Method of resolving dynamics of cellular transitions

Abstract

A method of analyzing cell fate over a course of time is disclosed. The method comprises analyzing cells which have been temporally labelled at a predetermined location at at least two different time points using at least two different labels, each of said at least two different labels corresponding to one of said at least two different time points, wherein said cells temporally labelled at said at least two different time points have localized to a target location prior to said analyzing.

Description

[0001] METHOD OF RESOLVING DYNAMICS OF CELLULAR TRANSITIONS

[0002] RELATED APPLICATION / S

[0003] This application claims the benefit of priority of Israel Patent Application No. 309573 filed on December 20, 2024, the contents of which are incorporated herein by reference in their entirety.

[0004] FIELD AND BACKGROUND OF THE INVENTION

[0005] The present invention, in some embodiments thereof, relates to a method of temporally analyzing cellular state transitions over time.

[0006] A fundamental challenge when studying biological systems is the description of cell state dynamics. During transitions between cell states, a multitude of parameters may change - from the promoters that are active, to the RNAs and proteins that are expressed and modified. Cells can also adopt different shapes, alter their motility and change their reliance on cell-cell junctions or adhesion. These parameters are integral to how a cell behaves and collectively define the state a cell is in. Until presently, technical challenges have prevented the measuring of all of these parameters simultaneously and dynamically.

[0007] Single-cell RNA sequencing (scRNA-seq) has dramatically improved our ability to measure cell states at high resolution and scale. However, due to the destructive nature of the methodology, scRNA-seq can only capture static snapshots of gene expression at experimental endpoints, thus fundamentally omitting the temporal dimension. Despite this limitation, scRNA- seq profiling at multiple time points in dynamic processes has provided insight into transcriptional drivers of development and differentiation in myriad systems. Computational approaches, such as pseudotime and RNA velocity estimations, infer a temporal dimension in scRNA-seq experiments either by ordering cells along a trajectory based on gene expression similarity or estimating the time derivative of gene expression state by distinguishing spliced and unspliced mRNAs, respectively. While these approaches are extremely useful, these algorithmic tools are sensitive to user-defined parameters and lack an empirically measured ground truth, making it difficult to test imputed conclusions, particularly in complex in vivo settings where multiple trajectories may coexist. Alternatively, combining scRNA-seq with metabolic labeling of new mRNA molecules offers a more direct measurement of gene expression changes over time and has provided insight into rapid cellular responses to stimuli such as viral entry, glucocorticoid receptor activation, and T cell signaling. However, metabolic labeling is currently best suited for measuring processes that unfold over hours and it is difficult to apply in vivo and to differentiation processes that span multiple days. Background art includes Potter et al., Sci. Transi, Med. 13, eabb4582 (2021); Mortlock, et al., Frontiers in Immunology 12 (2022): 772332; Shahi, et al. Scientific reports 7.1 (2017): 44447; Kirschenbaum et al., 2024, Cell, 187, pages 1-17; and Tkachev et al., Sci. Transi, Med. 13, eabc0227 (2021).

[0008] SUMMARY OF THE INVENTION

[0009] According to an aspect of the present invention there is provided a method of analyzing cell fate over a course of time comprising analyzing cells which have been temporally labelled at a predetermined location at at least two different time points using at least two different labels, each of the at least two different labels corresponding to one of the at least two different time points, wherein the cells temporally labelled at the at least two different time points have localized to a target location prior to the analyzing, thereby analyzing cell fate over the course of time.

[0010] According to embodiments of the invention, the method further comprises temporally labelling the cells at the predetermined location prior to the analyzing.

[0011] According to embodiments of the invention, the temporally labelling the cells at the predetermined location is effected in vivo.

[0012] According to embodiments of the invention, the temporally labelling the cells at the predetermined location is effected ex vivo.

[0013] According to embodiments of the invention, the analyzing is effected in vivo.

[0014] According to embodiments of the invention, the analyzing is effected ex vivo.

[0015] According to embodiments of the invention, the target location is in the body of an animal.

[0016] According to embodiments of the invention, the analyzing is effected at least 24 hours after the later time point of the at least two different time points.

[0017] According to embodiments of the invention, the analyzing is effected at least 1 hour following the contacting additional cells.

[0018] According to embodiments of the invention, the temporally labelling the cells comprises:

[0019] (a) contacting cells of a cell type with a cell affinity agent which comprises a first label, the cell affinity agent being specific for the cell type, so as to generate first label-bound cells at the predetermined location; and subsequently

[0020] (b) contacting additional cells of the cell type with a cell affinity agent which comprises a second label, so as to generate second label-bound cells at the predetermined location.

[0021] According to embodiments of the invention, the cell affinity agent binds to an exterior of the cells. According to embodiments of the invention, the cell affinity agent is selected from the group consisting of an antibody, a nanobody and an aptamer.

[0022] According to embodiments of the invention, the cell affinity agent comprising the first label is of the same type as the cell affinity agent comprising the second label.

[0023] According to embodiments of the invention, the cell affinity agent comprising the first label binds to the target and the cell affinity agent comprising the second label.

[0024] According to embodiments of the invention, the cells have been temporally labelled at at least three different time points, wherein the cells temporally labeled at the at least three different time points have localized to the target location prior to the analyzing.

[0025] According to embodiments of the invention, the method further comprises, following step (b):

[0026] (c) contacting additional cells of the cell type with the cell affinity agent which comprises a third label, so as to generate third label-bound cells at the predetermined location.

[0027] According to embodiments of the invention, the cell affinity agent is an antibody.

[0028] According to embodiments of the invention, the first label and the second label comprise fluorescent moieties.

[0029] According to embodiments of the invention, the first label, the second label and the third label comprises fluorescent moieties.

[0030] According to embodiments of the invention, the first label and the second label comprise nucleic acid barcodes.

[0031] According to embodiments of the invention, the first label comprises a first antigen determinant and the second label comprises a second antigen determinant.

[0032] According to embodiments of the invention, the method further comprises contacting the first label-bound cells with an additional affinity agent which specifically binds to the first antigen determinant and which is labeled with a first nucleic acid barcode and contacting the second-label bound cells with an additional affinity agent which specifically binds to the second antigen determinant and which is labeled with a second nucleic acid barcode.

[0033] According to embodiments of the invention, the first label, the second label and the third label comprises nucleic acid barcodes.

[0034] According to embodiments of the invention, the step (b) is effected at least 6 hours following step (a).

[0035] According to embodiments of the invention, the cells comprise white blood cells.

[0036] According to embodiments of the invention, the cell affinity agent comprises an agent which binds specifically to CD45. According to embodiments of the invention, the predetermined location is in the blood circulation of an animal.

[0037] According to embodiments of the invention, the target location is in a tissue or a tumor microenvironment of an animal.

[0038] According to embodiments of the invention, the analyzing comprises cell transcriptome analysis or cell proteome analysis.

[0039] According to embodiments of the invention, the analyzing comprises imaging.

[0040] According to embodiments of the invention, the analysis is effected on the single cell level.

[0041] According to an aspect of the invention there is provided a kit for analyzing cell fate over a course of time comprising: a first pair of affinity agents which comprise:

[0042] (i) a first cell affinity agent which binds to an outer surface of a specific cell, the first cell affinity agent comprising a first antigenic determinant:

[0043] (ii) a second cell affinity agent which binds to an outer surface of the specific cell, the second cell affinity agent comprising a second antigenic determinant; and a second pair of affinity agents which comprise:

[0044] (iii) a third affinity agent which binds specifically to the first antigenic determinant, the third cell affinity agent comprising a first nucleic acid barcode; and

[0045] (iv) a fourth affinity agent which binds specifically to the second antigenic determinant, the third cell affinity agent comprising a second nucleic acid barcode wherein the kit comprises no more than 10 pairs of affinity agents.

[0046] According to embodiments of the invention, the first cell affinity agent is an antibody labeled with a first fluorescent moiety, the first fluorescent moiety comprising the first antigenic determinant and the second cell affinity agent is an antibody labeled with a second fluorescent moiety, eht second fluorescent moiety comprising the second antigenic determinant.

[0047] According to embodiments of the invention, the third cell affinity agent and the fourth cell affinity agent are antibodies.

[0048] According to embodiments of the invention, the kit further comprises agents for performing single cell transcriptome analysis.

[0049] Unless otherwise defined, all technical and / or scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the invention pertains. Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of embodiments of the invention, exemplary methods and / or materials are described below. In case of conflict, the patent specification, including definitions, will control. In addition, the materials, methods, and examples are illustrative only and are not intended to be necessarily limiting.

[0050] BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

[0051] Some embodiments of the invention are herein described, by way of example only, with reference to the accompanying drawings. With specific reference now to the drawings in detail, it is stressed that the particulars shown are by way of example and for purposes of illustrative discussion of embodiments of the invention. In this regard, the description taken with the drawings makes apparent to those skilled in the art how embodiments of the invention may be practiced.

[0052] In the drawings:

[0053] FIG. 1. Sequence of distinctly tagged binders (e.g. antibodies with tagged Fc fragments) against a cell-surface epitope is injected at specific temporal intervals. After some time, the tissue of interest is harvested, followed by single-cell isolation. Next, cells in the single-cell suspension are labeled with a mixture of tag-specific binders (e.g. antibodies). Tag-specific binders harbor tag-specific oligonucleotide barcodes. Next, cells are processed as single-cells, and along with molecular (metabolomic, proteomic, transcriptomic, epigenetic) phenotyping, the timestamp barcode is sequenced. Lastly, temporal barcodes are integrated with the molecular phenotype of each cell, allowing for chronological alignment of molecular phenotypes.

[0054] FIGs. 2A-G.

[0055] Zman-seq facilitates tracing of transcriptomic cell states in vivo across time. A: Schematic illustration of the Zman-seq experimental and computational pipeline.

[0056] B: Median fluorescence intensity bar chart of CD45+PBMC after increasing concentrations of CD45-PE antibody intravenous injections, measured with flow cytometry. Error bars indicate mean and 90% CI (n = 3).

[0057] C: Median fluorescence intensity boxplots of CD45+PBMC after mixing with blood from mice injected with CD45-PE antibody for increasing durations. Staining intensity of PBMC decreases with longer incubation times in the donor mouse. Line in box indicates the median, boxplot extends to the 75thpercentile (n = 3), red line through mean values.

[0058] D: Histology of GBM brain tumors (GL261) from mice intravenously injected with CD45 antibody 24 hours (BB515, green) and 15 minutes (PE, red) before tumor extraction. Bar chart showing the percentage of extravascular (defined based on endothelia stained with CD31, blue) cells harboring the 24 hour or 15 minute staining. Error bars indicate mean and 90% CI (n = 3) and the p-value was calculated using a two-sided Wilcoxon test. E: Flow cytometry density plots of CD45+cells from GL261 tumor after intravenous fluorescent CD45 antibody injections 24, 36, 48, 60 hours before tumor extraction.

[0059] F: Flow cytometry dot plots and box plots of CD45+cells from GE261 tumor showing the in vivo CD45 antibody time labeling (CD45-PE / BB515 / BUV737 / BV711) of microglia and CD45highinfiltrating leukocytes. Infiltrating leukocytes harbor significantly more temporal labeling compared to microglia (p=0.02434, n=3).

[0060] G: Metacell graph projection of CD45+ cells from blood, colon and lung pooled (left panel) and separated by organ (right panels), consisting of 91 MC representing 8,976 pooled single-cells sorted from 3 mice submitted to Zman-seq (12, 24, 36, 48 hour time stamps), and a heatmap of time stamped / non- stained CD45+ cell ratio.

[0061] FIGs. 3A-G

[0062] A: Two-dimensional graph projection of reclustered lymphocytes consisting of 37 metacells representing 2,431 single-cells including NK cells, CD8 T cells, and CD4 T cells.

[0063] B: Cumulative sum distribution plot of each time bin (12h, 24h, and 36h) in each metacell. Each line represents a metacell and the color denotes the normalized area under the curve which we define as cTET. The dotted boxes show the distribution of normalized frequency of each time bin within the earliest cTET metacell (top left) and the longest cTET metacell (bottom right).

[0064] C: Projection of normalized cTET values onto the NK subpopulations as annotated in A.

[0065] D: Base trajectory inference of the time development along the cell types from the earliest cTET chemotactic NK to the latest cTET dysfunctional NK. The shaded color represents the cell type clusters and the four trajectory points correspond to the cluster average cTET. The trajectory arrows are drawn based on the ordering of average cTET (Methods).

[0066] E: Heatmaps of significantly up- (left panel) or down-regulated (right panel) genes across cTET in NK cells. Each row is normalized and smoothed according to the cTET of each metacell and shows the transitioning of genes from chemotactic NK to dysfunctional NK as ordered by cTET. F: Lollipop plot (left panels) and MC map (right panel) displaying the predicted time-dependent interaction between NK cells and the TME. Lollipop plots show the normalized gene expression of the two ligands Tgfbl and Ccll2 across cell types (potential senders). MC maps are colored by the scaled weighted cumulative target gene expression of TGF-βi- and CCL12-regulated target genes in NK cells.

[0067] G: Heatmap (middle panel) showing the interaction potential for the 20 most prioritized ligands and their respective target genes in NK cells that correlate with time. Ligands have been prioritized by their capability to explain gene expression changes across time in NK cells. Heatmap (left panel) indicates the predicted activity for each ligand. Ligands have been manually grouped into biologically relevant sub types. Line plots (top panel) display scaled gene expression of target genes in NK cells across time (red = downregulated, blue = upregulated). cTET = continuous tumor exposure time

[0068] Figures 4A-H.

[0069] Zman-seq reveals temporal NK cell trajectories in the tumor.

[0070] A: Marker gene projections of log2 normalized footprint expression on NK cells only, showing the transitioning from chemotactic NK to dysfunctional NK.

[0071] B: Schematic analytical workflow and methods developed for Zman-seq. The top panel illustrates the approach to define cTET using the cumulative distribution (CDF) profile for each metacell. The bottom panel depicts the construction of temporally-resolved trajectory across time and identification of significant marker genes across time.

[0072] C: The CDF profile for each cell type is plotted separately, with the mean CDF plotted as a dotted line across the center. Each line depicts a metacell and is colored by the cTET.

[0073] D: Heatmaps of significant TF activity along cTET in the NK metacells only. Each row is normalized and smoothed according to the cTET of each metacell and shows the transitioning of TF activity from chemotactic NK to dysfunctional NK as ordered by cTET. Line plots (right panel) show the normalized TF activity in the NKs across time for SMAD2 and SMAD3. The TF activity has been smoothed using loess fitting and the 95% confidence interval is shown as the shaded area.

[0074] E: Schematic representation of applied interaction analysis. Ligands acting on leukocytes in the TME are predicted by correlating time-dependent target genes in leukocytes with documented ligand-target gene databases.

