Genome-scale imaging of 3d organization and transcriptional activity of chromatin

Abstract

The present invention relates generally to genomics. Some embodiments relate to imaging the 3D organization of a genome or a portion of a genome in sequence space at high throughput. Some embodiments relate to imaging the 3D organization of a genome or a portion of a genome in the context of transcriptional activity and nuclear structure. Further, certain embodiments relate to chromatin structure, 3D chromatin organization, chromosome- chromosome interactions and chromatin-nuclear structure interactions, and their relationship to transcription, among others. Further, various embodiments relate to imaging methods that allow mapping of the 3D organization of a genome or a portion of a genome in the context of nuclear structure and transcriptional activity. Some embodiments relate to large-scale multiplexed fluorescence in situ hybridization methods for imaging chromatin loci and / or nascent RNA transcripts on a chromosome or genome scale.

Description

[0001] This application is a divisional application, with its parent application number being 202080087114.8, filed on December 18, 2020, and entitled "3D organization of chromatin and genome-scale imaging of transcriptional activity".

[0002] Related applications

[0003] This application claims the benefit of U.S. Provisional Patent Application Serial No. 62 / 954,720, entitled "Genome-Scale Imaging of the 3D Organization and Transcriptional Activity of Chromatin," filed December 30, 2019, and U.S. Provisional Patent Application Serial No. 63 / 060,947, entitled "Genome-Scale Imaging of the 3D Organization and Transcriptional Activity of Chromatin," filed August 4, 2020. Each of these is incorporated herein by reference in its entirety. Technical Field

[0004] This invention generally relates to genomics. Some embodiments involve imaging the 3D organization of the genome in the context of transcriptional activity and nuclear structure. Furthermore, some embodiments involve chromatin organization and chromatin-nuclear structure interactions and their relationship to transcription. Background Technology

[0005] The three-dimensional (3D) organization of the genome regulates many fundamental cellular functions, from gene expression to DNA replication. Biochemical and imaging measurements reveal complex chromatin structures across a wide range of scales. Recently, high-throughput chromosome conformation capture methods, such as Hi-C and other sequencing-based methods, have greatly enriched our knowledge of 3D genome organization, revealing chromatin structures such as loops, domains, and compartments from a whole-genome perspective. These powerful sequencing-based methods also have limitations. For example, they provide information on the connections between chromatin locus pairs but not on the direct spatial locations of these loci. Furthermore, most whole-genome insights into chromatin organization are built on population-averaged contact maps of millions of cells. Despite continuous improvements in single-cell Hi-C methods, the capture efficiency of chromatin contacts in single cells and / or the cell throughput of these methods remain relatively low, making the study of 3D genome organization in single cells a challenging task. Moreover, although methods combining Hi-C with other measurements have emerged, such as providing characterization of chromatin contacts in the context of interacting proteins, nuclear structures, or DNA modifications, multimodal measurements via sequencing remain challenging. It is noteworthy that a method that allows for genome-scale measurements of both chromatin organization and transcriptional activity in the same cell has not yet emerged, which is highly desirable because understanding how chromatin organization regulates transcription and how transcription, in turn, affects chromatin organization is crucial.

[0006] On the other hand, imaging-based methods offer direct measurements of the spatial location of chromatin loci in individual cells with high detection efficiency. In particular, fluorescence in situ hybridization (FISH) allows for highly specific detection of chromatin loci in fixed cells, and recently, the CRISPR (Clustered Short Palindromic Repeats) system has greatly enhanced our ability to image specific chromatin loci in living cells. Chromatin imaging can also be combined with RNA and protein imaging to reveal interactions between chromatin organization and transcriptional activity or interacting protein factors. However, current imaging methods are limited in throughput in sequence space, traditionally allowing only a few different genomic loci to be studied at a time. Genome-scale imaging requires a significant increase in the number of genomic loci that can be imaged in individual cells. Therefore, new improvements are needed. Summary of the Invention

[0007] This invention generally relates to genomics. Some embodiments involve high-throughput imaging of the 3D organization of a genome or a portion of a genome in sequence space. Some embodiments involve imaging the 3D organization of a genome or a portion of a genome in the context of transcriptional activity and nuclear structure. Furthermore, some embodiments involve chromatin structure, 3D chromatin organization, transchromosomal interactions and chromatin-nuclear structural interactions, and their relationship to transcription. The subject matter of this disclosure, in some cases, relates to interrelated products, alternative solutions to specific problems, and / or a variety of different uses of one or more systems and / or articles.

[0008] Some aspects generally relate to systems and methods for imaging chromatin, for example, using multiplex FISH and, in some cases, multiple error-resistant FISH (MERFISH). Furthermore, some aspects generally relate to systems and methods for imaging and / or identifying at least 100 or at least 500 different genomic loci in a single cell. Some aspects generally relate to systems and methods for imaging chromatin, for example, using FISH.

[0009] In one set of implementations, the method includes associating multiple nucleic acid targets of the genome with multiple codewords, wherein the codewords include multiple locations and a value for each location; exposing a sample containing the genome to multiple nucleic acid probes; determining the binding of the nucleic acid probe within the sample for each of the multiple nucleic acid probes; generating a codeword corresponding to the binding of the multiple nucleic acid probes within the sample; and determining the identity of the nucleic acid targets based on the assigned codewords.

[0010] In another set of embodiments, the method includes determining the location of newly generated RNA within the cell nucleus; applying a ribonuclease to the cell nucleus; and determining the location of DNA within the cell nucleus.

[0011] In one set of embodiments, the method includes imaging chromatin in cells using MERFISH. In another set of embodiments, the method includes imaging at least 100 or at least 500 different genomic loci in a single cell.

[0012] According to one set of embodiments, the method includes associating a plurality of nucleic acid targets of a genome with a plurality of codewords; exposing a sample containing cells suspected of containing the genome to a plurality of nucleic acid probes, wherein at least some of the plurality of nucleic acid probes contain a first portion comprising a target sequence and a second portion comprising one or more readout sequences, wherein each readout sequence represents a value at a position within a plurality of codewords; exposing the sample to one or more adaptors in a round, wherein each adaptor contains a first portion substantially complementary to one of the readout sequences and a second portion comprising an identification sequence; exposing the sample to one or more readout probes in a round to determine one or more identification sequences, wherein each readout probe contains a first portion comprising a sequence substantially complementary to one of the identification sequences and a second portion comprising a signal transduction entity; determining signal transduction entities at at least some positions in the sample; and inactivating signal transduction entities at at least some positions in the sample; repeating the steps of exposing the sample to one or more adaptors and one or more readout probes in a round, determining signal transduction entities, and inactivating signal transduction entities, wherein one or more different signal transduction entities are used in each round; determining codewords at positions based on the determined signal transduction entities in the sample; and determining nucleic acid targets in the sample based on the codewords.

[0013] In another set of embodiments, the method includes associating a plurality of nucleic acid targets of the genome with a plurality of codewords; exposing a sample containing cells suspected of containing the genome to a plurality of nucleic acid probes, wherein at least some of the plurality of nucleic acid probes contain a first portion comprising a target sequence and a second portion comprising one or more readout sequences, wherein each readout sequence represents a value at a position within a plurality of codewords; exposing the sample to one or more adaptors in a round, wherein each adaptor contains a first portion substantially complementary to one of the readout sequences and a second portion comprising an identification sequence; exposing the sample to one or more readout probes in a round to determine one or more identification sequences, wherein each readout probe contains a first portion comprising a sequence substantially complementary to one of the identification sequences and a second portion comprising a signal transduction entity; determining signal transduction entities at at least some positions in the sample; and inactivating signal transduction entities at at least some positions in the sample; repeating the steps of exposing the sample to one or more adaptors and one or more readout probes in a round, determining signal transduction entities, and inactivating signal transduction entities, wherein at least one of the signal transduction entities is used for more than one round; determining codewords at positions based on the determined signal transduction entities in the sample; and determining nucleic acid targets in the sample based on the codewords.

[0014] In another set of embodiments, the method includes exposing a sample containing cells suspected of containing a genome to a plurality of nucleic acid probes, wherein at least some of the plurality of nucleic acid probes contain a first portion comprising a target sequence and a second portion comprising one or more readout sequences; exposing the sample to one or more adapters in a round, wherein each adapter contains a first portion substantially complementary to one of the readout sequences and a second portion comprising an identification sequence; exposing the sample to one or more readout probes in a round to determine one or more identification sequences, wherein each readout probe contains a first portion comprising a sequence substantially complementary to one of the identification sequences and a second portion comprising a signal transduction entity; determining signal transduction entities at at least some locations in the sample; and inactivating signal transduction entities at at least some locations in the sample; repeating the steps of exposing the sample to one or more adapters and one or more readout probes in a round, determining signal transduction entities, and inactivating signal transduction entities, wherein one or more different signal transduction entities are used in each round; and determining a nucleic acid target in the sample based on the signal transduction entities determined in each round.

[0015] In another set of embodiments, the method includes exposing a sample containing cells suspected of containing a genome to a plurality of nucleic acid probes, wherein at least some of the plurality of nucleic acid probes contain a first portion containing a target sequence and a second portion containing one or more readout sequences; exposing the sample to one or more readout probes in a round to determine one or more readout sequences, wherein each readout probe contains a first portion containing a sequence substantially complementary to one of the readout sequences and a second portion containing a signal transduction entity; determining signal transduction entities at at least some locations in the sample; and inactivating signal transduction entities at at least some locations in the sample; repeating the steps of exposing the sample to one or more readout probes in a round, determining signal transduction entities, and inactivating signal transduction entities, wherein one or more different signal transduction entities are used in each round; and determining a nucleic acid target in the sample based on the signal transduction entities determined in each round.

[0016] In another set of embodiments, the method includes exposing a sample containing cells suspected of containing a genome to a plurality of nucleic acid probes in one round, wherein at least some of the plurality of nucleic acid probes contain a first portion comprising a target sequence and a second portion comprising a signal transduction entity; identifying the signal transduction entity at at least some locations in the sample; and inactivating the signal transduction entity at at least some locations in the sample; repeating the steps of exposing the sample to the plurality of nucleic acid probes in one round, identifying the signal transduction entity, and inactivating the signal transduction entity, wherein one or more different signal transduction entities are used in each round; and determining a nucleic acid target in the sample based on the signal transduction entity determined in each round.

[0017] In one set of embodiments, the method includes associating a plurality of nucleic acid targets of the genome with a plurality of codewords, wherein the codewords include a plurality of positions and a value for each position, and the codewords form an error-checking and / or error-correcting code space, and wherein the plurality of nucleic acid targets are separated by at least 100,000 nucleotides within the genome; exposing a cell nucleus containing the genome to a plurality of nucleic acid probes, wherein at least some of the plurality of nucleic acid probes contain a first portion containing a target sequence and a second portion containing one or more read sequences, wherein each read sequence represents a value at a position within the codeword; for each of the plurality of nucleic acid probes, determining the binding of the nucleic acid probe within the cell nucleus; creating a codeword corresponding to the binding of the plurality of nucleic acid probes within the cell nucleus, wherein the numerical value of the codeword is based on the read sequences present on the nucleic acid probe; for at least some codewords, matching the codewords with valid codewords, wherein if no match is found, the codeword is discarded or error correction is applied to the codeword to form a valid codeword, the valid codeword being a plurality of codewords assigned to the plurality of nucleic acid targets; and using the valid codewords corresponding to the binding of the plurality of nucleic acid probes within the cell nucleus to determine the nucleic acid abundance and / or spatial distribution within the cell nucleus.

[0018] In another set of embodiments, the method includes associating a plurality of nucleic acid targets of the genome with a plurality of codewords, wherein the codewords include a plurality of positions and a value for each position, and the codewords form an error-checking and / or error-correcting code space, and wherein the plurality of nucleic acid targets of the genome are distributed such that each chromosome of the genome contains no more than 200 nucleic acid targets; exposing a cell nucleus containing the genome to a plurality of nucleic acid probes, wherein at least some of the plurality of nucleic acid probes contain a first portion containing a target sequence and a second portion containing one or more read sequences, wherein each read sequence represents a value at a position within a codeword; for each of the plurality of nucleic acid probes, determining the binding of the nucleic acid probe within the cell nucleus; creating a codeword corresponding to the binding of the plurality of nucleic acid probes within the cell nucleus, wherein the numerical value of the codeword is based on the read sequences present on the nucleic acid probes; for at least some codewords, matching the codewords with valid codewords, wherein if no match is found, the codeword is discarded or error correction is applied to the codeword to form a valid codeword, the valid codeword being a plurality of codewords assigned to the plurality of nucleic acid targets; and using the valid codewords corresponding to the binding of the plurality of nucleic acid probes within the cell nucleus to determine the nucleic acid abundance and / or spatial distribution within the cell nucleus.

[0019] According to another set of embodiments, the method includes associating 500 to 1500 nucleic acid targets of the genome with a plurality of codewords, wherein the codewords include a plurality of positions and a value for each position, and the codewords form an error-checking and / or error-correcting code space; exposing a cell nucleus containing the genome to a plurality of nucleic acid probes, wherein at least some of the plurality of nucleic acid probes contain a first portion containing a target sequence and a second portion containing one or more read sequences, wherein each read sequence represents a value at a position within the codeword; for each of the plurality of nucleic acid probes, determining the binding of the nucleic acid probe within the cell nucleus; creating codewords corresponding to the binding of the plurality of nucleic acid probes within the cell nucleus, wherein the numerical value of the codeword is based on the read sequences present on the nucleic acid probes; for at least some codewords, matching the codewords with valid codewords, wherein if no match is found, the codeword is discarded or error correction is applied to the codeword to form a valid codeword, the valid codewords being the plurality of codewords assigned to the plurality of nucleic acid targets; and using the valid codewords corresponding to the binding of the plurality of nucleic acid probes within the cell nucleus to determine the nucleic acid abundance and / or spatial distribution within the cell nucleus.

[0020] In another set of embodiments, the method includes associating multiple nucleic acid targets of the genome with multiple codewords, wherein the codewords include multiple positions and a value for each position, and the codewords form an error-checking and / or error-correcting code space, and wherein the multiple nucleic acid targets are separated by at least 100,000 nucleotides within the genome; exposing a cell nucleus containing the genome to multiple nucleic acid probes; and determining the abundance and / or spatial distribution of nucleic acids within the cell nucleus by using error-checking and / or error-correcting detection techniques to determine the binding of the multiple nucleic acid probes within the cell nucleus.

[0021] In another set of embodiments, the method includes associating multiple nucleic acid targets of the genome with multiple codewords; exposing a sample containing cells suspected of containing the genome to multiple nucleic acid probes, wherein at least some of the multiple nucleic acid probes contain a first portion comprising a target sequence and a second portion comprising one or more read sequences, wherein each read sequence represents a value at a position within the multiple codewords; exposing the sample to multiple adaptors, wherein at least some adaptors contain a first portion substantially complementary to one or more read sequences and a second portion comprising one or more identification sequences; exposing the sample to one or more read probes in a round to determine one or more identification sequences, wherein at least some read probes contain a first portion comprising a sequence substantially complementary to one of the identification sequences and a second portion comprising a signal transduction entity; determining signal transduction entities at at least some positions in the sample; and inactivating signal transduction entities at at least some positions in the sample; repeating the steps of exposing the sample to one round, determining signal transduction entities, and inactivating signal transduction entities, wherein no more than 10 different signal transduction entities are used in all rounds; determining codewords at positions based on the determined signal transduction entities in the sample; and determining nucleic acid targets in the sample based on the codewords.

[0022] According to another embodiment, the method includes associating multiple nucleic acid targets of the genome with multiple codewords; exposing a sample containing cells suspected of containing the genome to multiple nucleic acid probes, wherein at least some of the multiple nucleic acid probes contain a first portion comprising a target sequence and a second portion comprising one or more read sequences, wherein each read sequence represents a value at a position within the multiple codewords; exposing the sample to multiple adaptors, wherein at least some adaptors contain a first portion substantially complementary to one or more read sequences and a second portion comprising one or more identification sequences; exposing the sample to one or more read probes in a round to determine one or more identification sequences, wherein at least some read probes contain a first portion comprising a sequence substantially complementary to one of the identification sequences and a second portion comprising a signal transduction entity; determining signal transduction entities at at least some positions in the sample; and inactivating signal transduction entities at at least some positions in the sample; repeating the steps of exposing the sample to one round, determining signal transduction entities, and inactivating signal transduction entities, wherein at least one of the signal transduction entities is used in more than one round; determining a codeword at a position based on the determined signal transduction entity in the sample; and determining a nucleic acid target in the sample based on the codeword.

[0023] According to another set of implementation methods, the method includes determining the location of newly generated RNA within the cell nucleus; determining the location of DNA within the cell nucleus; and determining the location of nuclear spots within the cell nucleus.

[0024] In another set of embodiments, the method includes determining the location of nascent RNA within the cell nucleus; determining the location of DNA within the cell nucleus; and determining the location of a protein within the cell nucleus. In yet another set of embodiments, the method includes determining the location of nascent RNA within the cell nucleus; determining the location of DNA within the cell nucleus; and determining the location of a nucleic acid within the cell nucleus, wherein the nucleic acid is neither nascent RNA nor DNA.

[0025] Some aspects include methods of manufacturing one or more embodiments described herein. Additionally, some aspects include methods of using one or more embodiments described herein.

[0026] Other advantages and novel features of this disclosure will become apparent from the following detailed description of various non-limiting embodiments when considered in conjunction with the accompanying drawings. Brief description of the attached figures

[0027] Non-limiting embodiments of the invention will be described by way of example with reference to the accompanying drawings, which are schematic and not intended to be drawn to scale. In the drawings, each identical or nearly identical component shown is generally represented by a single number. For clarity, not every component is labeled in every drawing, and not every component of each embodiment of the invention is shown where it is not necessary to provide explanation so that those skilled in the art can understand the disclosure.

[0028] Figure 1A-1I Genome-scale chromatin imaging is shown according to certain implementation schemes;

[0029] Figures 2A-2E Cross-chromosomal contact enrichment is shown in another embodiment;

[0030] Figures 3A-3H In another implementation, genome-scale imaging of chromatin and transcriptional activity is shown against the backdrop of nuclear structure;

[0031] Figures 4A-4F This illustrates cross-chromosome interactions between active chromatin in another embodiment;

[0032] Figures 5A-5E A saturable amplification system is shown in one embodiment;

[0033] Figures 6A-6B The contact frequency matrix is ​​shown in yet another implementation scheme;

[0034] Figures 7A-7C This illustrates subchromosomal structures derived from genome-scale imaging and compared with overall Hi-C data in yet another implementation scheme;

[0035] Figure 8In another embodiment, the reproducibility of chromatin imaging experiments between repeated experiments is shown;

[0036] Figures 9A-9B This shows different spatial distributions within a single cell in certain implementations;

[0037] Figures 10A-10B Imaging of nascent RNA transcripts in other implementation schemes is shown;

[0038] Figure 11 This demonstrates the association between the compartment-B locus and the nuclear lamina in some implementation schemes;

[0039] Figure 12 The association between the compartment-A locus and nuclear spots was shown in some implementation schemes;

[0040] Figures 13A-13C The changes in the association between the nuclear lamina and nuclear spots are shown after transcriptional repression in another implementation:

[0041] Figure 14 In another embodiment, the local density of transchromosomal A loci near each imaged locus is shown;

[0042] Figures 15A-15B This is shown as an enrichment of active-active transchromosomal interactions between chromatin loci in another embodiment;

[0043] Figures 16A-16B This is shown as an enrichment of active-active transchromosomal interactions in another embodiment;

[0044] Figure 17A-17M This is shown in one implementation, through sequential hybridization and characterization of high-resolution whole-chromosome tracing of chromatin domains in a single cell;

[0045] Figure 18A-18I In another implementation, the compartmental structure of a single chromosome and the relationship between transcriptional activity and local chromatin content are shown;

[0046] Figures 19A-19H This demonstrates the dependence of interdomain interactions on A / B composition and genomic distance in yet another implementation scheme;

[0047] Figures 20A-20H This is shown in yet another implementation of genome-scale chromatin imaging via large-scale multiplex FISH;

[0048] Figures 21A-21E This shows the enrichment of active-active chromatin interactions in cross-chromosomal interactions according to one embodiment;

[0049] Figure 22A-22JThis demonstrates multimodal genome-scale imaging of chromatin and transcriptional activity against the background of nuclear structure, according to another implementation scheme;

[0050] Figures 23A-23D This demonstrates the correlation between transcriptional activity and local enrichment of transchromatin active chromatin in yet another implementation scheme;

[0051] Figure 24A-24N The results show high-resolution whole-chromosome tracing via sequential hybridization in yet another implementation, as well as overall statistics of Chr21 structural features compared to Hi-C;

[0052] Figure 25A-25G The overall A / B compartment analysis of Chr21 and Chr2 is shown in yet another implementation scheme;

[0053] Figure 26A-26J This demonstrates the measurement of RNA and DNA FISH probe crosstalk in yet another implementation.

[0054] Figure 27A-27J Genome-scale imaging performed using combined FISH in one implementation: localization error, reproducibility, and comparison with Hi-C;

[0055] Figures 28A-28B According to another embodiment, the compartment-A and compartment-B loci show different spatial distributions in the cell nucleus;

[0056] Figures 29A-29F The effects of transcriptional repression on transchromatin interactions and nucleosome association rates at chromatin loci are shown in yet another implementation; and

[0057] Figures 30A-30D This is shown in yet another implementation of enrichment of transchromatin active chromatin interactions in different nuclear environments. Invention Details

[0058] This invention generally relates to genomics. Some embodiments involve high-throughput imaging of the 3D organization of a genome or a portion of a genome in sequence space. Some embodiments involve imaging the 3D organization of a genome or a portion of a genome in the context of transcriptional activity and nuclear structure. Furthermore, some embodiments involve chromatin structure, 3D chromatin organization, transchromatin interactions, and chromatin-nuclear structure interactions, and their relationship to transcription. Additionally, various embodiments involve imaging methods that allow mapping of the 3D organization of a genome or a portion of a genome in the context of nuclear structure and transcriptional activity. Some embodiments involve large-scale multiplex fluorescence in situ hybridization methods for imaging chromatin loci and / or nascent RNA transcripts at the chromosome or genome scale. In some cases, hundreds of genomic loci can be imaged simultaneously. In some cases, the transcriptional activity of approximately 1000 genomic loci and / or approximately 1000 genes within these loci can be imaged simultaneously under various nuclear structures. In some cases, chromatin domains and compartments can be observed. In some cases, extensive transchromatin interactions rich in active chromatin interactions in a transcription-related manner can be observed. In some cases, transcription-dependent chromatin interactions with nuclear spots and nuclear lamina can be observed in the genome.

[0059] The three-dimensional (3D) organization of chromatin regulates many genomic functions. The lack of tools that allow direct visualization of chromatin organization in its natural context at both the chromosome and genome scales hinders our understanding of 3D genome organization. Therefore, in some embodiments, a multiplex FISH approach is described, involving sequential imaging across multiple hybridization rounds, e.g., such that each round uses monochrome, dual-color, or tricolor imaging to target one, two, or three genomic loci. In other embodiments, a combined FISH approach is described, in which numerous chromatin loci are simultaneously imaged in each round, and their distinct identities are determined based on the combination of rounds in which they appear. This is typically based on MERFISH and other methods, such as those discussed in International Patent Application Publication No. WO2016 / 018960 entitled “Systems and Methods for Determining Nucleic Acids”; and International Patent Application Publication No. WO2016 / 018963 entitled “Probe Library Construction,” each of which is incorporated herein by reference in its entirety. Methods such as those discussed in this paper can be used to image different chromatin loci in single cells and can provide in-depth understanding of chromatin structure, their relationship with transcription, their interaction with nuclear proteins, and so on.

[0060] Some aspects generally involve systems and methods for imaging, for example, chromosomes or chromatin in cells, using multiplex FISH or other techniques (in some cases MERFISH, including those described herein). Furthermore, certain embodiments generally involve systems and methods for imaging and / or identifying at least 100 different genomic loci, at least 500 different genomic loci, or at least 1,000 different genomic loci in a single cell. In some cases, other parts of the cell or the nucleus can be identified, such as RNA present in the nucleus, such as nascent RNA, nuclear spots, nucleoli, nuclear lamina, other nuclear structures, or proteins. As a non-limiting example, for the nucleus, the location of chromosomes or chromatin, nascent RNA, nuclear spots, nucleoli, and / or nuclear lamina can be identified.

[0061] Some implementations involve identifying a sample, which may include cell cultures, cell suspensions, biological tissues, biopsies, organisms, etc. The sample may also be cell-free, but in some cases may still contain nucleic acids. If the sample contains cells, the cells can be human cells or any other suitable cells, such as mammalian cells, fish cells, insect cells, plant cells, etc. In some cases, more than one cell may be present.

[0062] In the sample, the targets to be identified may include nucleic acids, proteins, etc. For example, these may be present within the cell nucleus within the sample. In some embodiments, intracellular chromatin can be identified, for example, relative to the cell's nuclear structures, including nuclear spots, nucleoli, nuclear lamina, or nuclear structures or proteins. In some cases, chromatin loci and / or RNA transcripts can be identified intracellularly, for example, at the chromosome or genome scale.

[0063] An example of this approach will now be discussed. However, it should be understood that this approach is presented as an explanation rather than a limitation; other aspects and implementations are also discussed herein. In one set of implementations, nucleic acids are measured within a cell, such as the cell nucleus. These typically include DNA (e.g., genomic DNA, which may exist as chromatin, for example, packaged with proteins such as histones) and RNA (e.g., at the beginning of the transcription phase when DNA is transcribed into RNA; such RNA in the nucleus is sometimes referred to as nascent RNA). Compared to techniques for detecting RNA, which may be present anywhere within the cell, DNA is highly packed within the cell nucleus, making it more difficult to determine its structure. For example, DNA can be packaged within the cell as chromosomes or chromatin, and such DNA can often be tangled or tightly packed together within the cell nucleus. Therefore, in some implementations, DNA targets can be selected for spatial separation.

[0064] In some cases, samples are hybridized with nucleic acid probes in multiple rounds, with one or more rounds using monochromatic or multicolor imaging to target one or more nucleic acid targets. In some cases, the identification of nucleic acid targets is determined based on which round and / or which color channel they were imaged in. In some cases, the location of nucleic acid targets is determined. In some cases, at least 50, at least 100, at least 500, at least 1000, at least 5000, or at least 10,000 nucleic acid targets are determined. In some cases, nucleic acid targets are genomic loci. In some cases, nucleic acid targets are genomic loci and / or newly generated RNA transcripts. In some cases, the location of genomic loci is used to determine the three-dimensional organization of chromatin or the genome in cells.

[0065] In some cases, primary nucleic acid probes are designed to target nucleic acids within cells, such as the cell nucleus. Each probe contains a target sequence that binds to one of the nucleic acid targets. The probe may also contain a portion containing one or more "readout sequences," which can be used to identify and locate the primary nucleic acid probe. In some embodiments, the primary nucleic acid probe may contain multiple readout sequences. These can be read individually using one or more rounds of secondary nucleic acid probes (called readout probes, which are capable of binding to the readout sequences of the primary nucleic acid probe). The readout probes may also contain signal transduction entities, such as fluorescent entities, which can be determined using various microscopy techniques. In some cases, multiple rounds of readout probes can be applied sequentially, such that one type of readout probe is applied to the sample and fluorescence is determined within the sample, then the readout probe or the signal transduction entity on the readout probe is inactivated or removed and the next type of readout probe is applied. In some cases, the location within the sample may be associated with multiple readout probes, and this information can be digitized for analysis.

[0066] In some cases, multiple rounds of readout probes can be applied sequentially, such that in each round more than one type of readout probe is applied to the sample and / or multicolor imaging is used to determine fluorescence within the sample, followed by inactivation or removal of the readout probes and / or signal transduction entities on the readout probes, and application of the next set of more than one type of readout probes. In some cases, the location within the sample may be associated with multiple readout probes, and this information can be digitized for analysis.

[0067] In some cases, the location of primary nucleic acid probes and nucleic acid targets can be determined using one or more rounds of readout probes. For example, there can be at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 150, at least 200, at least 250, at least 300, at least 400, at least 500, at least 600, at least 700, at least 800, at least 900, at least 1000 rounds of readout probes. Therefore, in some cases, a sample can be exposed to multiple rounds of applied readout probes, probes within the sample can be identified (e.g., using signal transduction entities, as described herein), and secondary nucleic acid probes can be removed or inactivated.

[0068] Furthermore, it should be understood that readout probes do not have to be all different. In some cases, the same readout probe from more than one round can be used, for example, as a control to determine whether any degradation and / or migration has occurred in the sample due to the effects of supplying multiple rounds of nucleic acids or other chemicals, such as over time.

[0069] In some cases, samples are hybridized with nucleic acid probes in multiple rounds, with each round being imaged in monochrome or multicolor. In some cases, the identity of nucleic acid targets is determined based on the rounds in which they are imaged and / or the combination of color channels. In some cases, the location of nucleic acid targets is determined. In some cases, at least 50, at least 100, at least 500, at least 1000, at least 5000, or at least 10,000 nucleic acid targets are determined. In some cases, nucleic acid targets are genomic loci. In some cases, nucleic acid targets are genomic loci and / or newly generated RNA transcripts. In some cases, the location of genomic loci is used to determine the three-dimensional organization of chromatin or the genome in cells.

[0070] In some cases, primary nucleic acid probes (also called coding probes) are designed to target nucleic acids within cells, such as the cell nucleus. Each probe contains a target sequence that binds to one of the nucleic acid targets. The probe may also contain portions comprising one or more "readout sequences" that can be used to identify and locate the primary or coding nucleic acid probe. In some embodiments, the primary or coding nucleic acid probe may contain multiple readout sequences. These can be read individually using one or more rounds of readout probes, each capable of binding to the readout sequence of the primary or coding nucleic acid probe. The readout probe may also contain a signal transduction entity, such as a fluorescent entity, which can be determined using various microscopy techniques. In some cases, multiple rounds of readout probes can be applied sequentially, such that one type of readout probe is applied to a sample and fluorescence is determined within the sample, then the readout probe or the signal transduction entity on the readout probe is inactivated or removed, and the next type of readout probe is applied. In some cases, the location within the sample may be associated with multiple readout probes, and this information can be digitized for analysis. In some cases, multiple rounds of readout probes can be applied sequentially, such that in each round more than one type of readout probe is applied to the sample and multicolor imaging is used to determine the fluorescence within the sample. Then, the readout probe or the signal transduction entity on the readout probe is inactivated or removed, and the next set of more than one type of readout probe is applied. In some cases, the location within the sample may be associated with multiple readout probes, and this information can be digitized for analysis.

