High-throughput analysis of antibody binding and specificity

Abstract

The present disclosure provides a high-throughput assay system for studying binding between two proteins, such as a target protein and a target-binding protein (TBP). The assay system combines a microfluidic system for single-cell analysis with overlap-extension reverse transcription PCR, enabling parallel screening of multiple proteins and their binding targets. The present disclosure provides a high-throughput library-by-library interaction platform that utilizes antigens displayed on cell surfaces and antibodies displayed on soluble ribosomes.

Description

[Technical Field]

[0001] 1. CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of U.S. Provisional Patent Application No. 63 / 480,918, filed January 20, 2023, and U.S. Provisional Patent Application No. 63 / 612,234, filed December 19, 2023, the disclosures of which are incorporated by reference herein in their entireties for all purposes.

[0002] 2. Sequence Listing This application contains a Sequence Listing that has been submitted electronically in XML format, and is incorporated herein by reference in its entirety. The XML copy created on XX is named 28152-53884-SEQLIST.xml and is XX bytes in size. [Background technology]

[0003] 3. Background of the invention Antibody therapeutics are increasingly being used to treat intractable diseases such as cancer. However, the antibody drug discovery process is costly and laborious. The process involves identifying an antigen, isolating an antibody, and characterizing and selecting an antibody with the desired activity against the antigen. The workload for antibody development and characterization can be enormous, yet the chance of obtaining a functional and effective antibody is slim.

[0004] As highlighted by the coronavirus disease 2019 (COVID-19) pandemic, monoclonal antibodies can be important tools for preventing and controlling infectious diseases in high-risk individuals. The extraordinary specificity of antibodies, which provides safety and efficacy, also means that diverse and evolving targets can escape treatment. For example, the initial antibody bamlanivimab developed by Eli Lilly lost nearly all efficacy against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) strains harboring the E484K mutation in the spike protein, which arose less than a year after the start of the pandemic. Instead, binary antibody cocktails, serum-derived polyclonal antibodies, or recombinant polyclonal libraries show improved longevity of binding and protection against evolving pathogens.

[0005] The natural immune response is highly polyclonal, making it an ideal reservoir for mining diverse binders. While methods have been developed to capture and sequence native antibody pairings in these diverse libraries, a simple, high-throughput method for assessing the binding specificity of individual antibodies en masse is lacking. One approach, called LIBRA-seq, uses DNA-barcoded antigens to stain primary B cells, which are then harvested through a single-cell sequencing workflow to obtain in silico linked antibody and antigen sequences. While this method has been successfully adapted and used to identify neutralizing CoV-2 antibodies, a key challenge is the need for individually expressed, purified, and barcoded antigens. A different strategy utilizes existing biological systems in which two components selectively fuse. For example, alpha-seq involves encoding two interacting libraries in opposing yeast mating types, mating them, and then sequencing diploid cells. ENTER-seq and RAPTR display separate libraries on lentiviral particles and mammalian cells, and the transduced cells are sequenced to identify interacting pairs. While these methods have high selectivity and sensitivity, each fusion only provides information about a single pairing event, and mapping more complex interaction networks requires the analysis of large numbers of cells.

[0006] Ribosome display is an alternative display technology that allows for genotype-phenotype linkage of soluble proteins. In its simplest form, proteins are translated from mRNA lacking a stop codon, preventing dissociation of the protein-ribosome-mRNA complex. This method has been further improved over time with the use of additional stall sequences and recombinant expression reagents. Human open reading frame (ORF) libraries were generated on a ribosome display platform, paired with covalently linked DNA barcodes, and used to determine the antigen specificity of autoreactive antibodies from patient samples. While this technology allows for screening of an antigen library against an antibody library, antibody sequences are not recovered, limiting its use as a profiling method. SMI-Seq uses a cDNA-barcoded protein library displayed on gel-immobilized ribosomes that is incubated with a second barcoded library. Fluorescent sequencing is used to identify colocalized barcodes from interacting proteins; the methodology is innovative but highly complex and has limited scalability.

[0007] Thus, there is a continuing need for improved methods for the generation and characterization of recombinant antibodies. In particular, there is a need for high-throughput assays for large numbers of antibodies and antigens to expedite the discovery process for the development of therapeutic antibodies. Summary of the Invention [Means for solving the problem]

[0008] 4. Summary of the Invention The present disclosure provides a high-throughput library-by-library interaction platform that utilizes antigens displayed on cell surfaces and antibodies displayed on soluble ribosomes. Such a polyclonal mapping assay system, also referred to herein as "PolyMap," enables the study of binding patterns between various classes of proteins, such as target proteins (e.g., antigens, ligands, or receptors), and their cognate binding partners (e.g., antibodies, receptors, or ligands). This can be achieved, for example, by combining a microfluidic system for single-cell analysis with overlap extension reverse transcription PCR to enable parallel screening of numerous proteins and their binding targets. Such methods can be used for high-throughput analysis of target-binding proteins, for example, characterization of individual antibodies in polyclonal mixtures without the need for isolation and purification, and for identification of binding partners between libraries of antibodies and libraries of antigens (e.g., libraries of naturally occurring antigen variants). In contrast to previous studies, the antigen library is expressed on the surface of mammalian cells and supports the native structures of a wide range of human and viral proteins without the need for purification. Antibody libraries are expressed by ribosome display as soluble single-chain variable fragments (scFv), barcoded through their complementarity-determining region triplex sequences (CDR3H), and used to stain antigen-presenting cells in bulk. Single cells, each potentially binding thousands of antibody-ribosome-mRNA (ARM) complexes, are encapsulated with uniquely barcoded RNA capture beads, and the barcoded cDNA is sequenced to reveal antibody-antigen pairings. PolyMap is compatible with synthetic library generation methods, enabling not only profiling but also protein engineering applications. PolyMap can be used, for example, to identify CoV-2-targeting antibodies with unique antigen-binding specificities.

[0009] Specifically, the present disclosure provides a method for high-throughput analysis of target binding proteins ("TBPs"), the method comprising: providing a library of target-decorated cells, each of which displays a target of interest on its membrane; contacting the library of target-decorated cells with a plurality of target binding protein ("TBP")-ribosome-mRNA (TRM) complexes, thereby inducing binding between the target-decorated cells and the TRM complexes; generating a plurality of emulsion microdroplets, each microdroplet containing a single one of the target-decorated cells, one or more TRM complexes bound to the single cell, and a lysis reagent that induces lysis of the single cell; capturing RNA released from the single cell on a solid surface or within a semipermeable shell; and generating a library of hybrid polynucleic acids comprising sequences from transcripts identifying the expressed targets of the single cells and / or sequences from the mRNA of the TRM complexes bound to the cell. In some embodiments, the method includes generating a library of hybrid polynucleic acids comprising sequences from transcripts identifying expressed targets of the single cell and sequences from mRNA of a TRM complex associated with the cell. In some embodiments, the method includes generating a library of hybrid polynucleic acids comprising a) sequences from transcripts identifying expressed targets of the single cell and sequences (e.g., barcode sequences) from oligonucleotides immobilized on particles (e.g., beads) associated with the single cell; or b) sequences from mRNA of a TRM complex associated with the single cell and sequences (e.g., barcode sequences) from oligonucleotides immobilized on particles (e.g., beads) associated with the single cell.In some embodiments, the method includes generating a library of hybrid polynucleic acids comprising a first set of hybrid polynucleic acids and a second set of hybrid polynucleic acids, wherein a) the hybrid polynucleic acids of the first set of hybrid polynucleic acids comprise sequences from transcripts identifying expressed targets of the single cell and sequences (e.g., barcode sequences) from oligonucleotides immobilized on particles (e.g., beads) associated with the single cell; and b) the hybrid polynucleic acids of the second set of hybrid polynucleic acids comprise sequences from mRNA of a TRM complex bound to the single cell and sequences (e.g., barcode sequences) from oligonucleotides immobilized on particles (e.g., beads) associated with the single cell. In some embodiments, the method further comprises sequencing the library of hybrid polynucleic acids. In some embodiments, the method further comprises identifying target-TBP pairs based on sequencing of the library of hybrid polynucleic acids. In some embodiments, the method further comprises identifying a target binding protein specific to the target of interest. In some embodiments, the method further comprises identifying the binding affinity or specificity of the target binding protein specific to the target of interest.

[0010] In some embodiments, the method further comprises determining a read distribution of a plurality of TBPs in a plurality of TRM complexes across two or more targets of interest, hi some embodiments, the method further comprises normalizing the read distribution based on an input distribution of the plurality of TBPs.

[0011] In some embodiments, at least one target of interest comprises or is conjugated to a domain capable of inducing expression of an activation marker in target-decorated cells when the target of interest binds to TBP. In some embodiments, generating a plurality of monodisperse or polydisperse emulsion microdroplets comprises sorting the target-decorated cells based on the presence or absence of an activation marker.

[0012] In some embodiments, the target-decorated cells express a fusion protein comprising a target of interest and a transmembrane domain. In some embodiments, the target-decorated cells comprise a construct encoding the fusion protein comprising the target of interest and a transmembrane domain. In some embodiments, the construct further comprises a barcode sequence. In some embodiments, the construct further comprises a sequence encoding a fluorescent protein. In some embodiments, the construct further comprises a sequence encoding a surface marker or a detectable tag.

[0013] In some embodiments, the method further comprises isolating a plurality of emulsion microdroplets containing target decorated cells by detecting expression of a fluorescent protein.

[0014] In some embodiments, the library of target-decorated cells includes one cell clone displaying one target of interest. In some embodiments, the library of target-decorated cells includes two, three, or four cell clones, where each cell clone displays a unique target of interest that is different from the other cell clones. In some embodiments, the library of target-decorated cells includes at least five cell clones, where each cell clone displays a unique target of interest that is different from the other cell clones. In some embodiments, the library of target-decorated cells includes at least 10 cell clones, where each cell clone displays a unique target of interest that is different from the other cell clones. In some embodiments, the library of target-decorated cells includes at least 100 cell clones, where each cell clone displays a unique target of interest that is different from the other cell clones. In some embodiments, the library of target-decorated cells includes at least 1000 cell clones, where each cell clone displays a unique target of interest that is different from the other cell clones.

[0015] In some embodiments, the transcript of the isolated single cell comprises a coding sequence for a target of interest, hi some embodiments, the transcript of the isolated single cell comprises a barcode sequence that uniquely identifies the target expressed by the cell.

[0016] In some embodiments, each of the TRM complexes comprises a target binding protein. In some embodiments, each of the TRM complexes comprises an scFv (e.g., each of the TRM complexes comprises a TPB comprising an scFv). In some embodiments, each of the TRM complexes comprises a heavy chain variable region (e.g., each of the TRM complexes comprises a TPB comprising a heavy chain variable region). In some embodiments, each of the TRM complexes comprises a light chain variable region (e.g., each of the TRM complexes comprises a TPB comprising a light chain variable region).

[0017] In some embodiments, the plurality of TRM complexes comprises 1-5 unique TRM complexes, each of which comprises a unique target binding protein that is different from the rest of the unique TRM complex. In some embodiments, the plurality of TRM complexes comprises 6-10 unique TRM complexes, each of which comprises a unique target binding protein that is different from the rest of the unique TRM complex. In some embodiments, the plurality of TRM complexes comprises at least 10 unique TRM complexes, each of which comprises a unique target binding protein that is different from other unique TRM complexes. In some embodiments, the plurality of TRM complexes comprises at least 100 unique TRM complexes, each of which comprises a unique target binding protein that is different from other unique TRM complexes. In some embodiments, the plurality of TRM complexes comprises at least 1000 unique TRM complexes, each of which comprises a unique target binding protein that is different from other unique TRM complexes.

[0018] In some embodiments, each TRM complex comprises a target binding protein and an mRNA encoding the target binding protein. In some embodiments, the mRNA comprises a coding sequence for a complementarity determining region (CDR) of the target binding protein. In some embodiments, the mRNA comprises a coding sequence for a CDR H3 of the target binding protein. In some embodiments, the mRNA comprises a barcode sequence that uniquely identifies the TBP or target binding protein encoded by the mRNA.

[0019] In some embodiments, RNA capture is performed using oligonucleotides immobilized on beads. In some embodiments, RNA capture is performed using oligonucleotides immobilized on beads, where each bead has a diameter greater than 10 μm, 0.5-10 μm, less than 1 μm, or about 1 μm. In some embodiments, each bead is a solid bead or a porous bead.

[0020] In some embodiments, according to any of the above methods using oligonucleotides immobilized on beads, the oligonucleotides immobilized on the beads comprise a barcode sequence. In some embodiments, the barcode sequence in the oligonucleotides immobilized on the beads is unique to a given bead. For example, in some embodiments, 1) a first bead comprises an oligonucleotide immobilized on the first bead, wherein the oligonucleotide immobilized on the first bead comprises a first barcode sequence; and 2) a second bead comprises an oligonucleotide immobilized on the second bead, wherein the oligonucleotide immobilized on the second bead comprises a second barcode sequence, wherein the first barcode sequence and the second barcode sequence are different. This can be extended to any number of beads. In some embodiments, the hybrid polynucleic acid further comprises a barcode sequence derived from the oligonucleotide immobilized on the bead.

[0021] In some embodiments, the hybrid polynucleic acids are generated by overlap extension polymerase chain reaction (OE-PCR). In some embodiments, the generation of the hybrid polynucleic acids is preceded by first strand cDNA synthesis.

[0022] In some embodiments, the method further comprises generating a second set of monodisperse or polydisperse emulsion microdroplets containing bead-captured RNA released from the single cells prior to generating the library of hybrid polynucleic acids, hi some embodiments, the library of hybrid polynucleic acids is generated in the second set of monodisperse or polydisperse emulsion microdroplets.

