METHODS FOR DETERMINING SPATIAL AND TEMPORAL GENE EXPRESSION DYNAMICS IN SINGLE CELLS

• US 20190218276A1
• Filed: 10/27/2016
• Published: 07/18/2019
• Est. Priority Date: 03/21/2016
• Status: Active Grant

• Current Assignee: President and Fellows of Harvard College (Harvard Corporation), Massachusetts Institute of Technology, The Broad Institute, Inc.
• Original Assignee: The Broad Institute, Inc.
• Inventors: Regev, Aviv, Zhang, Feng, Habib, Naomi, Li, Yinqing, Heidenreich, Matthias, Swiech, Lukasz, Basu, Anindita, Weitz, David, Davidi, Inbal Avraham
• Granted Patent: US 12,060,412 B2
• Time in Patent Office: Days
• CPC Class Codes: C07K 16/1036 Retroviridae, e.g. leukemia viruses
• CPC Class Codes: C12N 15/1093 General methods of preparing gene libraries, not provided for in other subgroups
• CPC Class Codes: C12N 15/1096 cDNA Synthesis; Subtracted cDNA library construction, e.g. RT, RT-PCR
• CPC Class Codes: C12Q 1/686 Polymerase chain reaction [PCR]
• CPC Class Codes: C12Q 1/6869 Methods for sequencing
• CPC Class Codes: C12Q 2537/165 Mathematical modelling, e.g. logarithm, ratio
• CPC Class Codes: C12Q 2563/179 the label being a nucleic acid
• CPC Class Codes: C12Q 2565/514 characterised by the use of the arrayed oligonucleotides as identifier tags, e.g. universal addressable array, anti-tag or tag complement array
• CPC Class Codes: C12Q 2565/629 being a microfluidic device

Abstract

Transcriptomes of individual neurons provide rich information about cell types and dynamic states. However, it is difficult to capture rare dynamic processes, such as adult neurogenesis, because isolation from dense adult tissue is challenging, and markers for each phase are limited. Here, Applicants developed Nuc-seq, Div-Seq, and Dronc-Seq. Div-seq combines Nuc-Seq, a scalable single nucleus RNA-Seq method, with EdU-mediated labeling of proliferating cells. Nuc-Seq can sensitively identify closely related cell types within the adult hippocampus. Div-Seq can track transcriptional dynamics of newborn neurons in an adult neurogenic region in the hippocampus. Dronc-Seq uses a microfluidic device to co-encapsulate individual nuclei in reverse emulsion aqueous droplets in an oil medium together with one uniquely barcoded mRNA-capture bead. Finally, Applicants found rare adult newborn GABAergic neurons in the spinal cord, a non-canonical neurogenic region. Taken together, Nuc-Seq, Div-Seq and Dronc-Seq allow for unbiased analysis of any complex tissue.

Claims

1. A method of producing at least one high resolution map for visualizing different cell subtypes or cell states in a heterogeneous population of cells comprising:

(a) performing dimensionality reduction on single cell gene expression data obtained from the heterogeneous population of cells;

(b) producing a first set of clusters of cells by a method comprising measuring the dissimilarity between sets of genes in the dimensionality reduced single cell gene expression data and applying a first metric, wherein the clusters are in a dimensionality reduced space and the clusters comprise cells with a continuous trajectory;

(c) producing a set of informative genes by a method comprising scoring genes based on their expression across the first set of clusters of cells or a continuous trajectory of cells, wherein the informative genes are uniquely expressed in cells embedded in close proximity in the dimensionality reduced space; and

(d) producing at least one second set of clusters of cells or continuous trajectory of cells by a method comprising measuring the dissimilarity between the set of informative genes and applying a second metric, whereby visualization of a map of the second set of clusters or continuous trajectory of cells indicate cell subtypes or cell states.

2. The method according to claim 1, further comprising mapping the spatial location of the cell subtypes or cells having a cell state by performing RNA in situ hybridization (ISH) on whole tissue sections comprising said cell subtypes using probes specific for genes expressed in the cell subtypes, whereby the spatial location of cell subtypes is visualized in a biological sample.