[0071] In some cases, one or more rounds of readout probes can be used to determine the location of primary or coding nucleic acid probes and nucleic acid targets. For example, there can be at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 12, at least 16, at least 20, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 200, at least 300, at least 400, at least 500, at least 600, at least 700, at least 800, at least 900, at least 1,000 rounds of readout probes. Therefore, in some cases, a sample can be exposed to multiple rounds of applied readout probes to identify probes within the sample (e.g., using signal transduction entities, as described herein) and to remove or inactivate secondary nucleic acid probes.

[0072] In some implementations, primary or coding nucleic acid probes can be designed so that different combinations of readout sequences can be used to identify different targets within a sample, without requiring each readout sequence to be unique. As a non-limiting example, if the set of primary or coding nucleic acid probes targeting each nucleic acid target contains only two readout sequences, then for four possible readout sequences A, B, C, and D, up to six different targets can be identified, for example, corresponding to AB, AD, CB, CD, AC, and DB.

[0073] However, in some implementations, not all possible readout sequence combinations will be used. Instead, some combinations may not be assigned to any targets in the cell nucleus; for example, primary or coding nucleic acid probes with these combinations may not be used. In some cases, valid combinations of readout sequences for primary or coding nucleic acid probes can be arranged to form an error-checking and / or error-correcting code space. Using this method, the determination of readout sequences in a sample that do not correspond to valid primary nucleic acid probes can be determined as erroneous using error checking, and in some cases, even corrected using error correction, for example, to correspond to valid primary nucleic acid probes.

[0074] While these methods have been described previously, for example in International Patent Application Publication No. WO2016 / 018960 entitled "Systems and Methods for Determining Nucleic Acids" and International Patent Application Publication No. WO2016 / 018963 entitled "Probe Library Construction," these methods have not been applied to imaging DNA in the more confined environment within the cell nucleus. As previously mentioned, unlike other parts of the cell, the cell nucleus contains a very high percentage of nucleic acids, including almost all of the genomic DNA, and typically high concentrations of RNA (e.g., nascent RNA).

[0075] Therefore, to access DNA within the cell nucleus, targets for primary or coding nucleic acid probes can be selected, allowing binding within the nucleus to occur in a spatially segregated manner. For example, targets can be selected such that they are separated within the genomic space, for instance, at least 10,000 bp, at least 30,000 bp, at least 100,000 bp, at least 300,000 bp, at least 1,000,000 bp, or such that the genomic space contains no more than 100, no more than 200, no more than 300, no more than 500, no more than 1,000, no more than 5,000, no more than 10,000, no more than 50,000, or no more than 100,000 nucleic acid targets. In some cases, more than one type of fluorescent probe or "color" can also be used, for example, to allow for the determination of more targets within the cell nucleus.

[0076] In some embodiments, cells and / or cell nuclei may also be modified to allow such probes to reach the nucleic acids therein. For example, cells may be permeabilized or “fixed” to allow nucleic acid probes to enter. Furthermore, in some embodiments, DNA may be denatured, for example by heating, to allow primary or coding nucleic acid probes to more easily access the DNA. This is generally not suitable for RNA assays because RNA is single-stranded, while DNA is typically double-stranded. Additionally, in some embodiments, RNA within the cell nucleus must be removed and / or inactivated before DNA can be studied, for example, to prevent DNA-targeting probes from binding to RNA. For example, in some cases, an enzyme (e.g., ribonuclease) may be applied to the cell nucleus to prevent RNA interference with DNA assays.

[0077] Furthermore, it is important to note that in some embodiments, RNA within the cell nucleus can also be measured. This can be particularly valuable, for example, when studying the spatial location of DNA and RNA within the cell nucleus and how they are related. Therefore, in one set of embodiments, RNA within the cell nucleus can be measured prior to the removal or inactivation of RNA as described above, for example, similar to those described above for genomic DNA.

[0078] Furthermore, in some embodiments, intracellular proteins, such as those in the cell nucleus, can also be measured. Examples include, but are not limited to, nuclear spots, nucleoli, or histones. A variety of methods for protein measurement can be used. For example, in one set of embodiments, immunofluorescence assays can be used. In another set of embodiments, a "sandwich assay" can be used, in which a primary antibody capable of specifically binding to the nucleoprotein is applied, followed by a secondary antibody capable of specifically binding to the primary antibody, wherein the secondary antibody contains a signal transduction entity, such as a fluorescent entity. Such protein measurements can be performed on the same sample as described above or on the same cell nucleus, for example, before or after the measurement of nucleic acids within the cell nucleus. Thus, in some cases, proteins and nucleic acids within the cell nucleus can be measured, for example, spatially.

[0079] The above discussion is a non-limiting example of an implementation scheme that can be used to determine nucleic acids within the cell nucleus, such as genomic DNA and / or nascent RNA. However, other implementation schemes are also possible. Therefore, more generally, these aspects relate to a variety of systems and methods targeting nucleic acids.

[0080] As mentioned, in some embodiments, one, two, or more of DNA, RNA, and proteins can be measured within cells, such as the cell nucleus. The nucleic acid to be measured within the cell nucleus can include, for example, DNA (e.g., genomic DNA), RNA, or other nucleic acids present within the cell (or other sample). The nucleic acid can be endogenous to the cell or added to the cell. For example, the nucleic acid can be viral or artificially produced. In some cases, the nucleic acid to be measured can be expressed by the cell. In some embodiments, the nucleic acid is RNA. The RNA can be coding and / or non-coding RNA. For example, RNA can encode a protein. Non-limiting examples of RNAs that can be studied within cells include mRNA, siRNA, rRNA, miRNA, tRNA, lncRNA, snoRNA, snRNA, exRNA, piRNA, etc.

[0081] In one set of implementations, all or at least significant portions of the cellular genome can be determined. The determined genomic segments can be continuous or distributed across the genome. For example, in some cases, at least four genomic segments are determined within the cell; in others, at least three, four, seven, eight, twelve, fourteen, fifteen, sixteen, twenty-two, thirty, thirty-one, thirty-two, fifty, sixty-three, sixty-four, seventy-five, seventy-five, sixty-six, sixty-seven, sixty-four, seventy-five, seventy-five, sixty-seven ... 500, at least 1,000, at least 1,500, at least 2,000, at least 2,500, at least 3,000, at least 4,000, at least 5,000, at least 7,500, at least 10,000, at least 12,000, at least 15,000, at least 20,000, at least 25,000, at least 30,000, at least 40,000, at least 50,000, at least 75,000, or at least 100,000 genomic segments.

[0082] In some cases, the entire genome of a cell can be determined. It should be understood that the genome typically includes all DNA molecules produced within the cell, not just chromosomal DNA. Therefore, for example, in some cases, the genome may also include mitochondrial DNA, chloroplast DNA, plasmid DNA, etc., for example, in addition to (or instead of) chromosomal DNA. In some embodiments, at least about 0.01%, at least about 0.1%, at least about 1%, at least about 5%, at least about 10%, at least about 15%, at least about 20%, at least about 25%, at least about 30%, at least about 40%, at least about 50%, at least about 60%, at least about 70%, at least about 80%, at least about 90%, or 100% of the cell's genome can be determined.

[0083] Furthermore, in some implementations, most nucleic acids within the cell or cell nucleus can be studied. For example, in some cases, RNA within the cell nucleus, such as newly generated RNA, can be measured. Additionally, in some cases, sufficient amounts of RNA present within the cell can be measured to generate a partial or complete transcriptome of the cell. In some cases, at least four types of RNA (e.g., mRNA, newly generated RNA, etc.) within the cell or cell nucleus are measured, and in other cases, at least 3, at least 4, at least 7, at least 8, at least 12, at least 14, at least 15, at least 16, at least 20, at least 22, at least 30, at least 31, at least 32, at least 50, at least 63, at least 64, at least 72, at least 75, at least 100, at least 127, at least 128, or at least 140 can be measured within the cell or cell nucleus. At least 255, at least 256, at least 500, at least 1,000, at least 1,500, at least 2,000, at least 2,500, at least 3,000, at least 4,000, at least 5,000, at least 7,500, at least 10,000, at least 12,000, at least 15,000, at least 20,000, at least 25,000, at least 30,000, at least 40,000, at least 50,000, at least 75,000, or at least 100,000 types of RNA.

[0084] In some cases, the cellular transcriptome can be measured. It should be understood that the transcriptome typically includes all RNA molecules produced within the cell, not just mRNA. Therefore, for example, in some cases, the transcriptome may also include rRNA, tRNA, siRNA, etc. In some embodiments, at least about 0.01%, at least about 0.1%, at least about 1%, at least 5%, at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, or 100% of the cellular transcriptome can be measured. Furthermore, in some cases, the nuclear transcriptome can be measured.

[0085] Furthermore, in some embodiments, other targets to be measured may include targets linked to nucleic acids, proteins, etc. For example, in one set of embodiments, a binding entity capable of recognizing the target may be conjugated to a nucleic acid probe. The binding entity can be any entity capable of recognizing the target (e.g., specifically or non-specifically). Non-limiting examples include enzymes, antibodies, receptors, complementary nucleic acid chains, aptamers, etc. For example, an oligonucleotide-linked antibody can be used to measure the target. The target is capable of binding to the oligonucleotide-linked antibody, and the oligonucleotide is measured as discussed herein.

[0086] The determination of targets, such as nucleic acids in cells or other samples, can be qualitative and / or quantitative. Furthermore, the determination can also be spatial; for example, the location of nucleic acids or other targets within cells or other samples can be determined in two or three dimensions. In some embodiments, the location, quantity, and / or concentration of nucleic acids or other targets within cells or other samples can be determined.

[0087] As mentioned, in one set of implementations, DNA within the cell nucleus, such as the cell's genomic DNA, can be studied, for example, using nucleic acid probes as discussed herein, including, for example, sequential imaging or combined imaging utilizing error detection and / or error correction codes.

[0088] In some implementations, intracellular or nuclear DNA targets or codes associated with DNA targets can be selected such that the targets are spatially separated in each round of imaging, for example, in genomic space or in physical space based on chromatin organization knowledge (e.g., organizing chromosomes into compact regions). For example, this could be useful for identifying different intracellular targets within the cell nucleus, for example, using techniques such as those discussed herein.

[0089] Targets within the genomic space can be selected using any suitable technique (e.g., randomly or with a substantially uniform probability distribution). In some embodiments, targets can be selected individually to ensure spatial separation. Furthermore, in some embodiments, targets can be selected as those within the genome of interest, e.g., for a specific study.

[0090] For example, in some implementations, targets can be selected within the genomic space such that the cell nucleus will have no more than a certain number of nucleic acid targets. For example, targets can be selected such that the genomic space contains no more than 100,000, no more than 10,000, no more than 8,000, no more than 6,000, no more than 5,000, no more than 4,000, no more than 3,000, no more than 2,000, no more than 1,500, no more than 1,000, no more than 900, no more than 800, no more than 700, no more than 600, no more than 500, no more than 400, no more than 300, no more than 200, no more than 100 nucleic acid targets, no more than 30 nucleic acid targets, or no more than 10 nucleic acid targets. Furthermore, in some embodiments, targets can be selected such that the genomic space contains at least 10, at least 30, at least 50, at least 100, at least 200, at least 300, at least 500, at least 1,000, at least 1,500, at least 2,000, at least 3,000, at least 5,000, at least 10,000, at least 100,000, etc. In some embodiments, any combination of these is also possible; for example, there can be 30 to 100, 3,000 to 5,000, 500 to 1,500, etc., nucleic acid targets. Targets such as those discussed herein can be selected (e.g., selectively, randomly, etc.).

[0091] As another example, in some implementations, targets can be selected such that chromosomes within the genome have no more than a certain number of nucleic acid targets (e.g., genomic loci). For example, targets can be selected such that each chromosome has no more than 10,000, no more than 1,000, no more than 500, no more than 400, no more than 300, no more than 200, no more than 150, no more than 125, no more than 100, no more than 90, no more than 80, no more than 70, no more than 60, no more than 50, no more than 40, no more than 30, no more than 20, or no more than 10 nucleic acid targets. In some cases, targets can be selected to have at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 125, at least 150, at least 200, at least 300, at least 400, at least 1,000, at least 10,000, etc. In some cases, combinations of these can be chosen; for example, chromosomes can have 30 to 50, 40 to 100, 50 to 60, 30 to 80, etc. Furthermore, different chromosomes can independently have the same or different numbers of nucleic acid targets, for example, including the ranges described herein.

[0092] Targets can be selected (e.g., selectively, randomly, etc.), such as those discussed herein. As a non-limiting example, nucleic acid targets within the genome can be selected to have specific structural or functional properties, such as promoters, enhancers, and loci bound by specific nuclear structural proteins. In some cases, some or all of the nucleic acid targets can be nucleic acid targets specific to their respective chromosomes.

[0093] In yet another implementation, targets can be selected to be separated by a minimum number of nucleotides, for example, to promote the spatially separated distribution of targets. For example, targets can be selected within the genome such that each target is separated by at least 1,000, at least 3,000, at least 5,000, at least 10,000, at least 30,000, at least 50,000, at least 100,000, at least 300,000, at least 500,000, at least 1,000,000, at least 3,000,000, at least 5,000,000, at least 10,000,000 nucleotides. Furthermore, in some embodiments, targets can be selected within the genome such that each target is spaced no more than 10,000,000, 5,000,000, 3,000,000, 1,000,000, 500,000, 300,000, 100,000, 50,000, 30,000, 100,000, 50,000, 30,000, or 10,000 nucleotides apart. In some embodiments, any combination of these is also possible; for example, targets may be spaced 30,000 to 100,000, 3,000,000 to 5,000,000, 500,000 to 1,000,000 nucleotides apart. Such targets can be selected, for example, selectively, randomly, etc., as discussed herein.

[0094] Furthermore, in one set of embodiments, RNA within the cell nucleus, such as newly generated RNA of the cell, can be studied, for example, in place of or attached to the nuclear DNA as described above. In some cases, for example, RNA within the cell nucleus can be measured, and then DNA within the cell nucleus can be measured.

[0095] In some cases, RNA can be removed or inactivated after RNA assays and before DNA assays. This facilitates the separation of DNA and RNA assays, for example, because RNA signals that could complicate DNA assays are avoided. Examples of methods for removing or inactivating RNA include the use of ribonucleases, such as endonucleases or exonucleases. Specific, non-limiting examples include ribonuclease A, ribonuclease H, ribonuclease III, ribonuclease L, ribonuclease P, ribonuclease PhyM, ribonuclease T1, ribonuclease T2, ribonuclease U2, ribonuclease V, PNPase, ribonuclease PH, ribonuclease R, ribonuclease D, ribonuclease T, oligoribonucleases, exonucleases I and II, etc.

[0096] However, it should be understood that in other embodiments, DNA may be measured before RNA, and / or both may be measured simultaneously. For example, DNA may be removed or inactivated after measurement using deoxyribonucleases such as exonucleases or endonucleases. Examples include, but are not limited to, DNase I, DNase II, DNase IV, UvrABC endonucleases, etc. As another example, DNA can be degraded by exposure to restriction endonucleases. Many such nucleases are commercially available.

[0097] RNA within the cell nucleus can be determined using any suitable technique, and can be determined using the same or different techniques used to determine DNA within the cell nucleus. In one embodiment, MERFISH can be used to determine RNA. See, for example, International Patent Application Publication No. WO2016 / 018960 entitled "Systems and Methods for Determining Nucleic Acids"; and International Patent Application Publication No. WO2016 / 018963 entitled "Probe Library Construction," each of which is incorporated herein by reference in its entirety. In another embodiment, multiple nucleic acid probes can be used to determine RNA, for example, as discussed herein. For example, in some embodiments, nucleic acids such as those encoding nucleic acid probes, primary amplified nucleic acids, secondary amplified nucleic acids, etc., can be used to determine RNA, as discussed below. In some cases, nucleic acid probes may define error detection and / or error correction codes, for example, as discussed herein.

[0098] In some implementations, DNA, such as genomic DNA, can be determined using nucleic acids, such as coding nucleic acid probes, primary amplified nucleic acids, secondary amplified nucleic acids, etc., as described herein. In some cases, nucleic acid probes can define error detection and / or error correction codes, for example, as discussed herein.

[0099] Furthermore, in one set of implementations, for example, in addition to nucleic acids present in the cell nucleus, techniques such as those described above can be used to study proteins within the cell nucleus. Examples of proteins that can be studied include, but are not limited to, nuclear spots, nucleoli, nuclear lamina, or histones. Spots are structures rich in pre-messenger RNA splicing factors and may be located in the interchromatinous region of the nucleoplasm of mammalian cells. The nucleolus is a structure that forms around a highly transcribed genomic locus encoding ribosomal RNA (rRNA) and may be rich in rRNA and the associated transcriptional mechanisms. The nuclear lamina is a protein structure associated with the inner nuclear membrane and may be rich in intermediate filaments (laminins) and transcriptionally inactivated chromatin. Histones are proteins used to wrap or fold DNA into more compact complexes within the cell nucleus, forming chromatin.

[0100] Various methods can be used to determine proteins. For example, in one set of embodiments, immunofluorescence assays can be used. In another set of embodiments, a "sandwich assay" can be used, in which a primary antibody capable of specifically binding to nucleoproteins is applied, followed by a secondary antibody capable of specifically binding to the primary antibody, wherein the secondary antibody contains a signal transduction entity, such as a fluorescent entity, or a detectable oligonucleotide, for example, using a complementary oligonucleotide linked to a fluorescent entity. Such protein assays can be performed on the same samples as described above or on the same cell nuclei, for example, before or after the determination of nucleic acids within the cell nucleus. Thus, in some cases, proteins and nucleic acids within the cell nucleus can be determined, for example, spatially.

[0101] As mentioned, in the various embodiments described herein, such as those described herein, multiple nucleic acid probes can be used to determine one or more targets within cells or other samples, such as within the cell nucleus. Probes may comprise nucleic acids (or entities capable of hybridizing with nucleic acids, e.g., specifically), such as DNA, RNA, LNA (locked nucleic acid), PNA (peptide nucleic acid), and / or combinations thereof. Examples of nucleic acid probes include, but are not limited to, those described in International Patent Application Publication No. WO2016 / 018960 entitled “Systems and Methods for Determining Nucleic Acids”; and International Patent Application Publication No. WO2016 / 018963 entitled “Probe Library Construction,” each of which is incorporated herein by reference in its entirety. In some cases, additional components may also be present in the nucleic acid probe, for example, as discussed below. Furthermore, any suitable method can be used to introduce the nucleic acid probe into the cell, for example, to target its nucleus.

[0102] For example, in some embodiments, cells are immobilized prior to the introduction of nucleic acid probes, for example, to maintain the nucleic acid or other targets within the cell, such as in their position within the cell nucleus. Techniques for cell immobilization are known to those skilled in the art. As a non-limiting example, chemicals such as formaldehyde, paraformaldehyde, glutaraldehyde, ethanol, methanol, acetone, acetic acid, etc., can be used to immobilize cells. In one embodiment, HEPES-glutamate buffer-mediated organic solvent (HOPE) can be used to immobilize cells.

[0103] Furthermore, in some cases, cells (or other samples) may be fixed more than once, for example, during a relatively long experimental period. For instance, samples can be refixed after the start of the experiment, such as after exposing cell nuclei to multiple nucleic acid probes. Cells or other samples can be fixed at least every 7 days, at least every 4 days, at least every 2 days, at least daily, at least every 12 hours, at least every 6 hours, at least every 3 hours, etc. In some cases, this can be done between different rounds of exposure to nucleic acid probes (e.g., primary or secondary nucleic acid probes). In some cases, samples can be fixed a certain number of times, such as 2, 3, 4, 5, 6, 7, 8, 9, 10 times, or any other suitable number of times. If multiple fixations occur, these fixations can be performed independently using the same or different fixation techniques.

[0104] Nucleic acid probes can be introduced into cells (or other samples) using any suitable method. In some cases, cells may be fully permeabilized, allowing the probes to be introduced into the cells by flowing a fluid containing the probes around them. In some cases, cells may be fully permeabilized as part of a fixation process; in other embodiments, cells may be permeabilized by exposure to certain chemicals such as ethanol, methanol, Triton, etc. Furthermore, in some embodiments, techniques such as electroporation or microinjection can be used to introduce nucleic acid probes into cells or other samples.

[0105] Therefore, certain aspects generally involve nucleic acid probes introduced into cells (or other samples). Depending on the application, probes can contain any of a variety of entities capable of hybridizing with nucleic acids (typically via Watson-Crick base pairing), such as DNA, RNA, LNA, PNA, etc. Nucleic acid probes typically contain at least a portion of a target sequence capable of binding to a target (e.g., a nucleic acid target). In some cases, binding can be specific (e.g., through complementary binding). When introduced into cells or other systems, the target sequence may be able to bind to a specific target (e.g., nascent RNA, genomic DNA, mRNA, or other nucleic acids discussed herein). As discussed below, nucleic acid probes may also contain one or more readout sequences.

[0106] In some cases, more than one type of nucleic acid probe can be applied to a sample, for example, sequentially or simultaneously. For example, there can be at least 2, at least 5, at least 10, at least 25, at least 50, at least 75, at least 100, at least 300, at least 1,000, at least 3,000, at least 10,000, at least 30,000, at least 100,000, at least 300,000, or at least 1,000,000 distinguishable nucleic acid probes that can be applied to a sample (e.g., to cells to target their nuclei). In some cases, nucleic acid probes can be added sequentially. However, in some cases, more than one type of nucleic acid probe can be added simultaneously.

[0107] Nucleic acid probes may include one or more target sequences, which may be located anywhere within the probe. The target sequence may contain a region substantially complementary to a portion of a target (e.g., a nucleic acid target, which may be located within the cell nucleus). For example, in some cases, these portions may be at least 50%, at least 60%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 92%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or 100% complementary, for example, to produce specific binding. Typically, complementarity is determined based on Watson-Crick nucleotide base pairing.

[0108] In some cases, the length of the target sequence may be at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, at least 60, at least 65, at least 75, at least 100, at least 125, at least 150, at least 175, at least 200, at least 250, at least 300, at least 350, at least 400, or at least 450 nucleotides. In some cases, the length of the target sequence may not exceed 500, 450, 350, 250, 200, 175, 150, 125, 100, 75, 60, 65, 55, 45, 35, 20, or 10 nucleotides. Combinations of any of these are also possible; for example, the target sequence can have a length of 10 to 30 nucleotides, 20 to 40 nucleotides, 5 to 50 nucleotides, 10 to 200 nucleotides, or 25 to 35 nucleotides, 10 to 300 nucleotides, etc.

[0109] In some implementations, nucleic acid targets or codes associated with nucleic acid targets within cells or cell nuclei can be selected such that the targets are spatially separated in each round of imaging, for example in genomic space or in physical space based on prior knowledge of chromatin organization (e.g., organizing chromosomes into compact regions).

[0110] Furthermore, in some cases, the target sequence of a nucleic acid probe can be determined by referencing a target suspected of being present in cells or other samples (e.g., in the cell nucleus). For example, nucleic acid targets for proteins (e.g., nuclear spots, nuclear lamina, etc.) can be determined using the protein sequence, for example, by identifying the nucleic acid expressed to form the protein. In some cases, only a portion of the nucleic acid encoding the protein is used, for example, having the length discussed above.

[0111] According to certain implementation schemes, more than one target sequence can be used to identify a specific target. For example, multiple probes can be used sequentially and / or simultaneously, which are capable of binding or hybridizing to the same or different regions of the same target. Hybridization generally refers to the annealing process in which complementary single-stranded nucleic acids associate through Watson-Crick nucleotide base pairing (e.g., hydrogen bonds, guanine-cytosine, and adenine-thymine) to form double-stranded nucleic acids.

[0112] In some embodiments, the nucleic acid probe may also contain one or more "readout" sequences. The readout sequences can be used to identify the nucleic acid probe, for example, by association with a signal transduction entity, as discussed below. In some embodiments, the nucleic acid probe may contain 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16 or more, 20 or more, 24 or more, 32 or more, 40 or more, 48 or more, 50 or more, 64 or more, 75 or more, 100 or more, 128 or more readout sequences. The readout sequences can be located anywhere within the nucleic acid probe. If more than one readout sequence is present, the readout sequences may be positioned adjacent to each other and / or scattered with other sequences.

[0113] The readout sequence can be of any length. If more than one readout sequence is used, the readout sequences can independently have the same or different lengths. For example, the length of the readout sequence can be at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, at least 60, at least 65, at least 75, at least 100, at least 125, at least 150, at least 175, at least 200, at least 250, at least 300, at least 350, at least 400, or at least 450 nucleotides. In some cases, the length of the readout sequence may not exceed 500, 450, 400, 350, 300, 250, 200, 175, 150, 125, 100, 75, 60, 65, 60, 55, 50, 45, 40, 35, 30, 20, or 10 nucleotides. Any combination of these is also possible; for example, the readout sequence may have a length of 10 to 30 nucleotides, 20 to 40 nucleotides, 5 to 50 nucleotides, 10 to 200 nucleotides, or 25 to 35 nucleotides, 10 to 300 nucleotides, etc.

[0114] In some implementations, the readout sequence can be arbitrary or random. In some cases, the readout sequence is selected to reduce or minimize homology with other components of the cell or other sample, for example, such that the readout sequence itself does not bind or hybridize with other nucleic acids suspected of being present in the cell or other sample. In some cases, homology may be less than 10%, less than 8%, less than 7%, less than 6%, less than 5%, less than 4%, less than 3%, less than 2%, or less than 1%. In some cases, homology of fewer than 20 base pairs, fewer than 18 base pairs, fewer than 15 base pairs, fewer than 14 base pairs, fewer than 13 base pairs, fewer than 12 base pairs, fewer than 11 base pairs, or fewer than 10 base pairs may exist. In some cases, such base pairs are consecutive.

[0115] Furthermore, in some implementations, some or all readout sequences can be selected such that they do not exhibit specific binding to each other and / or to genomes or other nucleic acids suspected of being present in the sample. For example, a group of readout sequences can be “blasted” or tested for specific binding or complementarity. In some cases, readout sequences may not exhibit specific binding to each other, and / or the group of readout sequences may not have any one readout sequence in the group that is complementary to another readout sequence in the same group by more than 5, 6, 7, 8, 9, 10, etc.

[0116] In one set of embodiments, the nucleic acid probe group may contain a number of readout sequences, which may be the same as the number of nucleic acid targets to be measured in the sample; for example, each unique readout sequence corresponds to a unique target. In another set of embodiments, the nucleic acid probe group may contain a number of readout sequences, which may be fewer than the number of nucleic acid targets to be measured in the sample. Those skilled in the art will recognize that if there is one signal transduction entity and n readout sequences, it is generally possible to uniquely identify 2 n - One different nucleic acid target. However, it is not necessary to use all possible combinations. For example, a nucleic acid probe array can target 12 different nucleic acid targets but contain no more than 8 readout sequences. As another example, a nucleic acid probe array can target 140 different nucleic acid targets but contain no more than 16 readout sequences. Different nucleic acid targets can be identified separately by using different combinations of readout sequences within each probe. For example, a nucleic acid probe array can contain 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, or more readout sequences. In some cases, nucleic acid probe arrays can each contain the same number of readout sequences, although in other cases, different numbers of readout sequences may be present on various probes.

[0117] As a non-limiting example, a first nucleic acid probe may contain a first target sequence, a first readout sequence, and a second readout sequence, while a second, different nucleic acid probe may contain a second target sequence, the same first readout sequence, and a third readout sequence instead of the second readout sequence. Such probes can therefore be distinguished by identifying various readout sequences present or associated with a given probe or location, as discussed herein. For example, probes can be sequentially identified and encoded using "code words," as discussed below. Optionally, error detection and / or error correction can also be performed on the code words.

[0118] As another non-limiting example, a first nucleic acid probe group may comprise a first target sequence, a first readout sequence, and a second readout sequence, while a second, different nucleic acid probe group may comprise a second target sequence, the same first readout sequence, and a third readout sequence instead of the second readout sequence. Such probes can therefore be distinguished by identifying the various readout sequences present or associated with a given probe or location, as discussed herein. For example, probe groups can be sequentially identified and encoded using "code words," as discussed. Optionally, error detection and / or error correction can also be performed on the code words.

[0119] Furthermore, in some embodiments, nucleic acid probe groups can be prepared using only two or three of the four naturally occurring nucleotide bases, for example, by omitting all "G" or all "C" in the probe group. In some embodiments, sequences lacking "G" or "C" can form very few secondary structures and can contribute to more uniform and faster hybridization. Therefore, in some cases, nucleic acid probes may contain only A, T, and G; only A, T, and C; only A, C, and G; or only T, C, and G.

[0120] On the one hand, the readout sequence on a nucleic acid probe may be able to bind (e.g., specifically) to a corresponding recognition sequence on the primary amplified nucleic acid. Therefore, when the nucleic acid probe recognizes a target in a biological sample, such as a DNA or RNA target, the primary amplified nucleic acid can also associate with the target via the nucleic acid probe, where an interaction occurs between the readout sequence of the nucleic acid probe and the corresponding recognition sequence on the primary amplified nucleic acid, such as complementary binding. For example, the recognition sequence may be able to recognize the target readout sequence but essentially cannot recognize or bind to other non-target readout sequences. Depending on the application, the primary amplified nucleic acid may also contain any of a variety of entities capable of hybridizing with nucleic acids, such as DNA, RNA, LNA, and / or PNA. For example, such entities may form some or all of the recognition sequence.

[0121] In some cases, the recognition sequence can be substantially complementary to the target readout sequence. In some cases, the sequence can be at least 50%, at least 60%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 92%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or 100% complementary. Typically, complementarity is determined based on Watson-Crick nucleotide base pairing. The structure of the target readout sequence can include those previously described.

[0122] In some cases, the length of the identification sequence can be at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, at least 60, at least 65, at least 75, at least 100, at least 125, at least 150, at least 175, at least 200, at least 250, at least 300, at least 350, at least 400, or at least 450 nucleotides. In other cases, the length of the identification sequence may not exceed 500, 450, 350, 250, 200, 175, 150, 125, 100, 75, 60, 55, 45, 35, 20, or 10 nucleotides. Any combination of these is also possible; for example, the length of the recognition sequence can be 10 to 30 nucleotides, 20 to 40 nucleotides, 5 to 50 nucleotides, 10 to 200 nucleotides, or 25 to 35 nucleotides, 10 to 300 nucleotides, etc.