[0075] F: Dot plot displaying the gene expression of prioritized ligands that best explain temporal changes in target gene expression in NK cells (as presented in Figure 3F and G). Dot size indicates percentage of cells expressing the respective gene and color indicates mean normalized gene expression for each cluster.

[0076] G: Heatmap showing the interaction potential between prioritized ligands and their cognate receptor in NK cells.

[0077] H: The effect of TGFβ-blocking antibody on the expression of TNFa and fFNy, and the degranulation (CD 107a) of human NK cells co-cultured with human glioma stem cells.

[0078] FIGs. 5A-F.

[0079] TREM2 antagonistic antibody reprograms the TME by disrupting monocyte-to-TAM transition. A: Volcano plot depicting the differential genes between the Argl TAMs of the IgG treatment against the Acp5 TAMs of the aTREM2 treatment. The significant genes (p-value < 0.001 and log2 fold change > 0.5) were colored purple if they were enriched in IgG treatment and blue if enriched in aTREM2 treatment.

[0080] B: Metacell graph projection of monocytes and TAMs from both IgG and aTREM2 axis, denoting the trajectory of the aTREM2-treated monocyte-to-TAM differentiation based on cTET. The shaded color represents the cell clusters and each of the three points in the trajectory corresponds to the cluster average cTET.

[0081] C: Heatmaps of significant genes along cTET in the aTREM2 treated monocyte-to-TAM differentiation. Each row is normalized and smoothed according to the cTET of each metacell and shows the transitioning of genes from monocyte to Acp5 TAMs as ordered by cTET.

[0082] D: Heatmaps of significant TF activity along cTET in the aTREM2 treated monocyte-to-TAM differentiation. Each row is normalized and smoothed according to the cTET of each metacell and shows the transitioning of TF activity from monocyte to Acp5 TAMs as ordered by cTET.

[0083] E: Dotplot displaying the normalized gene expression of in vitro differentiated BMDMs cultured with GL261 cell culture supernatant in the presence of TREM2 blocking antibody or isotype or PBS control. Dot size indicates percentage of cells expressing the respective gene and color indicates scaled mean gene expression per group.

[0084] F: Circos plot (left panel) indicating the interaction between chemokine ligands produced by 7rem2-expressing cells (senders) and receiver cell types. Colour indicates the log2 fold change enrichment for each ligand upon Trem2 inhibition. Outer semi-circle displays the interaction potential between each ligand and the top 20 affected target genes per cell type. Line plots (right panel) showing scaled expression of chemokine ligands produced by monocyte-derived phagocytes (monocytes, MoMacs and TAMs) across time for aTREM2-treated mice (blue) and mice receiving isotype control (purple). Expression has been smoothed using a polynomial regression model and mean and 90% confidence intervals (shaded area) are shown. BMDMs = bone-marrow-derived macrophages

[0085] FIGs. 6A-B. Single-cell RNAseq of CD28 / CD3 activated and in vitro timestamped T-cells (with CD45-FITC / PE / APC), detected with oligonucleotide -barcoded anti-FITC / PE / APC antibodies. A, Single-cell RNAseq recapitulates a complex mixture of T-cell states. B, Temporal barcoding highlights different T-cell states.

[0086] FIGs. 7A-C. Titration of in vivo timestamping with CD45-FITC / PE / APC antibodies, detected ex vivo with barcoded anti-FITC / PE / APC antibodies. Readout: qPCR CT values. FIGs. 8A-B. Barcode-based Zman-seq highlights state transitions of NK cells (A), recapitulating temporal gene expression signatures previously published with fluorophore-based Zman-seq (B).

[0087] DESCRIPTION OF SPECIFIC EMBODIMENTS OF THE INVENTION

[0088] The present invention, in some embodiments thereof, relates to a method of temporally analyzing cellular state transitions over time.

[0089] Before explaining at least one embodiment of the invention in detail, it is to be understood that the invention is not necessarily limited in its application to the details set forth in the following description or exemplified by the Examples. The invention is capable of other embodiments or of being practiced or carried out in various ways.

[0090] Deciphering the cell-state transitions underlying immune adaptation across time is fundamental for advancing biology. Empirical in vivo genomic technologies that capture cellular dynamics are currently lacking.

[0091] The present inventors have now developed a single-cell technology which records transcriptomic dynamics across time by introducing time stamps into circulating cells, tracking them in tissues for days.

[0092] The technology, referred to herein as Zman-seq is based on a physical barrier separating two biological compartments such that labeling (pulse) of cells only occurs in one compartment, allowing for cell tracking across compartments (chase). By taking advantage of this pulse-chase principle, cells can be temporally labelled in a first location (e.g. vasculature) and can be shielded from consecutive rounds of labeling once they leave the first location and enter a second location (e.g. tissue / tumor). Cells within the second location may be profiled, and stamps on individual cells may be captured using index-sorting to infer the time of infiltration. This enables definition of the time each cell spent in the second location. Combining second location exposure time measurements with single-cell transcriptomic profiling, Zman-seq generates temporal maps of a defined microenvironment.

[0093] Whilst reducing the present invention to practice, the technology was applied in order to resolve cell-state and molecular trajectories of the dysfunctional immune microenvironment in glioblastoma. Within 24 hours of tumor infiltration, cytotoxic natural killer cells transitioned to a dysfunctional program regulated by TGFB1 signaling. Infiltrating monocytes differentiated into immunosuppressive macrophages, characterized by the upregulation of suppressive myeloid checkpoints Trem2, I118bp, and Argl, over 36 to 48 hours. Treatment with an antagonistic anti- TREM2 antibody reshaped the tumor microenvironment by redirecting the monocyte trajectory toward pro-inflammatory macrophages.

[0094] Zman-seq is a broadly applicable technology, enabling empirical measurements of differentiation trajectories, which can enhance the development of more efficacious therapies.

[0095] According to an aspect of the present invention, there is provided a method of analyzing cell fate over a course of time comprising analyzing cells which have been temporally labelled at a predetermined location at at least two different time points using at least two different labels, each of the at least two different labels corresponding to one of the at least two different time points, wherein the cells temporally labelled at the at least two different time points have localized to a target location prior to the analyzing, thereby analyzing cell fate over the course of time.

[0096] The term “temporally labelled” refers to a label which indicates a time that the labeled cell is at the predetermined location.

[0097] In one embodiment, the cells which are analyzed are those of a type that can be selectively labelled over other cells of a different type present in the cell environment.

[0098] Exemplary cells include, but are not limited to T cells, B cells, tumor cells, myeloid cells, red blood cells, embryonic cells, stem cells, fetal cells or maternal cells, nerve cells, neuroglial cells, muscle cells (e.g. skeletal muscle cells, cardiac muscle cells, or smooth muscle cells, skin cells and bacterial cells.

[0099] In one embodiment, the cell expresses a unique cell marker and can be distinguished from other cells in the cell environment by virtue of expression of the unique cell marker or a unique repertoire of cell markers.

[0100] In one embodiment, the unique cell marker is a cell surface marker.

[0101] The cell surface marker may include a carbohydrate, lipid, protein, extracellular protein, cell surface protein, B cell receptor, T cell receptor, major histocompatibility complex, tumor antigen, receptor, intracellular protein, or any combination thereof. In some embodiments of the present invention, surface cell targets include CDla, CDlb, CDlc, CDld, CDle, CD2, CD3d, CD3e, CD3g, CD4, CD5, CD6, CD7, CD8a, CD8b, CD9, CD 10, CDl la, CDl lb, CD11c, CDl ld, CDwl2, CD13, CD14, CD15u, CD15s, CD15su, CD16b, CD17, CD18, CD19, CD20, CD21, CD22, CD23, CD24, CD25, CD26, CD27. CD28. CD29. CD30, CD31, CD32, CD33, CD34, CD35, CD36, CD37, CD38, CD39, CD40, CD41, CD42a, CD42b, CD42c, CD42d, CD43, CD44. CD45 RA-CD 45RB, CD45RC, CD45RO. CD46, CD47, CD48, CD49a, CD49b. CD49c. CD49d, CD49e, CD49f, CD50, CD51, CD52, CD53, CD54, CD55, CD56, CD57, CD58, CD59, CD60a, CD60b, CD60c. CD61, CD62 5262 62L, CD, P, CD, CD64. CD65s, CD66a, CD66b, CD66c, CD66d, CD66e, CD66f, CD68, CD69, CD70, CD71, CD72, CD73, CD74, CD75s, CD77, CD79a, CD79b, CD80, CD8L CD82, CD83, CD84, CD85a, CD85d, CD85j, CD85k, CD86, CD87, CD88, CD89 CD45RB, CD45RC, CD45RO. CD46, CD47, CD48, CD49a, CD49b. CD49c, CD49d, CD49e, CD49f. CD50, CD51 , CD52, CD53. CD54, CD55, CD56, CD57, CD58, CD.59, CD60a, CD60b, CD60c, CD61, CD62E, CD L, CD P, CD, CD64, CD65s CD66a, CD66b, CD66c. CD66d, CD66e, CD66f, CD68, CD69, CD70, CD71. CD72, CD73, CD74, CD75s, CD77, CD79a. CD79b, CD80, CD81, CD82. CD83, CD84, CD85a. CD85d, CD85J, CD85k, CD86. CD87, CD88, CD89, CD80, CD201 , CD202b, CD203c, CD204, CD205, CD206, CD207, CD208, CD209, CD210, CDw210b, CD212, CD213al, CD213a2, CD215, CD217a, CD218b, CD220, CD221, CD222, CD223, CD224. CD225, CD226, CD227. CD228, CD229, CD230, CD231 , CD232. CD233, CD234, CD235a, CD235b, CD236R, CD, CD239, CD240CE, CD240DCE, CD240D, CD, CD242, CD243, CD244. CD245. CD246, CD247, CD248, CD249. CD252, CD253, CD254, CD256, CD266. CD267, CD268, CD269, CD270, CD223, CD273, CD274, CD275, CD276, CD277, CD279, CD283. CD280, CD28L CD282, CD289, CD2. CD276, CD 277: CD 293. CD294, CD295, CD296, CD297, CD298, CD299. CD300a, CD300c, CD300e, CD301 , CD302, CD303, CD304, CD305, CD306, CD307a, CD307b, CD307c. CD307d. CD307e, CD308, CD309, CD312, CD314. CD315, CD316, CD317, CD318, CD319. CD320, CD321, CD322. CD324,

[0102] CD325, CD326, CD327, CD328, CD329, CD331, CD332, CD333, CD334, CD335, CD336,

[0103] CD337, CD338, CD339. CD340, CD344, CD349. CD350, CD351, CD352, CD353, CD354.

[0104] CD355, CD357, CD358, CD360. CD361, CD362, CD363, CD365, CD366, CD367, CD368,

[0105] CD369, CD370, CD371, BCMA, protein, beta-2.

[0106] In another embodiment, the unique cell marker is expressed in the interior of the cell.

[0107] Methods of labelling cells are known in the art and include use of cell affinity agents (e.g. labelling agents) such as antibodies, aptamers, affibodies, DARPins and nanobodies.

[0108] The cell affinity agents binds to the target cell with at least 10 fold affinity than to another cell which dos not

[0109] In some embodiments, the antibody or fragment thereof comprises Fab, fab F (ab:) 2 Fv, scFv, dsFv, bispecific antibodies (diabodies), tri specific antibodies (triabodies), tetraspecific antibodies (tetrabodies), multispecific antibodies formed from antibody fragments, single domain antibodies (sdabs), single chains comprising complementary scFv (tandem scFv) or bispecific tandem scFv, fv constructs, disulfide-linked Fv (disulide-linked Fv).

[0110] As used herein, the term “aptamer” refers to double stranded or single stranded RNA molecule that binds to specific molecular target, such as a protein. Various methods are known in the art which can be used to design protein specific aptamers. The skilled artisan can employ SELEX (Systematic Evolution of Ligands by Exponential Enrichment) for efficient selection as described in Stoltenburg R, Reinemann C, and Strehlitz B (Biomolecular engineering (2007) 24(4):381-403).

[0111] The labelling agents may comprise a detectable moiety or may be a binding target of a second labelling agent which binds to the first labelling agent, wherein the second labelling agent comprises the detectable moiety (e.g. a primary antibody binds to the cell and a secondary antibody which is labelled binds to the primary antibody).

[0112] For example, a first cell population may be labeled with a first cell affinity agent. The first cell affinity agent may comprise a first antigen determinant. The first antigenic determinant may be a fluorescent moiety, such as those described above.

[0113] The second cell population may be labeled with the cell affinity agent which comprises a second antigen determinant which is not identical to the first antigen determinant. The second antigenic determinant may be a second fluorescent moiety.

[0114] Upon removal of the cells from the target location, they are contacted with additional affinity agents. At least one of the additional affinity agents binds specifically to the first antigen determinant, and at least one of the additional affinity agents binds to the second antigen determinant. The additional affinity agents are specifically labeled - e.g. barcoded (with a nucleic acid tag). Identity of the tag can be determined upon single cell sequencing as further described below.

[0115] In one embodiment, the cells at a predetermined location are labeled at a single time point using a first label. After a predetermined amount of time, the cells are analyzed at the target location. In this way, it is possible to analyze cell fate over the course of the predetermined amount of time and compare them with other cells of the same type which have not been labelled.

[0116] In another embodiment, the cells at a predetermined location are labeled at at least two different time points using at least two different labels. The two different labeling agents are the same type of cell labelling agent - e.g. both are antibodies, both labelling agents being attached to a different detectable moiety. In another embodiment, the two different labeling agents recognize the same target, both labelling agents being attached to a different detectable moiety. In still another embodiment, the two different labeling agents are of the same type and recognize the same target, both labelling agents being attached to a different detectable moiety.

[0117] The detectable label, moiety or marker may be detectable based on: such as fluorescence emission, absorbance, fluorescence polarization, fluorescence lifetime, fluorescence wavelength, absorbance wavelength, stokes shift (Stokes shift), light scattering, mass, molecular mass, redox, acoustic, raman, magnetic, radio frequency, enzymatic reactions (including chemiluminescence and electrochemiluminescence), or combinations thereof. In another embodiment, the detectable label is a nucleic acid barcode, the identity of which can be detected upon sequencing of the cell transcriptome.

[0118] The barcode sequence of this aspect of the present invention serves to identify the time at which the labeled cell is at the predetermined location.

[0119] The barcode sequence may be between 3-400 nucleotides, more preferably between 3-200 and even more preferably between 3-100 nucleotides. Thus, for example, the barcode sequence may be 6 nucleotides, 7 nucleotides, 8, nucleotides, nine nucleotides or ten nucleotides.