3. The method according to claim 1, further comprising mapping the spatial location of the cell subtypes or cells having a cell state by comparing gene expression data for each cell type to landmark gene expression patterns in tissue samples, whereby the spatial location of cell subtypes is visualized in a biological sample.

4. The method according to claim 1, wherein producing the second set of clusters of cells in step (d) comprises producing more than one set of clusters, wherein each set of clusters is produced by using the highest scoring informative gene and each successive cluster is produced by adding the next highest scoring informative gene.

5. The method according to claim 1, further comprising normalization of the single cell gene expression data, wherein gene expression of one cell is normalized to another using not highly expressed genes, or estimation of missed detection probability, wherein an expectation-maximization algorithm is applied.

7. The method according to claim 1, wherein scoring informative genes comprises applying a Moran'"'"'s I analysis and/or a Manhattan distance analysis, or wherein dimensionality reduction comprises PCA and/or tSNE or both.

9. The method according to claim 1, wherein the heterogeneous population of cells is derived from a section of a tissue or a tumor from a subject, or is a population of cells grown in tissue culture in particular neurons or immune cells.

10. The method according to claim 9, wherein the section is obtained by microdissection.

11. The method according to claim 9, wherein the tissue is nervous tissue optionally from the brain, spinal cord, or retina.

16. The method according to claim 1, wherein the gene expression data is obtained from single cell sequencing or from single nucleic sequencing.

18. The method according to claim 16, wherein single nuclei sequencing comprises:

(a) treating the heterogeneous population of cells with a reagent that stabilizes RNA;

(b) extracting nuclei from the cells;

(c) sorting single nuclei into separate reaction vessels;

(d) extracting RNA from the single nuclei;

(e) generating a cDNA library; and

(f) sequencing the library,

whereby gene expression data from single cells is obtained.

19. The method according to claim 18, wherein single nuclei are sorted into single wells of a plate by FACS or wherein the sorting comprises microfluidics.

22. The method according to claim 18, wherein single nuclei sequencing comprises a method of high-throughput single nuclei sequencing, said method comprising:

(a) treating the heterogeneous population of cells with a reagent that stabilizes RNA;

(b) extracting nuclei;

(c) generating a suspension of isolated nuclei, wherein the suspension comprises a nuclear pore blocking polymer;

(d) optionally, enriching the nuclei suspension by FACS or magnetic-activated cell sorting (MACS);

(e) applying the nuclei suspension to a reverse emulsion microfluidic device configured for single nuclei, wherein single nuclei are individually compartmentalized with a single uniquely barcoded capture bead in an emulsion drop;

(f) extracting mRNA onto the barcoded capture beads;

(g) generating a barcoded cDNA library; and

(h) sequencing the library using paired-end sequencing,

whereby gene expression data from single nuclei is obtained.

23. The method according to claim 22, wherein the nuclei suspension comprises 10⁵-10⁶ nuclei or wherein 10⁴-10⁵ nuclei are sequenced.

25. The method according to claim 22, wherein the nuclear pore blocking polymer comprises a poloxamer, or wherein the reagent that stabilized RNA comprises the properties of RNAlater, or both.

63. The method according to claim 18, wherein nuclei are further detectable by a fluorescent signal, whereby individual nuclei may be further sorted.

64. The method according to claim 63, wherein cells express a nuclear localized fluorescent protein.

65. The method according to claim 64, wherein the nuclear localized fluorescent protein comprises a fluorescent protein fused to a KASH protein domain.

66. The method according to claim 63, wherein single nuclei are immunostained with an antibody with specific affinity for an intranuclear protein.

67. The method according to claim 66, wherein the antibody is specific for NeuN.

68. The method according to claim 63, wherein nuclei are stained with a nuclear stain optionally comprising DAPI, Ruby red, trypan blue, Hoechst or propidium iodine.

