Compositions and methods for analyzing cells in mammalian tissues

Abstract

Methods of processing multicellular tissue specimens for microscopic imaging for expansion microscopy and processing image datasets obtained therefrom. Cells in the multicellular tissue section are cellularly barcoded using a plurality of epitope tags, wherein the barcoding comprises at least 10, 25, 50, 100, 250, 500, 1000, 2500, 5000, 10000, 25000, or 100,000, or more spectral color combinations, and image datasets are processed using a series of trained neural networks enabling robust segmentation and precise mapping of cellular identity.

Description

El 1BIO-001-PCTCOMPOSITIONS AND METHODS FOR ANALYZING CELLS IN MAMMALIAN TISSUES

[0001] The present application claims the benefit of United States Provisional Application No. 63 / 727,149, filed December 2, 2024, and to United States Provisional Application No. 63 / 864,994, filed August 15, 2025, and to United States Provisional Application No. 63 / 888,123, filed September 25, 2025, each of which is incorporated by reference herein and from each of which priority is claimed.FIELD

[0002] The present disclosure relates to compositions and methods for analyzing cells in a mammalian neural tissue, such as for detecting connectivity between cells in a brain tissue sample.BACKGROUND

[0003] US Patent Application No. 2019 / 0071,666 (Zador et al) discloses “a composition comprising a plurality of labeled neurons, each of which is labeled by an expression construct that encodes a unique barcoded nucleic acid”.

[0004] US Patent Application No. 2020 / 0299340 (Brown et al.,) discloses “a fusion protein comprising a scaffold protein and a series of two or more epitopes, where the distinct epitopes are recognized by distinct antibodies, and where the series of epitopes forms a detectable protein tag.”

[0005] Shen et al., Nature Communications (2020) 11 :4632 | doi.org / 10.1038 / s41467- 020-18422-8 states “we devise a light microscopy approach for connectivity analysis of defined cell types called spectral connectomics. We combine multicolor labeling (Brainbow) of neurons with multi-round immunostaining Expansion Microscopy (miriEx) to simultaneously interrogate morphology, molecular markers, and connectivity in the same brain section.”

[0006] An et al., Society for Neuroscience 2022 Poster discloses development progress towards “a scalable set of protein epitopes that can be safely expressed in combinations in neurons, so that each cell gets a unique combination of epitopes, and can be distinguished during serial staining, imaging, and washing steps.”El 1BIO-001-PCTSUMMARY

[0007] The present disclosure is based, at least in part, on discoveries and findings regarding methods for imaging brain circuit connectivity mapping in mammals that reduce from high error rates and large volume acquisition requirements of many previous methods.

[0008] In some aspects and embodiments of the disclosure a composition, comprising a plurality of different fusion proteins is provided, wherein each fusion protein in the plurality of fusion proteins independently comprises: (a) a fluorescent protein comprising an amino acid sequence at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence selected from the group consisting of SEQ ID NO: 40-43; and (b)one or more peptide epitopes comprising an amino acid sequence selected from the group consisting of SEQ ID NO: 1-31; wherein the composition in total comprises at least 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or all of the peptide epitopes of SEQ ID NO: 1-31.

[0009] In certain aspects and embodiments of the disclosure, provided is a composition comprising a plurality of antibodies, wherein the plurality of antibodies comprises antibodies that in total selectively bind to at least 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or more peptide epitopes comprising the amino acid sequence selected from the group consisting of SEQ ID NO: 1-31.

[0010] In some aspects a composition is provided wherein the composition includes a plurality of viral particles, wherein the plurality of viral particles comprise plurality of nucleic acids encoding different fusion proteins as disclosed elsewhere herein.

[0011] In one aspect, the present application provides a method for detecting connectivity between cells in a brain tissue sample is provided wherein the method comprises: (a) expressing the protein composition and / or the nucleic acid composition of any embodiment in a brain tissue sample; (b) contacting the brain tissue sample with:(i) the antibody composition of any embodiment herein under conditions to promote binding of the antibodies to the peptide epitopes to form detectable antibody-epitope complexes; andEl 1BIO-001-PCT(ii) antibodies selective for synaptic markers under conditions to promote binding of the antibodies to the synaptic markers to form detectable antibody-synaptic marker complexes; and(c) obtaining images of the detectable antibody-epitope complexes and the detectable synaptic markers in the brain sample; and (d) analyzing the images to identify connectivity between cells in the brain tissue sample.

[0012] In one aspect, the present application provides a method of processing a multicellular tissue specimen for microscopic imaging, comprising: embedding the multicellular tissue specimen in a first expandable polymer matrix and expanding the first expandable matrix to isotropically expand the multicellular tissue specimen, thereby providing an expanded multicellular tissue specimen; embedding the expanded multicellular tissue specimen in a stabilizing polymer matrix, thereby providing a stabilized expanded multicellular tissue specimen; sectioning the expanded multicellular tissue specimen to provide one or more multicellular tissue sections; and for one or more of the multicellular tissue sections, individually embedding the multicellular tissue section in a second expandable polymer matrix and expanding the second expandable matrix to isotropically expand the multicellular tissue section, thereby providing an expanded multicellular tissue section, and passivate the expanded multicellular tissue section to provide a processed multicellular tissue sections; wherein a plurality of cells in the multicellular tissue specimen and / or in the multicellular tissue section are cellularly barcoded using a plurality of epitope tags, wherein the barcoding comprises at least 10, 25, 50, 100, 250, 500, 1000, 2500, 5000, 10000, 25000, 100,000, or more spectral color combinations visually discernable by microscopic imaging.El 1BIO-001-PCT

[0013] In certain aspects, the multicellular tissue specimen is a neuron-containing tissue. By way of example, the multicellular tissue specimen may be a central nervous system specimen from a vertebrate.

[0014] In certain embodiments, the multicellular tissue specimen may also be stained with a morphology stain.

[0015] In certain embodiments, one or more of the processed multicellular tissue sections are microscopically imaged to provide one or more multicellular tissue images. Examples of applicable imaging methods include, but are not limited to, bright field microscopy, dark field microscopy, phase contrast microscopy, electron microscopy, fluorescence microscopy, reflection microscopy, interference microscopy and confocal microscopy. By way of example, a plurality of the processed multicellular tissue sections may be serial sections, and the serial sections are microscopically imaged to provide a set of serial multicellular tissue images. In certain embodiments, a three-dimensional reconstruction of all or a portion of the multicellular tissue specimen is prepared from the set of serial multicellular tissue images.

[0016] In certain embodiments, the method further comprises identifying one or more intercellular connections in the one or more multicellular tissue images. By way of example, the one or more intercellular connections comprise one or more neuronal synapses.

[0017] In certain embodiments, the barcoding comprises at least 100,000 spectral color combinations visually discernable by microscopic imaging.

[0018] In one aspect, the present application provides an image acquisition and processing method, comprising: on a computer system,(i) acquiring a raw digital image of a multicellular tissue section using an expansion microscopic imaging method, wherein a plurality of cells in the multicellular tissue section are cellularly barcoded using a plurality of epitope tags, wherein the barcoding comprises at least 10, 25, 50, 100, 250, 500, 1000, 2500, 5000, 10000, 25000, or 100,000, or more spectral color combinations visually discernable in the expansion microscopic imaging method, and wherein the raw digital image is assembled from a setEl IBIO-OOl-PCT of stitched image tiles of the multicellular tissue section obtained from the expansion microscopic imaging method;(ii) repeating step (i) for a plurality of multicellular tissue sections to provide a dataset of raw digital images comprising a plurality of individual raw digital images;(iii) registering the individual raw digital images in the dataset of raw digital images to provide a registered dataset of raw digital images;(iv) using a subset of the individual raw digital images in the dataset of raw digital images, providing a dataset of ground truth sparse residual digital images by, for each individual raw digital image in the subset, displaying the raw digital image on an electronic display and manually entering annotated boundaries of cellular structures identified according to barcode color, averaging barcode color intensity within each annotated boundary to provide a color-averaged digital image, and subtracting the averaged digital image from the raw digital image to provide a ground truth sparse residual digital image;(v) using the raw digital images in the subset and the corresponding ground truth sparse residual digital images to train a 3D convolutional neural network predict boundaries of cellular structures within raw digital images;(vi) providing a dataset of enhanced digital images by applying the trained 3D convolutional neural network to each raw digital image in the dataset of raw digital images to provide a predicted dense residual image corresponding to the raw digital image, and calculating a sum of the raw digital image and the corresponding predicted dense residual image to provide a corresponding to the enhanced digital image;(vii) using a first subset of enhanced digital images in the dataset of enhanced digital images to train a 3D convolutional neural network to predict boundary affinities and local shape descriptors of cellular structures within enhanced digital images, and providing datasets of predicted boundary affinity images and local shape descriptor images byEl 1BIO-001-PCT applying the trained 3D convolutional neural network to each enhanced digital image in the dataset of enhanced digital images to provide a predicted boundary affinity image and local shape descriptor image corresponding to the enhanced digital image;(viii) using a second subset of enhanced digital images in the dataset of enhanced digital images to train a multidimensional convolutional neural network (e.g., a 2-D, 2.5-D, or 3- D CNN) to estimate boundary probabilities of cellular structures within enhanced digital images and providing a database of boundary probability images of cellular structures by applying the trained 3D convolutional neural network to each enhanced digital image in the dataset of enhanced digital images to provide a boundary probability image of cellular structures corresponding to the enhanced digital image;(ix) using a third subset of enhanced digital images in the dataset of enhanced digital images to train a multilayer perceptron to provide a uniform embedding image of cellular structures within enhanced digital images, and applying the trained 3D convolutional neural network to each enhanced digital image in the dataset of enhanced digital images to provide uniform embedded image of cellular structures corresponding to the enhanced digital image;(x) providing a dataset of combined affinity images by calculating a dot product of each uniform embedded image in the dataset of uniform embedded images to provide a pseudoaffinity image corresponding to the uniform embedded image and calculating a product of each pseudoaffinity image within the dataset of pseudoaffinity images with its corresponding predicted boundary affinity image and local shape descriptor image to provide a combined affinity image corresponding to the pseudoaffinity image; and(xi) providing a dataset of segmented images using the dataset of combined affinity images.

[0019] In certain aspects, the multicellular tissue specimen is a neuron-containing tissue. By way of example, the multicellular tissue specimen may be a central nervous system specimen from a vertebrate.

[0020] In certain embodiments, the multicellular tissue specimen may also be stained with a morphology stain.El 1BIO-001-PCT

[0021] In certain embodiments, one or more of the processed multicellular tissue sections are microscopically imaged to provide one or more multicellular tissue images. Examples of applicable imaging methods include, but are not limited to, bright field microscopy, dark field microscopy, phase contrast microscopy, electron microscopy, fluorescence microscopy, reflection microscopy, interference microscopy and confocal microscopy. By way of example, a plurality of the processed multicellular tissue sections may be serial sections, and the serial sections are microscopically imaged to provide a set of serial multicellular tissue images. In certain embodiments, a three-dimensional reconstruction of all or a portion of the multicellular tissue specimen is prepared from the set of serial multicellular tissue images.

[0022] In certain embodiments, the barcoding comprises 10, 25, 50, 100, 250, 500, 1000, 2500, 5000, 10000, 25000, or 100,000, or more spectral color combinations visually discernable in the expansion microscopic imaging method.BRIEF DESCRIPTION OF THE FIGURES

[0023] Fig. 1 : A diagram of an exemplary workflow.

[0024] Fig. 2: Results if the combination of cellular morphology with multiplexed barcode readouts using the exemplary workflow.

[0025] Fig. 3 : A diagram of a second exemplary workflow.

[0026] Fig. 4: A six-channel dataset obtained using the second exemplary workflow.

[0027] Fig. 5: A single plane of data acquired from using the second exemplary workflow.

[0028] Fig. 6: A diagram of a third exemplary workflow.

[0029] Fig. 7A: A morphology stain acquired using the third exemplary workflow.

[0030] Fig. 7B: A barcode stain used as the registration channel acquired using the third exemplary workflow.

[0031] Fig. 7C: A barcode stain at 8* expansion, highlighting improved dendritic spine filling and signal smoothing.El 1BIO-001-PCT

[0032] Fig. 7D: A synaptic marker stain acquired at intermediate expansion and registered back to the full-resolution morphology channel.

[0033] Fig. 8: An overview of image acquisition, stitching and registration of raw digital images of a multicellular neuronal tissue section.

[0034] Fig. 9: A further overview of image acquisition, stitching and registration of raw digital images of a multicellular neuronal tissue section.

[0035] Fig. 10: Ground truth for networks computed by computing average barcodes sparsely.

[0036] Fig. 11 : Residual barcodes calculated as the difference between the average and the raw barcode signals.

[0037] Fig. 12: Dense residual barcodes predicted by an agnostic residual u-net trained on sparse labels.

[0038] Fig. 13: Final enhanced barcodes predicted by adding dense residual barcodes to raw barcodes.

[0039] Fig. 14: Predicted affinities and LSDs (combined losses) from a trained 3d multi-task (separate output heads) u-net using multi-channel enhanced barcodes as input.

[0040] Fig. 15: Barcode probabilities from a trained 5d u-net with increased receptive field using multi-channel enhanced barcodes as input.

[0041] Fig. 16: A multilayer perceptron (MLP) trained to project enhanced barcodes into a higher dimensional (more channels) embedding using multi-channel enhanced barcodes as input.

[0042] Fig. 17: A summarized processing flow.

[0043] Fig. 18: A second summarized processing flow.