[0023] In some embodiments, contacting the library of target decorated cells with a plurality of TBP-ribosome-mRNA (TRM) complexes is performed in a buffer containing 25-100 mM Mg2+. In some embodiments, the buffer contains 50 mM Mg2+. In some embodiments, the buffer contains 50 mM MgCl2. In some embodiments, the buffer contains HEPES, NaCl, 50 mM MgCl2, and BSA. In some embodiments, the buffer contains HEPES, NaCl, 50 mM MgCl2, polysorbate 20, heparin, and BSA. In some embodiments, the buffer contains 20 mM HEPES, 50 mM NaCl, 50 mM MgCl2, 0.01% polysorbate 20, 2.5 mg / ml heparin, and 0.5% BSA. In some embodiments, the buffer comprises 20 mM HEPES, 50 mM NaCl, 50 mM MgCl, 0.01% polysorbate 20, 2.5 mg / ml heparin, and 0.05% BSA. In some embodiments, the buffer further comprises an RNase inhibitor.

[0024] In one aspect, the disclosure provides a kit for high-throughput analysis of target-binding proteins, the kit including: a plurality of constructs, each construct encoding a fusion protein comprising a unique target of interest and a transmembrane domain; a plurality of TBP-ribosome-mRNA (TRM) complexes, each TRM complex comprising a unique target-binding protein; and a buffer.

[0025] In some embodiments, the kit further comprises a host cell. In some embodiments, the kit comprises a single construct encoding one target of interest.

[0026] In some embodiments, the kit comprises 2-10 unique constructs, wherein each unique construct encodes a unique target of interest. In some embodiments, the kit comprises at least 10 unique constructs, wherein each unique construct encodes a unique target of interest. In some embodiments, the kit comprises at least 100 unique constructs. In some embodiments, the kit comprises at least 1000 unique constructs.

[0027] In some embodiments, the kit includes one unique TRM complex, where one unique TRM complex includes one target binding protein. In some embodiments, the kit includes at least 10 unique TRM complexes, where each unique TRM complex includes a unique target binding protein. In some embodiments, the kit includes at least 100 unique TRM complexes. In some embodiments, the kit includes at least 1000 unique TRM complexes.

[0028] In some embodiments, the kit further comprises reagents for overlap extension polymerase chain reaction (OE-PCR).

[0029] In some embodiments, the buffer comprises 50 mM Mg. In some embodiments, the buffer comprises 50 mM MgCl. In some embodiments, the buffer comprises HEPES, NaCl, 50 mM MgCl, and BSA. In some embodiments, the buffer comprises HEPES, NaCl, 50 mM MgCl, polysorbate 20, heparin, and BSA. In some embodiments, the buffer comprises 20 mM HEPES, 50 mM NaCl, 50 mM MgCl, 0.01% polysorbate 20, 2.5 mg / ml heparin, and 0.5% BSA. In some embodiments, the buffer comprises 20 mM HEPES, 50 mM NaCl, 50 mM MgCl, 0.01% polysorbate 20, 2.5 mg / ml heparin, and 0.05% BSA. In some embodiments, the buffer further comprises an RNase inhibitor.

[0030] In another aspect, the present disclosure provides a library of hybrid polynucleic acids produced by the methods disclosed herein.

[0031] In yet another aspect, the present disclosure provides a method for high-throughput analysis of receptors, the method comprising: providing a library of ligand-decorated cells, each of which displays a ligand of interest on its membrane; contacting the library of ligand-decorated cells with a plurality of receptor-ribosome-mRNA (RRM) complexes, thereby inducing binding between the ligand-decorated cells and the RRM complexes; generating a plurality of monodisperse or polydisperse emulsion microdroplets, each microdroplet containing a single one of the ligand-decorated cells, one or more RRM complexes bound to the single cell, and a lysis reagent that induces lysis of the single cell; capturing RNA released from the single cell on a solid surface or within a semipermeable shell; and generating a library of hybrid polynucleic acids comprising sequences from transcripts of the single cells and / or sequences from the mRNA of the RRM complexes.

[0032] The present disclosure also provides a method for high-throughput analysis of receptors, the method comprising the steps of: providing a library of receptor-decorated cells, each of which displays a receptor of interest on its membrane; contacting the library of receptor-decorated cells with a plurality of ligand-ribosome-mRNA (LRM) complexes, thereby inducing binding between the receptor-decorated cells and the LRM complexes; generating a plurality of emulsion microdroplets, each microdroplet containing a single one of the receptor-decorated cells, one or more LRM complexes bound to the single cell, and a lysis reagent that induces lysis of the single cell; capturing RNA released from the single cell on a solid surface or within a semipermeable shell; and generating a library of hybrid polynucleic acids comprising sequences from transcripts of the single cells and / or sequences from the mRNA of the LRM complexes.

[0033] The present disclosure also provides a method for high-throughput analysis of antibodies, the method comprising the steps of: providing a library of antigen-decorated cells, each of which displays an antigen of interest on a membrane; contacting the library of antigen-decorated cells with a plurality of antibody-ribosome-mRNA (ARM) complexes, thereby inducing binding between the antigen-decorated cells and the ARM complexes; generating a plurality of emulsion microdroplets, each microdroplet containing a single one of the antigen-decorated cells, one or more ARM complexes bound to the single cell, and a lysis reagent that induces lysis of the single cell; capturing RNA released from the single cells on a solid surface or within a semipermeable shell; and generating a library of hybrid polynucleic acids comprising sequences from transcripts of the single cells and / or sequences from the mRNA of the ARM complexes. [Brief explanation of the drawings]

[0034] 5. A brief description of some figures in the drawing These and other features, aspects, and advantages of the present invention will become better understood with reference to the following description and accompanying drawings.

[0035] [Figure 1] Figure 1 outlines a high-throughput method for the analysis of target-binding proteins (antibodies). (1a) A library of target (antigen)-decorated cells is prepared by expressing the target library on cells; (1b) Multiple TBP-ribosome-mRNA (TRM) complexes are prepared; (2) The library of target-decorated cells and multiple TRM complexes are mixed in bulk to induce binding between the target-decorated cells and the TRM complexes; (3) Cells are individually encapsulated in droplets with lysis buffer and RNA capture beads; (4) The RNA capture beads are isolated and reencapsulated in a second droplet where the target barcode and antibody CDR3 sequences can be ligated and amplified; and (5) the amplified DNA is isolated and prepared in bulk for sequencing, e.g., using Illumina NGS methods. Bioinformatics analysis is used to normalize the data and map TBP:target interactions.

[0036] [Figure 2] Figure 2 shows FACS detection of cells expressing spike protein (CoV1-S or CoV2-S) from one of the six tested constructs (V1, V2, V3, V4, V5 and V6) using mAbs against the spike protein.

[0037] [Figure 3] Figure 3 shows the expression of CoV2-S antigen in Expi293 or CHOZN cell lines after stable transfection with the CoV2-S coding sequence (Expi293+CoV2-S, CHOZN+CoV2-S, and CHOZN+CoV2-S(HP+F)) described in Example 1. CoV2-S expression was detected with bamlamivimab (MFI).

[0038] [Figure 4] Figure 4 provides the structure of an exemplary construct backbone (p2G-FRT-GS) for generating target-decorated cells.

[0039] [Figure 5] Figure 5 provides the structure of an exemplary construct (T7-based expression plasmid) for generating the TBP-ribosome-mRNA complex (TRM complex).

[0040] [Figure 6] Figure 6 shows the detection of binding between CoV1 or CoV2 spike WT or variants (CoV2-K444T, CoV2-E484K, CoV2-F486K) on target decorated cells and TRM complexes containing the scFv of one of four mAbs (bamianivimab, casirivimab, imdevimab, or ipilimumab). TRM bound to CoV2-S variants was detected by strep-tag staining.

[0041] [Figure 7] Figure 7 shows the detection of binding between CoV1 or CoV2 spike WT or variants (CoV2-K444T, CoV2-E484K, CoV2-F486K) on target decorated cells and TRM complexes containing the scFv of one of four mAbs (bamianivimab, casirivimab, imdevimab, or ipilimumab). TRMs bound to CoV2-S variants were isolated and detected by TaqMan RT-qPCR using an antibody (TBP)-specific probe.

[0042] [Figure 8]Figure 8 shows the specificity of binding between the CoV2-S antigen and the TRM complex by measuring the ratio between specific staining (RNA isolated from the binding between the CoV2-S antigen and the TRM complex containing casirivimab scFv) and nonspecific staining (RNA isolated from the binding between the CoV2-S antigen and ipilimumab scFv). The TRM complex was generated under different cell-free translation (TL) conditions with different template concentrations and reaction times.

[0043] [Figure 9] Figure 9 shows the detection of binding between CoV1 or CoV2 spike WT or variants (CoV2-K444T, CoV2-E484K, CoV2-F486K) on target decorated cells and a mixture of five TRM complexes, where each TRM complex contains the scFv of one of four mAbs (bamianivimab, casirivimab, imdevimab, or ipilimumab). TRMs bound to CoV2-S variants were isolated and detected by TaqMan RT-qPCR using an antibody (TBP)-specific probe.

[0044] [Figure 10] Figure 10 provides Polymap scores representing binding between five target-decorated cell lines, each expressing CoV1 or CoV2 spike WT or variants (CoV2-K444T, CoV2-E484K, CoV2-F486K), and five TRM complexes, each containing the scFv of one of five mAbs (Bamianivimab, Casirivimab, Imdevimab, Ipilimumab, or Pembrolizumab). Polymap scores were calculated from the average percentage scaled by the average reads from cells for that target, as described in Example 6.

[0045] [Figure 11]Figure 11 shows overlap extension reverse transcription PCR of scFv RNA in the TRM complex and the barcode sequence in the target decorated cells. OE-RT-PCR generates a product containing the VH region sequence and the barcode sequence.

[0046] [Figure 12] Figure 12 outlines the PolyMap platform workflow, starting with a library of antigens expressed on the surface of mammalian cells, which are then incubated with a soluble library of antibody scFvs in a ribosome-display format. Stained single cells are encapsulated with uniquely barcoded RNA capture beads (barcodes are represented by stars, triangles, and diamonds) and lysed. The beads are isolated and used to generate cDNA, which is then further amplified with gene-specific primers for sequencing. Analysis of the cell barcode, antigen barcode, and antibody CDRH3 is used to generate a map of antibody-antigen interactions.

[0047] [Figure 13] Figure 13 shows CoV-2 spike expression in various cell lines. Results are presented as flow cytometry data showing surface expression levels of CoV-2 S in stable Expi293™ and CHOZN® cell lines. The variant labeled "stabilized" contains six proline mutations and removal of the furin site.

[0048] [Figure 14A-B]Figures 14A-14D show results for various antigen expression systems. Figure 14A shows an antigen expression construct containing a CMV promoter with a translation-enhancing element (SG), signal peptide (SP), and transmembrane (TM) region for surface display, and a unique barcode (BC) in the 3' untranslated region (UTR). An FRT site allows integration and translation of the glutamine synthetase gene (GS) for selection. Figure 14B shows surface expression of different spike variants stably expressed in CHOZN cells, as measured by monoclonal antibody and flow cytometry. Figure 14C shows a representative DNA fragment encoding an antibody scFv ribosome display library, containing a T7 promoter, strep tag II, a post-protein (TolA) spacer sequence, and a secM ribosome stall sequence. A 40-mer polyA tail is internally encoded by the bovine growth hormone (BGH) polyadenylation signal in the 3' UTR. Figure 14D shows RNA recovered from spike variant cells stained with clinical antibodies as ARM complexes, as measured by RT-qPCR. The table shows expected results based on literature and affinity studies. [Figure 14C-D]Figures 14A-14D show results for various antigen expression systems. Figure 14A shows an antigen expression construct containing a CMV promoter with a translation-enhancing element (SG), signal peptide (SP), and transmembrane (TM) region for surface display, and a unique barcode (BC) in the 3' untranslated region (UTR). An FRT site allows integration and translation of the glutamine synthetase gene (GS) for selection. Figure 14B shows surface expression of different spike variants stably expressed in CHOZN cells, as measured by monoclonal antibody and flow cytometry. Figure 14C shows a representative DNA fragment encoding an antibody scFv ribosome display library, containing a T7 promoter, strep tag II, a post-protein (TolA) spacer sequence, and a secM ribosome stall sequence. A 40-mer polyA tail is internally encoded by the bovine growth hormone (BGH) polyadenylation signal in the 3' UTR. Figure 14D shows RNA recovered from spike variant cells stained with clinical antibodies as ARM complexes, as measured by RT-qPCR. The table shows expected results based on literature and affinity studies.

[0049] [Figure 15] Figure 15 shows the optimization of ARM complex production, as demonstrated by RT-qPCR of RNA recovered from CoV-2 S WT cells stained with the ARM complex mixture produced under various conditions. Specific probes for "positive" (casirivimab) and "nonspecific" RNA were used to quantify the recovery rate of each, and the ratios were calculated here. The STD ivTXTL kit includes all components for transcription, translation, and termination of the reaction, while the ΔRF kit omits factors that aid in ribosome release from mRNA. ARM complexes were translated for 10 or 30 minutes before staining (TL).

[0050] [Figure 16-1]Figure 16 shows biolayer interferometry (BLI) of clinical antibodies against CoV-2 spike variants. Association and dissociation curves at three concentrations are shown along with globally calculated curve fits. [Figure 16-2] Figure 16 shows biolayer interferometry (BLI) of clinical antibodies against CoV-2 spike variants. Association and dissociation curves at three concentrations are shown along with globally calculated curve fits. [Figure 16-3] Figure 16 shows biolayer interferometry (BLI) of clinical antibodies against CoV-2 spike variants. Association and dissociation curves at three concentrations are shown along with globally calculated curve fits.