[0120] In one embodiment, the label may be a fluorophore, chromophore, enzyme substrate, catalyst, redox label, radiolabel, acoustic label, ram an (SERS) tag, mass tag, isotopic tag, magnetic particles, microparticles, nanoparticles, oligonucleotides, or any combination thereof. In some embodiments, the label is a fluorophore (l.e., fluorescent label, fluorescent dye, etc.). Fluorophores of interest may include, but are not limited to, dyes suitable for use in analytical applications (e.g., flow cytometry, imaging, etc.), such as acridine dyes, anthraquinone dyes, arylmethane dyes, diarylmethane dyes (e.g., diphenylmethane dyes), chlorophyll-containing dyes, triarylmethane dyes (e.g., triphenylmethane dyes), azo dyes, diazo dyes, nitrodyes, nitroso dyes, phthalocyanine dyes, cyanine dyes, asymmetric cyanine dyes, quinone-imine dyes, oxazine dyes, diaminoazine dyes, saffron dyes, indamine, indophenol dyes, fluoro dyes, oxazine dyes, oxadone dyes, thiazine dyes, thiazole dyes, xanthene dyes, fluorene dyes, pyronine dyes, fluoro dyes, rhodamine dyes, phenanthridine dyes, and combinations of two or more of the foregoing (e.g., in-line) dyes, polymer dyes having one or more monomeric dye units, and mixtures of two or more of the foregoing dyes. A large number of dyes are commercially available from various sources such as: for example, molecular Probes (Eugene, OR), dyoniics GmbH (Jena, Germany), sigma- Aldrich (St. Louis, MO), siriben, inc. (Santa Barbara, calif.) and Excion (Dayton, OH). For example, the fluorophore may comprise 4-acetamido-4 ’-isothiocyanatestilbene-2, 2’ -disulfonic acid; acridine and derivatives such as acridine, acridine orange, acridine yellow, acridine red and acridine isothiocyanate; allophycocyanin; phycoerythrin; polymethylalginate (peridinin) -chlorophyll protein; 5- (2‘ -aminoethyl) aminonaphthalene- 1 -sulfonic acid (EDANS); 4-amino-N- [ 3- vinylsulfonyl) phenyl ]Naphthalimide-3, 5 disulfonic acid (Lucifer Yellow VS); n- (4-anilino- 1 - naphthyl) maleimide; anthranilamide (anthranilamide): brilliant Yellow; coumarin and derivatives such as coumarin, 7-ainino-4-methylcoumarin (AMC, coumarin 120), 7-amino-4- trifluoromethyicoumarin (coumarin 151); cyanine and its derivatives such as tetrachlorotetrabromofluorescein (cyanosine), cy3, cy3.5, cy5, cy5.5, and Cy7;4', 6-diamidino-2- phenylindole (DAPI); 5', 5 "-dibromoPyrogallol sulfophthalein (bromophthalic trimellite red); 7- diethylamino-3- (4' -isothiocyanatophenyl) -4-methylcoumarin; diethylaminocoumarin; diethylene triamine pentaacetate; 4,4 '-diisothiocyanidine dihydro-stilbene-2, 2' -disulfonic acid; 4,4 '-diisocyanatostilbene-2, 2' -disulfonic acid; 5- [ dimethylamino ]]Naphthalene-l-sulfonyl chloride (DNS, dansyl chloride): 4- (4’ -dimethylaminophenylazo) benzoic acid (DABCYL); 4- dimethylaminophenylazo phenyl-4' -isothiocyanate (DABITC); eosin and derivatives such as eosin and eosin isothiocyanate; erythrosine and derivatives, such as erythrosine B and erythrosine isothiocyanate: ethidium; fluorescein and derivatives such as 5-carboxyfluorescein (FAM), 5- (4, 6-dichlorotriazin-2-yl) aminofluorescein (DTAF), 2’7’ -dimethoxy-4 '5’ -dichloro-6- carboxyfluorescein (JOE), fluorescein Isothiocyanate (FITC), chlorotriazinyl fluorescein, naphthofluorescein and qflitc (XRITC); fluorescent amine; IR.144; IR1446: green Fluorescent Protein (GFP); reef Coral Fluorescent Protein (RCFP); lissamineTMThe method comprises the steps of carrying out a first treatment on the surface of the Lissamine rhodamine, lucifer yellow (Lucifer yellow); malachite isothiocyanate green; 4-methylumbelliferone; o-cresolphthalein; nitrotyrosine; secondary fuchsin; nile red; Oregon green (Oregon green); phenol red: b- phycoerythrin; o-phthalaldehyde; pyrene and derivatives such as pyrene, pyrene butyrate and succinimidyl 1 -pyrene butyrate; reactive Red 4 (Reactive Red 4, cibacron ™ Bright red 3B- ase:Sub>A); rhodamine and derivatives such as 6-carboxy-X-Rhodamine (ROX), 6-carboxy rhodamine (R6G), 4, 7-dichloro rhodamine lissamine (lissamine), rhodamine B sulfonyl chloride, rhodamine (rhodi), rhodamine B, rhodamine 123, rhodamine isothiocyanate X, sulforhodamine B, sulforhodamine 101, sulfonyl chloride derivatives of sulforhodamine 101 (Texas red), N' - tetramethyl-6-carboxy rhodamine (TAMRA), tetramethyl rhodamine and Tetramethyl Rhodamine Isothiocyanate (TRITC); riboflavin: rosolic acid and terbium chelate derivatives; xanthenes; dye conjugated polymers (i.e., polymer- attached dyes), such as fluorescein isothiocyanate-dextran, as well as dyes combining two or more dyes (e.g., in tandem), polymer dyes having one or more monomeric dye units and two or more of the above dyes Mixtures of the various, or combinations thereof.

[0121] The detectable moiety may be selected from a set of spectrally distinct detectable moieties. The spectrally distinct detectable portion comprises a detectable portion having distinguishable emission spectra, even though their emission spectra may overlap. Non-limiting examples of detectable moieties include xanthene derivatives: fluorescein, rhodamine, Oregon green, eosin and texas red; cyanine derivatives: cyanines, indocarbocyanines, oxacarbocyanines, thiacarbocyanines and merocyanines; squaraine derivatives and ring-substituted squaraines, including Seta, seTau. and Square dyes: naphthalene derivatives (dansyl and prodan derivatives); coumarin derivatives; oxadiazole derivatives: pyridmyl oxazoles, nitrobenzooxadiazoles and benzoxadiazoles; anthracene derivative: anthraquinones, including DRAQ5, DRAQ7, and CyTRAK orange; pyrene derivatives; cascades blue; oxazine derivatives: nile red, nile blue, cresyl violet, oxazine 170; acridine derivative: proflavine (proflavin), acridine orange, acridine yellow; arylmethine (arylmethine) derivatives: gold amine, crystal violet, malachite green; and tetrapyrrole derivatives: porphine, phthalocyanine, bilirubin. Other non-limiting examples of detectable moieties include hydroxycoumarin, aminocoumarin, methoxycoumarin, cascade Blue, pacific orange, lucifer yellow, NBD, R-Phycoerythrin (PE), PE-Cy5 conjugate, PE-Cy7 conjugate, red 613, perCP, truRed, fluorX, fluorescein, BODIPY-FL, cy2, cy3B, cy3.5, cy5, cy5.5, cy7, TRIFC, X- rhodamine, lissamine rhodamine B, texas Red, allophycocyanin (APC), APC-Cy7 conjugate, hoechst 33342, DAPI, hoechst 33258, SYTOX Blue, chromomycin A3, mithramycin, YOYO-1, ethidium bromide, acridine orange, TOX green, TOTO-1, TO- PRO: cyanine monomers, thiazole orange, cyTRAK orange, propidium Iodide (PI), LDS 751 , 7-AAD, SYTOX orange, TOTO-3, TO-PRO-3, DRAQ5, DRAQ7, indo-1, fluo-3, fluo-4, DCFH, DHR and SNARE

[0122] In some embodiments, fluorophores of interest can include, but are not limited to, dyes suitable for use in analytical applications (e.g., flow cytometry, imaging, etc.), such as acridine dyes, anthraquinone dyes, and the like, Arylmethane dyes, diarylmethane dyes (e.g., diphenylmethane dyes), chlorophyll-containing dyes, triarylmethane dyes (e.g., triphenylmethane dyes), azo dyes, diazo dyes, nitrodyes, nitroso dyes, phthalocyanine dyes, cyanine dyes, asymmetric cyanine dyes, quinone-imine dyes, oxazine dyes, diaminoazine dyes, safranine dyes, indamine, indophenol dyes, fluoro dyes, oxazine dyes, oxadone dyes, thiazine dyes, thiazole dyes, xanthene dyes, fluorene dyes, pyronine dyes, fluoro dyes, rhodamine dyes, phenanthridine dyes, and dyes combining two or more dyes (e.g., in tandem), and polymer dyes having one or more monomeric dye units, as well as mixtures of two or more of the foregoing dyes. For example, the fluorophore may be 4-acetamido-4 '-isothiocyanatestilbene-2, 2' -disulfonic acid; acridine and derivatives such as acridine, acridine orange, acridine yellow, acridine red and acridine isothiocyanate; allophycocyanin; phycoerythrin; polymethylalginate (peridinin) -chlorophyll protein; 5- (2* -aminoethyl) aminonaphthalene- 1- sulfonic acid (EDANS); 4-amino-N- [ 3- vinylsulfonyl) phenyl JNaphthalimide-3, 5 di sulfonic acid (Lucifer Yellow VS); n- (4-anilino-l- naphthyl) maleimide; anthranilamide (anthranilamide); brilliant Yellow7; coumarin and derivatives such as coumarin, 7-amino-4-methylcoumarin (AMC, coumarin 120), 7-amino-4- trifluoromethylcoumarin (coumarin 151 ); cyanine and its derivatives such as tetrachlorotetrabromofluorescehi, cy3, cy5, cy5.5, and Cy7;4’, 6-diamidino-2-phenylindole (DAPI); 5', 5 "-dibromo-phloroglucinol sulfonephthal ein (bromophthalic trimellitol red); 7- diethylamino-3- (4' -isothiocyanatophenyl) -4-methylcoumarin; diethylaminocoumarin; diethylene triamine pentaacetate; 4,4 '-diisothiocyanidine dihydro-stilbene-2, 2' -disulfonic acid; 4,4 '-diisocyanatostilbene-2, 2' -disulfonic acid; 5- [ dimethylamino ]] Naphthalene- 1 -sulfonyl chloride (DNS, dansyl chloride); 4- (4* -dimethylaminophenylazo) benzoic acid (DABCYL); 4- dimethylaminophenylazo phenyl-4' -isothiocyanate (DABITC); eosin and derivatives such as eosin and eosin isothiocyanate; erythrosine and derivatives, such as ery thro sine B and erythrosine isothiocyanate: ethidium; fluorescein (Lu)And derivatives such as 5-carboxyfluorescein (FAM), 5- (4, 6-dichlorotriazin-2-yl) aminofluorescein (DTAF), 2’7' -dimethoxy-4 5' -dichloro-6- carboxyfluorescein (JOE), fluorescein Isothiocyanate (FITC), chlorotriazinyl fluorescein, naphthofluorescein and qflitc (XRITC); fluorescent amine: IR144; IR1446: green Fluorescent Protein (GFP): reef Coral Fluorescent Protein (RCFP); lissamine1MThe method comprises the steps of carrying out a first treatment on the surface of the Lissamine rhodamine, lucifer yellow; malachite isothiocyanate green; 4-methylumbelliferone; o-cresolphthalein; nitro tyro sine; secondary fuchsin; nile red; Oregon green; phenol red; b-phycoerythrin; o-phthaialdehyde; pyrene and derivatives such as pyrene, pyrene butyrate and succinimidyl 1 -pyrene butyrate: reactive Red 4 (Reactive Red 4, cibacronlMBright red 3B-ase:Sub>A); rhodamine and derivatives such as 6- carboxy-X-Rhod amine (ROX), 6-carboxy rhodamine (R6G), 4, 7-dichloro rhodamine lissamine (lissamine), rhodamine B sulfonyl chloride, rhodamine (rhodid), rhodamine B, rhodamine 123, rhodamine X isothiocyanate, sulforhodamine B, sulforhodamine 101, sulfonyl chloride derivatives of sulforhodamine 101 (tex as red), N’ -tetramethyl-6-carboxy rhodamine (TAMRA), tetramethyl rhodamine, and Tetramethyl Rhodamine Isothiocyanate (TRITC); riboflavin; rosolic acid and terbium chelate derivatives: xanthenes: dye conjugated polymers (i.e., polymer- attached dyes), such as fluorescein isothiocyanate-dextran, as well as dyes that combine two or more of the above dyes (e.g., in tandem), polymer dyes having one or more monomeric dye units, and mixtures of two or more of the above dyes.

[0123] A set of spectrally distinct detectable moieties may. for example, comprise five different fluorophores, five different chromophores, five combinations of fluorophores and chromophores, four different combinations of fluorophores and non-fluorophores, four combinations of chromophores and non-chromophores, or four combinations of fluorophores and chromophores and non-fluorophores. In some embodiments, the detectable moiety may be one of the following spectrally distinct moieties: 1, 2, 3, 4, 5, 6, 7. 8, 9. 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, or a number or range between any two of these values.

