety. In some embodiments, each spatially tagged oligonucleotide is an RNA oligonucleotide, wherein each RNA oligonucleotide consists essentially of: a spatial barcode about 12 nucleotides (nt) in length; a UMI about 6-8 nt in length, and a cleavable moiety. In some embodiments, the cleavable moiety is a chemically- or photo-labile cleavable moiety.

[0090] The methods further comprise generating an expression profile for cells in a tissue sample. Tissue of interest and preparation of a tissue sample of interest for generating expression profiles of cells can be as described in Section (I I )(b).

[0091] Generating an expression profile for cells in a tissue sample comprises tagging the cells in the tissue sample with the spatial barcodes of the microarray, and sequencing RNAs in the tagged cells. Generating an expression profile can further comprise spatially mapping the expression profile for each tagged cell across the tissue based on the identity of the nucleic acid tag in tagged cells. Tagging cells of the tissue of interest, sequencing RNAs in the tagged cells, and spatially mapping the expression profile can be as described in Section ll(c) herein below.

[0092] In some embodiments, the method further comprises correlating transcriptional data with multiscale 3D features of organs and tissues from high resolution microscopy data obtained from the tissue. Constructing a comprehensive 3D volumetric model of the tissue sample and correlating transcriptional data with multiscale 3D features of organs and tissues can be as described in Section ll(d) herein below.

[0093] In some embodiments, a spatial transcriptom ics workflow of the instant disclosure can be as described in FIG. 2.(b) Sample preparation

[0094] The tissue sample can be obtained from any organism or combination of organisms, including but not limited to humans, non-human mammals, non-mammalian vertebrates, invertebrates, insects, plants, fungi, and prokaryotic organisms such as bacteria and archaea. The tissue can be collected from model organisms, wild-type specimens, genetically modified organisms, or environmental sources.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0095] The tissue sample can be obtained from an entire organism or from a part thereof, including any tissue, organ, or combination of tissues and organs. The sample can be a whole organism at any developmental stage (e.g., embryo, juvenile, adult in mammals), or a dissected portion thereof. Plant tissues are also encompassed and can include a whole plant, roots, stems, leaves, floral organs, meristems, seeds, or vascular tissues.

[0096] In some embodiments, the tissue sample is of human origin. In other embodiments, the sample may be derived from non-human mammals such as mouse, rat, rabbit, pig, dog, cat, or non-human primates. In additional embodiments, the sample can be obtained from non-mammalian vertebrates such as birds (e.g., chicken), amphibians (e.g., Xenopus laevis), or fish (e.g., Danio rerio, commonly known as zebrafish). Representative tissues in vertebrate organisms include, but are not limited to, brain, heart, liver, kidney, lung, pancreas, lymph node, skin, muscle, gastrointestinal tract, reproductive organs, neural tissues, endocrine organs, immune organs such as spleen or thymus, and blood-derived cells. The sample can be from an isolated organ, a system of organs, or a mixture of tissues. Embryonic or fetal tissues, as well as tumor tissues and resected lesions, can also be used.

[0097] Tissue can also be invertebrate tissues, including whole organisms such as Caenorhabditis elegans, Drosophila melanogaster, mollusks, or annelids or tissues thereof. Non-limiting examples of tissues include nervous tissue, gut, body wall, and reproductive tissues.

[0098] Insect-derived tissue samples can include whole insects, or specific organs or structures such as antennae, salivary glands, fat body, reproductive tissue, or digestive organs, collected from species such as bees, mosquitoes, ants, and beetles.

[0099] Fungal tissues can include unicellular yeast such as Saccharomyces cerevisiae, pathogenic fungi such as Candida albicans, or filamentous fungi including Neurospora crassa and Aspergillus spp. Samples can consist of single bacterial cells, mycelia, spores, or fruiting bodies.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0100] Prokaryotic tissues or structures can include bacterial colonies or biofilms, such as those formed by Escherichia coli, Staphylococcus aureus, or Pseudomonas aeruginosa. Archaea or extremophile bacteria can also be sampled from environmental or engineered systems. Other prokaryotic tissues or structures are known to individuals of skill in the art.

[0101] In certain embodiments, the tissue sample can comprise additional organisms that are naturally or experimentally associated with the primary tissue source. For example, a plant root can be colonized by mycorrhizal fungi or rhizobacteria, or pathogenic organisms such as viruses, fungi and bacteria, or a human gut biopsy may include commensal, pathogenic, or probiotic bacteria. Tissue samples can also include viral particles, symbiotic microorganisms, or biofilms formed in situ.

[0102] The tissue can be a whole tissue, a dissected portion thereof, or a biopsy taken from a larger structure. Biopsies can be collected via surgical, needle, punch, or endoscopic methods and can include tissue fragments or aspirates.

[0103] The tissue can be healthy, diseased, inflamed, infected, cancerous, fibrotic, degenerative, or in a transitional state. The tissue sample can also represent a baseline or control condition in comparative studies with pathological tissues.

[0104] In some embodiments, the tissue is a plant or a plant tissue, organ, or a combination of tissues or organs, or a plant part. The plant tissue can include a variety of cell types including germ cells and somatic cells. Non-limiting examples of plant cells or cell types include parenchyma cells, sclerenchyma cells, collenchyma cells, xylem cells, and phloem cells. Plant parts can include, but are not limited to, stems, roots, ovules, stamens, leaves, embryos, meristematic regions, callus tissue, gametophytes, sporophytes, pollen, microspores, and the like. The plant can be a monocot plant or a dicot plant. For instance, the plant can be soybean; maize; sugar cane; beet; tobacco; wheat; barley; poppy; rape; sunflower; alfalfa; sorghum; rose; carnation; gerbera; carrot; tomato; lettuce; chicory; pepper; melon; cabbage; oat; rye; cotton; millet; flax; potato; pine; walnut; citrus (including oranges, grapefruit etc.); hemp; oak; rice; petunia; orchids; Arabidopsis', broccoli; cauliflower; brussels sprouts; onion; garlic; leek; squash;PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center pumpkin; celery; pea; bean (including various legumes); strawberries; grapes; apples; cherries; pears; peaches; banana; palm; cocoa; cucumber; pineapple; apricot; plum; sugar beet; lawn grasses; maple; teosinte; Tripsacum; Coix; triticale; safflower; peanut; cassava, duckweed, and olive.

[0105] In some embodiments, the tissue is duckweed. Duckweeds are aquatic plants with a simple body architecture. Duckweeds grow near-exponentially via clonal propagation, and thus have enormous economic potential (not to mention fascinating biology). Duckweed clonal growth results in interconnected asexual individuals that, as a population, represent a convenient and near continuous range of maturity, ideal for studying developmental progression. Plant vegetative growth initiates in the shoot apical meristem (SAM). Therefore, for all plants, the SAM is the source of aboveground organs, including leaves, during vegetative growth due to the group of actively dividing cells that contains a centralized stem cell niche from which cells are generated and contributed to organs. The molecular processes by which this differentiation from stem cells occurs is highly regulated, and involves complex genetic networks, signaling molecules, and tight spatiotemporal control of key transcription factors. Ultimately, the regulation of cell proliferation and the localized modification of the cell wall mechanical properties give rise to first microscopic and then macroscopic changes in shape.Multiple species of duckweed have good reference genomes (-150 Mbp for Spirodela, 500 Mbp for Wolffia) and chromosome-level assemblies. The Spirodela genome shares over 8,000 common gene families with other flowering and model species such as Arabidopsis and rice. Duckweeds have potential in phytoremediation by treatment of wastewater via active removal / recycling of organic nitrogen and phosphate nutrients.

[0106] In some embodiments, the tissue is a Wolffia sp. tissue. In some embodiments, the tissue is a whole Wolffia microscopica plant. Under laboratory conditions, Wolffia microscopica can double in ~23 h. The small size of Wolffia (-500 pm3) provides an opportunity to view how cells operate and organize within an entire multicellular organism and provide insights into developmental processes.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0107] In some embodiments, the tissue sample is a plurality of tissue sections sectioned from the tissue sample. In some embodiments, for the purposes of the methods of the instant disclosure, the tissue sample contacting the microarray has a depth of 1 to 3 intact cells, thereby facilitating downstream RNA sequencing steps that require intact cells.

[0108] In some embodiments, tissue is fixed. Fixation of tissue has the benefit of preserving RNA quality for downstream sequencing steps. Methods of fixing tissue can and will vary depending on the tissue and the intended use of the tissue, and are known or can be developed by individuals of skill in the art. Any method for fixing tissue can be used in methods of the instant disclosure, provided the method preserves RNA quality for downstream sequencing steps of the instant disclosure.

[0109] In some embodiments, tissue of interest is fixed before embedding, sectioning, and contacting the sections of tissue with the surface of the microarray to preserve its cellular and molecular structure. Fixation stabilizes RNA molecules and maintains tissue architecture, which is essential for accurate spatial transcriptom ics.

[0110] Methods and protocols for fixing and tissue sectioning in preparation for spatial transcriptom ics can and will vary depending on the tissue source, the tissue or organ, and the subsequent methods needed for performing RNA transcriptome sequencing. For instance, formaldehyde or paraformaldehyde can be used for fixation, crosslinking proteins and nucleic acids within the cells. Other fixation methods can include methanol fixation, ethanol fixation, acetone fixation, glutaraldehyde fixation, and zinc-based fixatives, depending on the preservation needs of the sample and downstream compatibility with transcriptom ic analysis. The tissue can be immersed in the fixative solution for a period that can range from hours to overnight, depending on the tissue type and size.

[0111] In plants, fixation can enable digestion of the cell wall at high temperatures at which enzymatic activity is optimal and stabilizes the plant cell cytoplasm, rendering cells resistant to mechanical shear force while maintaining high quality RNA. MethodsPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center and protocols for fixing plant tissue for scRNA-seq are known to individuals of skill in the art.

[0112] After fixation, the tissue can undergo dehydration through a graded series of alcohol washes, gradually increasing in concentration. This step can remove water from the tissue, preparing it for embedding. In some protocols, dehydration can be followed by a clearing step to replace alcohol with a reagent such as xylene or other clearing agents, facilitating infiltration of the embedding medium. For paraffin embedding, this can be followed by infiltration with molten paraffin. For cryo-embedding, dehydration can be omitted or replaced with cryoprotection using sucrose or other cryoprotectants to prevent ice crystal formation during freezing. The specific sequence of steps can and will vary depending on the embedding medium used and the downstream analytical requirements.

[0113] During tissue preparation, the tissue can be cleared using solutions like xylene, which helps to further prepare the tissue for embedding by removing residual alcohol and making the tissue more permeable to the embedding medium. Clearing renders tissues more optically transparent and removes lipids or other light-scattering components that may interfere with downstream imaging and visualization. In the context of plant tissues, clearing is particularly important due to the presence of rigid cell walls, dense cellular structures, and internal air pockets, which can obstruct optical access and spatial resolution. Effective clearing enhances light penetration and contrast in subsequent imaging steps and facilitates uniform infiltration of embedding media into internal tissue compartments. In addition to xylene, other commonly used clearing agents include HistoClear, benzyl alcohol / benzyl benzoate (BABB), ethyl cinnamate, and proprietary reagents such as Scale, CLARITY, and CUBIC. The choice of clearing agent can and will depend on the tissue type, compatibility with fixation and embedding protocols, and downstream applications such as imaging or molecular profiling.

[0114] The dehydrated and cleared tissue can then be embedded in a suitable medium for sectioning. Embedding provides structural support to the tissue, enablingPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center the preparation of thin, uniform sections. For embedding, the tissue can be placed in a mold with melted paraffin and allowed to solidify. Alternatively, for cryo-sectioning, the tissue might be embedded in an OCT (Optimal Cutting Temperature) compound and rapidly frozen. Methods of embedding tissue for further processing can and will vary depending on the tissue sample and additional processing steps in the methods of the instant disclosure and are known to individuals of skill in the art.