[0051] [Figures 17A-C]Figures 17A-17E show the results of screening five monoclonal antibodies and four SARS-CoV-2 spike variants. For all experiments, ARM complexes were generated from an equimolar mixture of the five mAbs. Figure 17A: Cell lines expressing the indicated spike variants (columns) were stained separately with the ARM complex mixture, and then RNA was harvested and amplified for Illumina sequencing. Data shown in the heatmap are the percentage of reads for each antibody sequence. Figure 17B: Single-cell sorting. Equal amounts of spike variant cell lines were pooled and stained with the ARM complex mixture, and then single cells were sorted into wells of a 96-well plate. The antibody variable heavy chain region and antigen barcode were amplified and sequenced, and the antigen barcode was used to determine the identity of the spike variant cell line in each well. Data are shown as a heatmap of the average percentage of reads for each antibody sequence associated with each antigen cell line. Figure 17C: Data for individual cells from Figure 17B, where each cell is represented as five data points, one for each antibody. Data shown is the percentage of each antibody associated with each cell. Figure 17D: Drop-seq binding profile. Equal amounts of spike variant cell lines were pooled and stained with an ARM complex mixture. RNA from individual cells was captured onto beads, and the antibody variable heavy chain region and antigen barcode were amplified and then sequenced along with the concatenated Drop-seq cell barcode. After merging the lists of antigen and antibody cell barcodes, 91 cells were analyzed for antibody binding. Data are shown as a heatmap of the average percent of reads relative to the Drop-seq data. Figure 17E: Data for individual cells from Figure 17D, where each cell is represented as five data points, one for each antibody. Data shown is the percentage of each antibody associated with each cell. [Figure 17D-E]Figures 17A-17E show the results of screening five monoclonal antibodies and four SARS-CoV-2 spike variants. For all experiments, ARM complexes were generated from an equimolar mixture of the five mAbs. Figure 17A: Cell lines expressing the indicated spike variants (columns) were stained separately with the ARM complex mixture, and then RNA was harvested and amplified for Illumina sequencing. Data shown in the heatmap are the percentage of reads for each antibody sequence. Figure 17B: Single-cell sorting. Equal amounts of spike variant cell lines were pooled and stained with the ARM complex mixture, and then single cells were sorted into wells of a 96-well plate. The antibody variable heavy chain region and antigen barcode were amplified and sequenced, and the antigen barcode was used to determine the identity of the spike variant cell line in each well. Data are shown as a heatmap of the average percentage of reads for each antibody sequence associated with each antigen cell line. Figure 17C: Data for individual cells from Figure 17B, where each cell is represented as five data points, one for each antibody. Data shown is the percentage of each antibody associated with each cell. Figure 17D: Drop-seq binding profile. Equal amounts of spike variant cell lines were pooled and stained with an ARM complex mixture. RNA from individual cells was captured onto beads, and the antibody variable heavy chain region and antigen barcode were amplified and then sequenced along with the concatenated Drop-seq cell barcode. After merging the lists of antigen and antibody cell barcodes, 91 cells were analyzed for antibody binding. Data are shown as a heatmap of the average percent of reads relative to the Drop-seq data. Figure 17E: Data for individual cells from Figure 17D, where each cell is represented as five data points, one for each antibody. Data shown is the percentage of each antibody associated with each cell.

[0052] [Figure 18] Figure 18 shows RNA enrichment under two different staining conditions. Enrichment fold of a mixture of 14 antibody clones based on sequencing before and after staining a CoV-2 S WT cell line with two different ARM complex concentrations.

[0053] [Figure 19-1] Figure 19 shows an alignment of selected CoV-2 spike variant sequences. Amino acids corresponding to the receptor binding domain (RBD) are indicated. [Figure 19-2] Figure 19 shows an alignment of selected CoV-2 spike variant sequences. Amino acids corresponding to the receptor binding domain (RBD) are indicated.

[0054] [Figure 20A] Figures 20A and 20B show the binding profiles of a library of anti-SARS-CoV-2 antibodies to a library of spike variants. Figure 20A: The antibody repertoire used to generate the ARM complex was determined by sequencing heavy chain fragments from the RNA mixture that served as input for the in vitro translation reaction. The top 100 clones are plotted along with the individual and cumulative percentages of each clone in the library. Figure 20B: Drop-seq. Cell lines expressing the indicated spike variants (columns) were pooled, stained with the ARM complex mixture, then encapsulated with barcoded beads and run through the Drop-seq workflow. The percentage of normalized reads for each antibody across all cell lines is plotted. The top 40 antibodies (based on total antibody reads) are shown. [Figure 20B]Figures 20A and 20B show the binding profiles of a library of anti-SARS-CoV-2 antibodies to a library of spike variants. Figure 20A: The antibody repertoire used to generate the ARM complex was determined by sequencing heavy chain fragments from the RNA mixture that served as input for the in vitro translation reaction. The top 100 clones are plotted along with the individual and cumulative percentages of each clone in the library. Figure 20B: Drop-seq. Cell lines expressing the indicated spike variants (columns) were pooled, stained with the ARM complex mixture, then encapsulated with barcoded beads and run through the Drop-seq workflow. The percentage of normalized reads for each antibody across all cell lines is plotted. The top 40 antibodies (based on total antibody reads) are shown.

[0055] [Figure 21] Figure 21 shows antigen cellular distribution from a Drop-seq library experiment. After ARM complex staining, encapsulation with barcoded Drop-seq beads (ChemGenes Corporation), and antigen amplification workflow, the identity of the recovered antigens was determined by sequencing the associated antigen barcodes.

[0056] [Figure 22A]Figures 22A and 22B show functional validation of antibody binding patterns identified by PolyMap. Figure 22A: Cell lines expressing the indicated spike variants (x-axis) were separately stained with an anti-CoV-2 library ARM complex mixture, and then RNA was harvested and amplified for Illumina sequencing. Data shown in the heatmap are the log2 fold change in enrichment relative to input for each antibody sequence associated with each cell line. Antibodies with log2 fold change values ​​below any of the three negative control antibodies are colored white. Figure 22B: A subset of individual antibody clones validated by flow cytometry and compared to previous data. For each antibody, CDR3H sequences and heatmaps for three datasets are shown: top, Drop-seq (from Figure 20B); middle, single antigen staining (from Figure 22A); bottom, flow cytometry. For flow cytometry, full-length monoclonal antibodies were expressed in CHO cells, and supernatants were used to stain individual spike variant cell lines. The heatmap shows the MFI of the secondary antibody signal. [Figure 22B-1] Figures 22A and 22B show functional validation of antibody binding patterns identified by PolyMap. Figure 22A: Cell lines expressing the indicated spike variants (x-axis) were separately stained with an anti-CoV-2 library ARM complex mixture, and then RNA was harvested and amplified for Illumina sequencing. Data shown in the heatmap are the log2 fold change in enrichment relative to input for each antibody sequence associated with each cell line. Antibodies with log2 fold change values ​​below any of the three negative control antibodies are colored white. Figure 22B: A subset of individual antibody clones validated by flow cytometry and compared to previous data. For each antibody, CDR3H sequences and heatmaps for three datasets are shown: top, Drop-seq (from Figure 20B); middle, single antigen staining (from Figure 22A); bottom, flow cytometry. For flow cytometry, full-length monoclonal antibodies were expressed in CHO cells, and supernatants were used to stain individual spike variant cell lines. The heatmap shows the MFI of the secondary antibody signal. [Figure 22B-2] Figures 22A and 22B show functional validation of antibody binding patterns identified by PolyMap. Figure 22A: Cell lines expressing the indicated spike variants (x-axis) were separately stained with an anti-CoV-2 library ARM complex mixture, and then RNA was harvested and amplified for Illumina sequencing. Data shown in the heatmap are the log2 fold change in enrichment relative to input for each antibody sequence associated with each cell line. Antibodies with log2 fold change values ​​below any of the three negative control antibodies are colored white. Figure 22B: A subset of individual antibody clones validated by flow cytometry and compared to previous data. For each antibody, CDR3H sequences and heatmaps for three datasets are shown: top, Drop-seq (from Figure 20B); middle, single antigen staining (from Figure 22A); bottom, flow cytometry. For flow cytometry, full-length monoclonal antibodies were expressed in CHO cells, and supernatants were used to stain individual spike variant cell lines. The heatmap shows the MFI of the secondary antibody signal.

[0057] [Figure 23] Figure 23 shows RNA enrichment of clones with different affinities. Enrichment of five clones in a mixture based on sequencing before and after staining a CoV-2 S WT cell line. Affinities were determined by BLI.

[0058] [Figure 24] Figure 24 shows the distribution of antigen sequences from bulk-transfected cells. gDNA was isolated from the transfected cells and subjected to sequencing analysis to determine the relative number of times each unique antigen barcode appeared. DETAILED DESCRIPTION OF THE INVENTION

[0059] 6. Detailed Description of the Invention 6.1.Definition As used herein, the term "target binding protein" or TBP refers to a protein comprising one or more target binding domains that specifically bind to a target. In some embodiments, the target binding domain binds to a target, or a fragment thereof, with specificity and affinity similar to that of a naturally occurring antibody. In some embodiments, the TBP comprises an antibody. In some embodiments, the TBP consists essentially of an antibody. In some embodiments, the TBP comprises an alternative scaffold. In some embodiments, the TBP consists essentially of an alternative scaffold. In some embodiments, the TBP comprises an antibody fragment. In some embodiments, the TBP consists essentially of an antibody fragment. In some embodiments, the TBP comprises a receptor or fragment thereof that binds to the target. In some embodiments, the TBP is a receptor or fragment thereof that binds to the target. In some embodiments, the TBP comprises a ligand or fragment thereof. In some embodiments, the TBP is a ligand or fragment thereof.

[0060] The term "antibody" is used herein in its broadest sense and includes a specific type of immunoglobulin molecule that contains one or more antigen-binding domains that specifically bind to an antigen or epitope. Antibodies specifically include intact antibodies (e.g., intact immunoglobulins), antibody fragments, and multispecific antibodies. An example of an antigen-binding domain is a V H -V L The antigen-binding domain is formed by a dimer. Antibodies are a type of ABP.

[0061] The term "alternative scaffold" refers to a molecule in which one or more regions can be diversified to generate one or more antigen-binding domains that specifically bind to an antigen or epitope. In some embodiments, the antigen-binding domains bind to an antigen or epitope with similar specificity and affinity as naturally occurring antibodies. Exemplary alternative scaffolds include those derived from fibronectin (e.g., Adnectins™), beta-sandwich (e.g., iMabs), lipocalins (e.g., Anticalins®), EETI-II / AGRP, BPTI / LACI-D1 / ITI-D2 (e.g., Kunitz domains), thioredoxin peptide aptamers, protein A (e.g., Affibodies®), ankyrin repeats (e.g., DARPins), gamma-B-crystallin / ubiquitin (e.g., affilin), CTLD3 (e.g., tetranectin), fynomers, and (LDLR-A modules) (e.g., avimers). Additional information regarding alternative scaffolds is provided in Binz et al., Nat. Biotechnol., 2005 23:1257-1268; Skerra, Current Opin. in Biotech., 2007 18:295-304; and Silacci et al., J. Biol. Chem., 2014, 289:14392-14398; each of which is incorporated by reference in its entirety. An alternative scaffold is a type of TBP.

[0062] The term "target binding domain" refers to a portion of TBP that is capable of specifically binding to a target.

[0063] An "antibody fragment" comprises a portion of an intact antibody, such as the antigen-binding or variable region of the intact antibody. Antibody fragments include, for example, Fv fragments, Fab fragments, F(ab')2 fragments, Fab' fragments, scFv (sFv) fragments, and scFv-Fc fragments.

[0064] "Single-chain Fv" or "sFv" or "scFv" antibody fragments contain the V H Domains and V L Includes domain. VH and V L are generally linked by a peptide linker. See Plueckthun A. (1994). In some embodiments, the linker is (GGGGS) n In some embodiments, n=1, 2, 3, 4, 5, or 6. See antibodies derived from Escherichia coli. (Rosenberg M. & Moore GP (Eds.), The Phytrmacology of Monoclonal Antibodies vol. 113 (pp. 269-315). Springer-Verlag, New York), incorporated by reference in its entirety.

[0065] "Affinity" refers to the strength of the sum total of non-covalent interactions between a single binding site of a molecule (e.g., ABP) and its binding partner (e.g., antigen or epitope). Unless otherwise specified, as used herein, "affinity" refers to the intrinsic binding affinity, which reflects a 1:1 interaction between members of a binding pair (e.g., ABP and antigen or epitope). The affinity of a molecule X for its partner Y is determined by the dissociation equilibrium constant (K D ) The kinetic components that contribute to the dissociation equilibrium constant are described in more detail below. Affinity can be measured by common methods known in the art, including those described herein. Affinity can be determined, for example, using surface plasmon resonance (SPR) technology (e.g., BIACORE®) or biolayer interferometry (e.g., FORTEBIO®).

[0066] With respect to the binding of TBP to a target molecule, the terms "bind," "specific binding," "specifically binds to," "specific for," "selectively binds," and "selective for" a particular antigen (e.g., a polypeptide target) or epitope on a particular antigen refer to binding that is measurably different from nonspecific or nonselective interactions (e.g., non-target molecules). Specific binding can be measured, for example, by measuring binding to a target molecule and comparing it to binding to a non-target molecule. Specific binding can also be determined by competition with a control molecule that mimics the recognized epitope on the target molecule. Specific binding is then indicated if binding of TBP to the target molecule is competitively inhibited by the control molecule.