[0044] Fig. 19: An example showing how the uniform embedding helps to prevent a merge.El 1BIO-001-PCT

[0045] Fig. 20: An exemplary workflow of how barcode information can be used to detect and correct errors in a segmentation of image data, including merging incorrectly disconnected portions of the same object i.e. automated proofreading.DETAILED DESCRIPTION

[0046] All references cited are herein incorporated by reference in their entirety. Within this application, unless otherwise stated, the techniques utilized may be found in any of several well-known references such as: Molecular Cloning: A Laboratory Manual (Sambrook, et al., 1989, Cold Spring Harbor Laboratory Press), Gene Expression Technology (Methods in Enzymology, Vol. 185, edited by D. Goeddel, 1991. Academic Press, San Diego, CA), “Guide to Protein Purification” in Methods in Enzymology (M.P. Deutshcer, ed., (1990) Academic Press, Inc.); PCR Protocols: A Guide to Methods and Applications (Innis, et al. 1990. Academic Press, San Diego, CA), Culture of Animal Cells: A Manual of Basic Technique, 2ndEd. (R.I. Freshney. 1987. Liss, Inc. New York, NY), Gene Transfer and Expression Protocols, pp. 109-128, ed. E.J. Murray, The Humana Press Inc., Clifton, N.J.), Dang, B. et al. SNAC-tag for sequence-specific chemical protein cleavage. Nat. Methods 16, 319-322 (2019), and the Ambion 1998 Catalog (Ambion, Austin, TX).

[0047] As used herein, the singular forms "a", "an" and "the" include plural referents unless the context clearly dictates otherwise.

[0048] As used herein, the amino acid residues are abbreviated as follows: alanine (Ala; A), asparagine (Asn; N), aspartic acid (Asp; D), arginine (Arg; R), cysteine (Cys; C), glutamic acid (Glu; E), glutamine (Gin; Q), glycine (Gly; G), histidine (His; H), isoleucine (He; I), leucine (Leu; L), lysine (Lys; K), methionine (Met; M), phenylalanine (Phe; F), proline (Pro; P), serine (Ser; S), threonine (Thr; T), tryptophan (Trp; W), tyrosine (Tyr; Y), and valine (Vai; V).

[0049] Any N-terminal methionine residue in any polypeptide of the disclosure may be present or may be deleted.

[0050] As used herein, an ‘antibody” may comprise a full length antibody or an antigen binding fragment thereof. Fragments with antigen-binding activity include, but are not limited to, Fab', F(ab')2, Fab, Fv and rlgG, single chain Fv fragments (scFv),El 1BIO-001-PCT bivalent or bispecific molecules, diabodies, triabodies, and tetrabodies, and single domain molecules such as VH and VL that are capable of specifically binding to an antigen.

[0051] As used herein, an “epitope tag” refers to a protein expressed within a target cell (e.g., from a genetic construct) that comprises a defined epitope sequence recognized by a cognate binding protein (e.g., an antibody directed to the epitope sequence). By way of example, fluorescent proteins such as green fluorescent protein (GFP) are soluble proteins that, when expressed by a neuron, distribute evenly throughout the cytoplasm of the neuron, effectively “filling” the neuronal structure. Such a fluorescent protein or other space filling protein is expressed as a fusion with an epitope sequence (also called a peptide tag sequence herein) such as one of the sequences shown in Table 1 below. These epitope tags can then be “interrogated” using antibodies that recognize and bind to the expressed epitope tags via the epitope sequence (referred to herein as a “cognate” antibody). Antibody that binds to the epitope tag within the neuron may be detected directly (e.g., by means of a labeled primary antibody) or indirectly (e.g., by means of a labeled secondary antibody) in the methods described herein.

[0052] The term “segment” as used herein with regard to an image or volume being processed according to the methods described herein refers to a region of an image or volume (e.g., a voxel) that has been computationally identified as representing a distinct object, structure, or component based on pixel or voxel similarity, spatial continuity, or predefined criteria. In some embodiments, a segment is a cell or part of a cell.

[0053] The term “node” as used herein with regard to an image or volume being processed according to the methods described herein refers to a spatially localized point (3D coordinate) on a graph representation of a segment’s centerline. Similarly, an “edge” as used herein refers to means a connection between two nodes.

[0054] The term “expansion microscopy” as used herein technique that enables resolution imaging of preserved cells and tissues on conventional microscopes via isotropic or nonisotropic physical expansion of the specimens before imaging. Preserved biological samples are embedded within a swellable matrix such as a hydrogel and physically expanded by exposure to an expansion solvent such as water, leading to optically transparent samples which allow for nanoscale resolution and aberration-freeEl 1BIO-001-PCT microscopy imaging on conventional diffraction limited microscopes. See, e.g., Hiimpfer et al., J Cell Sci. 2024 Apr 17;137(7):jcs260765. doi: 10.1242 / jcs.260765.

[0055] As used herein with regard to convolutional neural networks, a 2-D approach considers convolutions across two spatial dimensions (e.g., height and width), while a 3- D approach considers a 3-dimensional volume. A 2.5-D approach considers 3-D data by processing 2D slices independently or in a limited 3D context. In the context of CNNs, "multi-dimensional" typically refers to any CNN that operates on data with two or more dimensions.

[0056] All embodiments of any aspect of the disclosure can be used in combination, unless the context clearly dictates otherwise.

[0057] Current methods are also limited in their ability to acquire biomolecular information together with cellular connectivity in the same sample. Approaches based on genetically labeling (barcoding) single neurons have the potential to substantially reduce error rates and reduce acquisition volumes, since cell segments can be associated via barcodes without directly tracing intervening morphology. A further potential advantage of barcoding is that barcodes can be efficiently detected optically, and thus can be combined with efficient methods for optical biomolecular measurements. However, a primary issue with barcoding is the inability of genetically expressed barcode molecules (e.g. uniquely identifying sets of RNA molecules, fluorescent proteins, epitopes) to traffic to the entirety of the neuron, thus complicating accurately establishing connected cells and cell extensions.

[0058] In some aspects, disclosed herein is a method for analyzing a mammalian tissue, comprising: (a) contacting the mammalian tissue with a plurality of different vectors for expressing multiple different epitope tags in cells of the mammalian tissue, wherein each different vector can encode a different epitope tag, and wherein two or more cells in the mammalian tissue can each express a different combination of different epitope tags; (b) contacting the mammalian tissue with a first plurality of binders recognizing a first subset of the multiple different epitope tags; (c) detecting first signals associated with the first plurality of binders in the mammalian tissue; (d) contacting the mammalian tissue with a second plurality of binders recognizing a second subset of the multiple different epitope tags which can be different from the first subset; (e) detectingEl 1BIO-001-PCT second signals associated with the second plurality of binders in the mammalian tissue; and (f) generating a codeword for each of the two or more cells in the mammalian tissue, wherein the codeword can comprise signal codes corresponding to the presence or absence of the first signals and signal codes corresponding to the presence or absence of the second signals, and the codeword for a particular cell corresponding to the combination of different epitope tags expressed in the cell.

[0059] In some embodiments, the mammalian tissue can be a neural tissue, and the each of two or more cells can be independently selected from the group consisting of a neuron, an oligodendrocyte, an astrocyte, an ependymal cell, a microglia, a Schwann cell, and a satellite cell. In some embodiments, the mammalian tissue can be a brain tissue or a spinal cord tissue, and the two or more cells can be neurons. In any of the embodiments herein, i) the mammalian tissue can be a cell culture comprising cultured neurons, optionally wherein the cell culture can be a patient-derived cell culture; ii) the mammalian tissue can be a cultured tissue, optionally wherein the cultured tissue can be a cultured brain tissue; or iii) the mammalian tissue can be in a live mammalian individual and the plurality of different vectors can be introduced into the individual to contact with the mammalian tissue.

[0060] In any of the embodiments herein, the plurality of different vectors can be viral vectors. In any of the embodiments herein, the plurality of different vectors can be AAV vectors. In any of the embodiments herein, the plurality of different vectors can comprise stoichiometric ratios of the different vectors or non-stoichiometric ratios of the different vectors. In any of the embodiments herein, the plurality of different vectors can comprise more than 10 different vectors. In any of the embodiments herein, the plurality of different vectors can comprise about 30 different vectors. In any of the embodiments herein, the plurality of different vectors can comprise about 100 different vectors.

[0061] In any of the embodiments herein, the multiple different epitope tags can comprise more than 10 different epitope tags. In any of the embodiments herein, the multiple different epitope tags can comprise about 30 different epitope tags. In any of the embodiments herein, the multiple different epitope tags can comprise about 50 different epitope tags. In any of the embodiments herein, the multiple different epitope tags can comprise a peptide epitope recognized by a cognate antibody for interrogation and labeling. In any of the embodiments herein, the multiple different epitope tags can eachEl 1BIO-001-PCT be between about 6 and about 30 amino acid residues in length. In any of the embodiments herein, each different vector can encode a different fusion protein comprising the epitope tag linked to a scaffold protein. In some embodiments, the scaffold protein can be common among the different fusion proteins encoded by the plurality of different vectors, or the scaffold protein can be different among the different fusion proteins encoded by two or more different vectors of the plurality of different vectors. In any of the embodiments herein, the scaffold protein can be a fluorescent protein, optionally wherein the scaffold protein can be an eGFP, a mNeonGreen, mGreenLantern, or a momomeric GFP.

[0062] In any of the embodiments herein, each different vector can encode a different fusion protein comprising the epitope tag linked to a localization domain, optionally wherein the localization domain can be common among the different fusion proteins encoded by the plurality of different vectors, or optionally wherein the localization domain can be different among the different fusion proteins encoded by two or more different vectors of the plurality of different vectors. In any of the embodiments herein, each different vector can comprise a promoter operably linked to a sequence encoding the epitope tag and / or scaffold protein. In some embodiments, the promoter can be a CAG promoter or a Sindbis virus subgenomic promoter.

[0063] In any of the embodiments herein, the mammalian tissue can be the brain of an mammalian individual, wherein: each different epitope tag can be expressed randomly in about 1%, about 5%, about 10%, about 15%, about 20%, about 25%, about 30%, about 35%, about 40%, about 45%, or about 50% of the neurons labeled by the different epitope tags, or each different epitope tag can be expressed randomly in over 50% of the neurons labeled by the different epitope tags, optionally wherein each different epitope tag is expressed randomly in about 55%, about 60%, about 65%, about 70%, about 75%, about 80%, about 85%, about 90%, about 95%, or about 99% of the neurons labeled by the different epitope tags. In any of the embodiments herein, at least or about 10, at least or about 100, at least or about 103, at least or about 104, at least or about 105, at least or about 106, at least or about 107, at least or about 108, at least or about 109, at least or about 1010, or at least or about 1011neurons in the mammalian tissue can each express a unique combination of different epitope tags.El 1BIO-001-PCT

[0064] In any of the embodiments herein, the method can comprise: contacting the mammalian tissue with a third plurality of binders recognizing a third subset of the multiple different epitope tags; detecting third signals associated with the third plurality of binders in the mammalian tissue, wherein the codeword can further comprise signal codes corresponding to the presence or absence of the third signals. In some embodiments, the method can comprise: contacting the mammalian tissue with a fourth plurality of binders recognizing a fourth subset of the multiple different epitope tags; detecting fourth signals associated with the fourth plurality of binders in the mammalian tissue, wherein the codeword can further comprise signal codes corresponding to the presence or absence of the fourth signals. In some embodiments, the method can comprise: contacting the mammalian tissue with a fifth plurality of binders recognizing a fifth subset of the multiple different epitope tags; detecting fifth signals associated with the fifth plurality of binders in the mammalian tissue, wherein the codeword further can comprise signal codes corresponding to the presence or absence of the fifth signals. In any of the embodiments herein, each plurality of binders can comprise binders recognizing two, three, four, five, or more different epitope tags. In some embodiments, the method can comprise: contacting the mammalian tissue with a sixth plurality of binders recognizing a sixth subset of the multiple different epitope tags; detecting sixth signals associated with the sixth plurality of binders in the mammalian tissue, wherein the codeword further can comprise signal codes corresponding to the presence or absence of the sixth signals. In some embodiments, the method can comprise: contacting the mammalian tissue with a seventh plurality of binders recognizing a seventh subset of the multiple different epitope tags; detecting seventh signals associated with the seventh plurality of binders in the mammalian tissue, wherein the codeword further can comprise signal codes corresponding to the presence or absence of the seventh signals. In some embodiments, the method can comprise: contacting the mammalian tissue with a eighth plurality of binders recognizing a eighth subset of the multiple different epitope tags; detecting eighth signals associated with the eighth plurality of binders in the mammalian tissue, wherein the codeword further can comprise signal codes corresponding to the presence or absence of the eighth signals. In some embodiments, the method can comprise: contacting the mammalian tissue with a nineth plurality of binders recognizing a nineth subset of the multiple different epitope tags; detecting nineth signals associated with the nineth plurality of binders in the mammalian tissue, wherein the codeword further can comprise signal codes corresponding to the presence or absence of the ninethEl 1BIO-001-PCT signals. In some embodiments, the method can comprise: contacting the mammalian tissue with a tenth plurality of binders recognizing a tenth subset of the multiple different epitope tags; detecting tenth signals associated with the tenth plurality of binders in the mammalian tissue, wherein the codeword further can comprise signal codes corresponding to the presence or absence of the tenth signals. In some embodiments, the method can comprise: contacting the mammalian tissue with up to a fifteenth plurality of binders recognizing up to a fifteenth subset of the multiple different epitope tags; detecting up to fifteenth signals associated with the up to fifteenth plurality of binders in the mammalian tissue, wherein the codeword further can comprise signal codes corresponding to the presence or absence of the fifteenth signals. In some embodiments, the method can comprise: contacting the mammalian tissue with up to a twentieth plurality of binders recognizing up to a twentieth subset of the multiple different epitope tags; detecting up to twentieth signals associated with the up to twentieth plurality of binders in the mammalian tissue, wherein the codeword further can comprise signal codes corresponding to the presence or absence of the twentieth signals. In some embodiments, the method can comprise: contacting the mammalian tissue with up to a twenty-fifth plurality of binders recognizing up to a twenty -fifth subset of the multiple different epitope tags; detecting up to twenty-fifth signals associated with the up to twenty-fifth plurality of binders in the mammalian tissue, wherein the codeword further can comprise signal codes corresponding to the presence or absence of the twenty-fifth signals. In some embodiments, the method can comprise: contacting the mammalian tissue with up to a fiftieth plurality of binders recognizing up to a fiftieth subset of the multiple different epitope tags; detecting up to fiftieth signals associated with the up to fiftieth plurality of binders in the mammalian tissue, wherein the codeword further can comprise signal codes corresponding to the presence or absence of the fiftieth signals.