[0115] Once embedded, the tissue block can be sectioned. The goal is to produce thin tissue sections that can be placed on a microarray or slide. Sectioning techniques can include paraffin sectioning, cryosectioning, vibratome sectioning, or resin sectioning, among others, and may be selected based on the tissue type and intended downstream applications. The thickness of the section can and will vary depending on the size of cells to be subjected to transcriptom ic profiling. More specifically, slices of the tissue are sufficiently thin to ensure that one or two layers of intact cells are retained post-sectioning for proper single cell RNA sequencing.

[0116] The thin tissue sections can be carefully mounted onto microarrays designed for spatial genomics analysis. Microarrays can be treated to enhance tissue adhesion and can contain a predefined grid or array pattern that facilitates the mapping of transcriptom ic data to specific tissue locations.

[0117] During tissue preparation, the tissue can be labeled prior to or during processing, or it may already contain endogenous labels. For instance, tissues can express fluorescent proteins such as GFP, RFP, or YFP under the control of cell-type- specific promoters, allowing for intrinsic visualization of specific cell populations, molecules withing the tissue, or structures. Exogenous labeling methods can include immunofluorescence using antibodies conjugated to fluorophores, in situ hybridization with labeled probes, chemical dyes that stain cellular components (e.g., DAPI for nuclei, phalloidin for F-actin), or lectins for labeling cell walls in plant tissues. Labeling can be used to guide tissue sectioning, assist in registration of serial sections, or provide anatomical context for transcriptom ic data. Imaging of labeled tissues can enable the creation of high-resolution spatial maps, which can be combined with spatialPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center transcriptom ics data, proteomics, or morphological data to digitally reconstruct and analyze the three-dimensional organization of the tissue. This integration facilitates comprehensive characterization of cell types, spatial relationships, and tissue architecture. Additionally, staining procedures, such as hematoxylin and eosin (H&E) staining, can be performed to visualize the tissue morphology and facilitate the identification of different cell types within the sections.

[0118] Finally, the prepared tissue sections on microarray slides are ready for transcriptom ic analysis. This can involve permeabilizing the tissue to enable migration of oligonucleotides into cells. The optimal permeabilization time can be determined experimentally to ensure maximum oligonucleotide migration into cells without compromising tissue integrity. In some embodiments, cells are then removed from the array, separated, and sorted into individual wells on a plate using a Namocell sorter; the Namocell Hana sorter can fill a 384-well plate one cell per well in 6 minutes so is well suited to this task, and it works well with fixed plant cells.

[0119] In some embodiments, before proceeding with transcriptom ic analysis, the mounted tissue sections can undergo additional RNA preservation steps to further stabilize RNA molecules.

[0120] In some embodiments, additional manipulation of the tissue can be used to further increase the resolution of the method. For instance, the tissue can be expanded before contacting cells of the tissue sample with the nucleic acid microarray. In some embodiments, the tissue can be expanded using “expansion microscopy” (“ExM”), a technique to overcome diffraction limitations in microscopy. ExM infiltrates an expandable polyelectrolyte gel matrix into a sample to “grow” the sample 2- to 10- fold or more in a linear direction. This technique expands plant sections for spatial transcriptom ics, further increasing the accessible cellular resolution of transcripts. In some embodiments, the tissue sample is linearly expanded 4 to 5-fold.(c) Generating an expression profilePCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0121] Methods of the instant disclosure comprise generating expression profiles for cells in a tissue sample by tagging the cells with spatial barcodes of a microarray of the instant disclosure, and sequencing RNAs in the tagged cells. The methods can further comprise spatially mapping the expression profile for each tagged cell across the tissue sample based on the identity of the nucleic acid tag sequenced.

[0122] Tagging cells in a tissue sample with the spatial barcodes comprises contacting the cells of the tissue sample with the surface of the nucleic acid microarray, releasing the spatially tagged oligonucleotides to form free spatially tagged oligonucleotides, and diffusing the free spatially tagged oligonucleotides into cells in contact with the nucleic acid microarray to thereby tag each cell with a spatial barcode. As explained herein above, diffusing the free spatially tagged oligonucleotides into cells facilitates a broader and more inclusive capture strategy, enabling the acquisition and sequencing of a wide array of RNA types, including small RNAs (sRNA) and other non- mRNA entities. For instance, diffusing free spatially tagged oligonucleotides into cells using methods and microarrays of the instant disclosure can be used to sequence coding RNAs as well as small RNAs (sRNA), circular RNAs (circRNAs), cleaved or uncapped mRNAs, and non-polyadenylated RNAs, long-non-coding RNAs cleaved or uncapped mRNAs, non-polyadenylated RNAs, mixed host / pathogen (eukaryote / prokaryote), and any combination thereof.

[0123] When the oligonucleotides are releasably affixed to the microarray via a cleavable moiety, tagging cells in a tissue sample with the spatial barcodes can comprise contacting the cells of the tissue sample with the surface of the nucleic acid microarray, cleaving the cleavable moiety in each spatially tagged oligonucleotide to form free spatially tagged oligonucleotides, and diffusing the free spatially tagged oligonucleotides into cells in contact with the nucleic acid microarray to thereby tag each cell with a spatial barcode. In some embodiments, the tissue sample is prepared for contacting with the oligonucleotides as described in Section ll(b).

[0124] Methods of cleaving oligonucleotides of the instant disclosure to form free spatially tagged oligonucleotides can and will vary depending on the cleavable moiety ofPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center the oligonucleotides. Cleavable moieties can be as described in Section l(c) herein above. In some embodiments, a cleavable moiety is a chemically labile moiety, a photo-labile moiety, or a combination thereof. In some embodiments, a cleavable moiety of oligonucleotides of the instant disclosure is a chemically labile moiety. In some embodiments, a cleavable moiety of oligonucleotides of the instant disclosure is a photo-labile moiety.

[0125] After cleaving, the free oligonucleotides are diffused into cells in contact with the oligonucleotides on the nucleic acid microarray to thereby tag each cell with a spatial barcode. Methods of diffusing nucleic acids into cells of a tissue sample are known to individuals of skill in the art. Non-limiting examples of methods of disusing oligonucleotides into cells include simply allowing the oligonucleotides to diffuse into the cells they contact and electrophoresis. In some embodiments, oligonucleotides are diffused into cells using electrophoresis to thereby spatially tag cells of the tissue.Uptake of oligonucleotides by the cells can be assisted using methods normally used for transformation of cells, including, without limitation, electroporation and lipid mediated delivery among other methods known to individuals of skill in the art.

[0126] The expression profiles of tagged cells can then be generated by profiling transcriptomes in each tagged cell using single cell transcript sequencing. A powerful, commonly used technology is a high-throughput, transcriptome profiling method known as RNA-sequencing (RNA-seq). This has become a useful tool to answer fundamental questions in biology and its adoption has been accelerated because it is cost-effective and accurate. RNA-seq provides a rich source of data because the technique can be adapted to many methods. In some embodiments, profiling transcripts in each tagged cell comprises sequencing coding and noncoding RNAs in each cell. Non-limiting examples of RNA-seq methods suitable for methods of the instant disclosure include single-cell RNA sequencing (scRNA-seq), single-cell isoform RNA sequencing (ScISo- Seq), single nucleus RNA sequencing (snRNA-seq), or any combination thereof.

[0127] In some embodiments, the expression profiles of tagged cells are generated using scRNA-seq. scRNA-seq comprises an approach that capturesPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center separated, single cells, and constructs a single barcoded library from each cell, that can be merged and sequenced with other cells. Since an scRNA-seq library can be made in solution, the complexity is relatively good, compared to spatial transcriptom ic methods that require capture of RNA on a surface of a to spatial transcriptom ic methods that require capture of RNA on a surface of a microarray.

[0128] In some embodiments, the expression profiles of tagged cells are generated using scRNA-seq protocols and pipelines capable of obtaining sequences of coding RNAs, noncoding RNAs, or a combination thereof in each cell. In some embodiments, the expression profiles of tagged cells are generated using scRNA-seq methods capable of dual sequencing of mRNA and small RNA. Several methods for dual sequencing of mRNA and small RNA from isolated single cells are known in the art. These start with total RNA from the lysed cells, ultimately as a final step, PCR amplification using both RNAseq and small RNAseq specific primers. In some embodiments, the expression profiles of tagged cells are generated using and scRNA- seq pipeline of FIG. 3.

[0129] In some embodiments, the expression profiles of tagged cells are generated using ScISo-Seq. ScISo-Seq comprises identifying and quantifying transcript isoforms within single cells. Unlike conventional scRNA-seq that primarily captures gene expression levels, ScISo-Seq delves deeper into the transcriptome to differentiate between alternative splicing events. This method enhances our understanding of the post-transcriptional modifications and the diversity of the proteome that can arise from a single gene, offering a more nuanced view of cellular function and differentiation.

[0130] In some embodiments, the expression profiles of tagged cells are generated using snRNA-seq. snRNA-seq is adapted for scenarios where obtaining high-quality, intact single cells might be challenging, such as in solid tissues or postmortem samples. By isolating and sequencing RNA from single nuclei instead of whole cells, snRNA-seq provides an alternative approach to uncover gene expression profiles. This method is particularly useful for analyzing cells from complex tissues,PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center enabling the study of cell types and states in contexts where traditional single-cell dissociation is not feasible.

[0131] Generating an expression profile can further comprise spatially mapping the expression profile for each tagged cell across the tissue -i.e. , generating a gene expression map- based on the identity of the nucleic acid tag in tagged cells. In some embodiments, spatially mapping RNA sequence data in the tissue sample comprises matching each cell in the tissue sample to the RNA sequence data based on the coordinate of the cell and the sequence of the oligonucleotide spatial tag in the sequenced cell. In some embodiments, tissue samples are stained and imaged before generating a sequence profile. In some embodiments tissue samples are stained and imaged after contacting the array and before generating a sequence profile. In such embodiments, matching each cell in the tissue sample to the RNA sequence data based on the coordinate of the cell and the sequence of the oligonucleotide spatial tag in the sequenced cell comprises converting the stained tissue sample into a map of cells that are overlaid by the array polygons, positional markers on the microarray, or both; and correlating the RNA sequence data to a microscopy image by positions relative to the array polygons, the positional markers, or both. Imaging a tissue sample can be as described in Section ll(b) herein above.

[0132] When the tissue sample is a plurality of tissue sections sectioned from the tissue sample, spatially mapped RNA sequence data in each section of the plurality of tissue sections can be assembled into a Z-stack of RNA sequence data to thereby generate a reconstructed 3-dimentional (3D) RNA sequence map. In some embodiments, methods of the instant disclosure can further comprise imaging each of the plurality of tissue sections and combining the imaging data with each gene expression map for each section to form a three-dimensional imaging and gene expression map for the tissue. The methods of the instant disclosure can be further integrating with array-based barcoding of tissue sections. Sectioning a tissue sample and imaging tissue can be as described in Section ll(b) herein above.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center(d) 3D volumetric model

[0133] In some embodiments of methods of the instant disclosure, the methodologies extend to constructing a comprehensive 3D volumetric model of the tissue sample, meticulously detailed down to the single-cell or subcellular level. The creation of such a 3D volumetric model serves multiple critical functions in the context of spatial genomics research. Firstly, the 3D volumetric model acts as a "ground truth" - a highly accurate and reliable benchmark that reflects the true architecture and organization of the tissue in three dimensions. This ground truth can be used to validate and calibrate spatial transcriptom ics data obtained using methods of the instant disclosure, ensuring that the observed gene expression patterns and cellular arrangements closely match the actual biological structure. It provides a verifiable basis against which the spatial arrangement and gene expression data derived from the microarray can be compared, allowing for the correction of potential discrepancies and the refinement of data interpretation strategies.

[0134] Secondly, the 3D volumetric model functions as a "specific reference volume" for each corresponding tissue array dataset, facilitating the precise integration of spatial transcriptom ics data with morphological and structural information. This specific reference volume aids in contextualizing the gene expression data within the physical confines of the tissue's architecture, enabling researchers to pinpoint the exact location of gene expression events within the 3D structure of the tissue. Such detailed mapping can be used to understand the spatial dynamics of gene regulation and function, offering insights into cellular differentiation, tissue development, and disease progression at an unprecedented level of detail.