[0123] In some embodiments, the primary amplified nucleic acid may also include one or more readout sequences capable of binding to the secondary amplified nucleic acid, as discussed below. For example, the primary amplified nucleic acid may contain 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16 or more, 20 or more, 32 or more, 40 or more, 50 or more, 64 or more, 75 or more, 100 or more, 128 or more readout sequences. The readout sequences can be located anywhere within the primary amplified nucleic acid. If more than one readout sequence is present, the readout sequences may be positioned adjacent to each other and / or scattered with other sequences. In one embodiment, the primary amplified nucleic acid includes a recognition sequence at its first end and multiple readout sequences at its second end.

[0124] In some cases, the length of the readout sequence within the primary amplified nucleic acid can be at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, at least 60, at least 65, at least 75, at least 100, at least 125, at least 150, at least 175, at least 200, at least 250, at least 300, at least 350, at least 400, or at least 450 nucleotides. In other cases, the length of the readout sequence can not exceed 500, 450, 400, 350, 300, 250, 200, 175, 150, 125, 100, 75, 60, 65, 60, 55, 50, 45, 40, 35, 30, 20, or 10 nucleotides. Any combination of these is also possible; for example, the readout sequence can have a length of 10 to 20 nucleotides, 10 to 30 nucleotides, 20 to 40 nucleotides, 5 to 50 nucleotides, 10 to 200 nucleotides, or 25 to 35 nucleotides, 10 to 300 nucleotides, etc.

[0125] There can be any number of readout sequences within the primary amplified nucleic acid. For example, there can be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 or more readout sequences present in the primary amplified nucleic acid. If there is more than one readout sequence within the primary amplified nucleic acid, the readout sequences can be the same or different. For example, in some cases, the readout sequences can all be the same.

[0126] In some implementations, the primary amplified nucleic acid cluster can be prepared using only two or three of the four naturally occurring nucleotide bases, for example, by omitting all "G" or all "C" bases in the cluster. In some implementations, sequences lacking "G" or "C" can form very few secondary structures and can contribute to more uniform and faster hybridization. Therefore, in some cases, the primary amplified nucleic acid may contain only A, T, and G; only A, T, and C; only A, C, and G; or only T, C, and G.

[0127] In some cases, more than one type of primary amplified nucleic acid can be applied to a sample, for example, sequentially or simultaneously. For example, there can be at least 2, at least 5, at least 10, at least 25, at least 50, at least 75, at least 100, at least 300, at least 1,000, at least 3,000, at least 10,000, or at least 30,000 distinguishable primary amplified nucleic acids applied to the sample. In some cases, primary amplified nucleic acids can be added sequentially. However, in some cases, more than one primary amplified nucleic acid can be added simultaneously.

[0128] In one set of implementations, the readout sequence on the primary amplified nucleic acid may be able to bind (e.g., specifically) to a corresponding recognition sequence on the secondary amplified nucleic acid. Thus, when the nucleic acid probe recognizes a target in a biological sample, such as a DNA or RNA target, the secondary amplified nucleic acid can also associate with the target via the primary amplified nucleic acid, where an interaction occurs between the readout sequence of the primary amplified nucleic acid and the corresponding recognition sequence on the secondary amplified nucleic acid, such as complementary binding. For example, the recognition sequence on the secondary amplified nucleic acid may be able to recognize the readout sequence on the primary amplified nucleic acid, but essentially cannot recognize or bind other non-target readout sequences. Depending on the application, the secondary amplified nucleic acid may also contain any of a variety of entities capable of hybridizing with nucleic acids, such as DNA, RNA, LNA, and / or PNA. For example, such entities may form some or all of the recognition sequence.

[0129] In some cases, the recognition sequence on the secondary amplified nucleic acid can be substantially complementary to the readout sequence on the primary amplified nucleic acid. In some cases, the sequences can be at least 50%, at least 60%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 92%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or 100% complementary.

[0130] In some cases, the length of the recognition sequence on the secondary amplified nucleic acid can be at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, at least 60, at least 65, at least 75, at least 100, at least 125, at least 150, at least 175, at least 200, at least 250, at least 300, at least 350, at least 400, or at least 450 nucleotides. In other cases, the length of the recognition sequence may not exceed 500, 450, 350, 250, 200, 175, 150, 125, 100, 75, 60, 55, 45, 35, 20, or 10 nucleotides. Any combination of these is also possible; for example, the length of the recognition sequence can be 10 to 30 nucleotides, 20 to 40 nucleotides, 5 to 50 nucleotides, 10 to 200 nucleotides, or 25 to 35 nucleotides, 10 to 300 nucleotides, etc.

[0131] In some embodiments, the secondary amplified nucleic acid may contain a signal transduction entity and / or may contain one or more readout sequences capable of binding to the signal transduction entity, as discussed herein. For example, the secondary amplified nucleic acid may contain 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16 or more, 20 or more, 32 or more, 40 or more, 50 or more, 64 or more, 75 or more, 100 or more, 128 or more readout sequences capable of binding to the signal transduction entity. The readout sequences can be located anywhere within the secondary amplified nucleic acid. If more than one readout sequence is present, the readout sequences may be positioned adjacent to each other and / or scattered with other sequences. In one embodiment, the secondary amplified nucleic acid contains a recognition sequence at a first end and multiple readout sequences at a second end. This structure may also be the same as or different from the structure of the primary amplified nucleic acid.

[0132] In some cases, the length of the readout sequence within the secondary amplified nucleic acid can be at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, at least 60, at least 65, at least 75, at least 100, at least 125, at least 150, at least 175, at least 200, at least 250, at least 300, at least 350, at least 400, or at least 450 nucleotides. In other cases, the length of the readout sequence may not exceed 500, 450, 400, 350, 300, 250, 200, 175, 150, 125, 100, 75, 60, 65, 60, 55, 50, 45, 40, 35, 30, 20, or 10 nucleotides. Any combination of these is also possible; for example, the readout sequence within the secondary amplified nucleic acid can have a length of 10 to 20 nucleotides, 10 to 30 nucleotides, 20 to 40 nucleotides, 5 to 50 nucleotides, 10 to 200 nucleotides, or 25 to 35 nucleotides, 10 to 300 nucleotides, etc.

[0133] Any number of readout sequences can exist within the secondary amplified nucleic acid. For example, there can be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or more readout sequences within the secondary amplified nucleic acid. If more than one readout sequence exists within the secondary amplified nucleic acid, the readout sequences can be the same or different. For example, in some cases, the readout sequences can all be the same. Furthermore, the primary and secondary amplified nucleic acids can independently contain the same or different numbers of readout sequences.

[0134] Secondary amplified nucleic acid clusters can be prepared using only two or three of the four naturally occurring nucleotide bases, and in some embodiments, for example, all "G"s or all "C"s are omitted from the cluster. In some embodiments, sequences lacking "G"s or "C"s can form very few secondary structures and can contribute to more uniform and faster hybridization. Therefore, in some cases, the secondary amplified nucleic acid may contain only A, T, and G; only A, T, and C; only A, C, and G; or only T, C, and G.

[0135] In some cases, more than one type of secondary amplified nucleic acid can be applied to a sample, for example, sequentially or simultaneously. For example, there can be at least 2, at least 5, at least 10, at least 25, at least 50, at least 75, at least 100, at least 300, at least 1,000, at least 3,000, at least 10,000, or at least 30,000 distinguishable secondary amplified nucleic acids applied to the sample. In some cases, secondary amplified nucleic acids can be added sequentially. However, in some cases, more than one secondary amplified nucleic acid can be added simultaneously.

[0136] Furthermore, in some implementations, this pattern can alternatively be repeated before the signal transduction entity, for example, using tertiary amplified nucleic acids, quaternary nucleic acids, etc., similar to those discussed above. Therefore, the signal transduction entity can bind to the terminal amplified nucleic acid. Thus, as a non-limiting example, the coding nucleic acid probe can bind to the target, the primary amplified nucleic acid binds to the coding nucleic acid probe, the secondary amplified nucleic acid binds to the primary amplified nucleic acid, the tertiary amplified nucleic acid binds to the secondary amplified nucleic acid, and the signal transduction entity binds to the tertiary amplified nucleic acid; or the coding nucleic acid probe can bind to the target, the primary amplified nucleic acid binds to the coding nucleic acid probe, the secondary amplified nucleic acid binds to the primary amplified nucleic acid, the tertiary amplified nucleic acid binds to the secondary amplified nucleic acid, the quaternary amplified nucleic acid binds to the tertiary amplified nucleic acid, and the signal transduction entity binds to the quaternary amplified nucleic acid, etc. Therefore, in all implementations, the terminal amplified nucleic acid is not necessarily a secondary amplified nucleic acid.

[0137] A non-limiting example of such a system is shown in Figure 5. Figures 5A-5E The creation of a saturable system is shown. Figure 5A An example of a nucleic acid probe is shown, in which nucleic acid probe 15 has been bound to the target RNA. Figure 5B This shows the primary amplified nucleic acid being used according to certain implementation schemes. Figure 5C The secondary amplified nucleic acid that can bind to the primary amplified nucleic acid is shown. Figure 5D This indicates that multiple signal transduction entities have bound to the readout sequence of the secondary amplified nucleic acid. Figure 5E This shows that if amplification is not applied, the nucleic acid probe can be exposed to a suitable secondary nucleic acid probe containing a signal transduction entity.

[0138] In some cases, other components may also be present in the nucleic acid probe or the amplified nucleic acid. For example, in one set of embodiments, one or more primer sequences may be present, for example, to facilitate enzymatic amplification. Those skilled in the art will understand primer sequences suitable for applications such as amplification (e.g., using PCR or other suitable techniques). Many such primer sequences are commercially available. Other examples of sequences that may be present within primary or coding nucleic acid probes include, but are not limited to, promoter sequences, operons, identification sequences, nonsense sequences, etc.

[0139] Typically, primers are single-stranded or partially double-stranded nucleic acids (e.g., DNA) that serve as the starting point for nucleic acid synthesis, allowing polymerases (e.g., nucleic acid polymerases) to extend the primers and replicate the complementary strand. Primers are (e.g., designed) to be complementary to and hybridize with a nucleic acid target. In some embodiments, primers are synthetic primers. In some embodiments, primers are non-naturally occurring primers. Primers typically have a length of 10 to 50 nucleotides. For example, primers may have a length of 10 to 40, 10 to 30, 10 to 20, 25 to 50, 15 to 40, 15 to 30, 20 to 50, 20 to 40, or 20 to 30 nucleotides. In some embodiments, primers have a length of 18 to 24 nucleotides.

[0140] In some respects, as previously described, certain embodiments use a code space encoding various binding events, and optionally, error detection and / or correction can be used to determine the binding of a nucleic acid probe to its target. In some cases, the nucleic acid probe group may contain certain "readout sequences" capable of binding certain amplified nucleic acids, as discussed above, and the location of the nucleic acid probe or target can be determined within the sample using a signaling entity associated with the amplified nucleic acid, for example, within a code space, such as that discussed herein. See also International Patent Application Publications WO2016 / 018960 and WO2016 / 018963, each incorporated herein by reference in its entirety. In some cases, the readout sequence group within the nucleic acid probe can be combined in various combinations, for example, such that a relatively small number of readout sequences can be used to determine a relatively large number of different nucleic acid probes, as discussed herein.

[0141] Therefore, in some cases, nucleic acid probe groups can each contain a certain number of readout sequences, some of which are shared among different nucleic acid probes, so that the total population of nucleic acid probes can contain a certain number of readout sequences. Nucleic acid probe groups can have any suitable number of readout sequences. For example, a nucleic acid probe group can have 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 readout sequences. In some implementations, more than 20 are also possible. Furthermore, in some cases, the nucleic acid probe group may have a total of 1 or more, 2 or more, 3 or more, 4 or more, 5 or more, 6 or more, 7 or more, 8 or more, 9 or more, 10 or more, 11 or more, 12 or more, 13 or more, 14 or more, 15 or more, 16 or more, 20 or more, 24 or more, 32 or more, 40 or more, 50 or more, 60 or more, 64 or more, 100 or more, 128 or more, although some or all probes may each contain more than one readout sequence, as discussed herein. Additionally, in some embodiments, the nucleic acid probe group may have no more than 100, no more than 80, no more than 64, no more than 60, no more than 50, no more than 40, no more than 32, no more than 24, no more than 20, no more than 16, no more than 15, no more than 14, no more than 13, no more than 12, no more than 11, no more than 10, no more than 9, no more than 8, no more than 7, no more than 6, no more than 5, no more than 4, no more than 3, or no more than two readout sequences. Any combination of these is also possible; for example, a nucleic acid probe group can contain a total of 10 to 15 readout sequences.

[0142] As a non-limiting example of a method for identifying a relatively large number of nucleic acid probes from a relatively small number of readout sequences contained within a nucleic acid probe, in a population of six different types of nucleic acid probes or six different groups of nucleic acid probes (e.g., each group of probes binds to a nucleic acid target), each type or group of nucleic acid probes contains one or more readout sequences, and the total number of readout sequences within the population may not exceed four. It should be understood that although four readout sequences are used in this example for illustrative purposes, a larger number of nucleic acid probes can be achieved in other embodiments, such as using 5, 8, 10, 16, 32, or more readout sequences, or any other suitable number of readout sequences described herein, depending on the application. For example, if each nucleic acid probe or group of nucleic acid probes contains two different readout sequences, then by using four such readout sequences (A, B, C, and D), up to six probes or six groups of probes can be identified individually. It should be noted that in this example, the ordering of the readout sequences on the nucleic acid probes or nucleic acid probe groups is not required; that is, "AB" and "BA" can be considered synonyms (although in other embodiments, the ordering of the readout sequences may be essential, and "AB" and "BA" may not necessarily be synonyms). Similarly, if 5 readout sequences (A, B, C, D, and E) are used in a nucleic acid probe group, up to 10 probes or 10 probe groups (e.g., AB, AC, AD, AE, BC, BD, BE, CD, CE, DE) can be identified individually. For example, those skilled in the art will understand that for k readout sequences in a population with n readout sequences on each group of each probe or group, up to [number missing] readout sequences can be generated. Different probes may be used, assuming that the ordering of the read sequences is not required; because not all probes or probe groups need to have the same number of read sequences and not all combinations of read sequences need to be used in every implementation, more or fewer different probes may be used in some implementations. Furthermore, it should be understood that in some implementations, the number of read sequences on each probe or probe group need not be the same. For example, some probes or probe groups may contain 2 read sequences, while other probes or probe groups may contain 3 read sequences. In some implementations, each probe group binds to a nucleic acid target.

[0143] In some aspects, the readout sequences and / or binding patterns of nucleic acid probes in a sample can be used to define error detection and / or error correction codes, for example, to reduce or prevent false identification or errors of nucleic acids. Thus, for example, if binding is indicated (e.g., determined using a signal transduction entity), the location can be identified with a "1"; conversely, if binding is not indicated, the location can be identified with a "0" (and in some cases, the reverse). Multiple rounds of binding determination (e.g., using different readout probes complementary to the readout sequences) can then be used to create "codewords," for example, for that spatial location. In some implementations, error detection and / or correction can be performed on the codewords. For example, codewords can be organized such that if no match is found for a given set of readout sequences or binding patterns of nucleic acid probes, a match can be identified as an error, and optionally, error correction can be applied to the sequences to determine the correct target of the nucleic acid probe. In some cases, a codeword may have fewer "letters" or positions than the total number of nucleic acids encoded by the codeword, for example, when each codeword encodes a different nucleic acid.

[0144] Such error detection and / or error correction codes can take many forms. Several such codes, such as Golay codes or Hamming codes, have previously been developed in other environments, such as the telecommunications industry. In one set of implementations, the readout sequence or binding pattern of the nucleic acid probe is assigned such that not every possible combination is assigned.

[0145] For example, if there are potentially 4 readout sequences and the nucleic acid probe or probe set contains 2 readout sequences, then a maximum of 6 nucleic acid probes or 6 probe sets (e.g., such that each probe set binds to the nucleic acid target) can be identified; however, the number of nucleic acid probes or probe sets used can be less than 6. Similarly, for k readout sequences in a population with n readout sequences on each nucleic acid probe or probe set, a number of readout sequences can be generated. Different probes or different probe sets can be used, but the number of nucleic acid probes or nucleic acid probe sets used can be greater than or less than [the number of probes used]. Any number. Furthermore, these can be randomly assigned, or assigned in a specific manner to increase the ability to detect and / or correct errors.

[0146] As another example, if multiple rounds of nucleic acid probes are used (e.g., multiple rounds of readout probes that can bind to readout sequences on primary or coding probes), the number of rounds can be chosen arbitrarily. If each target can give two possible outcomes in each round, such as being detected or not being detected, then for n rounds of probes, there can be at most 2^n possible outcomes. n There are 2 different targets, but the actual number of targets used may be less than 2. nAn arbitrary number. In another example, if each target can give more than two possible results in each round, such as being detected in different color channels, then for a probe in n rounds, there may be more than 2 n (e.g. 3) n 4 n (...) different targets. In some cases, the actual number of targets used may be any number less than this. Furthermore, these can be randomly assigned or assigned in a specific way to increase the ability to detect and / or correct errors.

[0147] Codewords can be used to define various code spaces. Each nucleic acid target is associated with a codeword. For example, in one set of embodiments, codewords can be assigned within a code space such that the assignments are separated by Hamming distances, which measure the number of incorrect "reads" in a given pattern that cause a codeword or associated nucleic acid target to be misinterpreted as a different valid codeword or nucleic acid target. In some cases, the Hamming distance can be at least 2, at least 3, at least 4, at least 5, at least 6, etc. Furthermore, in one set of embodiments, the assignments can be formed as Hamming codes, such as Hamming(7,4) codes, Hamming(15,11) codes, Hamming(31,26) codes, Hamming(63,57) codes, Hamming(127,120) codes, etc. In another set of implementations, the assignment can form SECDED codes, such as SECDED(8,4) codes, SECDED(16,4) codes, SCEDED(16,11) codes, SCEDED(22,16) codes, SCEDED(39,32) codes, SCEDED(72,64) codes, etc. In yet another set of implementations, the assignment can form extended binary Golay codes, perfect binary Golay codes, or ternary Golay codes. In yet another set of implementations, the assignment can represent a subset of possible values ​​taken from any of the above codes.

[0148] For example, error correction codes can be formed by encoding the target using only binary words containing a fixed or constant number of "1" bits (or "0" bits). For example, the code space might only include 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, etc., such as all codes having the same number of "1" bits or "0" bits. In another set of implementations, the assignment can represent a subset of the possible values ​​taken from the above codes for the purpose of addressing asymmetric readout errors. For example, in some cases, codes where the number of "1" bits can be fixed for all used binary words can eliminate biased measurements of words with different numbers of "1" bits when the rates at which "0" bits are measured as "1" or "1" bits are measured as "0" are different.

[0149] Therefore, in some implementations, once a codeword is determined (e.g., as discussed herein), it can be compared with valid nucleic acid codewords. If a match is found, the nucleic acid target can be identified or determined. If no match is found, an error in the codeword reading can be identified. In some cases, error correction can also be applied to determine the correct codeword, leading to the correct identity of the nucleic acid target. In some cases, codewords can be selected such that, assuming only one error exists, only one possible correct codeword is available, thus making the unique correct identity of the nucleic acid target possible. In some cases, this can also be generalized to larger codeword spacing or Hamming distances; for example, codewords can be selected such that, if two, three, or four errors exist (or more in some cases), only one possible correct codeword is available, thus making the unique correct identity of the nucleic acid target possible.

[0150] Error correction codes can be binary error correction codes, or they can be based on other numbering systems, such as ternary or quaternary error correction codes. For example, in one set of embodiments, more than one type of signaling entity can be used and assigned different numbers within the error correction code. Thus, as a non-limiting example, a first signaling entity (or in some cases more than one signaling entity) can be assigned "1", and a second signaling entity (or in some cases more than one signaling entity) can be assigned "2" ("0" indicates the absence of a signaling entity), and code words are assigned to define a ternary error correction code. Similarly, a third signaling entity can be additionally assigned "3" to produce a quaternary error correction code, and so on.

[0151] In one set of implementations, each nucleic acid target in the sample is assigned a codeword. For example, these codewords can be selected from one of the code spaces described herein. In some cases, the codewords form error detection and / or error correction codes. In some cases, the sample can be hybridized with a set of primary or coding nucleic acid probes. Some or all of the primary or coding probes may contain a target sequence that can bind to one of the nucleic acid targets and / or may also contain one or more readout sequences. The readout sequences on the set of primary or coding probes binding to each nucleic acid target can form a unique codeword corresponding to the codeword assigned to the nucleic acid target. The sample is then hybridized with the readout probes for one or more rounds. The readout probes may be able to bind to the readout sequences and / or may associate with signal transduction entities. The set of readout sequences can be associated with the nucleic acid targets, and thus the codeword assigned to the nucleic acid targets can then be identified, for example, by the binding of the readout probes.

[0152] In some cases, multicolor imaging can be used in each round to allow simultaneous imaging and identification of multiple readout probes associated with different signal transduction entities. In some cases, the location of nucleic acid targets is determined. In some cases, at least 50, at least 100, at least 500, at least 1000, at least 5000, or at least 10,000 nucleic acid targets are identified in this manner. In some cases, the nucleic acid targets are genomic loci. In some cases, the nucleic acid targets are genomic loci and / or newly generated RNA transcripts. In some cases, the location of genomic loci is used to determine the three-dimensional organization of chromatin in cells or the three-dimensional organization of the genome. In some cases, primary amplified nucleic acids and / or secondary amplified nucleic acids and / or tertiary amplified nucleic acids and / or quaternary amplified nucleic acids are used to amplify the signal from each readout sequence. In some cases, adaptors are used as described below.

[0153] On the one hand, multiple connectors can be used to facilitate the detection of targets within a sample. Such connectors can be used, for example, to allow the use of a relatively small number of distinguishable signal transduction entities while still allowing the determination of a relatively large number of targets in the sample. For example, at least 3, at least 4, at least 7, at least 8, at least 12, at least 14, at least 15, at least 16, at least 20, at least 22, at least 30, at least 31, at least 32, at least 50, at least 63, at least 64, at least 72, at least 75, at least 100, at least 127, at least 128, at least 140, at least 255, at least 256, at least 500, at least 1,000, at least 1,500, at least 2,000, at least 2,500, at least 3,000, at least 4,000, etc., can be determined in the sample. 00, at least 5,000, at least 7,500, at least 10,000, at least 12,000, at least 15,000, at least 20,000, at least 25,000, at least 30,000, at least 40,000, at least 50,000, at least 75,000, or at least 100,000 targets, while using a smaller number of signal transduction entities, such as no more than 20, no more than 15, no more than 10, no more than 5, no more than 4, no more than 3, or no more than 2 signal transduction entities.

[0154] In one set of implementations, multiple adaptors may be used. An adaptor may comprise a first portion substantially complementary to one or more readout sequences on a nucleic acid probe (e.g., a primary nucleic acid probe), and a second portion comprising one or more identification sequences. Thus, the adaptor sequence is capable of binding to a specific nucleic acid probe that can bind to a target in a sample. The identification sequence can then be used for binding, for example, via a readout probe or a secondary nucleic acid probe, such as those discussed herein. Therefore, in some cases, an adaptor may be present between a primary nucleic acid probe and a secondary nucleic acid probe. Figure 24A The image shows a non-limiting example of this situation.

[0155] In some cases, adaptors can be selected to allow the use of a relatively small number of signal transduction entities, as described above. For example, the identification sequence can serve as a readout sequence that the secondary nucleic acid probe can bind to. In one round of testing, a relatively small number of secondary nucleic acid probes can be used, for example, containing a signal transduction entity and a sequence substantially complementary to one of the identification sequences, and the signal transduction entity can be measured, for example, as discussed herein. The secondary nucleic acid probe can then be removed and / or inactivated prior to the next round of testing, for example, as described herein. The next and subsequent rounds can use the same or different signal transduction entities, for example, on secondary nucleic acid probes containing sequences substantially complementary to different identification sequences.

[0156] Furthermore, in some implementations, to reduce contamination or "crosstalk," the adaptor used in the previous round can be inactivated in some way. For example, a blocked nucleic acid probe containing a sequence substantially complementary to the previously identified sequence can be added, enabling it to bind to the previous adaptor, but because it is typically undetectable in the absence of a signal transduction entity. Thus, in subsequent detection rounds, the signal from the previous round can be minimized.

[0157] Therefore, in some cases, no more than a relatively small number of signal transduction entities can be used to determine a relatively large number of identification sequences. For example, no more than 20, no more than 15, no more than 10, no more than 5, no more than 4, no more than 3, or no more than 2 signal transduction entities can be used to determine at least 3, at least 4, at least 7, at least 8, at least 12, at least 14, at least 15, at least 16, at least 20, at least 22, at least 30, at least 31, at least 32, at least 50, at least 63, at least 64, at least 72, at least 75, at least 100, at least 127, at least 128, at least 140, or at least 25 5. At least 256, at least 500, at least 1,000, at least 1,500, at least 2,000, at least 2,500, at least 3,000, at least 4,000, at least 5,000, at least 7,500, at least 10,000, at least 12,000, at least 15,000, at least 20,000, at least 25,000, at least 30,000, at least 40,000, at least 50,000, at least 75,000, or at least 100,000 identification sequences.

[0158] The identification sequence can be of any length. If more than one identification sequence is used, the identification sequences can independently have the same or different lengths. For example, the length of the identification sequence can be at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, at least 60, at least 65, at least 75, at least 100, at least 125, at least 150, at least 175, at least 200, at least 250, at least 300, at least 350, at least 400, or at least 450 nucleotides. In some cases, the length of the identification sequence may not exceed 500, 450, 400, 350, 300, 250, 200, 175, 150, 125, 100, 75, 60, 65, 60, 55, 50, 45, 40, 35, 30, 20, or 10 nucleotides. Any combination of these is also possible; for example, the identification sequence can have a length of 10 to 30 nucleotides, 20 to 40 nucleotides, 5 to 50 nucleotides, 10 to 200 nucleotides, or 25 to 35 nucleotides, 10 to 300 nucleotides, etc.

[0159] In some implementations, the identification sequence can be arbitrary or random. In some cases, the identification sequence is selected to reduce or minimize homology with other components of cells or other samples, for example, such that the identification sequence itself does not bind or hybridize with other nucleic acids suspected of being present in cells or other samples. In some cases, homology may be less than 10%, less than 8%, less than 7%, less than 6%, less than 5%, less than 4%, less than 3%, less than 2%, or less than 1%. In some cases, homology of fewer than 20 base pairs, fewer than 18 base pairs, fewer than 15 base pairs, fewer than 14 base pairs, fewer than 13 base pairs, fewer than 12 base pairs, fewer than 11 base pairs, or fewer than 10 base pairs may exist. In some cases, such base pairs are consecutive.

[0160] Furthermore, in some implementations, some or all of the identification sequences can be selected such that they do not exhibit specific binding to each other and / or to suspected genomic or other nucleic acid sequences, such as readout sequences, present in the sample. For example, the identification sequence group can be “aligned” or its specific binding or complementarity can be tested. In some cases, the identification sequences may not exhibit specific binding to each other, and / or the identification sequence group may not have any identification sequence with more than 5, 6, 7, 8, 9, 10, etc., nucleotide complementarity with another sequence within the identification sequence group and / or the readout sequence group.

[0161] In some implementations, the sample is first hybridized with a group of primary or coding nucleic acid probes. One or more of the primary or coding probes contain a target sequence capable of binding to one of the nucleic acid targets, and may also contain one or more readout sequences. The sample is then hybridized multiple rounds with adaptor probes and readout probes. The adaptor probe may contain a sequence capable of binding to the readout sequence and also contain one or more identification sequences. The readout probe may be capable of binding to the identification sequence and also associate with a signal transduction entity. In some cases, multicolor imaging may be used in each round to allow simultaneous imaging and identification of multiple readout probes associated with different signal transduction entities.

[0162] As discussed herein, in some respects, such as by imaging, signal transduction entities are used to determine nucleic acid probes and / or generate codewords. Examples of signal transduction entities include those discussed herein. In some cases, multiple techniques can be used, for example, to spatially determine signal transduction entities within a sample. In some embodiments, the signal transduction entity may be fluorescent, and techniques for measuring fluorescence within the sample, such as fluorescence microscopy or confocal microscopy, can be used to spatially identify the location of the signal transduction entity within the cell. In some cases, the location of the entity in the sample can be determined in two or even three dimensions. Furthermore, in some embodiments, more than one signal transduction entity can be determined at once and / or sequentially (e.g., signal transduction entities with different colors or emissions).

[0163] Furthermore, in some implementations, the confidence level of the identified target, such as a nucleic acid target, can be determined. For example, the confidence level can be determined using the ratio of the number of exact matches to the number of matches with one or more one-bit errors. In some cases, only matches with a confidence ratio greater than a certain value can be used. For example, in some implementations, a match is accepted only if its confidence ratio is greater than about 0.01, greater than about 0.03, greater than about 0.05, greater than about 0.1, greater than about 0.3, greater than about 0.5, greater than about 1, greater than about 3, greater than about 5, greater than about 10, greater than about 30, greater than about 50, greater than about 100, greater than about 300, greater than about 500, greater than about 1000, or any other suitable value. Furthermore, in some implementations, a match is accepted only if the confidence ratio of the identified target is greater than approximately 0.01, 0.03, 0.05, 0.1, 0.3, 0.5, 1, 3, 5, 10, 30, 50, 100, 300, 500, 1000, or any other suitable value for the internal standard or false positive control.

[0164] In some implementations, the spatial location of the entity (and therefore the nucleic acid probe that may associate with the entity) can be determined at a relatively high resolution. For example, the location can be determined with a spatial resolution better than about 100 micrometers, better than about 30 micrometers, better than about 10 micrometers, better than about 3 micrometers, better than about 1 micrometer, better than about 800 nm, better than about 600 nm, better than about 500 nm, better than about 400 nm, better than about 300 nm, better than about 200 nm, better than about 100 nm, better than about 90 nm, better than about 80 nm, better than about 70 nm, better than about 60 nm, better than about 50 nm, better than about 40 nm, better than about 30 nm, better than about 20 nm, or better than about 10 nm, etc.