[0065] In some embodiments, in each cycle of binder recognition and signal detection, a signal associated with each different epitope tag detected at a particular neuron, or the absence of the signal at the neuron, can be recorded as a signal code at a bit in the codeword for the neuron. In some embodiments, the method can comprise two, three, four, five, or more cycles of binder recognition and signal detection, and in each cycle the plurality of binders can comprise two, three, four, five, or more different binders each recognizing a different epitope tag of the multiple different epitope tags. In some embodiments, the method can comprise five cycles of binder recognition and signalEl 1BIO-001-PCT detection, and in each cycle the plurality of binders can comprise three different binders each recognizing a different epitope tag of the multiple different epitope tags. In some embodiments, the method can comprise two, three, four, five, six, seven, eight, nine, ten, up to fifteen, up to twenty, up to twenty-five, up to fifty, or more cycles of binder recognition and signal detection, and in each cycle the plurality of binders can comprise two, three, four, five, six, seven, eight, nine, ten, up to fifteen, up to twenty, up to twenty- five, up to fifty, or more different binders each recognizing a different epitope tag of the multiple different epitope tags. In some embodiments, the method can comprise five cycles of binder recognition and signal detection, and in each cycle the plurality of binders can comprise three different binders each recognizing a different epitope tag of the multiple different epitope tags. In any of the embodiments herein, in a particular cycle of binder recognition and signal detection, each different binder can be detected in a different channel of fluorescent microscopy.

[0066] In any of the embodiments herein, the codeword for the neuron can be between 2 and 100 bits, optionally wherein the codeword for the neuron can be 15 bits, 30 bits, 50 bits, or 100 bits. In any of the embodiments herein, prior to a particular cycle of binder recognition and signal detection, the method can comprise a step of removing the plurality of binders of a previous cycle from the mammalian tissue, and / or extinguishing signals associated with the plurality of binders of the previous cycle.

[0067] In any of the embodiments herein, each plurality of binders can comprise primary antibodies or epitope-binding fragments thereof that bind to the epitope tags, optionally wherein the primary antibodies or epitope-binding fragments thereof can be detectably labeled. In some embodiments, each plurality of binders can further comprise secondary antibodies or epitope-binding fragments thereof that bind to the primary antibodies or epitope-binding fragments thereof, optionally wherein the secondary antibodies or epitope-binding fragments thereof can be detectably labeled. In any of the embodiments herein, each of the primary antibodies or epitope-binding fragments thereof or the secondary antibodies or epitope-binding fragments thereof can be conjugated to a nucleic acid tag, optionally wherein the nucleic acid tag can comprise one or more barcode sequences.

[0068] In any of the embodiments herein, the mammalian tissue can be a brain tissue and the method can comprise detecting a pre-synaptic marker, a post-synaptic marker,El 1BIO-001-PCT and / or a neurotransmitter marker in the mammalian tissue. In some embodiments, the pre- synaptic marker can be selected from the group consisting of piccolo, bassoon, CASK, one or more SNARE types, SNAP25, VAMP, and syntaxin. In any of the embodiments herein, the post-synaptic marker can be selected from the group consisting of Homer, post-synaptic density-95 (PSD95), neuroligin, SAP 102, SAPAP, SHANK, and calciumdependent protein kinase II. In any of the embodiments herein, the neurotransmitter marker can be selected from the group consisting of a marker for glutamatergic transmission, a marker for GABAergic transmission, a marker for dopaminergic transmission, a marker for cholinergic transmission, and a marker for serotonergic transmission. In any of the embodiments herein, the neurotransmitter marker can be selected from the group consisting of VGAT, GAB RAI, gephyrin, NMDA-1, and vGluTl.

[0069] In any of the embodiments herein, the mammalian tissue can be expanded or not expanded. In any of the embodiments herein, the mammalian tissue can be a brain tissue and the method can comprise generating a plurality of different codewords at cellular structures in the mammalian tissue. In some embodiments, the method can comprise identifying two or more cellular structures having the same codeword as belonging to the same neuron. In some embodiments, the method can identify two or more cellular structures each having a different codeword as belonging to different neurons. In any of the embodiments herein, the cellular structures can be selected from the group consisting of a nucleus or a portion thereof, a cell body or a portion thereof, an axon or a portion thereof, and a dendrite or a portion thereof. In any of the embodiments herein, the method can comprise embedding the brain tissue in a swellable polymer matrix and expanding the swellable polymer matrix and the brain tissue embedded therein.

[0070] In some aspects, disclosed herein is a composition comprising a plurality of different vectors at stoichiometric ratios, wherein each different vector can encode a fusion protein comprising a different epitope tag linked to a scaffold protein, and wherein the plurality of different vectors can be configured to express the fusion proteins in cells of a mammalian tissue. In some embodiments, the plurality of different vectors can be viral vectors. In some embodiments, the plurality of different vectors can be AAV vectors or sindbis virus vectors. In any of the embodiments herein, the composition can compriseEl 1BIO-001-PCT more than 10 different vectors each expressing a fusion protein comprising a different epitope tag, and two or more or all of the different epitope tags can be linked to the same scaffold protein. In some embodiments, the composition can comprise about 30 different vectors each expressing a fusion protein comprising a different epitope tag linked to a common scaffold protein. In any of the embodiments herein, the different epitope tags can comprise peptide tag sequences between about 6 and about 30 amino acid residues in length. In any of the embodiments herein, the scaffold protein can be a fluorescent protein.

[0071] In some aspects, disclosed herein is a plurality of different fusion proteins each comprising a different peptide tag sequence linked to a common scaffold protein, wherein the different peptide tag sequences can be between about 6 and about 30 amino acid residues in length and the common scaffold protein can be a fluorescent protein. In some embodiments, the plurality of different fusion proteins can comprise more than 10 different fusion proteins each comprising a different peptide tag sequence. In some embodiments, the plurality of different fusion proteins can comprise about 30 different fusion proteins each comprising a different peptide tag sequence.

[0072] In some aspects, disclosed herein is a mammalian tissue comprising a plurality of different vectors in contact with cells of the mammalian tissue, wherein each different vector can encode a fusion protein comprising a different epitope tag linked to a scaffold protein, and wherein the plurality of different vectors can be configured to express the fusion proteins in cells of the mammalian tissue. In some embodiments, the mammalian tissue can comprise more than 10 different vectors each encoding a different peptide tag sequence. In some embodiments, the mammalian tissue can comprise about 30 different vectors each encoding a different peptide tag sequence.

[0073] In some aspects, disclosed herein is a mammalian tissue comprising a plurality of different fusion proteins expressed in cells of the mammalian tissue, wherein each different fusion protein can comprise a different peptide tag sequence linked to a scaffold protein, and wherein two or more cells in the mammalian tissue can each express a different combination of peptide tag sequences selected from the plurality of different fusion proteins. In some embodiments, the mammalian tissue can be a brain tissue, and wherein: each different peptide tag sequence of the plurality of different fusion proteins is expressed randomly in about 1%, about 5%, about 10%, about 15%, about 20%, aboutEl 1BIO-001-PCT25%, about 30%, about 35%, about 40%, about 45%, or about 50% of the neurons labeled by the different peptide tag sequences, or each different peptide tag sequence of the plurality of different fusion proteins is expressed randomly in over 50% of the neurons labeled by the different peptide tag sequences, optionally wherein each different peptide tag sequence can be expressed randomly in about 55%, about 60%, about 65%, about 70%, about 75%, about 80%, about 85%, about 90%, about 95%, or about 99% of the neurons labeled by the different peptide tag sequences. In any of the embodiments herein, the mammalian tissue can comprise more than 10 different fusion proteins each comprising a different peptide tag sequence linked to a common scaffold protein expressed in cells of the mammalian tissue. In some embodiments, the mammalian tissue can comprise about 30 different fusion proteins each comprising a different peptide tag sequence linked to a common scaffold protein expressed in cells of the mammalian tissue.

[0074] In some aspects, disclosed herein is a set of binders, comprising: i) a first plurality of binders recognizing a first subset of multiple different epitope tags, wherein each binder in the first plurality can be configured to be detected in a different channel of fluorescent microscopy, and ii) a second plurality of binders recognizing a second subset of the multiple different epitope tags, wherein the second subset can be different from the first subset, and wherein each binder in the second plurality can be configured to be detected in a different channel of fluorescent microscopy, wherein the multiple different epitope tags can comprise peptide tag sequences between about 6 and about 30 amino acid residues in length. The set of binders can further comprise: iii) a third plurality of binders recognizing a third subset of the multiple different epitope tags, wherein each binder in the third plurality can be configured to be detected in a different channel of fluorescent microscopy, iv) a fourth plurality of binders recognizing a fourth subset of the multiple different epitope tags, wherein the second subset can be different from the first subset, and wherein each binder in the fourth plurality can be configured to be detected in a different channel of fluorescent microscopy, and v) a fifth plurality of binders recognizing a fifth subset of multiple different epitope tags, wherein each binder in the fifth plurality can be configured to be detected in a different channel of fluorescent microscopy. In any of the embodiments herein, each subset can comprise two, three, four, five, or more different epitope tags and can be nonoverlapping with another subset. In any of the embodiments herein, each binder can be an antibody or epitope binding fragment thereof.El 1BIO-001-PCT

[0075] In some aspects, disclosed herein is a mammalian tissue comprising a plurality of about 30 different fusion proteins expressed in cells of the mammalian tissue, wherein each of the plurality of fusion proteins can comprise a different epitope tag linked to a common scaffold protein, wherein two or more cells in the mammalian tissue each can express a different combination of epitope tags selected from the plurality of different fusion proteins, wherein the mammalian tissue can be a brain tissue and each different epitope tag of the plurality of different fusion proteins can be expressed randomly in about 50% of the neurons labeled by the different epitope tags, and wherein the mammalian tissue can be in contact with a plurality of binders recognizing a subset of the different epitope tags, wherein each binder in the plurality can be configured to be detected in a different channel of fluorescent microscopy.

[0076] In one aspect, the disclosure provides compositions, comprising a plurality of different fusion proteins, wherein each fusion protein in the plurality of fusion proteins independently comprises:(a) a fluorescent protein comprising an amino acid sequence at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence selected from the group consisting of SEQ ID NO: 40-43; and(b) one or more peptide epitopes comprising an amino acid sequence selected from the group consisting of SEQ ID NO: 1-31; wherein the composition in total comprises at least 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or more, or all of the peptide epitopes of SEQ ID NO: 1-31.

[0077] As used herein, a “plurality” means at least two.

[0078] The sequences of SEQ ID NO: 1-31 are provided in Table 1 below, and the sequences of fluorescent proteins SEQ ID NO:40-43 are provided in Table 2. The inventors have discovered that the compositions of the disclosure are superior to previously available markers for labeling cells in a brain tissue sample to permit mapping connectivity between cells in the brain tissue sample. Specifically, the inventors have surprisingly discovered that the recited fluorescent protein component of the fusion proteins are far superior than previously used detectable protein fusion in filling cells inEl 1BIO-001-PCT the brain tissue sample, including in axons and dendrites many millimeters from the cell body. The inventors have also identified the peptide epitope targets of SEQ ID NO: 1-31, and antibodies detecting them, as ideally suited for detection in cells in brain tissue samples relative to a starting set of approximately 250 peptide epitope-antibody pairs.

[0079] For mapping brain circuit connectivity (“connectomics”), detection of marker combinations (such as the compositions of the disclosure), also referred to herein as “barcoding”, the barcodes need to fill the cell, including in axons and dendrites many millimeters from the cell body. The inventors have demonstrated that the compositions and methods of the disclosure can be used to examine expression in distant areas of the brain given an initial injection site as a proxy for sufficient cell filling, and that the methods result in significantly improved cell-filling labeling of fusion proteins relative to previously available methods, which is required for reducing error rates with barcodes when mapping cell connections. The compositions and methods disclosed herein enable intrinsic error correction, permitting larger brain circuits to be accurately mapped due to fewer errors, and enable “targeted” circuit mapping of specific circuits by bridging spatial gaps (e.g., segments of cells in different brain areas can be connected using barcodes without tracing through the intervening volume).

[0080] Table 1 : Exemplary epitope sequences for use in epitope tagsEl IBIO-OOl-PCTEl 1BIO-001-PCT

[0081] Table 2: Exemplary proteins for use in epitope tagsEl IBIO-OOl-PCTEl IBIO-OOl-PCTEl 1BIO-001-PCT

[0082] In a specific embodiment, the fluorescent protein comprises an amino acid sequence at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 40 (eGFP).

[0083] The fluorescent protein and the one or more peptide epitope may be directly adjacent in the fusion protein, or may be separated by amino acid linkers. In oneEl 1BIO-001-PCT embodiment, the fluorescent protein and the one or more peptide epitope are directly adjacent in each fusion protein in the composition, without any intervening amino acid linker. In other embodiments, 1, 2, 3, 4, 5, or more, or all of the fusion proteins further comprise an amino acid linker separating the fluorescent protein and one or more of the peptide epitope. In embodiments where an amino acid linker is present, the linker may be of any length and amino acid composition as suitable for an intended use. In non-limiting embodiments, the linker may comprise a flexible GS linker, including but not limited to GGSGGS, or a kinked linker comprising one or more prolines, or SEQ ID NO: 1 (ALFA). In another embodiment, the amino acid linkers when present are between 1-12 amino acids in length.

[0084] Individual fusion proteins may comprise a single peptide epitope, or multiple (2, 3, 4, 5, 6, 7, 8, 9, 10, or more) different peptide epitopes. Fusion proteins with more peptide epitope copies have increased signal per protein expressed. In embodiments where the fusion protein comprises multiple different peptide epitopes, amino acid linkers may be present between all of the domains (i.e., between the fluorescent protein and each peptide epitope), linkers may be present between only some of the domains, or the domains may all be directly adjacent with no amino acid linkers separating them. In embodiments wherein linkers are present between multiple domains and there are at least 2 amino acid linkers, the linkers may be the same or may be different.