[0135] Furthermore, the integration of spatial transcriptom ics data with the 3D volumetric model allows for a holistic view of the tissue, combining the molecular specificity of gene expression profiles with the anatomical precision of the tissue's structure. This synergistic approach opens new avenues for exploring the complex interplay between gene expression and tissue morphology, shedding light on the mechanisms that govern tissue organization and function.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0136] The 3D volumetric model can be built using the tissue sample before contacting cells of a tissue sample with the surface of the nucleic acid microarray. Alternatively, the 3D model can be built for a distinct sample of the same tissue of interest. For instance, a different tissue can be used when tissue sample preparation for building a 3D volumetric model are not compatible with downstream tissue preparation methods for generating the spatial transcriptom ics data according to methods of the instant disclosure.

[0137] To generate comprehensive 3D volumetric models of tissues that accurately represent their structure at a single-cell or subcellular level, several imaging techniques can be utilized, prominently featuring histotomography and microtomography techniques. X-ray Microtomography (XRM), a pivotal tool in the realm of histotomography, enables the non-destructive, high-resolution examination of whole tissues, thereby capturing cellular details used for the accurate 3D reconstruction of tissue samples. This technique facilitates a comprehensive visualization of tissue structures in three dimensions without necessitating physical sectioning, making it instrumental in building precise 3D models. XRM can be complemented with Volume Electron Microscopy (vEM) — including Serial Block-Face Scanning Electron Microscopy (SBF-SEM) and Focused Ion Beam Scanning Electron Microscopy (FIB-SEM) — which provides high resolution at the subcellular level. As a cornerstone of microtomography, vEM techniques offer detailed insights into the ultrastructure of tissues, allowing for the meticulous reconstruction of 3D volumetric models that detail the complex organization of cellular and subcellular components. Other non-limiting examples of tools that can be used for building 3D volumetric models include Optical Projection Tomography (OPT) and Light Sheet Fluorescence Microscopy (LSFM). OPT, suited for imaging transparent specimens, bridges microscopic and macroscopic perspectives with its mesoscopic resolution, enhancing histotomography efforts. Meanwhile, LSFM offers rapid, high-resolution imaging of fluorescently labeled tissues, minimizing photodamage and enabling dynamic observation of tissue development and function, thereby complementing microtomography approaches.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0138] In some embodiments a 3D volumetric model of the tissue sample can be built using X-ray microscopy (XRM). In some embodiments, building a whole-tissue 3D model at a single-cell and sub-cellular level comprises combining imaging via XRM with volumetric electron microscopy (vEM) to capture sub-cellular resolution in the 3D model of the tissue. In some embodiments, vEM is serial block-face scanning electron microscopy (SBF-SEM).

[0139] In some embodiments, when methods of the instant disclosure further comprise building a 3D volumetric model of the tissue sample, the methods can further comprise building a whole-tissue 3D model at a single-cell and sub-cellular level of the tissue sample or of a different sample of the tissue by combining imaging via XRM with volumetric electron microscopy (vEM); fixing the tissue sample before contacting the tissue sample with the microarray and before or after building the whole-tissue 3D model of the tissue sample; cryo-sectioning the tissue sample into a plurality of tissue sections; staining the tissue sections, and transferring the tissue sections onto the array; imaging the tissue sections on the microarray before diffusing the oligonucleotides into cells of the tissue sample; spatially mapping the expression profile for each tagged cell across each tissue section by matching each cell in the tissue section to the sequence data based on the coordinate of the cell and the sequence of the oligonucleotide spatial tag in the sequenced cell; assembling the sequence data of the plurality of tissue sections into a Z-stack of RNA sequence data to thereby generate a reconstructed 3D RNA sequence map; and integrating the spatially mapped expression profile for each tagged cell in the 3D RNA sequence map across the tissue with the whole-tissue 3D model. Spatially mapping the expression profile for each tagged cell across each tissue section can comprise: converting the stained tissue sections into a map of cells that are overlaid by the array spots, the positional markers or both; and correlating the RNA-seq data to the microscopy image by positions relative to the positional markers.(e) EmbodimentsPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0140] In some embodiments, a method of spatially resolving gene expression in a tissue sample comprises the steps of: (a) providing or having provided a high-density, high-resolution nucleic acid microarray; (b) contacting cells of a tissue sample with the surface of the nucleic acid microarray; (c) diffusing the free spatially tagged oligonucleotides into cells in contact with the oligonucleotides on the nucleic acid microarray to thereby generate tagged cells; (d) dissociating the tissue sample to form a plurality of cells; (e) performing RNA sequencing in the dissociated tissue sample to generate an expression profile for each tagged cell, wherein the RNA sequencing further sequences the spatially tagged oligonucleotide in the tagged cell; (f) spatially mapping the expression profile for each tagged cell across the tissue sample based on the identity of the nucleic acid tag sequenced in (e). Importantly, the method achieves single-cell resolution.

[0141] In some embodiments, the microarray further comprises positional markers. In some embodiments, the oligonucleotides of the microarray are synthesized using micromirrors for maskless photolithography (maskless array synthesizer (MAS) arrays).

[0142] The microarray comprises a solid support comprising a flat surface, the flat surface comprising a two-dimensional (2D) grid of a plurality of polygons, wherein each polygon is assigned a specific coordinate in the grid; and a plurality of spatially tagged oligonucleotides affixed to the flat surface within each polygon, wherein all spatially tagged oligonucleotides comprise a spatial barcode that associates with a coordinate of the polygon. In some embodiments, the plurality of spatially tagged oligonucleotides are affixed to the flat surface within each polygon via a cleavable moiety and wherein diffusing the free spatially tagged oligonucleotides into cells comprises cleaving the cleavable moiety in each spatially tagged oligonucleotide to form free spatially tagged oligonucleotides. In some embodiments, RNA sequencing is single cell RNA sequencing (scRNA-seq).

[0143] Each polygon of the microarray can comprise a surface area equal to or smaller than a surface area of a single cell in a tissue sample. In some embodiments,PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center each polygon comprises a surface area of about 10 pm2or smaller. Further, the polygons can be delineated by spaces not exceeding about 0.8 micron in width. In some embodiments, the microarray comprises a grid of 10 pm2 polygons, wherein the polygons are distributed such that the center-to-center distance between adjacent polygons is no more than about 10.8 pm. In some embodiments, the polygon is a square. In some embodiments, the microarray comprises a grid of square polygons, wherein each square comprises a surface area of about 10 pm2or smaller. In some embodiments, the microarray comprises a grid of 10 pm2square polygons, wherein the squares are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm. In some embodiments, the microarray comprises a grid of about 2,075,600 10 pm2 square polygons comprising, wherein the squares are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm.

[0144] In some embodiments, only one cell in the tissue sample contacts each square. Additionally, in some embodiments, each cell contacted by the nucleic acid microarray comprises at least one spatially tagged oligonucleotide.

[0145] In some embodiments, all spatially tagged oligonucleotides in each polygon comprise the same spatial barcode. Further, all spatially tagged oligonucleotides within a polygon can further comprise a unique molecular identifier (UM I). In some embodiments, each spatially tagged oligonucleotide consists essentially of a spatial barcode, a UMI, and a cleavable moiety. The spatially tagged oligonucleotides can be RNA oligonucleotides. In some embodiments, each spatially tagged oligonucleotide is an RNA oligonucleotide, wherein each RNA oligonucleotide consists essentially of: a spatial barcode about 12 nucleotides (nt) in length; a UMI about 6-8 nt in length, and a cleavable moiety. In some embodiments, the cleavable moiety is a chemically- or photo-labile cleavable moiety. In some embodiments, each spatially tagged oligonucleotide is an RNA oligonucleotide comprising about 26 nt, in the range captured by sRNA library protocols, but not a biologically relevant length.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0146] In some embodiments, the spatially tagged oligonucleotides are diffused into the tissue sample using electrophoresis.

[0147] RNA sequencing can comprise sequencing coding and noncoding RNAs in each cell. In some embodiments, the noncoding RNAs are selected from small RNAs (sRNA), circular RNAs (circRNAs), cleaved or uncapped mRNAs, and nonpolyadenylated RNAs, long-non-coding RNAs cleaved or uncapped mRNAs, nonpolyadenylated RNAs, mixed host / pathogen (eukaryote / prokaryote), and any combination thereof.

[0148] In some embodiments, the tissue sample is expanded before contacting cells of the tissue sample with the nucleic acid microarray. For instance, the tissue sample can be expanded via injection of an expandable polyelectrolyte gel matrix into the tissue sample. In some embodiments, the tissue sample is linearly expanded 4 to 5-fold. In some embodiments, nucleic acids and proteins in the tissue sample are cross-linked to the expandable polyelectrolyte gel matrix before expansion.

[0149] In some embodiments, the tissue sample is a plurality of tissue sections sectioned from the tissue sample. In some embodiments, the tissue sample contacting the microarray has a depth of 1 to 3 cells.

[0150] In some embodiments, the tissue sample is fixed and sectioned into a plurality of tissue sections prior to contacting the array. The tissue or tissue section can be stained prior to (pre-infiltrated) or after contacting the array. In some embodiments, the tissue sample is fixed, cryo-sectioned into a plurality of tissue sections prior to contacting the array, then stained.

[0151] In some embodiments, methods of the instant disclosure further comprise imaging the tissue sample before diffusing the oligonucleotides into the cells. In some embodiments, the tissue is imaged on the microarray. In some embodiments, the method comprises fixing the tissue sample, cryo-section ing into a plurality of tissue sections, staining, transferring onto the array, and imaging. The imaging can correlate each cell on the array with a coordinate polygon.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0152] In some embodiments, spatially mapping RNA sequence data in the tissue sample comprises matching each cell in the tissue sample to the RNA sequence data based on the coordinate of the cell and the sequence of the oligonucleotide spatial tag in the sequenced cell. Spatially mapping RNA sequence data in the tissue sample can comprise matching each cell in the tissue sample to the RNA sequence data based on the coordinate of the cell and the sequence of the oligonucleotide spatial tag in the sequenced cell by: converting the stained tissue sample into a map of cells that are overlaid by the array polygons, the positional markers, or both; and correlating the RNA sequence data to the microscopy image by positions relative to the array polygons, the positional markers, or both.

[0153] In some embodiments, methods of the instant disclosure further comprise integrating the method with array-based barcoding of tissue sections. The tissue sample can be a plurality of tissue sections sectioned from the tissue sample and spatially mapped RNA sequence data in each section of the plurality of tissue sections is assembled into a Z-stack of RNA sequence data to thereby generate a reconstructed 3-dimentional (3D) RNA sequence map. In some embodiments, the method further comprises imaging each of the plurality of tissue sections and combining the imaging data with each gene expression map for each section to form a three-dimensional imaging and gene expression map for the tissue.

[0154] In some embodiments, the method further comprises building a wholetissue 3D volumetric model at a single-cell level of the tissue sample before step (b) or of a different sample of the tissue of interest, wherein the 3D model serves as a ground truth and specific reference volume for each corresponding tissue array dataset for subsequent integration. In some embodiments, building a whole-tissue 3D volumetric model comprises imaging via X-Ray microscopy (XRM). The methods can further comprise building a whole-tissue 3D model at a single-cell and sub-cellular level of the tissue of interest before contacting cells of a tissue sample with the surface of the microarray or building a whole-tissue 3D model at a single-cell and sub-cellular level of a different sample of the tissue of interest. Building a whole-tissue 3D model at aPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center single-cell and sub-cellular level can comprise combining imaging via XRM with volumetric electron microscopy (vEM) to capture sub-cellular resolution in the 3D model of the tissue. In some embodiments, vEM is serial block-face scanning electron microscopy (SBF-SEM). In some embodiments, the methods further comprise integrating the spatially mapped expression profile for each tagged cell in the 3D RNA sequence map across the tissue with the whole-tissue 3D model. In some embodiments, the method further comprises: building a whole-tissue 3D model at a single-cell and sub-cellular level of the tissue sample or of a different sample of the tissue by combining imaging via XRM with volumetric electron microscopy (vEM); fixing the tissue sample before contacting cells of a tissue sample with the surface of the nucleic acid microarray and before or after building the whole-tissue 3D model of the tissue sample, cryo-sectioning the tissue sample into a plurality of tissue sections, staining the tissue sections, and transferring the tissue sections onto the array; imaging the tissue sections on the microarray before diffusing the free spatially tagged oligonucleotides into cells; spatially mapping the expression profile for each tagged cell across each tissue section by matching each cell in the tissue section to the sequence data based on the coordinate of the cell and the sequence of the oligonucleotide spatial tag in the sequenced cell; assembling the sequence data of the plurality of tissue sections into a Z-stack of RNA sequence data to thereby generate a reconstructed 3D RNA sequence map; and integrating the spatially mapped expression profile for each tagged cell in the 3D RNA sequence map across the tissue with the whole-tissue 3D model. In such embodiments, spatially mapping the expression profile for each tagged cell across each tissue section comprises: converting the stained tissue sections into a map of cells that are overlaid by the array spots, the positional markers or both; and correlating the RNA-seq data to the microscopy image by positions relative to the positional markers.

