METHOD FOR ANALYZING CELL SAMPLE HETEROGENEITY

Abstract

The present invention relates to a method for analyzing a biological sample on a single cell level by compartmentalizing said sample using a grid and performing an optimized combinatorial indexing protocol within cells of the compartmentalized tissue while deducing cell-specific information regarding the cell identity and activity in the spatial context within the sample.

Claims

1-22. (canceled)

23. A method for analyzing a tissue sample comprising cells, wherein said method comprises the following steps: (a) Compartmentalizing said sample with a grid into at least two compartments, wherein the grid comprises at least two break-through holes which define said at least two compartments of the sample and wherein at least one of said at least two compartments of the sample comprises at least one cell containing a plurality of molecules; and (b) Labeling the plurality of molecules within said at least one cell in situ with at least a compartment-specific label, wherein before step (a) the sample is fixed and permeabilized, and wherein steps (a) and (b) are performed repeatedly and alternating.

24. The method according to claim 23, wherein before every repetition of step (a) at least one of the following steps is performed: rotating the grid in relation to the surface defined by the two largest dimensions of the sample around an axis perpendicular to said surface; rotating the grid in relation to said surface around an axis parallel to said surface; shifting the grid in relation and parallel to said surface; exchanging the grid by a further grid, wherein said further grid comprises at least two break-through holes which differ with respect to their number and/or at least one cross-sectional size and/or at least one cross-sectional geometry from the at least two break-through holes of any of the at least one grid applied in any of the previous steps (a); choosing one of the at least one grid applied in any of the previous steps (a).

25. The method according to claim 23, wherein at least two grids with at least two different grid geometries are applied in at least three successively performed rounds of compartmentalization and compartment-specific labeling.

26. The method according to claim 23, wherein the tissue sample is a cryosection of a tissue.

27. The method according to claim 26, wherein the cryosection has a slice thickness between 1 μm and 50 mm or between 5 μm and 15 μm.

28. The method according to claim 23, the method further comprising the following steps: (i) Obtaining a first suspension comprising at least a fraction of the at least one cell containing a plurality of molecules derived from step (b) and distributing at least a fraction of the first suspension over a first set of at least two vessels; and (ii) Labeling the plurality of molecules comprised in the suspensions contained in the first set of at least two vessels with a first vessel-specific label within said at least one cell.

29. The method according to claim 28, the method further comprising the following steps that are performed, repeatedly and alternating, at least once: (iii) Obtaining a further suspension, comprising at least a fraction of the at least one cell containing a plurality of molecules, by mixing at least a fraction of the suspensions derived from step (ii) in case step (iii) is performed for the first time, or from step (iv) in case steps (iii) and (iv) have been performed at least once, and distributing at least a fraction of the further suspension over a further set of at least two vessels; and (iv) Labeling the plurality of molecules comprised in the suspensions contained in the further set of at least two vessels with at least a further vessel-specific label.

30. The method according to claim 29, wherein lysis of the at least one cell is performed after step (b), or in case step (iii) and step (iv) are performed once: after step (iii) and before step (iv), or in case step (iii) and step (iv) are performed in multiple iterations: after step (iii) and before step (iv) of the last iteration.

31. The method according to claim 23, further comprising obtaining data for at least a fraction of the sequence of at least a fraction of the plurality of molecules, said data being obtained using a sequencing or genotyping method or a next generation sequencing method.

32. The method according to claim 23, wherein the method further comprises identifying compartment-specific and/or vessel-specific labels, obtaining cell-specific parameters, and obtaining information of relative spatial cell positions.

33. The method according to claim 23, wherein the method further comprises comparing results for said sample to results obtained for at least a second sample and/or information obtained by other methods at least for said sample.

34. The method according to claim 23, wherein the method further comprises obtaining and/or analyzing at least one biomarker indicative for a condition.

35. The method according to claim 23, wherein before step (a) the sample is fixed and permeabilized.

36. The method according to claim 23, wherein the method further comprises positioning the sample on a base of a grid system, positioning the grid on the surface of the sample that is opposite to the surface of the sample that is in contact with the base, and moving the grid towards the base until the grid is in contact with the base.