[0085] In various embodiments, the composition comprises at least 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or more different fusion proteins. In other embodiments, one or more of the fusion proteins comprises 2, 3, 4, 5, or more different peptide epitopes. In further embodiments, 2, 3, 4, 5, or more, or all of the fusion proteins comprises 2, 3, 4, 5, or more different peptide epitopes.

[0086] The domains in the fusion protein may be arranged in any manner appropriate for an intended use. In one embodiment, the fluorescent protein is N-terminal to the one or more peptide epitopes. In another embodiment, the fluorescent protein is C-terminal to the one or more peptide epitopes.

[0087] In another embodiment, 1, 2, 3, 4, 5, or more, or all of the fusion proteins further comprise a localization domain. The addition of localization domains to the fusion protein can increase overall signal in distal regions. Exemplary such localization domainsEl 1BIO-001-PCT include, but are not limited to, membrane localization domains (e.g. a farnesylation motif), ER localization domains, mitochondrial localization domains, and actin localization domains. In various non-limiting embodiments, the localization domain may comprise an amino acid sequence at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence selected from the group consisting of SEQ ID NO:50-54. The amino acid sequence of SEQ ID NO:50-54 are shown in Table 3. In one embodiments, the location of the localization domain in the fusion protein (i.e., N-terminal or C-terminal) is noted in Table 3. In embodiments where there is more than one peptide epitope, the fluorescent protein can be between any two or more peptide epitopes (e.g., peptide epitope — eGFP — peptide epitope, where the peptide epitope can be the same or different). In some embodiments, a fusion protein disclosed herein can comprise two, three, or more epitopes (e.g., short peptides), wherein an epitope can be linked to a scaffold (e.g., eGFP) via a linker, and any two adjacent epitopes can be linked via one or more linkers. In some embodiments, a fusion protein disclosed herein can comprise two or more copies of an epitope fused to the c terminus of a scaffold. In some embodiments, a fusion protein disclosed herein can have the formula eGFP- epitopel-linker-epitope2-linker-epitope3, where epitope 1, epitope2, and epitope3 can be the same epitope or epitopes. In some embodiments, there are multiple copies of a peptide epitope on a single fusion protein, for example, for example, for signal amplification during detection.

[0088] Table 3El 1BIO-001-PCT

[0089] The composition may comprise any number of fusion proteins as appropriate for an intended use. In various non-limiting embodiments, the composition comprises between 2 and 500, or between 2 and 250, or between 2 and 100 fusion proteins.

[0090] In some embodiments, a nucleic acid (e.g., in a viral vector) encoding a fusion protein disclosed herein can comprise the formula promoter-scaffold-epitope-enhancer- polyA.

[0091] In another embodiment, the disclosure provides compositions comprising a plurality of antibodies, wherein the plurality of antibodies comprises antibodies that in total selectively bind to at least 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25 26, 27, 28, 29, 30, or all 31 peptide epitopes comprising the amino acid sequence selected from the group consisting of SEQ ID NO: 1-31.

[0092] The antibody compositions of the disclosure can be used, for example, in the methods of the disclosure to detect expression of the peptide epitopes in cells of the brain tissue sample. As described below, in the methods of the disclosure, individual cells will express multiple peptide epitopes, and their immunohistochemical detection by the antibodies provides a barcode for the cell and its extensions.

[0093] The antibodies in the composition are detectably distinguishable, either by being directly labeled with distinguishable, detectable labels, such as fluorescent dyes or conjugated nucleic acids or by secondary labeling with secondary antibodies that are distinguishable, detectably labeled, or by tertiary labeling of secondary antibodies with tertiary probes that are distinguishable and detectably labeled. For example, 3 separate compositions of antibodies each comprising 5 antibodies can be provided and used. In some embodiments, each antibody in composition 1 is separately distinguishable from the other composition 1 antibodies, but does not necessarily need to be separately distinguishable from composition 2 or 3 antibodies.

[0094] In some embodiments, a method disclosed herein comprises detection of fluorescent dye labeled primary antibody. In some embodiments, a method disclosed herein comprises detection of fluorescent dye labeled secondary antibody. In some embodiments, a method disclosed herein comprises detection of fluorescent dye labeled tertiary probe (e.g., nanobody). In some embodiments, a method disclosed hereinEl 1BIO-001-PCT comprises detection using DNA-conjugated primary or secondary antibody, detected with complementary dye-labeled DNA probe. In some embodiments, a method disclosed herein comprises detection using a DNA-conjugated primary or secondary antibody, DNA amplification reaction (e.g. branched DNA assay, rolling circle amplification, hybridization chain reaction), and detection of amplified DNA with tertiary dye-labeled probe. In some embodiments, a method disclosed herein comprises detection using a Horseradish peroxidase labeled primary or secondary antibody, and tyramide signal amplification. In some embodiments, an antibody is preincubated with a DNA- conjugated nanobody that binds to the antibody, and used for detection in a method disclosed herein. Exemplary methods involving the use of DNA-conjugated nanobodies are described in Unterauer et al., Spatial proteomics in neurons at single-protein resolution, bioRxiv 2023.05.17.541210, incorporated herein by reference in its entirety.

[0095] Antibodies against the peptide epitopes are commercially available, as shown below in Table 4.

[0096] Table 4El IBIO-OOl-PCTEl IBIO-OOl-PCTEl IBIO-OOl-PCTEl 1BIO-001-PCT

[0097] The antibody composition may comprise additional antibodies as appropriate for an intended use. In various non-limiting embodiments, the antibody composition comprises between 2 and 200 antibodies, between 2 and 100 antibodies, or between 2 and 75 antibodies, or between 2 and 50 antibodies, or between 2 and 45 antibodies, or between 2 and 40 antibodies.

[0098] In one embodiment, all antibodies are mixed in the antibody composition. In other embodiments, the composition may comprise multiple (2 or more) separate mixtures, such as multiple mixtures provided in a kit. For example, a first mixture may comprise five antibodies that selectively bind a different peptide epitope selected from the group consisting of SEQ ID NO: 1-5 (i.e.: the first antibody selectively binds to the peptide epitope of SEQ ID NO: 1, the second antibody selectively binds to the peptide epitope of SEQ ID NO:2, etc.), a second mixture may comprise ten antibodies that selectively bind a different peptide epitope selected from the group consisting of SEQ ID NO:6-15, and a third mixture may comprise eight antibodies that selectively bind a different peptide epitope selected from the group consisting of SEQ ID NO: 16-23. It will be clear to those of skill in the art that many such antibody composition mixtures are possible. In these embodiments where the composition comprises multiple mixtures of the antibodies, the antibodies in each individual mixture are detectably distinguishable. In some embodiments where the composition comprises multiple mixtures of the antibodies, the antibodies in different mixtures may be detectably distinguishable, or they may be detectably indistinguishable. The latter embodiment may be used, for example, in methods of the disclosure that involve iterative cycles of immunostaining, fluorescence imaging, and destaining.

[0099] In another aspect, the disclosure provides a composition comprising plurality of nucleic acids encoding different fusion proteins, wherein each nucleic acid in the plurality of nucleic acids encodes a fusion protein that independently comprises:El 1BIO-001-PCT(a) a fluorescent protein comprising an amino acid sequence at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence selected from the group consisting of SEQ ID NO: 40-43; and(b) one or more peptide epitopes comprising an amino acid sequence selected from the group consisting of SEQ ID NO: 1-31; wherein the plurality of nucleic acids in total encodes at least 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or more, or all of the peptide epitopes of SEQ ID NO: 1-31.

[0100] The nucleic acid may be DNA, RNA, or modified versions thereof.

[0101] The composition comprises a plurality of nucleic acids encoding different fusion protein, wherein the fusion proteins can be any as disclosed above for the fusion protein compositions of the disclosure. Thus in one embodiment, the encoded fluorescent protein comprises an amino acid sequence at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 40 (eGFP). In other embodiments, the encoded fluorescent protein and the one or more peptide epitope may be directly adjacent in the encoded fusion protein, or may be separated by amino acid linkers. In one embodiment, the encoded fluorescent protein and the one or more peptide epitope are directly adjacent in each encoded fusion protein in the composition, without any intervening amino acid linker. In other embodiments, 1, 2, 3, 4, 5, or more, or all of the encoded fusion proteins further comprise an encoded amino acid linker separating the encoded fluorescent protein and one or more of the encoded peptide epitopes. In embodiments where an encoded amino acid linker is present, the encoded linker may be of any length and amino acid composition as suitable for an intended use. In non-limiting embodiments, the encoded linker may comprise a flexible GS linker, including but not limited to GGSGGS, or a kinked linker comprising one or more prolines, or SEQ ID NO: 1 (ALFA). In another embodiment, the encoded amino acid linkers when present are between 1-12 amino acids in length.

[0102] The encoded fusion proteins may encode a single peptide epitope, or multiple (2, 3, 4, 5, 6, 7, 8, 9, 10, or more) different peptide epitopes. In embodiments where the encoded fusion protein comprises multiple encoded peptide epitopes, encoded amino acidEl 1BIO-001-PCT linkers may be present between all of the encoded domains (i.e., between the encoded fluorescent protein and each encoded peptide epitope), encoded linkers may be present between only some of the encoded domains, or the encoded domains may all be directly adjacent with no encoded amino acid linkers separating them. In embodiments wherein encoded linkers are present between multiple encoded domains and there are at least 2 encoded amino acid linkers, the encoded linkers may be the same or may be different.

[0103] In various embodiments, the composition comprises at least 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or more nucleic acids encoding different fusion proteins. In other embodiments, one or more of the encoded fusion proteins comprises 2, 3, 4, 5, or more different encoded peptide epitopes. In further embodiments, 2, 3, 4, 5, or more, or all of the encoded fusion proteins comprises 2, 3, 4, 5, or more different encoded peptide epitopes.

[0104] The domains in the encoded fusion protein may be arranged in any manner appropriate for an intended use. In one embodiment, the encoded fluorescent protein is N-terminal to the one or more encoded peptide epitopes. In another embodiment, the encoded fluorescent protein is C-terminal to the one or more encoded peptide epitopes.

[0105] In another embodiment, 1, 2, 3, 4, 5, or more, or all of the encoded fusion proteins further comprise an encoded localization domain. Exemplary such encoded localization domains include, but are not limited to, membrane localization domains (e.g., a famesylation motif), ER localization domains, mitochondrial localization domains, and actin localization domains. In various non-limiting embodiments, the encoded localization domain may comprise an amino acid sequence at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence selected from the group consisting of SEQ ID NO:50-54. In one embodiments, the location of the encoded localization domain in the fusion protein (i.e., N-terminal or C-terminal) is noted in Table 3.

[0106] In one embodiment, each nucleic acid is operatively linked to a promoter. As used herein, “operatively linked” means capable of effecting the expression of the nucleic acid molecules. The promoter need not be contiguous with the nucleic acid sequences, so long as it functions to direct the expression thereof. Any promoter may be used as suitable for an intended purpose. Constitutive promoters drive expression in a largely cellEl 1BIO-001-PCT independent fashion, making them useful for generalized barcoding applications. However, different promoters drive different ranges of expression and are useful for tuning the amount of protein produced i.e. in cells that are poor protein factories (and thus need stronger promoters) or cells that produce too much GFP (and therefore need weaker promoters to reduce toxicity). These promoters are typically readily swappable parts, and include, but are not limited to: CAG, CMV, PGK, Efl -alpha, TRE (tet response element). Cell-type dependent promoters allow for barcoding applications to be applied only to specific cell types, or to use a specific epitope as a marker of cell type (e.g. if a specific tag had a CamKII promoter). Such promoters include: CamKII - excitatory neurons; Synapsin - neuron specific; homeobox Dlx5 / 6 - GABAergic neurons; Drdla - dopaminergic neurons; and GluR - glutamic neurons.

[0107] In one embodiment, the promoter independently (i.e., the promoter may be the same or different in different nucleic acids in the composition) comprises a nucleotide sequence selected from the group consisting of SEQ ID NO: 60-65. The nucleotide sequences of SEQ ID NO:60-65 are provided in Table 5.

[0108] Table 5El IBIO-OOl-PCTEl IBIO-OOl-PCTEl 1BIO-001-PCTincluding but not limited to polyadenylation signals, termination signals, and ribosome binding sites. In one embodiment, each nucleic acid further comprises an enhancer operatively linked to the promoter and to the coding region of the fusion protein. Any enhancer may be used as appropriate for an intended use. In one embodiment, the enhancer comprises the nucleotide sequence selected of SEQ ID NO: 70, as shown in Table 6.

[0110] Table 6El 1BIO-001-PCT

[0111] In another embodiment, the nucleic acids may further comprise recombinase targeting sites, to enable conditional genetic expression. For example, a nucleic acid can be made into a vector (e.g., FLEx vector) by flanking it with recombinase sites to permit expression of the vector only in cells that express the recombinase. Site specific recombinases (e.g. Cre, FlpO, Nigri, Panto, etc) allow for control of the number of cells expressing a signal while keeping the infection rate of the barcoding components at the desired levels. Further, site-specific recombinases allow conditional expression in particular genetically defined cell types. In various embodiments, the encoded fusion protein is flanked on the 5’ and 3’ end with recombinase targeting sites, with the promoter located 5’ to the 5’ flanking region and any enhancer located 3’ to the 3’ flanking region. Any recombinase targeting sites may be used as appropriate for an intended purpose. In non-limiting embodiments, the flanking regions are selected from (a) a 5’ flanking region comprising the nucleotide sequence of SEQ ID NO:80 and a 3’ flanking region comprising the nucleotide sequence of SEQ ID NO:81 (Cre recombinase targeting sites); (b) a 5’ flanking region comprising the nucleotide sequence of SEQ ID NO:82 and a 3’ flanking region comprising the nucleotide sequence of SEQ ID NO:83 (FlpO recombinase targeting sites); and (c) a 5’ flanking region comprising the nucleotide sequence of SEQ ID NO:84 and a 3’ flanking region comprising the nucleotide sequence of SEQ ID NO:85 (oNigri recombinase targeting sites). The sequences of SEQ ID NO:80- 85 are shown in Table 7.

