Magnetic activated cell sorting

Abstract

Certain embodiments of the invention provide methods and assays for increasing the percentage of droplets containing desired cells.

Description

[0001] MAGNETIC ACTIVATED CELL SORTING

[0002] CROSS-REFERENCE TO RELATED APPLICATION

[0003] This application claims priority to United States Provisional Application Number 63 / 708,556 that was filed on October 17, 2024. The entire content of the application referenced above is hereby incorporated by reference herein.

[0004] BACKGROUND

[0005] For droplet-based assays of single cells, one of the primary variables that affects the screening throughput is the percentage of droplets that contain cells. Because the droplets must be analyzed in series, empty droplets or, in the case of 2 or more cell assays, droplets that do not contain all of the required cells, must still be screened and can significantly decrease the effective screening rate of droplets containing all the components for the assay. Therefore, an improved method to increase the percentage of droplets containing cells is needed.

[0006] SUMMARY

[0007] Accordingly, provided herein in certain embodiments are methods for increasing the percentage of droplets containing desired cells in a population of droplets, comprising:

[0008] (1) attaching magnetic beads to a first population of desired cells in a solution to provide a first population of magnetic cells;

[0009] (2) removing magnetic beads not attached to cells from the solution;

[0010] (3) partitioning the desired magnetic cells into magnetic droplets;

[0011] (4) moving the magnetic droplets across a magnetic field;

[0012] (5) removing non-magnetic droplets; and

[0013] (6) removing the magnetic droplets from magnetic field to provide magnetic droplets that comprise the first population of desired cells.

[0014] Also provided herein in certain embodiments are assay and systems for performing the methods described herein.

[0015] BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1. Figure 1 depicts results of dMACS bead-based enrichment (see, e.g., Example

[0017] 1). Figure 2. Figure 2A-2B depicts examples of beta mercaptoethanol declumping images with 2 A at 10 minute time point and 2B at 20 minute time point. BME decreases cell clumping and debris in the channel, which improves droplet formation reliability (see, e.g., Example 2).

[0018] Figure 3. Figure 3A-3G depicts (see, e.g., Example 3). Figure 3A depicts an unenriched fraction. Figure 3B depicts a positive fraction. Figure 3C depicts a negative fraction. Figure 3D and 3E depict co-loading results.

[0019] Figure 4. Figure 4A-4G depicts (see, e.g, Example 4). Figure 4A depicts an unenriched fraction. Figure 4B depicts a positive fraction. Figure 4C depicts a negative fraction. Figure 4D- 4G depict loading results.

[0020] Figure 5. Figure 5A-5E depicts (see, e.g, Example 6) optimizations of certain aspects of the technology described herein.

[0021] Figure 6. Figure 6A-6C depicts certain optimization aspects described, e.g., in Example 7.

[0022] Figure 7. Figure 7 depicts certain additional workflows contemplated and described, e.g., in Example 7.

[0023] DETAILED DESCRIPTION

[0024] Provided herein are methods for increasing the percentage of droplets containing desired cells in a population of droplets. The method comprises (1) attaching magnetic beads to a first population of desired cells in a solution to provide a first population of magnetic cells; (2) optionally removing magnetic beads not attached to cells from the solution; (3) partitioning the desired magnetic cells into magnetic droplets; (4) moving the magnetic droplets across a magnetic field; (5) removing non-magnetic droplets; and (6) removing the magnetic droplets from magnetic field to provide magnetic droplets that comprise the first population of desired cells.

[0025] In certain embodiments, the method further comprises attaching magnetic beads to a second population of desired cells in a solution to provide a second population of magnetic cells and (6) further comprises removing the magnetic droplets from magnetic field to provide magnetic droplets that comprise the first population of desired cells and the second population of magnetic cells.

[0026] In certain embodiments, the method further comprises attaching magnetic beads to a third population of desired cells in a solution to provide a third population of magnetic cells and (6) further comprises removing the magnetic droplets from magnetic field to provide magnetic droplets that comprise the first population of desired cells and the second population of magnetic cells and the third population of desired cells.

[0027] In certain embodiments, the method comprises only labeling one of the cell populations when multiple (e.g., 2, 3 or more) cell populations are present.

[0028] In certain embodiments, the first, second and / or third population of desired cells is a population of antibody producing cells.

[0029] In certain embodiments, the first, second and / or third population of desired cells is a population of reporter cells.

[0030] In certain embodiments, the first, second and / or third population of desired cells is a population of cancer cells.

[0031] In certain embodiments, the method further comprises using a declumping agent.

[0032] In certain embodiments, the declumping agent is beta mercaptoethanol.

[0033] In certain embodiments, the method further comprises processing the non-magnetic droplets removed in (5) to increase the yield of magnetic droplets. This can be accomplished by performing the method again using the non-magnetic droplets removed in (5).

[0034] In certain embodiments, the method further comprises processing the magnetic droplets removed in (6) to increase the purity of magnetic droplets. This can be accomplished by performing the method again using the non-magnetic droplets removed in (6).

[0035] For droplet-based assays of single cells, one of the primary variables that affects the screening throughput is the percentage of droplets that contain cells. The percentage of droplets that contain cells can be predicted by a Poisson distribution and is proportional to the cell concentration and volume of the droplet. However, it is often not possible to simply increase the concentration of cells or increase the size of the droplet. Increasing the concentration of cells can lead to significant clogging of the microfluidic channel, preventing droplet production. Additionally, high cell concentrations will lead to droplets containing more than 1 cell of a single type, which can lead to assay fidelity concerns in some assays. For example, having a droplet with two cells producing two different antibodies can lead to a false positive if the droplet is sorted.

[0036] Increasing the size of the droplet will also increase the percentage of droplets containing more than 1 cell of a single type. In addition, droplet sorters have a maximum droplet size they are compatible with. Going beyond that droplet size will lead to a decrease in sorting efficiency, defined as the percentage of attempted sorts that are completed, and can also lead to clogging and other issues. Larger droplets can also require lower sort rates to be sorted, such as for the case when droplets are sorted with FACS and larger droplets require a larger nozzle that operates at lower maximum sort rates. Furthermore, larger droplets will have a lower concentration of key biomolecules released from cells (e.g., antibodies, cytokines, etc) or fluorescent molecules in the bulk droplet, both of which can decrease the ability to detect signal.