37. The method according to claim 23, wherein said molecules are naturally occurring, synthetic and/or engineered molecules comprising nucleic acids and/or polypeptides.

38. The method according to claim 37, wherein each of the compartment-specific labels has a compartment-specific nucleotide sequence of 1 nucleotide to 200 nucleotides in length or of 5 nucleotides to 25 nucleotides in length, and/or wherein each of the vessel-specific labels has a vessel-specific nucleotide sequence of 1 nucleotides to 200 nucleotides in length or of 5 nucleotides to 25 nucleotides in length.

39. The method according to claim 23, wherein step (b) comprises reverse transcribing the plurality of molecules within the at least one cell using compartment-specific labeled primers.

40. The method according to claim 23, wherein each compartment does not comprise more than one cell.

41. The method according to claim 23, wherein the plurality of molecules comprising at least one compartment-specific label is labeled with at least one vessel-specific label by ligation.

42. The method according to claim 23, wherein the plurality of molecules comprising at least one compartment-specific label is amplified by PCR.

43. The method according to claim 42, wherein at least one further vessel-specific label is added to the molecules by PCR.

44. The method according to claim 23, further comprising the steps of performing second strand synthesis of the plurality of molecules and/or tagmentation of the plurality of molecules.

Description

BRIEF DESCRIPTION OF THE FIGURES

[0134] FIG. 1: Workflow for analyzing a sample according to a preferred aspect of the present invention using a grid 200 and applying a combinatorial indexing protocol for labeling the plurality of molecules comprised in cells of the sample.

[0135] FIG. 2: Grid 200 according to the present invention.

[0136] FIG. 3: Grid system 300 according to the present invention.

[0137] FIG. 4: 2-dimensional visualization of single-cell expression data obtained for HEK293 cells of 16 wells of a 96 well plate upon analysis of the obtained sequence data using t-distributed stochastic neighbor embedding (t-SNE).

[0138] FIG. 5: Relationship between the number of transcripts and the number genes per cell based on single-cell expression data of HEK293 cells.

[0139] FIG. 6: Correlation of gene expression of HEK293 cells between data obtained by the method according to the present invention and data of a published dataset (Alles et al., BMC Biology, 2017, 15:44) that was sequenced at 50-fold higher depth.

[0140] FIG. 7: Set-up with two grids: [0141] 1. a first grid of the illustrated geometry was applied to the sample to attach the RT barcodes; [0142] 2. a second grid of the illustrated geometry was applied to the sample to attach a second barcode (ligation); [0143] 3. this effectively compartmentalizes the sample into 9 compartments. The matrix shows the number of transcripts found per barcode combination in the sequencing data. [0144] 4. the 9 barcode combinations comprise more than 98% of the transcripts quantified in the sequencing data.

[0145] Other aspects and advantages of the invention will be described in the following examples, which are given for purposes of illustration and not by way of limitation. Each publication, patent, patent application or other document cited in this application is hereby incorporated by reference in its entirety.

EXAMPLES

[0146] Methods and materials are described herein for use in the present disclosure; other, suitable methods and materials known in the art can also be used. The materials, methods, and examples are illustrative only and not intended to be limiting.

Example I: Manufacture of a Grid System

[0147] The grid system 300 comprised the grid 200, a standard microscope glass slide as the base 301 and a supporting device 303 for fixing the grid 200 on the base 301. The grid 200 and its supporting device 303 were designed using the Autodesk Fusion360 CAD software (v2.0.3706). The supporting device 303 was printed with a BIMSB-owned ultimaker 2+ using filaments made from polyacetide and polyhydroxy alkanoate. The grid 200 was printed using a Projet HD 3000 Plus 3D Production System with Visijet EX200 building material being thermostable up to 50° C. A total of four grids 200 was manufactured. Each grid 200 had a height 204 of 2 mm, a width 205 of 9.5 mm, and a length 206 of 11 mm. Furthermore, each grid 200 comprised two major surfaces 208 and 207 as well as 6×5 break-through holes 202 defined by 4 outer grid walls 201 and 20 inner grid walls 203. According to this embodiment, surfaces 207 and 208 have the same sizes and geometries. The break-through holes 202 were equally sized each enclosing a cross-sectional area of 1 mm.sup.2 defined by four 1 mm long sides of the inner grid walls 203. The break-through holes 202 were separated by the inner grid walls 203 having a width of 100 μm, 150 μm, 200 μm and 300 μm, respectively. It was tested and confirmed that the grids (i) do not melt upon exposure to high temperatures, (ii) do not leak i.e. that leakiness between compartments if present at all does not influence the obtained results by more than 5%, and (iii) can be applied to recover substantial fraction of cells after sample compartmentalization. Moreover, it has been confirmed that the grid 200 and/or the grid system 300 according to the methods of the present invention is easy handle without requiring for example robotics, though an automated system might optionally be used when applying the grid to a sample.