[0112] Table 7: The composition according to table 1, which enables conditional expression using recombinases (e.g., the composition can be sparsified using recombinases). The plasmid is of the form - [promoter]-[left flank]-[scaffold, with epitope]-[right flank]-[posttranscriptional enhancer].El 1BIO-001-PCT

[0113] In one embodiment, the nucleic acids comprise expression vectors. Any expression vector may be used as suitable for an intended purpose. In various embodiments, each nucleic acid comprises an expression vector, wherein the expression vector comprises a viral vector selected from the group consisting of an adenoviral vector, an adeno-associated viral (AAV) vector (including but not limited to, AAV1, AAV2, AAV9, AAV.PHP.eB); a Sindbis viral vector, a rabies viral vector, a yellow fever viral vector, a lentivirus viral vector, and an HSV vector. While AAV vectors are exemplified herein, the other listed vectors provide other benefits:• Sindbis virus is a strong RNA virus with a rapid onset of protein expression. Expressing protein barcodes in a short period of time allows for rapid barcoding of unstable or difficult systems e.g. ex vivo human tissue.• Rabies virus can deliver their payload to a starter cell and to upstream cells (retrograde transport), allowing the mapping of functional connections without direct observation of synapses, or complementing anterograde tracing to confirm putative circuits. For an example of this, see Rabies virus-based barcoded neuroanatomy resolved by single-cell RNA and in situ sequencing. See, e.g., Chen et al., High-Throughput Mapping of Long-Range Neuronal Projection Using In Situ Sequencing, Cell . 2019 Oct 17;179(3):772-786, incorporated herein by reference in its entirety.• Yellow fever vaccine is an anterograde transsynaptic tracer e.g. a starter cell will label post-synaptic partners with the same barcode. This allows projection tracing without directly observing a given synapse. For an example of this, refer to Anterograde transneuronal tracing and genetic control with engineered yellow fever vaccine YFV- 17D. See, e.g., Li et al., Anterograde transneuronal tracing and genetic control withEl 1BIO-001-PCT engineered yellow fever vaccine YFV-17D, Nat Methods . 2021 Dec; 18(12): 1542- 1551, incorporated herein by reference in its entirety.• Lentiviral delivery of barcodes permits random genetic integration and subsequent expression of a larger payload than AAV (up to lOkb). This permits the generation of more complex circuits and enhanced tactics for expression (e.g. loading the lentiviral cassette with multiple GFPs) as well as using more complex cell type specific promoters.• Adenovirus is a larger capacity DNA virus (up to 8.5 kb) that drives rapid strong protein expression. Expressing protein barcodes from such a virus would permit expression in short lived systems (e.g. ex vivo brain tissue).• HSV is a large capacity (wild type genome of 152kb, with a potential payload on the order of 100+ kb) retrograde tracer. Using such a system would permit delivery of multiple exogenous genes and and complex genetic circuits to aid in the process of retrograde tracing.

[0114] In a specific embodiment, each nucleic acid comprises an AAV expression vector, and each encoded fusion protein comprises eGFP.

[0115] The composition may comprise any number of nucleic acids encoding different fusion proteins as appropriate for an intended use. In various non-limiting embodiments, the composition comprises between 2 and 500, or between 2 and 250, or between 2 and 100 nucleic acids encoding different fusion proteins.

[0116] In another aspect, the disclosure provides a composition, comprising a plurality of viral particles, wherein the plurality of viral particles in total comprises the plurality of nucleic acids encoding different fusion proteins of any embodiment or combination of embodiments herein. In this embodiment, the plurality of nucleic acids are packaged in viral particles, which can be use, for example, to carry out the methods for brain circuit mapping disclosed herein. In one embodiment, viral particles comprising nucleic acids encoding different fusion proteins are present in approximately stoichiometric ratios (i.e., + / - 10% of stoichiometric ratio). In another embodiment, viral particles comprising nucleic acids encoding different fusion proteins are present in nonEl 1BIO-001-PCT stoichiometric ratios (i.e., + / - 100% of stoichiometric ratio). In a specific embodiment, each encoded fusion protein comprises eGFP.

[0117] In another aspect, the disclosure provides host cells comprising the composition of any embodiment herein. As disclosed herein, the methods of the disclosure involve expressing the fusion proteins in brain cells to permit brain circuit mapping. In one embodiment, host cells comprise the nucleic acid or viral composition of any embodiment disclosed herein. In one such embodiment, the plurality of nucleic acids are stably integrated into the cell genome. In another embodiment, the plurality of nucleic acids are stable within the nucleus as extrachromosomal DNA. In another embodiment, the plurality of nucleic acids are transiently transfected into the host cell. In one embodiment, the host cell is a mammalian host cell. In another embodiment, the mammalian host cell comprises a neuron, glial cell, or oligodendrocyte.

[0118] The disclosure also provides transgenic mammals, comprising a host cell in which the plurality of nucleic acids are stably integrated into the cell genome. In one embodiment, the transgenic mammal is a transgenic mouse. This comprises a strategy to express unique combinations of epitopes from a genetically integrated locus. This would effectively barcode every neuron (and cell) in the mouse, including both central and peripheral nervous systems.

[0119] The disclosure also provides kits, comprising one or more composition of the disclosure. The kits can be used, for example, in carrying out the methods of the disclosure. In one embodiment, the kit comprises (a) any embodiment of the nucleic acid compositions of the disclosure, and (b) any embodiment of the antibody compositions of the disclosure. In another embodiment, the kit comprises (a) any embodiment of the viral particle compositions of the disclosure, and (b) any embodiment of the antibody compositions of the disclosure. In another embodiment, the kit comprises (a) any embodiment of the host cell compositions of the disclosure, and (b) any embodiment of the antibody compositions of the disclosure.

[0120] In another embodiment of any of these embodiments, the kits further comprise one or more antibodies that detect synaptic markers. The methods of the certain embodiments of the disclosure include detecting the barcode generated in individual cells by expression of the fusion proteins in cells of brain tissue samples, and staining cellsEl 1BIO-001-PCT with antibodies that detect synaptic markers to infer circuit connectivity by spatial colocatization of pre- and post-synaptic markers. Any antibodies detecting synaptic markers may be present in the kits as appropriate for an intended use. In some embodiments, the synaptic markers may comprise antibodies that selectively bind synaptic targets selected from the group consisting of Amphiphysin, Ankyrin G, Bassoon, Dynamin 1 / 2 / 3, Gephyrin, Homer 1, MAP2, Munc 13-1, Parvalbumin, Piccolo, RIM1, Synaptophysin, VAMP2, vGluTl, PSD95, and Shank 2. Antibodies against these synaptic targets are commercially available, as noted in Table 8. In some embodiments, the kits include antibodies against both pre-synaptic and post-synaptic markers; the specificity of exemplary epitopes for pre-synaptic or post-synaptic locations is also provided in Table 8. In other embodiments, the kits comprise at least 2, 3, 4, 5, 6, 7, 8, or more antibodies detecting synaptic markers. The antibodies that selectively bind synaptic markers may be directly labeled with distinguishable, detectable labels, or by secondary labeling with secondary antibodies that are distinguishable, detectably labeled. In some embodiments, the antibodies that selectively bind synaptic markers are detectably distinguishable from antibodies that selectively bind to the protein epitopes. In other embodiments, the synaptic markers are not detectably distinguishable from antibodies that selectively bind to the protein epitopes; the methods of the disclosure comprise iterative immunostaining, and the synaptic marker-selective antibodies may be detected separately from the protein epitope selective antibodies.

[0121] Table 8. Exemplary synaptic markersEl IBIO-OOl-PCTEl 1BIO-001-PCT

[0122] In another aspect, the disclosure provides methods for detecting connectivity between cells in a brain tissue sample, comprising(a) expressing the protein composition and / or the nucleic acid composition of any embodiment in a brain tissue sample;(b) contacting the brain tissue sample with(i) the antibody composition of any embodiment herein under conditions to promote binding of the antibodies to the peptide epitopes to form detectable antibody-epitope complexes; and(ii) antibodies selective for synaptic markers under conditions to promote binding of the antibodies to the synaptic markers to form detectable antibody-synaptic marker complexes;(c) obtaining images of the detectable antibody-epitope complexes and the detectable synaptic markers in the brain sample; and(d) analyzing the images to identify connectivity between cells in the brain tissue sample, e.g., by detecting binding of the antibodies to the peptide epitopes as a barcode, wherein all cell segments sharing the same barcode are defined as connected even if they are spatially separated.El 1BIO-001-PCT

[0123] In some embodiments, the brain tissue sample can comprise cultured neural cells such as neurons. In some embodiments, the brain tissue sample is isolated from a mammalian subject and processed, such as by cryosectioning into tissue sections or tissue blocks.

[0123] In some embodiments, a tissue sample such as a brain tissue sample is chemically fixed (with e.g. paraformaldehyde) before contacting it with the antibody composition and antibodies selective for synaptic markers. In some embodiments, the sample is contacted by antibodies, then embedded in a gel such as a swellable hydrogel. In some embodiments, the sample is embedded in a gel such as a swellable hydrogel, then contacted by antibodies.

[0124] In one embodiment, the brain tissue sample is embedded in a swellable hydrogel. In some embodiments, a swellable hydrogel provided herein comprises a copolymer composition and one or more anchoring reagents. In some embodiments, the polymer composition comprises a copolymer of sodium acrylate, acrylamide, dimethylacrylamide (DMAA), and bis-acrylamide, and the biomolecules in the tissue are anchored to the polymer by inclusion of methacrolein in the polymerization solution. In some embodiments, In some embodiments, the polymer composition comprises a copolymer of sodium acrylate, acrylamide, and bis-acrylamide, and the biomolecules in the tissue are anchored to the polymer by inclusion of methacrolein in the polymerization solution, and dimethylacrylamide is not included. In some embodiments, In some embodiments, the polymer composition comprises a copolymer of acrylamide and bis- acrylamide, and the biomolecules in the tissue are anchored to the polymer by inclusion of methacrolein in the polymerization solution, and dimethylacrylamide and sodium acrylate are not included. In some embodiments, the anchoring reagent comprises methacrolein. In some embodiments, the anchoring reagent comprises any one or more of Acryloyl-X, SE, (6-((acryloyl)amino)hexanoic Acid, Succinimidyl Ester, and / or Methacrylic acid N-hydroxysuccinimide, with or without methacrolein.

[0125] Steps (b)(i) and (ii) may be carried out in any order, or may be carried out at the same time. When steps (b)(i) and (b)(ii) are not carried out at the same time, then steps (c) and (d) may be carried out before either of (b)(i) or (b)(ii).El 1BIO-001-PCT

[0126] In one embodiment, steps (b)-(d) are carried out iteratively. In one nonlimiting example, contacting step (b)(i) is carried out first, followed by steps (c) and (d) to obtain fluorescence images of the detectable antibody-epitope complexes and analyze the images, followed by stripping of the sample of the antibodies selective for peptide epitopes. Then the same brain tissue sample is contacted with the antibodies selective for synaptic markers, followed by steps (c) and (d) to obtain fluorescence images of the detectable antibody-synaptic marker complexes and analyze the images.

[0127] In another exemplary embodiment, contacting step (b)(i) comprises two or more iterative steps. By way of non-limiting example, step (b)(i) may comprise contacting the brain tissue sample with embodiments of the antibody composition that are present in two or more mixtures as described above. In this embodiment, contacting step (b)(i) is first carried out with a first mixture of the antibody composition (including a first set of antibodies selective for a subset of the peptide epitopes), followed by steps (c) and (d) to obtain and analyze fluorescence images of the first set of detectable antibodyepitope complexes, followed by stripping the sample of the first mixture of antibodies selective for peptide epitopes. Then the same brain tissue sample is contacted with a second mixture of the antibody composition ((including a first set of antibodies selective for a subset of the peptide epitopes), followed by steps (c) and (d) to obtain and analyze fluorescence images of the second set of detectable antibody-epitope complexes. Other embodiments will be clear to those of skill in the art based on the present disclosure.

[0128] Samples are embedded in a swellable hydrogel and expanded to improve resolution. Sets of proteins in the sample are decoded through N cycles of iterative immunostaining, fluorescence imaging, and destaining. For example, 5 bits may be read out per cycle, defined by the number of spectrally distinct laser lines on the microscope. Protein barcodes are read out across cycles. Endogenous proteins, such as synaptic markers are then read out in the same sample. The joint barcode (morphology) and synaptic data improves reconstructing connectivity.

[0129] Stripping antibody from the samples / destaining can be carried out using any suitable technique.

[0130] As disclosed herein, the methods of the disclosure provide significant improvements in detecting connectivity between cells in a brain tissue sample.El 1BIO-001-PCTSpecifically, the inventors have surprisingly discovered that the recited fluorescent protein component of the fusion proteins are far superior than previously used detectable proteins in filling cells in the brain tissue sample, including in axons and dendrites many millimeters from the cell body. The inventors have also identified the peptide epitope targets of SEQ ID NO: 1-31, and antibodies detecting them, as ideally suited for detection in cells in brain tissue samples relative to a starting set of approximately 250 peptide epitopes. For mapping brain circuit connectivity (“connectomics”), detection of marker combinations (such as the compositions of the disclosure), also referred to herein as “barcoding”, the barcodes need to fill the cell, including in axons and dendrites many millimeters from the cell body. The inventors have demonstrated that the compositions and methods of the disclosure can be used to examine expression in distant areas of the brain given an initial injection site as a proxy for sufficient cell filling, and that the barcodes have intrinsic error correction that can be utilized to significantly reduce error rates in mapping cell connections both at the level of projections between brain areas and at the level of single cells relative to previously available methods. The compositions and methods disclosed herein enable intrinsic error correction, permitting larger brain circuits to be accurately mapped due to fewer errors, and enable “targeted” circuit mapping of specific circuits by bridging spatial gaps (i.e., segments of cells in different brain areas can be connected using barcodes without tracing through the intervening volume).