[0124] The excitation wavelength of the detectable moiety may vary, for example, as follows or about: 10 nm, 20 nm, 30 nm, 40 nm, 50 nm, 60 nm, 70 nm, 80 nm, 90 nm, 100 nm, 110 nm, 120 nm, 130 nm, 140 nm, 150 nm, 160 nm, 170 nm, 180 nm, 190 nm, 200 nm, 210 nm, 220 nm, 230 nm, 240 nm, 250 nm, 260 nm, 270 nm, 280 nm, 290 nm, 300 nm, 310 nm, 320 nm, 330 nm, 340 nm, 350 nm. 360 nm, 370 nm, 380 nm, 390 nm, 400 nm, 410 nm, 420 nm. 430 nm, 440 nm, 450 nm, 460 nm, 470 nm, 480 nm, 490 nm, 500 nm, 510 nm 520 nm, 530 nm, 540 nm, 550 nm, 560 nm, 570 nm, 580 nm, 590 nm. 600 nm, 610 nm, 620 nm, 630 nm, 640 nm, 650 nm, 660 nm, 670 nm, 680 nm, 690 nm, 700 nm, 710 nm, 720 nm, 730 nm, 740 nm, 750 nm, 760 nm, 770 nm, 780 nm, 790 nm. 800 nm, 810 nm, 820 nm, 830 nm, 840 nm, 850 nm, 860 nm. 870 nm, 880 nm, 890 nm, 900 nm, 910 nm, 920 nm, 930 nm, 940 nm, 950 nm, 960 nm, 970 nm, 980 nm, 990 nm, 1000 nm, or a number or range between any two of these values. The emission wavelength of the detectable moiety may also vary, for example, as follows or about: 10 nm, 20 nm, 30 nm, 40 nm, 50 nm, 60 nm, 70 nm, 80 nm, 90 nm, 100 nm, 110 nm, 120 nm, 130 nm, 140 nm, 150 nm, 160 nm, 170 nm, 180 nm, 190 nm, 200 nm, 210 nm, 220 nm, 230 nm, 240 nm, 250 nm, 260 nm, 270 nm, 280 nm, 290 nm, 300 nm, 310 nm, 320 nm, 330 nm, 340 nm, 350 nm, 360 nm, 370 nm, 380 nm, 390 nm, 400 nm, 410 nm. 420 nm, 430 nm, 440 nm, 450 nm, 460 nm, 470 nm, 480 nm, 490 nm, 500 nm, 510 nm 520 nm, 530 nm, 540 nm, 550 nm, 560 nm, 570 nm, 580 nm, 590 nm, 600 nm, 610 nm. 620 nm, 630 nm, 640 nm, 650 nm, 660 nm, 670 nm, 680 nm. 690 nm, 700 nm, 710 nm, 720 nm, 730 nm, 740 nm, 750 nm, 760 nm, 770 nm, 780 nm, 790 nm, 800 nm, 810 nm, 820 nm, 830 nm, 840 nm, 850 nm, 860 nm, 870 nm, 880 nm, 890 nm, 900 nm, 910 nm, 920 nm, 930 nm, 940 nm, 950 nm, 960 nm, 970 nm, 980 nm, 990 nm, 1000 nm, or a number or range between any two of these values.

[0125] In one embodiment, the temporal labelling is effected in vivo. In this embodiment, the predetermined location may be in the blood circulation. Thus, the different labels may be introduced into the animal by intravenous injection. Other types of administration are also contemplated depending on the predetermined location. Thus, for example the temporal labelling may be effected by any one of the following administration routes - oral, rectal, transmucosal, transnasal, intestinal or parenteral delivery, including intramuscular, subcutaneous and intramedullary injections as well as intrathecal, direct intraventricular, intracardiac, e.g., into the right or left ventricular cavity, into the common coronary artery, intravenous, intraperitoneal, intranasal, or intraocular injections.

[0126] It will be appreciated that the method can be effected by temporally labelling cells at more than one predetermined location, wherein the labeling correlates both with the time at which it is labeled and the predetermined location.

[0127] In one embodiment, the time between the at least two labelings is at least 2 hours, 6 hours, 12 hours, 24 hours, 48 hours . In another embodiment, the time between the at least two labelings does not exceed 2 days, 3 days, 4 days, 5 days, 6 days or longer. A sufficient time is waited until at least 50 %, 60 %, 70 %, 80 % 90 % of the labeled cells reach their target location (e.g. 6 hours, 12 hours, 24 hours, 48 hours from the last labeling). The temporal labelling at the predetermined location is such that only cells at the predetermined location are labeled and those in the target location are not labelled. Thus, if the identical cell type is present at the target location at the time of temporal labeling at the predetermined location, those cells are shielded from the labeling agent and are not amenable to labeling. The target location is not the predetermined location. In one embodiment, the target location is a tumor microenvironment. In another embodiment, the predetermined location is blood circulation and the target location is tumor microenvironment. In another embodiment, the predetermined location is an ex vivo cell population and the target location is a location in the body (e.g. a transplant of cells).

[0128] At least one part of the method may be carried out in an animal - e.g. mammal such as mouse, rat, rabbit. In one embodiment, at least one part of the method is carried out in a human.

[0129] Cell analysis is carried out at a time by which at least two populations of differentially temporally labeled cells have reached the target location. In one embodiment, cell analysis is carried out at a time by which at least one of the populations of the later tagged cells have been in the target location for at least 2, 4, 6, 8, 12, 24 hours.

[0130] Analysis of cells may be carried out in vivo. For example, the analysis may be carried out in vivo by imaging.

[0131] Analysis of cells may alternatively be carried out ex vivo. Cells may be removed from the body and then analyzed.

[0132] In one embodiment, the cells are analyzed at the single cell level.

[0133] Methods currently used for single cell isolation include: Dielectrophoretic digital sorting, enzymatic digestion, FACS, hydrodynamic traps, laser capture microdissection, manual picking, microfluidics, micromanipulation, serial dilution, and Raman tweezers.

[0134] Methods of analyzing cells (e.g. on the single cell level) include, but are not limited to transcriptome analysis, genomic analysis, proteomic analysis, histological analysis, microscopic analysis, metabolic analysis.

[0135] In one embodiment, the RNA of the cells is analyzed. General methods for RNA extraction are known in the art. RNA may be extracted from paraffin embedded tissues. RNA may be extracted from cultured cells and tissue samples using a commercial purification kit according to the manufacturer's instructions, e.g., using Qiagen RNeasy mini-columns, Master Pure™, Complete DNA Kit, EPICENTRE. RTM. RNA Purification Kit, and Ambion, Inc., Paraffin Block RNA Isolation Kit, Tel-Test RNA Stat-60. In certain embodiments, the extracted RNA is an RNA sample or an isolated RNA sample.

[0136] In another embodiment, analysis of RNA of cells is affected by converting RNA to cDNA.

[0137] Single cell transcriptome analysis

[0138] The process of single cell RNA sequencing is known in the art, and there are numerous notable methods which differ from one another in at least one of the following aspects: (i) cell isolation; (ii) cell lysis; (iii) reverse transcription; (iv) amplification; (v) transcript coverage; (vi) strand specificity; and (vii) UMI (unique molecular identifiers or tags that can be applied for the detection and quantification of unique transcripts). Another main point of compari son between the different methods is the coverage of the produced RNA transcript, whether it is a full length or nearly full-length transcript, a transcript corresponding to only the 3 ’-end, or the 5 ’-end. Acceptable methods for the production of a full-length RNA transcript include, but are not limited to the following methods: Tang, Quartz-seq, SUPeR-seq, Smart-seq, Smart-seq2, MATQ-seq. Methods for the production of a 3 ’-end include but are not limited to CEL-seq, CEL-seq2, MARS- seq, MARS-seq2, InDrop, Drop-seq, SPLiT-seq, Seq-Well, sci-RNA-seq, Quart-seq2, Chromium, Cytoseq, STRT-seq and STRT / C1. Methods for the production of a 5 ’-end include but are not limited to, Chromium and DroNUC-seq. Compared to 3'-end or 5’-end counting protocols, full- length scRNA-seq methods have incomparable advantages in isoform usage analysis, allelic expression detection, and RNA editing identification due to their improved transcript coverage.

[0139] Notably, droplet-based technologies (e.g., Drop-seq, InDrop and Chromium) can generally provide a lager throughput of cells and a lower sequencing cost per cell compared to wholetranscript scRNA-seq. Thus, droplet-based protocols are suitable for generating huge amounts of cells to identify the cell subpopulations of complex tissues or tumor samples. Several scRNA-seq technologies can capture both polyA+ and poly A- RNAs, such as SUPeR-seq and MATQ-seq. These protocols are useful for sequencing long noncoding RNAs (IncRNAs) and circular RNAs (circRNAs). Compared to traditional bulk RNA-seq technologies, scRNA-seq protocols suffer higher technical variations. In order to estimate the technical variances among different cells, spikeins (such as External RNA Control Consortium (ERCC) controls) and UMIs have been widely used in corresponding scRNA-seq methods. The RNA spike-ins are RNA transcripts (with known sequences and quantity) that are applied to calibrate the measurements of RNA hybridization assays, such as RNA-Seq, and UMIs can theoretically enable the estimation of absolute molecular counts. Notably, ERCC and UMIs are not applicable to all scRNA-seq technologies due to the inherent protocol differences. Spike-ins are used in approaches like Smart-seq2 and SUPeR-seq but are not compatible with droplet-based methods, whereas UMIs are typically applied to 3 '-end sequencing technologies (such as Drop-seq, InDrop and MARS-seq).

[0140] Additional methods of single cell transcriptome analysis are described in PCT Publication No. W02014 / 108850, WO 2018 / 222548, WO 2020 / 180778, WO WO2023199311, Patent Application No. 20100203597 and US Patent Application No. 20180100201, US Patent Application No. 2021 / 0047638 the contents of which are incorporated herein by reference.

[0141] Kits for analyzing cell fate over a course of time are contemplated.

[0142] Thus, according to another aspect of the invention there is provided a kit for analyzing cell fate over a course of time.

[0143] The kit comprises: a first pair of affinity agents which comprise:

[0144] (i) a first cell affinity agent which binds to an outer surface of a specific cell, the first cell affinity agent comprising a first antigenic determinant:

[0145] (ii) a second cell affinity agent which binds to an outer surface of said specific cell, the second cell affinity comprising a second antigenic determinant; and a second pair of affinity agents which comprise:

[0146] (iii) a third affinity agent which binds specifically to said first antigenic determinant, said third cell affinity agent comprising a first nucleic acid barcode; and

[0147] (iv) a fourth affinity agent which binds specifically to said second antigenic determinant, said third cell affinity agent comprising a second nucleic acid barcode

[0148] Typically, the kit does not comprise more than ten pairs of affinity agents, such that the kit is limited to analyzing no more than ten different times in a single experiment.

[0149] In another embodiment, the kit does not comprise more than nine pairs of affinity agents, such that the kit is limited to analyzing no more than nine different times in a single experiment.

[0150] In another embodiment, the kit does not comprise more than eight pairs of affinity agents, such that the kit is limited to analyzing no more than eight different times in a single experiment.

[0151] In another embodiment, the kit does not comprise more than seven pairs of affinity agents, such that the kit is limited to analyzing no more than seven different times in a single experiment.

[0152] In another embodiment, the kit does not comprise more than six pairs of affinity agents, such that the kit is limited to analyzing no more than six different times in a single experiment.

[0153] In another embodiment, the kit does not comprise more than five pairs of affinity agents, such that the kit is limited to analyzing no more than five different times in a single experiment. In another embodiment, the kit does not comprise more than four pairs of affinity agents, such that the kit is limited to analyzing no more than four different times in a single experiment.

[0154] In another embodiment, the kit does not comprise more than three pairs of affinity agents, such that the kit is limited to analyzing no more than three different times in a single experiment.

[0155] In another embodiment, the kit does not comprise more than two pairs of affinity agents, such that the kit is limited to analyzing no more than two different times in a single experiment.

[0156] Examples of cell affinity agents are provided herein above.

[0157] In a particular embodiment, each of the affinity agents of the kit are antibodies.

[0158] In another embodiment, at least two of the cell affinity agents of the kit target that identical cell type.

[0159] The antigenic determinants comprised in the cell affinity agents include any moiety on the cell affinity agent that can be specifically recognized by another affinity agent.

[0160] In one embodiment, the moiety is a fluorescent moiety, including for example an antibody that binds specifically to Allophycocyanin (APC) - anti-APC, an antibody that binds specifically to Phycoerythrin (PE) - anti-PE, and an antibody that binds specifically to fluorescein (FITC) - anti-FITC (clone FIT-22).

[0161] The kit may comprise other agents which can be used to carry out single cell sequencing. Such reagents include for example T4 RNA ligase, RNAseH, DNase, sequencing adaptors and / or a reverse transcriptase.

[0162] Preferably, each of these components are packaged in separate packaging. The containers of the kits will generally include at least one vial, test tube, flask, bottle, syringe or other containers, into which a component may be placed, and preferably, suitably aliquoted. Where there is more than one component in the kit, the kit also will generally contain a second, third or other additional container into which the additional components may be separately placed. However, various combinations of components may be comprised in a container. When the components of the kit are provided in one or more liquid solutions, the liquid solution can be an aqueous solution. However, the components of the kit may be provided as dried powder(s). When reagents and / or components are provided as a dry powder, the powder can be reconstituted by the addition of a suitable solvent.

[0163] A kit will preferably include instructions for employing, the kit components as well the use of any other reagent not included in the kit. Instructions may include variations that can be implemented.

[0164] As used herein the term “about” refers to ± 10 % The terms "comprises", "comprising", "includes", "including", “having” and their conjugates mean "including but not limited to".

[0165] The term “consisting of’ means “including and limited to”.

[0166] The term "consisting essentially of" means that the composition, method or structure may include additional ingredients, steps and / or parts, but only if the additional ingredients, steps and / or parts do not materially alter the basic and novel characteristics of the claimed composition, method or structure.

[0167] As used herein, the singular form "a", "an" and "the" include plural references unless the context clearly dictates otherwise. For example, the term "a compound" or "at least one compound" may include a plurality of compounds, including mixtures thereof.

[0168] Throughout this application, various embodiments of this invention may be presented in a range format. It should be understood that the description in range format is merely for convenience and brevity and should not be construed as an inflexible limitation on the scope of the invention. Accordingly, the description of a range should be considered to have specifically disclosed all the possible subranges as well as individual numerical values within that range. For example, description of a range such as from 1 to 6 should be considered to have specifically disclosed subranges such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6 etc., as well as individual numbers within that range, for example, 1, 2, 3, 4, 5, and 6. This applies regardless of the breadth of the range.

[0169] Whenever a numerical range is indicated herein, it is meant to include any cited numeral (fractional or integral) within the indicated range. The phrases “ranging / ranges between” a first indicate number and a second indicate number and “ranging / ranges from” a first indicate number “to” a second indicate number are used herein interchangeably and are meant to include the first and second indicated numbers and all the fractional and integral numerals therebetween.

[0170] As used herein the term "method" refers to manners, means, techniques and procedures for accomplishing a given task including, but not limited to, those manners, means, techniques and procedures either known to, or readily developed from known manners, means, techniques and procedures by practitioners of the chemical, pharmacological, biological, biochemical and medical arts.

[0171] When reference is made to particular sequence listings, such reference is to be understood to also encompass sequences that substantially correspond to its complementary sequence as including minor sequence variations, resulting from, e.g., sequencing errors, cloning errors, or other alterations resulting in base substitution, base deletion or base addition, provided that the frequency of such variations is less than 1 in 50 nucleotides, alternatively, less than 1 in 100 nucleotides, alternatively, less than 1 in 200 nucleotides, alternatively, less than 1 in 500 nucleotides, alternatively, less than 1 in 1000 nucleotides, alternatively, less than 1 in 5,000 nucleotides, alternatively, less than 1 in 10,000 nucleotides.

[0172] It is appreciated that certain features of the invention, which are, for clarity, described in the context of separate embodiments, may also be provided in combination in a single embodiment. Conversely, various features of the invention, which are, for brevity, described in the context of a single embodiment, may also be provided separately or in any suitable subcombination or as suitable in any other described embodiment of the invention. Certain features described in the context of various embodiments are not to be considered essential features of those embodiments, unless the embodiment is inoperative without those elements.