[0067] As used herein, the term "TBP-ribosome-mRNA complex" or "TRM complex" refers to a complex containing a target-binding protein, a ribosome, and an mRNA. TRM complexes can be generated by any method known in the art. For example, TRM complexes can be generated using an in vitro cell-free system. In vitro ribosome display technology can produce stable protein (target-binding protein)-ribosome-mRNA (TRM) complexes by linking individual target-binding proteins to their corresponding mRNAs. TRM complexes can be formed by the deletion of a terminal stop codon from the mRNA, which causes the translating ribosome to stall at the end of the mRNA without releasing the nascent polypeptide. Protein-mRNA binding allows for the simultaneous isolation of the mRNA and the desired protein (target-binding protein) through affinity for an immobilized ligand. Protein-mRNA complexes that strongly bind to the ligand can be subjected to in situ reverse transcription-PCR (RT-PCR) to recover the DNA sequence encoding the protein, which can then be amplified in a PCR reaction to generate a template for further manipulation and protein synthesis.

[0068] As used herein, the term "droplet" refers to a small volume of liquid. Droplets are typically spherical, but may consist of cylindrical slugs spanning the entire diameter of a microfluidic channel. Droplets can be formed in air, oil, or aqueous solutions, depending on the composition of the material and the method of formation. Droplets occur in both monodisperse and polydisperse populations.

[0069] As used herein, the term "monodisperse" refers to the property of components characterized by uniform or nearly uniform size. For example, monodisperse droplets typically require a size dispersity of <5% for >90% of the droplets in a mixture. In many cases, monodisperse droplet populations are more stable than non-monodisperse, i.e., polydisperse, droplet populations. In some embodiments, the generation of monodisperse droplets requires some type of controlled microfluidic device.

[0070] As used herein, the term "cell clone" refers to at least two cells of a similar type or classification.

[0071] As used herein, the term "lysis" refers to the process of disrupting the cell membrane of a cell or cells by physical or chemical means. Lysis can be achieved by chemical detergents such as Triton® X-100, alkaline lysis buffer, heat, electric current, or physical disruption.

[0001] 6.2. Other Interpretation Rules Ranges recited herein are understood to be shorthand for all values ​​within the range, inclusive of the recited endpoints. For example, a range of 1 to 50 is understood to include any number, combination of numbers, or subrange from the group consisting of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, and 50.

[0072] 6.3. Methods for High-Throughput Analysis of Antibodies The present disclosure provides a method for high-throughput analysis of antibodies. Specifically, the method may include the following steps: (i) providing a library of target-decorated cells, each of which displays a target of interest on its membrane; (ii) contacting the library of target-decorated cells with a plurality of TBP-ribosome-mRNA (TRM) complexes, thereby inducing binding between the target-decorated cells and the TRM complexes; (iii) generating a plurality of emulsion microdroplets, each of which contains a single target-decorated cell, one or more TRM complexes bound to the single cell, and a lysis reagent that induces lysis of the single cell; (iv) capturing RNA released from the single cell on a solid surface or in a semipermeable shell; and (v) generating a library of hybrid polynucleic acids containing sequences from the transcripts of the single cell and / or sequences from the mRNA of the TRM complexes. The present disclosure further provides variations or modifications of the method for analyzing antibodies.

[0073] The hybrid polynucleic acid library can be analyzed or sequenced to provide information about the binding between the target of interest and the TRM complex. For example, TBPs that bind to the target of interest and / or their binding affinity or specificity can be studied based on the sequence. Thus, the method can further include identifying target-TBP pairs based on sequencing the hybrid polynucleic acid library. In some embodiments, multiple target-TBP pairs are identified by sequencing the hybrid polynucleic acid library. In some embodiments, more than two, more than three, more than four, more than five, more than six, more than seven, more than eight, more than nine, more than ten, more than twenty, or more target-TBP pairs are identified. In some embodiments, the method further includes identifying a target-binding protein specific to the target of interest. In some embodiments, the method further includes identifying the binding affinity or specificity of the target-binding protein specific to the target of interest.

[0074] 6.3.1. Target Decorative Cells The methods disclosed herein use target-decorated cells that express a fusion protein comprising a target of interest and a transmembrane domain. In some cases, the target-decorated cells express a transmembrane protein comprising the target of interest instead of the fusion protein. The target-decorated cells display the target of interest on their surface. In some embodiments, the fusion protein further comprises a signaling domain that facilitates delivery of the target of interest to the surface.

[0075] In some embodiments, the target is an antigen. In some embodiments, the target is a ligand of a receptor or a modification thereof. In some embodiments, the target is a ligand or a modification thereof.

[0076] In some embodiments, the target-decorated cells comprise a coding sequence of a target of interest. In some embodiments, a polynucleotide construct encoding the target of interest is transiently transfected into the target-decorated cells. In some embodiments, a polynucleotide construct encoding the target of interest is stably transfected into the target-decorated cells. In some embodiments, the polynucleotide construct is a lentiviral vector. In some embodiments, the polynucleotide construct comprises an FRT site that allows stable integration of the coding sequence of the target of interest into the landing pad. In some embodiments, the polynucleotide construct comprising an FRT site is transfected with a recombinase to allow stable integration.

[0077] In some embodiments, the target of interest is expressed from an exogenous polynucleotide in the target-decorated cells. In some such embodiments, the target-decorated cells comprise a construct encoding a fusion protein or transmembrane protein comprising the target of interest. In some embodiments, the construct comprises a coding sequence for the fusion protein or transmembrane protein operably linked to a regulatory sequence, such as a promoter. In some embodiments, the construct further comprises a barcode sequence. In some embodiments, the barcode sequence indicates the target of interest expressed or displayed on the target-decorated cells. In some embodiments, the construct further comprises a sequence encoding a reporter protein. In some embodiments, the reporter protein is a fluorescent protein. In some embodiments, the reporter protein is a surface marker or a detectable tag. In some such embodiments, the reporter protein (e.g., a fluorescent protein) can be used to identify or isolate target-decorated cells (including constructs or emulsion microdroplets containing such target-decorated cells). The emulsion microdroplets can be monodisperse or polydisperse.

[0078] In some embodiments, the construct contains an FRT site or other site for introducing the coding sequence of a target of interest. In some embodiments, the construct contains a self-cleaving site such as P2A or T2A. In some embodiments, the construct contains a gene (e.g., glutamine synthetase (GS)) that allows for selection of cells with stable integration of the construct.

[0079] In some embodiments, the target of interest is expressed from the genome of the target decorated cell. In some embodiments, the genome is genetically modified to include a coding sequence for a fusion protein or a transmembrane protein. In some embodiments, the target of interest is expressed from an endogenous sequence in the genome.

[0080] A library of target-decorated cells can comprise single cell clones that display a common target of interest. In some embodiments, a library of target-decorated cells comprises more than one cell clone, where each cell clone expresses a unique target of interest. In some embodiments, a library of target-decorated cells comprises more than one cell clone, where each cell clone expresses a unique variant of a particular target. In some embodiments, the targets of interest expressed on the library of target-decorated cells are different from each other. In some embodiments, the targets of interest expressed on the library of target-decorated cells are similar to each other. In some embodiments, the targets of interest expressed on the library of target-decorated cells have similar protein sequences. In some embodiments, the targets of interest expressed on the library of target-decorated cells are targets of the same or related proteins.

[0081] In some embodiments, the library of target-decorated cells comprises two, three, four, or more cell clones, where each cell clone displays a unique target of interest that is different from the other cell clones. In some embodiments, the library of target-decorated cells comprises at least five cell clones, where each cell clone displays a unique target of interest that is different from the other cell clones. In some embodiments, the library of target-decorated cells comprises at least 10 cell clones, where each cell clone displays a unique target of interest that is different from the other cell clones. In some embodiments, the library of target-decorated cells comprises at least 100 cell clones, where each cell clone displays a unique target of interest that is different from the other cell clones. In some embodiments, the library of target-decorated cells comprises between 100 and 1,000 cell clones, where each cell clone displays a unique target of interest that is different from the other cell clones. In some embodiments, the library of target-decorated cells comprises more than 1,000 cell clones, where each cell clone displays a unique target of interest that is different from the other cell clones. In some embodiments, the library of target-decorated cells comprises 30-1000 cell clones, where each cell clone displays a unique target of interest that is distinct from the other cell clones.

[0082] The methods described herein can use a variety of targets of interest. In some embodiments, a target specific to a disease (e.g., cancer or an immune disease) is used. In some embodiments, a target specific to a pathogen (e.g., a bacterium or virus) is used. In some embodiments, the target is a ligand for a receptor. In some embodiments, the target is a receptor, such as a multi-spanning membrane protein.

[0083] Various cell types can be used as target decorative cells in the methods described herein. The cells can be prokaryotic (e.g., bacterial cells) or eukaryotic (e.g., mammalian or plant cells, fungi). The cells can be primary cells or cell lines. In some embodiments, the cells are cancer cells. In some embodiments, the cells are derived from multicellular organisms, including, for example, birds, plants, and mammals such as humans, cows, sheep, apes, monkeys, pigs, rats, mice, dogs, and cats. In some embodiments, the cells are derived from unicellular organisms, including, for example, bacteria and yeast. The cells can be prokaryotic, such as E. coli, or eukaryotic, such as unicellular eukaryotes (e.g., yeast or other fungi), plant cells (e.g., tobacco or tomato plant cells), animal cells (e.g., human cells, monkey cells, hamster cells, rat cells, mouse cells, or insect cells), or hybridomas. Exemplary cells include CS-9 cells, the COS-7 line of monkey kidney cells (ATCC CRL 1651) (see Gluzman et al., 1981, Cell 23:175), L cells, C127 cells, 3T3 cells (ATCC CCL 163), Chinese hamster ovary (CHO) cells or their derivatives, such as Veggie CHO and related cell lines that grow in serum-free medium (see Rasmussen et al., 1998, Cytotechnology 28:31), HeLa cells, the BHK (ATCC CRL 10) cell line, the CV1 / EBNA cell line derived from the African green monkey kidney cell line CV1 (ATCC CCL 70) (see McMahan et al., 1991, EMBO J. 10:2821), human embryonic kidney cells, such as 293, 293 EBNA or MSR. 293, human epidermal A431 cells, human Colo205 cells, other transformed primate cell lines, normal diploid cells, cell lines derived from in vitro culture of primary tissue, primary explants, HL-60, U937, HaK, or Jurkat cells. In some embodiments, the cells are Expi293 cells or CHOZN cells. Typically, the cells are cultured cells that can be transformed or transfected with a nucleic acid encoding a polypeptide, which can then be expressed in the cell.

[0084] 6.3.2. TBP-ribosome-mRNA complex (TRM complex) The disclosed methods use multiple TBP-ribosome-mRNA (TRM) complexes that comprise target binding proteins. In some embodiments, each of the TRM complexes comprises a TBP that comprises an scFv. In some embodiments, each of the TRM complexes comprises a TBP that comprises a heavy chain variable region. In some embodiments, each of the TRM complexes comprises a TBP that comprises a light chain variable region. In some embodiments, each of the TRM complexes comprises a single-chain binding fragment (e.g., a V H / κ) containing TBP.

[0085] In some embodiments, the plurality of TRM complexes comprises one unique TRM complex comprising a target binding protein. In some embodiments, the plurality of TRM complexes comprises 1-5 unique TRM complexes, wherein each of the unique TRM complexes comprises a unique target binding protein that is different from the rest of the unique TRM complex. In some embodiments, the plurality of TRM complexes comprises 6-10 unique TRM complexes, wherein each of the unique TRM complexes comprises a unique target binding protein that is different from the rest of the unique TRM complex. In some embodiments, the plurality of TRM complexes comprises at least 10 unique TRM complexes, wherein each of the unique TRM complexes comprises a unique target binding protein that is different from other unique TRM complexes. In some embodiments, the plurality of TRM complexes comprises at least 100 unique TRM complexes, wherein each of the unique TRM complexes comprises a unique target binding protein that is different from other unique TRM complexes. In some embodiments, the plurality of TRM complexes comprises at least 1000 unique TRM complexes, wherein each of the unique TRM complexes comprises a unique target binding protein that is different from other unique TRM complexes. In some embodiments, the plurality of TRM complexes comprises at least 10,000 unique TRM complexes, wherein each of the unique TRM complexes comprises a unique target binding protein that is distinct from other unique TRM complexes.

[0086] In some embodiments, the plurality of TRM complexes comprises one or more target binding proteins with known targets, hi some embodiments, the plurality of TRM complexes comprises one or more target binding proteins with unknown targets.

[0087] In some embodiments, the multiple TRM complexes comprise more than one unique target binding protein, each binding to a unique epitope on the same target. In some embodiments, the multiple TRM complexes comprise more than one unique target binding protein, each binding to the same epitope.

[0088] In some embodiments, each TRM complex comprises a target binding protein and an mRNA encoding the target binding protein. In some embodiments, the mRNA comprises a coding sequence for a complementarity determining region (CDR) of the target binding protein. In some embodiments, the mRNA comprises a coding sequence for a CDR3 of the target binding protein. In some embodiments, the mRNA comprises a coding sequence for a V H In some embodiments, the mRNA comprises a coding sequence for the region. In some embodiments, the mRNA comprises a barcode sequence. In some embodiments, the barcode sequence identifies the target binding protein encoded by the mRNA.

[0089] 6.3.3. Analysis Process 6.3.3.1 Contact of target-decorated cells with TBP-ribosome-mRNA (TRM) complexes The method disclosed herein includes contacting target-decorated cells with a TRM complex. This step is carried out under conditions that allow binding between the target-decorated cells and the TRM complex. The conditions can be adjusted or optimized to allow the target-binding proteins of the TRM complex to specifically bind to their targets. For example, the conditions can be adjusted or optimized by changing the composition of the buffer.

[0090] In some embodiments, the contacting step comprises contacting a 2+In some embodiments, the buffer contains 25 mM to 100 mM Mg 2+ In some embodiments, the buffer contains 50 mM Mg 2+ In some embodiments, the buffer contains 25 mM Mg 2+ In some embodiments, the buffer contains 75 mM Mg 2+ In some embodiments, the buffer contains 100 mM Mg 2+ Includes.