[0131] The compositions for use in the methods may be any embodiment or combination of embodiments disclosed above. In one embodiment, the detectable synaptic markers may comprise antibodies that selectively bind synaptic targets selected from the group consisting of Amphiphysin, Ankyrin G, Bassoon, Dynamin 1 / 2 / 3, Gephyrin, Homer 1, MAP2, Munc 13-1, Parvalbumin, Piccolo, RIM1, Synaptophysin, VAMP2, vGluTl, PSD95, and Shank 2. Antibodies against these synaptic targets are commercially available, as noted in Table 8. In some embodiments, the kits include antibodies that detect both pre-synaptic and post-synaptic markers; the specificity of exemplary antibodies detecting pre-synaptic or post-synaptic proteins is also provided in Table 8. In other embodiments, the kits comprise at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 50, 100, or more antibodies detecting synaptic markers. The antibodies that selectively bind synaptic markers may be directly labeled with distinguishable, detectable labels, or by secondary labeling with secondary antibodies that are distinguishable, detectably labeled. In some embodiments, the antibodies that selectively bind synapticEl 1BIO-001-PCT markers are detectably distinguishable from antibodies that selectively bind to the protein epitopes. In other embodiments, the synaptic markers are not detectably distinguishable from antibodies that selectively bind to the protein epitopes; the methods of the disclosure comprise iterative immunostaining, and the synaptic marker-selective antibodies may be detected separately from the protein epitope selective antibodies.

[0132] In one embodiment, the brain tissue sample is injected with the viral particle composition of any embodiment herein prior to step (a). In one embodiment, different viral particles comprising nucleic acids encoding different fusion proteins are present in approximately stoichiometric ratios (i.e., + / - 10% of stoichiometric ratio). In a specific embodiment, each encoded fusion protein comprises eGFP. In another embodiment, the plurality of different viral particles comprise AAV particles. In another embodiment, the injecting results in each peptide epitope being expressed in a random subset of cells in the brain tissue sample. In some embodiments, the cells can express unique combinations of small epitope tags on a stable scaffold.

[0133] The following is a list of preferred embodiments.Embodiment 1. A method of processing a multicellular tissue specimen for microscopic imaging, comprising: embedding the multicellular tissue specimen in a first expandable polymer matrix and expanding the first expandable matrix to isotropically expand the multicellular tissue specimen, thereby providing an expanded multicellular tissue specimen; embedding the expanded multicellular tissue specimen in a stabilizing polymer matrix, thereby providing a stabilized expanded multicellular tissue specimen; sectioning the expanded multicellular tissue specimen to provide one or more multicellular tissue sections; and for one or more of the multicellular tissue sections, individually embedding the multicellular tissue section in a second expandable polymer matrix and expanding the second expandable matrix to isotropically expand theEl 1BIO-001-PCT multicellular tissue section, thereby providing an expanded multicellular tissue section, and passivate the expanded multicellular tissue section to provide a processed multicellular tissue sections; wherein a plurality of cells in the multicellular tissue specimen and / or in the multicellular tissue section are cellularly barcoded using a plurality of epitope tags, wherein the barcoding comprises at least 10, 25, 50, 100, 250, 500, 1000, 2500, 5000, 10000, 25000, 100,000, or more spectral color combinations visually discernable by microscopic imaging.Embodiment 2. A method of processing a multicellular tissue specimen for microscopic imaging, comprising: embedding the multicellular tissue specimen in a first expandable polymer matrix and expanding the first expandable matrix to isotropically expand the multicellular tissue specimen, thereby providing an expanded multicellular tissue specimen; embedding the expanded multicellular tissue specimen in a stabilizing polymer matrix, thereby providing a stabilized expanded multicellular tissue specimen; wherein a plurality of cells in the multicellular tissue specimen and / or in the multicellular tissue section are cellularly barcoded using a plurality of epitope tags, wherein the barcoding comprises at least 10, 25, 50, 100, 250, 500, 1000, 2500, 5000, 10000, 25000, 100,000, or more spectral color combinations visually discernable by microscopic imaging.Embodiment 3. A method according to embodiment 1 or 2, wherein the cells in the multicellular tissue specimen and / or in the multicellular tissue section are contacted with a plurality of binders recognizing a first subset of the plurality of epitope tags, and detecting first signals associated with the first plurality of binders to generate the one or more spectral color combinations.Embodiment 4. A method according to one of embodiments 1-3, wherein the multicellular tissue specimen is a neuron-containing tissue.El 1BIO-001-PCTEmbodiment 5. A method according to embodiment 2, wherein the multicellular tissue specimen is a central nervous system specimen from a vertebrate.Embodiment 6. A method according to one of embodiments 1-5, further comprising labeling the multicellular tissue section with a morphology stain.Embodiment 7. A method according to one of embodiments 1-6, wherein one or more of the processed multicellular tissue sections are microscopically imaged to provide one or more multicellular tissue images.Embodiment 8. A method according to embodiment 7, wherein a plurality of the processed multicellular tissue sections are serial sections, and the serial sections are microscopically imaged to provide a set of serial multicellular tissue images.Embodiment 9. A method according to embodiment 7 or 8, further comprising identifying one or more intercellular connections in the one or more multicellular tissue images.Embodiment 10. A method according to embodiment 9, wherein the one or more intercellular connections comprise one or more neuronal synapses.Embodiment 11. A method according to one of embodiments 1-10, wherein the barcoding comprises at least 100,000 spectral color combinations visually discernable by microscopic imaging.Embodiment 12. A method according to one or more of embodiments 1-11, one or more multicellular tissue images are obtained by one or more imaging methods selected from the group consisting of bright field microscopy, dark field microscopy, phase contrast microscopy, electron microscopy, fluorescence microscopy, reflection microscopy, interference microscopy and confocal microscopy.Embodiment 13. A method according to one or more of embodiments 8-12, wherein a three-dimensional reconstruction of all or a portion of the multicellular tissue specimen is prepared from the set of serial multicellular tissue images.El 1BIO-001-PCTEmbodiment 14. An image acquisition and processing method, comprising: on a computer system,(i) acquiring a raw digital image of a multicellular tissue section using an expansion microscopic imaging method, wherein a plurality of cells in the multicellular tissue section are cellularly barcoded using a plurality of epitope tags, wherein the barcoding comprises at least 10, 25, 50, 100, 250, 500, 1000, 2500, 5000, 10000, 25000, 100,000, or more spectral color combinations visually discernable in the expansion microscopic imaging method, and wherein the raw digital image is assembled from a set of stitched image tiles of the multicellular tissue section obtained from the expansion microscopic imaging method;(ii) repeating step (i) for a plurality of multicellular tissue sections to provide a dataset of raw digital images comprising a plurality of individual raw digital images;(iii) registering the individual raw digital images in the dataset of raw digital images to provide a registered dataset of raw digital images;(iv) using a subset of the individual raw digital images in the dataset of raw digital images, providing a dataset of ground truth sparse residual digital images by, for each individual raw digital image in the subset, displaying the raw digital image on an electronic display and manually entering annotated boundaries of cellular structures identified according to barcode color, averaging barcode color intensity within each annotated boundary to provide a color- averaged digital image, and subtracting the averaged digital image from the raw digital image to provide a ground truth sparse residual digital image;(v) using the raw digital images in the subset and the corresponding ground truth sparse residual digital images to train a 3D convolutional neural network predict boundaries of cellular structures within raw digital images;El IBIO-OOl-PCT(vi) providing a dataset of enhanced digital images by applying the trained 3D convolutional neural network to each raw digital image in the dataset of raw digital images to provide a predicted dense residual image corresponding to the raw digital image, and calculating a sum of the raw digital image and the corresponding predicted dense residual image to provide a corresponding to the enhanced digital image;(vii) using a first subset of enhanced digital images in the dataset of enhanced digital images to train a 3D convolutional neural network to predict boundary affinities and local shape descriptors of cellular structures within enhanced digital images, and providing datasets of predicted boundary affinity images and local shape descriptor images by applying the trained 3D convolutional neural network to each enhanced digital image in the dataset of enhanced digital images to provide a predicted boundary affinity image and local shape descriptor image corresponding to the enhanced digital image;(viii) using a second subset of enhanced digital images in the dataset of enhanced digital images to train a multidimensional convolutional neural network to estimate boundary probabilities of cellular structures within enhanced digital images and providing a database of boundary probability images of cellular structures by applying the trained 3D convolutional neural network to each enhanced digital image in the dataset of enhanced digital images to provide a boundary probability image of cellular structures corresponding to the enhanced digital image;(ix) using a third subset of enhanced digital images in the dataset of enhanced digital images to train a multilayer perceptron to provide a uniform embedding image of cellular structures within enhanced digital images, and applying the trained 3D convolutional neural network to each enhanced digital image in the dataset of enhanced digital images to provide uniform embedded image of cellular structures corresponding to the enhanced digital image;(x) providing a dataset of combined affinity images by calculating a dot product of each uniform embedded image in the dataset of uniform embedded images to provide a pseudoaffinity image corresponding to the uniform embedded image and calculating a product of each pseudoaffinity image within the dataset of pseudoaffinity images with its corresponding predicted boundary affinity image and local shape descriptor image to provide a combined affinity image corresponding to the pseudoaffinity image; andEl 1BIO-001-PCT(xi) providing a dataset of segmented images using the dataset of combined affinity images.Embodiment 15. A method according to embodiment 14, wherein the multicellular tissue specimen is a neuron-containing tissue.Embodiment 16. A method according to embodiment 15, wherein the multicellular tissue specimen is a central nervous system specimen from a vertebrate.Embodiment 17. A method according to one of embodiments 14-16, further comprising labeling the multicellular tissue section with a morphology stain.Embodiment 18. A method according to embodiment 14-17, further comprising identifying one or more intercellular connections in the one or more multicellular tissue images.Embodiment 19. A method according to embodiment 18, wherein the one or more intercellular connections comprise one or more neuronal synapses.Embodiment 20. A method according to one of embodiments 14-19, wherein the barcoding comprises at least 10, 25, 50, 100, 250, 500, 1000, 2500, 5000, 10000, 25000, or 100,000 spectral color combinations visually discernable in the expansion microscopic imaging method.Embodiment 21. A method according to one or more of embodiments 14-20, one or more multicellular tissue images are obtained by one or more imaging methods selected from the group consisting of bright field microscopy, dark field microscopy, phase contrast microscopy, electron microscopy, fluorescence microscopy, reflection microscopy, interference microscopy and confocal microscopy.Embodiment 22. A method according to one or more of embodiments 14-21, wherein a three-dimensional reconstruction of all or a portion of the multicellular tissue specimen is prepared from the dataset of segmented images.Embodiment 23. A method according to one of embodiments 14-22, wherein the cellular barcoding is performed according to the method of one of embodiments 1-13.El 1BIO-001-PCTEmbodiment 24. A method of image acquisition and processing that includes automated proofreading, comprising: on a computer system,(i) acquiring a raw digital image of a multicellular tissue section using an expansion microscopic imaging method, wherein a plurality of cells in the multicellular tissue section are cellularly barcoded using a plurality of epitope tags, wherein the barcoding comprises at least 10, 25, 50, 100, 250, 500, 1000, 2500, 5000, 10000, 25000, 100,000, or more spectral color combinations visually discernable in the expansion microscopic imaging method, and wherein the raw digital image is assembled from a set of stitched image tiles of the multicellular tissue section obtained from the expansion microscopic imaging method;(ii) repeating step (i) for a plurality of multicellular tissue sections to provide a dataset of raw digital images comprising a plurality of individual raw digital images;(iii) registering the individual raw digital images in the dataset of raw digital images to provide a registered dataset of raw digital images;(iv) using a subset of the individual raw digital images in the dataset of raw digital images, providing a dataset of ground truth sparse residual digital images by, for each individual raw digital image in the subset, displaying the raw digital image on an electronic display and manually entering annotated boundaries of cellular structures identified according to barcode color, averaging barcode color intensity within each annotated boundary to provide a color- averaged digital image, and subtracting the averaged digital image from the raw digital image to provide a ground truth sparse residual digital image;(v) using the raw digital images in the subset and the corresponding ground truth sparse residual digital images to train a 3D convolutional neural network predict boundaries of cellular structures within raw digital images;El IBIO-OOl-PCT(vi) providing a dataset of enhanced digital images by applying the trained 3D convolutional neural network to each raw digital image in the dataset of raw digital images to provide a predicted dense residual image corresponding to the raw digital image, and calculating a sum of the raw digital image and the corresponding predicted dense residual image to provide a corresponding to the enhanced digital image;(vii) using a first subset of enhanced digital images in the dataset of enhanced digital images to train a 3D convolutional neural network to predict boundary affinities and local shape descriptors of cellular structures within enhanced digital images, and providing datasets of predicted boundary affinity images and local shape descriptor images by applying the trained 3D convolutional neural network to each enhanced digital image in the dataset of enhanced digital images to provide a predicted boundary affinity image and local shape descriptor image corresponding to the enhanced digital image;(viii) using a second subset of enhanced digital images in the dataset of enhanced digital images to train a multidimensional convolutional neural network to estimate boundary probabilities of cellular structures within enhanced digital images and providing a database of boundary probability images of cellular structures by applying the trained 3D convolutional neural network to each enhanced digital image in the dataset of enhanced digital images to provide a boundary probability image of cellular structures corresponding to the enhanced digital image;(ix) using a third subset of enhanced digital images in the dataset of enhanced digital images to train a multilayer perceptron to provide a uniform embedding image of cellular structures within enhanced digital images, and applying the trained 3D convolutional neural network to each enhanced digital image in the dataset of enhanced digital images to provide uniform embedded image of cellular structures corresponding to the enhanced digital image;(x) providing a dataset of combined affinity images by calculating a dot product of each uniform embedded image in the dataset of uniform embedded images to provide a pseudoaffinity image corresponding to the uniform embedded image and calculating a product of each pseudoaffinity image within the dataset of pseudoaffinity images with its corresponding predicted boundary affinity image and local shape descriptor image to provide a combined affinity image corresponding to the pseudoaffinity image;El 1BIO-001-PCT(xi) providing a dataset of segmented images using the dataset of combined affinity images, wherein a segment in the dataset of segmented images is a cell or part of a cell;(xii) masking the raw registered data to a set of pixels within each segment and an average barcode per segment calculated by averaging the per channel values within the set of pixels,(xiii) simplifying the segments in the dataset of segmented images into a series of nodes and edges using a Tree-structure Extraction Algorithm for Accurate and Robust Skeletons (TEASAR) algorithm, wherein the average barcode per segment is assigned to each node within the segment;(xiv) calculating a pairwise cosine distance between the average barcode at each node to all other nodes;(xv) for nodes within a predefined barcode distance, and optionally within a predefined spatial distancejoining the nodes together with an edge and relabeling the segments corresponding to the joined nodes to be a common segment.Examples

[0134] The following examples are included for illustrative purposes only and are not intended to limit the scope of the present disclosure.