[0155] In some embodiments, the tissue sample is an animal or plant tissue sample. In some embodiments, the tissue sample is a whole Wolffia sp. plant.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0156] In some embodiments, a method of spatially mapping gene expression across three dimensions in a tissue sample comprises: (a) fixing and cryo-sectioning the tissue sample into a plurality of tissue sections that are 1-3 cells deep; (b) spatially mapping RNA sequence data in each section to generate a set of gene expression maps for each section; (c) assembling the gene expression maps for each section into a Z-stack of RNA sequence data to thereby generate a reconstructed 3D RNA sequence map. Spatially mapping RNA sequence data in each section can be as described in Sections ll(a) to ll(c) herein above. In some embodiments, the method can further comprise imaging each tissue section and combining the imaging data with each gene expression map for each section to form a three-dimensional imaging and gene expression map for the tissue. In some embodiments, the method can further comprise building a whole-tissue 3D volumetric model at a single-cell level of the tissue sample before step (b) or of a different sample of the tissue of interest. The 3D model can serve as a ground truth and specific reference volume for each corresponding tissue array dataset for subsequent integration. Building a whole-tissue 3D volumetric model can be as described in Section ll(d) herein above.

[0157] In some embodiments, the tissue is an animal or plant tissue. In some embodiments, the tissue is a plant tissue. In some embodiments, the tissue sample is a whole Wolffia sp. plant.HIGH-DENSITY, HIGH-RESOLUTION NUCLEIC ACID MICROARRAY

[0158] An additional aspect of the instant disclosure encompasses a high- density, high-resolution nucleic acid microarray for positionally mapping gene expression in a tissue sample. The microarray can be as described in Section I herein above. The microarray comprises: (a) a solid support comprising a flat surface, the flat surface comprising a two-dimensional (2D) grid of a plurality of polygons, wherein each polygon is assigned a specific coordinate in the grid; and (b) a plurality of spatially tagged oligonucleotides affixed to the flat surface within each polygon, wherein all spatially tagged oligonucleotides comprise a spatial barcode that associates with aPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center coordinate of the polygon. The plurality of spatially tagged oligonucleotides can be affixed to the flat surface within each polygon via a cleavable moiety and diffusing the free spatially tagged oligonucleotides into cells can comprise cleaving the cleavable moiety in each spatially tagged oligonucleotide to form free spatially tagged oligonucleotides. In some embodiments, the microarray further comprises positional markers. In some embodiments, each polygon comprises a surface area equal to or smaller than a surface area of a single cell in a tissue. Each polygon can comprise a surface area of about 10 pm2 or smaller. Further, the polygons can be delineated by spaces not exceeding about 0.8 micron in width. In some embodiments, the microarray comprises a grid of 10 pm2 polygons, wherein the polygons are distributed such that the center-to-center distance between adjacent polygons is no more than about 10.8 pm. In some embodiments, the polygons is a square shaped.

[0159] In some embodiments, the microarray comprises a grid of polygons, wherein each square comprises a surface area of about 10 pm2 or smaller. In some embodiments, the microarray comprises a grid of 10 pm2 squares, wherein the squares are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm. In some embodiments, the microarray comprises a grid of about 2,075,600 10 pm2 square polygons comprising, wherein the squares are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm. All spatially tagged oligonucleotides in each polygon can comprise the same spatial barcode. In some embodiments, each spatially tagged oligonucleotide consists essentially of a spatial barcode, a unique molecular ID, and a cleavable moiety.

[0160] In some embodiments, the spatially tagged oligonucleotides are RNA oligonucleotides, DNA oligonucleotides, or RNA / DNA oligonucleotides. In some embodiments, spatially tagged oligonucleotides are RNA oligonucleotides. Each spatially tagged oligonucleotide can be an RNA oligonucleotide, wherein each RNA oligonucleotide consists essentially of: (a) a spatial barcode about 12 nucleotides (nt) in length; (b) a IIMI about 6-8 nt in length, and (c) a cleavable moiety.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center

[0161] In some embodiments, the cleavable moiety is a chemically- or photo- labile cleavable moiety. In some embodiments, the oligonucleotides of the microarray are synthesized using micromirrors for maskless photolithography (maskless array synthesizer (MAS) arrays).III. Kits

[0162] A further embodiment of the present disclosure provides kits comprising one or more high-resolution high-density nucleic acid microarray detailed above in Section I herein above. The kits can further comprise tissue preparation reagents, microscopy reagents, cell growth media, selection media, and the like, or any combination thereof. The kits provided herein generally include instructions for carrying out the methods detailed herein above. Instructions included in the kits may be affixed to packaging material or may be included as a package insert. While the instructions are typically written or printed materials, they are not limited to such. Any medium capable of storing such instructions and communicating them to an end user is contemplated by this disclosure. Such media include, but are not limited to, electronic storage media (e.g., magnetic discs, tapes, cartridges, chips), optical media (e.g., CD ROM), and the like. As used herein, the term “instructions” may include the address of an internet site that provides the instructions.DEFINITIONS

[0163] Unless defined otherwise, all technical and scientific terms used herein have the meaning commonly understood by a person skilled in the art to which this invention belongs. The following references provide one of skill with a general definition of many of the terms used in this invention: Singleton et al., Dictionary of Microbiology and Molecular Biology (2nd ed. 1994); The Cambridge Dictionary of Science and Technology (Walker ed., 1988); The Glossary of Genetics, 5th Ed., R. Rieger et al. (eds.), Springer Verlag (1991 ); and Hale & Marham, The Harper Collins Dictionary ofPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent CenterBiology (1991 ). As used herein, the following terms have the meanings ascribed to them unless specified otherwise.

[0164] When introducing elements of the present disclosure or the preferred aspects(s) thereof, the articles "a", "an", "the" and "said" are intended to mean that there are one or more of the elements. The terms "comprising", "including" and "having" are intended to be inclusive and mean that there may be additional elements other than the listed elements.

[0165] As used herein, the term “MAS” refers to a “maskless array synthesizer”, such as the one using the technology described by NimbleGen Systems, Madison, Wis. (Singh-Gasson et al., 1999, Nature Biotechnology 17: 974-978). A “MAS derived microscope slide” refers to a microscopic slide on which an array of oligonucleotides has been synthesized using a maskless array synthesizer.

[0166] The terms “nucleic acid,” “oligonucleotide,” and “polynucleotide” refer to a deoxyribonucleotide or ribonucleotide polymer, in linear or circular conformation. For the purposes of the present disclosure, these terms are not to be construed as limiting with respect to the length of a polymer. The terms may encompass known analogs of natural nucleotides, as well as nucleotides that are modified in the base, sugar and / or phosphate moieties. In general, an analog of a particular nucleotide has the same base-pairing specificity, i.e., an analog of A will base-pair with T. The nucleotides of a nucleic acid or polynucleotide may be linked by phosphodiester, phosphothioate, phosphoram idite, phosphorodiamidate bonds, or combinations thereof.

[0167] The term "nucleotide" refers to deoxyribonucleotides or ribonucleotides. The nucleotides may be standard nucleotides (i.e., adenosine, guanosine, cytidine, thymidine, and uridine) or nucleotide analogs. A nucleotide analog refers to a nucleotide having a modified purine or pyrimidine base or a modified ribose moiety. A nucleotide analog may be a naturally occurring nucleotide (e.g., inosine) or a non- naturally occurring nucleotide. Non-limiting examples of modifications on the sugar or base moieties of a nucleotide include the addition (or removal) of acetyl groups, amino groups, carboxyl groups, carboxymethyl groups, hydroxyl groups, methyl groups,PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center phosphoryl groups, and thiol groups, as well as the substitution of the carbon and nitrogen atoms of the bases with other atoms (e.g., 7-deaza purines). Nucleotide analogs also include dideoxy nucleotides, 2’-O-methyl nucleotides, locked nucleic acids (LNA), peptide nucleic acids (PNA), and morpholinos.

[0168] Techniques for determining nucleic acid and amino acid sequence identity are known in the art. Typically, such techniques include determining the nucleotide sequence of the mRNA for a gene and / or determining the amino acid sequence encoded thereby, and comparing these sequences to a second nucleotide or amino acid sequence. Genomic sequences may also be determined and compared in this fashion. In general, identity refers to an exact nucleotide-to-nucleotide or amino acid-to- amino acid correspondence of two polynucleotides or polypeptide sequences, respectively. Two or more sequences (polynucleotide or amino acid) may be compared by determining their percent identity. The percent identity of two sequences, whether nucleic acid or amino acid sequences, is the number of exact matches between two aligned sequences divided by the length of the shorter sequences and multiplied by 100. An approximate alignment for nucleic acid sequences is provided by the local homology algorithm of Smith and Waterman, Advances in Applied Mathematics 2:482- 489 (1981 ). This algorithm may be applied to amino acid sequences by using the scoring matrix developed by Dayhoff, Atlas of Protein Sequences and Structure, M. 0. Dayhoff ed., 5 suppl. 3:353-358, National Biomedical Research Foundation, Washington, D.C., USA, and normalized by Gribskov, Nucl. Acids Res. 14(6):6745-6763 (1986). An exemplary implementation of this algorithm to determine percent identity of a sequence is provided by the Genetics Computer Group (Madison, Wis.) in the "BestFit" utility application. Other suitable programs for calculating the percent identity or similarity between sequences are generally known in the art, for example, another alignment program is BLAST, used with default parameters. For example, BLASTN and BLASTP may be used using the following default parameters: genetic code=standard; filter=none; strand=both; cutoff=60; expect=10; Matrix=BLOSUM62; Descriptions=50 sequences; sort by = HIGH SCORE; Databases=non-redundant,PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent CenterGenBank+EMBL+DDBJ+PDB+GenBank CDS translations+Swiss protein+Spupdate+PIR. Details of these programs may be found on the GenBank website. With respect to sequences described herein, the range of desired degrees of sequence identity is approximately 80% to 100% and any integer value therebetween. Typically the percent identities between sequences are at least 70-75%, preferably 80- 82%, more preferably 85-90%, even more preferably 92%, still more preferably 95%, and most preferably 98% sequence identity.

[0169] As various changes could be made in the above-described cells and methods without departing from the scope of the invention, it is intended that all matter contained in the above description and in the examples given below, shall be interpreted as illustrative and not in a limiting sense.EXAMPLES

[0170] All patents and publications mentioned in the specification are indicative of the levels of those skilled in the art to which the present disclosure pertains. All patents and publications are herein incorporated by reference to the same extent as if each individual publication was specifically and individually indicated to be incorporated by reference.

[0171] The publications discussed throughout are provided solely for their disclosure before the filing date of the present application. Nothing herein is to be construed as an admission that the invention is not entitled to antedate such disclosure by virtue of prior invention.