6. (canceled)

8. (canceled)

12-15. -15. (canceled)

17. (canceled)

20. (canceled)

21. (canceled)

24. (canceled)

26. (canceled)

27. A method of producing a temporally phased single-cell sequencing library of at least one cell type or subtype, wherein said sequencing library comprises cells along a continuous trajectory of cell developmental stages comprising:

(a) treating more than one population of cells of a single cell type or subtype, or optionally a heterogeneous cell type, with a nucleoside analogue, wherein the nucleoside analogue is incorporated into replicating DNA and is configured for labeling with a detectable marker;

(b) isolating a first population of cells at one time point and isolating at least one other population of cells at a later time point, optionally, isolating single nuclei from the isolated populations of cells;

(c) staining the nucleoside analogue incorporated into replicated DNA with the detectable marker within each population of cells or single nuclei isolated from each population of cells, wherein the DNA is stained with the detectable marker;

(d) sorting the stained and/or unstained cells or optionally, sorting the stained and/or unstained single nuclei into separate reaction vessels; and

(e) sequencing the RNA from the sorted single cells or optionally, sorted single nuclei, whereby single cell gene expression data is obtained for cells at different stages of maturation.

28. The method according to claim 27, wherein the treating more than one population of cells of a single cell type or subtype, or optionally a heterogeneous cell type with a nucleoside analogue is performed in at least one subject.

29. The method according to claim 28, wherein the subject is a mouse.

30. The method according to claim 28, wherein isolating one population of cells comprises dissection of a tissue from the subject, or the population of cells is grown in tissue culture, and wherein the tissue is optionally nervous tissue in particular nervous tissue isolated from the brain, spinal cord, or retina or immune cells.

36. The method according to claim 27, wherein the gene expression data is obtained from single cell sequencing, or wherein the gene expression data is obtained from single nucleic sequencing.

38. The method according to claim 36, wherein single nuclei sequencing comprises:

(a) treating the populations of cells with a reagent that stabilizes RNA;

(b) extracting nuclei from the cells;

(c) sorting single nuclei into separate reaction vessels;

(d) extracting RNA from the single nuclei;

(e) generating a cDNA library; and

(f) sequencing the library,

whereby gene expression data from single cells is obtained.

39. The method according to claim 38, wherein single nuclei are sorted into single wells of a plate by FACS or wherein the sorting comprises microfluidics.

42. The method according to claim 36, wherein single nuclei sequencing comprises a method of high-throughput single nuclei sequencing, said method comprising:

(a) treating the population of cells with a reagent that stabilizes RNA;

(b) extracting nuclei;

(c) generating a suspension of isolated nuclei, wherein the suspension comprises a nuclear pore blocking polymer;

(d) optionally, enriching the nuclei suspension by FACS or magnetic-activated cell sorting (MACS);

(e) applying the nuclei suspension to a reverse emulsion microfluidic device configured for single nuclei, wherein single nuclei are individually compartmentalized with a single uniquely barcoded capture bead in an emulsion drop;

(f) extracting mRNA onto the barcoded capture beads;

(g) generating a barcoded cDNA library; and

(h) sequencing the library using paired-end sequencing,

whereby gene expression data from single nuclei is obtained.

43. The method according to claim 42, wherein the nuclei suspension comprises 10⁵-10⁶ nuclei or wherein 10⁴-10⁵ nuclei are sequenced.

45. The method according to claim 42, wherein the nuclear pore blocking polymer comprises a poloxamer, or wherein the reagent that stabilizes RNA comprises the properties of RNAlater, or both.