[0037] To address these needs, an improved system and methods (sometimes referred to herein as dMACS; magnetic activated cell sorting in droplets) are described to enrich for droplets that contain cells, allowing one to use a lower initial cell concentration and / or smaller droplets to obtain increased percentages of droplets containing single cells. This provides improvements as compared to existing technology as existing technologies have attempted to solve the current problems using the following modifications but have encountered the following limitations:

[0038] Increase cell concentration: increasing cell concentration can lead to device clogging and / or more than 1 cell per droplet. Further, cells used may be rare and / or expensive (e.g., B cells from a mouse) and it will not be possible to increase the cell concentration sufficiently high.

[0039] Increase droplet size to increase average cells per drop: increasing the droplet size could lower the sorting efficiency of the droplets. Most droplet sorters are only recommended to be used up to a certain droplet size. Additionally, increasing the droplet size can lower the sort rate of the droplets and thus decrease the maximum throughput of the screen. Furthermore, increasing the droplet size can lower the concentration of key biomolecules in the droplet, such as antibodies produced by a cell.

[0040] Inertial microfluidics: inertial microfluidics has been used to increase the percentage of droplets that contain single cells. However, the technique appears unreliable between experiments, evidenced by its lack of widespread adoption in the field. It also requires high flowrates, which are incompatible with formation of some droplet types.

[0041] Sort droplets with droplet sorter: a droplet sorter can be used to increase the percentage of droplets that have single cells. However, droplets must be screened in series, which takes considerable time. The techniques described herein are designed to enrich for droplets containing cells ahead of a droplet sorter to increase the throughput of the screen.

[0042] Certain aspects of the invention are summarized below, which provide improvements over methods currently used.

[0043] General protocol: Attach magnetic bead(s) to cells; remove surplus beads from solution; partition cells into droplets; move droplet across magnetic field; collect non-magnetic droplets (e.g., as waste); remove remaining droplets from magnetic field; and collect enriched droplets containing cells.

[0044] Choice of cell concentration

[0045] • 2 or more cell screens

[0046] • Cell concentration can be chosen to optimize the percentage of droplets that contain both cell types after dMACS. In general, the cell type not labeled with a magnetic bead should be used at high concentration, while the cell type labeled with the magnetic bead can be used at low or moderate concentration. The percentage of co-loaded droplets can be reasonably predicted using a Poisson distribution based on cell concentration and droplet size, then the droplets that do not contain the labeled cell can be removed from the distribution and the percentages recalculated of droplets that contain each of the different possible combinations of cells.

[0047] • Choice of beads

[0048] • The beads used must bind to cells. The magnetic bead bound to the cell are chosen, e.g., to bind with receptors on the cell surface, an antibody on the cell surface, a modification made to the cell surface, etc.

[0049] • To demonstrate this concept, cells were used that were labeled with biotin and beads coated with streptavidin were used. NHS- biotin was used to biotinylate the cells.

[0050] • Other means to label cells with beads can also be used.

[0051] • The beads used should not clump and bind to other beads. Bead clumping can result in cell clumping. The media is selected to avoid having components that will lead to cell clumping. Cell and bead clumping can lead to clogging of the microfluidic channel when making droplets and lead to more than one cell per droplet.

[0052] • The beads used are used at a sufficiently high concentration such that the cells have a sufficient number of magnetic beads so that they will be sorted in the droplet when placed into the magnetic field. The concentration of beads should not be so high such that free beads are in solution, not bound to cells, as these can also be encapsulated into droplets and sorted, thereby decreasing the purity of the final sample. In certain embodiments, using smaller beads helps prevent unbound beads from being sorted in droplets. It has been demonstrated that increasing cell concentration during bead binding to cells can greatly help decrease the number of free beads in solution.

[0053] • The beads used are sufficiently magnetizable so that they can sort cells when in droplets. This can be controlled, e.g., by increasing the percentage of magnetic material in the bead, increasing the strength of the magnetic material, using larger beads, etc.

[0054] • dMACS was successful when cells were labeled with Stemcell Streptavidin Rapidspheres, but not with Miltenyi Streptavidin particles. The Stemcell particles are on the order of 1 um in size while the Miltenyi particles are on the order of 1 - 100 nm in size. Therefore, the Stemcell particles were likely much more magnetizable, which helped with the magnetic sorting of the droplets.

[0055] • Removing surplus beads from solution

[0056] • Labeling cells with beads with cells at a very high concentration and beads at a minimal concentration has led to improved success in preventing excess beads from being in solution. For example, cells were labeled at 1.5E8 cells / ml in 100 pl of media (PBS + 2 mM EDTA + 0.1% beta-mercaptoethanol) and used 2 pl (~2% v / v) of Streptavidin Rapidsphere beads. The cell suspension was washed into 1 ml of cell growth media for a final cell concentration of 1.5E7 cells / ml.

[0057] • Using higher concentration of beads (10% v / v) and / or lower concentrations of cells (5E7) during labeling led to higher amounts of free beads in solution and lower purity of droplets containing cells after magnetic selection.

[0058] • Other methods to filter out unlabeled beads, e.g., with inertial microfluidics or microfluidic Deterministic Lateral Displacement (DLD), may also be used. It may also be feasible to filter out unlabeled beads using a size-exclusion filter or density gradient.

[0059] • Droplets • Various types of droplets can be used with dMACS e.g., single emulsion droplets, double emulsion droplets, nanovials, hydrogels, core-shell particles, semi-permeable capsules, etc).

[0060] • dMACS has been successfully used with a type of double emulsion droplet, demonstrating that this system can be translated to other droplet types.

[0061] • The droplets used must be sufficiently stable to enable sorting via magnetic field. Unstable droplets could be broken when the cells are pulled against the magnetic field, which leads to cells pulling against the droplet to move the droplet. Droplet stability can be adjusted, e.g., with surfactants, medias, size, etc.

[0062] • The droplets used must be of a suitable size for sorting on the magnetic column. Smaller droplets will be easier to sort than larger droplets, due to their decreased mass and drag.

[0063] • dMACS has been successfully used on a magnetic column purchased from Miltenyi that is meant for cells up to 50 pm in size, using droplets around 55 pm in size.