[0172] The following examples are included to demonstrate the disclosure. It should be appreciated by those of skill in the art that the techniques disclosed in the following examples represent techniques discovered by the inventors to function well in the practice of the disclosure. Those of skill in the art should, however, in light of the present disclosure, appreciate that many changes could be made in the disclosure and still obtain a like or similar result without departing from the spirit and scope of thePCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center disclosure, therefore all matter set forth is to be interpreted as illustrative and not in a limiting sense.Example 1. Overview

[0173] The fundamental, new insight driving this project is that microarrays synthesized by photolithography offer unparalleled advantages for spatial transcriptom ics, particularly when coupled with high resolution 3D imaging techniques for reconstruction of the organism. While the last few years have brought many advances in spatial transcriptom ics, all of these methods fall short in one aspect or another: poor resolution, too expensive, poor capture of mRNA, no small RNA (sRNA) data, proprietary methods, etc. Integration of these data with 3D volumes is also problematic, in part because the optimal spatial transcriptom ic data sets don’t yet exist. The inventors devised a new process to make a technological leap with spatial transcriptom ics and integrate this with a new generation of imaging methods, delivering a complete package of methods and technologies for transcriptional analysis and characterization of multicellular organisms, essentially creating a four-dimensional map - 3D space plus developmental sequence.

[0174] The study and functional characterization of specialized tissue and cellular domains in multicellular organisms is critical to understand their gene networks and biological systems. The recent development of spatially resolved gene expression profiling allows for high-throughput, in situ transcript profiling. However, several implementations have low cellular resolution and single-cell RNA-seq (scRNA-seq) approaches lack spatial definition. Photolithography-based microarrays are “programmable” (can be custom manufactured) at extremely high resolution, and when coupled with scRNA-seq and imaging approaches, could yield the ultimate goal of many labs - cost-effective single-cell, spatially resolved transcriptional data. The examples herein demonstrate how the devised methods advance the methodology by developing, implementing, and integrating several proven technologies that have not previously been linked together, yielding scRNA-seq in a spatially resolved manner. CapturePCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center efficiency for spatial transcriptom ics has been a major challenge, due to the requirement of surface interactions on array-based methods, and one of the insights by the inventors is that cleavable oligos released from the array surface can penetrate cell sections to provide a molecular tag for cells in situ (in a section), prior to separation. Another insight is that integration of the spatial transcriptom ics data with 3D organismal models is not yet a solved problem. An additional insight is that few, if any, methods have been reported to capture sRNAs or any RNAs other than mRNA, in an unbiased, genomewide basis for spatial transcriptom ics. The integration of the set of technologies described herein is highly synergistic, and has yet to be achieved due to the extensive experimentation needed to achieve the desired results. This project, for the first time, couples three spatial technologies with multi-scale imaging approaches to provide cellular and organismal context to multi-dimensional spatial transcriptome data. Via the integration of these high-throughput spatial RNA analysis methods, the project measures where in cells and tissues, and at what level, genes and sRNAs are expressed in whole organisms, to put transcriptional data into a spatial context, at single- cell resolution.

[0175] The work in these examples is performed using a model for plant development, duckweed. Plants offer multiple advantages. Insights from plants have made many important contributions to animal studies. Basic research and discoveries utilizing plant systems include major advances such as the connection between sRNAs and gene silencing , the discovery of RNA polymerase diversification, and the characterization of the phased nature of reproductive sRNAs that pre-dates by several years the identification of phasing of metazoan piRNAs. The viability of plant mutants in sRNA pathways, the sophistication of plant genetics, and the accessibility of tissue for biochemical analyses underpin the rapid progress in sRNA biology and support plants as valuable models for genetics and RNA biology. Plants also have more complex populations of sRNAs than animals, the result of additional RNA polymerases (RNA Pol IV and Pol V, unique to land plants) that yield heterochromatic siRNAs that function in DNA methylation and genome defense. In the context of these examples, thePCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center extracellular matrix of plant cells (mainly the cell wall) is consistent across tissues and easily broken down with enzymatic treatments, allowing for single-cell isolation.

[0176] To fully implement the methods, the inventors focus on several major advances. (1 ) A method in which barcodes from the array surface are infused into the tissue section; the section with infused, positional barcodes is subsequently separated into single cells for scRNA-seq. The barcodes are sequenced within each cell, along with endogenous RNAs, yielding precise positional information and scRNA-seq data for both mRNA and sRNA. (2) “Expansion microscopy” (“ExM”), a technique to overcome diffraction limitations in microscopy is implemented. ExM infiltrates an expandable polyelectrolyte gel matrix into a sample to “grow” the sample 4- to 5-fold in a linear direction. In the examples herein, this technique to expand plant sections for spatial transcriptom ics, further increasing the accessible cellular resolution of transcripts. (3) These novel, powerful transcriptional methods are coupled with whole-organism, nondestructive, 3D X-ray microtomography as a major tool in creating intermediate 3D models / framework for mapping transcriptional data. These X-ray-generated 3D models are further correlated, augmented and annotated with critical insights from high- resolution subcellular details provided by “volume Electron Microscopy” (“vEM”) where certain features or cell-types are not resolvable or otherwise contrasted via X-ray strategies. Notably XRM and vEM have recently been shown capable of producing coaligned multi-scale 3D models. The biological aim of the examples herein is to generate these multiscale data sets in the context of duckweed development. The examples describe the production of the ultimate in fine-scale transcriptional analysis - single- cell, whole-organism mRNA and sRNA data in their native 3D context over a developmental series. To corroborate / validate these transcriptional data, single molecule FISH analysis are applied to a subset of genes.Organismal system

[0177] In order to develop robust methodologies, the inventors focus on duckweed. Duckweeds are aquatic plants with a simple body architecture and growPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center near-exponentially via clonal propagation, and thus have enormous economic potential (not to mention fascinating biology). Under laboratory conditions, Wolffia microscopica (the target species) can double in ~23 h. The small size of Wolffia (~500 pm3) provides an opportunity to view how cells operate and organize within an entire multicellular organism and provide insights into developmental processes. Duckweed clonal growth results in interconnected asexual individuals that, as a population, represent a convenient and near continuous range of maturity, ideal for studying developmental progression. Plant vegetative growth initiates in the shoot apical meristem (SAM), and thus for all plants, the SAM is the source of above-ground organs, including leaves, during vegetative growth due to the group of actively dividing cells that contains a centralized stem cell niche from which cells are generated and contributed to organs. The molecular processes by which this differentiation from stem cells occurs is highly regulated, and involves complex genetic networks, signaling molecules, and tight spatiotemporal control of key transcription factors. Ultimately, the regulation of cell proliferation and the localized modification of the cell wall mechanical properties give rise to first microscopic and then macroscopic changes in shape.

[0178] Multiple species of duckweed have good reference genomes (~150 Mbp for Spirodela, 500 Mbp for Wolffia) and chromosome-level assemblies; the Spirodela genome shares over 8,000 common gene families with other flowering and model species such as Arabidopsis and rice. Duckweeds have potential in phytoremediation by treatment of wastewater via active removal / recycling of organic nitrogen and phosphate nutrients.Transcriptome profiling at a near single-cell level

[0179] Multicellular organisms, including plants, are able to upregulate and downregulate a plethora of genes throughout development, abiotic and biotic stresses, and circadian cycles. Identifying and quantifying these differentially expressed genes are important for understanding growth, development, and responses. A powerful, commonly used technology is a high-throughput, transcriptome profiling method knownPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center as RNA-sequencing (RNA-seq). In recent years, this has provided a cost-effective and accurate tool to answer fundamental questions in plant biology. RNA-seq also provides a rich source of data because the technique can be adapted to many methods. Singlecell RNA-seq, abbreviated to scRNA-seq, is a modern approach that captures separated, single cells, and constructs a single barcoded library from each cell, that can be merged and sequenced with other cells. Since the scRNA-seq library is made in solution, the complexity is relatively good, compared to spatial transcriptom ic methods that require capture of RNA on a surface.

[0180] RNA-seq is being used to understand regulatory networks in development, responses to stresses, etc., via computational modeling. Modeled networks are used to predict new or additional functions for previously described genes or their products. Additionally, gene networks can help to predict potential interactions among genes and provide a systemic view of the molecular mechanisms of a biological process.However, a complete understanding of these biological processes requires functional characterization of specialized tissue domains. Gene expression profiling in specific organs, tissues, and cell types has been an active research area for decades. As with other organisms, transcriptional analysis in plants has advanced substantially over the last decade, albeit with some differences due to the nature and structure of plant cells, including the cell wall, vacuoles, and chloroplasts. For example, to disrupt the cell wall, and in contrast to mammalian systems, plant tissue is first flash-frozen in liquid nitrogen in order to extract RNA, and an enzymatic process is needed to remove the cell wall to isolate cellular contents (this generates protoplasts). These steps are actually beneficial for the proposed transcriptional analysis, as it is possible to flash-fix cells, then protoplast the cells for scRNA-seq, yielding high quality RNA and complex single-cell libraries.Spatially-resolved transcriptome profiling

[0181] These examples achieve scRNA-seq resolution of spatial data within or starting from intact tissues. While methods have been previously established in manyPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center organisms that allow for transcriptional analysis at the level of individual tissue sections or cell types, method described herein have advantages. Previously established methods include the following: (1 ) fluorescence-activated cell sorting (FACS) of fluorescently labeled cells, used in animals and plants, of both transgenic and antibody labeled cell types, most often in animals from cells in a suspension or soft tissues; (2) isolation of nuclei tagged in specific cell types (INTACT), developed in plants, but applied to multiple animal systems; and the ‘classic’, (3) laser capture microdissection (LCM) of tissue sections, developed in animals, but widely deployed in studies of other kingdoms. More modern approaches including Drop-seq or the Chromium system (10X Genomics) use isolated single cells. While all of these methods have been useful, they each have inherent limitations. The FACS and INTACT methods can require transgenic organisms - challenging in some species - or antibodies that can produce variable results. Additionally, FACS, Drop-seq, and the Chromium system require separation of cells (often chilled, live cells, introducing a major and unwanted treatment and variable), reducing spatial data from the cells. LCM is laborious and can require many hours to collect just hundreds of cells, and it can be limited in quality by the cell type. Technologies that do offer high spatial resolution for RNA localization lack breadth in the number of transcripts characterized. Until recently, one of the few feasible ways to get precise spatial data was to perform fluorescent in situ hybridization (FISH) microscopy, which presents throughput limitations.

[0182] A strategy that would provide more comprehensive, cell-type specific data is spatially-resolved, high-resolution and high-throughput transcriptom ic analysis.“Spatial transcriptom ics” was originally developed for mammalian systems and applied to plants, and is a recent technology that integrates microarrays with RNA-seq to make innumerable barcoded libraries, indexed by the X / Y coordinate location from which the RNA is derived. This is performed by placing tissue sections on microarrays covered with reverse transcription oligo(dT) primers, each having unique positional barcodes. After fixing and imaging the sample on the array, the mRNA is captured (by interactions on the surface of the array - a weakness that we avoid) and reversed transcribed intoPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center cDNA before subsequent use in library preparation and RNA-seq. Finally, the gene expression data are mapped to specific locations within the tissue section via the barcodes embedded within the array. The method has been commercialized as the 10X Genomics Visium platform. This approach is a powerful tool, yet it has limitations, including array resolution, which, as the Visium system, consists of 55 pm spots with a 100 pm center-to-center spacing, gridded on a 6.5 mm x 6.5 mm array. Animal and plant cells are similarly sized, and may be as small as 10 pm; thus, with the Visium system, many cells are measured together, and there are broad, unsampled gaps in a tissue section. Resolution is being improved by both the 10X company with the release of the “HD” Visium system (consisting of <10 pm spots - but 10X has provided few technical details). Also problematic is the cost of the Visium system, which is roughly $1200 for one slide including kits (our proposed method should be much cheaper).Example 2. Spatially-resolved transcriptional analysis at single-cell resolution, using MAS arrays

[0183] In this example, spatial transcriptom ics are adapted to MAS-produced microarrays. These arrays are made using digital micromirrors for light-activated oligo synthesis (FIG. 1). Newer MAS instruments developed for oligonucleotide synthesis use 1080p arrays (Table 1) that substantially increases the resolution compared to previous instruments ~14 pm center-to-center spots (gridded squares). These arrays are sufficiently dense to match single-cell sizes of animals and plants. The MAS arrays are rectangular and the 1080p arrays are comprised of 2,075,600 densely packed spots; an array can be split into gasket-separated quadrants allowing us to multiplex samples - in this case, for duckweed sections of up to 1 mm dimensions and 1 mm spacing, 16 duckweed slices can be loaded / replicated more than for one array (four per quadrant). The arrays are synthesized in situ and covalently immobilized onto the array. Oligos sequences on the arrays for spatial transcriptom ics vary in the spatial barcode that designates their X-Y position on the array and a randomized UniquePCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent CenterMolecular Identifier (UMI) (FIG. 2). Thus, for spatial transcriptom ics, the MAS instrument yields high quality, long oligos.Example 3. Linking spatial transcriptomics with scRNA-seq for high efficiency “spatial scRNA-seq” libraries

[0184] The current approaches for standard spatial transcriptomics depend on diffusion of cellular RNA onto the array surface, followed by in situ library construction, elution off the array surface, and sequencing. Although this is currently the ‘state of the art’, there are drawbacks that severely limit efficiency: (1 ) the RNA from the tissue sections must diffuse onto the surface, and this RNA can anneal with oligos only at the array surface, and (2) after annealing, the oligos are enzymatically cleaved off of the surface (a requirement for subsequent library construction steps.). The efficiency of the USER enzyme (Uracil-Specific Excision Reagent, from NEB) in cleaving the oligos is low, rarely more than 50%, leaving much of the annealed RNA (already a small subset of RNA in the cells) stuck on the array surface. These combined limitations result in poor library yields, and all three issues are a drawback of a surface-based process, as implemented in the Visium system and related approaches.