[0173] Various embodiments and aspects of the present invention as delineated hereinabove and as claimed in the claims section below find experimental support in the following examples.

[0174] EXAMPLES

[0175] Reference is now made to the following examples, which together with the above descriptions illustrate some embodiments of the invention in a non-limiting fashion.

[0176] EXAMPLE 1

[0177] MATERIALS AND METHODS

[0178] Cell lines

[0179] GL261 mouse glioma cells (kindly provided by the laboratory of T. Weiss, University Hospital Zurich)87were grown in Dulbecco’s Modified Eagle Medium (DMEM, Gibco, 41965- 039) at 37° C with 5% CO2, supplemented with 1% penicillin- streptomycin (Biological Industries, 03-031-1B), 2mM L-glutamine (Sartorius, 03-020-1B), ImM sodium pyruvate (Sartorius, 03-042- 1B) and 10% heat inactivated fetal bovine serum (GE, Cat# SH30071.03).

[0180] Mouse primary bone marrow derived macrophages (BMDMs)

[0181] Primary bone marrow cells from female C57BL / 6 mouse were cultured with C10 medium containing RPMI medium supplemented with 10% FBS (GE, Cat# SH30071.03), ImM sodium pyruvate (Sartorius, Cat# 03-042-1B), 1% x100 non-essential amino acids (Biological Industries, Cat# 013401B), 10mM HEPES buffer, 2mM L-glutamine (Sartorius, Cat# 030201B) and 50μM b-mercaptoethanol (Gibco, Cat# 35310-010), 30 ng / ml MCSF (Peprotech, Cat# 300-25-100), maintained at 37°C and 5% CO2.

[0182] Mouse models

[0183] For all Zman-seq experiments female (8-13 weeks old) wildtype mice (C57BL / 6) (Harlan) were housed in the Weizmann Institute animal facility under pathogen-free conditions and 12-h light / 12-h dark cycle. Food and water was provided ad libitum. For antibody decay measurements C57BL / 6 Ub-GFP transgenic mice (Strain #004353) and B6.SJL mice (Strain #033076) were used. All experiments were conducted as approved by the Institutional Animal Care and Use Committee. Mice were monitored daily for 3 days after tumor implantation. Tumor bearing mice were sacrificed before (<14 days post implantation) the development of symptoms for scRNAseq experiments.

[0184] METHOD DETAILS

[0185] Animal treatments

[0186] Glioma-bearing mice were treated intraperitoneally (i.p.) with aTREM2 antibody (clone 178; Fc-mutated recombinant mAb, 200pg / mouse) on day 2 and 740. Anti-hILTI (clone 135.5; Fc- mutated recombinant mAb specific for human ILT1, a receptor absent in mice) was used as a control (200 pg / mouse).

[0187] Brain tumor implantation

[0188] Eleven-week-old female mice were anesthetized with isoflurane (Isoflurane, USP Terrell, Piramal) (3-4% induction, 1.5-2% maintenance), received local 2 mg / kg lidocaine (Vetmarket 2%, 162097) and systemic 0.05 mg / kg buprenorphine (Bupaq, 0.3mg / ml, Richter Pharma) analgesia. Mice were intracerebrally injected with 2xl04GL261 cells in 2p 1 PBS in the right anterior striatum (coordinates from Bregma: +0.5 mm antero-posterior, +2.2 mm medio-lateral, -3.5 mm dorso- ventral). Mice were kept on a heating pad during the procedure and until mice were awake and mobile. After surgery mice were monitored and treated with meloxicam 5mg / kg (Loxicom, 5 mg / ml, Norbrook) subcutaneously daily, for three days.

[0189] In vivo time stamping

[0190] Mice were briefly anesthetized with isoflurane (3-4%) and were injected with 100 μl of fluorescent antibody solution intravenously (containing 1.2 or 4 pg antibody corresponding to 60 pg / kg or 200 pg / kg body weight (bw) respectively). Time stamping was done with CD45 antibodies, clone 30-F11 with following fluorophores: PE (Biolegend, Cat# 103106), BB515 (BD Biosciences, Cat# 564590), BUV737 (BD Biosciences, Cat# 748371), BV711 (BD Biosciences, Cat# 563709). We consistently observed a broad range of fluorescence intensities of PBMC cells for all fluorophores, already 15 minutes after in vivo CD45-antibody injection.

[0191] In order to calibrate antibody concentrations for in vivo timestamping we injected 20, 60 and 200 ug / kg bw CD45-PE labeled antibodies and analyzed the PBMCs 15 minutes later via FACS (Figure 2B).

[0192] We established at what rate injected CD45 antibodies become unavailable for binding CD45 in the blood by sampling the blood (50 μl) of mice 5, 15, 30, 60, and 90 minutes after injecting 200 ug / kg bw CD45-PE. We indirectly measured the amount of unbound CD45-PE antibody by mixing the samples with 50 μl blood from a GFP-mouse for 15 minutes, followed by washing and measuring the PE emission of GFP+cells via FACS (Figure 2C).

[0193] Estimating the in vivo labeling decay

[0194] The decay of anti-CD45-PE antibody from stained cells was measured by transferring CD45-PE stained CD45.1+ cells (2xl06in 100 μl PBS) from congenic mice (B6.SJL) to CD45.2 mice. Three mice received CD45-PE stained CD45.1 cells, one control mouse received CD45.1 cells without PE stain, one mouse did not receive cell transfer. CD45.1 PBMCs were isolated from the blood of 5 female mice with Ficoll density gradient and were stained for 20 minutes on 4°C with CD45-PE antibody (Biolegend, Cat# 103106, 1:100). At 4, 8, 24, 48, 96, 144 hours after cell transfer, 100 μl of blood was drawn. Red blood cells were lysed with RBC lysis solution (Sigma, Cat# R7757) washed with FACS buffer and stained with antibodies against CD45.1 (Biolegend, Alexa 488, clone A20, Cat# 110717), CDl lb (Biolegend, BV605, clone MI / 70, Cat#101237), CD19 (Biolegend, PE / Cy7, clone 6D5, Cat# 313655), TCR b chain (Biolegend, APC / Cy7, clone H57-597, Cat #109219), LY-6c (Biolegend, PerCP / Cy5.5, clone HK1.4, Cat# 128012) and DAPI live / dead stain. Cells were washed with FACS buffer and measured with BD ESR II flow cytometer. Data was analyzed with FlowJo and R.

[0195] Flow cytometry analysis of tumor infiltrating and resident CD45+ cells

[0196] We tested for how long in vivo stained leukocytes can be detected in the tumor by injecting mice with distinct fluorescent CD45 antibodies 60, 48, 36 and 24 hours (PE, BV711, BB515, BUV737 respectively) before mice were sacrificed (Figure 2E). Mice were transcardially perfused with PBS to wash out the blood. Tumors were dissociated into single cell suspensions, stained for CD45-APC (BD Biosciences, Cat# 559864) and were recorded with BD Symphony 6 flow cytometer. To test whether microglia gets stained by 48, 36, 24 and 12 hour CD45 time stamping, tumor bearing hemispheres were dissociated into single cells and stained 1:100 with CD45-BV421 (BD Biosciences, #Cat 563890), CDl lb-BV605 (Biolegend, Cat# 101237), P2RY12-Alexa647 (Biolegend, Cat# 848017), F4 / 80-APC / eFluor780 (eBioscience, Cat# 47480182), EY6C- PerCP / Cy5.5 (Biolegend, #Cat 128012) (Figure 2F) and recorded with BD Symphony 6. Data was analyzed with Flowjo and R.

[0197] Single cell isolation and sorting for RNA sequencing

[0198] Tumor-bearing mice were sacrificed 13 days after tumor inoculation. Mice were deeply anaesthetized by intraperitoneal injection of ketamin 10 mg / ml (Ketavet Veterinary, 10Omg / ml, Zoetis) and xylazine 2 mg / ml (Sedaxylan 20mg / ml, Eurovet). Mice were transcardially perfused with 10 ml of ice cold PBS. After brain extraction, tumors were dissected to reduce the amount non-tumorous brain tissue in downstream analyses. Samples were blade-minced on ice and transferred to 1.3 ml digestion buffer (0.0125 mg / ml dNase type I (Roche) and 1 mg / ml collagenase IV (Worthington) in RPMI-1640 (Gibco, 21875-034). Samples were incubated for 20 min at 37 °C, during this incubation samples were mechanically dissociated with syringes (gauge 26). Cells were then transferred to ice and filtered through a 100-pm cell strainer, washed with ice-cold FACS buffer (0.5% BSA, 2 mMEDTA in PBS), and centrifuged (5 min, 4 °C, 350g). Supernatant was aspirated and samples were resuspended in 37% Percoll (Sigma, Cat# GE17089101) density gradient, followed by centrifugation at 800 g, 30 min at 4 °C. The cell pellet was washed PBS, followed by Fc-blocking (1:200) (TruStain FcX, Clone 93, BioLegend) and staining with CD45-APC (30-F11, BD Biosciences). DAPI 0.1 pg / ml (BioLegend) was used for live / dead stain. Cells were filtered through a 70 pm cell strainer. Cell sorting was done on BD FACS Symphony 6 flow cytometer (BD Biosciences). Dead cells and doublets were excluded by gating, after which CD45+cells were gated based on APC fluorescence, followed by gating for the in vivo CD45 stains. Single cells were sorted into 384-well capture plates either by the ex vivo CD45-APC stain, or by sorting cells harboring any of the in vivo CD45 stains. Capture plates contained 100 nl of lysis solution, 3 μl of mineral oil and 20 nM barcoded poly(T) reverse transcription primers for scRNA-seq. After sorting, plates were first centrifuged and then snap- frozen on dry ice and stored at -80 °C. Cells were analyzed using BD FACSDiva software (BD Bioscience) and FlowJo software (FlowJo LLC).

[0199] Colon, lung and blood processing for Zman-seq

[0200] Mice (5-weeks-old) were treated with the Zman-seq (n = 3, +1 non- stained control) protocol with timestamping antibody injections 48 (CD45-PE), 36 (CD45-BB515), 24 (CD45-BUV737), 12 (CD45-BV711) hours prior to tissue harvesting. Blood was sampled from the heart, then mice were transcardially perfused with PBS. Blood was washed and incubated with 1 ml lx ACK lysis buffer (Thermo Fisher, Cat# A1049201) for 5 min at RT. After washing and centrifugation, leukocytes were resuspended in FACS buffer and kept on ice until staining.

[0201] Lungs from perfused mice were cut into small pieces and digested for 60 minutes at 37°C with digestion buffer containing collagenase IV (final 0.2 mg / ml) and DNase I (final 0.05 mg / ml) in RPMI-1640 medium supplemented with 10% FBS. Next, tissue pieces were dissociated with syringes. Cell suspensions were passed through a 70 pm strainer, washed and centrifuged. Cells were resuspended in FACS buffer and kept on ice until staining.

[0202] Colon was opened longitudinally, cleaned, cut into 0.5 cm long sections and washed with PBS. Colon tissues were incubated in PBS with 5 mM EDTA and 2 mM DTT at 37°C for 20 min to detach epithelial cells. Colon pieces were vigorously shaken up and down for 15 seconds by hand. The remaining tissue was washed with PBS twice to remove EDTA. Next, tissue was then cut into small pieces and was added to the digestion solution and incubated at 37°C for 60 minutes. Digested tissues were dissociated with syringes, passed through a 70 pm strainer, washed and centrifuged. Cells were resuspended in FACS buffer and kept on ice until staining. Leukocytes were stained with CD45-APC stain and DAPI. CD45-APC+cells were sorted with BD FACS Symphony 6 sorter into barcoded capture plates for scRNAseq.

[0203] Single cell library preparation

[0204] The scRNA-seq libraries were created using a revised version of the massively parallel scRNAseq technique91,92. The process involved capturing polyadenylated mRNA from individual cells that had been sorted into 384-well plates, and then barcoding the mRNA during the reverse transcription into cDNA. The resulting cDNA was then pooled for each plate, and underwent fragmentation and amplification to generate sequencing -ready libraries for Illumina sequencing. Quality and DNA concentration tests were performed on each library created from a plate.

[0205] Read alignment

[0206] The scRNA-seq libraries were pooled at equimolar concentrations and sequenced on an Illumina NovaSeq 6000 sequencer with a sequencing depth ranging from 10k to 50k reads per cell. Reads were condensed into original molecules by counting same unique molecular identifiers (UMI). We ensure that the batches for analysis showed a low level cross single-cell contamination (less than 3%) by statistics on the detected spurious UMI in empty wells. Alignment of reads was done using the MARS-seq2.0 pipeline. In short reads were filtered for low quality reads, subsequently mapped to the mouse reference genome mm10 using HIS AT (version 0.1.6), excluding reads with multiple mapping positions91. The UCSC genome browser was used as a reference to assign exonic reads to genes. Cell UMI uniqueness was tested for 3kb aria. In cases where exons of different genes shared a genomic position on the same strand, reads were considered as a single gene with a concatenated gene symbol.

[0207] Time bin assignment

[0208] To assign a tumor exposure time bin to each cell, we classified the fluorescent FACS signal of each time stamp into stained or non-stained. As the fluorescence signal is variable in each cell due to their diversity in size and cell types as well as autofluorescence, the fluorescence cutoffs could not be generalized. To this end, we additionally sorted CD45+ unstained cells to build our classification model. For each fluorescence time stamp, we trained a second order individual generalized linear model (GLM) with forward scatter (FSC), side scatter (SSC), and DAPI channels as model predictors. Our assumption is that the forward and sideward scatters account for the diverse cell types and ensure that we have an unbiased GLM model across different cell types. Formally, let each fluorescence channel be denoted as Yiand x = [xi... , xpbe the vector predictor variables (where p = 7 for DAPI, three forward scatter channels and sideward scatter channels respectively). We define the generalized linear model of second order interactions with gaussian link as: where e~lV(0,l) is the independent identically distributed normal error. We check that the residuals of the model are normally distributed by assessing plotting the residuals against the fitted values as well as Q-Q plots. Then cells with measured fluorescent values of p < 0.001 were considered as stained. Finally, we assign a tumor exposure time bin to each cell based on the positive antibody stain, which in case of multiple positive stains was determined by the last stain the cell was exposed to in the circulation.

[0209] We also developed an approach for feeding the GLM with unstained FACS events without index sorting. For this, we first manually defined the cutoff for unstained cells in stained samples. Then we applied the same generalized linear model on unstained samples falling into the manually defined cutoff.