[0064] • The dMACS methods, systems and assays described herein are also applicable for use with nanovial technology and with hydrogel bead technology.

[0065] • Choice of magnetic column

[0066] • The magnetic field strength must be high enough to sort droplets containing magnetic beads bound to cells.

[0067] • A Miltenyi Large Cell magnetic column was used successfully, which has a very high magnetic field strength compared to other magnetic columns from Stemcell (EasySep) or Thermofisher (Dynamag). The Miltenyi column has the particles travel tightly in between magnetized column beads, where magnetic particles will get trapped in the column and non magnetic particles will flow through. In contrast, the Stemcell and Thermofisher columns are essentially magnets that capture particles on the side or bottom of the column. • The magnetic column should ideally have sufficient space to store all positively selected droplets containing cells, or else the yield will be negatively affected.

[0068] • A high initial cell loading protocol was used, wherein dMACS led to 37% of sorted droplets contained a cell labeled with a magnetic bead. In this case, the Miltenyi Large Cell column could not hold all of the positively selected droplets, leading to the negative pass- through fraction containing a large number of droplets containing cells and lower yield. The Miltenyi column is said to hold about 80 ul of liquid and about 500 ul of droplets was passed through it, of which 185 ul (37%) should have a cell and be captured. A larger column that holds more droplets would be a good way to increase the yield. Alternatively, repeating dMACS additional times on the flow-through fraction is a way to increase the yield, which has also been demonstrated. This process can be automated, such as with the AutoMACS system or another custom system.

[0069] • Use for more than 1 cell assays

[0070] • 2 cell screens

[0071] • For screens that require two cells per droplet, droplets can be selected for those containing 2 cells with dMACS. One cell should labeled with magnetic beads, while the other cell type is unlabeled. Droplets containing the bead-labeled cell will be enriched for, which will also enrich for the droplets containing both cell types. This effect will be enhanced if the bead-labeled cells are at low concentration and the unlabeled cells are at high concentration, as a high percentage of the droplets containing a bead-labeled cell will also contain an unlabeled cell. This has been demonstrated in experiments and mathematically.

[0072] • Both cells could also be labeled with beads, and magnetic bead concentration or magnetizability could be selected so as to only sort droplets containing 2 cells, to better enrich for 2 cell assays.

[0073] • dMACS could also be performed separately on droplets for each cell type, resulting in pure populations of droplets containing cell 1 and droplets containing cell 2. These droplets could then be merged using common microfluidic droplet merging methods, resulting in a high population of co-loaded droplets. This could work for 3 or more cells per droplet, too.

[0074] • 3 or more cell screens

[0075] • For screens which require two cells per droplet, droplets can also be selected for droplets containing 3 or more cells, following the same principals as the 2 cell screen.

[0076] • Cell type(s)

[0077] • dMACs is applicable to any cell type to which magnetic beads can be bound.

[0078] • Other component(s), whether they are inert or active, excipient or active pharmaceutical ingredient, may also be used.

[0079] • Media viscosity during magnetic selection could be modified so as to give droplets more or less time to pass by the magnet, which could affect purity.

[0080] • dMACS can be used in assays or screens that have even higher throughput or to enable higher throughput.

[0081] The invention will now be illustrated by the following non-limiting Examples.

[0082] Example 1.

[0083] To test single cell loading into droplets, dMACS bead-based enrichment will be evaluated in this Example.

[0084] To stain Jurkat NFAT-GFP cells, cells are gently pipetted several times to reduce cell clumping. Jurkat NFAT-GFP cells are then counted, centrifuged at 300 x gravity (g) at 20 degrees Celsius and resuspended in a solution of Phosphate Buffered Saline (PBS) containing a dilution (1 to 1000) of CellTrace Yellow (CTY). Jurkat NFAT-GFP cells are placed at 20 degrees Celsius in the absence of light for 20 minutes rotating on a rotary mixer. Following incubation, an equivalent volume of RPMI containing 10% FBS is added to quench the dye. This suspension is incubated for 5 minutes at 20 degrees Celsius in the absence of light. Cells are spun down at 300 x g for 5 minutes at 20 degrees Celsius and resuspended in ,5-lmL of RPMI containing 10% FBS and counted.

[0085] Following CTY incubation, Jurkat NFAT-GFP cells are biotinylated. Jurkat NFAT-GFP cells are washed with 10 milliliters (mL) of cold PBS and centrifuged at 300 x gravity (g) for 5 minutes at 20 degrees Celsius. Jurkat NFAT-GFP cells are washed for a second time with lOmL of cold PBS and centrifuged at 300 x gravity (g) for 5 minutes at 20 degrees Celsius. lOmL of cold Sulfo-NHS-biotin PBS solution at a concentration of .1 milligram (mg) per mL is added to the pellet of the second wash. Cells are incubated at 4 degrees Celsius for 10 minutes in the absence of light. Quenching is then performed with PBS containing 2 percent FBS and 1 mM EDTA. Quenched cells are then centrifuged at 300 x g at 20 degrees Celsius for 5 minutes. The supernatant is discarded, and the pellet is washed with lOmL of PBS with 2 percent FBS and ImM EDTA. Prior to centrifuging the cells at this step, cells are counted to determine cell concentration for bead labeling. Cells are then centrifuged at 300 x g at 20 degrees Celsius for 5 minutes and resuspended in PBS with 2 percent FBS and 1 mM EDTA at a concentration of 50,000,000 cells per mL in a final volume of 684 pl.

[0086] Jurkat NFAT-GFP cells are then labeled with beads, Stemcell Streptavidin Rapidspheres, that bind to the biotin on the surface of the cells. Beads are added at 2% by volume to the cells (13.68 pl of beads to 684 pl of cells) and allowed to bind the cells by incubating for 15 minutes at 20 degrees Celcius. Cells are then diluted lOx to dilute excess beads by adding 6156 pl of total media (5472 pl of RPMI with 10 percent FBS and 684 pl of Optiprep). Because cells and beads were clumpy, which would lead to clogging the microfluidic channel, the cell and bead suspension was passed through a 70 pm filter to yield the solution “Aql”.