[0091] In some embodiments, the contacting step is carried out in a buffer containing MgCl2. In some embodiments, the buffer contains 25 mM to 100 mM MgCl2. In some embodiments, the buffer contains 50 mM MgCl2. In some embodiments, the buffer contains 25 mM MgCl2. In some embodiments, the buffer contains 75 mM MgCl2. In some embodiments, the buffer contains 100 mM MgCl2.

[0092] In some embodiments, the buffer further comprises HEPES. In some embodiments, the buffer further comprises a salt. In some embodiments, the buffer further comprises NaCl. In some embodiments, the buffer further comprises polysorbate 20. In some embodiments, the buffer further comprises heparin. In some embodiments, the buffer further comprises BSA. In some embodiments, the buffer further comprises an RNase inhibitor.

[0093] In some embodiments, the buffer comprises HEPES, NaCl, 50 mM MgCl2, and BSA. In some embodiments, the buffer comprises HEPES, NaCl, 50 mM MgCl2, polysorbate 20, heparin, and BSA. In some embodiments, the buffer comprises 20 mM HEPES, 50 mM NaCl, 50 mM MgCl2, 0.01% polysorbate 20, 2.5 mg / ml heparin, and 0.5% BSA. In some embodiments, the buffer comprises 20 mM HEPES, 50 mM NaCl, 50 mM MgCl2, 0.01% polysorbate 20, 2.5 mg / ml heparin, and 0.05% BSA. In some embodiments, the buffer further comprises an RNase inhibitor.

[0094] 6.3.3.2 Microdroplet generation In the method provided herein, the sample containing target decorated cells and TRM complexes obtained from the contact step can be isolated into microdroplets.Monodisperse emulsions can be formed on microfluidic chips, or polydisperse emulsions can be formed using machines such as the IKA Utra-Turrax Tube Drive system.In some embodiments, a microfluidic system (e.g., Dolomite microfluidics) with three pressure pumps is used.Methods for generating microdroplets are known in the art, for example, as disclosed in International Publication No. 2016200577A1, the entire contents of which are incorporated by reference.

[0095] In some embodiments, a microfluidic device is used to generate single-cell emulsion droplets. The microfluidic device ejects single cells in an aqueous reaction buffer into a hydrophobic oil mixture. The device can generate thousands of emulsion microdroplets per second. After the emulsion microdroplets are generated, the device ejects the emulsion mixture into a trough. The mixture can be pipetted or collected into standard reaction tubes for downstream processing, such as thermal cycling.

[0096] Custom microfluidic devices for single-cell analysis are routinely fabricated in academic and commercial laboratories (Kintses et al., 2010 Current Opinion in Chemical Biology 14:548-555). For example, chips can be made from polydimethylsiloxane (PDMS), plastic, glass, or quartz. In some embodiments, fluids move through the chip by pressure or the action of a syringe pump. Single cells can even be manipulated on programmable microfluidic chips using custom dielectrophoresis devices (Hunt et al., 2008 Lab Chip 8:81-87). In one embodiment, a pressure-based PDMS chip composed of flow-focusing geometries fabricated with soft lithography techniques is used (e.g., Dolomite Microfluidics, Royston, UK) (Anna et al., 2003 Applied Physics Letters 82:364-366). The stock design can typically generate 10,000 monodisperse water-in-oil microdroplets per second in the size range of 10-150 μm in diameter. In some embodiments, the hydrophobic phase consists of a fluorinated oil containing an ammonium salt of carboxy-perfluoropolyether, which ensures optimal conditions for molecular biology and reduces the probability of droplet coalescence (Johnston et al., 1996 Science 271:624-626). To measure the periodicity of cell and droplet flow, images are recorded at 50,000 frames per second using standard techniques such as a Phantom V7 camera or Fastec InLine (Abate et al., 2009 Lab Chip 9:2628-31).

[0097] Microfluidic systems can optimize droplet size, input cell density, chip design, and cell loading parameters so that more than 98% of droplets contain single cells. There are three common methods for achieving such statistics: (i) extreme dilution of the cell solution; (ii) fluorescent selection of droplets containing single cells; and (iii) fluorescent cell sorting to reduce extreme cell dilution and negative rates to control the multiple hit rate. In some embodiments, the target-decorated cells contain a coding sequence for a reporter protein, such as a fluorescent protein. In some such embodiments, the reporter protein can be used to detect and identify droplets containing target-decorated cells. In certain embodiments, the fluorescent protein is used to isolate droplets containing single target-decorated cells.

[0098] In some embodiments, the flow of input cells is aligned with the periodicity of droplet formation so that more than 98% of droplets contain a single cell (Edd et al., 2008 Lab Chip 8:1262-1264; Abate et al., 2009 Lab Chip 9:2628-31). In these microfluidic devices, a dense suspension of cells is forced through a high aspect ratio channel such that the cell diameter is a large fraction of the channel width. Several input channel widths and flow rates can be tested to arrive at an optimal solution.

[0099] In certain embodiments of the present invention, microfluidic chips are used to isolate 10, 100, 1000, 10,000, 100,000, 1 million, or 1 billion single cells from a heterogeneous pool of target decorated cells. In some embodiments, the methods of the present invention use single cells in reaction vessels rather than emulsion droplets. Examples of such reaction vessels include 96-well plates, 0.2 mL tubes, 0.5 mL tubes, 1.5 mL tubes, 384-well plates, 1536-well plates, etc.

[0100] In some embodiments, the isolated target decorated cells in the microdroplets are bound to TRM complexes. In some embodiments, the isolated target decorated cells in the microdroplets are not bound to TRM complexes.

[0101] In some embodiments, the isolated target decorated cells are encapsulated with a lysis reagent, hi some embodiments, the isolated target decorated cells are under conditions that allow the lysis reagent to induce lysis of the target decorated cells.

[0102] 6.3.3.3 Capture of RNA released from single cells The methods described herein further include capturing RNA released from the target decorated cells. In some such embodiments, RNA molecules can be captured using beads containing polynucleic acid probes targeting the RNA molecules. In one embodiment, the beads are spherical beads comprising agarose, glass, chemical polymers, or magnetic materials. In some embodiments, the beads are made of materials such as latex, glass, or silica, ranging in size from 0.1 microns to 1 mm. In some embodiments, the beads have a diameter greater than 10 μm. In some embodiments, the beads have a diameter greater than 100 μm. In some embodiments, the beads have a diameter less than 1 mm. In some embodiments, the beads have a diameter of 0.5-10 μm, less than 1 μm, or about 1 μm. In one embodiment, the probe comprises biotin, and the beads comprise streptavidin attached to the surface of the beads. In some embodiments, the beads are solid or porous beads.

[0103] In some embodiments of the present invention, target-decorated cells are encapsulated in droplets with beads containing bound polynucleic acid probes with sequences complementary to the desired polynucleic acid targets in the single cells. In certain embodiments, the probes are 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, 120, 130, or more nucleotides in length. The probes are RNA, DNA, locked nucleic acid (LNA), peptide nucleic acid (PNA), glycol nucleic acid (GNA), or any nucleic acid analog. The polynucleic acid targets are either DNA or RNA. The reagent mixture present in these bead-containing embodiments is specifically designed to lyse the cells and promote polynucleic acid hybridization, thus allowing the beads to capture the desired DNA or RNA targets.

[0104] In some embodiments of the present invention, the polynucleic acid probe targets the constant region of an immunoglobulin. In some embodiments, the polynucleic acid probe targets the constant region of an IgK or IgG. In some embodiments, the polynucleic acid probe targets an mRNA sequence in a TRM complex. In some embodiments, the polynucleic acid probe comprises a sequence complementary to an mRNA sequence or a fragment thereof in a TRM complex.

[0105] In some embodiments of the present invention, the polynucleic acid probe targets a transcript in a target decorated cell. In some embodiments, the polynucleic acid probe targets a transcript of a target of interest in a target decorated cell. In some embodiments, the polynucleic acid probe targets a transcript of a fusion protein comprising the target of interest and a transmembrane domain. In some embodiments, the polynucleic acid probe targets a transcript of the transmembrane domain of the fusion protein. In some embodiments, the polynucleic acid probe comprises an oligo(dT) polynucleotide that can hybridize to the poly(A) tail of an mRNA.

[0106] In some embodiments, the polynucleic acid probe targets a barcode sequence. In some embodiments, the polynucleic acid probe targets both the TRM complex (e.g., an mRNA within the TRM complex) and the transcript of the fusion protein or target of interest.

[0107] In some embodiments, a mixture of several polynucleic acid probes is used. In some such embodiments, a first set of polynucleic acid probes targeting a TRM complex can be combined with a second set of polynucleic acid probes targeting a fusion protein or transcript of a target of interest.

[0108] In some embodiments, 5'-amino modified polynucleic acid probes are bound to carboxylic acid beads using 2-(N-morpholino)ethanesulfonic acid (MES) buffer (Kojima et al., 2005, Nucleic Acids Research 33:el50). In other embodiments, biotinylated polynucleic acid probes are bound to streptavidin-coated beads.

[0109] In some embodiments, the methods of the present invention use single cells in reaction vessels rather than emulsion droplets. Examples of such reaction vessels include 96-well plates, 0.2 mL tubes, 0.5 mL tubes, 1.5 mL tubes, 384-well plates, 1536-well plates, etc. Various other designs of microfluidic chips can also be used to isolate single cells (Marcus et al., 2006, Anal Chem 78:3084-3089).

[0110] In some embodiments, cells or a subpopulation of cells are added to a reaction vessel together with beads containing bound polynucleic acid probes and a lysis buffer. The lysis buffer lyses the cells, allowing the polynucleic acid probes to bind to the desired polynucleic acid targets from the cell(s). Beads hybridized to the polynucleic acid targets are isolated from the lysis buffer as single beads or subpopulations of beads into a reaction vessel and contacted with a PCR mixture to allow amplification and / or fusion of the polynucleic acid targets.

[0111] In certain embodiments of the present invention, the aqueous phase of the droplet emulsion containing the beads and their bound targets is recovered using a solvent such as ethyl ether. The beads are isolated into an emulsion using a PCR mix so that, on average, a single bead is isolated into a single emulsion microdroplet (DeKosky et al., 2015, Nat Med 21:86-91). Monodisperse emulsions can be formed on a microfluidic chip, or polydisperse emulsions can be formed using a machine such as the IKA Utra-Turrax Tube Drive system. PCR reagents amplify multiple DNA or RNA targets of interest. In some embodiments of the present invention, the PCR product is a ligated transcript of the fusion protein and the target of interest in the target-decorated cell and the mRNA of the TRM complex.

[0112] In some embodiments, the step of capturing the RNA released from the target-decorated cells is omitted. In some cases, the target-decorated cells are encapsulated in a semipermeable shell, thereby enabling a workflow without the step of capturing the RNA released from the target-decorated cells. In some such embodiments, the following steps for generating a library of hybrid polynucleotide acids are carried out in microdroplets containing the RNA released from the target-decorated cells. The amplification reaction can be carried out in microdroplets loaded with reagents for reverse transcription and PCR (e.g., OE-RT-PCR) reactions.

[0113] 6.3.3.4 Generation of a Library of Hybrid Polynucleic Acids The methods described herein further include generating a library of hybrid polynucleic acids comprising sequences from transcripts from a single cell and / or sequences from mRNA of a TRM complex. In some embodiments, the hybrid polynucleic acid comprises sequences from transcripts from a single cell and sequences from mRNA of a TRM complex. In other embodiments, the hybrid polynucleic acid comprises sequences from transcripts from a single cell or sequences from mRNA of a TRM complex. The sequences from the transcripts from the single cell can identify targets expressed from the single cell. The sequences from the mRNA can be from a TRM complex bound to the single cell and can therefore be used to identify the TRM complex. In some embodiments, the hybrid polynucleic acid described herein further comprises sequences associated with the single cell. For example, in some embodiments, the hybrid polynucleic acid further comprises barcode sequences from oligonucleotides attached to particles, such as beads, associated with the single cell. In some such embodiments, the hybrid polynucleic acids comprise: a) a sequence from a transcript identifying an expressed target of the single cell and a sequence (e.g., a barcode sequence) from an oligonucleotide immobilized on a particle (e.g., a bead) associated with the single cell; or b) a sequence from an mRNA of a TRM complex bound to the single cell and a sequence (e.g., a barcode sequence) from an oligonucleotide immobilized on a particle (e.g., a bead) associated with the single cell. In some embodiments, the library of hybrid polynucleic acids comprises a first set of hybrid polynucleic acids and a second set of hybrid polynucleic acids, wherein: a) the hybrid polynucleic acids of the first set of hybrid polynucleic acids comprise a sequence from a transcript identifying an expressed target of the single cell and a sequence (e.g., a barcode sequence) from an oligonucleotide immobilized on a particle (e.g., a bead) associated with the single cell; and b) the hybrid polynucleic acids of the second set of hybrid polynucleic acids comprise a sequence from an mRNA of a TRM complex bound to the single cell and a sequence (e.g., a barcode sequence) from an oligonucleotide immobilized on a particle (e.g., a bead) associated with the single cell.In some such embodiments, the library of hybrid polynucleic acids comprises sequences associated with a plurality of cells, and for each of the plurality of cells, the barcode sequence of the oligonucleotides attached to particles associated with the given cell is unique compared to the barcode sequences of the oligonucleotides attached to particles associated with other cells, thereby allowing the hybrid polynucleic acids comprising sequences associated with a given cell to share a common barcode that is distinct from the barcodes shared by other hybrid polynucleic acids comprising sequences associated with other cells.

[0114] Hybrid polynucleic acids can be generated from RNA isolated from microdroplets. Specifically, hybrid polynucleic acids can be generated from RNA released from target-decorated cells and TRM complexes isolated into the same microdroplets. In some such embodiments, the target of interest in the target-decorated cells and the target-binding protein of the TRM complex can have high affinity and specificity. In some embodiments, RNA is captured using beads containing polynucleic acid probes that target RNA molecules. In some embodiments, the beads contain polynucleic acid probes that contain barcode sequences. In some embodiments, the barcode sequences of the polynucleic acid probes are unique to each microdroplet.