Example II: Optimized Combinatorial Indexing

[0148] In this example, the sample was an adult mouse brain. The sample was embedded in a cryoprotectant and sliced into cryosections of 10 μm thickness. The tissue was fixed and permeabilized 101 with methanol. A grid 200 manufactured as described above was applied to the cryosection and tightened to the tissue and the base 301 by a support device 303 with screws 302. By applying the grid 200, the tissue was compartmentalized 102 into 1 mm.sup.2 compartments, corresponding to approximately 1,000 cells. mRNA molecules were labeled with compartment-specific labels within cells 103 by a new established in situ reverse transcription protocol using 30 different labels. Thus, reverse transcription was carried out directly on the methanol fixed tissue slice for preserving spatial information by labeling molecules with a compartment-specific label. Nucleotide sequences added during reverse transcription to the plurality of mRNA molecules each comprised a compartment-specific label flanked on one side by a ligation linker region and on the other side by a ligation barcode linked to an oligo(dT) primer. The compartmentalized tissue was digested using papain for dissociating the tissue compartments and obtaining a single cell solution 104 that was distributed over 96 wells of a 96 well plate 105. mRNA molecules were labeled by ligation using splint oligos and 96 different labels corresponding to one label sequence per well 106. The vessel-specific labels differed in their length with the length being in the range of 6 to 18 nucleotides and were flanked by a ligation linker region and a PCR handle. Cell solutions were obtained from the wells and mixed 104 before distribution over 96 wells of a further 96 well plate 105. Second strand synthesis was performed and cells lysed for obtaining purified double stranded cDNA molecules for tagmentation. Upon tagmentation, cDNA molecules were labeled with 96 further vessel-specific labels by PCR 106 with each of the 96 further vessel-specific label sequences being specific to one of the 96 further wells. Labeling was performed at both ends of the cDNA molecules by PCR with nucleotide sequences comprising said further vessel-specific label linked to a sequencing adapter such that the further vessel-specific label were added to the cDNA molecules.

[0149] Thus, at each of the three labeling steps 103,106,106 several hundred cells were labeled with the same label sequence. However, as the cells were mixed and distributed over a new set of vessels, i.e. wells, between the labeling steps, at each labeling step a different set of cells was labeled with the same label sequence. Using 30 compartment-specific label sequences in the reverse transcription reaction 103, 96 vessel-specific label sequences for ligation 106 and 96 vessel-specific label sequences during PCR 106, a total of 276,480 possible combinations of label sequences could be obtained by these three labeling steps. Moreover, labels were designed to be diverse with respect to sequence and length for relatively easy and efficient recovery after sequencing and for obtaining high quality data. Additionally, the optimized combinatorial indexing protocol does not require any expensive biotin-streptavidin purification step, thus lowering costs and protocol complexity.

Example III: Quality Aspects

[0150] mRNA molecules of HEK293 cells were labeled by combinatorial indexing as described in Example II and sequenced at very shallow depth 107. Sequence data analysis 108 comprised the steps described in the following. Standard quality controls of the sequenced reads were performed using the publicly available FastQC software (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). Cell-specific gene expression was estimated by mapping the reads to the reference genome 109 using the STAR software (Dobin et al., Bioinformatics, 2013, 29(1):15-21), followed by obtaining the labeling information per mRNA molecule and quantifying the gene expression based on the amount of reads per cell using the Picard (https://broadinstitute.github.io/picard/) and the Drop-Seq (http://mccarrolIlab.com/dropseq/; Macosko et al., Cell, 2015, 161(5):1202-1214) tools.