[0087] To produce the droplets, Aql (5,000,000 per ml Jurkat NFAT-GFP cells labeled with CTY and beads in RPMI containing 10 percent FBS with 10 percent Optiprep) and Aq2 (OKT3 antibody at lOnM in RPMI containing 10 FBS with 10 percent Optiprep) solutions are mixed in a microfluidic chip using 2 percent Poloxamer 188 in RPMI containing 10 percent FBS as sheath fluid and 2% RAN-008 surfactant in HFE-7500 as the oil.

[0088] To prepare the dMACS column for cell enrichment, a Miltenyi Large Cell Separation Column is placed on the magnet. A 23-gauge needle is attached to the column to serve as a flow resistor. 500 pL of PBS is added to the top of the column to wet the beads in the column, and the flow through is collected in a FACS tube and discarded.

[0089] A new FACS tube is placed under the column to collect the negative fraction of cells added to the column. Droplets are dispersed with a Pl 000 pipette and 500-1000 pL of the droplets suspension are added to the top of the column. The cell suspension is gently mixed with the Pl 000 pipette at the top of the column during flow through to prevent droplet settling and clogging. After the droplet suspension has flowed through the column, the column is washed three times with 500 pL of 2 percent Poloxamer 188 (Pl 88) in RPMI containing 10 percent FBS. All wash flow through is collected in the negative fraction FACS tube at the bottom of the column.

[0090] Once the flow stops after the third wash, the column is removed from the magnet and the negative fraction FACS tube is replaced with a new FACS tube to be used to collect the positive fraction. The 23-gauge needle flow resistor is removed from the magnetic column. ImL of 2 percent Pl 88 in RPMI containing 10 percent FBS is added to the top of the column and immediately the column plunger is used in a slow and controlled manner to push the remaining droplets in the column into the positive fraction FACS tube, ensuring that excess air is not pushed through the column at the end of collection to minimize bubble formation.

[0091] Droplets at each stage of selection were imaged on the EVOS M7000 microscope. Positive, negative, and unenriched fractions of the droplet flow through were analyzed for the detection of CTY-labeled cell loaded droplets. The unenriched fraction that was not loaded onto the column had 3% (13 out of 431 droplets) of droplets with CTV-labeled cells. Positive selection increased this percentage to 23.6% (145 out of 614 droplets), a 7.8-fold increase over the unenriched fraction. The negative fraction had 0.5% (2 out of 393 droplets) of droplets with CTV-labeled cells. dMACS enrichment increases the occurrence of cell-loaded droplets, however cell loading and cell and bead clumping will be decreased in further iterations to increase the percentage of droplets containing cells. Notably, bead clumps were also enriched for in the positive fraction after dMACS and thus decreasing bead clumping should increase positive selection for cell-containing droplets.

[0092] Example 2.

[0093] To decrease cell clumping and increase the amount of cells that can be loaded into droplets, beta mercaptoethanol was tested as a declumping agent. In this experiment beta mercaptoethanol was added to the media after staining, while in later experiments this was optimized to have declumping occur during staining.

[0094] To stain Jurkat NFAT-GFP cells, cells are gently pipetted several times to reduce cell clumping. Jurkat NFAT-GFP cells are then counted, centrifuged at 300 x gravity (g) at 20 degrees Celsius and resuspended in a solution of Phosphate Buffered Saline (PBS) containing a dilution (1 to 1000) of CellTrace Yellow (CTY). Jurkat NFAT-GFP cells are placed at 20 degrees Celsius in the absence of light for 20 minutes rotating on a rotary mixer. Following incubation, an equivalent volume of RPMI containing 10 percent FBS is added to quench the dye. This suspension is incubated for 5 minutes at 20 degrees Celsius in the absence of light. Cells are spun down at 300 x g for 5 minutes at 20 degrees Celsius and resuspended in ,5-lmL of RPMI containing 10 percent FBS and counted.

[0095] Two conditions were tested. In condition 1, cells were subsequently treated with 0.1 percent beta mercaptoethanol for 15 minutes at room temperature in RPMI containing 10 percent FBS. In condition 2, cells were not treated with beta mercaptoethanol. A 70 pm filter was used to remove remaining large clumps from each condition. Cells in each condition were subsequently resuspended into “Aql” media. For condition 1, Aql consisted of 5,000,000 per ml Jurkat NFAT-GFP cells labeled with CTY and beads in RPMI containing 10 percent FBS with 10 percent Optiprep and 0.1 percent beta mercaptoethanol. For condition 2, Aql consisted of 5,000,000 per ml Jurkat NFAT-GFP cells labeled with CTY and beads in RPMI containing 10 percent FBS with 10 percent Optiprep.

[0096] To produce droplets, Aql and Aq2 (OKT3 antibody at lOnM in RPMI containing 10 FBS with 10 percent Optiprep) solutions are mixed in a microfluidic chip using 2 percent Poloxamer 188 in RPMI as sheath fluid and 2% RAN-008 surfactant in HFE-7500 as the oil.

[0097] The amount of time droplets could be made for condition 1 and 2 was observed. A new channel was used for each condition. For condition 1, where cells were treated with beta mercaptoethanol, droplets could be made for over 20 minutes. The channel did not clog and did not show signs of clogging soon. In contrast, for condition 2, where cells were not treated with beta mercaptoethanol, the channel clogged by 10 minutes of droplet making. This result demonstrated that beta mercaptoethanol improved the ability to minimize clogging of cells during droplet creation.

[0098] Example 3.

[0099] To increase cell co-loading within droplets, dMACS bead-based enrichment will be evaluated in this Example.

[0100] To label Jurkat NFAT-GFP cells, cells are gently pipetted several times to reduce cell clumping. Jurkat NFAT-GFP cells are then counted, centrifuged at 300 x gravity (g) at 20 degrees Celsius and resuspended in a solution of Phosphate Buffered Saline (PBS) and a dilution (1 to 1000) of CellTrace Yellow (CTY) or a dilution (1 to 500) of CellTrace Violet (CTV). Jurkat NFAT-GFP cells are placed at 4 degrees Celsius in the absence of light for 20 minutes. This was found to decrease cell clumping as compared to 20 degrees Celsius.