[0115] In some embodiments, PCR is used to generate hybrid polynucleic acids. In some embodiments, the hybrid polynucleic acids are generated by high-throughput amplification in a reaction vessel. As used herein, the term "reaction vessel" refers to any entity that provides physical separation of reactants into separate compartments. A reaction vessel can be a plastic compartment, a microfluidic chamber, or, for example, a droplet of aqueous reaction solution. When droplets are used, the method described herein further comprises generating droplets containing a reaction mixture with bead-isolated RNA.

[0116] In some embodiments, first strand cDNA synthesis precedes the generation of hybrid polynucleic acids. In some embodiments, oligo(dT), random primers, or a combination thereof are used to prime the reverse transcription reaction. In some embodiments, oligo(dT) is used to initiate synthesis, preferably at the 3' end of the RNA fragment.

[0117] In some embodiments, the method for overlap extension PCR is used to generate the fusion amplicon product of the transcript of target decorated cell and the sequence from TRM complex in a single reaction tube or microdroplet.In some embodiments, the fusion amplicon product further comprises the sequence (e.g., barcode sequence) from the oligonucleotide fixed on the particle (e.g., bead) associated with target decorated cell.The method for amplifying and linking amplicon is disclosed in Johnson et al., 2005 Genome Research 15:1315-24; U.S. Patent No. 7,749,697; PCT Publication WO 2012 / 083225; and PCT Publication WO 2013 / 096643, each of which is incorporated herein by reference in its entirety.

[0118] In some embodiments, at least two nucleic acid target sequences (e.g., a first target sequence in a transcript of a target-decorated cell and a second target sequence in a TRM complex) are selected and designated as target sequences. Forward and backward primers are designed for each of the two nucleic acid target sequences, and the primers are used to amplify the target sequences. A "minor" amplicon is generated by amplifying the two nucleic acid target sequences separately, and then fused by amplification to create a fusion amplicon, also known as a "major" amplicon. In one embodiment, the "minor" amplicon is a nucleic acid sequence amplified from the first target sequence, and the "major" amplicon is a fusion complex, e.g., a recombinant fusion polynucleotide, generated from the sequence amplified between multiple target sequences. In some embodiments, the target sequences further include a third target sequence in an oligonucleotide immobilized on a particle (e.g., a bead) associated with the target-decorated cell.

[0119] The methods described herein use "internal" primers (i.e., a reverse primer for a first target sequence and a forward primer for a second target sequence) that contain one domain that hybridizes to the minor amplicon and a second domain that hybridizes to the second minor amplicon. The "internal" primers are limiting reagents such that during the exponential phase of PCR, the internal primers are used up, driving the overlapping domain within the minor amplicon to anneal and generate the major amplicon.

[0120] PCR primers are designed for the target of interest using standard parameters, e.g., a melting temperature (Tm) of approximately 55-65°C, with a length of 20-50 nucleotides. Primers are used with standard PCR conditions, e.g., 1 mM Tris-HCl pH 8.3, 5 mM potassium chloride, 0.15 mM magnesium chloride, 0.2-2 μM primers, 200 μM dNTPs, and a thermostable DNA polymerase. Many commercially available kits are available for performing PCR, such as Platinum Taq (Life Technologies), Amplitaq Gold (Life Technologies), Titanium Taq (Clontech), Phusion polymerase (Finnzymes), and HotStartTaq Plus (Qiagen). Any standard thermostable DNA polymerase, such as Taq polymerase or Stoffel fragment, can be used for this step.

[0121] In one embodiment, a set of nucleic acid probes (or primers) including a first probe, a second probe, a third probe, and a fourth probe is used to amplify a first target nucleic acid sequence and a second target nucleic acid sequence to form a fusion complex. The first probe includes a sequence complementary to the first target nucleic acid sequence (e.g., the 5' end of the first target nucleic acid sequence). The second probe includes a sequence complementary to the first target nucleic acid sequence (e.g., the 3' end of the first target nucleic acid sequence) and a second sequence complementary to an exogenous sequence. In some embodiments, the exogenous sequence is a non-human nucleic acid sequence and is not complementary to any of the target nucleic acid sequences. For example, the exogenous sequence may be a polynucleotide sequence encoding a polypeptide sequence rich in Ser and Gly amino acids that links the heavy chain variable region and light chain variable region in an scFv (see, e.g., PCT / US1992 / 001478). The first and second probes are forward and reverse primers for a first target nucleic acid sequence.

[0122] The third probe contains a sequence complementary to the portion of the second probe that is complementary to the exogenous sequence and a sequence complementary to the second target nucleic acid sequence (e.g., the 5' end of the second target nucleic acid sequence). The fourth probe contains a sequence complementary to the second target nucleic acid sequence (e.g., the 3' end of the second target nucleic acid sequence). The third and fourth probes are forward and reverse primers for the second target nucleic acid sequence.

[0123] The second and third probes are also called the "inner" primers of the reaction (i.e., the reverse primer for the first locus and the forward primer for the second locus) and are concentration-limited (e.g., 0.01 µM for the inner primers and 0.1 µM for all other primers). This drives amplification of the major amplicon preferentially over the minor amplicon. The first and fourth probes are called "outer" primers.

[0124] The first and second nucleic acid sequences are independently amplified, with the first nucleic acid sequence being amplified using the first and second probes, and the second nucleic acid sequence being amplified using the third and fourth probes.Then, the complementary sequence regions of the amplified first and second nucleic acid sequences are hybridized, and the hybridized sequence is amplified using the first and fourth probes to generate a fusion complex.This is called overlap extension PCR amplification.

[0125] During overlap extension PCR amplification, complementary regions of the amplified first and second nucleic acid sequences act as primers for extension on both strands and in each direction by DNA polymerase molecules. In subsequent PCR cycles, the outer primers prime the complete fusion sequence so that the fusion complex is replicated by the DNA polymerase. This method produces multiple fusion complexes.

[0126] In some embodiments, the fusion complex is cloned into an expression vector. In some embodiments, the expression vector is a plasmid or a phagemid. In one embodiment, the regulatory sequence (e.g., promoter) is inserted into the expression vector by a method selected from Gibson assembly, site-specific digestion and ligation, and targeted recombination.

[0127] In some embodiments, the coding sequence for the target-binding protein in the fusion complex is cloned separately into an expression vector. The expression vector may comprise the coding sequence for the target-binding protein operably linked to a regulatory sequence capable of directing expression of the target-binding protein.

[0128] In one aspect, the present disclosure provides a library of hybrid polynucleic acids produced by the methods described herein.

[0129] In some embodiments, the methods described herein do not include generating a hybrid polynucleic acid library containing sequences from transcripts identifying the expressed target of a single cell and sequences from the mRNA of a TRM complex bound to that cell. Instead of generating such hybrid polynucleic acids, the methods may include generating a first library of polynucleic acids containing sequences from the transcripts from the single cell and a first barcode, and a second library of polynucleic acids containing sequences from the mRNA of the TRM complex and a second barcode. The first barcode and the second barcode can be used to identify the target expressed from the single target-decorated cell and the TRM complex bound to the single target-decorated cell. For example, the first barcode and the second barcode from a single microdroplet can be the same or related. The methods may further include analyzing the first library of polynucleic acids and the second library of polynucleic acids to identify the target of interest and a target-binding protein capable of binding to the target of interest.

[0130] 6.3.3.5 Analysis of Hybrid Polynucleic Acids The methods described herein can further include analyzing the library of hybrid polynucleic acids. In some embodiments, the analysis includes sequencing. DNA sequencing is used to verify the construct or library of constructs. In some embodiments, DNA sequencing is performed using massively parallel methods from suppliers such as Illumina or single clone sequencing methods such as Sanger sequencing from suppliers such as Applied Biosystems. The sequence of the hybridized polynucleic acid can include the sequence of a target of interest and / or the sequence of a TRM complex, or a portion thereof. Thus, the sequence can be used to identify the target of interest expressed on the surface of the target-decorated cell and the TRM complex bound to the target of interest. The sequence of the hybridized polynucleic acid can further include the sequence, or a portion thereof, of an oligonucleotide immobilized on a particle (e.g., a bead) associated with the target-decorated cell. Thus, the pair of target of interest and bound TRM complex from a given cell can be identified.

[0131] The analysis can include identifying the target of interest and the TRM complex isolated in the microdroplet. The TRM complex that binds to the target of interest can be isolated in the same microdroplet together with the target-decorated cells that express the target of interest. Thus, in some embodiments, the analysis includes identifying the target of interest and the target binding protein that binds to the target of interest. In some cases, the analysis includes identifying the target of interest and the target binding protein that can bind to the target of interest. In some embodiments, the analysis can further include identifying multiple TRM complexes that bind to the target of interest in the same microdroplet, for example, the microdroplet contains particles (e.g., beads) to which oligonucleotides containing barcode sequences are immobilized.

[0132] In some embodiments, a target-decorated cell can bind to two or more TRM complexes. In some such embodiments, multiple TRM complexes are isolated in the same microdroplet along with the target-decorated cell. In some embodiments, RNA captured from a microdroplet can contain sequences from more than one TRM complex. Thus, hybridized polynucleic acids generated from a single microdroplet can contain one or more pairs of a target of interest and a target-binding protein. Thus, in some embodiments, multiple target-binding proteins that bind to the same target of interest are identified. In some embodiments, hybridized polynucleic acids generated from a single microdroplet can contain barcode sequences from oligonucleotides immobilized on particles (e.g., beads) in the microdroplet.

[0133] In some embodiments, the methods described herein can be performed under conditions of varying degrees of stringency, and the identified pairs of target of interest and TRM complexes under the various conditions can be analyzed to test their binding affinity.

[0134] In some embodiments, the methods described herein are used to identify a subset of target-binding proteins that have high binding affinity for a target of interest. In some embodiments, the methods are used to identify a subset of target-binding proteins that have a desired binding affinity for a target of interest.

[0135] In some embodiments, the analysis involves quantitation of specific pairs of targets of interest and target-binding proteins. The quantitative data can be used to measure the affinity and specificity of the target-binding proteins for the targets of interest.

[0136] In some embodiments, the analysis involves identifying all target-binding proteins that bind to a particular target of interest. In some embodiments, the analysis involves quantitating target-binding proteins that bind to a particular target of interest. Quantitative data can be used to identify target-binding proteins with desired affinity and / or specificity for the target of interest. In some embodiments, quantitation can be used to determine the binding affinity and / or specificity of the target-binding protein for the target of interest by comparing data from known target-TBP pairs. In some embodiments, the above process can be used for affinity maturation of antibodies by testing the binding of various modifications of the antibody to the target of interest. Specific modified antibodies can be selected based on their binding activity (e.g., affinity and specificity) for the target of interest.

[0137] In some embodiments, the analysis includes identifying all targets of interest that bind to a particular target-binding protein. In some embodiments, the analysis includes quantitating targets of interest that bind to a particular target-binding protein. In some embodiments, the methods described herein are used to characterize individual antibodies in a polyclonal mixture without the need for isolation and purification. In some embodiments, the methods are used for high-throughput screening of binding partners between a diverse library of antibodies and a diverse library of targets (e.g., a library of naturally occurring target variants). Quantitation can be used to identify targets of the target-binding protein.

[0138] In some embodiments, the disclosed methods are used to analyze the binding of polyclonal antibodies to a target of interest. In some embodiments, the disclosed methods are used to analyze the binding of polyclonal antibodies in a sample from a donor to a target of interest in order to select donors. In some embodiments, the methods are used to analyze and monitor the quality or production yield and consistency of polyclonal antibodies to a target of interest. In some embodiments, the methods are used to characterize polyclonal antibodies. In some embodiments, the methods are used for QC monitoring of antibody libraries over time, assessment of library diversity, epitope mapping, or functional evaluation of antibodies against mutations.

[0139] In some embodiments, the methods described herein are used to identify antibody epitopes by identifying binding between an antibody and a library of targets, where the targets are variants of a target protein with one or more mutations at different sites. Epitopes can be identified by analyzing which mutations cause the antibody to lose binding to the target protein. In some embodiments, the methods can be used to identify epitopes of polyclonal antibodies. In some such embodiments, the diversity of polyclonal antibodies can be determined based on epitope mapping.

[0140] 6.3.4. Analysis of Interactions Between Ligands and Receptors The high-throughput analytical methods disclosed herein can be used to study the interaction between a ligand and a receptor. In some such embodiments, the target can be a receptor or a fragment thereof, and the ABP can be a ligand or a fragment thereof. Alternatively, the target can be a ligand or a fragment thereof, and the ABP can be a receptor or a fragment thereof. In some embodiments, the method comprises the following steps: (i) providing a library of ligand-decorated cells, each of which displays a ligand of interest on its cell membrane; (ii) contacting the library of ligand-decorated cells with a plurality of receptor-ribosome-mRNA (RRM) complexes, thereby inducing binding between the ligand-decorated cells and the RRM complexes; (iii) generating a plurality of emulsion microdroplets, each microdroplet containing a single one of the ligand-decorated cells, one or more RRM complexes bound to the single cell, and a lysis reagent that induces lysis of the single cell; (iv) capturing RNA released from the single cell on a solid surface or within a semipermeable shell; and (v) generating a library of hybrid polynucleic acids comprising sequences from transcripts of the single cells and / or sequences from the mRNA of the RRM complexes. In some embodiments, the method comprises the following steps: (i) providing a library of receptor-decorated cells, each of which displays a receptor of interest on its membrane; (ii) contacting the library of receptor-decorated cells with a plurality of ligand-ribosome-mRNA (LRM) complexes, thereby inducing binding between the receptor-decorated cells and the LRM complexes; (iii) generating a plurality of emulsion microdroplets, each microdroplet containing a single one of the receptor-decorated cells, one or more LRM complexes bound to the single cell, and a lysis reagent that induces lysis of the single cell; (iv) capturing RNA released from the single cell on a solid surface or within a semipermeable shell; and (v) generating a library of hybrid polynucleic acids comprising sequences from transcripts of the single cells and / or sequences from the mRNA of the LRM complexes.