[0165] Various techniques exist for optically determining or imaging the spatial location of entities, such as using fluorescence microscopy. In some implementations, more than one color can be used. In some cases, spatial location can be determined with super-resolution or a resolution better than the wavelength of light or the diffraction limit. Non-limiting examples include STORM (Stochastic Optical Reconstruction Microscopy), STED (Stimulated Emission Depletion Microscopy), NSOM (Near-Field Scanning Optical Microscopy), 4Pi Microscopy, SIM (Structured Illumination Microscopy), SMI (Spatial Modulation Illumination) Microscopy, RESOLFT (Reversible Saturated Optical Linear Fluorescence Transition Microscopy), GSD (Ground State Depletion Microscopy), SSIM (Saturated Structured Illumination Microscopy), SPDM (Spectroscopic Precision Distance Microscopy), Photoactivated Localization Microscopy (PALM), Fluorescence Photoactivated Localization Microscopy (FPALM), LIMON (3D Optical Microscopy Nanoscale Microscopy), Super-Resolution Optical Wave Imaging (SOFI), Dilatation Microscopy, etc. See, for example, U.S. Patent No. 7,838,302 entitled “Sub-Diffraction Limit Image Resolution and Other Imaging Techniques”, issued November 23, 2010 by Zhuang et al.; U.S. Patent No. 8,564,792 entitled “Sub-diffraction Limit Image Resolution in Three Dimensions”, issued October 22, 2013 by Zhuang et al.; or International Patent Application Publication No. WO2013 / 090360 entitled “High Resolution Dual-Objective Microscopy”, published June 20, 2013 by Zhuang et al., each of which is incorporated herein by reference in its entirety.

[0166] As illustrative and non-limiting examples, in one set of embodiments, a sample can be imaged using a high numerical aperture, an oil immersion objective with 100X magnification, and light collected on an electron-multiplying CCD camera. In another example, a sample can be imaged using a high numerical aperture, an oil immersion lens with 40X magnification, and light collected using a wide-field scientific CMOS camera. In various non-limiting embodiments, using different combinations of objectives and cameras, a single field of view may correspond to a sample area of ​​not less than 1x1 μm, 10x10 μm, 40x40 μm, 80x80 μm, 120x120 μm, 240x240 μm, 340x340 μm, or 500x500 μm, etc. Similarly, in some embodiments, a single camera pixel may correspond to a sample area of ​​not less than 10x10 nm, 20x20 nm, 40x40 nm, 80x80 nm, 120x120 nm, 160x160 nm, 240x240 nm, or 300x300 nm, etc. In another example, the sample can be imaged using an air lens with a low numerical aperture and 10X magnification, and light collected using an sCMOS camera. In another embodiment, the sample can be optically sliced ​​by illumination via one or more scanning diffraction-limited focal points generated by a scanning mirror or rotating disk, and the collected sample can be passed through one or more pinholes. In yet another embodiment, the sample can also be illuminated by a thin sheet produced via any of a variety of methods known to those skilled in the art.

[0167] In one embodiment, the sample can be illuminated by a single Gaussian mode laser line. In some embodiments, the illumination profile can be flattened by passing these laser lines through a multimode optical fiber vibrating via piezoelectric or other mechanical means. In some embodiments, the illumination profile can be flattened by passing the single-mode Gaussian beam through various refractive beam shapers such as piShaper or a series of stacked Powell lenses. In yet another set of embodiments, the Gaussian beam can be passed through various different diffusers, such as frosted glass or engineered diffusers, which in some cases can be rotated at high speed to remove residual laser spots. In yet another embodiment, laser illumination can be achieved through an array of small lenses to produce an overlay image of illumination with a near-flat illumination field.

[0168] In some implementations, the centroid of the spatial location of an entity can be determined. For example, the centroid of a signal-transmitting entity can be determined within an image or series of images using image analysis algorithms known to those skilled in the art. In some cases, algorithms can be selected to determine non-overlapping individual emitters and / or partially overlapping individual emitters in a sample. Non-limiting examples of suitable techniques include maximum likelihood algorithms, least squares algorithms, Bayesian algorithms, compressed sensing algorithms, etc. Combinations of these techniques may also be used in some cases.

[0169] In some implementations, one or more signal transduction entities can be measured. For example, a signal transduction entity may bind to a readout probe or recognition entity on a secondary amplified nucleic acid (or other terminal amplified nucleic acid). Non-limiting examples of signal transduction entities include fluorescent entities (fluorophores) or phosphorescent entities, such as those discussed herein. The signal transduction entity can then be measured, for example, to measure a nucleic acid probe or target. In some cases, the measurement may be spatial, for example, in two or three dimensions. Furthermore, in some cases, the measurement may be quantitative, for example, measuring the amount or concentration of the signal transduction entity and / or target.

[0170] In one set of implementations, signal transduction entities may be attached to secondary amplified nucleic acids (or other terminal amplified nucleic acids). The signal transduction entities may be attached to the secondary amplified nucleic acids (or other terminal amplified nucleic acids) before or after the secondary amplified nucleic acids associate with a target within the sample. For example, the signal transduction entities may be initially attached to the secondary amplified nucleic acids, or attached after the secondary amplified nucleic acids have been applied to the sample. In some cases, signal transduction entities are added, followed by a reaction to attach them to the amplified nucleic acids.

[0171] In one set of embodiments, signal transduction entities can be attached to a nucleotide sequence via bonds that can be cleaved to release the signal transduction entity. For example, after determining the distribution of nucleic acid probes within a sample, the signal transduction entity can be released or inactivated before another round of nucleic acid probes and / or nucleic acid amplification. Therefore, in some embodiments, the bond can be a cleavable bond, such as a disulfide bond or a photocleavable bond. Examples of photocleavable bonds are discussed in detail herein. In some cases, such bonds can be cleaved, for example, upon exposure to a reducing agent or light (e.g., ultraviolet light). See below for further details. In some cases, signal transduction entities are inactivated by photobleaching. Other examples of systems and methods for inactivating and / or removing signal transduction entities are discussed in more detail herein.

[0172] In some implementations, the use of primary and secondary amplified nucleic acids can be used to generate a maximum number of signaling entities capable of binding to a given nucleic acid probe. For example, there may be a maximum number of signaling entities capable of binding a nucleic acid probe, for instance, due to a maximum number of readout probes with signaling entities capable of binding a limited number of secondary amplified nucleic acids, a maximum number of secondary amplified nucleic acids capable of binding a limited number of primary amplified nucleic acids, and / or a maximum number of primary amplified nucleic acids capable of binding a limited number of readout sequences on a nucleic acid probe. While each potential site does not actually need to be filled with signaling entities, the structure indicates a saturation limit for signaling entities beyond which any other signaling entities that might happen to exist cannot associate with the nucleic acid probe or its target.

[0173] Therefore, certain implementations generally involve systems and methods for amplifying signals indicating a saturable nucleic acid probe or its target (i.e., an upper limit of saturation such that there exists a saturation limit on how many signaling entities can associate with the nucleic acid probe or its target). Typically, this number is greater than 1. For example, the upper limit of signaling entities can be at least 2, at least 3, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 40, at least 50, at least 75, at least 100, at least 125, at least 150, at least 175, at least 200, at least 250, at least 300, at least 400, at least 500, etc. In some cases, the upper limit may be less than 500, less than 400, less than 300, less than 250, less than 200, less than 175, less than 150, less than 125, less than 100, less than 75, less than 50, less than 40, less than 30, less than 25, less than 20, less than 15, less than 10, less than 5, etc. In some cases, the upper limit can be determined as the maximum number of signal transduction entities that can bind to secondary amplified nucleic acids, multiplied by the maximum number of secondary amplified nucleic acids that can bind to primary amplified nucleic acids, multiplied by the maximum number of primary amplified nucleic acids that can bind to a nucleic acid probe that binds to the target. In contrast, techniques such as rolling circle amplification or hairpin unwinding allow for uncontrolled signal amplification; that is, amplification can continue without a predetermined endpoint or saturation limit when sufficient reagents are available. Therefore, such techniques do not have a theoretical upper limit on the number of signal transduction entities that can associate with nucleic acid probes or their targets.

[0174] However, it should be understood that the average number of signal transduction entities that actually bind to the nucleic acid probe or its target does not actually need to be the same as its upper limit; that is, the signal transduction entities may not actually be completely saturated (although they can be). For example, saturation (or the number of bound signal transduction entities relative to the maximum number that can bind) may be less than 97%, less than 95%, less than 90%, less than 85%, less than 80%, less than 75%, etc., and / or at least 50%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 97%, etc. In some cases, allowing more time for binding to occur and / or increasing the reagent concentration may increase the saturation.

[0175] Due to a potential upper limit on the number of signaling entities that actually bind to a nucleic acid probe or its target, binding events distributed within a sample (e.g., spatially) may exhibit substantially consistent size and / or brightness compared to uncontrolled amplification, such as those discussed above. For example, due to the limited number of secondary amplified nucleic acids that can bind to primary amplified nucleic acids, secondary amplified nucleic acids larger than a fixed distance from the nucleic acid probe or its target cannot be detected, which may limit the “spot size” or diameter of fluorescence indicating binding from signaling entities.

[0176] In some implementations, at least 50%, at least 60%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, or at least 95% of binding events can exhibit substantially the same brightness, size (e.g., apparent diameter), color, etc., which can make it easier to distinguish binding events from other events (e.g., nonspecific binding, noise, etc.).

[0177] Furthermore, in some cases, the signal transduction entity may be inactivated. For example, in some embodiments, a primary nucleic acid probe or readout probe that can associate with the signal transduction entity (e.g., using amplified nucleic acid) can be applied to a sample that can identify a first readout sequence (e.g., on a primary or coding nucleic acid probe), thus inactivating the signal transduction entity before applying a secondary nucleic acid probe or readout probe (e.g., that can associate with the signal transduction entity (e.g., using amplified nucleic acid)) to the sample. If multiple signal transduction entities are used, the same or different techniques can be used to inactivate the signal transduction entities, and some or all of the multiple signal transduction entities can be inactivated, for example, sequentially or simultaneously.

[0178] Inactivation can be caused by the removal of the signal transduction entity (e.g., from a sample, or from a nucleic acid probe), and / or by chemically altering the signal transduction entity in some way (e.g., by photobleaching, bleaching, or chemically altering the structure of the signal transduction entity, such as by reduction). For example, in one set of embodiments, the fluorescent signal transduction entity can be inactivated by chemical or optical techniques, such as oxidation, photobleaching, chemical bleaching, rigorous washing or enzymatic digestion, or by exposure to enzymes, dissociation of the signal transduction entity from other components (e.g., probes), chemical reactions of the signal transduction entity (e.g., with reactants capable of altering the structure of the signal transduction entity), etc. For example, bleaching can be achieved by exposure to oxygen, reduction occurring, or the signal transduction entity can be chemically cleaved from the nucleic acid probe and flushed away by fluid flow.

[0179] In some embodiments, various nucleic acid probes may associate with one or more signal transduction entities, such as amplified nucleic acids discussed herein. If more than one nucleic acid probe (or secondary nucleic acid probe or readout probe) is used, the signal transduction entities may be the same or different. In some embodiments, the signal transduction entity is any entity capable of emitting light. For example, in one embodiment, the signal transduction entity is fluorescent. In other embodiments, the signal transduction entity may be phosphorescent, radioactive, absorbent, etc. In some cases, the signal transduction entity is any entity that can be determined within a sample at a relatively high resolution (e.g., at a resolution better than the wavelength of visible light or the diffraction limit). The signal transduction entity may be, for example, a dye, a small molecule, a peptide, or a protein. In some cases, the signal transduction entity may be a single molecule. If multiple secondary nucleic acid probes or readout probes are used, the nucleic acid probes may associate with the same or different signal transduction entities.

[0180] Non-limiting examples of signal transduction entities include fluorescent entities (fluorophores) or phosphorescent entities, such as cyanine dyes (e.g., Cy2, Cy3, Cy3B, Cy5, Cy5.5, Cy7, etc.), Alexa Fluor dyes, Atto dyes, photoswitched dyes, photoactivated dyes, fluorescent dyes, metal nanoparticles, semiconductor nanoparticles, or "quantum dots".

[0181] In one set of embodiments, the signal transduction entity can be attached to an oligonucleotide sequence via a bond that can be cleaved to release the signal transduction entity. In another set of embodiments, the fluorophore can be conjugated to the oligonucleotide via a cleavable bond, such as a photocleavable bond. Non-limiting examples of photocleavable bonds include, but are not limited to, 1-(2-nitrophenyl)ethyl, 2-nitrobenzyl, biotin phosphoramidite, acrylate phosphoramidite, diethylaminocoumarin, 1-(4,5-dimethoxy-2-nitrophenyl)ethyl, cyclododecyl(dimethoxy-2-nitrophenyl)ethyl, 4-aminomethyl-3-nitrobenzyl, (4-nitro-3-(1-chlorocarbonyloxyethyl)phenyl)methyl-S-acetylthioester, (4-nitro-3-(1-thlorocarbonyloxyethyl))phenyl)methyl-3-(2-pyridinedithiopropionic acid), and 3-(4,4'-dimethoxytriphenylmethyl)-1-(2-nitrophenyl) )-propane-1,3-diol-[2-cyanoethyl-(N,N-diisopropyl)]-phosphoramide, 1-[2-nitro-5-(6-trifluoroacetylhexamidomethyl)phenyl]-ethyl-[2-cyanoethyl-(N,N-diisopropyl)]-phosphoramide, 1-[2-nitro-5-(6-(4,4'-dimethoxytriphenylmethyloxy)butamidomethyl)phenyl]-ethyl-[2-cyanoethyl-(N,N-diisopropyl)]-phosphoramide, 1-[2-nitro-5-(6-(N-(4,4'-dimethoxytriphenylmethyl))-biotinaminohexanoylamino-methyl)phenyl]-ethyl-[2-cyanoethyl-(N,N-diisopropyl)]-phosphoramide, or similar linkers. The oligonucleotide sequence can be, for example, a primary or secondary (or other) amplified nucleic acid, such as those discussed herein.

[0182] In another set of embodiments, the fluorophore can be conjugated to an oligonucleotide via a disulfide bond. The disulfide bond can be cleaved by a variety of reducing agents, such as, but not limited to, dithiothreitol, dithioerythritol, β-mercaptoethanol, sodium borohydride, thioredoxin, glutarate, trypsinogen, hydrazine, diisobutylaluminum hydride, oxalic acid, formic acid, ascorbic acid, phosphorous acid, stannous chloride, glutathione, mercaptoacetate, 2,3-dimercaptopropanol, 2-mercaptoethylamine, 2-aminoethanol, tris(2-carboxyethyl)phosphine, bis(2-mercaptoethyl)sulfone, N,N'-dimethyl-N,N'-bis(mercaptoacetyl)hydrazine, 3-mercaptopropionate, dimethylformamide, thiopropyl-agarose, tri-n-butylphosphine, cysteine, ferric sulfate, sodium sulfite, phosphite, hypophosphite, thiophosphate, etc., and / or any combination of these. Oligonucleotides can be, for example, primary nucleic acid probes, encoding nucleic acid probes, readout probes, primary or secondary (or other) amplified nucleic acids, such as those discussed herein.

[0183] In another embodiment, the fluorophore can be conjugated to an oligonucleotide via one or more phosphate-thioester modified nucleotides, wherein the sulfur modification replaces bridging and / or non-bridging oxygen. In some embodiments, the fluorophore can be cleaved from the oligonucleotide by adding a compound such as, but not limited to, iodoethanol, iodine mixed in ethanol, silver nitrate, or mercuric chloride. In yet another set of embodiments, the signaling entity can be chemically inactivated by reduction or oxidation. For example, in one embodiment, sodium borohydride can be used to reduce chromophores such as Cy5 or Cy7 to a stable non-fluorescent state. In another set of embodiments, the fluorophore can be conjugated to the oligonucleotide via an azo bond, and the azo bond can be cleaved with 2-[(2-N-arylamino)phenylazo]pyridine. In yet another set of embodiments, the fluorophore can be conjugated to the oligonucleotide via a suitable nucleic acid segment that can be cleaved upon appropriate exposure to a DNase (e.g., an exonuclease or an endonuclease). Examples include, but are not limited to, deoxyribonuclease I or deoxyribonuclease II. In one set of embodiments, cleavage can occur via a restriction endonuclease. Non-restrictive examples of potentially suitable restriction endonucleases include BamHI, BsrI, NotI, XmaI, PspAI, DpnI, MboI, MnlI, Eco57I, Ksp632I, DraIII, AhaII, SmaI, MluI, HpaI, ApaI, BclI, BstEII, TaqI, EcoRI, SacI, HindII, HaeII, DraII, Tsp509I, Sau3AI, PacI, etc. More than 3000 restriction enzymes have been studied in detail, of which more than 600 are commercially available. In another set of embodiments, the fluorophore can be conjugated to biotin, and the oligonucleotide can be conjugated to avidin or streptoavidin. The interaction between biotin and avidin or streptoavidin allows the fluorophore to bind to the oligonucleotide, while, upon sufficient exposure to an excess of addition, free biotin may "override" the conjugation, leading to the unbinding of the fluorophore from the oligonucleotide. In another set of embodiments, a corresponding "toe-hold-probe" can be used to remove the probe, which contains the same sequence as the secondary or readout probe, plus an additional number of bases homologous to the primary or coding probe (e.g., 1-20 additional bases, such as 5 additional bases). These probes can remove the labeled secondary or readout probe via strand substitution interactions. The oligonucleotide can be, for example, a primary nucleic acid probe, a coding nucleic acid probe, a readout probe, a primary or secondary (or other) amplified nucleic acid, such as those discussed herein.

[0184] As used herein, the term "light" generally refers to electromagnetic radiation having any suitable wavelength (or equivalently, frequency). For example, in some embodiments, light may include wavelengths in the optical or visual range (e.g., having wavelengths between about 400 nm and about 700 nm, i.e., "visible light"), infrared wavelengths (e.g., having wavelengths between about 300 micrometers and 700 nanometers), ultraviolet wavelengths (e.g., having wavelengths between about 400 nanometers and about 10 nanometers), and so on. In some cases, as discussed herein, more than one entity may be used, i.e., entities that are chemically different or distinguishable (e.g., structurally). However, in other cases, entities may be chemically identical or at least substantially chemically identical.

[0185] In one set of embodiments, the signal transduction entity is "switchable," meaning that the entity can switch between two or more states, in at least one of which emits light of a desired wavelength. In other states, the entity may not emit light, or may emit light of a different wavelength. For example, the entity can be "activated" to a first state capable of producing light of the desired wavelength and "deactivated" to a second state in which it cannot emit light of the same wavelength. If an entity can be activated by incident light of a suitable wavelength, then it is "photoactivated." As a non-limiting example, Cy5 or Alexa 647 can be switched between fluorescent and dark states in a controlled and reversible manner by light of different wavelengths; that is, 633nm (or 642nm, 647nm, 656nm) red light can switch or deactivate Cy5 or Alexa 647 to a stable dark state, while 405nm green light can switch or activate Cy5 or Alexa 647 back to a fluorescent state. In some cases, the entity can reversibly switch between two or more states, for example, upon exposure to appropriate stimuli. For example, a first stimulus (e.g., light of a first wavelength) can be used to activate the switchable entity, while a second stimulus (e.g., light of a second wavelength) can be used to deactivate the switchable entity to, for example, a non-luminescent state. Any suitable method can be used to activate the entity. For example, in one embodiment, incident light of a suitable wavelength can be used to activate the entity to emit light; that is, the entity is “photoswitched.” Thus, a photoswitched entity can switch between different luminescent or non-luminescent states by, for example, incident light of different wavelengths. The light can be monochromatic (e.g., generated using a laser) or multicolor. In another embodiment, the entity can be activated by stimulation with an electric and / or magnetic field. In other embodiments, the entity can be activated by exposure to a suitable chemical environment (e.g., by adjusting pH or inducing a reversible chemical reaction involving the entity). Similarly, any suitable method can be used to deactivate the entity, and the methods for activating and deactivating the entity need not be the same. For example, the entity can be deactivated by exposure to incident light of a suitable wavelength, or the entity can be deactivated by waiting for a sufficient amount of time.

[0186] Typically, a person skilled in the art can identify a “switchable” entity by determining the conditions under which an entity in a first state can emit light when exposed to an excitation wavelength, switching the entity from the first state to a second state (e.g., when exposed to light of a switched wavelength), and then showing that the entity in the second state can no longer emit light (or emits light with a greatly reduced intensity) when exposed to an excitation wavelength.

[0187] In one set of embodiments, as discussed, the switchable entity can switch when exposed to light. In some cases, the light used to activate the switchable entity can come from an external source such as a light source like a laser source, another light-emitting entity nearby, etc. In some cases, the second light-emitting entity can be a fluorescent entity, and in some embodiments, the second light-emitting entity itself can also be a switchable entity.

[0188] In some implementations, the switchable entity includes a first luminescent portion (e.g., a fluorophore) and a second portion that activates or "switches" the first portion. For example, upon exposure to light, the second portion of the switchable entity can activate the first portion, causing the first portion to emit light. Examples of activator portions include, but are not limited to, Alexa Fluor 405 (Invitrogen), Alexa Fluor 488 (Invitrogen), Cy2 (GE Healthcare), Cy3 (GE Healthcare), Cy3B (GE Healthcare), Cy3.5 (GE Healthcare), or other suitable dyes. Examples of luminescent portions include, but are not limited to, Cy3B (GE Healthcare), Cy5, Cy5.5 (GE Healthcare), Cy7 (GE Healthcare), Alexa Fluor 647 (Invitrogen), Alexa Fluor 680 (Invitrogen), Alexa Fluor 700 (Invitrogen), Alexa Fluor 750 (Invitrogen), Alexa Fluor 790 (Invitrogen), DiD, DiR, YOYO-3 (Invitrogen), YO-PRO-3 (Invitrogen), TOT-3 (Invitrogen), TO-PRO-3 (Invitrogen), or other suitable dyes. See, for example, U.S. Patent No. 7,838,302, which is incorporated herein by reference in its entirety. In some cases, the first luminescent portion can subsequently be inactivated by any suitable technique, such as directing 647 nm red light to the Cy5 portion of the molecule.

[0189] In some implementations, multiple nucleic acid probes with different sequences are used, and the distribution of each nucleic acid probe is analyzed sequentially based on its binding pattern, and used to create a "codeword" for each location. By selecting nucleic acid probes that define an appropriate code space, obvious errors in observed binding patterns can be identified and / or discarded and / or corrected to identify the correct codeword, thereby identifying the correct target of the nucleic acid probe within the sample. This error-resistant and error-correction system was originally introduced for multiplex error-resistant fluorescence in situ hybridization (MERFISH) and has since been used in various related techniques. See, for example, International Patent Application Publications WO2016 / 018960 and WO2016 / 018963, each incorporated herein by reference in its entirety.

[0190] As mentioned, in some implementations, this technique can be combined with error correction, for example, as used in MERFISH or other similar techniques. For example, the codeword can be based on the binding (or non-binding) of multiple readout probes capable of binding to readout sequences on primary or encoding nucleic acid probes, and in some cases, the codeword can define error correction codes to help reduce or prevent false identification of nucleic acid probes. In some cases, a relatively small number of readout probes can be used to identify a relatively large number of different targets, for example, by using various combined methods. Fluorescence microscopy, wide-field fluorescence microscopy, epifluorescence microscopy, confocal microscopy, or light sheet microscopy can be used for image acquisition. Image acquisition techniques such as STORM or other super-resolution imaging methods can also be used to image such samples and facilitate the identification of nucleic acid probes. For further details on techniques such as MERFISH, see, for example, U.S. Patent Nos. 9,712,805 or 10,073,035, or International Patent Application Publication Nos. WO2008 / 091296 or WO2009 / 085218, each of which is incorporated herein by reference in its entirety. In some cases, expansion microscopy can also be used, in which the sample expands before imaging. See, for example, International Patent Application Publication No. WO2018 / 089445 entitled “Matrix Imprinting and Clearing” or International Patent Application Publication No. WO2018 / 089438 entitled “Multiplexed Imaging Using MERFISH and Expansion Microscopy”, each of which is incorporated herein by reference in its entirety.

[0191] On the other hand, this relates to a computer-implemented method. For example, a computer and / or automated system capable of automatically and / or repeatedly executing any of the methods described herein can be provided. As used herein, an "automated" device means a device capable of operating without human supervision; that is, an automated device can perform a function some time after anyone has taken any action to facilitate the function (e.g., by initiating the process by inputting instructions to a computer). Typically, an automated device can perform repetitive functions after this point in time. In some cases, the processing steps may also be recorded on a machine-readable medium.

[0192] For example, in some cases, a computer can be used to control the imaging of a sample, such as using fluorescence microscopy, wide-field fluorescence microscopy, epifluorescence microscopy, confocal microscopy, light sheet microscopy, diffraction-limited optical microscopy, STORM, or other super-resolution techniques, such as those described herein. In some cases, a computer can also control operations such as drift correction, physical registration, hybridization and cluster alignment in image analysis, cluster decoding (e.g., fluorescence cluster decoding), error detection or correction (e.g., as discussed herein), noise reduction, identification of foreground features from background features (e.g., noise or debris in an image), etc. As an example, a computer can be used to control the activation and / or excitation and / or inactivation of signal transducing entities within a sample, and / or the acquisition of images of the signal transducing entities. In one set of embodiments, light with various wavelengths and / or intensities can be used to excite the sample, and a computer can be used to correlate the wavelength sequence of light used to excite the sample with an acquired image of the sample containing the signal transducing entity. For example, a computer can apply light with various wavelengths and / or intensities to a sample to generate a different average number of signal-transmitting entities in each region of interest (e.g., one active entity per location, two active entities per location, etc.). In some cases, this information can be used to construct an image and / or determine the location of the signal-transmitting entities, and in some cases, this is achieved at high resolution, as described above.

[0193] In some aspects, the sample is located on a microscope. In some cases, the microscope may include one or more channels, such as fluid or microfluidic channels, to guide or control the inflow or outflow of fluid into or out of the sample. For example, in one embodiment, nucleic acid probes (such as those discussed herein) can be introduced and / or removed from the sample by allowing fluid to flow into or out of the sample through one or more channels. In some cases, there may also be one or more chambers or reservoirs for holding fluid, for example, in fluid communication with the channels and / or with the sample. Those skilled in the art will be familiar with channels, including fluid or microfluidic channels, for moving fluid into or out of a sample.

[0194] The following documents are incorporated herein by reference: International Patent Application Publication No. WO 2018 / 218150 entitled "Systems and Methods for High-Throughput Image-Based Screening"; WO 2016 / 018960 entitled "Systems and Methods for Determining Nucleic Acids"; WO 2016 / 018963 entitled "Probe Library Construction"; WO 2018 / 089445 entitled "Matrix Imprinting and Clearing"; WO2018 / 089438 entitled "Multiplexed Imaging Using MERFISH and Expansion Microscopy"; and U.S. Patent Application Serial No. 62 / 836,578 entitled "Imaging-Based Pooled CRISPR Screening" and 62 / 779,333 entitled "Amplification Methods and Systems for MERFISH and Other Applications". The following documents are also incorporated herein by reference in their entirety: U.S. Patent Nos. 2017 / 0220733 and 2017 / 0212986.

[0195] In addition, U.S. Patent Application Publication No. 62 / 954,720 entitled “Genome-Scale Imaging of the 3D Organization and Transcriptional Activity of Chromatin” and U.S. Patent Application Publication No. 63 / 060,947 entitled “Genome-Scale Imaging of the 3D Organization and Transcriptional Activity of Chromatin” are each incorporated herein by reference in their entirety.

[0196] The following examples are intended to illustrate certain embodiments of the present invention, but do not represent the full scope of the invention.

[0197] Example

[0198] Example 1

[0199] The following examples demonstrate a large-scale multiplex FISH approach for imaging 3D organization of chromatin at the genome scale in a single cell, and further demonstrate the ability to place 3D genome organization in its native structural and functional context by combining chromatin and nascent transcript imaging (both at the genome scale) with nuclear structure identification.

[0200] The first embodiment reports a large-scale multiplex FISH method that allows for genome-wide imaging of chromatin organization in single cells. Using this method, imaging and identification of >1,000 distinct genomic loci (approximately 2,000 chromatin loci, calculated based on chromosomal homologous pairs) in the human genome were confirmed in a single cell. Furthermore, simultaneous imaging of these genomic loci was confirmed using nascent RNA transcripts of >1,000 genes present at these loci against a background of various nuclear structures, including nuclear spots, nucleoli, and nuclear lamina. This method can be used to explore the relationship between chromatin organization, transcriptional activity, and the nuclear environment in single cells.

[0201] To achieve genome-wide chromatin imaging, a combinatorial FISH method was designed, inspired by previous error-resistant FISH methods developed for transcriptome imaging, but with significant modifications specifically designed for chromatin imaging, taking into account the aggregation properties of chromatin (i.e., adjacent loci in the genome sequence are spatially close) and the regional organization of chromosomes (i.e., different chromosomes tend to occupy different spatial regions). See, for example, WO2016 / 018960 entitled “Systems and Methods for Determining Nucleic Acids”; WO2016 / 018963 entitled “Probe Library Construction”, each of which is incorporated herein by reference in its entirety. To allow for combinatorial imaging, each genome locus is assigned a unique 100-bit binary code with a Hamming weight of 2, i.e., each barcode contains two “1” bits and 98 “0” bits. Figure 1AThe bit values ​​in these barcodes determine the presence (1) or absence (0) of a signal at each locus in successive imaging rounds. To avoid simultaneously imaging spatially close chromatin regions at the same bit, a subset of these 100-bit Hamming-weighted 2-bit barcodes was further selected to encode target genomic loci, and the allocation of these barcodes was optimized such that loci with a "1" bit at the same barcode position were maximally separated in genomic space. This strategy minimizes detection errors caused by overlapping signals from nearby chromatin loci. Furthermore, since the vast majority of possible 100-bit binary codes are invalid (i.e., not assigned to any target locus), this design allows for the identification and discarding of detection errors, further improving measurement accuracy.

[0202] Barcodes are physically imprinted onto target genomic loci using highly diverse coding probe libraries. Each coding probe library contains a 40-nt target region for binding to one of the target loci and a 20-nt readout sequence selected from 100 pre-designed readout sequences. Figure 1A Each readout sequence corresponds to one of 100 bits, and the coding probe set for each genomic locus (approximately 400 probes per locus) contains only two distinct readout sequences, corresponding to the two bits in the barcode assigned to that locus that read "1". After binding to the coding probe, the barcode imprinted on the chromatin locus is detected by sequential hybridization of fluorescently labeled readout probes, each probe being complementary to one of the 100 readout sequences. Figure 1A Each hybridization round introduces two distinct readout probes and images are performed in two color channels, thus enabling the imaging and identification of all approximately 1000 genomic loci after 50 hybridization rounds. Figure 1A-1C In contrast, direct sequential methods for imaging 1000 genomic loci would require 500 rounds of hybridization using two-color imaging. Since each chromosome in a diploid cell has two homologs, a clustering algorithm is used to further assign homologous identities to the imaged loci, taking advantage of the tendency of chromosomes to occupy different regions in each cell nucleus.