[0101] Following CTY incubation, labeled Jurkat NFAT-GFP cells are washed with 10 milliliters (mL) of cold PBS and centrifuged at 300 x gravity (g) at 20 degrees Celsius. Jurkat NFAT-GFP cells are washed for a second time with lOmL of cold PBS and centrifuged at 300 x gravity (g) at 20 degrees Celsius. lOmL of cold Sulfo-NHS-biotin PBS solution at a concentration of .1 milligram (mg) per mL is added to the pellet of the second wash. Cells are incubated at 4 degrees Celsius for 10 minutes in the absence of light. Quenching is then performed with a solution of RPMI containing 10 percent FBS. Quenched cells are then centrifuged at 300 x g at 20 degrees Celsius for 5 minutes. The supernatant is discarded and the pellet is washed with lOmL of PBS with 2mM EDTA. Prior to centrifuging the cells at this step, cells are counted to determine cell concentration for bead labeling. Cells are then centrifuged at 300 x g at 20 degrees Celsius for 5 minutes and resuspended in PBS with 2 mM EDTA and 0.1 percent beta mercaptoethanol at a concentration of 50,000,000 cells per mL.

[0102] To bead label the CTY-labeled Jurkat NFAT-GFP cells, a 2 percent by volume solution of Stemcell Streptavidin Rapidspheres is added to cell suspensions. Cells are incubated with the Rapidspheres for 15 minutes at 20 degrees Celsius. Cells are then diluted to one tenth of the cell labeling concentration during resuspension in 10% Optiprep in RPMI containing 10 percent FBS and 0.1 percent beta mercaptoethanol.

[0103] To produce droplets, Aql (5,000,000 per ml Jurkat NFAT-GFP cells labeled with CTY and beads in RPMI containing 10 percent FBS with 10 percent Optiprep, 0.1 percent beta mercaptoethanol) and Aq2 (10,000,000 per ml Jurkat NFAT-GFP cells labeled with CTV in RPMI containing 10 percent FBS with 10 percent Optiprep, 0.1 percent beta mercaptoethanol) solutions are mixed in a microfluidic chip using 2 percent Pol oxamer 188 in RPMI containing 10 percent FBS as sheath fluid and 2% RAN-008 in HFE-7500 as oil.

[0104] To prepare the dMACS column for cell enrichment, a Miltenyi Large Cell Separation Column is placed on the magnet. A 23-gauge needle is attached to the column to serve as a flow resistor. 500 pL of PBS is added to the top of the column to wet the beads in the column, and the flow through is collected in a FACS tube and discarded.

[0105] A new FACS tube is placed under the column to collect the negative fraction of cells added to the column. Droplets are dispersed with a Pl 000 pipette and 500-1000 pL of the droplets suspension are added to the top of the column. The cell suspension is gently mixed with the Pl 000 pipette at the top of the column during flow through to prevent droplet settling and clogging. After the droplet suspension has flowed through the column, the column is washed three times with 500 pL of 2 percent Poloxamer 188 in RPMI containing 10 percent FBS. All wash flow through is collected in the negative fraction FACS tube at the bottom of the column.

[0106] Once the flow stops after the third wash, the column is removed from the magnet and the negative fraction FACS tube is replaced with a new FACS tube to be used to collect the positive fraction. The 23-gauge needle flow resistor is removed from the magnetic column. ImL of 2 percent Pl 88 in RPMI containing 10 percent FBS is added to the top of the column and immediately the column plunger is used in a slow and controlled manner to push the remaining droplets in the column into the positive fraction FACS tube, ensuring that excess air is not pushed through the column at the end of collection to minimize bubble formation.

[0107] Droplets at each stage of selection were imaged on the EVOS M7000 microscope.

[0108] Positive, negative, and unenriched fractions of the droplet flow through were analyzed for the detection of CTV and CTY co-loaded droplets. The unenriched fraction that was not loaded onto the column had 0.4% (3 out of 752 droplets) of droplets with both CTV and CTY-labeled cells. Positive selection increased this percentage to 7% (18 out of 258 droplets), a 17.5-fold increase over the unenriched fraction. The negative fraction had 0.0% (0 out of 336 droplets) of droplets with CTV and CTY-labeled cells. dMACS enrichment increases the occurrence of co-loaded droplets; however total cell numbers will be increased in the next interaction to increase the percentage of co-loaded droplets post-enrichment.

[0109] Example 4.

[0110] To increase cell co-loading within droplets, increased cell loading, increased Optiprep percentage, and a lower bead concentration with higher cell concentration during labeling is evaluated in this Example.

[0111] To label Jurkat NFAT-GFP cells, cells are gently pipetted several times to reduce cell clumping. Jurkat NFAT-GFP cells are then counted, centrifuged at 300 x gravity (g) at 20 degrees Celsius and resuspended in a solution of Phosphate Buffered Saline (PBS) containing 0.1 percent beta mercaptoethanol and a dilution (1 to 1000) of CellTrace Yellow (CTY) or a dilution (1 to 500) of CellTrace Violet (CTV). Jurkat NFAT-GFP cells are placed at 4 degrees Celsius in the absence of light for 20 minutes.

[0112] Following CTY incubation, labeled Jurkat NFAT-GFP cells are washed with 10 milliliters (mL) of cold PBS and centrifuged at 300 x gravity (g) at 20 degrees Celsius. Jurkat NFAT-GFP cells are washed with 10 milliliters (mL) of cold PBS and centrifuged at 300 x gravity (g) at 20 degrees Celsius. Jurkat NFAT-GFP cells are washed for a second time with lOmL of cold PBS and centrifuged at 300 x gravity (g) at 20 degrees Celsius. lOmL of cold Sulfo-NHS-biotin PBS solution at a concentration of .1 milligram (mg) per mL is added to the pellet of the second wash. Cells are incubated at 4 degrees Celsius for 10 minutes in the absence of light. Quenching is then performed with a solution of RPMI containing 10 percent FBS. Quenched cells are then centrifuged at 300 x g at 20 degrees Celsius for 5 minutes. The supernatant is discarded and the pellet is washed with lOmL of PBS with 2mM EDTA. Prior to centrifuging the cells at this next step, cells are counted to determine cell concentration for bead labeling. Cells are then centrifuged at 300 x g at 20 degrees Celsius for 5 minutes and resuspended in PBS with 2 mM EDTA and 0.1 percent BME at a concentration of 150,000,000 cells per mL.