47. The method according to claim 27, further comprising producing at least one high resolution map for visualizing the temporal position or cell developmental stage of cells of a specific cell type, subtype or cell state during proliferation comprising:

(a) optionally, performing a method of producing at least one high resolution map for visualizing different cell subtypes or cell states in a heterogeneous population of cells comprising;

(i) performing dimensionality reduction on single cell gene expression data obtained from the heterogeneous population of cells;

(ii) producing a first set of clusters of cells by a method comprising measuring the dissimilarity between sets of genes in the dimensionality reduced single cell gene expression data and applying a first metric, wherein the clusters are in a dimensionality reduced space and the clusters comprise cells with a continuous trajectory;

(iii) producing a set of informative genes by a method comprising scoring genes based on their expression across the first set of clusters of cells or a continuous trajectory of cells, wherein the informative genes are uniquely expressed in cells embedded in close proximity in the dimensionality reduced space; and

(iv) producing at least one second set of clusters of cells or continuous trajectory of cells by a method comprising measuring the dissimilarity between the set of informative genes and applying a second metric, whereby visualization of a map of the second set of clusters or continuous trajectory of cells indicate cell subtypes or cell states,

whereby heterogeneous cells are clustered by cell type, subtype, or cell state;

(b) performing dimensionality reduction on the single cell gene expression data from the stained cells of a single cell type, subtype or cell state within each population of cells or the stained single nuclei of a single cell type or subtype isolated from each population of cells;

(c) measuring the dissimilarity between sets of genes in the dimensionality reduced single cell gene expression data and applying a first metric, whereby a continuous trajectory is visualized in the dimensionality reduced space from an early time point to a later time point;

(d) producing a set of informative genes by a method comprising scoring genes based on their expression across the continuous trajectory, wherein the informative genes are uniquely expressed in cells embedded in close proximity in the dimensionality reduced space, optionally, wherein lowly expressed genes are filtered out; and

(e) producing at least one set of clusters of cells by a method comprising measuring the dissimilarity between the set of informative genes and applying a second metric, whereby visualization of the set of clusters in the dimensionality reduced space indicate the gene expression profiles of cells based on a temporal position or developmental stage.

48. The method according to claim 47, wherein producing the set of clusters of cells in step (d) comprises producing more than one set of clusters, wherein the first set of clusters is produced by using the highest scoring informative gene and each successive set of clusters is produced by adding the next highest scoring informative gene.

49. The method according to claim 47, further comprising a) normalization of the single cell gene expression data, wherein gene expression of one cell is normalized to another using not highly expressed genes, or b) estimation of missed detection probability, wherein an expectation maximization algorithm is applied, or both a) and b).

51. The method according to claim 47, wherein scoring informative genes comprises applying a Moran'"'"'s I analysis and/or a Manhattan distance analysis, or wherein dimensionality reduction comprises PCA and/or tSNE, or both.

53. The method according to claim 27, wherein the nucleoside analogue comprises EdU (5-ethynyl-2′

-deoxyuridine).

31-35. -35. (canceled)

37. (canceled)

40. (canceled)

41. (canceled)

44. (canceled)

46. (canceled)

50. (canceled)

52. (canceled)

54. A method for creating a composite single nuclei sequencing library comprising:

(a) merging one uniquely barcoded RNA capture microbead with a single nuclei in an emulsion droplet having a diameter from 50 μ

m to 210 μ

m, wherein the single nuclei is blocked with a nuclear pore blocking polymer;

(b) extracting RNA onto the RNA capture microbead;

(c) performing a reverse transcription reaction to convert the mRNA to first strand cDNA that is covalently linked to the RNA capture microbead;

or conversely reverse transcribing within droplets and thereafter breaking droplets and collecting cDNA-attached beads;

(d) preparing and sequencing a single composite RNA-Seq library, containing cell barcodes that record the cell-of-origin of each RNA, and unique molecular identifiers (UMI) that distinguish among RNAs from the same cell.

56. The method of claim 54 or 55, wherein the method of amplifying the cDNA-attached beads is template switch amplification.

57. The method of claim 54 or 55, wherein the method of amplifying the cDNA-attached beads is T7 linear application.

58. The method of claim 54 or 55, wherein the method of amplifying the cDNA-attached beads is exponential isothermal amplification.

59. The method of claim 54 or 55, wherein the emulsion droplet is formed via co-encapsulation comprising RNA capture microbead and composite single nuclei.