[0203] In this embodiment, 1,041 genomic loci were selected for imaging, each approximately 30 kb in size, uniformly covering the 22 autosomes and X chromosome in human lung fibroblast (IMR90) cells. Each chromosome was required to contain at least 30 target loci; therefore, the number of loci imaged per chromosome homolog ranged from 30 to 80, depending on chromosome length. These 1,041 genomic loci in approximately 5,400 individual cells across five biological replications were imaged with approximately 80% detection efficiency per locus. Considering that each chromosome has two homologs, approximately 1,700 chromatin loci were detected per cell. Figure 1D-1E).

[0204] To obtain a population-average view of chromatin organization, the spatial distance between each pair of imaged chromatin loci is calculated in each cell, and then the median distance and contact frequency between each pair of loci are determined in all imaged cells. Figure 1F and Figure 6A The contact frequencies between chromatin locus pairs within the same chromosome, identified from imaging data, showed a high correlation with the contact frequencies detected by overall Hi-C, with a Pearson correlation coefficient of 0.91. Figure 6B The imaging data captured chromatin structures at multiple scales, from chromosome organization into regions ( Figure 1F and Figure 6A The formation of compartments A and B within the chromosome arm () Figure 7A This is also consistent with the compartments identified through overall Hi-C measurement. Figures 7B-7C Furthermore, the imaging results showed high reproducibility among independent biological replicates. Figure 8 ).

[0205] By exploring the chromatin organization within individual cells, chromosomes occupy different regions within each cell while ( Figure 1F-1G They also show significant overlap with each other. Figure 1G-1H On average, about 80% of the convex hull volume occupied by any given chromosome is shared with other chromosomes in the same cell. Figure 1I This indicates a high degree of transchromosomal interaction. Since these interactions are not fully explored, the following analysis focuses on them.

[0206] Figure 1A-1I Genome-scale chromatin imaging was shown. Figure 1A The imaging scheme is shown. Target genomic loci were assigned error-resistant barcodes, such as 100-bit binary barcodes with a Hamming weight of 2 (i.e., two reads of "1" in the 100 bits). The barcodes were imprinted onto the genomic loci using coding oligonucleotide probes that identified the loci and associated each locus with two distinct readout sequences, corresponding to the two reads of "1" in the barcode assigned to that locus. Each locus was labeled with a total of 400 coding probes, but only four are shown. Fluorescent readout probes complementary to the readout sequences were added sequentially and imaged, allowing the reads of "1" at each locus to determine its barcode identity. Approximately 1000 genomic loci were imaged. Figure 1BRepresentative images from multiple imaging rounds within the nucleus of a single cell are shown. Fluorescent signals from chromatin loci of the readout probe are shown in lighter shades, while the signal of 4',6-diamidinyl-2-phenylindole (DAPI), used as a nuclear marker, is shown in darker shades. Scale bar: 5 micrometers. Figure 1C It is a small spatial region centered on a single chromatin locus in all imaging rounds. Figure 1B A magnified image of the frame (indicated by the image). Locus identity was determined based on two readout probes (1 and 13) that emitted the signal. Scale bar: 300 nm. Figure 1D It is a 3D rendering of all detected chromatin loci in a single cell, grayscaled according to the chromosome they belong to. Adjacent loci in the genome sequence are connected by thin lines. Figure 1E Showing with Figure 1D Chromatin loci in the same cell, but the two homologs of the chromosome shown are displayed in a different shade of gray than all other loci. Figure 1F This is a median distance matrix calculated from approximately 5,400 single cells. For each pair of loci, it displays the median of all observed 3D spatial distances between the loci. Figure 1G The image shows an example of the locations of multiple chromosomal regions in a single cell. The shaded areas represent the convex hull around each chromosome, which is used as the operational definition of the chromosomal region. Figure 1H Showing Figure 1G The same cell distance matrix is ​​shown below. It displays the spatial distances between each pair of chromatin loci. The chromosome order is shown in the heatmap below, where two homologs of each chromosome are displayed. Figure 1I This is a quantification of the volume fraction of each chromosomal region shared by at least one other chromosome in the same cell. The median (center line), 25th to 75th percentiles (boxes), and 5th to 95th percentiles (whiskers) are shown. n = 10,910 chromosome copies (5,455 cells, with two homologous copies per cell for each chromosome).

[0207] Figures 6A-6B The contact frequency matrix derived from genome-scale imaging is shown, along with a comparison with overall Hi-C data. Figure 6A The contact frequency matrix for all 1041 genomic loci imaged in this embodiment is shown. The contact frequency between a pair of loci is calculated by dividing the number of times the distance between the loci measured is less than 500 nm by the total number of distances measured between the two loci. Figure 6B A correlation plot showing the contact frequencies between intrachromosomal locus pairs derived from imaging data and those derived from the overall Hi-C experiment is presented, binned in 500 kb and centered on the target locus. The Pearson correlation coefficient is 0.91.

[0208] Figures 7A-7C The subchromosomal structures derived from genome-scale imaging are shown, along with comparisons to overall Hi-C data. Figure 7A The image shows a contact frequency matrix generated from imaging data of one arm of chromosome 22. The bars below the matrix show the assignment of each locus to compartment A or B based on this matrix. Figure 7B The contact frequency matrix for the same arm of chromosome 22 is shown, calculated from Hi-C data, binned in 500kb units and centered on the target loci. The bars below the matrix show the A and B compartment assignments for each locus based on this matrix, using [data / methods - not specified in the original text]. Figure 7A The same procedure is used for allocation. The A / B compartment allocation derived from imaging data and Hi-C data is the same. Figure 7C A plot showing the correlation frequency between locus pairs on chromosome 22 derived from imaging data and those derived from the whole-cell Hi-C experiment is presented. The Pearson correlation coefficient is 0.91.

[0209] Figure 8 The reproducibility of chromatin imaging experiments between replicates is shown. The figure shows the correlation of pairwise distances between chromatin loci observed in two independent biological replicates of 1041 genomic loci imaging experiments. The Pearson correlation coefficient between replicates is 0.98. The clouds in the upper right represent cross-chromosomal pairwise distances, and the clouds in the lower left represent intrachromosomal pairwise distances.

[0210] Example 2

[0211] This embodiment investigated how cross-chromosomal interactions depend on the epigenetic properties of chromatin. Previous Hi-C and imaging analyses have shown that chromatin is segregated into compartments A and B, enriched with active and inactive chromatin, respectively. Different mechanisms can mediate active-active-inactive-inactive chromatin interactions, such as HP1-mediated heterochromatin condensation and transcription factor and cofactor-mediated active chromatin condensation. In this embodiment, each imaged genomic locus was classified into compartments A and B using an established calling method based on published overall Hi-C data. 38% of the imaged loci belonged to compartment A, which tended to be relatively rich in genes and active chromatin markers, such as H3K27Ac, ​​while 62% belonged to compartment B, which tended to be rich in inactive chromatin markers, such as H3K9me3. To examine whether the degree of cross-chromosomal interactions between active and inactive chromatin differed, genomic loci were ordered in a cross-chromosomal contact frequency matrix, placing all A loci adjacent to each other, and similarly, all B loci together. The matrix shows that, compared to compartment B loci, compartment A loci exhibit a significantly stronger tendency for cross-chromosomal interactions with each other. Figure 2A-2BIn contrast, B-compartment loci did not show similar transchromosomal affinity to each other, but rather a slightly higher probability of transchromosomal interaction with A-compartment chromatin. Figure 2A-2B In other words, cross-chromosomal AA interactions appear to be significantly stronger than AB interactions, while AB interactions appear to be slightly stronger than BB interactions. This contrasts sharply with intrachromosomal cis interactions, where the A and B compartments tend to separate, leading to an enrichment of AA and BB interactions relative to AB interactions.

[0212] Next, the epigenetic dependence of cross-chromosomal interactions was examined at the single-cell level. In individual cells, A-compartment and B-compartment loci exhibited different spatial distributions, with locus A showing a tendency to be more centrally located in the nucleus than locus B. Figure 2C and Figures 9A-9B There is also a considerable degree of mixing between loci A and B. Figure 2C and Figures 9A-9B For each imaged locus on each chromosome, calculate the local densities of its A and B loci from all other chromosomes, and determine the ratio of these two densities (hereinafter referred to as the cross-chromosome A / B density ratio). Figure 2C This quantity provides a measure of the local enrichment of transchromosomal active chromatin near the locus. The majority (62%) of the imaged loci belonged to compartment B, resulting in an overall bias of less than 1 in the A / B ratio. To control for this bias, the distributions of the observed transchromosomal A / B density ratios for loci A and B were compared to those obtained in a randomized control group, where the A and B identities of the imaged loci were randomly shuffled among the imaged loci, while keeping the number of A and B loci constant. Notably, the observed transchromosomal A / B density ratio for loci A was significantly higher than that for loci B, which in turn was higher than that obtained in the randomized control group. Figure 2D This trend was observed in most single cells. Figure 2E These single-cell analyses further support the concept that cross-chromosomal interactions preferentially enrich interactions between active chromatin.

[0213] Figures 2A-2E This demonstrates the interaction between cross-chromosomal contacts that preferentially enrich active chromatin. Figure 2AThe normalized cross-chromosomal contact frequency matrix is ​​shown. It displays the contact frequencies between each cross-chromosomal locus pair (locus pairs on different chromosomes). The loci are rearranged such that compartment A loci appear first, followed by compartment B loci; therefore, the top-left block represents interactions between compartment A loci pairs, and the bottom-right block represents interactions between compartment B loci pairs. Each entry in the matrix is ​​normalized by the median contact frequency of all locus pairs originating from the same chromosome pair to account for different baseline levels of interaction between chromosome pairs. Figure 2B The cross-chromosomal contact frequency distributions of locus A pairs (AA, right; n=72,771 locus pairs), locus B pairs (BB, left; n=193,753 locus pairs), and pairs consisting of one A and one B locus (AB, n=237,986 locus pairs) are shown, originating from... Figure 2A The matrix shown is represented as a histogram in the top plot and as a box plot in the bottom plot, displaying the median (center line), the 25th to 75th percentiles (boxes), and the 5th to 95th percentiles (whiskers). Figure 2C The distribution of A-compartment and B-compartment loci in a single cell is shown. The left figure represents the location of all detected loci within a single z-plane in a single nucleus. A-compartment loci are shown at the top of the scale bar, while B-compartment loci are shown at the bottom. In the right figure, the shading of each locus represents the ratio of the local density of A and B loci across the chromosome, consistent with the shading scale bar shown on the right. Figure 2D The distribution of local cross-chromosomal A / B density ratios for imaged genomic loci is shown. For each locus, the median A / B density ratio across all cells was determined, and the distributions of different loci are shown for A-compartment loci (n=382 loci) and B-compartment loci (n=623 loci). Due to the different forms of genome assembly used in this study and the Hi-C dataset used for compartment recall, 36 of the 1041 imaged loci were not assigned A / B identities. Dark gray histograms represent randomized controls where A and B compartment identities were randomly shuffled while keeping the total number of A loci and the total number of B loci constant. Figure 2E The distribution of enriched A / B density ratios across chromosomes relative to randomized controls is shown. For each imaged cell, the median A / B density ratio of all A loci is divided by the median A / B density ratio of the randomized controls, as shown. Figure 2D As described above, the distribution of this value across all imaged cells (n=5,455 cells) is presented. The distribution of the same enrichment at the B locus is shown (n=5,455 cells). Lines mark values ​​of 1, indicating no enrichment.

[0214] Figures 9A-9B The A-compartment and B-compartment loci show different spatial distributions in a single cell. Figure 9A The left image shows example images of compartment A and compartment B loci displayed in a single z-plane of a single cell. The right image shows the distribution of distances from compartment A and compartment B loci to the nuclear periphery in these single cells. The nuclear periphery was identified as the convex hull surrounding all detected chromatin loci. The histogram shows the distribution of distances from the nuclear periphery to points uniformly sampled within the convex hull around the detected chromatin loci. Figure 9B The population mean distribution of the distances from compartment A loci and compartment B loci to the nuclear periphery is shown. n=382 A loci; n=623 B loci.

[0215] Example 3

[0216] To place the 3D organization of chromatin within the context of its functional activity and other nuclear structures, this embodiment extends the imaging approach to allow for the simultaneous measurement of chromatin organization, transcriptional activity at numerous genomic loci, and landmark nuclear structures. Specifically, 1,041 genomic loci are imaged together with newly transcribed RNA from each of the 1,137 genes located at these loci, simultaneously with important nuclear structures, including nuclear spots and nucleoli. Figure 3A ).

[0217] To allow for imaging of DNA, RNA, and nuclear structures within the same cell, multiplex imaging of intron RNAs of 1,137 genes was performed using a combined imaging strategy similar to that described above for chromatin. Figure 3A Considering that not all genes are transcribed in every individual cell, the density of transcriptional foci should not be as high as the density of chromatin loci. Therefore, RNA was encoded using a 54-bit Hamming weighted 2 code, and 1,137 possible barcodes were selected to encode genes, similar to how barcodes for chromatin imaging are selected to minimize the chance of imaging spatially adjacent genes at the same locus. After RNA imaging, the RNA transcript was enzymatically digested (this step was also performed in our single-modality chromatin imaging experiments), and multiplex DNA FISH was performed as described above to image 1,041 genomic loci. Figure 3A Decoding of genomic loci and nascent RNA transcripts is largely independent, with the additional constraint that transcripts co-localize with their respective genomic loci. This procedure further improves the detection accuracy of transcribed RNA and allows for estimation of the detection efficiency (~90%) of transcriptional bursts at each genomic locus. Finally, nuclear spots and nucleoli are imaged using immunofluorescence targeting known molecular components of these structures. Figure 3AThe location of the nuclear lamina is estimated by calculating the convex hulls of all imaged genomic loci and determining the boundaries of the convex hulls. In summary, these multimodal measurements allow for an integrated single-cell view of 3D genome structure, transcriptional activity, and nuclear organization. Figure 3B These multimodal imaging experiments were performed on approximately 3,700 individual cells in two biological replicates. Chromatin imaging data from these multimodal experiments were also included in the aforementioned five replicates (approximately 5,400 cells) for 3D genome organization analysis.

[0218] From these multimodal experiments, measurements of newly generated RNA transcripts quantified the frequency of transcriptional bursts for each gene, representing the fraction of cells that were active transcriptional genes. Figure 3C ) and median burst size from RNA intron signal brightness ( Figure 3D These measurements showed a high correlation in repeated experiments. Figures 10A-10B The burst frequency exhibited a bimodal behavior, with high burst frequency genes mainly located in compartment A, while low burst frequency genes were present in both compartments. Figure 3C Furthermore, a cutoff spatial distance of 250 nm was used to estimate whether specific chromatin loci were associated with the nucleosome, and a higher association frequency was found between B-compartment loci and the nuclear lamina. Figure 11 ) and higher association frequency between A compartment locus and nuclear spots ( Figure 12 These results are consistent with previous observations that inactive and active chromatin preferentially associate with the nuclear lamina and nuclear spots, respectively. For individual loci, their local cross-chromosomal A / B density ratio is negatively correlated with the frequency of nuclear lamina association. Figure 3E ), and is positively correlated with the frequency of nuclear spots ( Figure 3F Finally, the nucleolus showed preferential association with centromeres, telomeres of certain chromosomes, and chromosomes containing ribosome-encoding genes. Figure 3G These biological results provide further validation for multimodal measurements.

[0219] Notably, for virtually all imaged loci, nuclear lamina associations reduced the observed transcriptional activity. Figure 3H Nuclear spot associations are associated with higher transcriptional activity at most imaged loci. Figure 3H Furthermore, after treatment with α-amanitin to suppress transcription, the association rates of almost all loci with the nuclear lamina generally increased, while the association rates with nuclear spots generally decreased. Figures 13A-13C These results expand upon previous imaging studies of nuclear relocation of single or multiple genomic regions following transcriptional activation or repression, and provide a genome-scale view of the relationship between transcriptional activity and its interaction with nuclear structures.

[0220] Figures 3A-3H Genome-scale imaging of chromatin and transcriptional activity against the background of nuclear structure is shown. Figure 3A This is an illustration of a multimodal imaging scheme that combines imaging of chromatin (left), nascent RNA transcripts (middle), and nucleosomes (right) to generate an integrated view of chromatin organization and functional activity against the background of nuclear structure. Imaging was performed on approximately 1000 genomic loci, nascent RNA transcripts of approximately 1100 genes at target loci, and two types of nucleosomes (nuclear spots and nucleoli). Representative raw images of each imaging modality are shown below: chromatin loci (left), nascent RNA transcripts (middle), and nucleosomes (right: nuclear spots; left: nucleoli) from multiple imaging rounds. Scale bar: 5 micrometers. Figure 3B This is a 3D rendering of chromatin loci, transcription bursts, and nucleosomes in a single cell. Left: All detected chromatin loci, grayscaled by chromosome (based on the chromosome index shown below). Middle: All detected intron RNAs shown as spheres; shading indicates the identity of the imaging gene, and sphere size indicates the size of the transcription burst. Chromatin loci are shown in the background. Right: Volume-filled representation of detected nucleosomes. Nucleoli and nuclear spots are shown in different shades. The nuclear lamina is identified as the convex hull surface surrounding all detected chromatin loci. Figure 3C-3D It is the transcription burst frequency of genes located in compartment A (n=494 genes) and compartment B (n=625 genes). Figure 3C ) and burst size ( Figure 3D The distribution of ). Figure 3E-3F The local cross-chromosome A / B density ratio of each imaged genomic locus is used as the finding with the nuclear lamina ( Figure 3E ) and nuclear spots ( Figure 3F A scatter plot of the frequency of associated loci as a function of their relative frequencies. A locus is considered associated with a nuclear structure if the measured distance between the locus and the nuclear structure is less than 250 nm. The cross-chromosome A / B density ratio shown in the figure is the median value for all imaged cells. Figure 3G The frequency of association between all imaged genomic loci and nucleoli is shown, ordered by genomic location. Vertical lines indicate centromere locations, and parentheses highlight chromosomes containing ribosome-coding genes (rDNA). Figure 3H The effect of nuclear structural association on transcription is shown. When comparing cell populations where loci are nuclear lamina-associated relative to non-nuclear lamina-associated (left) and speckled association relative to non-spotted association (right), the circles represent the fold change in transcription burst frequency at each locus. Dashed lines highlight no change, and solid lines represent the median fold change for each case.

[0221] Figures 10A-10B The reproducibility of the nascent RNA transcript imaging experiments between repeated experiments was demonstrated. Figures 10A-10BThe burst frequency of each gene was shown. Figure 10A ) and burst size ( Figure 10B The correlation between repeated experiments of RNA imaging was measured. The Pearson correlation coefficients were 0.94 and 0.81, respectively.

[0222] Figure 11 The preferential association of compartment B loci with the nuclear lamina is shown. The distribution of association rates between compartment A loci (n=382 loci) and compartment B loci (n=623 loci) and the nuclear lamina is shown. A locus is operationally defined as associated with the nuclear lamina if its distance from the nucleus periphery is less than 250 nm.

[0223] Figure 12 The preferential association between compartment A loci and nuclear spots is shown. The distribution of association rates between compartment A loci (n=382 loci) and compartment B loci (n=623 loci) and nuclear spots is also shown. A locus is operationally defined as associated with a spot if the distance between the locus and the nearest spot is less than 250 nm.

[0224] Figures 13A-13C The changes in the association between the nuclear lamina and nuclear spots were shown after transcriptional repression. Figures 13A-13B Untreated cells were shown. Figure 13A ) and cells treated with the transcription inhibitor α-amanitin ( Figure 13B The image shows a representative image of an individual nucleus with imaged chromatin loci, nucleolus, and nuclear spots. Figure 13C The chart shows the fold change in the association rates of each locus with the nuclear lamina (left) and nuclear spots (right) when transcription was inhibited by α-amanitin. Data points for each genomic locus are shown as circles; solid lines represent the median fold change for all loci in each case, and dashed lines indicate no change.

[0225] Example 4

[0226] In this embodiment, these multimodal single-cell measurements were used to further characterize transchromatographic interactions in the context of transcriptional activity and nuclear structure. Given the observed preferential enrichment of interactions between A-compartment loci in transchromatographic interactions, it was tested whether these interactions were associated with chromatin transcriptional activity. To this end, the local density of A and B chromatin from other chromosomes and the transchromatographic A / B density ratio for each locus in each cell were calculated, and the median of these quantities for the two cell populations was determined: (i) cells where the locus under consideration exhibited transcriptional activity (i.e., RNA burst signaling), and (ii) cells where the locus at least temporarily exhibited transcriptional silencing (i.e., cells where the locus was at least temporarily silencing). Figure 4A It is noteworthy that, in addition to observing that compartment A loci showed a higher local transchromosomal A / B density ratio than compartment B loci (…), Figure 2D-2EEven for the same locus, a consistent trend was observed towards higher transchromatic A density and A / B density ratio during transcription when the locus was active. Figure 4B and Figure 14 This correlation observed between transcriptional activity and transchromatographic interactions is consistent with two possible explanations: higher epigenetic or transcriptional activity of chromatin loci increases their rate of transchromatographic interactions, or the localization of loci in environments rich in active chromatin enhances their transcriptional activity.

[0227] Figures 4A-4F This shows that preferential transchromosomal interactions between active chromatin are transcription-related and are disrupted by treatments that interfere with condensate formation. Figure 4A These are single-cell images of chromatin loci and transcriptional activity. Left: Location of all imaged A compartments (above the scale bar) and B (below) loci in a single z-plane from a single nucleus. Middle: Local cross-chromosomal A / B density ratio of the same loci based on the grayscale scale bar. Right: Same as the middle image, where detected transcriptional bursts are superimposed and displayed as circles. Figure 4B The comparison of local cross-chromosome A / B density ratios for each locus in the transcriptional and silencing states is shown. For each genomic locus containing at least one imaged gene, the cross-chromosome A / B density ratios were calculated for cells with active transcription (called transcription) and cells without transcription (called silencing). The cross-cell median is shown for each state. Loci are sorted by their A / B density ratios in the silencing state, and A / B density ratios are plotted for both silencing and transcriptional states. Figure 4C The normalized cross-chromosomal contact frequency matrix of cells treated with α-amanitin to suppress transcription is shown. Figure 2A As described above, the matrix is ​​ordered and standardized. Figure 4D The distribution of contact frequencies for AA, BB, and AB, shown as a box plot, is illustrated in [the image / plot]. Figure 2B As described in [the document]. There are n=72,771 locus pairs for AA, n=193,753 locus pairs for BB, and n=237,986 locus pairs for AB. Figure 4E-4F and Figures 4C-4D Same, but cells were treated with 1,6-hexanediol.

[0228] Figure 14 The local density of transchromosomal A loci near each imaging locus is shown when the locus is in an active transcriptional or silent state. For each locus, cells are divided into two groups depending on whether the locus is active transcriptional or silent. The median local density of A loci in these two groups of cells (transcribed and silent) is shown. Loci are ordered based on their local transchromosomal A locus density in the silent state.

[0229] Because nuclear spots are among the most prominent nuclear bodies containing loci of active transcription, it is unknown whether association with nuclear spots can provide a simple explanation for the observed preferential occurrence of active-active chromatin interactions across chromosomes. Interestingly, when the analysis was limited to loci unrelated to nuclear spots, the same trend of enrichment of AA relative to AB and BB interactions was observed in cross-chromosomal contacts. Figure 15A ) and active transcription loci showed the same trend of higher local A / B density ratios compared to silent loci. Figure 15B It is noteworthy that these trends were observed even when considering only loci associated with the nuclear lamina and therefore in environments rich in B-compartment chromatin. Figures 16A-16B The latter result also suggests that the observed enrichment of activity-activity cross-chromosomal interactions cannot simply explain the fact that active chromatin is more concentrated in the center of the cell nucleus.

[0230] Figures 15A-15B This shows the enrichment of active-active transchromosomal interactions in chromatin loci unrelated to nuclear spots. Figure 15A This shows the cross-chromosomal contact frequencies between A locus pairs (AA), B locus pairs (BB), and pairs consisting of one A and one B locus (AB), considering only cells where two loci are not associated with nuclear spots. Contact frequencies are shown as follows: Figure 2A The distribution is standardized. It is represented as a box plot, such as... Figure 2B As described in the text. For AA, there are n=72,771 locus pairs (left), for BB, there are n=193,753 locus pairs (right), and for AB, there are n=237,986 locus pairs (middle). For comparison, the median of all data, regardless of the blob association state, is shown as a triangle. Figure 15B The fold change in the local cross-chromosomal A / B density ratio between transcriptional and silencing states is shown for loci unrelated to nuclear spots. For each genomic locus, the fold change in the local cross-chromosomal A / B density ratio between transcriptional and silencing states was calculated, considering only cells where the locus was not associated with a nuclear spot. The median A / B density ratio for each state (transcriptional or silencing) of each locus was determined, with the fold change between the two states shown on the left (each circle corresponds to a genomic locus). Regardless of the nuclear spot association state of the locus, the corresponding fold change derived from all data is shown on the right for comparison. Dashed lines indicate no change, and solid lines indicate the median fold change for all loci in each case.

[0231] Figures 16A-16B This study showed enrichment of active-active transchromosomal interactions in chromatin loci associated with the nuclear lamina. Figure 16AThe cross-chromosomal contact frequencies are shown for A locus pairs (AA), B locus pairs (BB), and pairs consisting of one A and one B locus (AB), considering only locus pairs associated with the nuclear lamina (within 250 nm). Contact frequencies are as follows: Figure 2A The distribution is standardized. It is represented as a box plot, such as... Figure 2B As described above. There are n=72,771 locus pairs for AA (left), n=193,753 locus pairs for BB (right), and n=237,986 locus pairs for AB (middle). For comparison, the median of all data, regardless of the nuclear lamina association status, is shown as a triangle. In these cases, the variance is relatively large because only a relatively small fraction of loci are associated with the nuclear lamina, but the differences between different types of AA, BB, and AB pairs are statistically significant (P < 10⁻¹⁰). Figure 16B The fold change in the local cross-chromosomal A / B density ratio of loci associated with the nuclear lamina is shown between transcriptional and silencing states. For each genomic locus, the fold change in the local cross-chromosomal A / B density ratio between transcriptional and silencing states is calculated, considering only cells where that locus is associated with the nuclear lamina. The median A / B density ratio for each state (transcriptional or silencing) of each locus is determined, and the fold change between the two states is shown on the left (each circle corresponds to one locus). Outliers (33 loci above and 18 below the presented bar) are omitted to allow for clearer visualization of the median fold change. Regardless of the nuclear lamina association state, the fold change derived from all data is shown on the right for comparison. Dashed lines indicate no fold change, and solid lines indicate the median fold change for all loci in each case.

[0232] The results showed that transchromatic interactions preferentially occur between active chromatin loci, and this behavior was consistently observed in multiple different nuclear settings. The next step was to explore what might lead to this preferential, widespread activity-activity chromatin interaction. Since RNA polymerase II (Pol II) contains low-complexity domains (LCDs) and can form aggregates, transcriptional repression of Pol II was tested to determine whether it could lead to these preferential transchromatic interactions. This was achieved by using the transcriptional repressor α-amanitin, which causes Pol II dissociation and degradation. Although transcription was eliminated and nuclear structure and its association with chromatin were altered (…),… Figures 13A-13C However, treatment with α-amanitin did not significantly reduce preferential transchromatic interactions between active chromatin. Figures 4C-4DThis indicates that additional or other active chromatin-binding factors are involved in these transchromatic interactions. The results suggest that several other proteins associated with active chromatin contain LCDs that may potentially mediate aggregate formation. Therefore, the aim was to more generally interfere with aggregate formation using 1,6-hexanediol (a drug known to disrupt hydrophobic interactions between LCDs). Notably, after treating cells with 2% 1,6-hexanediol for 45 minutes, the preferential enrichment of active chromatin interactions in transchromatic contacts was largely eliminated. Figure 4E-4F This suggests a potential role in the formation of aggregates during the establishment or maintenance of these interactions.

[0233] In summary, these examples develop a large-scale multiplex FISH approach for imaging 3D chromatin organization at the genome scale in single cells and further demonstrate the ability to place 3D genome organization within its native structural and functional context by combining chromatin and nascent transcript imaging (both at the genome scale) with nuclear structure identification. This provides a comprehensive view of nuclear organization within a single cell. While target loci are consistently selected across all chromosomes here to provide an unbiased view of overall 3D genome organization, this approach can also be used to target genomic loci with specific structural and functional properties, such as promoters, enhancers, and loci that bind to specific nuclear structural proteins, to investigate interactions between these loci and their relationship to transcription and other chromatin functions. The broad applicability of this approach to a wide range of questions related to genome organization can elucidate the mechanisms controlling chromatin organization and the role of chromatin structure in regulating genome function.

[0234] Example 5

[0235] This embodiment illustrates various materials and methods that can be used in the above embodiments.

[0236] Target genomic regions. For chromatin imaging, genomic loci were selected for imaging in the following manner. For each human autosome and X chromosome, a 30kb segment was selected at approximately 3Mb intervals. If this interval resulted in fewer than 30 selected loci on a given chromosome, the interval for that chromosome was reduced until at least 30 loci were selected on all chromosomes. This resulted in a total of 1,041 target genomic loci for imaging, with the number of loci on a single chromosome ranging from 30 to 80. Coding probes (approximately 400 oligonucleotide probes) were then designed for each 30kb segment for combined FISH imaging.

[0237] For imaging of nascent RNA transcripts, all intron-containing genes that completely or partially overlapped with target genomic loci were selected. Coding probes were then designed for the introns of all these RNAs, resulting in approximately 20 coding probes per RNA, with the targeting sequences of the coding probes positioned as close as possible to the transcription start site (TSS). A total of 1,137 genes were targeted.

[0238] Barcode Design for Combined FISH Imaging. Binary barcodes for imaging 1,041 genomic loci were selected as follows: First, all possible 100-bit binary barcodes with a Hamming weight of 2 (i.e., each barcode contains two "1" bits and 98 "0" bits) were generated, and 1,041 barcodes were randomly selected from this list. The selected barcodes were then randomly assigned to the 1,041 genomic loci. Next, barcodes were randomly swapped between used and unused code libraries and between loci from different chromosomes to minimize the difference in the number of loci appearing (i.e., reading "1") at different positions on each chromosome. This resulted in approximately equal numbers of loci imaged per position on each chromosome. To optimize the association between barcodes and loci within each chromosome, barcodes were allowed to be swapped between loci within the same chromosome, and optimization was performed for the maximum minimum genomic distance between loci with barcodes reading "1" at the same code position. When comparing code assignments with the same minimum genomic distance, the code assignment that minimizes the coefficient of variation of the genomic distance is selected (so that the genomic distance has a larger mean and a smaller standard deviation).