[0151] With approximately 1,000 reads per cell, a total of 7,725 genes could be recovered in a total of 875 cells. Based on the obtained cell-specific expression data it could be demonstrated that (i) no batch effects exist between different wells (FIG. 4), (ii) there is a linear relationship between the number of genes and the number transcripts per cell (FIG. 5), and that (iii) the estimated gene expression correlates well with the gene expression reported for a state-of-the-art published dataset (Alles et al., BMC Biology, 2017, 15:44) sequenced at 50-fold higher depth (FIG. 6).

Example IV: Multiple Indexing Rounds

[0152] The following experiment was performed for investigating the performance of multiple rounds of barcoding on tissue, and especially of reverse transcription (RT) and ligation directly on tissue. Exemplarily, two grids having the same geometry were used and successfully applied to the tissue in varying orientation in relation to the latter.

[0153] Cell lysis, tagmentation and polymerase chain reaction (PCR) were performed on all cells in parallel. As a read out quantitative PCR (qPCR) was used for investigating barcoding and for estimating leakiness of the grids. Sequencing was used to analyze the experiment.

[0154] All cell pelleting steps in the following were performed in Maxymum Recovery tubes (Corning MCT-150-L-C).

[0155] I. Fixation of Tissue with Methanol (MeOH) [0156] 1. Thaw tissue slice containing boxes from −80° C. to room temperature [0157] 2. Wash, e.g. two, grids 2× with 100% ethanol (EtOH) to remove e.g. hydrophobic rests, wash with water and air dry [0158] 3. Thaw RT primers, Superscript IV buffer, DTT, dNTPs (ThermoFisher Scientific 18091050, kit) on ice [0159] 4. Fix slice in cold MeOH (stored at −20° C.), 20 min at −20° C. [0160] 5. Dry slices at room temperature and wash 3× with cold PBS/Ribolock (1:80) [0161] 6. Fix first grid on tissue slice

[0162] II. RT

[0163] Total of 15 μl reaction per grid compartment; total of 3 reactions: [0164] 1. Put RT primer in PCR plate—same orientation as in the grid afterwards, e.g.:

TABLE-US-00001   1 2 3 [0165] 2. Mix remaining Mastermix separately according to the following scheme and then add 10 μl to RT primer:

TABLE-US-00002 Per E.g. for well 3 wells = 3.2x RT primer 100 μM 5.0 — 10 mM dNTPs 2.5 8.0 5x Superscirpt IV Buffer 4.0 12.8 Ribolock (ThermoFisher 2.5 8.0 Scientific EO0381) 100 mM DTT 1.0 3.2 Superscript IV 2.5 8.0 [0166] 3. Add 15 μl of Mastermix to each grid compartment [0167] 4. Close first grid with qPCR plate foil seal [0168] 5. Incubate at 25° C. for 30 min, then at 37° C. for 30 min, and then at 45° C. for 20 min; e.g. by placing the first grid on a copper plate in a 96 well PCR cycler, tightening the lid of the PCR cycler, using “Block control, highest possible sample volume”, and “hot lid control: 50° C.” [0169] 6. During incubation: [0170] Prepare ligation plate: [0171] 1. Prepare ligation barcode plate: [0172] a. Anneal ‘Ligation linker oligonucleotide 1″ and “Ligation barcode oligonucleotides 1/2/3” [0173] b. Dilute barcode stocks to 50 μM using the following scheme:

TABLE-US-00003 For 10.0 μl 20 μM barcode mix x4 Ligation linker oligonucleotide 4.0 16.0  1 50 μM Ligation barcode 4.0 — oligonucleotides 1/2/3 50 uM CutSmart Buffer 10x 1.0 4.0 Water 1.0 4.0 [0174] c. Heat to 95° C. for 5 min, ramp down to 20° C. with −0.1° C./s (e.g. using an Eppendorf Mastercycler X50s for ligation annealing) [0175] 2. Ligation reaction, add 10 μl of the following ligation mix to the annealed 10 μl in the barcode plate:

TABLE-US-00004 20 μl Reaction x4 T4 DNA Ligation Buffer 2.0 8.0 T4 DNA Ligase (NEB 0.5 2.0 M0202M) (10 U/μl final) Water 7.5 30.0 [0176] 7. When incubation with RT mix is done, disassemble grid-slice-slide structure and wash once with cold PBS/Ribolock [0177] 8. Place second grid on the tissue slice and fill 20 μl of ligation mix into each compartment, with the second grid being placed e.g. in the following orientation:

TABLE-US-00005 1 2 3 [0178] 9. Incubate 1 h at room temperature [0179] 10. Meanwhile: [0180] Prepare papain for tissue digestion for 15 min before RT is finished: [0181] Aliquot powder of one Worthington Papain Vial (Worthington LK003178) in 3 Eppendorf tubes/or use one aliquot [0182] Add 500 μl EBSS (Earle's Balanced Salt Solution, Sigma Aldrich E2888) [0183] Incubate for 10 min at 37° C., e.g. in a cell culture incubator for O.sub.2:CO.sub.2 equilibration, check if color changed from very pink to orange [0184] 11. Unscrew second grid and wash slice carefully but fast with 1×PBS/Ribolock [0185] 12. Dry surrounding of the tissue slice with Q-tip, and circle with a hydrophobic pen [0186] 13. Prepare single cell suspension: [0187] Place microscope slide in a plastic box that can be closed with a lid and place a water-soaked Kimwipe next to it (to avoid drying out) [0188] Add 150-300 μl prepared, equilibrated papain solution to the slice and incubate for 15 min at 37° C., e.g. in a cell culture incubator, with the plastic box being closed [0189] Wash slice off into an Eppendorf tube, if slice parts stick scratch off with pipette tip [0190] Completely harvest slice into an Eppendorf tube, use up to 300 μl PBS/Ribolock (1:80)+0.01% BSA to aid washing off [0191] Incubate for another 15 min at 37° C. on a shaker with 12 rpm [0192] Spin down cells 3000 g for 15 min at 4° C. [0193] Discard supernatant and resuspend in 1000 μl PBS/Ribolock+BSA [0194] Spin down cells: 3000 g, 15 min, 4° C. [0195] Resuspend in 110 μl PBS/Ribolock [0196] Take out 6 μl and mix with 6 μl Trypan blue for cell counting [0197] Count cells: _cells/ml (DEAD) [0198] 14. Add 30 μl 4× Dropseq Lysis Buffer (see below)+DTT

TABLE-US-00006 4x DropSeq Lysis Buffer Mix for (final conc.) 1 ml 50% Ficoll PM-400 24% (6% x4) 480 μl 20% Sarkosyl 0.8% (0.2% x4)  40 μl 0.5M EDTA 80 mM (20 mM x4) 160 μl 1M Tris pH 7.5 120 mM 120 μl 1M DTT (add 200 mM (50 mM x4) 200 μl fresh) [0199] a. Incubate 30 min at room temperature, shaking on a vortexer (e.g. stage 4) [0200] b. Fill up to 500 μl with water by adding 366 μl [0201] c. Add 329.4 μl AMPure XP beads (Beckmann Coultier A63881) (0.9×) per well mix thoroughly (5× with pipette), incubate at room temperature for 5 min, bind to magnet, wash with 1000 μl 80% EtOH (take of 96-well plate from magnet, add EtOH, put on magnet, shift right, shift left, wait for beads to bind to magnet, take off EtOH) [0202] d. Repeat washing with 1000 μl 80% EtOH, let beads air dry (at appearance of first crack, proceed with elution) [0203] e. Elute in 200 μl water (add water, resuspend by pipetting, incubate 5 min at room temperature, put back on magnet, recover supernatant) [0204] f. Repeat AMPure bead clean-up, add 160 μl beads (0.8×), elute in 100 μl water

[0205] III. Second Strand Synthesis [0206] 1. Perform Second Strand Synthesis, 1 reaction, e.g. using the following scheme