[0210] MetaCell Construction

[0211] We used the R package MetaCell20to analyze the single-cell data in the paper. We first removed specific mitochondrial genes, immunoglobulin genes, ribosomal genes, and genes linked with poorly supported transcriptional models (such as those annotated with the suffix “Rik” and so on). We also discarded cells with less than 300 UMIs from the subsequent analysis. Informative genes with high dispersion were selected using the variance-to-mean parameter T_vm > 0.1 and minimum total UMI count > 100. To construct the metacells, we used standard parameters of K = 100 and 750 bootstrap iterations to resample 75% of the cells in each iteration to ensure homogeneity within each metacell. Cells across treatment conditions were combined for metacell construction in the treatment data (Figures 5A-F) for global characterization. Metacells were then manually annotated based on analysis of marker genes and known cell type markers.

[0212] In the initial clustering of the metacells, we identified the major clusters of myeloid and lymphoid cells, as well as doublet contaminations. We then removed the doublet metacells and manually annotated the clustering of the myeloid and lymphoid clusters separately. We combined the clusters if there were no significant difference in their marker genes and annotated them with known cell type markers (e.g. Clec9a - cDCl, Clqa - TAMs, Argl - Argl TAM, Ace and Chil3 - Monocytes, Foxp3 - Treg, Prfl and granzymes - NKs, and so on). The 2D projection of metacells is computed using a force-directed layout algorithm on the regularized similarity metacell graph GM as described in the MetaCell algorithm20. For the single cell projection, first we compute a raw similarity matrix using Pearson’s correlation and then construct a weighted adjacency matrix to define a directed cell graph G. Cells are then positioned by taking the average metacell coordinates of filtered neighbor cells from G.

[0213] MetaCell Time Trajectory Construction

[0214] To estimate a continuous tumor exposure time (cTET) profile for each metacell, we used the distribution of each time point within each metacell to calculate the corresponding cumulative distribution (CDF) and area under curve which we define as cTET (top panel of Figure 4B). Specifically, let us denote a distribution of timepoints TPi where i e {0, 12, 24, 36, 48} time bins. For each metacell, the CDF of P(X ≤ i is defined as the normalized proportion of each time point TPi. Then we estimate the cTET values as curve (AUC) of the CDF. Since earlier time bins correspond to a larger AUC, we renormalize the cTET such that cTET c [0,1] and that smaller cTET indicates an earlier time bin.

[0215] We then refine the cTET profile to construct the trajectory of cell differentiation along tumor exposure time (bottom panel of Figure 4B). First, we construct a base reference trajectory Tr= that shows the transitioning of cTET by taking the average cTET in each of the annotated cell type cluster. The trajectory is ordered by the average cTET and we plot the arrows according to the cTET values. Subsequently, we performed a correlation-based unsupervised clustering on the top informative cell type specific genes for k ∈ [3 ... N ] as described previously93. Briefly, we used the dispersion ratio to select the top informative genes and performed a hierarchical clustering on the normalized Spearman correlation distance matrix to obtain unbiased cell type clusters. Then for each k, we construct a clustering trajectory Tkby ordering the cluster average cTET. We define the refined cTET profile by the combined trajectory as follows:

[0216] Temporally-resolved Trajectory Analysis

[0217] We utilized the refined cTET trajectory to delve deeper into the molecular changes across time in the transitioning of cell types (e.g. chemotactic NKs to dysfunctional NKs). We used the Spearman’s rank correlation to first identify top genes which are significantly correlated with cTET (p < 0.05). The expression of the top genes were then smoothed across cTET using loess fitting and each gene was re-normalized to [0, 1] across all metacells. Groups of up-regulated, down-regulated, and transiently-regulated gene modules were then visualized as heatmaps by clustering genes of similar patterns together.

[0218] To further investigate the underlying mechanisms that drive the gene regulations, we infer the TF activity of each metacell, followed by the same correlation analysis and loess smoothing across cTET as mentioned above. The TF activity was estimated using the Dorothea31database with genes of confidence levels of “A” and “B”. The top correlated TF activities are shown as heatmaps showing the shift of TF regulation across time.

[0219] Differential gene analysis:

[0220] We used a pseudo-bulk approach to identify differential genes between the Acp5 TAMs (enriched in the aTREM2 antibody treatment) and Argl TAMs (enriched in the control antibody treatment). The gene counts were first summed within each individual mouse as pseudo-bulk samples, resulting in 3 aTREM2 treated samples and 3 control-treated samples. Then DESeq296was applied to normalize the counts and to perform differential gene analysis. The top genes were visualized in Figure 5D as volcano plot with a threshold of p-value < 0.001 and log2 fold change > 0.5.

[0221] QUANTIFICATION AND STATISTICAL ANALYSIS

[0222] Plots and analyses were generated with R and are detailed in the respective methods sections including the statistical tests used. Statistical details such as error estimation and number of replicates are provided in the respective figure legends.

[0223] Microscopy images were analyzed with open-source ImageJ and CellProfiler. RESULTS Zman-seq resolves in vivo trans criptomic cell states across time

[0224] Cells respond to external stimuli by changing their functional state over time. To dissect how cells are reprogrammed in vivo upon exposure to a new environment, we developed Zman-seq, a technology combining fluorescent temporal tracking of cells with scRNA-seq (Figure 2A). Zman- seq is based on a physical barrier separating two biological compartments such that labeling (pulse) of cells only occurs in one compartment, allowing for cell tracking across compartments (chase). We took advantage of this pulse-chase principle and fluorescently labeled the circulating immune compartment with a temporal sequence of fluorophores in a syngeneic model for GBM (GL261). Leukocytes could be fluorescently labeled while in the vasculature but were shielded from consecutive rounds of labeling once they left the circulation and entered the tissue / tumor. Immune cells within the tumor were profiled using fluorescence-activated cell sorting (FACS), and fluorescent stamps on individual cells were captured using index-sorting to infer the time of tumor infiltration. This enabled us to define the time each immune cell spent in the tumor - termed tumor exposure time. Combining tumor exposure time measurements with single-cell transcriptomic profiling, Zman-seq generates temporal maps of the tumor microenvironment.

[0225] In order to establish an in vivo labeling protocol, we titrated the staining of peripheral blood mononuclear cells (PBMC) by intravascular injections of fluorescent anti-CD45 antibodies. Concentrations ranging from 20 to 200 pg / kg body weight (bw) stained the entirety of PBMCs (99.8 % ± 0.26, n = 9), with the mean fluorescent intensity increasing with antibody dose (Figure 2B). To ensure that Zman-seq faithfully measures tumor exposure time between a labeling event in the circulation and harvest in the tumor, it was critical to validate that our fluorescent antibody injections specifically label circulating immune cells and not tumor-resident immune cells. Reasoning that any non-specific labeling of tumor-resident immune cells would require extravasation of antibodies into the tumor and would therefore be highly influenced by plasma antibody concentrations, we first measured the kinetics of unbound antibody levels in the circulation (Figure 2C). We found that the amount of unbound antibody available for staining decays rapidly, dropping below detection level 60 minutes after injection, likely a consequence of high blood CD45 antigen levels and in agreement with published data18. Building on this, we then performed histological analysis on GL261 tumors from mice injected with fluorescent anti-CD45 antibodies 15 minutes (when the unbound plasma- antibody concentration was still sufficiently high) and 24 hours before harvest. This analysis demonstrated significantly more extravascular localization of the 24-hour labeled cells compared to the 15-minute labeled cells: while 84.8% (±5.9% confidence interval, n=3) of the 15 minute signal was intravascular, after 24 hours 80% (±13.6%) of the signal shifted to the extravascular compartment (Figure 2D). Importantly, there was no non-specific labeling of tumor-resident immune cells at the 15 minute time point. We next analyzed the effective time window of Zman-seq labeling by quantifying fluorescently labeled circulating CD45+cells that extravasate to the tumor over a range of labeling time points spanning biologically relevant tumor-immune exposure times. To this end, we injected anti-CD45 antibodies every 12 hours between 60 and 24 hours prior to GL261 tumor harvest using a set of distinct fluorophores for every time point. We detected fluorescently labeled leukocytes in the tumor from all injection time points, demonstrating the applicability of Zman-seq to processes unfolding over multiple days in the tumor (Figure 2E). To evaluate the rate of false-positive Zman-seq labeling we analyzed CD45± cell labeling in the brain after 12, 24, 36 and 48 hour time stamps. We found that microglia, resident brain macrophages which express CD45 and are entirely of yolk sac origin, were not labeled with any of the anti-CD45 time stamp injections highlighting the selectivity of the method to unambiguously label circulating leukocytes (Figure 2F). To further estimate the time window labelled cells can be tracked using Zman-seq, we labeled PBMCs from B6.SJL mice carrying the CD45.1 allele with anti-CD45-PE antibody ex vivo and transferred the cells into CD45.2 recipient mice. Notably, we observed that anti-CD45-PE labeled CD45.1 leukocytes could be reliably detected in the blood across immune cell subsets for 96 hours and longer. Taken together, the rapid decay of unbound anti-CD45 antibody in circulation, combined with the accumulation of sparsely distributed labeled immune cells in the tumor and no off-target labeling of microglia, are consistent with the desired properties of specific and stable high resolution (~30 min) stamping of circulating immune cells.

[0226] Specific detection of labels from each time point enabled grouping of leukocytes into tumor exposure time bins based on their fluorescent profiles recorded during single-cell FACS indexsorting. Tumor exposure time was determined based on the last antibody a cell was exposed to in the circulation. For example, a cell was considered in the 36 hour tumor exposure time bin if it was only labeled with the 36 hour fluorophore but also if it was additionally labeled with the 48- and / or 60- hour fluorophores - meaning that the cell was in circulation and exposed to antibody injections between 60 - 36 hours before it entered the tumor and was shielded from the 24 hour injection. To assess whether our empirical time measurement methodology could be generalized to peripheral organs beyond the central nervous system, we benchmarked Zman-seq in the physiological steady state of colon and lung tissues by injecting 12, 24, 36, and 48 hour anti-CD45 time stamps. The profiled leukocytes showed remarkable organ-specific cell state adaptation compared to circulating leukocytes, highlighting the potential of Zman-seq to investigate time- resolved immune cell adaptation (Figure 2G). We detected robust time stamp signals across all immune cell states except for embryonically-derived alveolar macrophages, tissue-resident intestinal innate-like lymphoid cells and intestinal IgA plasma cells -highlighting the exclusion of time stamps from long-term tissue resident cells of the colon and the lung and thus the broad applicability of Zman-seq to profile cellular dynamics of diverse tissues (Figure 2G). Collectively, we demonstrate that time-stamping using sequential in vivo injections of fluorescently-labeled antibodies facilitates a robust methodology to trace cellular and molecular kinetics of leukocytes across different organs in healthy and diseased states.

[0227] Temporal dynamics of the TME in glioblastoma

[0228] GBM is the most frequent and aggressive lethal primary brain malignancy in adults with a gold standard radio-chemotherapy protocol that has not been significantly improved for two-decades, thus remaining a major unmet need19. The lack of successful immunotherapy approaches in GBM highlights a need to better understand and curtail immunosuppressive differentiation trajectories in the TME. For this purpose, we used Zman-seq (labeling 12, 24, and 36 hour time stamps) and measured single-cell RNA profiles of 10,583 high-quality leukocytes sorted from the TME of the murine syngeneic GL261 glioma model, simultaneously with tumor exposure time stamps. We grouped these cells by their transcriptomes into 139 metacells (MC) containing 55-170 cells / metacell20. We identified a large myeloid compartment consisting of monocytes (Plac8\ Chil3+p TAMs (Trem2+, Argl+) and dendritic cells (DCs) (H2-0a+'). Lymphocytes in the tumor comprised CD4+T helper cells, CD8+cytotoxic T cells, T regulatory cells and natural killer (NK) cells. NK cells could be further subsetted into chemotactic (Slpr5+), cytotoxic / effector (Prfl+Gzma+, Gz.inb+), and dysfunctional (Itgal+Xcll+Eomes-, Pmepal+), with the latter resembling signatures of NK cells exposed to TGF-p21-24. Leukocytes assigned to 12, 24 or 36 hour tumor exposure time bins clustered distinctly, suggesting an association between tumor exposure time and immune cell functional states. Analysis of the distribution of cell types in the 12, 24, and 36 hour tumor exposure time bins corresponded with specific cell states. To better resolve cellular states, we increased the metacell resolution and reclustered the myeloid and lymphoid cell compartments separately. Indeed, chemotactic (SlprS+) and cytotoxic (Prfl+Gzma+, Gz.inb+) NK cells showed high enrichment in early (12 hour) time points, while dysfunctional NK cells (Itgal+Ctla2a\ Gzmc+, Prfl / Gzma / Gzmb-low) were enriched in later (24-36 hour) time points - suggesting that Zman-seq captured biologically plausible and relevant information of the NK response to tumor signals21,22. For instance, Ducimetiere and colleagues described a set of NK clusters: mature / effector (Gz / na\ Zeb2+, Prfl+, Klrgl+cNK_5), immature (Ctla2a+, Ccr2+cNK_4), and TGF-P-imprinted (Gzmc+, Pmepal+, Xcll+, Ctla2a+cNK_6) NK cells found only in immunologically cold tumors, which aligned well with the clusters we defined with distinct temporal kinetics. The myeloid compartment showed a similar association of differentiation and time, with monocytes (Ear2+, Ace+, Chil3+, Plac8+) harboring early (12 hour) time stamps, and tumor associated macrophages (TAMs; ClqaP Trem2\ Argl+) harboring late time stamps, consistent with the expected transition of circulating monocytes to TAMs in the tumor. These cell state changes upon exposure to the TME were associated with extensive molecular dynamics. For instance, in NK cells the expression of the homing receptor Slpr5 decreased with time, while Xcll expression, a chemoattractant for DCs, increased. Similarly, in the myeloid compartment, the expression of Plac8, highly expressed in monocytes24,25, decreased with tumor exposure time, while the expression of the myeloid checkpoint Trem2 increased. In summary, we demonstrate that Zman-seq is capable of resolving fine-grained temporal gene expression trends during assembly of the TME, thereby providing a technology to resolve immune escape mechanisms evoked by the tumor. Molecular trajectories ofNK cell dysfunction in GBM

[0229] A limitation of most single-cell transcriptome atlases is the inability to empirically define cellular trajectories over time. To overcome this limitation in our GBM model, we annotated immune cell states at high resolution and leveraged our temporal labeling to track transcriptional trajectories between states (Figure 3A, 4A). To translate single-cell time stamps to a time axis over which to observe differentiation trajectories, we developed a robust statistical method to assign continuous tumor exposure time (cTET) values on the metacell level (Figure 4B, Methods). For each metacell, we calculated a cumulative distribution function (CDF) based on the frequency of cells in 12, 24, and 36 hour bins (Figure 3B, 4B-C). For each metacell-CDF, the area under the curve (AUC) represents the continuous tumor exposure time. Overlaying the AUC profile on the metacell map elucidated the gradual shift of cell states across time (Figure 3C). We then used the average cTET for each annotated cell cluster (Figure 3A) to define the trajectory underlying these observed cell state transitions (Figure 3D). Next, we temporally ordered metacells by cTET and identified the most significant genes correlating with time (Figure 3E). This approach revealed multiple gene expression changes unfolding at different rates in response to tumor exposure. Genes that were upregulated immediately upon tumor exposure included the TGF-P response genes Pmepal and Srgn, NK immaturity markers Ccr2 and Tcf722, as well as the inhibitory checkpoint receptor Tigit26. Upregulation of TGF-P driven Car2 and Ctla2a, previously shown to hamper NK anti-tumor activity, closely followed, along with Xcll and Itgal22. Gzmc expression, also shown to be TGF-P driven, increased rapidly during the final stage of the trajectory - suggesting that major biomarkers of NK exposure to GBM TME are TGF-P-associated signaling modules and are indicators of the duration of TGF-P exposure. Genes that were immediately downregulated upon tumor exposure include homing receptors (Slpr5, Cx3crl) and NK effector and maturity markers (Klrgl, Cmalf1'29. Expression of inflammatory cytokines and cytotoxic molecules (Ccl3, Gzmb, Gzma, Prfl) decreased more gradually30(Figure 3E).