[0239] Similarly, barcodes were chosen for imaging the nascent RNA transcripts of 1,137 genes, but a 54-bit Hamming distance 2 code was used instead of a 100-bit Hamming distance 2 code.

[0240] Encoding probe design. The encoding probes used for chromatin imaging were synthesized from an oligonucleotide library purchased from Twist Biosciences. Each oligonucleotide in this library used the following subsequence (from 5' to 3'):

[0241] 1. A 20-nucleotide (nt) forward initiation region used for PCR amplification and reverse transcription (RT)

[0242] 2.20-nt readout sequence, corresponding to one of the positions that will be used to image the genomic locus targeted by the probe.

[0243] 3.40-nt target sequence, designed to bind uniquely to a single target genomic locus.

[0244] 4. Additional copies of the above 20-nt readout sequence

[0245] 5. 20-nt reverse promoter sequence for PCR amplification

[0246] Select forward and reverse priming sequences from the previously generated list of random 20-nt sequences optimized for PCR.

[0247] The readout sequence was selected using the following procedure. First, a list of 30-nt sequences with minimal homology to the human genome was created. Then, a subset of these sequences was sorted based on the observed signal-to-noise ratio (SNR), and the top 100 were selected as DNA readout probes. Finally, the readout sequence was selected by reverse complementation of the last 20 nt of each of these sequences.

[0248] Similarly, 40-nt target sequences were selected. In short, the following procedure was repeated for each genomic region of interest (see the discussion on “Target Genomic Regions” above). First, a list of all 40-nt sequences complementary to the genomic region of interest was created (starting from each possible base in the target region). Sequences were then filtered by requiring them to be within defined melting temperatures and GC content ranges. The remaining sequences were then further filtered by restricting the allowed homology with the human genome, human transcriptome, and databases containing repetitive sequences. Homology was determined by creating a table of all possible 17-nt sequences and their occurrences in target databases (e.g., the human genome, human transcriptome) and calculating the total number of exact 17-nt matches with a given candidate sequence. Finally, after the final filtering step, target sequences were selected from the remaining sequences such that there was no genomic overlap between any pair of target sequences.

[0249] To generate full-length probes, each 40-nt target sequence selected for each target genomic locus was alternately assigned to two groups spanning the entire target locus. Each of these groups was associated with a single readout sequence, corresponding to one of two positions in which the locus would be imaged. Each target sequence was then ligated to two identical copies of the readout sequence assigned to its group, and then to forward and reverse PCR primers.

[0250] The probes used for RNA imaging are similarly designed, except that each probe contains three copies of the same readout sequence, one at the 5' end of the target region and two at the 3' end. The readout sequences used for RNA imaging are orthogonal to those used for DNA imaging and are selected from a list of test readout sequences with the same order.

[0251] Coding probe synthesis. The coding probe is amplified from the template library described above (see “Coding Probe Design” above). This is accomplished using an amplification protocol involving the following steps:

[0252] 1. The initial oligonucleotide library was amplified using finite-cycle PCR for approximately 20 cycles. The reverse primers used in this step also introduced the T7 promoter sequence via primer extension.

[0253] 2. The obtained product was purified by column purification and further amplified and converted into RNA by high-yield in vitro transcription reaction.

[0254] 3. RNA products are converted into single-stranded DNA through reverse transcription.

[0255] 4. The product from the previous step is subjected to alkaline digestion (to remove residual RNA and primer DNA) and then purified by column chromatography.

[0256] 5. If necessary, the product from the previous step is dried in a vacuum and resuspended in water to achieve the desired primary probe concentration.

[0257] All primers were purchased from Integrated DNA Technologies (IDT).

[0258] Cell culture and hybridization with coding probes. Cell preparation was performed as follows. IMR-90 cells were purchased from the American Type Culture Collection (ATCC, CCL-186) and cultured according to recommended protocols. To avoid potential chromatin structural alterations, all cells in this study were seeded at the following densities within 6 weeks of the start of culture.

[0259] To prepare for DNA imaging, cells were plated on 40 mm, round #1.5 coverslips (Bioptechs, 0420-0323-2) at a density of approximately 500,000 cells per coverslip. Cells were allowed to grow at 37°C and 5% CO2 for approximately 2 days until confluence. In transcriptional repression experiments, the cell culture medium was replaced with fresh medium containing 100 μg / mL α-amanitin (Sigma-Aldrich, A2263) 6 hours before cell fixation. For experiments using 1,6-hexanediol (Sigma-Aldrich, 240117), coverslips were coated with 10 μg / mL fibronectin (Sigma-Aldrich, F1141) before cell plating, and the medium was replaced with fresh medium containing 2% w / v 1,6-hexanediol for 45 minutes. The cultures were then fixed with 4% paraformaldehyde (PFA) in PBS at room temperature for 10 minutes, followed by washing 2-3 times in PBS. Cell permeabilization was then performed in two steps: first, cells were treated with 0.5% v / v Triton-X (Sigma-Aldrich, T8787) in PBS at room temperature for 10 minutes. Then, cells were treated with 0.1M hydrochloric acid (HCl) at room temperature for 5 minutes, followed by washing 2-3 times in PBS. After HCl treatment, cells were treated with a solution of 0.1 mg / mL ribonuclease A (ThermoFisher, EN0531) dissolved in PBS at 37°C for 30-45 minutes to remove potential sources of off-target RNA binding. Following this treatment, cells were incubated for approximately 10 minutes in a prehybridization buffer consisting of 2x saline-sodium citrate buffer (SSC; Ambion, AM9763) and 50% formamide (Ambion, AM9342). Next, invert a cell coverslip and place it on a drop of 50 μL hybridization buffer (2xSSC, 50% formamide, 10% dextran sulfate (Sigma-Aldrich, D8906), containing a coding probe mixture with or without 10 μg human Cot-1 DNA (ThermoFisher, 15279011) at a total concentration of approximately 25 μmol) in a 60 mm culture dish. Partially immerse the dish in a water bath at approximately 90°C for 3 minutes, then incubate in a humidified chamber at 47°C for 16–36 hours. After incubation with the coding probe, wash the sample in 2xSSC and in 40% formamide for 30 minutes, then fix with 4% PFA in 2xSSC at room temperature for 10 minutes.The sample was then incubated with reference beads (ThermoFisher F8805 or ThermoFisher F8792) in 2xSSC for 2–3 minutes, and stained with 1 μmol of 4',6-diamidinyl-2-phenylindole (DAPI; ThermoFisher D1306) in 2xSSC for 5–10 minutes, and then stored in 2xSSC until imaging.

[0260] For experiments involving RNA imaging, all buffers used from cell fixation onwards contain a ribonuclease inhibitor (NEB M0314 or Fisher Scientific N2615) diluted 1:10–1:1,000. RNA staining is performed as described above until treatment with HCl. Following this step, cells are incubated in prehybridization buffer for 10 minutes, then a cell coverslip is inverted and placed in a drop of hybridization buffer containing approximately 1 μmol of coding probes targeting RNA introns, as described for DNA staining. However, in this case, no 90°C heat denaturation is performed; cells are immediately incubated in a humidified chamber at 47°C for 16–36 hours. After incubation with the coding probes, the sample is washed in formamide solution and fixed with PFA as described above for DNA. It is then incubated with reference beads and stained with 1 μmol of DAPI, and then stored in 2xSSCs until imaging. After RNA imaging, the sample is removed from the microscope, the cells are treated with ribonuclease A, and then DNA hybridization is performed in the same manner as described above for DNA imaging without RNA imaging.

[0261] The readout probes used for FISH imaging undergo sequential hybridization. All fluid exchange in this part of the protocol is achieved using a custom-designed fluid system in which a coverslip is mounted in an FCS2 flow chamber (Bioptechs, 060319-2). The fluid system utilizes three to four computer-controlled eight-way valves (Hamilton, MVP, and HVXM8-5) and a computer-controlled peristaltic pump (Gilson, MINIPLUS3). Together, these components allow control over the rate and type of fluid flow at any given time.

[0262] Each round of hybridization uses the following steps:

[0263] 1. As described below, flow in a hybridization buffer containing a set of oligonucleotide probes specific to each round.

[0264] 2. Incubate at room temperature for 10 minutes.

[0265] 3. Flowing wash buffer

[0266] 4. Incubate for approximately 200 seconds.

[0267] 5. Flow Imaging Buffer

[0268] An imaging buffer was prepared comprising 60 mM Tris pH 8.0, 10% w / v glucose, 1% glucose oxidase scavenger solution (containing ~100 mg / mL glucose oxidase (Sigma-Aldrich, G2133) and 1:3 diluted catalase (Sigma-Aldrich, C3155)), 0.5 mg / mL 6-hydroxy-2,5,7,8-tetramethyltryptane-2-carboxylic acid (Trolox; Sigma-Aldrich, 238813), and 50 μmol Trolox Quinone (generated by UV irradiation of a Trolox solution). Trolox was dissolved in methanol and then added to the solution. After preparation, the imaging buffer was coated with a layer of mineral oil approximately 0.5 cm thick to prevent exposure to oxygen.

[0269] Hybridization buffer and wash buffer consist of 35% and 30% formamide in 2xSSC, respectively, and the hybridization buffer also contains 0.01% v / v Triton-X. The hybridization buffer is stored separately for each hybridization round and contains two or three sets (one for DNA and one for RNA imaging) of readout probes. Fluorescent signals are introduced in one of two ways:

[0270] 1. For DNA imaging, each round of hybridization buffer includes two fluorescent readout probes, one labeled with Cy5 or Alexa647 and the other with Alexa750. The fluorescent readout probes use: 1) fluorescently labeled oligonucleotides complementary to a readout sequence common to all coding probes used for localization imaging, added at a concentration of 100 nM; or 2) a combination of an adaptor oligonucleotide having a sequence complementary to the readout sequence, linked to an additional readout sequence (called a secondary readout sequence) common to all adaptors (more precisely, common to all adaptors in each color channel) and orthogonal to all other used readout sequences, and a fluorescently labeled oligonucleotide probe complementary to this secondary readout sequence. The adaptor and secondary readout probe are premixed at a 1:1.5 ratio and added to a final concentration of ~100 nM. For some experiments, the adaptor and readout probe are hybridized to the sample sequentially. This allows the use of lower concentrations of the more expensive secondary readout probe.

[0271] 2. For RNA imaging, each round of hybridization buffer contains three adaptor oligonucleotides (to be detected in three different color channels), each binding to a different readout sequence, and each containing an additional secondary readout sequence. All adaptors corresponding to the same color channel share the same secondary readout sequence. Each round includes two discrete hybridization steps: first, the adaptors are infused, hybridized, and excess material is washed away. Then, three fluorescent readout probes (labeled Cy3, Cy5, and Alexa750, respectively) complementary to the secondary readout sequences on the adaptors are sequentially infused. The fluorescent readout probes used for RNA imaging contain disulfide bonds linking the fluorophore to the secondary oligonucleotides to allow for efficient signal removal between rounds. After fluorescent readout hybridization, the imaging buffer is infused and the signal is collected.

[0272] Before the next round of hybridization of readout probes or adapter probes, the fluorescence signal from the readout probes or secondary readout probes in the current round is removed, as described in "Signal Removal Between Hybridization Rounds" below.

[0273] Prior to the first round of hybridization, an imaging round was performed to obtain DAPI signals and identify nuclear boundaries. Then, the entire set of 1,041 genomic loci was imaged in 50 rounds of hybridization, with two color channels per round. In each round, the genomic loci were imaged in 3D by stepping along the z-dimensional plane. Newborn RNA transcripts of 1,137 genes were also imaged in 3D in three colors over 18 rounds. Additional rounds were used to relabel the genomic locus set and assess color difference and bleed between color channels. Imaging approximately 60 fields of view, containing a total of approximately 1,000–2,000 cells, took approximately 3 days.

[0274] The 3-4 valve system allows for the loading of up to 20-30 different hybridization solutions. As a result, after all the fluid system channels were exhausted, the sample chamber was bypassed, and all channels used for hybridization were washed with a 30% formamide aqueous solution. Next, the chambers were reconnected, and the next set of hybridization and imaging rounds was performed.

[0275] Antibody labeling and imaging. Antibody imaging is performed immediately after RNA or DNA imaging. After imaging is completed using the above protocol, the sample undergoes the following steps:

[0276] 1. Incubate the sample with blocking solution (PBS containing 0.1% v / v Tween-20 (Sigma-Aldrich P9416) and 1% w / v bovine serum albumin (BSA; Jackson Immunoresearch 001-000-162)) for 30 minutes.

[0277] 2. Incubate the sample with the primary antibody diluted in the blocking buffer for 1 hour.

[0278] 3. Wash the sample three times in PBS containing 0.1% Tween-20, for 5 minutes each time.

[0279] 4. Repeat steps 2 and 3 for the fluorescently labeled secondary antibody.

[0280] All buffer exchanges were performed on the microscope using the microfluidic system described above. The Cy5 color channel was used for imaging, and photobleaching was used to eliminate signal between consecutive antibody labels.

[0281] The following groups of primary and secondary antibodies were used:

[0282] 1. For nuclear dot imaging, a primary antibody against SC35 (Abcam, ab11826) (a splicing factor commonly used as a nuclear dot marker) diluted 1:200 from the stock solution was used, and a Cy5-labeled donkey anti-mouse secondary antibody (Jackson Immunoresearch, 715-175-150) diluted 1:1,000 from the stock solution concentration was used.

[0283] 2. For nucleolar imaging, antifibrin antibody (Abcam, ab5821) diluted 1:200 from the stock solution was used, as was donkey anti-rabbit secondary antibody labeled with Alexa 657 dye (Jackson Immunoresearch, 711-605-152) diluted 1:1,000 from the stock solution concentration.

[0284] Signal removal between hybridization rounds. Before imaging each round, the signal from the previous round (or endogenous background, in the case of the first round) is extinguished. This is achieved by photobleaching the signal. Photobleaching is performed by changing the buffer to 2xSSC and illuminating each field of view for 10 seconds with the maximum available power of 647 and 750 lasers (and 560 lasers when imaging RNA). In RNA imaging experiments, the buffer used for bleaching also contains 50 mM tris(2-carboxyethyl)phosphine (TCEP; Sigma-Aldrich, C4706) to cleave the disulfide bonds connecting the fluorophore and the readout probe. The DAPI signal disappears due to the high formamide concentration in the hybridization and washing buffers.

[0285] Image acquisition. Images were acquired using a custom-designed microscope system. This system was constructed around the body of a Nikon Ti-U microscope with a Nikon CFI Plan Apo Lambda 60x oil immersion objective lens with a 1.4 NA aperture. Illumination was based on one of two options:

[0286] 1. Solid-state single-mode lasers with the following wavelengths: 405 nm (Coherent, Obis 405 nm LX 200mW), 560 nm (MPB Communications, 2RU-VFL-P-2000-560-B1R), 647 nm (MPB Communication, 2RU-VFL-P-1500-647-B1R), and 750 nm (MPB Communication, 2RU-VFL-P-500-750-B1R). In this case, the outputs of the 560-nm, 647-nm, and 750-nm lasers are controlled by acousto-optic tunable filters (AOTF), while the 405-nm laser is directly controlled via its laser control box. Custom-designed dichroic filters (Chroma, zy405 / 488 / 561 / 647 / 752RP-UF1) and emission filters (Chroma, ZET405 / 488 / 461 / 647-656 / 752m) are used to separate excitation and emission illumination.

[0287] 2. The Lumencor CELESTA optical engine (an illumination system based on fiber-coupled solid-state lasers) has the following wavelengths: 405nm, 446nm, 477nm, 520nm, 546nm, 638nm, and 749nm. This system is used in conjunction with a five-bandpass dichroic filter (IDEX, FF421 / 491 / 567 / 659 / 776-Di01-25x36) and a five-bandpass filter (IDEX, FF01-441 / 511 / 593 / 684 / 817-25).

[0288] Image acquisition was performed using a scientific CMOS camera (either a factory-calibrated Hamamatsu FLASH4.0 or Hamamatsu C13440 for single-molecule imaging). Sample position in three dimensions was controlled using an XYZ stage (Ludl). A custom autofocus system was used to maintain a constant focal plane over extended periods. This was achieved by comparing the relative positions of two IR laser beams (Thorlabs, LP980-SF15) reflected from the glass-fluid interface and imaged on a separate CMOS camera (Thorlabs, uc480).

[0289] For each experiment, approximately 60 fields of view (FOV) were selected for imaging, avoiding areas with sparse cells (we typically determined each FOV to contain 10–50 cells). Each camera FOV had 1,000 x 1,000 pixels (camera pixels correspond to 153 nm in each dimension of the imaging plane) or 2,048 x 2,048 pixels (camera pixels correspond to 108 nm in each dimension of the imaging plane).

[0290] After each round of hybridization (see “Sequential Hybridization of Readout Probes for FISH Imaging” above), z-stack images of each FOV were acquired in 3 or 4 colors: 647 nm and 750 nm illumination (or 560 nm, 647 nm, and 750 nm illumination for RNA imaging in the case of combined DNA and RNA imaging) were used to acquire FISH images, and the reference beads were imaged using 560 nm illumination (or 405 nm illumination in the case of combined RNA and DNA imaging). For the first round of imaging, the DAPI signal was imaged using 405 nm illumination, while for antibody imaging, the 647 nm excitation channel was used after RNA or DNA imaging. Successive z-slices were spaced 85, 100, or 150 nm apart, covering the entire nuclear volume of all imaged cells. At each z-position, images were acquired in all channels before moving the stage, and at a rate of ~10 Hz.

[0291] Image analysis and point fitting for DNA and RNA imaging. The following analysis pipeline is applied to the field of view (FOV) of each imaging to obtain the three-dimensional (3D) locations of all loci of interest:

[0292] 1. Reference points were fitted in all imaging rounds and used for image alignment.

[0293] 2. In the first round of imaging (before the first round of hybridization), the DAPI signal is used to identify the boundaries of individual cell nuclei and for image registration between RNA and DNA imaging.

[0294] 3. Fit a 3D Gaussian function to each identified diffraction confinement point within the nucleus to identify their centroid and brightness above the local background.

[0295] 4. Using custom algorithms and software (described in detail in the sections “Decoding Algorithm for Fitted DNA Points” and “Decoding Algorithm for Fitted RNA Points”), the fitted points are compared with other localizations within the same nucleus in all rounds of hybridization to identify the loci from which they originate.

[0296] Point fitting for DNA and RNA imaging. Signals from individual FISH imaging rounds were fitted using 3D Gaussian fitting. To make the analysis more manageable, the number of fitted points for each image used for decoding was fixed at 125 (approximately 3 times the expected number of different loci in noise-free conditions).

[0297] Drift correction. Reference bead fitting is performed in the same manner as described above. Then, a set of reference bead positions is compared across multiple rounds of hybridization, and a rigid transformation is applied to minimize the sum of squared differences in the relative positions of the beads.

[0298] Color correction. Multicolor imaging is performed by independently labeling the same set of genomic loci in each imaging channel and comparing the signals of the same loci in different channels to correct bleed and color difference.

[0299] Nucleus segmentation. DAPI images from the first round of imaging are used to identify the volume of individual cell nuclei and allow for cell segmentation. This is achieved by building and training a convolutional neural network that takes the maximum projection of the DAPI image onto the xy-plane as input.

[0300] Additional Image Analysis: Image Registration Between DNA and RNA Imaging. In experiments involving both DNA and RNA imaging, the DAPI signal was first used for coarse image registration (to camera pixel accuracy) between two sets of images via 2D image correlation (all images in each set were aligned with the DAPI image using reference beads). After performing the first round of RNA decoding (see “Decoding Algorithm for Fitted RNA Points” below), a finer alignment was calculated by assuming that the displacement between the location of nascent RNA and the DNA locus containing it should be averaged to zero when considering all imaged genes and cells in the field of view. Accordingly, an additional rigid transformation was calculated to minimize the average displacement between the imaged nascent RNA and its corresponding DNA locus, and this was used as the final alignment.

[0301] Nuclear bodies are identified using immunofluorescence imaging. The locations of nuclear bodies (nuclear spots and nucleoli) are extracted from the immunofluorescence signal by applying a threshold to the intensity of the immunofluorescence signal, thus generating a pixelated mask that identifies high immunofluorescence signals. This mask is then considered as a set of pixelated locations that "contain" the nuclear body.

[0302] Decoding algorithm for fitted DNA points. Identification and 3D localization of each genomic locus are achieved through the following steps:

[0303] 1. Generate a list of the drift and aberration correction locations of all identified points in each round of imaging.

[0304] 2. For each detected point in each imaging round, it was found that all points from other rounds were within a set cutoff distance (approximately 150 nm in x, y, and z) from their location. All these point pairs were retained for further analysis, regardless of whether the barcodes generated by the point pairs (based on which circle and color channel they appeared in) corresponded to valid barcodes (barcodes assigned to genomic loci).

[0305] 3. For each pair of points, calculate three quality metrics:

[0306] A. Displacement between the 3D positioning of two points

[0307] B. Brightness difference between the two points

[0308] C. Average brightness of the two points

[0309] The brightness of each point is standardized by the median brightness of all points in the corresponding bit.

[0310] 4. Then, based on whether the point pairs correspond to valid barcodes (and therefore possibly genomic loci), they are divided into two groups. Within each group, the distribution of the quality metric is calculated. For convenience, the distribution of the quality metric for point pairs from invalid barcodes is called the "invalid distribution," and the distribution from all valid barcodes is called the "valid distribution."

[0311] 5. For each point pair, the three quality metrics from Step 3 are combined into a single metric by calculating the combined Fisher p-value for each candidate point pair against the “effective distribution” in Step 4. This is considered the overall quality score for each point pair and is calculated for each pair as follows: for each of the three metrics, the scores of other point pairs in the “effective distribution” are calculated using the lower quality metric, and these scores are multiplied together. Then, the expectation-maximization procedure is used to sequentially select the two point pairs with the highest quality scores corresponding to each target chromatin locus, and the “effective distribution” is updated again. This optimization procedure is repeated until convergence. After convergence, the last set of point pairs (each corresponding to a chromatin locus) is used to determine the 3D spatial location of the locus.

[0312] 6. Following step 5, an improved K-means algorithm is used to separate chromatin loci belonging to the same chromosome into two homologs. In contrast to the standard K-means clustering algorithm, which divides points into two groups and minimizes the radius of gyration within each group, points are progressively switched between groups to first maximize the score assigned to points in each homolog, and then minimize the radius of gyration of each homolog.

[0313] 7. After separating the two homologs, their centroids and the distance of each point pair from step 2 are calculated as the centroids of their parent chromosomes. In addition to the three measures considered in step 3, the distance from the chromosome center is added as another quality measure, and steps 3-6 are repeated.

[0314] 8. Finally, the point pairs from step 7 are filtered to remove point pairs with quality scores similar to the "invalid distribution".

[0315] The remaining point pairs after step 8 are used to determine the final location of chromatin loci and trace chromatin structure.

[0316] Decoding algorithm for fitted RNA points. The following procedure is used to decode signals from RNA imaging rounds:

[0317] 1. Generate a list of the drift and aberration correction locations of all identified points in each round of imaging.

[0318] 2. For each detected point in each imaging round, find all points from other rounds within a set cutoff distance from its location, and retain them as candidate RNA bursts if these point pairs form valid barcodes.

[0319] 3. After initial image registration (based on DAPI images) and drift and aberration correction, the position of each of these candidate RNA bursts is compared with the position of the DNA locus containing the relevant gene, and if they are within a set threshold distance, they are retained.

[0320] 4. Refine the registration between DNA and RNA images based on the displacement between the initially decoded RNA localizations (from step 3) and the locations of the DNA loci containing them, as described in the “Image Registration Between DNA and RNA Images” section above.

[0321] 5. Again, compare the locations of all candidate RNA bursts with the locations of the DNA loci containing the genes they decode, this time using refined image registration. If, at this stage, the nascent RNA is located within the cutoff distance of the DNA locus containing it, it is considered a detected transcription burst.

[0322] Additional analysis: Identification of the nuclear lamina. The location of the nuclear lamina was estimated by generating the minimum 3D convex hull around the locations of all decoded chromatin loci in a given cell (using Python's scipy package).

[0323] Spatial distance. The spatial distance between any pair of loci is simply calculated as the Euclidean distance between their fitted 3D Gaussian centers, multiplied by an appropriate ratio involving camera pixels and z-step to physical distance. In the case of distance from the nucleus body, the minimum Euclidean distance to all identified nucleus body “locations” or the minimum distance to the convex hull surface defining the nucleus layer is calculated.

[0324] The contact frequency matrix is ​​derived from imaging. To calculate the contact frequency between any given pair of loci, the number of measured distances between that pair that are less than a set threshold is calculated. This number is then divided by the total number of distances measured for that pair of loci.

[0325] Local density analysis. To calculate the transchromosomal local density of A-compartment and B-compartment loci at each decoding location, the spatial distance between each pair of chromatin loci per cell was calculated. For each locus, the local A / B density ratio was calculated as follows:

[0326] 1. The density contribution of each other locus was calculated from different chromosomes by evaluating the Gaussian function value of the distance between two loci with a standard deviation of 500 nm (adjusted to account for cell size variability).

[0327] 2. Then the total A density at the locus is calculated as the sum of the Gaussian function values ​​of all cross-chromosomal A loci, and the total B density is calculated in a similar manner.

[0328] 3. The A / B density ratio of a transchromosomal A-compartment locus is obtained by dividing the total density of transchromosomal B-compartment loci by the density of the transchromosomal B-compartment locus.

[0329] Estimating detection efficiency in multiplex RNA imaging. The estimation of detection efficiency for transcriptional burst events is performed as follows:

[0330] 1. Consider all targeted genomic loci containing genes whose RNA introns are imaged. For any of these genomic loci, if the gene is transcribed, its corresponding RNA signal should appear in two predefined positions. Know the rate at which each of these two positions is not detected ( p This allows for the derivation of RNA detection efficiency. A group of genomic loci colocalized with RNA signals (within approximately 150 nm) in at least one of the two expected loci in their respective genes were identified.

[0331] 2. The fraction of loci colocalized with RNA signals from the complete set of chromatin loci identified in step 1. f It is determined from only one (not two) corresponding position in its gene. Based on measurements... f (8.4%), which should be equal to ,estimate p (4.4%).

[0332] 3. Use equations The overall detection efficiency for detecting co-location signals in two bits was calculated and found to be approximately 92%.

[0333] Hi-C data analysis. Hi-C data for IMR-90 cells were acquired and loaded using a pipette (straw) method. To identify A / B compartments within individual chromosomes, established, publicly available protocols were followed. To compare the contact frequencies derived from imaging data with Hi-C, target-region-centered bins were created, and Hi-C data for these bins were obtained by summing the read counts from higher-resolution Hi-C data.

[0334] Example 6

[0335] The three-dimensional (3D) organization of chromatin regulates many genomic functions. However, understanding 3D genomic organization is hampered by a lack of tools for directly visualizing chromatin conformations in their natural environment. This paper reports an imaging platform for visualizing chromatin organization across multiple scales in single cells with high genomic throughput. First, multiple imaging of hundreds of genomic loci via sequential hybridization is demonstrated, allowing for high-resolution conformational tracking across the entire chromosome. Next, a combined imaging approach for genome-scale chromatin tracking is developed, and simultaneous imaging of >1000 genomic loci and >1000 genes of newly generated transcripts, as well as iconic nuclear structures, is confirmed. Using this platform, chromatin domains, compartments, and transchromosomal interactions, and their relationship to single-cell transcription, were characterized. This high-throughput, multi-scale, and multimodal imaging technique has broad applications, providing a comprehensive view of chromatin organization within its natural structural and functional context.

[0336] The 3D organization of the genome regulates many fundamental cellular functions, from gene expression to DNA replication. Biochemical and imaging measurements have revealed complex chromatin structures across a wide range of scales. In particular, high-throughput chromosome conformation capture methods (such as Hi-C and other sequencing-based methods) have revealed chromatin structures, such as domains and compartments, from a whole-genome perspective. Notably, chromatin is segmented into genomic regions with enhanced self-interactions, called topologically related domains (TADs), which appear as blocky structures on Hi-C contact maps. These TADs range in size from hundreds of kilobases (kb) to several megabases (Mb), typically contain co-regulated genes, and have boundaries consistent with regulatory epigenetic elements. At a larger scale, chromatin is divided into two main compartments, called compartments A and B, which are enriched with active and inactive chromatin, respectively, as shown by alternating “lattice” patterns in Hi-C maps, consistent with previous imaging-based observations that gene-rich and gene-deficient segments of chromatin tend to be spatially separated. Recent imaging experiments have shown that A-compartment and B-compartment chromatin do indeed tend to spatially separate within a single cell. The physiological significance of A / B compartmentalization is related to its changes during development and between cell types.

[0337] Overall, high-throughput sequencing-based methods have greatly enriched our knowledge of 3D genome organization. Nevertheless, these powerful methods also have limitations. For example, they provide contact information for chromatin locus pairs but not the direct spatial location of these loci. Furthermore, most insights into chromatin organization are built on population-averaged contact maps of millions of cells. Despite continuous improvements in single-cell Hi-C methods, the capture efficiency of chromatin contacts in single cells and / or the cell throughput of these methods remain relatively low, making the study of 3D genome organization in single cells a challenging task. Moreover, while methods combining Hi-C with other measurements, such as those characterizing chromatin contacts in the context of interacting proteins, nuclear structures, or DNA modifications, multimodal measurements via sequencing remain challenging. Notably, methods that allow for genome-scale measurements of chromatin organization and transcriptional activity within the same cell are still lacking, although such methods are needed to further understand how chromatin organization regulates transcription and how transcription, in turn, influences chromatin organization.