[0185] The inventors provide a substantial, yet simple, rethinking of this process, which can provide major advantages. The approach is to diffuse the barcodecontaining RNA oligos off the arrays and up into a tissue section, and then remove the intact section, separate the individual cells within the section into individual wells of plates, and perform scRNA-seq in solution, capturing the barcodes that were carriedPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center with the separated cells (FIG. 3). RNA oligos that are 26 nt in length are provided, in the range captured by sRNA library protocols, but not a biologically relevant length; the oligos include a 12 nt spatial barcode, a spacer, and a 6 or 8 nt UMI. This length of oligo is easily synthesized with high fidelity using the MAS process, including a labile, cleavable linker at the base. scRNA-seq library construction from separated cells is both relatively mature as an approach and far more efficient than current spatial methods using array surface chemistry. In the provided approach, only a small number of barcoded oligos need to be captured and passed through to the sequencing results in order to localize the cells on the array. The barcoded RNA oligos can be efficiently cleaved from the surface of the array using a chemically or photo-labile, cleavable linker at the surface of the array, far more efficient than the current USER enzyme, and oligos can be pre-cleaved on the surface before applying the section for maximal efficiency. Moreover, short, synthetic RNAs are easily synthesized on the array, with each 10 pm spot containing a unique positional barcode + UMI. This process yields “spatial scRNA- seq”.

[0186] The approach provided by the inventors builds on protocols that first fix tissues and then separate the fixed cells by dissolving the extracellular matrix (cell wall), followed by single-cell library construction. The immediate fixation preserves RNA quality. Thicker sections of fixed tissue are made, retaining one or two layers of cells that are intact, post-sectioning. Alternatively, and to further enhance both spatial resolution and oligo accessibility, a modified protocol for ExM in plants is used by expanding vibratome-generated sections for microarrays. The oligos are moved into cells, cells are then removed from the array, separated, and sorted into individual wells on a plate using a Namocell sorter; the Namocell Hana sorter can fill a 384-well plate one cell per well in 6 minutes so is well suited to this task, and it works well with fixed plant cells. Finally, from these sorted single cells, scRNA-seq libraries are generated using either CEL-seq2, or one of the emerging methods for capturing both mRNA and sRNA, or optimized methods can be provided. A key step is the capture from each cell of the synthetic barcode(s) from the array that indicates its location. To summarize,PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent CenterRNA is synthesized on the array, cleaved, and diffused into the cells, and the barcode is captured as part of a sRNA library made from the same cell (FIG. 3). This approach fits with an additional aspect of the work, which is to sequence both mRNA and sRNA in a spatial manner. Several methods have been published for dual capture of mRNA and sRNA, although they all lack spatial resolution.

[0187] From the scRNA-seq data, the barcodes are used to precisely localize each single cell, mapping the cells back to the array and to the tissue section that is imaged after it is applied to the array, relative to laser-engraved “fiducials”, physical markers on the array for positional alignment in both the image and sequencing data. Subsequent steps are consistent with standard spatial transcriptom ics processes. Ultimately, the inventors adapt this spatial scRNA-seq process to other types of RNAs, including sRNA or cleaved, uncapped mRNA, or even mixed host / pathogen (eukaryote / prokaryote) that would be hard to capture with array surface chemistry for spatial analysis but are facile in solution for a single-cell approach.

[0188] To summarize the approach: samples (whole duckweed) would be fixed, cryo-sectioned, stained, transferred onto the array, and imaged. Live plant tissue is preinfiltrated or post-stained to define plant cell walls, assisting with automated postalignment steps (i.e. for correlation from sections to spatial RNA data to whole plants from XRM imaging) and then frozen plant tissue is embedded before longitudinally cryosectioning. In parallel, the possibility to fix whole Wolffia specimens stained with XRM contrasting agents is tested and 3D XRM volumes are collected, followed by sectioning and other array steps. This serves as a ground truth and specific reference volume for each corresponding tissue array dataset for subsequent integration. A series of fixation / sectioning conditions are tested to confirm optimal retention of transcripts, sectioning properties and morphology for arrays. As the smallest duckweeds in size, Wolffia sp. Is used for initial efforts to serial section the entire plant. Occasional sections or partial sections can be lost, wrinkled or torn. Since each section is significantly thinner than an entire plant cell (or meristem), spatial transcript mapping is not impaired, and any missing data corrected or interpolated. T he plant tissue samplesPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center are fixed onto the MAS arrays and stained to preserve and visualize cell morphological structures.

[0189] Stained tissue sections are then converted into a map of cells that are overlaid by the array spots. The fiducials mentioned above allow correlation of the RNA-seq data to the microscopy image by positions relative to the fiducials; additionally, the outline or edge of the arrays spots that yield RNA reads is matched to sectioned tissue areas. A variety of solutions and strategies allow the integration of most images in a multi-scale, multi-modal workspace. For example, ec-CLEM, an open source plugin and 2D / 3D correlative solution, can be applied. Correlated serial section images are processed for robust automated section alignment and pseudo-3D volumetric datasets are created. However, volumes reconstructed from serial sections can suffer from significant tissue loss, distortion, and other artifacts that compromise analysis of the actual 3D structure, which is addressed through combined XRM / vEM in Example herein below. These initial experiments use a relatively simple 2D section of the plant sections (see below) that are imaged by light / brightfield microscopy; when the RNA-seq data are obtained. These are overlayed using a virtual grid of 10 pm spots, using the tissue and data edges to correlate the data with cells and tissue features. Software can be used to facilitate registration of image and transcriptional data, which is developed as needed.

[0190] After imaging on the array surface, the tissue is treated directly on the glass slides to liberate the RNA followed by reverse transcription (“RT”) to generate cDNAs that are barcode-tagged and collected for RNA-seq library construction. A standard spatial transcriptom ics protocol is then performed.Example 4. Increased resolution of spatial transcriptomics by cellular expansion

[0191] ExM is a technique which infiltrates an expandable polyelectrolyte gel matrix into a sample; after cross-linking nucleotides and proteins to the matrix, the matrix is expanded to “grow” the sample by around to 4- to 5-fold in a linear direction, for example growing a 70 pm section to 300 pm 14. It is used to get around thePCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center diffraction limit in microscopy by expanding a sample for imaging; we propose to use the same technique to expand plant sections for spatial transcriptom ics. Protocols for expansion microscopy in plants have been developed using an enzymatic mix to break down the cell wall after fixation. Commercial reagents that cross-link RNA to the expansion gel can keep RNA in place during sample preparation. Use of ExM with RNA allows expansion of the plant tissues substantially with RNAs fixed in relative 3D positions. For traditional spatial transcriptom ics, one challenge to using this method is how to liberate the RNA for diffusion onto the array after expansion. However, the methods described in the instant examples are preferable, allowing RNA release from the crosslink after sorting as single cells, in solution. This method is beneficial for the duckweed meristem as it is particularly small and poorly described.Example 5. Spatial analysis of duckweed

[0192] This approach combines the optimized methods of the approaches above and applies the techniques to entire (but sectioned) duckweed plants, assembling and refining the methods for spatial transcriptom ics of an entire organism. More than five biological replicates are used, a high degree of replication, to avoid issues of noise from variation in growth and sectioning. Further, since a duckweed plant itself contains fronds at different stages of development, there will be replication of a developmental series. To get more replication over a more continuous distribution of developmental progression, ten or more full replicates can be needed. Wolffia microscopica is grown on six-well plates, and entire plants processed for 3D spatial scRNA-seq.Example 6. Data analysis of spatial transcriptomic data

[0193] Data analysis comprises two steps: the first step is processing of the raw data, which for each library includes (1 ) standard RNA-seq clean up to remove sequenced adapters, (2) deconvolution of the reads to specific array spots (X and Y coordinates on the array), and stripping the corresponding barcode out of the sequence, (3) for each spot on the array, use the Unique Molecular Identifier (UM I) to collapsePCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center reads to remove amplification duplicates, (4) create for this non-redundant, spot-based set of reads a “pseudolibrary” (file of data for that X-Y position on the array). Then, for this pseudolibrary, (5) map the reads to a cell or plant location based on the XRM / optical microscopy of the section to assign it a physical location in the plant tissue, and (6) map the reads to the reference genome, sum by genes, and normalize by total spot reads per physical plant location to generate an abundance list of genes. Once each 2D slice is analyzed, a Z-stack of 3D abundances is created by overlaying these 2D slices and assigning X-Y-Z coordinates with corresponding XRM ground truth datasets (See examples above and below) within the plant tissue.Example 7. X-ray microscopy and vEM for 3D volumes of entire duckweed plants at cellular / subcellular resolution

[0194] In this example and Example 7, benchtop X-ray microscopy is used to build whole-organism models at a single-cell and vEM (specifically, serial block-face scanning electron microscopy (SBF-SEM) for sub-cellular resolution), to be used to integrate the spatial transcriptom ic data into 3D models.

[0195] XRM is a powerful tool for high-quality, high-resolution, non-destructive morphometric measurements, including quantifying cell shape and tissue organization, but its utility for whole-organism histotomography has only been demonstrated at national synchrotrons, greatly restricting its potential widespread adoption. In this example, a benchtop ZEISS Xradia 520 Versa XRM instrument that provides submicron 3D imaging resolution (down to 200 nm) over large fields of view (at least 5 cubic mm), is used for this purpose, as recently demonstrated in principle in plant systems. Likewise, vEM extends the resolution range of XRM down to nm scale, all EM contrasted cell structures are visible and enhanced heavy metal staining and resin embedment for vEM improves overall XRM contrast and resolution. Here, intact 3D reconstructions of an entire organism are generated with each cell segmented (3D histotomography), to develop general multi-scale developmental models. These models are used in examples below as frames to map spatial transcriptom ic data. ThePCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center combination of data provides a novel and complete picture of the organism by combining morphological and transcriptom ic information to enable understanding gene function at a single-cell level. A question of particular interest is the level of variation in cellular patterning among simple, clonal, organisms with very regular macromorphological features.

[0196] In this exampe, the XRM data yields (1 ) a model for integrating 3D transcriptom ics, (2) a tool to aid in developmental assessment of duckweed, and (3) a ground truth reference for correlative cryo and vibratome sections. Analyses combining correlative light and electron microscopy (CLEM) have been performed previously while other work has developed methods to reconstruct 3D structures from 2D serial-section ribbons. However, these volumes are distorted along the Z-axis (which is a different resolution than X-Y) and are physically sliced. In contrast, 3D reconstructions from X- ray tomography and correlated vEM using SBF-SEM with an in situ microtome to image a rigid block surface are complementary, isometric, arise from intact / whole-mount specimens, and as described subsequently, sample preparation procedures can highlight different facets of the cell.