[0230] To identify the drivers of these transcriptional dynamics, we next estimated the transcription factor (TF) activities of each metacell based on the average gene expression signature of the gene targets of each TF using the dorothea database31(Figure 4D, Methods). This analysis revealed anticorrelated activity over time of SMAD3 and SMAD4, critical downstream mediators of TGF-P signaling. SMAD3 relays TGF-P signaling, leading to reduced cytotoxicity in NK cells30, while SMAD4 counteracts this effect32. Consistently, Zman-seq showed that SMAD4 was initially active for a sustained period counteracting TGF- signaling, followed by a rapid drop at the end of the temporal trajectory that coincided with a spike in SMAD3 activity orchestrating TGF-P signaling and leading to the emergence of the TGF-p imprinted dysfunctional NK cell state. Additionally, the activity of TFs involved in NK differentiation, maturation and interferon-y signaling decreased along the temporal trajectory at distinct rates: decrease in GATA3 activity was gradual, while FOSL2 and GATA2 was faster33’35.

[0231] A core focus in immuno-oncology is to understand interactions in the tumor-immune environment leading to ineffective anti-tumor immunity. Zman-seq provides a critical analytical tool to identify ligands that reprogram immune cells into a tumor-tolerant state upon TME exposure (Figure 4E). To infer putative upstream ligands that may drive the observed immune-escape gene programs in NK cells, we prioritized ligands based on their ability to explain the temporal gene expression changes detected using Zman-seq. Consistent with the TGF-P responsive gene expression module and TF activity dynamics we observed, this analysis revealed TGF-p 1 as a major ligand driving the NK dysfunctional trajectory in the TME (Figure 3F-G, 4F-G). Additional predicted ligands driving NK dynamics included the macrophage-derived chemokine CCL12 and cell-cell contact ligands ICAM1 and APOE (Figure 3F-G, 4F-G). However, TGF-P 1 explained almost exclusively the temporal transition of Itgallowto ItgalhighNK cells with concomitant reduction in cytotoxic activity (Gzmb, Prfl ; Figure 3F). Late dysfunctional NK cells had high Tgfbl expression, though contribution by resident non-immune cells of the microenvironment, such as tumor cells, stromal cells or microglia, also appears likely36(Figure 3F). We next used functional assays on primary human patient samples to ask whether the data showing that TGF-P signaling in the GBM TME inhibits NK cell cytotoxicity translates to human pathology in GMB patients. By co-culturing human glioma stem cells with primary human NK cells in vitro we evaluated degranulation (CD 107a), TNFa and ZFNy expression in response to TGF-P blockade (Figure 4H)37. Consistent with the dynamics we observed using Zman-seq in the mouse model, we found a strong negative association between TGF-P signaling and the production of cytotoxic molecules and inflammatory cytokines in human NK cells (Figure 4H). Taken together, we demonstrate a principled analytical framework for Zman-seq data and demonstrate how it can resolve the complex trajectory of NK cells and the signaling and TF circuitry of the TME, leading to a gradual decline in NK cytotoxic activity in contribution to tumor immune escape.

[0232] Temporal molecular trajectories of the mononuclear phagocyte system in the TME

[0233] Tumor associated macrophages are a major component of the immunosuppressive TME and are key to tumor persistence38. TAMs in GBM originate from infiltrating circulating monocytes and local microglia39. Characterizing the molecular trajectories of the mononuclear phagocyte system in the TME is crucial for better understanding the signals and pathways associated with immune dysfunction in GBM and solid tumors in general. We found a strong correlation between tumor exposure time and Trem2 expression, an immunosuppressive signaling hub for macrophages17. Previous studies have shown that genetic ablation of Trem2 or blocking TREM2 with monoclonal antibodies modifies TAMs and evoke anti-tumor immunity40-42. These studies raise the question of whether blocking TREM2 affects mature Trem2+TAM function or acts earlier in the differentiation of monocytes into TAMs. To address this question and to define monocyte-to-TAM differentiation in the presence and absence of TREM2 signaling, we treated 6 mice with an antagonistic antibody targeting TREM2 (aTREM2) and 6 mice with a control antibody at days 2 and 7 post tumor implantation, similar to Molgora and colleagues40. Eleven days after tumor implantation, leukocytes were time stamped 12, 24, 36, and 48 hours before tissue harvesting. Control mice recapitulated the TME composition observed previously, including activated and dysfunctional NK states, infiltrating monocytes, DCs, and TAMs. The myeloid compartment of mice treated with the aTREM2 antibody contained macrophage cell states highly separated from those in the control antibody-treated mice. In mice treated with the control antibody, Zman-seq revealed a well-defined temporal trajectory of monocyte to TAMs: infiltrating Chil3high Plac8highmonocytes harbored the earliest time stamps and gradually differentiated to Arg7high, Gpnmbbl&bregulatory macrophage populations, described in the context of immunosuppressive myeloid cells in cancer41. The expression of genes promoting immunosuppression in the TME showed distinct temporal patterns along this trajectory. The first changes were related to differentiation from the monocyte state by decreasing the expression of Plac8 and Chil3. In parallel, Sall and Ccl3, genes involved in inflammatory processes43,44, rapidly decayed in monocytes. This was followed by a rapid increase in the expression of macrophage programs and pro-tumorigenic factors such as C3, as well as the pro-angiogenic factor Vegfa and its receptor Fltl45-47. Simultaneously, Pirb (LILRB3), an inhibitory receptor which binds to MHC class I instructing immunosuppressive TAM differentiation48, was transiently upregulated at the monocyte-macrophage junction. This was followed by robust activation of tumor macrophages programs with upregulation of immunosuppressive pathways Argl, Cd.274 (PDL1) and Il18bp24,49, followed by Trem2 and Gpnmb, a hallmark for suppressive Mregs, peaking in the terminal sections of the temporal trajectory .

[0234] Analysis of the TF circuits regulating this myeloid trajectory revealed distinct temporal patterns, including the downregulation of pro-inflammatory and upregulation of immunosuppressive TF circuits. Downregulation of NF-KB associated RELA activity, together with POU2F1 involved in inflammatory macrophage activation, were among the first TF responses upon monocyte tumor exposure50,51. This was followed by loss of STAT2 and IRF9 activity, key nodes in antiviral pathways52,53. The final step of reprogramming toward TAM included upregulation of immunosuppressive STAT3 activity, followed by monotonical upregulation of HIF1A and CREB 1 activity, involved in hypoxic suppressive TAM function and suppressive macrophage polarization, respectively54-56. The terminal portion of our analyzed trajectory was characterized by high NFE2L2 activity, involved in suppressing inflammatory cytokine secretion in macrophages57.

[0235] We next applied these temporal data to analyze the upstream ligands that may drive the differentiation process from infiltrating monocytes to TAMs, potentially representing attractive targets for immunotherapy (Figure 4E). The ligands that best explained the trajectory from monocytes to TAMs included pro-tumorigenic TGF-pl and ANXA1 and pro-inflammatory AP0A1, while chemokines (for example CCE2, CCE3, CCE8) contributed less58-60. The cumulative expression of target genes regulated by TGF-βi and ANXA1 was relatively low in monocytes and increased along the temporal trajectory until its peak in Arg7hlghand GpnmbhighTAMs. The number of time dependent genes affected by a given ligand was highly variable: For instance, TGF-βi or CCE3 and CCE2 contributed to a broad set of target genes, while ANXA1 only affected the upregulation of monocyte recruitment via CCE12, and the downregulation of inflammatory factors Illb and Junb61. Argl , a metabolic immunosuppressive hallmark gene of TAMs differentiation, was influenced by only a few pro-tumorigenic ligands: TGF-pl, IE13 and AGT62,63. To assess whether TGF- pi could indeed orchestrate the observed cell trajectory from circulating monocytes to pro-tumorigenic TAMs, we differentiated murine bone marrow derived myeloid cells in the presence or absence of TGF- i and confirmed the emergence of regulatory macrophage cell states expressing R18bp and Gpnmb in vitro.

[0236] To evaluate to what extent the Zman-seq resolved trajectory of the mononuclear phagocyte compartment in GE261 -bearing mice reflected human pathology in GBM patients, we next analyzed publicly available transcriptomic data of human GBM patients. We tested how genes controlling the myeloid trajectory across time in murine GBM relate to myeloid states found in human GBM patients. Murine genes defining early cell states of the trajectory aligned with monocyte and inflammatory TAM states in humans, while late genes overlapped with regulatory lipid- and phagocytic- TAMs - suggesting robust translation of the murine temporal results to human data. Overall, Zman-seq enabled in vivo characterization of the molecular trajectory of monocyte differentiation to regulatory macrophages, which are a major driver of the immune dysfunction in the GBM TME. Our analysis lays the foundation for improved understanding of myeloid reprogramming in the tumor and identification of effective molecules that antagonize regulatory macrophage differentiation in the TME. Blocking TREM2 signaling redirects the monocyte-to-macrophage trajectory in the TME

[0237] Next, we assessed how TREM2 antagonistic immunotherapy influenced the differentiation of the mononuclear phagocyte system in GBM. Three of the six mice treated with aTREM2 antibody showed marked differences in their respective TME compared to control mice (indicating an immunological response), while three mice showed more subtle perturbation. As expected, Trem2- expressing myeloid cells, such as TAMs and monocyte-derived macrophages, responded most to the TREM2 antagonistic therapy, indicating that the efficacy of the therapy may be attributed to direct effects on Trem2 -expressing cells. We compared myeloid cells of mice treated with aTREM2 to myeloid cells of mice treated with the control antibody, and used Zman-seq to investigate the immune-dynamics of the two treatment groups. aTREM2 treatment resulted in a distinct TAM population compared to the control group, characterized by downregulation of immunosuppressive genes such as Argl, Pirb (LILRB3), Ill8bp, Vegfa and Cd.274 (PDL1) and upregulation of pro-inflammatory genes (Ccl3, Ccl4, Cd81 and Cd83), in line with previous reports40,41,64’65. These two distinct TAM populations developed through a striking bifurcation of monocyte-to-TAM trajectory (Figure 5A-B). We utilized this dataset as a ground truth to perform a detailed analytical benchmarking and compared the performance of computational approaches to our empirical time measurement methodology (summarized in Methods). In short, trajectories predicted by commonly used pseudotime and RNA velocity algorithms showed variable results, strongly depending on the selection of genes. These comparisons highlight the limitations of current computational approaches to resolve cellular trajectories and demonstrate the synergy of combining analytical and empirical time measurements in single cell transcriptomic studies.

[0238] To better understand the mechanism of action of anti-TREM2 immunotherapy, we next investigated the trajectory bifurcation elicited by the treatment. The roots of both the control and aTREM2 trajectories were similar, consisting of infiltrating monocytes that did not express Trem2. In both groups, initial state transitions entailed the downregulation of monocyte transcripts (Chil3, Plac8) (Figure 5C). We observed immediate and late transcriptomic changes distinguishing the control and aTREM2 trajectories. Early changes upon TREM2-inhibition, included the rapid upregulation of pro-inflammatory factors Ccl4, Cd81, and inhibition of early upregulation of immunosuppressive factors seen in the control group (e.g. Pirb, Vegfa)45'41. In monocyte-derived- macrophages, aTREM2 elicited a transient upregulation of an interferon-induced gene module including Ifit2 and lfil366 . This was followed by TAMs upregulating additional pro-inflammatory markers (Ccl3, Acp5, Cd72)64'65'61, and downregulating immunosuppressive genes (Argl, Ill8bp, 117 r). TF activity analysis identified regulators of the proinflammatory trajectory induced by TREM2 inhibition (Figure 5D). Early on, FOSL1, part of the AP-1 TF complex enhancing inflammation by inhibiting Argl transcription68, showed high activity in monocytes and maintained this throughout the trajectory of macrophages treated with aTREM2 antibody. TFs showing high transient activity in monocyte states included components of the NF-KB complex, such as RELA (p65) and NFKB 1 (p50), forming a heterodimer and promoting pro-inflammatory transcriptional programs69. ATF2, indicated in inflammatory macrophages70, showed a similar early transient activity as NFKB1. Slightly later, in the monocyte-macrophage junction, proinflammatory STAT2 and IRF9 involved in interferon signaling, showed a transient activity peak. In parallel aTREM2 treatment downregulated the immunosuppressive activities of STAT3 and STAT652,71. In the final TAM program, aTREM2 treatment induced pro-inflammatory NFKB2 activity (p52)69and inhibited the upregulation of immunosuppressive HIF1A and NFE2L2 activity54,57.

[0239] Differential gene expression analysis of TAMs during TREM2 inhibition revealed differentially abundant transcripts involved in metabolic functions, such as Satl, Argl and Hifla72(Figure 5A, C). Similar changes in metabolic profile have previously been associated with the conversion of regulatory T cells to highly inflammatory T helper 17 cells73. It therefore appears plausible that an analog metabolic mechanism coexists during the conversion of monocytes to immunosuppressive TAMs and that this metabolic pathway is remodeled following TREM2 blockade. To assess whether the observed effects in TAMs were directly mediated by the aTREM2 antibody, we performed in vitro differentiation of bone-marrow-derived macrophages using varying doses of GBM-conditioned medium in the presence or absence of the aTREM2 antibody. Supplementing cell culture medium with varying doses of GL261 supernatant elicited a dose-dependent effect on the state of bone-marrow-derived macrophages, mimicking the in vivo TME. Using these in vitro GBM TME conditions, we recapitulated several of the in vivo effects of TREM2 blockade in primary murine macrophages (Figure 5E). Moreover, we observed increased Satl expression in bone-marrow-derived myeloid cells treated with aTREM2 antibody, indicating that the metabolic changes may be a direct consequence of TREM2 blockade (Figure 5E).