[0338] On the other hand, imaging-based methods offer direct measurements of the spatial location of chromatin loci in individual cells with high detection efficiency. In particular, fluorescence in situ hybridization (FISH) allows for highly specific detection of chromatin loci in stationary cells, and recently, the CRISPR (Clustered Short Palindromic Repeats) system has significantly enhanced the ability to image specific chromatin loci in live cells. Chromatin imaging can also be combined with RNA and protein imaging to study interactions between chromatin organization and transcriptional activity or interacting protein factors. However, current imaging methods have limited throughput in genomic (sequence) space, traditionally allowing only a few different genomic loci to be studied at a time. Recently developed methods involve chromatin tracking through successive rounds of FISH imaging, each round targeting one or two genomic loci using monochrome or dual-color imaging. This approach allows imaging of dozens of different chromatin loci in a single cell and has been used to provide insights into chromatin structure and its relationship to transcription. However, due to the still limited number of genomic loci that can be imaged simultaneously in an individual cell, a high-resolution view of the entire chromosome in a single cell remains lacking, let alone a genome-scale view of chromatin organization within an individual cell.

[0339] This paper reports a multiscale, multiplexed FISH imaging platform that allows simultaneous imaging of hundreds to >1,000 different genomic loci in a single cell at varying resolutions and genome coverage. First, sequential imaging methods have made significant progress, allowing imaging of hundreds of genomic loci, and this approach has been used to provide high-resolution views of the entire chromosome, elucidating chromatin domains and compartmental structures, their relationships, and the relationship between chromatin organization and transcription in single cells. Next, a large-scale multiplexed FISH method based on combined markers and imaging has been developed, allowing imaging of more genomic loci with fewer hybridization rounds. Using this method, it has been shown that >1,000 genomic loci in an individual cell can be simultaneously imaged, along with the nascent RNA transcripts of >1,000 genes located at these loci and in iconic nuclear structures (including nucleoli and nucleoli), allowing chromatin organization to be placed within its native structural and functional context. This method is used to explore the relationships between transchromosomal interactions, transcriptional activity, and nuclear structure in single cells.

[0340] To allow for systematic observation of chromatin structure across multiple scales, an imaging platform using a custom-designed microscope and fluidic apparatus (see Example 19) was developed for direct visualization of chromatin in sequence space at extremely high throughput, up to the genome scale. This platform incorporates two complementary approaches ( Figure 17A First, imaging relatively small chromatin structures makes it difficult to distinguish the different loci contained within them in any single image. Previously reported sequential imaging strategies have been extended to allow tracking hundreds of chromatin loci in a single cell. In this approach, chromatin is imaged one locus at a time (or 2-3 loci at a time, using 2-3 color imaging passes) across multiple imaging rounds. Figure 17A (Left). This method is demonstrated by using it to track the conformation of an entire chromosome in a single cell at high resolution. Secondly, to image chromatin structures dispersed over regions significantly larger than the diffraction limit (e.g., structures throughout the cell nucleus), a more efficient combinatorial strategy was developed, in which numerous chromatin loci are imaged simultaneously in each round, and their distinct identities are determined based on the different round combinations in which they appear. Figure 17A (Right). The latter approach allows imaging of a large number of genomic loci in far fewer imaging rounds. This approach is used to provide a genome-scale view of chromatin organization in the context of single-cell transcriptional activity and important nuclear structures.

[0341] Figure 17A-17M High-resolution whole chromosomes were shown by sequential hybridization and characterization of chromatin domains in single cells. Figure 17AThis diagram illustrates a multi-scale chromatin tracing platform. Left: Schematic diagram of chromatin tracing across an entire chromosome via sequential hybridization and imaging. When the target chromatin structure is comparable to or smaller than the diffraction-limited resolution, each imaging round images a single chromatin locus in each color channel. After all rounds of imaging, 3D chromatin traces can be generated for each copy of the targeted chromosome. Right: Schematic diagram of genome-scale imaging via combined FISH. When the target loci are expected to be scattered in a space much larger than the diffraction-limited resolution, such as when loci are dispersed throughout the cell nucleus, multiple loci can be imaged and resolved in each round, and the identity of each locus can be derived from barcodes based on the combination of imaging rounds in which loci were detected. This method significantly reduces the number of rounds required to image the same number of loci compared to sequential imaging methods.

[0342] Example 7

[0343] High-resolution chromatin tracing of the entire chromosome. This section describes high-resolution whole-chromosome tracing performed using sequential imaging methods. Figure 17A ,Left; Figure 24A First, we focused on human chromosome 21 (Chr21). The non-repetitive portion of the chromosome (Chr21: 10.4–46.7 Mb) was divided into >600 contiguous segments (i.e., >600 genomic loci), each 50 kb in length. A library of primary oligonucleotide probes was designed, each probe containing a variable target sequence for hybridization with the chromosome and a unique readout sequence for each 50 kb locus. Figure 24A All primary probes binding to a specific 50-kb locus share the same readout sequence, therefore the readout sequence can be used to identify each locus by hybridization with complementary readout probes labeled with fluorescent dyes. Figure 24A However, due to the high cost of dye-labeled oligonucleotides, identifying these genomic loci using >600 different fluorescent readout probes would be extremely expensive. To overcome this challenge, a two-step labeling strategy was designed to detect different readout sequences using a common set of three dye-labeled oligonucleotide probes (called readout probes, one for each color channel), mediated by unlabeled adaptor probes that convert each locus-specific readout sequence into one of the three common readout sequences. Figure 24AUsing this strategy, >600 chromatin loci in Chr21 of human lung fibroblasts (IMR-90) were sequentially imaged using >200 rounds of hybridization with adaptors and readout probes, with three-color imaging in each round. To achieve stable imaging across such a large number of hybridization rounds, the imaging protocol was further optimized in the following ways: (i) to maintain sample integrity and primary probe binding stability, samples were periodically refixed with formaldehyde during imaging; (ii) to ensure complete removal of fluorescence signals after each imaging round and minimize the accumulation of residual signals over hundreds of labeling rounds, a combined chemical lysis and photobleaching method was used to remove fluorescence signals from readout probes, and unlabeled readout probes were added after each imaging round to prevent any unoccupied binding sites on the adaptor probes; (iii) to minimize interference and optimize experimental time, duration of each hybridization round, and flow rate of the fluid system.

[0344] After imaging, the centroid position of each chromatin locus was determined in 3D, and the conformation of each homologous copy of Chr21 in each cell was reconstructed. Figure 17B To estimate the stability of the sample and imaging instrument across multiple hybridization rounds, the same chromatin loci were re-imaged after different numbers of hybridization / imaging rounds, and the displacement between the original locus location and the corresponding re-imaged locus location was used as a measure of measurement accuracy. The median displacement between the original and re-imaged loci increased from approximately 70 nm when 11 hybridization rounds separated the two imaging instances to approximately 120 nm when the initial and re-imaged instances were separated by approximately 250 hybridization rounds. Figure 24B-24C Among them, loci showing larger displacements during re-imaging also exhibited lower fluorescence signal intensity. Figure 24D It is noteworthy that even after >250 rounds of hybridization, the median shift error was significantly smaller than the median distance between adjacent chromatin loci (~250 nm). Figure 24B Furthermore, the median pairwise distances between imaged loci showed high reproducibility across biological replicates. Figure 24E In these experiments, the detection efficiency of chromatin loci was >90% (i.e., >90% of the target chromatin loci were detected in each chromosome).

[0345] To obtain a population-average view of the chromatin conformation of Chr21, pairwise interactions between imaged loci were quantified by calculating their median spatial distance and the probability that the loci were close to each other in approximately 3,500 imaged cells. Figure 17C and Figure 24F-24I A high correlation was observed between the median pairwise distance from the imaging data and previously published Hi-C data across all length scales present in Chr21 (Pearson correlation 0.89). Figure 24F-24G For shorter genomic distances, the consistency is particularly high (Pearson correlation coefficient is 0.97); Figure 24H-24I To select a cutoff distance below which two loci are considered close, Pearson correlation coefficients were calculated between Hi-C data across a range of cutoff distances and neighbor frequencies (i.e., the fraction of instances where two loci are close) derived from imaging data. The Pearson correlation coefficients remained high over a wide range of cutoff distances, but peaked at 0.88 at a cutoff distance of approximately 400–500 nm. Figure 24J Therefore, 500 nm was chosen as the cutoff distance to generate a neighboring frequency map throughout the operation (for a more detailed explanation of the rationale behind the choice of cutoff distance, see Example 19).

[0346] Both median distance and nearest neighbor frequency plots show a blocky TAD structure. Figure 17C and Figure 24F and 24H The TAD boundaries identified from the distance and proximity frequency maps of the imaging data are highly similar to those determined from the overall Hi-C data. Figure 24K Furthermore, it was confirmed that the locus localization error (~100 nm) in the chromatin traces had little impact on the identification and accuracy of domain boundaries. Figure 24K ).

[0347] Figure 17B This image shows a 3D structural rendering and spatial distance matrix of two copies of Chr21 in a single IMR90 cell imaged using a sequential hybridization method. Left: Two copies of Chr21 in a single cell overlaid on a DAPI image of the cell nucleus. Scale bar: 5 μm. Top right: 3D rendering of all detected chromatin loci (spheres) in the two Chr21 copies according to their genomic coordinates along the chromosome (genomic locations are shown on the right). Flexible lines connect adjacent chromatin loci. Scale bar: 1 μm. Bottom right: Paired spatial distance matrices corresponding to the chromosome copies shown above (no imaging of genomic regions without a suitable reference genome or containing highly repetitive sequences).

[0348] Figure 17C This shows the overall neighbor frequency matrix of Chr21 and the preferential localization of single-cell domain boundaries at the CTCF / RAD21 binding site. Top: Neighbor frequency matrix of Chr21 derived from imaging data. Each matrix element is defined as the frequency of a cutoff distance of less than 500 nm between a pair of loci. Middle: Enlarged version of the neighbor frequency matrix for the 10 Mb portion of Chr21. Bottom: Probability of single-cell domain boundary formation on each 50-kb genomic segment imaged. Triangles indicate CTCF and RAD21 ChIP-seq peaks.

[0349] Figure 24A-24N High-resolution whole chromosome traced by sequential hybridization is shown, along with overall statistics on Chr21 structural features compared to Hi-C. Figure 24A The labeling and imaging scheme for sequential hybridization using adaptor probes is shown. First, the sample is hybridized with primary probes, each containing a target sequence and a readout sequence that allow specific binding to target genomic loci. Each locus is labeled with a total of 350–500 primary probes, but only one is shown. Each target genomic locus is assigned a unique readout sequence (shown in various colors) that is common to all primary probes binding to that locus. The readout sequences are then detected using successive rounds of hybridization. During each round of hybridization, the readout sequence corresponding to the target locus (one for each of the three color channels Alexa750, Alexa647, and Cy3) is labeled with an oligonucleotide adaptor probe, each probe consisting of two parts: a segment complementary to the locus-specific readout sequence and a segment containing a color channel-specific common readout sequence. Each color channel contains a unique common readout sequence, which is shared by all adaptors visualized in the same color channel. The common readout sequence is then hybridized with a dye-conjugated complementary readout probe in the corresponding color channel. This procedure allows imaging of three genomic loci in three color channels during each round of hybridization. After each round of imaging, the fluorescent dye linked to the readout probe via disulfide bonds is cleaved from the shared readout probe by TCEP, and any unoccupied readout sequences on the adaptor are blocked by the unlabeled shared readout probe to prevent crosstalk between hybridization rounds. This process iterates for hundreds of rounds until all readout sequences have been detected, thus completing the detection of all targeted genomic loci.

[0350] Figure 24B The displacement of the locus during a single experiment is shown. At the beginning and end of the experiment (interval > 250 rounds of hybridization), consecutive 50 kb segments within a 900 kb region of Chr21 (chr21: 32.45–33.35 Mb) were imaged. The displacement distribution between the re-imaged points and their original image counterparts is shown. For comparison, the distance distribution between adjacent 50 kb segments within the same 900 kb region measured in the original imaging rounds is shown.

[0351] Figure 24C Box plots showing chromatin gene locus shifts between the original and re-imaging rounds, separated by different numbers of hybridization rounds. Median (center line), 25th–75th percentiles (boxes), and 10th–90th percentiles (whiskers) are shown.

[0352] Figure 24DBox plots showing fluorescence signals at chromatin loci with low (<500 nm) and high (>500 nm) shift errors between original and re-imaging experiments with an interval >250 rounds of hybridization are displayed. Median (center line), 25th–75th percentiles (boxes), and 10th–90th percentiles (whiskers) are shown.

[0353] Figure 24E The comparison of median spatial distances between chromatin loci is shown between two replicate experiments. For two biological replication experiments of Chr21, the median spatial distances between imaged chromatin locus pairs were calculated and plotted against each other. The Pearson correlation coefficient between the data measured in the two replicates is [value missing]. =0.98.

[0354] Figure 24F The comparison shows the median spatial distance matrix (left), the neighbor frequency matrix (middle), and the overall Hi-C contact matrix (right) derived from imaging for Chr21. For imaging data, two chromatin loci are considered close when the spatial distance between them is less than a cutoff distance of 500 nm. The Hi-C contact matrix is ​​binned in 50 kb blocks and centered on the target region.

[0355] Figure 24G The plot shows a re-logarithmic scatter plot of the number of contacts derived from overall Hi-C and the median pairwise distance derived from imaging of individual pairs of chromatin loci. The line represents a linear regression of the data (slope = -4.43). The Pearson correlation coefficient between the imaging and Hi-C data is [value missing]. =0.89.

[0356] Figure 24H and Figure 24F Same, but for the 3-Mb region in Chr21 (chr21: 30.30-33.38Mb). TAD boundaries are marked with lines.

[0357] Figure 24I and Figure 24H Same, but for the 3-Mb region shown in (H). Slope = -4.51. =0.97.

[0358] Figure 24J The Pearson correlation between the Hi-C contact map and imaging-derived proximity frequency maps generated using different cutoff distances is shown. To generate the proximity frequency map, a cutoff distance is chosen, and two loci with a distance less than this cutoff value are considered close. The proximity frequency between a pair of loci is then calculated by dividing the number of times the measured distance between the loci is less than the cutoff distance by the total number of measured distances between the two loci.

[0359] Figure 24K The normalized insulation score as a function of genomic location on Chr21 is shown, calculated as follows: 1) median pairwise distance from imaging (top), 2) neighbor frequency from imaging (middle), and 3) Hi-C contact reading (bottom). To calculate the insulation score, a fixed-length (250 kb) genomic segment upstream and an identical-length segment downstream of the location of interest were first selected. The normalized insulation score was then defined as the difference between the pairwise distances between median segments and the pairwise distances within median segments, normalized by the sum of these two median distances. The TAD boundary was defined as the local maximum of the normalized insulation score along the chromosome, identified by the standard peak call algorithm (see Example 19). The vertical dashed line is the global TAD boundary called from the Hi-C data. The median distance (black line, top) and neighbor frequency (black line, middle) after perturbing the locus location with a 3D Gaussian noise term with a standard deviation of 100 nm are also shown in the top and bottom figures, consistent with the estimated localization measurement error.

[0360] Figure 24L The image shows chromatin domains in the 10-Mb region of Chr21 (chr21: 28.2–38.1 Mb) in two example single cells. The image also shows the pairwise distances of two individual copies of Chr21 in a single cell (top, middle) and the population median pairwise distances from all imaged cells (bottom).

[0361] Example 8

[0362] Chromatin domains within a single chromosome. At the single-cell level, chromosomes are observed to be divided into domains that exhibit blocky characteristics in the single-cell spatial distance matrix. Figure 24L These domain and locus distances demonstrate high variability between cells. Figure 24L-24M This is consistent with the significant intercellular variation in chromatin contact observed in single-cell Hi-C data. Similar domain structures in single cells have been previously observed when imaging small (~2Mb) regions of chromosomes at similar resolution. However, in these previously measured small regions, a significant proportion of cells did not show clear single-cell domain boundaries, and it remains uncertain whether domains were not formed within those cells or whether the entire imaging region was within a single domain. Furthermore, due to the small size of these previously imaged regions, many domains were artificially truncated at the ends of the imaged genomic regions, making it impossible to accurately characterize certain fundamental domain properties, such as their physical and genomic size. The high genome throughput of this study provides a whole-chromosome view of these single-cell domain structures, revealing their significant presence throughout the chromosome in virtually all imaged cells, thus allowing for a more systematic characterization of their properties.

[0363] First, the genomic locations of these single-cell domain boundaries were identified, and the probability of boundary formation at each 50-kb genomic locus was quantified. Although a non-zero probability of boundary formation was observed at all imaged genomic loci, domain boundaries preferentially located near the binding sites of CTCF and adhesion proteins. Figure 17C-17D ).

[0364] In addition to intercellular variations at domain boundary locations, significant heterogeneity was observed in other features of these single-cell domains, ranging from the physical size of the domains to the degree of insulation or interaction between them. Figure 17E-17H Specifically, single-cell domains were observed in their genome size ( Figure 17I ) and physical dimensions measured by radius of gyration ( Figure 17E and 17J Both the genome size distribution and the physical size distribution of these domains are variable. The distribution of genome size and the physical size distribution of these domains are insensitive to an estimated locus localization error of approximately 100 nm. Figure 17I-17J It is noteworthy that domains defined by the same genomic region or with the same genomic size exhibit significant variations in their physical size between cells. Figure 17E and 24N Interestingly, domains defined by interacting CTCF / adhesion protein binding sites tend to be smaller in physical size than domains not defined by such genomic loci. Figure 17K Furthermore, the degree of physical separation between adjacent structural domains also varies considerably. Figure 17F and 17L Some of these adjacent structural domains are completely separated and connected only by joint areas, while others show partial overlap and less distinct boundaries. Figure 17F Furthermore, even domains completely separated from their adjacent domains may partially overlap spatially with non-adjacent domains separated by small or large genomic distances. Figure 17G Finally, it was observed that the two chromatin loci at the ends of these single-cell domains also exhibited variable distances from each other, and did not show a tendency to be closer to each other compared to chromatin loci separated by similar genomic distances within the domain, regardless of whether the domain was defined by CTCF / adhesion protein binding sites. Figure 17H and 17M ).

[0365] Figure 17D The average probability (grey) of domain boundary formation in a single cell at a genomic location centered on the CTCF / Rad21 binding site or the overall TAD boundary.

[0366] Figure 17EThese are examples of two single-cell chromatin domains with the same genomic coordinates that occupy a larger (top) or smaller (bottom) volume in physical space. Left: 3D rendering of the chromatin domains, where green spheres represent imaged genomic loci within the domain, and flexible linkers connect adjacent loci in the genomic sequence. Gray dots represent imaged loci in the rest of the chromosome. Scale bar: 1 micrometer. Right: The left side shows the pairwise distance matrix of the chromatin domains (marked with lines), with flanking regions.

[0367] Figure 17F Examples of two pairs of chromatin domains with high (top) and low (bottom) insulation scores are shown. Left image: 3D rendering of the chromatin domains, as shown. Figure 17E As shown. Scale bar: 250nm. Right figure: The left side shows the pairwise distance matrix of chromatin structural domains, where the rendered structural domains are marked with their respective colors.

[0368] Figure 17G Two instances of long-range contacts between chromatin domains with partially overlapping volumes are shown. Left: 3D rendering of chromatin domains, as shown in (E). Shading represents different domains. Scale bar: 250 nm. Right: Pairwise distance matrix of chromatin domains, where rendered domains are labeled with their corresponding colors. Gray space represents the intervals within a 22.85 Mb genome distance.

[0369] Figure 17H Examples of chromatin domains flanking the CTCF binding site are shown, illustrating the small (top) and large (bottom) distances between CTCF sites. Left: 3D rendering of the chromatin domains, as shown in (E), but with the CTCF locus located at the end of the domain. Scale bar: 250 nm. Right: Pairwise distance matrix of chromatin domains, correspondingly labeling the domains and the boundary CTCF.

[0370] Figure 17I The distribution of genome size for measured chromatin domains in Chr21 within single cells is shown. The black line represents the distribution of genome size for chromatin domains in Chr21 within single cells, derived from simulated data that accounts for a 100 nm localization error. In this simulation, the locations of the imaged loci were affected by 3D Gaussian noise with a standard deviation of 100 nm, similar to our measurement error.

[0371] Figure 17J The distribution of the measured physical dimensions (e.g., defined by the radius of gyration) of chromatin domains in Chr21 within a single cell is shown. The black lines represent the distribution of the physical dimensions of chromatin domains in Chr21 within a single cell, derived from simulated data that takes into account a 100 nm localization error. Figure 17I As shown.

[0372] Figure 17K The median radius of gyration is shown as a function of genome size as a function of chromatin domains, where boundary loci contain interacting CTCF / Rad21 sites and boundary loci do not contain CTCF / Rad21 sites. Error bars represent 95% confidence intervals derived through resampling.

[0373] Figure 17L The distribution of insulation scores between adjacent structural domains is shown, with the domain boundaries appearing at CTCF / Rad21 binding sites and non-CTCF / Rad21 binding sites.

[0374] Figure 17M The median normalized end-to-end distance of domains is shown as a function of genome size, where boundary loci contain interacting CTCF / Rad21 sites and boundary loci do not contain CTCF / Rad21 sites. The normalized end-to-end distance is defined as the end-to-end distance of a domain divided by the median distance between pairs of similar loci separated by the same genome distance but located within a single domain. Error bars represent 95% confidence intervals derived through resampling.

[0375] Figure 24M The standard deviation matrix of spatial distances between Chr21 loci is shown. For each pair of regions, the standard deviation of the distances between the corresponding locus pairs across all single chromosome copies is shown.

[0376] Figure 24N Box plots show the physical dimensions of chromatin domains of different genome sizes in Chr21, as measured by radius of gyration. For each genome size, the median (center line), 25th–75th percentiles (boxes), and 10th–90th percentiles (whiskers) are shown.

[0377] Example 9

[0378] Chromatin compartments within a single chromosome. Next, a high-resolution view of the entire chromosome is used to examine how chromatin loci within compartments A and B are arranged in a single cell. First, using the previously described algorithm, principal component analysis (PCA) of the Pearson correlation matrix derived from the Chr21 proximity frequency map from the imaging data is used to determine the overall A / B compartment boundaries. Figure 18A ; Figure 25A The compartment boundaries obtained from the imaging data are highly similar to those determined from previously published overall Hi-C data. Figure 25A Below, the compartment boundaries obtained from the global proximity frequency map are used to assign A / B identities to individual loci in individual cells.

[0379] The >10-fold increase in resolution compared to previous studies allows for detailed viewing of the organization of A-compartment and B-compartment loci within individual chromosomes. High variability in the arrangement of A and B loci was observed between individual chromosome copies from cell to cell. Figure 18B While in some chromosomes the A and B loci are separated into largely non-overlapping spatial regions, other chromosomes show significant spatial overlap between the A and B loci. Interestingly, the A compartmental locus on the same chromosome is sometimes divided into multiple "microcompartments" ( Figure 18B ).

[0380] To quantify the spatial separation of A and B loci in an individual chromosome, a method based on local density was designed, and for each imaged locus, the local densities of other A and B loci were calculated. Figure 25B As expected, loci A tend to be surrounded by loci A on average, and the same is true for loci B. Figure 25C Based on the purity of loci observed in the spatial volume containing the majority of A or B loci, the A / B segregation score for each individual chromosome was further defined. Figure 18C Complete physical separation of loci A and B was expected to result in a segregation score of 1, while complete mixing of loci A and B resulted in a segregation score of 0.5 (see Example 19). It was observed that for the vast majority of Chr21 copies in cells, the segregation score was significantly higher than that of the randomized control centered at 0.5 (obtained by randomly moving the compartment boundaries along the genome axis while maintaining constant compartment size). Figure 18C This indicates a tendency for loci A and B to spatially separate in a single cell. It is also noteworthy that the spatial separation of loci A and B is often incomplete. Figure 18C This may reflect an incomplete spatial separation of active and inactive chromatin, but it could also be partly due to intercellular variability in epigenetic modifications, which could make the overall A / B compartment identity not entirely representative of the depiction of active / inactive chromatin in individual cells. Notably, the degree of A / B separation was found to be cell cycle dependent: A / B separation was stronger in G2 / S phase cells compared to G1 phase cells. Figure 25D This is consistent with previous findings that A / B compartments gradually develop during the cell cycle.

[0381] Notably, Chr21 is one of the smallest chromosomes, measuring only about 48 Mb, and is divided into only a few contiguous A and B regions. To extend the findings to larger chromosomes and investigate their generalization, one of the largest chromosomes, chromosome 2 (Chr2), was imaged, which exhibits numerous transitions (approximately 50 transitions) between A and B compartments along its genome sequence. Specifically, Chr2 was traced by marking and imaging 50-kb segments along its genome sequence at 250-kb intervals. Based on the imaging data ( Figure 18D ; Figure 25E Using the same method described above, compartments A and B in chromosome p and q arms were retrieved, and quantitative consistency with compartments A and B determined from previously published overall Hi-C data was observed. Figure 25F-25G At the single-chromosome level, various spatial arrangements of loci A and B were observed again, ranging from almost complete spatial separation to significant spatial overlap between loci A and B. Figure 18E Interestingly, some chromosomes exhibit a "sandwich" structure, with the A locus situated between two layers of B loci, likely due to the preferential association of the B loci with the nuclear lamina near the top and bottom of the nucleus. Quantitatively, the distribution of A / B segregation scores in individual copies of Chr2 again showed the overall trend of segregation of A and B loci in individual chromosomes compared to a randomized control. Figure 18F The spatial separation in Chr2 appears to be smaller than that in Chr21. Figure 18C and 18F ).

[0382] Figure 18A-18I It shows the compartmental structure in a single chromosome and the relationship between transcriptional activity and local chromatin content. Figure 18A The Pearson correlation matrix of genome distance normalized to neighbor frequencies for Chr21, derived from our imaging data, is shown. Two loci are considered close if the distance between them is less than a cutoff distance of 500 nm. The two bars at the bottom show the A / B calls (shown as compartment A and compartment B) derived from the neighbor frequency matrix and the G banding of each genomic locus in the chromosome.

[0383] Figure 18B This image shows a 3D rendering of an individual copy of Chr21 in a single cell, with loci A and B displayed as spheres. Flexible lines connect adjacent loci in the genomic sequence. The bars at the bottom show the A / B calls for each genomic locus in the chromosome, derived from the neighbor frequency matrix. Scale bar: 1 micrometer.

[0384] Figure 18CThe distribution of A / B segregation scores for individual copies of Chr21 is shown. To calculate the A / B segregation score, an A (or B) dense volume is defined for each chromosome by thresholding the local A (or B) density so that 2 / 3 of the A (or B) loci are contained within the volume (note that for chromosomes showing spatial overlap between A and B loci, the A and B dense volumes may overlap). The purity of loci within a chromosome's A (or B) dense volume is defined as the fraction of all loci within the volume that are A (or B) loci, and the A / B segregation score for a chromosome copy is defined as the average purity of the A and B volumes. Histograms represent the distribution of A / B segregation scores for randomized controls, where the boundaries between consecutive A and B regions are randomly shifted along the genomic sequence while maintaining the number and size of A and B regions. n = ~7,500 chromosomes.

[0385] Figure 18D The Pearson correlation matrix of genome distance-normalized neighbor frequencies for the p and q arms of Chr2, derived from our imaging data, is shown, along with the corresponding A / B calls and G banding, as follows: Figure 18A As shown.

[0386] Figure 18E The image shows a 3D rendering of an individual copy of Chr2, such as... Figure 18B As shown. Scale: 1 micrometer.

[0387] Figure 18F The distribution of Chr2 A / B segregation scores in single cells is shown, as follows: Figure 18C As shown. n=~3, 100 chromosomes.

[0388] Figure 25A-25G Showing the overall A / B compartment analysis of Chr21 and Chr2. Figure 25A The compartment calls based on Chr21 principal component analysis are shown. The first principal component (PC1) calculated from the Pearson correlation matrix of genome distance normalized neighbor frequencies derived from imaging (top panel) and overall Hi-C (bottom panel) experiments is shown, with PC1 values ​​> 0 corresponding to compartment A and PC1 values ​​< 0 corresponding to compartment B.

[0389] Figure 25B This image shows 3D renderings of the A-compartment loci (A locus) and B-compartment loci (B locus) in a single copy of Chr21, along with the A / B density ratio. Left: A and B loci of a representative copy of Chr21. A and B compartment calls from the overall proximity frequency map derived from the imaging are shown in the bottom bar. Right: The same chromosome, but each locus is colored by its local A / B density ratio.

[0390] Figure 25CThe average A and B density scores for each imaging locus in Chr21 are shown across all imaging cells. The figure below represents the A or B compartment calls for each locus derived from the neighbor frequency map derived from the imaging.

[0391] Figure 25D Histograms showing the distribution of A / B segregation scores for individual copies of Chr21 in cells during the G1 and G2 / S phases of the cell cycle.

[0392] Figure 25E The adjacent frequency matrix is ​​shown, derived from the Chr2 imaging (left) and the overall Hi-C contact matrix (right). The Hi-C contact matrix is ​​binned in 50kb increments, but only the contact from the imaging segment is shown (selected at 250kb intervals).

[0393] Figure 25F Showing with Figure 25A The same PC analysis, but for the p-arm (top) and q-arm (bottom) of Chr2.

[0394] Figure 25G Showing with Figure 25C Same average A and B density analysis, but for Chr2.

[0395] Example 10

[0396] Relationship between transcription and local A / B chromatin content. To investigate whether chromatin compartmentalization is associated with active transcription in a single chromosome, oligonucleotide probes targeting the first intron of 86 genes contained in Chr21 were designed and sequentially hybridized to image the nascent RNA transcripts of these genes, followed by chromatin tracing. Furthermore, to more accurately detect gene spatial locations, 5kb genomic loci centered at the transcription start site (TSS) of each target gene were imaged. To prevent RNA probe binding to genomic DNA and vice versa, RNA probe hybridization was performed without heating and denaturing the double-stranded genome, followed by RNA molecule digestion using ribonuclease (this step is also included when imaging chromatin separately), and then chromatin tracing. This strategy confirmed that crosstalk between RNA and DNA signals was negligible. Figure 26A-26J ).

[0397] Typically, a subset of the genes being imaged shows transcriptional activity in any individual cell. Figure 18GThe study investigated how transcriptional activity is associated with the local chromatin environment. To characterize local A / B chromatin content, the local densities near the A and B loci for each gene were calculated, and their ratio (hereinafter referred to as the A / B density ratio) was used as an indicator of local enrichment of active chromatin. It was found that for approximately 80% of the genes studied, the local A / B density ratio at the TSS was higher when the gene was actively transcribed than when the gene was not activated. Figure 18H As a corollary, in cells where the TSS of a gene has a high local A / B density ratio, the gene activation rate is also often higher. Figure 18I These results suggest that cells with higher enrichment of the same gene at locus A and / or de-enrichment at locus B in the vicinity of the gene tend to have higher transcriptional activity. This increased transcriptional activity may be due to local enrichment of transcriptional mechanisms and / or de-enrichment of silencing factors. Alternatively, given that transcriptional mechanisms and cofactors can form condensates, it is possible that active transcriptional chromatin associated with transcriptional mechanisms has a stronger tendency to interact with other active chromatin. Note that these two possible mechanisms are not mutually exclusive but can work synergistically to reinforce each other.