60. The method of claim 54 or 55, wherein the RNA is mRNA.

61. The method of claim 54 or 55, wherein the diameter of the emulsion droplet is 50-100 μ

m or 70-90 μ

m.

62. The method of claim 54 or 55, wherein the diameter of the RNA capture microbeads is from 10 μ

m to 95 μ

m.

55. A method for creating a composite single-cell sequencing library comprising:

(a) merging one uniquely barcoded RNA capture microbead with a single nuclei in an emulsion droplet having a diameter from 50 μ

m to 210 μ

m, wherein the single nuclei is blocked with a nuclear pore blocking polymer;

(b) extracting RNA onto the RNA capture microbead;

(c) breaking droplets and pooling beads in solution;

(d) performing a reverse transcription reaction to convert the RNA to first strand cDNA that is covalently linked to the RNA capture microbead;

or conversely reverse transcribing within droplets and thereafter breaking droplets and collecting cDNA-attached beads;

(e) preparing and sequencing a single composite RNA-Seq library, containing cell barcodes that record the cell-of-origin of each RNA, and unique molecular identifiers (UMI) that distinguish among RNAs from the same cell.

69. (canceled)

70. An apparatus for creating a composite single nuclei sequencing library via a microfluidic system, comprising:

a oil-surfactant inlet comprising a filter and a carrier fluid channel, wherein said carrier fluid channel further comprises a resistor;

an inlet for an analyte configured for nuclei comprising a filter and a carrier fluid channel, wherein said carrier fluid channel optionally further comprises a resistor;

an inlet for mRNA capture microbeads and lysis reagent comprising a filter and a carrier fluid channel, wherein said carrier fluid channel optionally further comprises a resistor;

said carrier fluid channels have a carrier fluid flowing therein at an adjustable or predetermined flow rate;

wherein each said carrier fluid channels merge at a junction; and

said junction being connected to a mixer, which contains an outlet for drops.

71. The apparatus of claim 70, wherein the analyte comprises a nuclear pore blocking polymer, an enzyme and a nuclei.

72. The apparatus of claim 70, wherein said junction is connected to said mixer by a fluid carrier channel with a constriction for droplet pinch-off.

73. The apparatus of claim 70, wherein the analyte is a nuclei or a host pathogen cell or nuclei.

75. The apparatus of claim 70, wherein the lysis reagent comprises an anionic surfactant, such as sodium lauroyl sarcosine, or a chaotropic salt, such as guanidiunium thiocyanate.

76. The apparatus of claim 70, wherein the filter comprises square PDMS.

77. The apparatus of claim 70, wherein the resistor is serpentine having a length from 7000-9000 μ

m, width of 50-75 μ

m and depth of 100-150 mm, optionally wherein the resistor has a diameter of 50 μ

m.

79. The apparatus of claim 70, wherein the channels having a length of length of 8000-12,000 μ

m and width of 125-250 mm, and depth of 100-150 mm.

80. The apparatus of claim 79, wherein the channel diameter is 75 μ

m, 80 μ

m, 85 μ

m, 90 μ

m or 125 μ

m.

82. The apparatus of claim 70, wherein the mixer has a length of 7000-9000 μ

m and a width of 80-140 μ

m or 75-90 μ

m.

83. The apparatus of claim 82, wherein the mixer width is 70-90 μ

m, or 125 μ

m.

85. The apparatus of claim 70, wherein the oil-surfactant is a PEG block polymer, optionally BIORAD™

QX200 Droplet Generation Oil.