[0240] To obtain mechanistic insights into the environmental context in which aTREM2 treatment redirects TAM differentiation towards an inflammatory state, we investigated the potential downstream crosstalk of Trem2-expressing cells following aTREM2 treatment. We investigated ligand-receptor pairs which could explain the observed changes in target gene expression elicited by aTREM2 treatment in TAMs. Strikingly, a small set of ligands were predicted to explain a large portion of the downstream effects of TREM2 inhibition in TAMs (p = 3.9 x 10-102). These ligands were mostly restricted to the chemokines CCL3, CCL4, CCL5, CCL7, CCL8 and CXCL9 and complement factor C3 (Figure 5F). In addition to their predicted role, the same chemokines ranked among the most differentially expressed genes in TAMs following aTREM2 treatment (C3, Ccl3, Ccl4, Ccl5, Ccl7, Ccl8, Cxcl9) and demonstrated temporal expression kinetics, providing further evidence that this set of chemokines may be crucial for the myeloid remodeling and even for clinical outcomes (CCL5, CCL7; p = 0.016, p = 0.029 respectively) (Figure 5C, F). Chemokine circuits have been reported to play a fundamental role in the establishment and modulation of the TME74. Thus, we next assessed whether the set of differentially abundant chemokines that explained the majority of downstream signaling upon TREM2 inhibition in TAMs, may also affect neighboring leukocytes in a paracrine fashion. Interestingly, these chemokines and complement factor C3 showed widespread effects in modulating gene expression in receiver cells. The identified set of ligands explained a large fraction of the observed gene expression changes upon TREM2 inhibition in NK cells, monocytes and monocyte-derived macrophages providing a rationale for how modulating TAM differentiation may affect the organization of the TME (Figure 5F).

[0241] In summary, applying Zman-seq in a therapeutic context identified the capacity of TREM2 inhibition to reprogram the monocyte-to-regulatory macrophage trajectory and temporally resolved molecular circuits downstream of TREM2 that facilitate the transition of circulating monocytes to suppressive TAM.

[0242] EXAMPLE 2

[0243] MATERIALS AND METHODS

[0244] Tumor implantation

[0245] Eleven-week-old female mice were anesthetized with isoflurane (Isoflurane, USP Terrell, Piramal) (3-4% induction, 1.5-2% maintenance), received local 2 mg / kg lidocaine (Vetmarket 2%, 162097) and systemic 0.05 mg / kg buprenorphine (Bupaq, 0.3mg / ml, Richter Pharma) analgesia. Mice were intracerebrally injected with 2xl04GL261 cells in 2pl PBS in the right anterior striatum (coordinates from Bregma: +0.5 mm antero-posterior, +2.2 mm medio-lateral, -3.5 mm dorso- ventral). Mice were kept on a heating pad during the procedure and until mice were awake and mobile. After surgery mice were monitored and treated with meloxicam 5mg / kg (Loxicom, 5 mg / ml, Norbrook) subcutaneously daily, for three days.

[0246] Mice for antibody titration were injected with 10A6 MC38 tumor cells subcutaneously, and tumors were extracted after ~2 weeks. In vivo time stamping

[0247] Mice were briefly anesthetized with isoflurane (3-4%) and were injected with 100 pl of fluorescent antibody solution intravenously (containing 1.2 or 4 μg antibody corresponding to 60 pg / kg or 200 pg / kg body weight (bw) respectively). Time stamping was done with CD45 antibodies, clone 30-F11 with the following fluorophores: PE (36 hours), APC (24 hours), and FITC (12 hours).

[0248] For the antibody titration experiments, mice were injected 48 hours before MC38 tumor extraction, either with FITC, APC, or PE-CD45 (30F11) antibodies, each at three doses (3, 6, or 18 ug / kg).

[0249] Single-cell isolation and sorting for RNA sequencing

[0250] Tumor-bearing mice were sacrificed 13 days after tumor inoculation. Mice were deeply anaesthetized by intraperitoneal injection of ketamin 10 mg / ml (Ketavet Veterinary, 10Omg / ml, Zoetis) and xylazine 2 mg / ml (Sedaxylan 20mg / ml, Eurovet). Mice were transcardially perfused with 10 ml of ice cold PBS. After brain extraction, tumors were dissected to reduce the amount non-tumorous brain tissue in downstream analyses. Samples were blade-minced on ice and transferred to 1.3 ml digestion buffer (0.0125 mg / ml dNase type I (Roche) and 1 mg / ml collagenase IV (Worthington) in RPMI-1640 (Gibco, 21875-034). Samples were incubated for 20 min at 37 °C, during this incubation samples were mechanically dissociated with syringes (gauge 26). Cells were then transferred to ice and filtered through a 100-pm cell strainer, washed with ice-cold FACS buffer (0.5% BSA, 2 mMEDTA in PBS), and centrifuged (5 min, 4 °C, 350g). Supernatant was aspirated and samples were resuspended in 37% Percoll (Sigma, Cat# GE17089101) density gradient, followed by centrifugation at 800 g, 30 min at 4 °C. The cell pellet was washed PBS, followed by Fc-blocking (1:200) (TruStain FcX, Clone 93, BioEegend) and staining with CD45-BV711 (30-F11), and anti-APC (clone APC003), PE (clone PE001), FITC (clone FIT-22) barcoded antibodies (Totalseq-B barcodes, Biolegend). DAPI 0.1 pg / ml (BioLegend) was used for live / dead stain. Cells were filtered through a 70 pm cell strainer. CD45 bulk sorting was done on BD FACS Symphony 6 flow cytometer (BD Biosciences). Dead cells and doublets were excluded by gating, after which CD45+cells were gated.

[0251] Single-cell library preparation

[0252] The scRNA-seq libraries were created using the commercial 10X Genomics 3’ genomic and feature barcode pipeline. Quality and DNA concentration tests were performed on each library. Quantitative PCRfor antibody calibration

[0253] After bulk sorting, T-cells were stained with 1:50, 1:100, or 1:200 dilutions of barcoded anti-PE, -APC, or -PE antibodies. Barcodes were detected using reverse transcription (Superscript II RT, Thermo), followed by qPCR (Fast SYBR, Thermo) using primers complementary to the published 10X genomics 3’ feature barcode system.

[0254] In vitro T-cell stimulation and temporal barcoding

[0255] T-cells were isolated (MACS Pan T Kit II) and stained in triplicates with CD45-PE, CD45- APC, and CD45-FITC at 6-hour intervals, and stimulated with CD28 / CD3 activation beads at the same time. At the end of the experiment, all samples were stained with a mixture of barcoded anti- PE / APC / FITC antibodies, further, each sample was given an individual barcode via hashing antibodies (anti-mouse and MHCI hashing TotalseqB, Biolegend). Cells were sorted with flow cytometry and submitted to the 10X 3’ single-cell pipeline.

[0256] RESULTS

[0257] In vitro CD38 / CD3 T cell stimulation with CD45 timestamping and single-cell analysis

[0258] To investigate the temporal dynamics of T cell activation, T cells were stimulated in vitro using CD38 / CD3 in triplicates (Figures 6A-B). Sequential activation was paired with CD45-APC, PE, or FETC timestamping, followed by detection of these fluorophores using specific barcoded antibodies and hashing of individual samples. After pooling the samples, they were submitted to the 10x Genomics single-cell pipeline. Using this approach, the samples were successfully demultiplexed. The corresponding time stamps were identified and the dynamic changes in T cell states over time were tracked.

[0259] Titration of in vivo CD45 timestamping in MC38 tumor-bearing mice

[0260] Next, the present inventors optimized in vivo CD45 timestamping and detection by injecting either 3, 6, or 18 pg / kg of CD45-PE, -APC, or -FITC antibodies into MC38 tumorbearing mice 48 hours before tumor harvesting (Figures 7A-C). Corresponding barcoded anti-PE, APC or FITC antibodies were used at 1:200, 1:100, and 1:50 dilutions to detect the injected labels via qPCR. A gradual increase in detection sensitivity (reflected by reduced Ct values) was observed with increasing antibody concentrations. Notably, the 1:50 dilution exhibited the highest signal intensity, particularly for the PE fluorophore.

[0261] In vivo single-cell timestamping in GL261 brain tumors

[0262] In vivo single-cell timestamping was performed in GL261 brain tumors at 12, 18, and 24 hours, combined with ex vivo detection using barcoded antibodies. This approach allowed for the tracking of immune cell dynamics and gene expression changes over time in natural killer NK cells (Figures 8A-B), recapitulating the results of fluorophore-based Zman-seq (see Example 1) - further validating the system for in vivo temporal profiling of immune responses.

[0263] Although the invention has been described in conjunction with specific embodiments thereof, it is evident that many alternatives, modifications and variations will be apparent to those skilled in the art. Accordingly, it is intended to embrace all such alternatives, modifications and variations that fall within the spirit and broad scope of the appended claims.

[0264] It is the intent of the applicant(s) that all publications, patents and patent applications referred to in this specification are to be incorporated in their entirety by reference into the specification, as if each individual publication, patent or patent application was specifically and individually noted when referenced that it is to be incorporated herein by reference. In addition, citation or identification of any reference in this application shall not be construed as an admission that such reference is available as prior art to the present invention. To the extent that section headings are used, they should not be construed as necessarily limiting. In addition, any priority document(s) of this application is / are hereby incorporated herein by reference in its / their entirety.

Claims

WHAT IS CLAIMED IS:

1. A method of analyzing cell fate over a course of time comprising analyzing cells which have been temporally labelled at a predetermined location at at least two different time points using at least two different labels, each of said at least two different labels corresponding to one of said at least two different time points, wherein said cells temporally labelled at said at least two different time points have localized to a target location prior to said analyzing, thereby analyzing cell fate over the course of time.

2. The method of claim 1, further comprising temporally labelling said cells at said predetermined location prior to the analyzing.

3. The method of claim 2, wherein said temporally labelling said cells at said predetermined location is effected in vivo.

4. The method of claim 2, wherein said temporally labelling said cells at said predetermined location is effected ex vivo.

5. The method of any one of claims 1-4, wherein said analyzing is effected in vivo.

6. The method of any one of claims 1-4, wherein said analyzing is effected ex vivo.

7. The method of claim 4, wherein said target location is in the body of an animal.

8. The method of any one of claims 1-7, wherein the analyzing is effected at least 24 hours after the later time point of said at least two different time points.

9. The method any one of claims 2-8, wherein said temporally labelling said cells comprises:(a) contacting cells of a cell type with a cell affinity agent which comprises a first label, said cell affinity agent being specific for said cell type, so as to generate first label-bound cells at said predetermined location; and subsequently(b) contacting additional cells of said cell type with a cell affinity agent which comprises a second label, so as to generate second label-bound cells at said predetermined location.

10. The method of claim 9, wherein said analyzing is effected at least 1 hour following said contacting additional cells.

11. The method of claim 9, wherein said cell affinity agent binds to an exterior of the cells.

12. The method of claim 9, wherein said cell affinity agent is selected from the group consisting of an antibody, a nanobody and an aptamer.

13. The method of any one of claims 9-12, wherein said cell affinity agent comprising said first label is of the same type as said cell affinity agent comprising said second label.

14. The method of any one of claims 9-13, wherein said cell affinity agent comprising said first label and said cell affinity agent comprising said second label bind to an identical target molecule.

15. The method of any one of claims 1-13, wherein said cells have been temporally labelled at at least three different time points, wherein said cells temporally labeled at said at least three different time points have localized to said target location prior to said analyzing.

16. The method of claim 9, further comprising, following step (b):(c) contacting additional cells of said cell type with said cell affinity agent which comprises a third label, so as to generate third label-bound cells at said predetermined location.

17. The method of any one of claims 9-16, wherein said cell affinity agent is an antibody.

18. The method of any one of claims 9-17, wherein said first label and said second label comprise fluorescent moieties.

19. The method of claim 16, wherein said first label, said second label and said third label comprises fluorescent moieties.

20. The method of any one of claims 9-17, wherein said first label and said second label comprise nucleic acid barcodes.

21. The method of any one of claims 9-17, wherein said first label comprises a first antigen determinant and said second label comprises a second antigen determinant.

22. The method of claim 21, further comprising contacting said first label-bound cells with an additional affinity agent which specifically binds to said first antigen determinant and which is labeled with a first nucleic acid barcode and contacting said second-label bound cells with an additional affinity agent which specifically binds to said second antigen determinant and which is labeled with a second nucleic acid barcode.

23. The method of claim 16, wherein said first label, said second label and said third label comprises nucleic acid barcodes.

24. The method of any one of claims 9-22, wherein step (b) is effected at least 6 hours following step (a).

25. The method of any one of claims 1-24, wherein said cells comprise white blood cells.

26. The method of claim 25, wherein said cell affinity agent comprises an agent which binds specifically to CD45.

27. The method of any one of claims 1-26, wherein said predetermined location is in the blood circulation of an animal.

28. The method of any one of claims 1-27, wherein said target location is in a tissue or a tumor microenvironment of an animal.

29. The method of any one of claims 1-28, wherein said analyzing comprises cell transcriptome analysis or cell proteome analysis.

30. The method of claim 5, wherein said analyzing comprises imaging.

31. The method of any one of claims 1-30, wherein said analysis is effected on the single cell level.

32. A kit for analyzing cell fate over a course of time comprising: a first pair of affinity agents which comprise:(i) a first cell affinity agent which binds to an outer surface of a specific cell, the first cell affinity agent comprising a first antigenic determinant:(ii) a second cell affinity agent which binds to an outer surface of said specific cell, the second cell affinity agent comprising a second antigenic determinant; and a second pair of affinity agents which comprise:(iii) a third affinity agent which binds specifically to said first antigenic determinant, said third cell affinity agent comprising a first nucleic acid barcode; and(iv) a fourth affinity agent which binds specifically to said second antigenic determinant, said third cell affinity agent comprising a second nucleic acid barcode wherein the kit comprises no more than 10 pairs of affinity agents.

33. The kit of claim 32, wherein said first cell affinity agent is an antibody labeled with a first fluorescent moiety, said first fluorescent moiety comprising the first antigenic determinant and said second cell affinity agent is an antibody labeled with a second fluorescent moiety, said second fluorescent moiety comprising the second antigenic determinant.

34. The kit of claims 32 or 33, wherein said third cell affinity agent and said fourth cell affinity agent are antibodies.

35. The kit of any one of claims 32-34, further comprising agents for performing single cell transcriptome analysis.

Images