[0113] To bead label the CTY-labeled Jurkat NFAT-GFP cells, a 2 percent by volume solution of Stemcell Streptavidin Rapidspheres is added to the cell suspension. Cells are incubated with the Rapidspheres for 15 minutes at 20 degrees Celsius. 2mL of RPMI containing 10 percent FBS is added to wash the cell suspension. Cells are then imaged on the EVOS M7000 microscope prior to centrifuging cells at 300 x g for 5 minutes at 20 degrees Celsius and removing the supernatant. Cells are then diluted to one tenth of the cell labeling concentration during resuspension in 15% Optiprep in RPMI containing 10 percent FBS.

[0114] To produce droplets, Aql (15,000,000 Jurkat NFAT-GFP cells labeled with CTY and beads in RPMI containing 10 percent FBS with 15 percent Optiprep, 0.1 percent BME) and Aq2 (20,000,000 Jurkat NFAT-GFP cells labeled with CTV in RPMI containing 10 percent FBS with 15 percent Optiprep, 0.1 percent BME) solutions are mixed in a microfluidic chip using 2 percent Poloxamer 188 in RPMI containing 10 percent FBS as sheath fluid and 2 percent RAN- 008 surfactant in HFE-7500 as oil.

[0115] To prepare the dMACS column for cell enrichment, a Miltenyi Large Cell Separation Column is placed on the magnet. A 23-gauge needle is attached to the column to serve as a flow resistor. 500 pL of PBS is added to the top of the column to wet the beads in the column, and the flow through is collected in a FACS tube and discarded.

[0116] A new FACS tube is placed under the column to collect the negative fraction of cells added to the column. Droplets are dispersed with a Pl 000 pipette and 500-1000 pL of the droplets suspension are added to the top of the column. The cell suspension is gently mixed with the Pl 000 pipette at the top of the column during flow through to prevent droplet settling and clogging. After the droplet suspension has flowed through the column, the column is washed three times with 500 pL of 2 percent Poloxamer 188 in RPMI containing 10 percent FBS. All wash flow through is collected in the negative fraction FACS tube at the bottom of the column.

[0117] Once the flow stops after the third wash, the column is removed from the magnet and the negative fraction FACS tube is replaced with a new FACS tube to be used to collect the positive fraction. The 23-gauge needle flow resistor is removed from the magnetic column. ImL of 2 percent Pl 88 in RPMI containing 10 percent FBS is added to the top of the column and immediately the column plunger is used in a slow and controlled manner to push the remaining droplets in the column into the positive fraction FACS tube, ensuring that excess air is not pushed through the column at the end of collection to minimize bubble formation.

[0118] Droplets at each stage of selection were imaged on the EVOS M7000 microscope. Positive, negative, and unenriched fractions of the droplet flow through were analyzed for the detection of CTV and CTY co-loaded droplets. The unenriched fraction that was not loaded onto the column had 14.7% (53 out of 360 droplets) of droplets with both CTV and CTY-labeled cells and 36.9% (133 out of 360 droplets) of droplets with CTY-labeled cells containing beads. Positive selection increased this percentage of co-loaded droplets to 33.9% (100 out of 295 droplets), a 2.3-fold increase over the unenriched fraction, and CTY-labeled cells in droplets to 88.5% (261 out of 295 droplets), a 2.4-fold increase over the unenriched fraction. This result shows that dMACS enrichment increases the occurrence of co-loaded droplets which can enable larger scale antibody screens.

[0119] The negative fraction had 11.7% (40 out of 342 droplets) of droplets with CTV and CTY-labeled cells, a 0.8-fold decrease compared to the unenriched fraction, and 23.4% (80 out of 342 droplets) of droplets with CTY-labeled cells, a 0.6-fold decrease compared to the unenriched fraction, indicating a lower yield for this experiment despite the high purity in the positive fraction. This result is in contrast to previous results with lower cell loading and indicates that the magnetic column used to trap magnetically -labeled cells in droplets was not large enough for the higher number of droplets that could be captured in this experiment.

[0120] To determine whether the yield could be increased with multiple rounds of dMACS, a second round of dMACS was performed using the droplets from the negative fraction as the input. Positive selection increased this percentage of co-loaded droplets to 35.5% (93 out of 262 droplets), a 3-fold increase over the input fraction, and CTY-labeled cells in droplets to 80.5% (93 out of 262 droplets), a 3.4-fold increase over the input fraction. The negative fraction had a percentage of co-loaded droplets at 7.6% (27 out of 355 droplets), a 0.7-fold decrease compared to the input fraction, and percentage of CTY-labeled cells in droplets at 17.2% (61 out of 355 droplets), a 0.7-fold decrease compared to the input fraction. Because positive selection continued to enrich for droplets with cells and the negative fraction continued to have a lower percentage of droplets with cells, it is believed that multiple rounds of dMACS could be used to increase the yield of cell-containing droplets in the case of high initial cell loading. Alternatively, a larger magnetic column could be used to increase the maximum number of droplets that can be captured in a single round of dMACS.

[0121] Example 5.

[0122] A different bead type is evaluated in this Example: Miltenyi Streptavidin Microbeads. Miltenyi Streptavidin Microbeads are much smaller than the Stemcell Streptavidin Rapidspheres used in the other Examples and may have decreased clumping and decreased propensity to be sorted in droplets when not on a cell.

[0123] To label Jurkat NFAT-GFP cells, cells are gently pipetted several times to reduce cell clumping. Jurkat NFAT-GFP cells are then counted, centrifuged at 300 x gravity (g) at 20 degrees Celsius for 5 minutes and resuspended in a solution of Phosphate Buffered Saline (PBS) and a dilution (1 to 1000) of CellTrace Yellow (CTY). Jurkat NFAT-GFP cells are placed on ice in the absence of light for 20 minutes. An equivalent volume of RPMI with 10 percent FBS is then added to quench the dye, after which the cell suspension is centrifuged at 300 x g at 20 degrees Celcius for 5 minutes and the supernatant is removed.