87. The apparatus of claim 70, wherein the carrier fluid is water-glycerol mixture.

74. (canceled)

78. (canceled)

81. (canceled)

84. (canceled)

86. (canceled)

88. A method of determining cells responsive to a therapeutic agent comprising:

(a) contacting a tissue of interest with a therapeutic agent and a nucleoside analogue, wherein the nucleoside analogue is incorporated into replicating DNA and is configured for labeling with a detectable marker;

(b) isolating the tissue of interest, optionally, isolating single nuclei from the tissue of interest;

(c) staining the nucleoside analogue incorporated into replicated DNA with the detectable marker within the tissue of interest or single nuclei isolated from the tissue of interest, wherein the DNA is stained with the detectable marker;

(d) sorting the stained and/or unstained cells or optionally, sorting the stained and/or unstained single nuclei into separate reaction vessels;

(e) sequencing the RNA from the sorted single cells or optionally, sorted single nuclei, whereby single cell gene expression data is obtained; and

(f) determining cell subtypes and/or cell subtypes of a specific developmental stage, whereby cells responsive to a therapeutic agent are determined.

89. The method according to claim 88, wherein the nucleoside analogue is contacted with the tissue of interest before the therapeutic agent, optionally between 2 and 14 days before the therapeutic agent.

91. The method according to claim 88, wherein the nucleoside analogue and therapeutic agent is administered from an implantable device or localized injection.

92. The method according to claim 88, wherein the tissue of interest is a tumor.

93. The method according to claim 88, wherein the cell subtypes are determined by a method comprising:

(a) performing dimensionality reduction on the single cell gene expression data obtained from the heterogeneous population of cells;

(b) producing a first set of clusters of cells by a method comprising measuring the dissimilarity between sets of genes in the dimensionality reduced single cell gene expression data and applying a first metric, wherein the clusters are in a dimensionality reduced space and the clusters comprise cells with a continuous trajectory;

(c) producing a set of informative genes by a method comprising scoring genes based on their expression across the first set of clusters of cells or a continuous trajectory of cells, wherein the informative genes are uniquely expressed in cells embedded in close proximity in the dimensionality reduced space; and

(d) producing at least one second set of clusters of cells or continuous trajectory of cells by a method comprising measuring the dissimilarity between the set of informative genes and applying a second metric, whereby visualization of a map of the second set of clusters or continuous trajectory of cells indicate cell subtypes or cell states;

optionally wherein the cell subtypes of a specific developmental stage are determined by determining the cell subtype of cells that are stained and unstained.

95. The method according to claim 93, wherein the cell subtypes of a specific developmental stage are determined by a method comprising:

(a) performing dimensionality reduction on the single cell gene expression data from the stained cells of a single cell type, subtype or cell state within each population of cells or the stained single nuclei of a single cell type or subtype isolated from each population of cells;

(b) measuring the dissimilarity between sets of genes in the dimensionality reduced single cell gene expression data and applying a first metric, whereby a continuous trajectory is visualized in the dimensionality reduced space from an early time point to a later time point;

(c) producing a set of informative genes by a method comprising scoring genes based on their expression across the continuous trajectory, wherein the informative genes are uniquely expressed in cells embedded in close proximity in the dimensionality reduced space, optionally, wherein lowly expressed genes are filtered out; and

(d) producing at least one set of clusters of cells by a method comprising measuring the dissimilarity between the set of informative genes and applying a second metric, whereby visualization of the set of clusters in the dimensionality reduced space indicate the gene expression profiles of cells based on a temporal position or developmental stage, optionally comprising producing more than one set of clusters, wherein the first set of clusters is produced by using the highest scoring informative gene and each successive set of clusters is produced by adding the next highest scoring informative gene.

97. The method according to claim 88, further comprising a) normalization of the single cell gene expression data, wherein gene expression of one cell is normalized to another using not highly expressed genes, or b) estimation of missed detection probability, wherein an expectation maximization algorithm is applied, or both a) and b).

99. The method according to claim 88, wherein scoring informative genes comprises applying a Moran'"'"'s I analysis and/or a Manhattan distance analysis or wherein dimensionality reduction comprises PCA and/or tSNE or both.

101. The method according to claim 88, wherein the nucleoside analogue comprises EdU (5-ethynyl-2′

-deoxyuridine).

90. (canceled)

94. (canceled)

96. (canceled)

98. (canceled)

100. (canceled)