[0141] The hybrid polynucleic acid library can be analyzed or sequenced to provide information about the binding between the ligand of interest and the RRM complex or between the receptor of interest and the LRM complex. For example, the receptors that bind to the ligand of interest and / or their binding affinity or specificity can be studied based on the sequences. Thus, the method can further include identifying ligand-receptor pairs based on sequencing the hybrid polynucleic acid library. In some embodiments, the method further includes identifying a ligand-binding protein specific to the ligand of interest. In some embodiments, the method further includes identifying the binding affinity or specificity of the ligand-binding protein specific to the ligand of interest.

[0142] 6.4. Kits for High-Throughput Analysis of Antibodies The present disclosure also provides kits for use in high-throughput analysis of antibodies.

[0143] The kit can include multiple polynucleotide constructs, each construct encoding a fusion protein comprising a unique target of interest and a transmembrane domain, hi some embodiments, the kit includes multiple polynucleotide constructs, each construct comprising a coding sequence for a transmembrane domain and an insertion site for adding a coding sequence for a target of interest.

[0144] In some embodiments, the kit further comprises a plurality of TBP-ribosome-mRNA (TRM) complexes, each TRM complex comprising a unique target binding protein. In some embodiments, the plurality of TRM complexes comprises 1, 2, 3, 4, 5, or more unique target binding proteins. In some embodiments, the plurality of TRM complexes comprises more than 5 unique target binding proteins. In some embodiments, the plurality of TRM complexes comprises more than 10 unique target binding proteins. In some embodiments, the plurality of TRM complexes comprises more than 50 unique target binding proteins. In some embodiments, the plurality of TRM complexes comprises more than 100 unique target binding proteins. In some embodiments, the plurality of TRM complexes comprises more than 500 unique target binding proteins. In some embodiments, the plurality of TRM complexes comprises more than 1000 unique target binding proteins. In some embodiments, the plurality of TRM complexes comprises more than 10,000 unique target binding proteins.

[0145] In some embodiments, the kit further comprises reagents for PCR. In some embodiments, the kit further comprises reagents for overlap extension polymerase chain reaction (OE-PCR). In some embodiments, the kit further comprises reagents for first strand cDNA synthesis. In some embodiments, the kit further comprises a polymerase.

[0146] In some embodiments, the kit further comprises a buffer that can be used in the step of contacting the target decorated cells with the TRM complex. In some embodiments, the buffer contains Mg 2+ In some embodiments, the buffer contains 25 mM to 100 mM Mg 2+ In some embodiments, the buffer contains 50 mM Mg 2+ In some embodiments, the buffer contains 25 mM Mg 2+ In some embodiments, the buffer contains 75 mM Mg 2+ In some embodiments, the buffer contains 100 mM Mg 2+ Includes:

[0147] In some embodiments, the buffer contains MgCl2. In some embodiments, the buffer contains 25 mM to 100 mM MgCl2. In some embodiments, the buffer contains 50 mM MgCl2. In some embodiments, the buffer contains 25 mM MgCl2. In some embodiments, the buffer contains 75 mM MgCl2. In some embodiments, the buffer contains 100 mM MgCl2.

[0148] In some embodiments, the buffer further comprises HEPES. In some embodiments, the buffer further comprises a salt. In some embodiments, the buffer further comprises NaCl. In some embodiments, the buffer further comprises polysorbate 20. In some embodiments, the buffer further comprises heparin. In some embodiments, the buffer further comprises BSA.

[0149] In some embodiments, the buffer comprises HEPES, NaCl, 50 mM MgCl2, and BSA. In some embodiments, the buffer comprises HEPES, NaCl, 50 mM MgCl2, polysorbate 20, heparin, and BSA. In some embodiments, the buffer comprises 20 mM HEPES, 50 mM NaCl, 50 mM MgCl2, 0.01% polysorbate 20, 2.5 mg / ml heparin, and 0.5% BSA. In some embodiments, the buffer comprises 20 mM HEPES, 50 mM NaCl, 50 mM MgCl2, 0.01% polysorbate 20, 2.5 mg / ml heparin, and 0.05% BSA. In some embodiments, the buffer further comprises an RNase inhibitor. [Example]

[0150] 7. Working Example 7.1. Example 1: Development of target-decorated cell lines and constructs for target display Six different constructs (V1–V6) were tested for the generation of target-decorated cells. The constructs had different leader sequences (mouse IgM, native CoV1-S or CoV2-S, or tPA) and transmembrane domains (cmyc-PDGFR, native CoV1-S or CoV2-S, or native sequences with 19 or 37 amino acid C-terminal truncations), as specified in Table 1. [Table 1]

[0151] The constructs were transiently transfected into expi293 cells, and transgene expression was assessed 24 hours later by flow cytometry using a mAb against the spike protein (Figure 2). This study showed that constructs with native leaders performed well, particularly those with native leaders and a 19-amino acid truncated transmembrane domain (nativeΔ19).

[0152] The constructs were further modified to generate three stable cell lines expressing the CoV2-S antigen. First, the V1 construct was packaged into a lentivirus and used to generate stable cell lines in Expi293 cells. Cells transfected with V1-lentivirus were isolated by positive c-myc signal on FACS. (Expi293-CoV2-S) The second and third constructs were modified to be compatible with CHOZN cells. In particular, the second construct was modified to contain V5 components, a native leader, and FRT sites with the P2A and glutamine synthetase genes (GS), allowing selection of stable integrants via glutamine-depleted medium. This construct also utilizes the 2G-UNIC translation-enhancing element in conjunction with the CMV promoter. A third construct (CHOZN-CoV2-S) was generated containing the same components as the second construct, plus six stabilizing proline mutations to prevent cleavage (Hsieh et al., Science 369, 2020) and a furin site mutation (Peacock et al., Nature Microbiology 6, 2021). (CHOZN-CoV2-S(HP+F)) The second and third constructs were transfected into CHOZN cells containing a single landing pad at a locus shown to have high and consistent gene expression, generating stable cell lines. In this step, a recombinase was used to insert a single copy of the construct containing the FRT site into the landing pad.

[0153] All three cell lines and corresponding negative controls were stained with different concentrations of bamlamivimab (Jones et al., BioRxiv 9OCT20), a human IgG1 with high affinity for CoV2-S. As shown in Figure 3, CHOZN cells (CHOZN+CoV2-S and CHOZN+CoV2-S(HP+F)) showed significantly better surface expression of CoV2-S than Expi293 cells (Expi293+CoV2-S). Furthermore, stabilizing mutations (HexaPro, HP) and furin mutations (F) increased surface display.

[0154] Based on this study, a construct backbone (p2G-FRT-GS) was selected to generate target-decorated cells. As shown in Figure 4, the construct backbone contains (i) a CMV promoter with a 2G translation-enhancing element, (ii) a blue fluorescent protein stuffer (to identify library background during cloning), (iii) a 20-nucleotide unique barcode (e.g., HHSWNNHHCTGGNNHHSWHH (SEQ ID NO: 114), NNNWSHHHNNHHHNNWSNNN (SEQ ID NO: 115), or similar), (iv) an FRT site with 2A-GS (to select in-frame integrants into the landing pad line), and (v) a bacterial origin of ampicillin resistance and for propagation in E. coli.

[0155] Target libraries can be introduced into the construct backbone in several ways. For small libraries, each insert can be synthesized with a unique barcode and cloned by Gibson assembly or insertion into the HindIII / NheI site. For larger libraries, a backbone library with unique barcodes must first be generated by inserting a library of barcodes, possibly with gene blocks or oligos with degenerate nucleotides, into the MluI / NheI site. For scanning mutagenesis, a library of single point mutants can be synthesized and then cloned into the library barcode backbone via HindIII / MluI or Gibson assembly. For other mutagenesis methods (e.g., random mutagenesis), a library of inserts can be generated using, for example, error-prone PCR or OE-PCR with primers with degenerate nucleotides, and then cloned into the barcoded backbone using HindIII / MluI. Note that the barcode library needs to be at least 10x larger than the insert library to ensure a unique barcode per insert.

[0156] 7.2. Example 2: Target decorated cell lines expressing CoV2 antigens All available CoV2 sequences as of May 2022 were downloaded from GISAID and analyzed. Analysis revealed 25 unique clades with 20 unique spike protein sequences. This list was further condensed to remove those with only a single amino acid difference, resulting in 17 spike sequences. These 17 were synthesized along with the CoV1 spike and cloned into a 2G-FRT-GS backbone with a unique barcode. The spike sequences contained hexapro and furin mutations to improve surface display and no leader sequence mutations.

[0157] DNA from all library members was pooled in equimolar ratios and transfected into the platform CHOZN strain, followed by selection in glutamine-deficient medium for several weeks. Alternatively, clones were transfected individually, selected, and then pooled. RNA from clones was isolated and sequenced to confirm library distribution.

[0158] 7.3. Example 3: TBP-Ribosome-mRNA Complex (TRM Complex) The five mAbs were cloned into T7-based expression plasmids with the structures provided in Figure 5. The T7-based expression plasmids contain (i) a T7 promoter and ribosome binding site, (ii) an origin of replication for ampicillin resistance and propagation in E. coli, (iii) a start codon followed by the variable light region (without the signal peptide), (iv) a flexible linker, (G4S)4 or similar, (v) the variable heavy region, (vi) a Strep tag, (vii) an unstructured spacer sequence derived, for example, from the E. coli protein tolA or other sequences of various lengths, (viii) a translation pause sequence derived from the E. coli protein secM, (ix) the absence of a stop codon, (x) an internally encoded poly(A) tract of 40 × A nucleotides, and (xi) a T7 transcription terminator (possibly several tandems or variants).

[0159] The entire plasmid, linearized plasmid, or a PCR fragment of the plasmid (from promoter to terminator) was used as a template for transcription or coupled transcription-translation. The NEB PURExpress kit, the NEB T7 Quick High Yield RNA kit, and the Promega TnT Rabbit Reticulocyte kit, with or without release factors, were also used in the reactions.

[0160] 7.4. Example 4: Binding between Target and TBP-Ribosome-mRNA Complex (TRM Complex) Cells expressing spike variants (CoV2-WT, CoV2-K444T, CoV2-E484K, CoV2-F486K, and CoV1-WT) were stained with different dilutions of TRM complexes, each containing the scFv of one of four mAbs (bamlanivimab, casirivimab, imdevimab, or ipilimumab). TRMs bound to CoV2-S variants were detected by strep-tag staining, and the results are presented in Figure 6.

[0161] In another set of experiments, cells expressing spike variants were stained with different dilutions of the TRM complex, and RNA was isolated from the stained cells and quantified using TaqMan RT-qPCR with an antibody (TBP)-specific probe. The results are provided in Figure 7. RNA was recovered from the stained cells as expected, and binding between TRM and CoV2-S was observed. CoV variants previously known not to bind to a given antibody did not show binding to the antibody (red line), as provided in Figure 6.

[0162] In subsequent experiments, we tested TRM complexes produced under different cell-free translation (TL) conditions. Specifically, TRM complexes were produced using different concentrations of RNA template at different TL reaction times. The data presented in Figure 8 show that the TRM complex produced using 150 nM RNA template at a short reaction time (0.5 h) provided significantly reduced nonspecific background and the greatest ratio between RNA detected by the TRM complex containing casirivimab scFv (i.e., specific staining) and RNA detected by the TRM complex containing ipilimumab scFv (i.e., nonspecific staining).

[0163] 7.5. Example 5: Binding between Target and Pooled TBP-Ribosome-mRNA Complexes (TRM Complexes) CoV2-S antigen-expressing cell lines were individually stained with a mixture of TRM complexes. The TRM complex mixture was generated by in vitro translation (0.5 h reaction time) from a pool of RNA templates (150 nM template). RNA was isolated from cells stained with the TRM complex mixture and quantified using TaqMan RT-qPCR with an antibody (TBP)-specific probe. The results, presented in Figure 9, demonstrate specific binding between the target and TRM complex pair, low background staining, and no signal for variants known not to bind to the given antibody (shown in red text). Specifically, bamlanivimab ("Bam") lost binding to the E484K mutant, casirivimab ("Casi") lost binding to the F486K mutant, and imdevimab ("Imdevi") lost binding to the K444T mutant. Ipilimumab ("Ipi") is not predicted to bind to any of the spike variants (it binds to CTLA4), and neither of the antibodies is predicted to bind to the CoV1 spike. These known binding specificities were confirmed in paired target and TRM complexes, as shown in Figure 9.

[0164] 7.6. Example 6: Single Cell Isolation and RNA Amplification Five target-decorated cell lines were pooled and stained with a mixture of five TRM complexes. Single cells were sorted and isolated into 96-well plates. Samples in individual wells were amplified and barcoded for sequencing. The first RT-PCR reaction added adapters to the mRNA transcripts, identifying the V of the antibody mRNA in the TRM complex. H Both the region spanning the target barcode from the cell line and the region spanning the target barcode from the cell line were amplified. A second PCR reaction was added to the amplicons indexing the unique barcodes in the rows and columns of a 96-well plate. Samples from all wells were pooled and sequenced using mi-seq.

[0165] Single-cell data were analyzed for binding distribution. First, the targets of individual cells within each well were determined. A threshold was applied to exclude empty wells and wells with more than one cell. For each well, the percentage of reads corresponding to each antibody was calculated. Polymap scores were calculated from the average percentage scaled by the average reads from cells for that target. Polymap scores were determined for CoV1 or CoV2 spike WT or variants (CoV2-K444T, CoV2-E484K, CoV2-F486K) and provided for TRM complexes containing scFvs of one of four mAbs (bamianivimab, casirivimab, imdevimab, ipilimumab, or pembrolizumab). Polymap scores were significantly lower for cell lines with targets not expected to bind to a given TRM complex (shown in red outline). Negative antibodies (ipilimumab or pembrolizumab (red text)) were present in low proportions across all targets.