[0197] For imaging, plants are prepared using a range of fixation and contrasting protocols with the aim of achieving the highest scan resolution while maintaining sample compatibility with spatial transcriptom ic sample preparation methods. Exact XRM scanning conditions are optimized depending on specific resolution, selected objective lens, and image analysis requirements. The XRM has multiscale imaging capability, allowing entire samples to be scanned at low magnification to identify regions-of- interest, which are then scanned at high magnification without removing the sample from the instrument. When heavy metal-stained resin blocks are used, select samples are then transferred for 3D SBF-SEM acquisition at 100 nm, 50 nm and / or 5-10 nm pixel resolutions. Resulting 3D volume data are integrated and analyzed directly for identification and measurement of specific cell types (computational segmentation).

[0198] A feature of the Wolffia model is the inherent developmental progression whereby the “mother frond” contains an apparently self-similar somatic progeny, thePCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center“daughter frond”, that emerges from the apical meristem (FIG. 4). The relative size of the mother and daughter fronds are used to stage the duckweeds along a developmental time series prior to developing the generalized 3D shape models described in the next section. The available XRM / vEM has a co-registered multi-scale imaging feature that is employed, first scanning the entire larger mother frond, which has expanded and much larger cells, at lower magnification, before zooming in to the daughter frond at higher magnification. This is used as an internal developmental calibration within samples, as well as across samples.Example 8. Analysis of XRM / vEM data to construct 3D volumetric models

[0199] Cell segmentation is conducted through various approaches depending on the quality of the image; therefore, this work is closely integrated with sample preparation and XRM scanning. For high quality images and cell wall signals that have high contrast, the inventors start from some of the most effective and easiest implemented methods such as thresholding based approaches and watershed segmentation that have been previously shown to work well for XRM data. However, the goal is to develop semi / automated approaches that consistently produce accurate cell segmentation for tens or hundreds of samples. Widely used deep learning methods such as 3D UNet or other convolutional neural networks are customized or used to identify cell boundaries.

[0200] A 3D volumetric model is constructed from XRM / vEM that represents the average profile of the organs and cells. For a static model from biologically replicated plants of the similar stage, this process comprises: (1 ) Compute an eigen contour of the whole organ or region of interest among the population. A Wolffia plant usually has a mother frond and a daughter frond emerging from it. All the organ shapes are registered, for example, based on the position of the daughter frond and then advanced morphometries are employed such as Elliptical Fourier Descriptor to measure mother frond and daughter frond contours by sequence of harmonics. The mean harmonics are then converted back to a 3D surface as an eigen contour. (2) Quantify cell profile.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent CenterCells are be identified individually after cell segmentation (above). Cell profiles are measured from many aspects such as cell counts, size, shape, phyllotaxis, position, distribution. Cell-cell connection networks are also built, recording cell organization information. (3) Classify cells into subdomains. The subdomain is trained and partitioned based on cell counts, geometrical position, morphology, neighbor graph topology, and biological function, and then incorporated with the cell types that are assigned through the patterns of gene expression (See example above) parsing the overlaid spatial transcriptome data cross subdomains. (4) Average cell profiles for each subdomain. Numerical features such as cell counts, size, shape could be easily averaged by taking their mean values. For comprehensive representation such as cellcell networks, registration methods are employed such as simple affine alignment or NP-hard graph matching methods to develop an average graph. This will enable calculation of the mean cell shape per node. (5) Fill average subdomain cells into eigen contour. A model is initiated by setting the average network of every subdomain as the backbone and filling cells at every node with the cell mean shape. Cell overlaps, gaps, and many other unreasonable cell organization outcomes are anticipated in the initial model. The model is then iteratively modified by changing the cell position and shape within the allowable variance until converging to a model that lacks overlaps and gaps. For a dynamic model that represents a developmental time series, the inventors start a static model from an early stage. Then, the cell profiles of the next stage are not only the average from plants at a fixed stage, but also incorporate statistical learning from a dynamic analysis for the growth trajectory from the time series data.Example 9. Analysis of spatial transcriptom ic data in 3D space

[0201] For the inference of biologically-relevant differences in spatial expression levels, approaches described for spatial transcriptom ics are initially used while integrating with the XRM / vEM data obtained in the examples above. There are now a growing number of software packages that integrate image data and spatial transcriptom ic data, most notably Squidpy. The primary aims from these experimentsPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center are to (1 ) generate high-resolution expression data, in this case single-cell, for individual genes across a complex organ or tissue, and (2) distinguish and assign unique cell types and developmental stages of those types based on patterns of gene expression in single cells. There are challenges that come with the analysis of single-cell RNA-seq data such as (1 ) dealing with technical noise and biological noise (cell-to-cell variation due to transcript isoform expression), (2) imputing missing data due to cell expression dropout by clustering across cell types based on spatial positions, and (3) visualization of weak gene expression correlation profiles. The computational approaches to address these challenges are by first applying standard normalization and scaling over the distribution of reads per cell to avoid overrepresentation of highly expressed transcripts. Highly variable patterns in gene expression are identified and used for Principal Component Analysis (PCA) to assess dimensionality. Principle components that capture high signal to noise ratios are used to directly inform downstream clustering approaches that expand upon k-nearest neighbors (KNN) algorithm to make use of graph embedding, where the weights assigned to edges connecting any given pair of cells are refined by the degree of shared overlap (Jaccard similarity index) between local neighborhoods embedded within the graph. Final cell clustering is achieved using community detection based on modularity optimization, where the goal of the algorithm is to maximize the difference between the observed and expected number of edges that interconnect nodes within the graph. Non-linear dimensionality reduction approaches such as t-SNE or UMAP are employed to visualize the arrangement and separability between clusters of similar cells in low-dimensional space, and to help identify biologically relevant differences in transcriptional programming between cell types of interest.Example 10. Integration of spatial transcriptomic and XRM / vEM data

[0202] For each set of materials from the examples above, analyzed by spatial transcriptom ics, the Z-stacks of transcriptional data mapped to physical sections are then overlayed, and / or their imperfect 3D reconstructions for an entire plant onto thePCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent CenterXRM / vEM-generated 3D framework using SIFT or other feature detection and alignment algorithms. Constrained conditions are also incorporated through some object linking methods to not only register the image slice, but also preserve the slice order. A cost function that measures the matching distance traversing all slices are defined and an optimal solution returns one of the best alignments between transcriptional data and 3D volumetric model.

[0203] To code each cell type, both with respect to the type and the X-Y-Z position relative to both cell- and tissue-level anatomical “markers”; correlative microscopy of the section images are combined with XRM datasets. Next is the alignment of sectioned tissues (above); the linear SIFT algorithm can originally be used to align one replicate of an entire plant with another replicate of an entire plant. One solution can be to take plants that are at almost the same stage (judged by size of mother and daughter fronds), and then computationally separate out subdomains / regions within each organ based on cell counts, size, phyllotaxis, etc. and overlay the spatial data from these subdomains. With this single-cell analysis, analyses for local (single cell) and broader (regions of the tissue) trends in gene expression can be formed that are currently routinely missed due to the use of whole organs. On a ‘per gene’ basis, this would look like a 3D heatmap of abundance levels, using quantitative analysis to identify statistically robust and unique patterns of expression.Example 11. Validation of gene expression by smFISH

[0204] From the spatial transcriptom ic analysis, numerous candidate genes and sRNAs that may play a role in development are identified, demonstrating complex patterns of expression over developmental time and across cell types. Identifying and comparing gene activity in 3D space (integrating with imaging data from Approach 2) helps to dissect complex gene networks. Demonstration that the techniques have yielded useful data require validation. In this component of the work, the expression of sets of mRNAs and sRNAs is validated by single molecule FISH (smFISH) analysis, providing high sensitivity and subcellular resolution of transcripts. smFISH remains thePCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center gold standard for sensitive RNA detection by microscopy. It has single molecule sensitivity and detects nearly 100% of transcripts fixed in samples.

[0205] About 20 genes and ~20 sRNAs (microRNAs, probably, as they are key regulators of gene expression involved in many aspects of development) are chosen from key pathways identified in examples herein above. A set of candidates from the spatial scRNA-seq analysis that have moderate expression levels are selected since they are easier to visualize, but distinctive patterns of expression within the analyzed tissues. smFISH methods that are easily applied to duckweed are used and clearing methods for FISH optimized for animal tissues to remove autofluorescence from proteins. Those methods are adapted to plant tissue and modified to remove the cell wall. Expansion microscopy can be used to separate the FISH spots to examine higher expressed RNA and to reduce the error rate.

Claims

PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent CenterCLAIMSWhat is claimed is:1 . A method of spatially mapping gene expression in a tissue sample, the method comprising: a. providing or having provided a high-density, high-resolution nucleic acid microarray, wherein the microarray comprises: i. a solid support comprising a flat surface, the flat surface comprising a two-dimensional (2D) grid of a plurality of polygons, wherein each polygon is assigned a specific coordinate in the grid; and ii. a plurality of spatially tagged oligonucleotides releasably affixed to the flat surface within each polygon, wherein all spatially tagged oligonucleotides comprise a spatial barcode that associates with a coordinate of the polygon; b. contacting cells of a tissue sample with the surface of the nucleic acid microarray; c. diffusing the free spatially tagged oligonucleotides into cells in contact with the oligonucleotides on the nucleic acid microarray to thereby generate tagged cells; d. dissociating the tissue sample to form a plurality of cells; e. performing RNA sequencing in the dissociated tissue sample to generate an expression profile for each tagged cell, wherein the RNA sequencing further sequences the spatially tagged oligonucleotide in the tagged cell; f. spatially mapping the expression profile for each tagged cell across the tissue sample based on the identity of the nucleic acid tag sequenced in (e) to thereby generate a gene expression map for the tissue sample; wherein the method achieves single-cell resolution.

2. The method of claim 1 , wherein the plurality of spatially tagged oligonucleotides are releasably affixed to the flat surface within each polygon via a cleavable moiety and wherein diffusing the free spatially tagged oligonucleotides into cells comprisesPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center cleaving the cleavable moiety in each spatially tagged oligonucleotide to form free spatially tagged oligonucleotides.

3. The method of claim 1 or claim 2, wherein RNA sequencing is single cell RNA sequencing (scRNA-seq).

4. The method of any of the preceding claims, wherein each polygon of the microarray comprises a surface area equal to or smaller than a surface area of a single cell in a tissue sample.

5. The method of any of the preceding claims, wherein each polygon comprises a surface area of about 10 pm2or smaller.

6. The method of any one of the preceding claims, wherein the polygons are delineated by spaces not exceeding about 0.8 micron in width.

7. The method of any one of any one of the preceding claims, wherein the microarray comprises a grid of 10 pm2polygons, wherein the polygons are distributed such that the center-to-center distance between adjacent polygons is no more than about 10.8 pm.

8. The method of any one of any one of the preceding claims, wherein the polygons is a square.

9. The method of any one of any one of the preceding claims, wherein the microarray comprises a grid of polygons, wherein each polygons comprises a surface area of about 10 pm2or smaller.

10. The method of any one of any one of the preceding claims, wherein the microarray comprises a grid of 10 pm2polygons, wherein the polygons are distributed such that the center-to-center distance between adjacent squares is no more than about 10.8 pm.11 . The method of any one of any one of the preceding claims, wherein the microarray comprises a grid of about 2,075,600 10 pm2polygons, wherein the polygons are distributed such that the center-to-center distance between adjacent polygons is no more than about 10.8 pm.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center12. The method of any one of any one of the preceding claims, wherein all spatially tagged oligonucleotides in each polygon comprise the same spatial barcode.

13. The method of any one of the preceding claims, wherein all spatially tagged oligonucleotides within a polygon further comprise a unique molecular identifier (UMI).

14. The method of any one of any one of the preceding claims, wherein each spatially tagged oligonucleotide consists essentially of a spatial barcode, a UMI, and a cleavable moiety.

15. The microarray of any one of any one of the preceding claims, wherein the spatially tagged oligonucleotides are RNA oligonucleotides.