[0398] Figure 18G A 3D rendering of a single copy of Chr21 is shown, displayed alongside the measured transcriptional bursts of the gene. Spheres represent all detected nascent RNA bursts within this chromosome. Scale bar: 500 nm.

[0399] Figure 18H The changes in the A / B density ratio at the transcription start site (TSS) of genes imaged between active and non-active states are shown (measured as log difference). For each gene, the median A / B density ratio was calculated at its TSS in the chromosome where the gene was being activated and in the chromosome where the gene was not activated. The log differences of these values ​​for the 84 genes imaged on Chr21 were sorted according to the magnitude of the change in the median A / B density ratio. 79% of the imaged genes showed an increase in the A / B density ratio at active activation compared to non-activation.

[0400] Figure 18I This study shows the change in excitation rate (measured as logarithmic difference) of imaging genes as the local environment of the TSS changes from a low (bottom quartile) to a high (top quartile) A / B density ratio. The logarithmic differences in excitation rates of 84 genes imaged on Chr21 are sorted numerically by excitation rate. 79% of the imaging genes showed a higher excitation rate at the top quartile compared to the bottom quartile when their TSS was at the top quartile.

[0401] Figure 26A-26J The measurement of crosstalk between RNA and DNA FISH probes is shown. Figure 26AThe image shows the fluorescence signal of a sample cell with its nucleus labeled with DAPI (top) and a FISH probe for the nascent RNA of the target gene (BRWD1) (bottom).

[0402] Figure 26B and Figure 26A They are the same, but their genes are different (SCAF4). Figure 26A and 26B The staining followed the RNA FISH protocol described in the “Cell Culture Preparation and Primary / Encoding Probe Hybridization” section of Example 19.

[0403] Figure 26C and 26D respectively with Figure 26A and 26B Similarly, except that the RNA FISH protocol was modified to include an additional ribonuclease treatment step to remove cellular RNA before adding the FISH probe. Figure 26C and Figure 26D The cells in Figure 26A and 26B Imaging under similar illumination conditions, and their fluorescence signals are similar to... Figure 26A and 26B The same contrast is displayed.

[0404] Figure 26E The number of spots per cell with a signal-to-noise ratio >3 was shown for untreated and ribonuclease-treated cells across five measured genes.

[0405] Figure 26F The image shows the fluorescence signal of an example cell with its nucleus labeled with DAPI (top) and a probe targeting a genomic locus (chr21: 15.2Mb-15.25Mb) (bottom).

[0406] Figure 26G and Figure 26F Same, but used for different loci (chr21: 14.95Mb-15Mb). Figure 26F and Figure 26G The staining followed the protocol described in the DNA FISH protocol described in the “Cell Culture Preparation and Primary / Encoding Probe Hybridization” section of Example 19.

[0407] Figure 26H and 26I respectively with Figure 26F and 26G The same except that the DNA FISH protocol was modified to omit the heat denaturation step and thus remove accessible genomic DNA sites. Figure 26H and 26I The cells in Figure 26F and 26GImaging under similar illumination conditions, and their fluorescence signals are similar to... Figure 26H and 26I The same contrast is displayed.

[0408] Figure 26J The number of spots per cell with a signal-to-noise ratio >3 is shown for cells treated with the thermal denaturation step and cells that omited this step.

[0409] Example 11

[0410] The relationships between chromatin domains and compartments within a single chromosome were investigated. Next, the interactions between chromatin domains in a single cell and how these interactions are associated with compartmental identity were examined. Due to the large size of Chr2 and the large number of compartmentalizations within it, analysis of Chr2 is expected to provide more insights, thus this chromosome was the focus of this study.

[0411] Although most single-cell domains in Ch2 are “pure” domains consisting entirely of A or B loci, a large proportion of single-cell domains cross the global A / B boundary and contain both A and B loci. Figures 19A-19C The presence of these “hybrid” domains suggests that domain formation in individual cells may not be closely related to chromatin properties that determine compartmental identity, but intercellular variability in epigenetic modifications may also lead to shifts in the active / inactive chromatin boundary in some single cells.

[0412] Next, we examine how domains interact with each other, focusing on how these interactions depend on the domain's A and B components and the genomic distance between domains. Domains make contact in both short and long genome separations, and this contact is represented by off-diagonal boxes in the spatial distance map of individual chromosomes. This contact pattern varies significantly between cells. Figure 19D Despite this heterogeneity, the interactions between domains appear to be regulated by the A / B composition of their underlying chromatin. Figure 19E The average contact frequency between domains primarily containing B loci is higher than that between domains primarily containing A loci, which in turn is higher than that between domains dominated by chromatin with different A / B identities. This average picture is consistent with the hierarchical structure of A and B chromatin interaction strength recently proposed based on chromatin structure modeling of A / B compartmentalization measured by Hi-C and global arrangement of A and B loci measured by imaging.

[0413] Further examination of domain contact frequencies as a function of genomic distance revealed a more complex situation. For simplicity, we focused on “pure” A and “pure” B domains at loci containing a single compartmental identity. As expected, for all constituent domain pairs, contact frequencies decreased with genomic distance ( Figure 19F However, at shorter genomic distances (up to ~75 Mb for Chr2), the contact frequency between B domain pairs (BB) is higher than that between A domain pairs (AA), while AA domain contacts outperform BB domain contacts at larger genomic separations. Figure 19F This result is consistent with the genomic distance dependence of AA and BB chromatin interactions reported in recent overall Hi-C studies and provides further insights into how preferential interactions between single-cell domains can lead to these overall trends. Notably, at relatively large genomic distances, the contact probability of the BB domain decays to a level similar to that between the A and B domains (AB), while the contact probability of the AA domain remains higher than that of the AB domain even at larger genomic separations. Figure 19F This results in AA domain interactions being significantly dominant at larger genomic distances. Figure 19G Furthermore, the contact domain pairs also exhibit varying degrees of spatial overlap, with some pairs showing contact between opposing surfaces. Figure 19F (Illustration), while other pairs show strong mixing ( Figure 19H (Illustration). Interestingly, compared to the AA domain pair, the BB domain pair shows a significantly stronger tendency to form these hybrid spheres. Figure 19H ).

[0414] Overall, these results indicate that preferential AA and BB domain interactions lead to spatial separation of chromatin compartments, and that the nature of these interactions is distinct. These differences in the nature of the interactions may stem from different molecular factors involved in AA and BB associations. For example, heterochromatin factors (e.g., HP1) are thought to be involved in BB interactions, while transcriptional activators or coactivators such as BRD4 and mediators may be involved in active chromatin interactions. Whether these different molecular factors are responsible for the observed differences in the genomic distance dependence of AA and BB domain interactions and their tendency to mix warrants further investigation.

[0415] Figures 19A-19H The study showed the dependence of interdomain interactions on its A / B composition and genomic distance. Figure 19AThe left image is a 3D rendering of a "mixed" chromatin domain containing loci A and B, flanked by "pure" domains containing only loci B in a single-cell Chr2 copy. Scale bar: 500 nm. The right image shows a pairwise distance matrix of the same region as shown on the left. Bars at the bottom and left of the matrix indicate A and B calls to the loci, with outlines highlighting the boundaries of the chromatin domains. A / B calls are determined from the overall proximity frequency map of Chr2.

[0416] Figure 19B and Figure 19A The same applies, but for two pure domains, one composed entirely of loci A and the other entirely of loci B, rather than mixed domains.

[0417] Figure 19C It is the distribution of the fraction of loci that serve as A loci in the single-cell chromatin domain of Chr2.

[0418] Figure 19D This is a single-cell spatial distance matrix for two example copies of Chr2. The first and third figures show the matrices for both complete chromosomes, while the second and last figures show magnified matrices of the regions highlighted in yellow in the first and third figures, respectively. Side bars show A / B compartment calls from the overall proximity frequency plot.

[0419] Figure 19E The diagram shows the domain contact probabilities of different A / B domain compositions in Chr2. The X and Y axes represent the fraction of loci within a domain that are A loci (0% corresponds to a pure B domain, and 100% corresponds to a pure A domain). Domains are defined as in contact if their insulation score is <2. See Example 19 for the calculation of the insulation score.

[0420] Figure 19F The domain contact probabilities in Chr2 between two pure A domains (AA), between two pure B domains (BB), and between a pure A and a pure B domain (AB) are shown, plotted as a function of the genomic distance between the two interacting domains. The inset includes a 3D rendering of example domain pairs showing long-range interactions with an insulation score of 2. Scale bar: 500 nm.

[0421] Figure 19G and Figure 19E The same, but for domain pairs with a genomic distance greater than 80 Mb.

[0422] Figure 19H and Figure 19FSimilar, but limited to domain pairs with high mixing (defined by a low insulation score <1). The illustration contains a 3D rendering of an example domain pair showing long-range interactions with high mixing (insulation score = 1). Scale bar: 500 nm.

[0423] Example 12

[0424] Genome-scale chromatin imaging. The sequential imaging method described above allows for high-resolution views of chromatin within individual chromosomes. This direct sequential imaging method is well-suited for imaging chromatin structures at or below the diffraction-limited resolution. However, in this method, the number of imaged genomic loci increases only linearly with the number of imaging rounds. For genome-scale chromatin imaging, since many genomic loci can be resolved and located simultaneously in the cell nucleus, a more efficient, non-linear scaling of the number of imaged loci relative to the number of imaging rounds is possible.

[0425] To achieve this goal, a combinatorial FISH method was designed, inspired by the multiple error-resistant FISH method previously developed for transcriptome imaging, but with significant modifications specifically designed for chromatin imaging, taking into account the aggregation properties of chromatin (i.e., adjacent loci in the genome sequence are spatially close) and the regional organization of chromosomes (i.e., different chromosomes tend to occupy different spatial regions). To allow for combinatorial imaging, each genomic locus is assigned a unique 100-bit binary barcode with a Hamming weight of 2, meaning each barcode contains two "1" bits and 98 "0" bits. Figure 20A The bit values ​​("1" or "0") in these barcodes determine the presence or absence of a signal at each locus in successive rounds of imaging. To avoid simultaneously imaging spatially close chromatin regions at the same bit, a subset of these 100-bit Hamming-weighted 2-bit barcodes was further selected to encode the targeted genomic loci, and the barcode allocation was optimized such that loci with a "1" bit at the same barcode position were spaced apart to the greatest extent possible in the genomic space (see Example 19). This strategy allows for minimizing detection errors caused by overlapping signals from nearby chromatin loci. Furthermore, since the vast majority of possible 100-bit binary codes are invalid (i.e., not assigned to any target locus), this design allows for the identification and discarding of detection errors, further improving measurement accuracy.

[0426] A highly diverse library of coding probes physically imprints barcodes onto targeted genomic loci. Each coding probe contains a target region for binding to one of the target loci and a readout sequence selected from 100 pre-designed readout sequences. Figure 20AEach readout sequence corresponds to one of 100 bits, and the coding probe set for each genomic locus (approximately 400 probes per locus) contains only two distinct readout sequences, corresponding to the two bits that read "1" in the barcode assigned to that locus. After binding to the coding probe, the barcode imprinted on the chromatin locus is detected by sequential hybridization of fluorescently labeled readout probes, each probe being complementary to one of the 100 readout sequences. Figure 20A In some cases, an adaptor probe strategy for high-resolution whole-chromosome tracking was also used here. Two different adaptor / readout probes were introduced in each hybridization round and imaged in dual-color channels, allowing for the reading of 2 sites in each hybridization round. This allowed for the imaging and identification of approximately 1000 genomic loci in just 50 hybridization rounds. Figures 20A-20C This represents approximately 10 times fewer hybridization rounds and therefore 10 times shorter experimental time compared to sequential imaging of the same number of loci with the same number of color channels. Since each chromosome in a diploid cell has two homologs, a clustering algorithm was used to further assign homologous identities to the imaged loci, taking advantage of the tendency of chromosomes to occupy different regions in each cell nucleus.

[0427] In this work, 1,041 genomic loci were selected for imaging, each approximately 30 kb in size, uniformly covering the 22 autosomes and X chromosome in IMR-90 cells. Another requirement was that each chromosome contained at least 30 target loci; therefore, the number of loci imaged per chromosome homolog ranged from 30 to 80, depending on chromosome length. These 1,041 genomic loci were imaged from approximately 5,400 individual cells across five biological replications, with a detection efficiency of approximately 80% per locus, resulting in approximately 1,700 chromatin loci detected per cell. Figure 20D-20E At the end of the combined imaging process, a small subset of genomic loci is re-imaged using sequential imaging, one locus at a time. The displacement between the locus location determined by combined imaging and the re-imaged location determined by sequential imaging is only ~50 nm. Figure 27A This demonstrates the high decoding accuracy of the combined imaging method and the minimal sample degradation / deformation during the imaging process.

[0428] To obtain a population-average view of chromatin organization, the spatial distance between each pair of imaged chromatin loci in each cell was calculated, and then the median distance and nearest neighbor frequency between each pair of loci in all imaged cells were determined. Figure 20F ; Figure 27B The proximity frequencies between chromatin loci pairs within the same chromosome, determined from imaging data, showed a high correlation with contact frequencies detected by overall Hi-C assays, with a Pearson correlation coefficient of 0.89. Figure 27C Furthermore, the imaging results showed high reproducibility among independent biological replicates. Figure 27D ).

[0429] By exploring the chromatin organization of individual cells, it was noted that chromosomes tend to occupy different regions within each cell ( Figure 20F-20G However, it also shows a large overlap between them. Figure 20G-20H These results are consistent with and extend the observations of earlier imaging studies. Given that this observation suggests a high degree of transchromosomal interactions, further analysis will focus on exploring them.

[0430] Figures 20A-20H Genome-scale chromatin was shown through large-scale multiplex FISH imaging. Figure 20A The imaging scheme is shown. Targeted genomic loci are assigned error-resistant barcodes, such as a subset of a 100-bit binary barcode with a Hamming weight of 2 (i.e., two read "1"s in the 100 bits). The barcodes are imprinted on the genomic loci using coding oligonucleotide probes that recognize the loci and associate two distinct readout sequences with each locus, corresponding to the two read "1" bits in the barcode assigned to that locus. Each bit is uniquely assigned a readout sequence. Each locus is labeled with a total of 400 coding probes, but only four are shown. Fluorescent readout probes complementary to the readout sequences are sequentially added and imaged, allowing the read "1" bits to be read at each locus, thus determining the barcode identity of that locus.

[0431] Figure 20B Representative images from multiple imaging rounds in the nucleus of a single cell are shown. Fluorescence signals from chromatin loci of the readout probe and the signal of 4',6-diamidinyl-2-phenylindole (DAPI), used as a nuclear marker, are displayed. Scale bar: 5 micrometers.

[0432] Figure 20C A magnified image of a small region centered on a chromatin locus (white box in B) is shown across all imaging rounds. Locus identity was determined based on two emitting readout probes (1 and 13). Scale bar: 300 nm.

[0433] Figure 20D This is a 3D rendering of all detected chromatin loci (spheres) in a single IMR-90 cell, color-coded according to the chromosomes they belong to (chromosome indexes are shown below the image). Adjacent loci in the genome sequence are connected by flexible lines. Approximately 1000 genomic loci were imaged.

[0434] Figure 20E Showing with Figure 20D The same chromatin loci in the cells are shown, but two homologs of the chromosomes shown are highlighted.

[0435] Figure 20F The median distance matrix calculated from approximately 5,400 single cells is shown. For each pair of loci, the median 3D spatial distances observed between loci across all cells are presented.

[0436] Figure 20G Example images are shown, illustrating the locations of multiple chromosomal regions within a single cell. Chromosomes are labeled as encoded, and shaded areas represent the convex hulls surrounding all imaged loci. For clarity, only one homolog is shown for each chromosome.

[0437] Figure 20H Showing Figure 20G The matrix shows the spatial distances between identical cells. It illustrates the spatial distances between each pair of chromatin loci. The chromosome order is shown below the matrix, with two homologs for each chromosome displayed.

[0438] Figure 27A-27J Genome-scale imaging using combined FISH is shown: localization error, reproducibility, and comparison with Hi-C. Figure 27A The shift distribution between the localization of genomic loci measured during a combined imaging run and the localization of the same loci individually re-imaged using sequential hybridization after the combined imaging was completed is shown. Ten genomic regions in Chr6 were re-imaged in approximately 2000 cells. The median shift was ~50 nm.

[0439] Figure 27B The nearest neighbor frequency matrix for all 1,041 genomic loci imaged by combined FISH imaging is shown. The nearest neighbor frequency between a pair of loci is calculated by dividing the number of times a cutoff distance of less than 500 nm occurs between the loci by the total number of measured distances between the two loci.

[0440] Figure 27C A correlation plot showing the proximity frequency between intrachromosomal locus pairs derived from our imaging data and the number of contacts derived from the overall Hi-C experiment is presented, binned in 500kb units and centered on the target locus. The Pearson correlation coefficient is 0.91. Available Hi-C data for IMR90 cells are scarce for cross-chromosomal contacts, thus hindering a reliable comparison of cross-chromosomal interactions between our imaging data and Hi-C data.

[0441] Figure 27DThis shows the correlation of pairwise distances between chromatin loci observed in two independent biological replicates in a genome-scale imaging experiment. The Pearson correlation coefficient between the replicates is 0.98. The upper right cloud represents cross-chromosomal pairwise distances, and the lower left cloud represents intrachromosomal pairwise distances.

[0442] Example 13

[0443] Enrichment of transchromosomal AA interactions. Next, we investigated how transchromosomal interactions depend on the A / B compartment identity of chromatin. Based on published global Hi-C data, each imaged genomic locus was classified into A and B compartments. Although genome-scale imaging data also allows for fairly accurate A / B compartment identification, giving approximately 80% agreement with identification based on Hi-C data (…),… Figure 27E However, due to the higher genomic resolution of the overall Hi-C data, Hi-C calls were used to classify the A / B compartment identities of the imaged loci. 38% of the imaged loci belonged to compartment A, while 62% belonged to compartment B. To examine whether the degree of transchromosomal interactions between active and inactive chromatin differed, the genomic loci in the transchromosomal proximity frequency matrix were rearranged, placing all A loci adjacent to each other, followed by all B loci. This matrix showed that the average trend of transchromosomal interactions between compartment A loci and another compartment A locus was stronger than that with individual compartment B loci. Figures 21A-21B This is consistent with previous observations of transchromatic interactions between active chromatin. Conversely, compared to loci A, loci B showed comparable or lower transchromatic affinity for each other. Figures 21A-21B In other words, transchromosomal AA interactions tend to be significantly stronger than AB interactions, and AB interactions, in turn, tend to be slightly stronger than BB interactions. A similar trend was observed for large cutoff distances used to construct proximity frequency maps, provided that a sufficient number of cells were included in the analysis. Figure 27F-27H It is noteworthy that this contrasts sharply with the overall hierarchy of cis-interactions within the same chromosome, where BB interactions show a stronger tendency to form than AA interactions, and AA interactions, in turn, show a stronger tendency to form than AB interactions. Figure 19E However, interestingly, this observed trend in cross-chromosomal interactions (AA>AB) is noteworthy. The trend observed in high-resolution Chr2 data for cis-chromosome interactions at large genomic distances is similar to that seen in BB data. Figure 19F-19G This trend was also observed in long-range cis-chromosome interactions in genome-scale data, with chromosome clusters across all imaging ( Figure 21C ).

[0444] Next, the relationship between transchromosomal interactions and chromatin compartmental identity at the single-cell level was examined. In individual cells, loci A and B exhibited different spatial distributions, with loci A showing a tendency to be more concentrated in the nucleus than loci B, as expected. Figure 21D (Figure 28). To further characterize transchromatin interactions, a density-based approach similar to that described earlier was used for high-resolution whole-chromosome tracing, except that only transchromosome interactions were considered here. In short, for each imaged locus in each chromosome, the local densities of loci A and B from all other chromosomes in the same cell were calculated, and the ratio of these two densities (hereinafter referred to as the trans-A / B density ratio) was determined. Figure 21D-21E This quantity provides a measure of the local enrichment of active chromatin across chromosomes near the locus. It was noted that the majority (62%) of the imaged loci belonged to compartment B, resulting in an overall bias of less than 1 for the A / B ratio. To control for this bias, the distribution of cross-A / B density ratios observed for loci A and B was compared, where the A and B identities of the imaged loci were obtained in a randomized control with the number of A and B loci kept constant while the number of A and B loci was randomly shuffled among the imaged loci. Notably, the cross-A / B density ratio observed for loci A was significantly higher than that observed for loci B, which in turn was higher than that derived from the randomized control. Figure 21E These single-cell analyses further support the idea that cross-chromosomal interactions preferentially enrich interactions between active chromatin.

[0445] Further inquiry was made into whether the enrichment of transchromosomal AA interactions required transcription. To address this issue, transcription was inhibited by treatment with α-amanitin, and it was found that this treatment did not lead to a significant reduction in the enrichment of transchromosomal AA interactions. Figures 29A-29C This observation is consistent with and expands upon previous research showing that the long-range and transchromatic interactions of activated β-globin loci with other active chromatin loci are not inhibited by transcriptional repression.

[0446] Figures 21A-21E This shows an enrichment of active-active chromatin interactions in cross-chromosomal interactions. Figure 21AThe normalized cross-chromosomal proximity frequency matrix is ​​shown. The proximity frequencies between each cross-chromosomal locus pair (a pair of loci on different chromosomes) are displayed, where locus pairs are considered close if the distance between them is less than a cutoff distance of 500 nm. The loci are rearranged such that compartment A loci appear first, followed by compartment B loci; therefore, the top-left block represents interactions between compartment A loci, and the bottom-right block represents interactions between compartment B loci. Each entry in the matrix is ​​normalized by the median proximity frequency of all locus pairs originating from the same chromosome pair to account for different baseline levels of interactions between chromosome pairs.

[0447] Figure 21B The distribution of cross-chromosomal proximity frequencies for locus A pairs (AA; n=72,771 locus pairs), locus B pairs (BB; n=193,753 locus pairs), and pairs consisting of one A and one B locus (AB; n=237,986 locus pairs) is shown, originating from... Figure 21A The matrix shown is represented by a histogram in the upper figure and a box plot in the lower figure, showing the median (center line), the 25th-75th percentiles (boxes), and the 5th-95th percentiles (whiskers).

[0448] Figure 21C The median proximity frequency between chromatin locus pairs within the same chromosome as a function of their genomic distance is shown, averaged for locus pairs separated by the same genomic distance across all chromosomes. The median contact frequency of locus A pairs (AA), locus B pairs (BB), and mixed pairs of loci A and B (AB) is shown.

[0449] Figure 21D The distribution of A-compartment and B-compartment loci in two single cells is shown. The left panel represents the location of all detected loci in a single z-plane within a single nucleus, containing both A-compartment and B-compartment loci. In the right panel, the shading of each locus represents the local density ratio of A and B loci across the chromosome, i.e., the A / B density ratio, consistent with the scale bar shown on the right.

[0450] Figure 21EThe distribution of local cross-A / B density ratios for imaged genomic loci is shown. For each locus, the median cross-A / B density ratio for all cells was determined, and the distribution of cross-A / B density ratios for loci A (n=382 loci) and for loci B (n=623 loci) is shown. Due to the Hi-C dataset used for compartment recall and the different forms of genome assembly used in this study, 36 of the 1041 imaged loci were not assigned A / B identities. Dark gray histograms represent randomized controls where A and B compartment identities were randomly shuffled while keeping the total number of A loci and the total number of B loci constant.

[0451] Figure 27E This is a bar chart showing the percentage of loci whose A / B compartmentalization is consistent between genome-scale imaging data and Hi-C data for each human autosome. On average, approximately 81% of loci per chromosome show consistency in A / B allocation between our imaging data and Hi-C data.

[0452] Figure 27F The median normalized cross-chromosomal AA, AB, and BB neighbor frequencies (e.g., ) are shown as a function of the cutoff distance used to assess proximity. Figure 21A and 21B (as defined in the text). The standardized approximations were calculated from ~5,400 IMR-90 cells. Median-normalized transchromosomal AA, AB, and BB neighbor frequencies after perturbing locus locations with a 3D Gaussian noise term with a 50 nm standard deviation are also shown, comparable to the estimated localization measurement error, such as... Figure 27A As shown.

[0453] Figure 27G and Figure 27F The results were similar, but when additional data from α-amanitin-treated cells were combined with those from untreated cells (approximately 9,500 cells in total), α-amanitin-treated cells showed similar enrichment of AA across chromosomes relative to neighboring frequencies of AB and BB as untreated cells. Figure 29A , 29B The merge re...

Claims

1. A method comprising: Multiple nucleic acid targets in the genome are associated with multiple codewords, where each codeword includes multiple positions and a value for each position; Exposing samples containing genomes to multiple nucleic acid probes; For each of the multiple nucleic acid probes, determine the binding of the nucleic acid probe within the sample; Generates code words corresponding to the binding of multiple nucleic acid probes within the sample; and The identity of the nucleic acid target is determined based on the assigned code word.

2. A method comprising: Determine the location of newly generated RNA within the cell nucleus; Applying ribonuclease to the cell nucleus; and To determine the location of DNA within the cell nucleus.

3. A method comprising: Determine the location of newly generated RNA within the cell nucleus; Determine the location of DNA within the cell nucleus; and To determine the location of proteins within the cell nucleus.

4. A method comprising: Determine the location of newly generated RNA within the cell nucleus; Determine the location of DNA within the cell nucleus; and Determine the location of nucleic acids within the cell nucleus, where the nucleic acids are neither newly generated RNA nor DNA.

5. A method comprising: MERFISH was used to image chromatin in cells.

6. A method comprising: Imaging at least 100 different genomic loci in a single cell.

7. A method comprising: Associate multiple nucleic acid targets in the genome with multiple code words; A sample containing cells suspected of containing a genome is exposed to multiple nucleic acid probes, wherein at least some of the multiple nucleic acid probes contain a first portion containing a target sequence and a second portion containing one or more readout sequences, wherein each readout sequence represents a value at a position within multiple codewords; The sample is exposed to one or more interposers in one round, wherein each interposer contains a first part that is substantially complementary to one of the readout sequences and a second part that contains an identification sequence; The sample is exposed to one or more readout probes in one round to determine one or more identification sequences, wherein each readout probe contains a first portion comprising a sequence substantially complementary to one of the identification sequences and a second portion comprising a signal transduction entity; Identify signal-transmitting entities at at least some locations in the sample; and Inactivate signal transduction entities at at least some locations in the sample; The steps of repeatedly exposing the sample to one or more adapters and one or more readout probes in a round, identifying the signal transduction entity and inactivating the signal transduction entity, wherein one or more different signal transduction entities are used in each round; The code word at the location is determined based on the signal transduction entity in the sample; and Nucleic acid targets in samples are identified based on code words.

8. A method comprising: Associate multiple nucleic acid targets in the genome with multiple code words; A sample containing cells suspected of containing a genome is exposed to multiple nucleic acid probes, wherein at least some of the multiple nucleic acid probes contain a first portion containing a target sequence and a second portion containing one or more readout sequences, wherein each readout sequence represents a value at a position within multiple codewords; The sample is exposed to one or more interposers in one round, wherein each interposer contains a first part that is substantially complementary to one of the readout sequences and a second part that contains an identification sequence; The sample is exposed to one or more readout probes in one round to determine one or more identification sequences, wherein each readout probe contains a first portion comprising a sequence substantially complementary to one of the identification sequences and a second portion comprising a signal transduction entity; Identify signal-transmitting entities at at least some locations in the sample; and Inactivate signal transduction entities at at least some locations in the sample; The steps of repeatedly exposing a sample to one or more adapters and one or more readout probes in one round, identifying signal transduction entities, and inactivating signal transduction entities are repeated, wherein at least one of the signal transduction entities is used for more than one round; The code word at the location is determined based on the signal transduction entity in the sample; and Nucleic acid targets in samples are identified based on code words.

9. A method comprising: A sample containing cells suspected of containing a genome is exposed to multiple nucleic acid probes, wherein at least some of the multiple nucleic acid probes contain a first portion containing a target sequence and a second portion containing one or more readout sequences; The sample is exposed to one or more interposers in one round, wherein each interposer contains a first portion substantially complementary to one of the readout sequences, and a second portion containing an identification sequence; The sample is exposed to one or more readout probes in one round to determine one or more identification sequences, wherein each readout probe contains a first portion comprising a sequence substantially complementary to one of the identification sequences and a second portion comprising a signal transduction entity; Identify signal-transmitting entities at at least some locations in the sample; and Inactivate signal transduction entities at at least some locations in the sample; The steps of repeatedly exposing the sample to one or more adapters and one or more readout probes in a round, identifying the signal transduction entity and inactivating the signal transduction entity, wherein one or more different signal transduction entities are used in each round; Nucleic acid targets in the sample are determined based on the signal transduction entities identified in each round.

10. A method comprising: A sample containing cells suspected of containing a genome is exposed to multiple nucleic acid probes, wherein at least some of the multiple nucleic acid probes contain a first portion containing a target sequence and a second portion containing one or more readout sequences; The sample is exposed to one or more readout probes in one round to determine one or more readout sequences, wherein each readout probe contains a first portion comprising a sequence substantially complementary to one of the readout sequences and a second portion comprising a signal transduction entity; Identify signal-transmitting entities at at least some locations in the sample; and Inactivate signal transduction entities at at least some locations in the sample; Repeat the steps of exposing the sample to one or more readout probes in a round, identifying the signal transduction entity, and inactivating the signal transduction entity, wherein one or more different signal transduction entities are used in each round; Nucleic acid targets in the sample are determined based on the signal transduction entities identified in each round.

11. A method comprising: A sample containing cells suspected of containing a genome is exposed to a round of multiple nucleic acid probes, wherein at least some of the multiple nucleic acid probes contain a first part containing a target sequence and a second part containing a signal transduction entity; Identify signal-transmitting entities at at least some locations in the sample; and Inactivate signal transduction entities at at least some locations in the sample; The steps of repeatedly exposing a sample to multiple nucleic acid probes in one round, identifying the signal transduction entity, and inactivating the signal transduction entity are repeated, wherein one or more different signal transduction entities are used in each round; Nucleic acid targets in the sample are determined based on the signal transduction entities identified in each round.

Images