[0124] Labeled Jurkat NFAT-GFP cells are then washed with 10 milliliters (mL) of cold PBS and centrifuged at 300 x gravity (g) at 20 degrees Celsius. Jurkat NFAT-GFP cells are washed for a second time with lOmL of cold PBS and centrifuged at 300 x gravity (g) at 20 degrees Celsius. lOmL of cold Sulfo-NHS-biotin PBS solution at a concentration of .1 milligram (mg) per mL is added to the pellet of the second wash. Cells are incubated at 4 degrees Celsius for 10 minutes in the absence of light. Quenching is then performed with a solution of RPMI containing 10 percent FBS. Quenched cells are then centrifuged at 300 x g at 20 degrees Celsius for 5 minutes. The supernatant is discarded and the pellet is washed with lOmL of PBS with 2mM EDTA. Prior to centrifuging the cells at this next step, cells are counted to determine cell concentration for bead labeling. Cells are then centrifuged at 300 x g at 20 degrees Celsius for 5 minutes and resuspended in PBS with 2 mM EDTA and 0.1 percent BME at a concentration of 100,000,000 cells per mL.

[0125] Two bead labeling conditions were tested, with condition 1 using a 10 percent by volume solution of beads and condition 2 using a 50 percent by volume solution of beads. In condition 1, to bead label the CTY-labeled Jurkat NFAT-GFP cells, Miltenyi Streptavidin Microbeads are added to the cell suspension to yield a 10 percent by volume solution of beads. Cells and beads are mixed well and are incubated with the Microbeads for 15 minutes at 4 degrees Celsius. 2mL of RPMI containing 10 percent FBS is added to wash the cell suspension. Cells are then centrifuged at 300 x g for 5 minutes at 20 degrees Celsius and the supernatant is removed. Cells are then diluted to 5,000,000 cells per mL in media of 10 percent Optiprep and 0.1 percent BME in RPMI containing 10 percent FBS to yield Aql for condition 1. In condition 2, to bead label the CTY-labeled Jurkat NFAT-GFP cells, Miltenyi Streptavidin Microbeads are added to the cell suspension to yield a 50 percent by volume solution of beads. Cells and beads are mixed well and are incubated with the Microbeads for 15 minutes at 4 degrees Celsius. 2mL of RPMI containing 10 percent FBS is added to wash the cell suspension. Cells are then centrifuged at 300 x g for 5 minutes at 20 degrees Celsius and the supernatant is removed. Cells are then diluted to 4,360,000 cells per mL in media of 10 percent Optiprep and 0.1 percent BME in RPMI containing 10 percent FBS to yield Aql for condition 2.

[0126] To produce droplets for each condition 1 and 2, Aql and Aq2 (10 nM OKT3 antibody in RPMI containing 10 percent FBS with 10 percent Optiprep) solutions are mixed in a microfluidic chip using 2 percent Pol oxamer 188 in RPMI as sheath fluid and 2 percent RAN- 008 surfactant in HFE-7500 as oil.

[0127] To prepare the dMACS column for cell enrichment, a Miltenyi Large Cell Separation Column is placed on the magnet. A 23-gauge needle is attached to the column to serve as a flow resistor. 500 pL of PBS is added to the top of the column to wet the beads in the column, and the flow through is collected in a FACS tube and discarded.

[0128] A new FACS tube is placed under the column to collect the negative fraction of cells added to the column. Droplets are dispersed with a Pl 000 pipette and 500-1000 pL of the droplets suspension are added to the top of the column. The cell suspension is gently mixed with the Pl 000 pipette at the top of the column during flow through to prevent droplet settling and clogging. After the droplet suspension has flowed through the column, the column is washed three times with 500 pL of 2 percent Pol oxamer 188 in RPMI. All wash flow through is collected in the negative fraction FACS tube at the bottom of the column.

[0129] Once the flow stops after the third wash, the column is removed from the magnet and the negative fraction FACS tube is replaced with a new FACS tube to be used to collect the positive fraction. The 23-gauge needle flow resistor is removed from the magnetic column. ImL of 2 percent Pl 88 in RPMI is added to the top of the column and immediately the column plunger is used in a slow and controlled manner to push the remaining droplets in the column into the positive fraction FACS tube, ensuring that excess air is not pushed through the column at the end of collection to minimize bubble formation.

[0130] For both condition 1 and condition 2, neither concentration of labeling beads led to an enrichment of droplets containing cells in the positive fraction, with almost all of the droplets ending up in the negative fraction. This result indicates that the Miltenyi Streptavidin Microbeads were not appropriate for use with dMACS, likely due to their smaller size and lower magnatizability.

[0131] Example 6.

[0132] This example and associated Figures provide optimizations to dMACS. It was determined that the final co-loaded droplet percentage after dMACS is dominated by the concentration of unlabeled cells. The concentration of labeled cells minimally affects the final percentage of co-loaded droplets; however, it does affect the yield. The calculations assume drops with an inner diameter of 45 pm and perfect yield and purity of dMACS. The calculations used to make these Figures ignored droplets that have more than 2 cells of each type.

[0133] As shown in Figure 5, the final concentration of labeled cells does not significantly change the final percentage of co-loaded drops with dMACS. Instead, the concentration of unlabeled cells is the most important parameter.

[0134] To create these figures, a spreadsheet was created to calculate the cell loading of a 45 pm inner diameter drop from the final cell concentrations in the drop. To do this, first the average number of cells per drop was calculated by multiplying the concentration of each cell type (unlabeled or labeled) by the volume of the drop, giving the average number of cells per drop X unlabeled and ^ labeled. Each X value can then be plugged into a Poisson distribution to calculate the percentage of total droplets with each combination of cell numbers; for example, what percentage of drops has 0 unlabeled cells and 1 labeled cells, or 2 unlabeled cells and 1 labeled cells. These numbers were calculated up to 2 cells per drop of each type. Droplets containing 1 or 2 of both cell types were designated as “co-loaded”, while those containing 0 of either cell type were not. Droplets containing more than 2 of either cell type were not included in the calculations. Figure 5 was made by setting the concentration of labeled cells constant and varying the concentration of unlabeled cells from 1E6 cells / ml to 2E7 cells / ml. These cell concentrations are reasonably achieved within droplets.

[0135] Example 7.