16. The method of any one of any one of the preceding claims, wherein each spatially tagged oligonucleotide is an RNA oligonucleotide, wherein each RNA oligonucleotide consists essentially of: a. a spatial barcode about 12 nucleotides (nt) in length; b. a UMI about 6-8 nt in length, and c. a cleavable moiety.

17. The method of any one of any one of the preceding claims, wherein the cleavable moiety is a chemically- or photo-labile cleavable moiety.

18. The method of any one of any one of the preceding claims, wherein the oligonucleotides of the microarray are synthesized using micromirrors for maskless photolithography (maskless array synthesizer (MAS) arrays).

19. The method of any one of the preceding claims, wherein the microarray further comprises positional markers.

20. The method of any one of the preceding claims, wherein only one cell in the tissue sample contacts each polygon.21 . The method of any one of the preceding claims, wherein each cell contacted by the nucleic acid microarray comprises at least one spatially tagged oligonucleotide.

22. The method of any one of the preceding claims, wherein step (c) comprises diffusing the spatially tagged oligonucleotides into the tissue sample using electrophoresis.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center23. The method of any one of the preceding claims, wherein the RNA sequencing in (e) comprises sequencing coding and noncoding RNAs in each cell.

24. The method of claim 23, wherein the noncoding RNAs are selected from small RNAs (sRNA), circular RNAs (circRNAs), cleaved or uncapped mRNAs, and nonpolyadenylated RNAs, long-non-coding RNAs cleaved or uncapped mRNAs, nonpolyadenylated RNAs, mixed host / pathogen (eukaryote / prokaryote), and any combination thereof.

25. The method of any one of the preceding claims, wherein each spatially tagged oligonucleotide is an RNA oligonucleotide comprising about 26 nt, in the range captured by sRNA library protocols, but not a biologically relevant length.

26. The method of any one of the preceding claims, wherein the tissue sample is expanded before contacting cells of the tissue sample with the nucleic acid microarray.

27. The method of claim 26, wherein the tissue sample is expanded using methods used in expansion microscopy.

28. The method of any one of claims 26 or 27, wherein the tissue sample is expanded via injection of an expandable polyelectrolyte gel matrix into the tissue sample.

29. The method of claim 28 wherein nucleic acids and proteins in the tissue sample are cross-linked to the expandable polyelectrolyte gel matrix before expansion.

30. The method of any one of claims 26 to 29, wherein the tissue sample is linearly expanded 4 to 5-fold.31 . The method of any one of the preceding claims, wherein the tissue sample is a plurality of tissue sections sectioned from the tissue sample.

32. The method of any one of the preceding claims, wherein the tissue sample contacting the microarray has a depth of 1 to 3 cells.

33. The method of any one of the preceding claims, wherein the tissue sample is fixed and sectioned into a plurality of tissue sections prior to contacting the array.

34. The method of any one of the preceding claims, wherein the tissue or tissue section is stained prior to (pre-infiltrated) or after contacting the array.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center35. The method of any one of the preceding claims, wherein the tissue sample is fixed, cryo-sectioned into a plurality of tissue sections prior to contacting the array, then stained.

36. The method of any of the preceding claims, further comprising imaging the tissue sample before step (c).

37. The method of any of the preceding claims, wherein the method comprises imaging the tissue sample on the array.

38. The method of any one of the preceding claims, wherein the method comprises fixing the tissue sample, cryo-sectioning into a plurality of tissue sections, staining, transferring onto the array, and imaging.

39. The method of any of claim 36 to claim 38, wherein the imaging correlates each cell on the array with a coordinate polygon.

40. The method of claim 36 to claim 38, wherein spatially mapping RNA sequence data in the tissue sample comprises matching each cell in the tissue sample to the RNA sequence data based on the coordinate of the cell and the sequence of the oligonucleotide spatial tag in the sequenced cell.41 . The method of claim 36 to claim 40, wherein spatially mapping RNA sequence data in the tissue sample comprises matching each cell in the tissue sample to the RNA sequence data based on the coordinate of the cell and the sequence of the oligonucleotide spatial tag in the sequenced cell by: a. converting the stained tissue sample into a map of cells that are overlaid by the array polygons, the positional markers, or both; and b. correlating the RNA sequence data to the microscopy image by positions relative to the array polygons, the positional markers, or both.

42. The method of any of the preceding claims, further comprising integrating the method with array-based barcoding of tissue sections.

43. The method of any of the preceding claims, wherein the tissue sample is a plurality of tissue sections sectioned from the tissue sample and spatially mapped RNA sequence data in each section of the plurality of tissue sections is assembled into aPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent CenterZ-stack of RNA sequence data to thereby generate a reconstructed 3-dimentional (3D) RNA sequence map.

44. The method of claim 43, further comprising imaging each of the plurality of tissue sections and combining the imaging data with each gene expression map for each section to form a three-dimensional imaging and gene expression map for the tissue.

45. The method of any of the preceding claims, further comprising building a wholetissue 3D volumetric model of the tissue sample at a single-cell level before step (b) or of a different sample of the tissue of interest, wherein the 3D model serves as a ground truth and specific reference volume for each corresponding tissue array dataset for subsequent integration.

46. The method of any of claim 45, wherein building a whole-tissue 3D volumetric model comprises imaging via X-Ray microscopy (XRM).

47. The method of claim 45, wherein building a whole-tissue 3D model at a single-cell and sub-cellular level comprises combining imaging via XRM with volumetric electron microscopy (vEM) to capture sub-cellular resolution in the 3D model of the tissue.

48. The method of claim 46, wherein vEM is serial block-face scanning electron microscopy (SBF-SEM).

49. The method of claim 42 to claim 48, further comprising integrating the spatially mapped expression profile for each tagged cell in the 3D RNA sequence map across the tissue with the whole-tissue 3D model.

50. The method of claim 42 to claim 48, wherein the method further comprises: a. building a whole-tissue 3D model at a single-cell and sub-cellular level of the tissue sample or of a different sample of the tissue by combining imaging via XRM with volumetric electron microscopy (vEM); b. fixing the tissue sample before step (b) and before or after building the whole-tissue 3D model of the tissue sample, cryo-sectioning the tissue sample into a plurality of tissue sections, staining the tissue sections, and transferring the tissue sections onto the array;PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center c. imaging the tissue sections on the microarray before step (c); d. spatially mapping the expression profile for each tagged cell across each tissue section by matching each cell in the tissue section to the sequence data based on the coordinate of the cell and the sequence of the oligonucleotide spatial tag in the sequenced cell by: i. converting the stained tissue sections into a map of cells that are overlaid by the array spots, the positional markers or both; and ii. correlating the RNA-seq data to the microscopy image by positions relative to the positional markers; e. assembling the sequence data of the plurality of tissue sections into a Z- stack of RNA sequence data to thereby generate a reconstructed 3D RNA sequence map; and f. integrating the spatially mapped expression profile for each tagged cell in the 3D RNA sequence map across the tissue with the whole-tissue 3D model.

51. The method of any of the preceding claims, wherein the tissue sample is an animal or plant tissue sample.

52. The method of any of the preceding claims, wherein the tissue sample is a whole Wolffia sp. plant.

53. A method of spatially mapping gene expression across three dimensions in a tissue sample, the method comprising: a. fixing and cryo-sectioning the tissue sample into a plurality of tissue sections that are 1 -3 cells deep; b. spatially mapping RNA sequence data in each section according to the method of any one of claims 1 to 43 to generate a set of gene expression maps for each section; and c. assembling the gene expression maps for each section into a Z-stack of RNA sequence data to thereby generate a reconstructed 3D RNA sequence map.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center54. The method of claim 54, further comprising imaging each tissue section and combining the imaging data with each gene expression map for each section to form a three-dimensional imaging and gene expression map for the tissue.

55. The method of claim 53 or claim 54, further comprising building a whole-tissue 3D volumetric model at a single-cell level of the tissue sample before step (b) or of a different sample of the tissue of interest according to the method of any one of claims 44 to 48, wherein the 3D model serves as a ground truth and specific reference volume for each corresponding tissue array dataset for subsequent integration.

56. The method of any one of claims 53 to 55, wherein the tissue is an animal or plant tissue.

57. The method of any one of claims 53 to 56, wherein the tissue is a plant tissue.

58. The method of any one of claims 53 to 57, wherein the tissue sample is a whole Wolffia sp. plant.

59. A high-density, high-resolution nucleic acid microarray for positionally mapping gene expression in a tissue sample, the microarray comprising: a. a solid support comprising a flat surface, the flat surface comprising a two- dimensional (2D) grid of a plurality of polygons, wherein each polygon is assigned a specific coordinate in the grid; and b. a plurality of spatially tagged oligonucleotides affixed to the flat surface within each polygon, wherein all spatially tagged oligonucleotides comprise a spatial barcode that associates with a coordinate of the polygon.

60. The microarray of claim 59, wherein the plurality of spatially tagged oligonucleotides are affixed to the flat surface within each polygon via a cleavable moiety and wherein diffusing the free spatially tagged oligonucleotides into cells comprises cleaving the cleavable moiety in each spatially tagged oligonucleotide to form free spatially tagged oligonucleotidesPCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center61 . The microarray of claim 59 or 60, wherein the microarray further comprises positional markers.

62. The microarray of any one of the preceding claims, wherein each polygon comprises a surface area equal to or smaller than a surface area of a single cell in a tissue.

63. The microarray of any one of the preceding claims, wherein each polygon comprises a surface area of about 10 pm2or smaller.

64. The microarray of any one of the preceding claims, wherein the polygons are delineated by spaces not exceeding about 0.8 micron in width.

65. The microarray of any one of any one of the preceding claims, wherein the microarray comprises a grid of 10 pm2polygons, wherein the polygons are distributed such that the center-to-center distance between adjacent polygons is no more than about 10.8 pm.

66. The microarray of any one of any one of the preceding claims, wherein the polygons is a square shaped.

67. The microarray of any one of any one of the preceding claims, wherein the microarray comprises a grid of polygons, wherein each polygons comprises a surface area of about 10 pm2or smaller.

68. The microarray of any one of any one of the preceding claims, wherein the microarray comprises a grid of 10 pm2polygons, wherein the polygons are distributed such that the center-to-center distance between adjacent polygons is no more than about 10.8 pm.

69. The microarray of any one of any one of the preceding claims, wherein the microarray comprises a grid of about 2,075,600 10 pm2polygons polygons comprising, wherein the polygons are distributed such that the center-to-center distance between adjacent polygons is no more than about 10.8 pm.

70. The microarray of any one of any one of the preceding claims, wherein all spatially tagged oligonucleotides in each polygon comprise the same spatial barcode.PCT PATENT APPDanforth Ref. DDPSC0134-401 -PCT Via Patent Center71 . The microarray of any one of any one of the preceding claims, wherein each spatially tagged oligonucleotide consists essentially of a spatial barcode, a unique molecular ID, and a cleavable moiety.

72. The microarray of any one of any one of the preceding claims, wherein the spatially tagged oligonucleotides are RNA oligonucleotides, DNA oligonucleotides, or RNA / DNA oligonucleotides.

73. The microarray of any one of any one of the preceding claims, wherein spatially tagged oligonucleotides are RNA oligonucleotides.

74. The microarray of any one of any one of the preceding claims, wherein each spatially tagged oligonucleotide is an RNA oligonucleotide, wherein each RNA oligonucleotide consists essentially of: a. a spatial barcode about 12 nucleotides (nt) in length; b. a UMI about 6-8 nt in length, and c. a cleavable moiety.

75. The microarray of any one of any one of the preceding claims, wherein the cleavable moiety is a chemically- or photo-labile cleavable moiety.

76. The microarray of any one of any one of the preceding claims, wherein the oligonucleotides of the microarray are synthesized using micromirrors for maskless photolithography (maskless array synthesizer (MAS) arrays).

77. A kit for spatially mapping gene expression in a tissue sample, the kit comprising: a. one or more high-resolution high-density nucleic acid microarray of claims 59-76; and b. optionally, tissue preparation reagents, microscopy reagents, cell growth media, selection media, or any combination thereof.

Images