[0135] International Application No: PCT / US2024 / 055771, which is hereby incorporated by reference in its entirety, discloses methods for barcoding of cellular samples.

[0136] Example 1

[0137] A first workflow diagram is shown in Fig. 1. This workflow begins with labeling barcode proteins using primary antibodies, followed by chemical crosslinking to secure them in place. The sample is then embedded in a first swellable hydrogel that allows isotropic expansion. After an initial water-driven expansion, the structure is stabilized in a secondary gel matrix and subsequently re-embedded in a second swellable gel to enable higher expansion factors. Once stabilized, secondary antibodies are introduced to visualize the barcodes, and a complementary morphology stain is applied toEl 1BIO-001-PCT reveal cellular structures. Finally, the gel is expanded again and imaged, providing both barcode and morphological information. In certain cases, proteins are more available pregelation and this allows them to be detected.

[0138] Exemplary results from this workflow are shown in Fig. 2 as a three-channel image. Panel A shows the morphology stain, providing structural context. Panel B displays one barcode stain, and Panel C shows a second, independent barcode stain. Together, these channels demonstrate the combination of cellular morphology with multiplexed barcode readouts using workflow 1.

[0139] Example 2

[0140] A second workflow diagram is shown in Fig. 3. This workflow starts with direct embedding of the sample in a first swellable hydrogel, followed by water-driven expansion. The expanded material is then stabilized with a secondary gel and reembedded in a second swellable gel to permit additional expansion. After this preparation, barcode proteins are labeled with primary antibodies. Secondary antibodies are then applied to reveal the barcode signals, and a morphology stain is introduced to capture cellular architecture. The sample is expanded further and imaged, producing data that combines barcode identity with detailed morphology. This approach shifts antibody labeling until after the major expansion steps, reducing steric hindrance and improving antibody penetration in dense tissue. By expanding first, then staining, it maximizes access of probes to epitopes while still achieving ultra-high expansion factors. The combination of barcode and morphology channels again supports robust cell segmentation and fine ultrastructural mapping.

[0141] Shown in Fig. 4 is a six-channel dataset from workflow 2. Panels are described counter-clockwise from the top-left. The first panel displays an NHS-derived ultrastructural stain of the morphology channel. The following five panels each show a distinct barcode channel. Together, these channels demonstrate multiplexed barcode labeling overlaid on fine structural context. Arrows highlight regions where barcode signals localize to thin structures, including axons and near-synaptic terminals, illustrating that different barcodes can distinguish neighboring features. Scale bar, 50 pm postexpansion, corresponding to ~3 pm biological distance in an 18-fold expanded tissue.El 1BIO-001-PCT

[0142] A single plane of data acquired in M3is shown in Fig. 5. Panel (a) is the Morphology pan-stain; panels (b) and (c) denote two unique protein stains. Arrows denote the track of a thin axon, detectable only by use of the barcode; and the point where two barcodes are present in the pre and post synaptic regions of the same synapse.

[0143] Example 3

[0144] A third workflow diagram is shown in Fig. 6. The sample is first embedded in a swellable gel and expanded by water uptake, then stabilized in a secondary gel. Barcode proteins are labeled with primary and secondary antibodies, and an initial round of imaging is performed. The sample is then re-embedded in a second swellable gel, and an additional staining step is carried out, incorporating at least one channel shared with the first imaging round alongside a morphology stain. After a second expansion and imaging, the shared channel is used to register mid-resolution barcode data to the higher-resolution morphological dataset. The combined barcode and morphology information enables accurate segmentation of individual cells.

[0145] This staged approach separates barcode acquisition from final morphology capture, allowing barcodes to be recorded under conditions that favor labeling efficiency while reserving maximum expansion and resolution for morphology. The use of a shared reference channel ensures accurate alignment between imaging rounds. This strategy yields high-quality barcode data linked directly to ultrastructural morphology, enabling robust segmentation and precise mapping of cellular identity.

[0146] Fig. 7A shows the morphology stain acquired at full expansion, providing fine ultrastructural detail. Fig. 7B presents a barcode stain acquired at full expansion, used as the registration channel. Fig. 7C displays a barcode stain at 8* expansion, highlighting improved dendritic spine filling and signal smoothing. Fig. 7D shows a synaptic marker stain acquired at intermediate expansion and registered back to the full-resolution morphology channel. Shown with arrows are thin spines, demonstrating that signals acquired at lower resolution (where they are brighter and easier to detect) correctly align to fine structures in the full-resolution morphology data. The scale bars indicate approximate pre-expansion size.

[0147] Example 4El 1BIO-001-PCT

[0148] Figs. 8 and 9 provide an overview of image acquisition, stitching and registration of raw digital images of a multicellular tissue section (Fig. 8), in this case a neuronal tissue, and processing the processed digital images to segment cells (in this case neurons) using machine learning algorithms as described herein.

[0149] In Fig. 10, ground truth for networks were computed by computing average barcodes sparsely. Raw barcode data (shape c,z,y,x where c = 18) was used to generate unique ground truth label data (zyx) that is sparse, i.e not all objects are painted in the volume. Using the ground truth labels, the average barcode intensities (channel wise) of the raw data (czyx) are computed.

[0150] In Fig. 11, “residual barcodes” were calculated by first subtracting the raw barcode data from the average barcode data. This produces “residual barcodes” or the difference between the average and the raw. Raw pixels that are lower (darker) than the average barcode would have a higher (brighter) residual (and vice versa).

[0151] In Fig. 12, an agnostic network was trained to predict dense residual barcodes. An agnostic residual u-net was trained on sparse labels to predict dense residual barcodes. Masked out regions (the gray background) were not considered in the loss computation as the label data is sparse so these regions are considered “unknown”. The network is then forced to learn what to predict in these regions using what it learns from the masked-in regions. In this u-net, the input to the convolutional block was added to the output of the convolutions. Following the x -> f(x) + x motif of residual networks (where + x is the identity function) has as been found to improve training stability and results especially for denoising tasks.

[0152] In Fig. 13, predicted dense residual barcodes were added to raw barcodes to get final enhanced barcodes. As noted above, pixels in the raw image that are dimmer than the average barcode will have a brighter residual and vice versa. Adding these together brings the final enhanced barcodes towards the mean. This helps to greatly reduce the variability of the barcode. Additionally, since the residual is added back to the barcodes, this ensures that the result retains information from the barcodes in areas in which the residual predictions are not optimal. This may help with finer objects and with overlapping objects that might have a strong signal in the raw data but a weaker signal in the residual, while still allowing the use of the residual result in cases where the barcodesEl 1BIO-001-PCT suffer (e.g large black holes inside objects). This step is important as the output is used as input for downstream nets and greatly affects the quality of downstream segmentations.

[0153] In Fig. 14, a 3d multi-task (separate output heads) u-net was trained to learn affinities and LSDs (combined losses) using multi-channel enhanced barcodes as input. The affinities were trained with a long-range neighborhood which both acts as an auxiliary learning task and is useful for edge weights during downstream mutex watershed. Additionally, the LSDs force the network to make more use of shapes to infer boundaries instead of only relying on pixel intensity information.

[0154] In Fig. 15, a 2.5d u-net with increased receptive field was trained to learn barcode probabilities - i.e. closer to 1 inside “barcodable” neurons, and closer to 0 outside (e.g. background, blood vessels, dim objects, etc.). These probabilities are then thresholded to use as a mask for downstream watershed.

[0155] In Fig. 16 a multilayer perceptron (MLP) was trained to project enhanced barcodes into a higher dimensional (more channels) embedding, with a larger range in the intensity values. This network only takes in channel information, no spatial information. It is designed to further amplify the barcodes and therefore is very sensitive to intensity variability (hence why it is valuable to first enhance well), and alignment / registration errors (hence why it is very valuable to have an airtight registration pipeline).

[0156] Figs. 17 and 18 summarize the processing flow discussed above. Given ground truth labels, one could simply train a network to predict foreground / background pixels. However, this becomes limited by the resolution of the data - e.g. objects may become falsely assigned to background. A solution to this is to instead assign a connectivity probability to each edge between voxels (known as an affinity graph). This effectively handles the resolution issue, and then the affinity graph can be used to generate segmentations via various clustering / agglomeration algorithms (e.g low affinity signals boundary -> split, high affinity signals inside -> merge). In addition to the short- range affinities, a network may also be trained to make use of long range affinity neighborhoods. This acts as a useful auxiliary learning task and then allows us to use the different neighborhoods for splitting / merging during mutex watershed (e.g merge with short, split with long).El 1BIO-001-PCT

[0157] For LSDs, a gaussian center around each voxel was computed and intersected with the underlying label. The center of mass of the intersected region was computed, as well as various statistics between the voxel and the center of mass. These statistics are designed to capture attributes descriptive to the object’s shape such as its size, direction it is moving, elongation, and the “normals” (e.g. vectors perpendicular to the surface). Learning these descriptors along with affinities helps the network to infer presence of a boundary in the absence of sufficient pixel intensity evidence.

[0158] According to the present example, raw data is first enhanced, and then predicted affinities and uniform embedding is applied. From the uniform embedding “pseudo affinities” are computed via the dot product. These affinities are then combined via the product. Since the pseudo affinities will split more, taking the product will also split more - which could be preferable in certain cases to reduce merges.

[0159] Fig. 19 depicts an example showing how the uniform embedding helps to prevent a merge. The top row shows a traditional segmentation pipeline. The white arrowhead points to a false merge caused by the affinity network failing to learn a boundary between these two neurons. However, since this is already a clear split in the barcodes of the present method, uniform embedding helps to amplify this and in the subsequent pseudo affinities a boundary between these two objects is indicated. Combining the two affinities resolves this merge.

[0160] Fig. 20 depicts an example workflow of how barcode information can be used to detect and correct errors in a segmentation of image data, including merging incorrectly disconnected portions of the same object i.e. automated proofreading. In the top row, the first illustration shows a set of segments with false splits from the initial segmentation. The second illustration shows the same segments converted into representative nodes and edges, where nodes are a spatially localized point (3D coordinate) on a graph representation of a segment’s centerline and edges are a connection between two nodes that follows the centerline of the segment. The third shows the region of spatial query around the end nodes of the segments. On the bottom row, the first illustration shows how nodes can be compared and edges added based on barcode distance between the nodes. The second shows the reconnected sets of nodes and edges. The third illustration shows the now-correctly merged segments.El 1BIO-001-PCT

[0161] It is to be understood that the disclosure is not limited in its application to the details of construction and to the arrangements of the components set forth in the following description or illustrated in the drawings. The disclosure is capable of embodiments in addition to those described and of being practiced and carried out in various ways. As such, those skilled in the art will appreciate that the conception upon which this disclosure is based may readily be utilized as a basis for the designing of other structures, methods and systems for carrying out the several purposes of the present disclosure. It is important, therefore, that the claims be regarded as including such equivalent constructions insofar as they do not depart from the spirit and scope of the present disclosure.

[0162] While the disclosure has been described and exemplified in sufficient detail for those skilled in this art to make and use it, various alternatives, modifications, and improvements should be apparent without departing from the spirit and scope of the disclosure. The examples provided herein are representative of preferred embodiments, are exemplary, and are not intended as limitations on the scope of the disclosure. Modifications therein and other uses will occur to those skilled in the art. These modifications are encompassed within the spirit of the disclosure and are defined by the scope of the claims.

[0163] It will be readily apparent to a person skilled in the art that varying substitutions and modifications may be made to the disclosure disclosed herein without departing from the scope and spirit of the disclosure.

[0164] All patent applications, patents, publications and other references mentioned in the specification are indicative of the levels of those of ordinary skill in the art to which the disclosure pertains and are each incorporated herein by reference. The references cited herein are not admitted to be prior art to the claimed disclosure.

[0165] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. In the case of conflict, the present specification, including definitions, will control.

[0166] The use of the articles "a", "an", and "the" in both the description and claims are to be construed to cover both the singular and the plural, unless otherwise indicatedEl 1BIO-001-PCT herein or clearly contradicted by context. The terms "comprising", "having", "being of' as in "being of a chemical formula", "including", and "containing" are to be construed as open terms (i.e., meaning "including but not limited to") unless otherwise noted. Additionally, whenever "comprising" or another open-ended term is used in an embodiment, it is to be understood that the same embodiment can be more narrowly claimed using the intermediate term "consisting essentially of or the closed term "consisting of'.

[0167] The term "about", "approximately", or "approximate", when used in connection with a numerical value, means that a collection or range of values is included. For example, "about X" includes a range of values that are ±20%, ±10%, ±5%, ±2%, ±1%, ±0.5%, ±0.2%, or ±0.1% of X, where X is a numerical value. In one embodiment, the term "about" refers to a range of values which are 10% more or less than the specified value. In another embodiment, the term "about" refers to a range of values which are 5% more or less than the specified value. In another embodiment, the term "about" refers to a range of values which are 1% more or less than the specified value.