TABLE-US-00007 1x 10x Second Strand Buffer 12.0 Second Strand Enzyme Mix (NEB E6111S) 4.0 Random hexamers (Fermentas SO142) 1.0 Water 3.0 [0207] Incubate for 1:30 h at 16° C.; e.g. in an Eppendorf Mastercycler [0208] 2. Purify dsDNA by adding 120 μl AMPure XP beads (1.0×): add, mix, incubate for 5 min at room temperature; put on magnet; wash twice with 1000 μl 80% EtOH; and elute in 35 μl [0209] 3. Take out 5 μl for qPCR to measure cDNA amount per well [0210] a. E.g. use qPCR stripes (e.g. by Applied Biosystems, MicroAmp strips+optical caps) [0211] b. Required: standard curve for organism of interest as shown below for primers for mouse and human:

TABLE-US-00008 TABLE qPCR scheme: qPCR − 3D-seq measure cDNA amount per well qPCR plate 1 2 3 4 A sample + primer 647/692 sample + primer 647/694 B water + primer 647/692 water + primer 647/694 C D E F G H cDNA + 7.5 ul SYBR = 11.25 ul per well expl. wells cDNA [ul] water [ul] SYBR [ul] sample 4 4.40 12.10 33.00 water 4 0.00 16.50 33.00 Primer − 2.5 uM mix stock = 3.75 ul per well 647/692 binds in ligation barcode + 3′ end of ActB 647/694 binds in ligation barcode + 3′ end of Tubb3 to make 2.5 uM mix stock from 100 uM stock: 647 20 ul 692 20 ul water 760 ul 

[0212] The sample and water (as control) are measured with 2 sets of primers. One primer binds the ligation linker sequence (647) the other primer binds at the 3′end of a gene (Actb or Tubb3, housekeeper genes were selected). This is then compared to a previously measured standard curve to estimate the concentration of molecules that have a ligation barcode attached. [0213] Calculate pg/μl e.g. as follows:

[00001]$\mathrm{\_\_pg}/\mu l\times 50\mu l=\mathrm{total}\mathrm{mass\_\_pg}=\mathrm{\_\_ng}$$5\mu l\mathrm{Amplicon}\mathrm{Tagmentation}\mathrm{Mix}\left(\mathrm{ATM}\right)\mathrm{per}600\mathrm{pg}->\mathrm{\_\_\mu }l\mathrm{ATM}\left(\times \mathrm{pg}\star 5\mu l/600\mathrm{pg}\right)\mathrm{need}\mathrm{to}\mathrm{be}\mathrm{added}\mathrm{in}\mathrm{the}\mathrm{tagmentation}\mathrm{reaction}$

[0214] IV. Tagmentation [0215] 1. Tagmentation for 5 min at 55° C., with a final volume of 75 μl according to the following scheme, mix well

TABLE-US-00009 1x Nextera TD Buffer 37.5 (IIlumina Nextera XT Library Preparation Kit FC- 131-1096, includes Buffer, Amplicon tagmentation mix and NT buffer) (2x) Amplicon Tagmentation mix Sample 30.0 Water Fill up to 75 μl [0216] 2. Add 5 μl NT buffer, mix by pipetting using e.g. 10 pipetting steps [0217] 3. Perform PCR, e.g. with Terra Polymerase for 12 cycles using the following schemes:

TABLE-US-00010 1x sample 80.0 10 μM N7xx (Illumina 2.0 compatible PCR primers, ordered from Eurofins) 10 μM S5xx (IDT for 2.0 Illumina Nextera DNA Unique Dual Indexes Set B 20027214) 5 μM Library 1.0 Amplification Primer 1 5 μM Library 1.0 Amplification Primer 2 Water 10.0 Terra Buffer 100.0 Terra polymerase 4.0 (Clontech 639271)

[0218] Terra Polymerase PCR Program:

TABLE-US-00011 98° C. 2 min 98° C. 10 s X cycles 60° C. 15 s (check qPCR) 68° C. 60 s 68° C. 5 min [0219] 4. Purify with AMPure beads by adding 160 μl of AMPure XP beads (0.8×) to the tube, and elute in 100 μl of water for combining all beads in one tube [0220] 5. Repeat bead clean-up by adding 80 μl AMPure beads (0.8×), and elute in 40 μl water [0221] 6. Continue with sequencing, e.g. on an Illumina Nextseq 550