[0136] As demonstrated mathematically herein, the concentration of unlabeled cells matters more than the concentration of labeled cells. Therefore one can design microfluidic droplet generation channels that prioritize loading unlabeled cells at very high concentrations over loading labeled cells. In a traditional channel design, two different cell types will meet at a junction prior to being encapsulated into droplets. Figure 6A shows this for the case where the droplets are double emulsions. Here, the flow rate Q of each stream containing the different cell type is typically equal, and cells are loaded at approximately the same concentration. The rationale behind this is that typically one wants a similar concentration of each cell in the final drop.

[0137] In an optimized channel design for dMACS, where the unlabeled cell concentration should be increased, the inlet can be modified so as to increase the flow volume devoted to the unlabeled cells. The flowrate Q unlabeled of the unlabeled cells can be increased above the flowrate Q labeled of the labeled cells, devoting more of the droplet to the unlabeled cells.

[0138] To help visualize this, an example is presented where both cells are loaded at 1E7 cells / ml. If the flowrates are equal, the final droplet volume will contain 50% of the volume from the unlabeled cell stream and 50% of the volume from the labeled cell stream. The final concentrations of each cell type will be equal at 5E6, or 50% of the original concentration. If instead Q unlabeled is 9-times higher than Q labeled, the final droplet volume will contain 90% of the volume from the unlabeled cell stream and 10% of the volume from the labeled cell stream. The final concentration of unlabeled cells will be 9E6 and the labeled cells will be 1E6, thus increasing the concentration of unlabeled cells as desired. Furthermore, the original concentration of unlabeled cells should be substantially higher than the labeled cells in order to maximize performance. One of the major issues with increasing the concentration of cells is clogging, which occurs when too many cells are pushed through too small of a volume. By increasing the flow volume devoted to unlabeled cells, this issue can be minimized and the concentration of unlabeled cells used can be increased.

[0139] As an alternative to using magnetic columns for dMACS, a microfluidic channel could be used to sort droplets containing magnetically labeled cells or beads, (see Figure 6B) A version of this channel would be designed to have one inlet for all droplets to enter and two outlets, with only one outlet near a magnet so as to bias the flow of droplets containing magnetic beads to that outlet for dMACS enrichment. The droplets would be sorted in a continuous manner as they pass the magnet into the correct channel depending on if they contain a magnetic bead or not.

[0140] Example 8. The dMACS technology can also be used to enable higher-throughput transcriptomics workflows downstream of a fluorescence-based functional screen, which in some embodiments will enable linking of mRNA contents in multiple cells on a per droplet basis.

[0141] Performing the amount of sequencing required for many beads is expensive. By linking the beads to cells, one can cut down on sequencing costs by avoiding drops without the required cell types.

[0142] Here, approximately one magnetic bead will be bound to the labeled cell using a method already discussed herein, such as a streptavidin-biotin linker. This magnetic bead will further be labeled with a poly-T oligonucleotide designed to capture mRNA, as well as a unique nucleotide barcode per bead. In some embodiments, unique molecular identifiers will be used to more accurately count the number of each mRNA captured.

[0143] Some embodiments will capture mRNA on a solid bead, while others will utilize porous hydrogels to capture more mRNA. Some embodiments will have a bead with a magnetic core, while others will have magnetic particles diffused throughout the bead.

[0144] In some embodiments with 2 or more cells in the same drop, the mRNA will be captured, barcoded, and sequenced from all the cells in the drops. This enables linking genes and gene expression between multiple cells on a per drop basis. This can be useful, for example, for antibody discovery of functional antibodies, where an antibody-producer cell can be sequenced and matched to gene expression changes in a function -reporting cell.

[0145] When done for functional gene expression screens, it will often make economic sense to first screen and sort for a fluorescent readout of function. Transcriptomics can then be done only on droplets containing all desired cell types that also show some amount of function, with the transcriptomics yielding additional data to better quantify differences in antibody function. All publications, patents, and non-patent documents are incorporated by reference herein, as though individually incorporated by reference. The invention has been described with reference to various specific and preferred embodiments and techniques. However, it should be understood that variations and modifications may be made while remaining within the spirit and scope of the invention.

Claims

CLAIMSWhat is claimed is:

1. A method for increasing the percentage of droplets containing desired cells in a population of droplets, comprising:(1) attaching magnetic beads to a first population of desired cells in a solution to provide a first population of magnetic cells;(2) optionally removing magnetic beads not attached to cells from the solution;(3) partitioning the desired magnetic cells into magnetic droplets;(4) moving the magnetic droplets across a magnetic field;(5) removing non-magnetic droplets;(6) removing the magnetic droplets from magnetic field to provide magnetic droplets that comprise the first population of desired cells.

2. The method of claim 1, wherein (1) of the method further comprises attaching magnetic beads to a second population of desired cells in a solution to provide a second population of magnetic cells and (6) further comprises removing the magnetic droplets from magnetic field to provide magnetic droplets that comprise the first population of desired cells and the second population of magnetic cells.

3. The method of claim 2, wherein (1) of the method further comprises attaching magnetic beads to a third population of desired cells in a solution to provide a third population of magnetic cells and (6) further comprises removing the magnetic droplets from magnetic field to provide magnetic droplets that comprise the first population of desired cells and the second population of magnetic cells and the third population of desired cells.

4. The method of any one of claims 1-3, wherein the first, second and / or third population of desired cells is a population of antibody producing cells.

5. The method of any one of claims 1-4, wherein the first, second and / or third population of desired cells is a population of reporter cells.

6. The method of any one of claims 1-5, wherein the first, second and / or third population of desired cells is a population of cancer cells.

7. The method of any one of claims 1-6, wherein the method further comprises using a declumping agent.

8. The method of claim 7, wherein the declumping agent is beta mercaptoethanol.

9. The method of any one of claims 1-8, wherein the method further comprises processing the non-magnetic droplets removed in (5) to increase the yield of magnetic droplets.

10. The method of any one of claims 1-9, wherein the method further comprises processing the magnetic droplets removed in (6) to increase the purity of magnetic droplets.

11. The method of any one of claims 1-10, wherein the method comprises labeling only one of the cell populations when multiple cell populations are present.

12. The method of any one of claims 1-11, wherein step (2) comprises removing magnetic beads not attached to cells from the solution.

13. The method of any one of claims 1-11, wherein step (2) does not comprise removing magnetic beads not attached to cells from the solution.

14. An assay that comprises using the method of any one of claims 1-13.

Images