[0168] Recitation of ranges of values are merely intended to serve as a shorthand method of referring individually to each separate value falling within the range, unless otherwise indicated herein, and each separate value is incorporated into the specification as if it were individually recited herein. A range used herein, unless otherwise specified, includes the two limits of the range. For example, the terms "between X and Y" and "range from X to Y, are inclusive of X and Y and the integers there between. On the other hand, when a series of individual values are referred to in the disclosure, any range including any of the two individual values as the two end points is also conceived in this disclosure.

[0169] The disclosure illustratively described herein suitably may be practiced in the absence of any element or elements, limitation or limitations which is not specifically disclosed herein. Thus, for example, in each instance herein any of the terms "comprising", "consisting essentially of and "consisting of' may be replaced with either of the other two terms. The terms and expressions which have been employed are used as terms of description and not of limitation, and there is no intention that in the use of such terms and expressions of excluding any equivalents of the features shown and described or portions thereof, but it is recognized that various modifications are possible within theEl 1BIO-001-PCT scope of the claims. Thus, it should be understood that although the present disclosure has been specifically disclosed by preferred embodiments and optional features, modification and variation of the concepts herein disclosed may be resorted to by those skilled in the art, and that such modifications and variations are considered to be within the scope of this disclosure as defined by the appended claims.

[0170] Other embodiments are set forth within the following claims.

Claims

El 1BIO-001-PCTCLAIMS1. A method of processing a multicellular tissue specimen for microscopic imaging, comprising: embedding the multicellular tissue specimen in a first expandable polymer matrix and expanding the first expandable matrix to isotropically expand the multicellular tissue specimen, thereby providing an expanded multicellular tissue specimen; embedding the expanded multicellular tissue specimen in a stabilizing polymer matrix, thereby providing a stabilized expanded multicellular tissue specimen; sectioning the expanded multicellular tissue specimen to provide one or more multicellular tissue sections; and for one or more of the multicellular tissue sections, individually embedding the multicellular tissue section in a second expandable polymer matrix and expanding the second expandable matrix to isotropically expand the multicellular tissue section, thereby providing an expanded multicellular tissue section, and passivate the expanded multicellular tissue section to provide a processed multicellular tissue sections; wherein a plurality of cells in the multicellular tissue specimen and / or in the multicellular tissue section are cellularly barcoded using a plurality of epitope tags, wherein the barcoding comprises at least 10, 25, 50, 100, 250, 500, 1000, 2500, 5000, 10000, 25000, 100,000, or more spectral color combinations visually discernable by microscopic imaging.

2. A method of processing a multicellular tissue specimen for microscopic imaging, comprising: embedding the multicellular tissue specimen in a first expandable polymer matrix and expanding the first expandable matrix to isotropically expand the multicellular tissue specimen, thereby providing an expanded multicellular tissue specimen;El 1BIO-001-PCT embedding the expanded multicellular tissue specimen in a stabilizing polymer matrix, thereby providing a stabilized expanded multicellular tissue specimen; wherein a plurality of cells in the multicellular tissue specimen and / or in the multicellular tissue section are cellularly barcoded using a plurality of epitope tags, wherein the barcoding comprises at least 10, 25, 50, 100, 250, 500, 1000, 2500, 5000, 10000, 25000, 100,000, or more spectral color combinations visually discernable by microscopic imaging.

3. A method according to claim 1 or 2, wherein the cells in the multicellular tissue specimen and / or in the multicellular tissue section are contacted with a plurality of binders recognizing a first subset of the plurality of epitope tags, and detecting first signals associated with the first plurality of binders to generate the one or more spectral color combinations.

4. A method according to one of claims 1-3, wherein the multicellular tissue specimen is a neuron-containing tissue.

5. A method according to claim 2, wherein the multicellular tissue specimen is a central nervous system specimen from a vertebrate.

6. A method according to one of claims 1-5, further comprising labeling the multicellular tissue section with a morphology stain.

7. A method according to one of claims 1-6, wherein one or more of the processed multicellular tissue sections are microscopically imaged to provide one or more multicellular tissue images.

8. A method according to claim 7, wherein a plurality of the processed multicellular tissue sections are serial sections, and the serial sections are microscopically imaged to provide a set of serial multicellular tissue images.

9. A method according to claim 7 or 8, further comprising identifying one or more intercellular connections in the one or more multicellular tissue images.El 1BIO-001-PCT10. A method according to claim 9, wherein the one or more intercellular connections comprise one or more neuronal synapses.

11. A method according to one of claims 1-10, wherein the barcoding comprises at least 100,000 spectral color combinations visually discernable by microscopic imaging.

12. A method according to one or more of claims 1-11, one or more multicellular tissue images are obtained by one or more imaging methods selected from the group consisting of bright field microscopy, dark field microscopy, phase contrast microscopy, electron microscopy, fluorescence microscopy, reflection microscopy, interference microscopy and confocal microscopy.

13. A method according to one or more of claims 8-12, wherein a three-dimensional reconstruction of all or a portion of the multicellular tissue specimen is prepared from the set of serial multicellular tissue images.

14. An image acquisition and processing method, comprising: on a computer system,(i) acquiring a raw digital image of a multicellular tissue section using an expansion microscopic imaging method, wherein a plurality of cells in the multicellular tissue section are cellularly barcoded using a plurality of epitope tags, wherein the barcoding comprises at least 10, 25, 50, 100, 250, 500, 1000, 2500, 5000, 10000, 25000, 100,000, or more spectral color combinations visually discernable in the expansion microscopic imaging method, and wherein the raw digital image is assembled from a set of stitched image tiles of the multicellular tissue section obtained from the expansion microscopic imaging method;(ii) repeating step (i) for a plurality of multicellular tissue sections to provide a dataset of raw digital images comprising a plurality of individual raw digital images;(iii) registering the individual raw digital images in the dataset of raw digital images to provide a registered dataset of raw digital images;El IBIO-OOl-PCT(iv) using a subset of the individual raw digital images in the dataset of raw digital images, providing a dataset of ground truth sparse residual digital images by, for each individual raw digital image in the subset, displaying the raw digital image on an electronic display and manually entering annotated boundaries of cellular structures identified according to barcode color, averaging barcode color intensity within each annotated boundary to provide a color-averaged digital image, and subtracting the averaged digital image from the raw digital image to provide a ground truth sparse residual digital image;(v) using the raw digital images in the subset and the corresponding ground truth sparse residual digital images to train a 3D convolutional neural network predict boundaries of cellular structures within raw digital images;(vi) providing a dataset of enhanced digital images by applying the trained 3D convolutional neural network to each raw digital image in the dataset of raw digital images to provide a predicted dense residual image corresponding to the raw digital image, and calculating a sum of the raw digital image and the corresponding predicted dense residual image to provide a corresponding to the enhanced digital image;(vii) using a first subset of enhanced digital images in the dataset of enhanced digital images to train a 3D convolutional neural network to predict boundary affinities and local shape descriptors of cellular structures within enhanced digital images, and providing datasets of predicted boundary affinity images and local shape descriptor images by applying the trained 3D convolutional neural network to each enhanced digital image in the dataset of enhanced digital images to provide a predicted boundary affinity image and local shape descriptor image corresponding to the enhanced digital image;(viii) using a second subset of enhanced digital images in the dataset of enhanced digital images to train a multidimensional convolutional neural network to estimate boundary probabilities of cellular structures within enhanced digital images and providing a database of boundary probability images of cellular structures by applying the trained 3D convolutional neural network to each enhanced digital image in the dataset of enhancedEl IBIO-OOl-PCT digital images to provide a boundary probability image of cellular structures corresponding to the enhanced digital image;(ix) using a third subset of enhanced digital images in the dataset of enhanced digital images to train a multilayer perceptron to provide a uniform embedding image of cellular structures within enhanced digital images, and applying the trained 3D convolutional neural network to each enhanced digital image in the dataset of enhanced digital images to provide uniform embedded image of cellular structures corresponding to the enhanced digital image;(x) providing a dataset of combined affinity images by calculating a dot product of each uniform embedded image in the dataset of uniform embedded images to provide a pseudoaffinity image corresponding to the uniform embedded image and calculating a product of each pseudoaffinity image within the dataset of pseudoaffinity images with its corresponding predicted boundary affinity image and local shape descriptor image to provide a combined affinity image corresponding to the pseudoaffinity image; and(xi) providing a dataset of segmented images using the dataset of combined affinity images.

15. A method according to claim 14, wherein the multicellular tissue specimen is a neuron-containing tissue.

16. A method according to claim 15, wherein the multicellular tissue specimen is a central nervous system specimen from a vertebrate.

17. A method according to one of claims 14-16, further comprising labeling the multicellular tissue section with a morphology stain.

18. A method according to claim 14-17, further comprising identifying one or more intercellular connections in the one or more multicellular tissue images.

19. A method according to claim 18, wherein the one or more intercellular connections comprise one or more neuronal synapses.El 1BIO-001-PCT20. A method according to one of claims 14-19, wherein the barcoding comprises at least 10, 25, 50, 100, 250, 500, 1000, 2500, 5000, 10000, 25000, or 100,000 spectral color combinations visually discernable in the expansion microscopic imaging method.

21. A method according to one or more of claims 14-20, one or more multicellular tissue images are obtained by one or more imaging methods selected from the group consisting of bright field microscopy, dark field microscopy, phase contrast microscopy, electron microscopy, fluorescence microscopy, reflection microscopy, interference microscopy and confocal microscopy.

22. A method according to one or more of claims 14-21, wherein a three-dimensional reconstruction of all or a portion of the multicellular tissue specimen is prepared from the dataset of segmented images.

23. A method according to one of claims 14-22, wherein the cellular barcoding is performed according to the method of one of claims 1-13.

24. A method of image acquisition and processing that includes automated proofreading, comprising: on a computer system,(i) acquiring a raw digital image of a multicellular tissue section using an expansion microscopic imaging method, wherein a plurality of cells in the multicellular tissue section are cellularly barcoded using a plurality of epitope tags, wherein the barcoding comprises at least 10, 25, 50, 100, 250, 500, 1000, 2500, 5000, 10000, 25000, 100,000, or more spectral color combinations visually discernable in the expansion microscopic imaging method, and wherein the raw digital image is assembled from a set of stitched image tiles of the multicellular tissue section obtained from the expansion microscopic imaging method;(ii) repeating step (i) for a plurality of multicellular tissue sections to provide a dataset of raw digital images comprising a plurality of individual raw digital images;(iii) registering the individual raw digital images in the dataset of raw digital images to provide a registered dataset of raw digital images;El IBIO-OOl-PCT(iv) using a subset of the individual raw digital images in the dataset of raw digital images, providing a dataset of ground truth sparse residual digital images by, for each individual raw digital image in the subset, displaying the raw digital image on an electronic display and manually entering annotated boundaries of cellular structures identified according to barcode color, averaging barcode color intensity within each annotated boundary to provide a color-averaged digital image, and subtracting the averaged digital image from the raw digital image to provide a ground truth sparse residual digital image;(v) using the raw digital images in the subset and the corresponding ground truth sparse residual digital images to train a 3D convolutional neural network predict boundaries of cellular structures within raw digital images;(vi) providing a dataset of enhanced digital images by applying the trained 3D convolutional neural network to each raw digital image in the dataset of raw digital images to provide a predicted dense residual image corresponding to the raw digital image, and calculating a sum of the raw digital image and the corresponding predicted dense residual image to provide a corresponding to the enhanced digital image;(vii) using a first subset of enhanced digital images in the dataset of enhanced digital images to train a 3D convolutional neural network to predict boundary affinities and local shape descriptors of cellular structures within enhanced digital images, and providing datasets of predicted boundary affinity images and local shape descriptor images by applying the trained 3D convolutional neural network to each enhanced digital image in the dataset of enhanced digital images to provide a predicted boundary affinity image and local shape descriptor image corresponding to the enhanced digital image;(viii) using a second subset of enhanced digital images in the dataset of enhanced digital images to train a multidimensional convolutional neural network to estimate boundary probabilities of cellular structures within enhanced digital images and providing a database of boundary probability images of cellular structures by applying the trained 3D convolutional neural network to each enhanced digital image in the dataset of enhancedEl IBIO-OOl-PCT digital images to provide a boundary probability image of cellular structures corresponding to the enhanced digital image;(ix) using a third subset of enhanced digital images in the dataset of enhanced digital images to train a multilayer perceptron to provide a uniform embedding image of cellular structures within enhanced digital images, and applying the trained 3D convolutional neural network to each enhanced digital image in the dataset of enhanced digital images to provide uniform embedded image of cellular structures corresponding to the enhanced digital image;(x) providing a dataset of combined affinity images by calculating a dot product of each uniform embedded image in the dataset of uniform embedded images to provide a pseudoaffinity image corresponding to the uniform embedded image and calculating a product of each pseudoaffinity image within the dataset of pseudoaffinity images with its corresponding predicted boundary affinity image and local shape descriptor image to provide a combined affinity image corresponding to the pseudoaffinity image;(xi) providing a dataset of segmented images using the dataset of combined affinity images, wherein a segment in the dataset of segmented images is a cell or part of a cell;(xii) masking the raw registered data to a set of pixels within each segment and an average barcode per segment calculated by averaging the per channel values within the set of pixels,(xiii) simplifying the segments in the dataset of segmented images into a series of nodes and edges using a Tree-structure Extraction Algorithm for Accurate and Robust Skeletons (TEASAR) algorithm, wherein the average barcode per segment is assigned to each node within the segment;(xiv) calculating a pairwise cosine distance between the average barcode at each node to all other nodes;(xv) for nodes within a predefined barcode distance, and optionally within a predefined spatial distancejoining the nodes together with an edge and relabeling the segments corresponding to the joined nodes to be a common segment.

Images