[0166] 7.7. Example 7: Single Cell Analysis Using a Microfluidic System Five target-decorated cell lines were pooled and stained with a mixture of five TRM complexes. Once stained with the TRM complexes, single cells were encapsulated in water-in-oil droplets with cell lysis buffer and RNA capture beads for antibody repertoire analysis. After the beads were released from the emulsion and washed to remove unbound RNA, the RNA-bound beads were isolated from the single cells. Because each cell expresses a single target (i.e., a SARS-CoV-2 spike variant), each bead was expected to bind the RNA of a single target as well as RNA from each TRM complex bound to that particular cell.

[0167] The single RNA-bound beads were then encapsulated in a second emulsion for antibody repertoire analysis. These second emulsions also contained PCR buffer, reagents, and primers for overlap-extension reverse transcription PCR (OE-RT-PCR). In the reaction mixture, the RNA was converted to cDNA, which was then probed with the target barcode and antibody V. H The fragments were amplified with primers that allowed the CDR3 to be physically linked to each other. The reverse primer for the antibody fragment contained a sequence complementary to the sequence on the forward primer for the target fragment. During subsequent rounds of PCR, these complementary fragments annealed, and the resulting product contained both the antibody sequence and the target sequence. Figure 11 shows the results of overlap extension reverse transcription PCR and the amplification of the antibody (V). H )-specific sequence and target (barcode)-specific sequence.

[0168] Following OE-RT-PCR, the fused TBP target sequence was further amplified with primers containing Illumina adapters and indexes for standard Illumina sequencing workflow. In our experiments, we successfully amplified and sequenced the following CDR3 and barcode sequences linked to the CoV2-S antigen: [Table 2]

[0169] 7.8. Example 8: High-Throughput Specificity Profiling of Antibody Clones Using Ribosome Display and Microfluidics High-throughput methods for pairwise interaction mapping of protein-protein interactions enable a deeper understanding of biological interactions, such as natural antibody responses. However, current technologies are limited by the number of interactions that can be screened at one time. To overcome this, a polyclonal mapping system (also referred to herein as "PolyMap") was developed. This system is based on bulk binding of a ribosome-displayed antibody library to a library of cell surface-expressed antigens, combined with single-cell analysis using droplet microfluidics. In this example, mapping of thousands of antigen-antibody interactions was demonstrated for a set of clinically relevant SARS-CoV-2 surface antigens against a diverse library of naturally occurring anti-SARS-CoV-2 antibodies, and antibodies with selective binding to a wide range of antigen clades were found. Although developed using an antibody-antigen library, PolyMap is well suited to screening other types of protein-protein interactions, including soluble ligands and their membrane-bound targets, such as cytokines and immune regulators and their receptors.

[0170] Designing a PolyMap workflow The PolyMap platform enables one-pot interaction screening of antibody and antigen libraries (Figure 12). Antibodies are expressed in a ribosome display format, using tethered mRNA to provide genotype-phenotype linkage of soluble proteins. Antigens are expressed at high levels on the surface of mammalian cells and can then be bound by soluble ARM complexes. Incubation of a library of ribosome-displayed antibodies with a library of antigen-expressing cells allows sampling of all possible antibody-antigen combinations in a single bulk step. To elucidate antibody-antigen interactions, cells stained with ARM complexes are washed and individually encapsulated in microdroplets containing lysis reagents and uniquely barcoded RNA capture beads. A series of molecular biology steps are used on the isolated beads to generate barcoded cDNAs of the linked antibody and antigen sequences. The barcoded transcripts are then read, quantified by deep sequencing, and analyzed using bioinformatics to map antibody-antigen pairwise binding specificities.

[0171] Establishment of antigen expression system This section concerns the establishment of a robust antigen surface display platform. In particular, the platform should possess robust surface expression of a single antigen per cell, a unique barcode for antigen identification, and a simple and efficient cloning and cell line generation process suitable for use with libraries. A series of mammalian expression vectors were evaluated to determine the effect of different signal peptides and transmembrane regions on surface expression of selected target SARS-CoV-2 surface antigens (CoV-2 S; data not shown). The native signal peptide and transmembrane region were retained because they performed at least as well as alternatives. A 19-amino acid C-terminal truncation was performed to remove the endoplasmic reticulum retention signal inherent in CoV-2 S.

[0172] Stable cell lines expressing CoV-2 S were generated by two different methods: (1) lentiviral transduction into Expi293™ cells or (2) flippase (Flp) recombinase-mediated integration into CHOZN® cells engineered with an Flp recognition target (FRT) landing pad for single-site integration. After sorting or selection of the cell lines, respectively, staining for CoV-2 S surface expression revealed a much higher relative signal from the CHOZN cell line (Figure 13). A modified construct containing six stabilizing proline mutations and a furin site mutation to abolish the fusogenic activity of the spike protein was also evaluated and found to further increase cell surface expression of the antigen (Figure 13).

[0173] The final antigen display vector uses a CMV promoter with translation-enhancing sequence elements (2G UNic®, ProteoNic) to drive efficient antigen expression, a signal peptide, a transmembrane domain for surface display, the previously described proline modifications, and a mutated furin site (Figure 14A). A 20-mer nucleic acid barcode unique to each antigen is included after a translation stop codon, followed by common flanking sequences, allowing antigen identification by PCR amplification and short-read sequencing. Restriction enzyme sites are included for linearization of the vector and easy ligation of the antigen library.

[0174] To generate a library of antigen-displaying cells, the antigen-encoding plasmid library was transfected, along with a plasmid expressing Flp recombinase, into a CHOZN cell line previously engineered to contain an FRT site integrated into a highly expressed genomic locus. This ensures single-copy integration and normalized expression levels across individual antigens. Successful integration of the plasmid results in expression of glutamine synthetase, allowing selection in glutamine-free medium.

[0175] For proof-of-concept studies, we selected the ancestral Wuhan (referred to as wild-type in this study; WT) CoV-2 spike protein and three individual point mutants (K444T, E484K, and F486K) known to significantly disrupt binding of specific monoclonal antibodies in clinical development. Four spike protein coding sequences, including the native leader sequence, stabilizing mutations, native transmembrane region, and deleted cytoplasmic tail, were cloned into platform vectors with unique barcodes. In this case, the plasmids were separately transfected into platform CHOZN cells, which were then selected and stained for spike surface expression using cross-reactive antibodies. Flow cytometry analysis demonstrated high and consistent expression across the four CoV-2-S variants (Figure 14B).

[0176] Establishment of a ribosome display system The ribosome display construct contains basic features previously reported by others, with some more recent enhancements and features unique to this application (Figure 14C). A T7 promoter drives expression of the scFv gene, followed by a strep tag II sequence. The absence of a stop codon and the addition of a 17-amino acid SecM stall sequence prevent the ribosome from dissociating from the mRNA and physically tether the newly synthesized protein to its coding sequence. The encoded polyA tail allows for cDNA synthesis, barcoding, and subsequent capture onto beads for sequencing. The CDR3H sequence provides a unique identifier for the antibody, and for the antigen plasmid, a common cleavage site and flanking regions are included for library cloning, amplification, and sequencing. The entire cassette is maintained within a pUC-based plasmid.

[0177] To generate ARM complexes, DNA fragments extending from the T7 promoter to the transcription terminator region were amplified by PCR and used as templates to generate large amounts of RNA. While it is possible to use circular plasmids as transcription templates, even constructs containing improved terminator designs were found to have significant terminator readthrough, so PCR fragments with defined ends were found to be important for RNA uniformity (data not shown). Using an in vitro expression system based on recombinant components and starting from a fixed input of purified RNA was determined to result in more consistent ARM complex yields than a coupled transcription-translation process, likely due to variations in transcription kinetics between variants. It was also determined that omitting ribosome release factors from the in vitro translation reaction and using a short 10-minute translation time further improved the stability and specificity of the ARM complex (Figure 15).

[0178] For initial testing, we selected the clinical SARS-CoV-2 monoclonal antibodies bamlanivimab, casirivimab, and imdevimab, which have known susceptibility to the spike point mutations E484K, F486K, and K444T, respectively. Biolayer interferometry (BLI) was used to confirm the binding affinity of these antibodies, presented as scFvs, to each spike variant expressed as a soluble receptor-binding domain (RBD)-Fc fusion (and lack of binding to specific point mutations) (Table 3, Figure 16). The scFvs were then cloned into ribosome display constructs, produced as ARM complexes, and individually incubated with antigen-expressing cells. Total RNA was isolated from stained cells and then analyzed by RT-qPCR to quantify antibody mRNA recovery. For each antibody, there was complete or partial loss of RNA to the expected spike variant, consistent with the measured binding affinity (Figure 14D). Although there was some variation in RNA recovery between different antibodies, likely due to differences in translation efficiency, the concentrations reported here represent RNA recovery of 1–4%, which is consistent with previous reports. [Table 3] Measured binding affinities of clinical antibodies to soluble spike variants by BLI. * is the off-rate below the detection limit.

[0179] Proof-of-concept study using a clinical-stage anti-SARS-CoV-2 antibody Preliminary studies were performed using "mini-libraries" of three clinical-stage anti-SARS-CoV-2 antibodies and two negative control antibodies (anti-CTLA-4 ipilimumab and anti-PD-1 pembrolizumab). In the simplest experiment, an equimolar RNA mixture was used to generate ARM complexes, which were then diluted to a concentration equivalent to 15 nM RNA and incubated separately with each antigen-expressing cell line. Total RNA from the stained cells was amplified and sequenced, and each antibody was identified using the CDR3H sequence. The percentage of total reads assigned to each antibody was tabulated for each antigen-expressing cell line. As expected, there were very few reads for the expected dropouts, and reads for both negative control antibodies across all samples were low (Figure 17A). Surprisingly, an outsized proportion of reads were derived from casirivimab, which is consistent with the higher RNA recovery from initial RT-qPCR results for this antibody (Figure 14D).

[0180] This experiment was repeated by mixing the four antigen cell lines together and bulk staining with a mini-antibody library. To pair antigen and antibody data, single cells were sorted into 96-well plates and lysed. Antigen barcodes and antibody heavy chains were amplified with arrayed primers containing column and row barcodes to enable unique identification of amplicons from each individual well. After identifying the antigens on each cell, the distribution of antibodies bound to that cell was calculated. As previously seen when staining individual cell lines, we observed strong predicted CoV-2 variant shedding, low background binding, and reproducibility between individual cells (Figures 17B and 17C).

[0181] While single-cell sorting and arrayed PCR are feasible for small sample sets, single-cell droplet-based approaches combined with high-throughput sequencing enable the analysis of larger sets of antibodies and / or antigens. For proof-of-concept studies, we utilized Drop-seq, a previously published microfluidic droplet technology that co-encapsulates single cells with barcoded beads (ChemGenes Corporation) in nanoliter droplets using a microfluidic chip and syringe pump setup. We used the same antibody-RNA mixture and four cell lines, but with a slightly modified staining strategy in which the ARM complex was diluted to a concentration equivalent to 4.5 nM RNA (this concentration is not expected to affect the results, as testing of the ARM complex binding to CoV-2 WT at two different concentrations showed similar levels of enrichment for many antibody clones; Figure 18). The Drop-seq workflow, including the cDNA amplification step, was performed as previously described (https: / / mccarrolllab.org / dropseq / ), and antigen barcodes and antibody heavy chain sequences were amplified using gene-specific primers with Illumina adapters for sequencing. Individual cells were first identified by Drop-Seq barcodes, and then antigen identity was determined by antigen barcodes. 91 cells were identified for antibody-antigen interaction mapping (Figures 17D, 17E). Drop-seq results were consistent with both individual cell line staining (Figure 17A) and single-cell sorting (Figures 17B, 17C).

[0182] Testing the anti-SARS-CoV-2 immune repertoire against a larger set of spike variants Next, PolyMap was applied to a larger library of antibodies and antigens. All publicly available CoV-2 S sequences (as of August 2023) were downloaded from GISAID.org and 16 unique sequences (nextstrain.org) representing phylogenetically distinct clades were selected (Figure 19, SEQ ID NOs: 126-138, 140, 142, and 143). These sequences and control antigens (CoV-1 S, CTLA-4, PD-1, and blue fluorescent protein [BFP]) were cloned into an antigen scaffold with unique barcodes and then individually transfected into a CHOZN landing pad line. Next, a diverse library of naturally paired scFvs was cloned into a ribosome display vector. This library was previously isolated from a convalescent COVID-19 donor and enriched for binders to the ancestral spike RBD protein. Sequencing of this library showed a broad variant distribution with 99% of the sequences derived from 300 unique clones, with the top clones constituting approximately 6.9% of the library (Figure 20A).

[0183] For Drop-seq microfluidic analysis of larger libraries, cells expressing each CoV-2 S sequence variant were pooled together and then stained with the ARM complex of the anti-CoV-2 antibody library described above. After identifying antigen and antibody sequences using Drop-seq cell barcodes, we generated a merged list of all antibodies and antigens associated with each cell barcode, ultimately analyzing 2,783 cells across 20 antigens (Figure 21). Because each antibody has an equal opportunity to bind to each antigen, we counted the total antibody reads for each clone across all cell lines, and then normalized the percentage of reads for each antibody by the number of associated cells. When examining the top 40 antibodies based on total number of antibody reads, we observed distinct binding patterns across antibody sequences (Figure 20B). Many of the antibody clones showed minimal enrichment for omicron spike variants (21K, 22A / D, 23B), which was expected since this library was generated in early 2020 from a donor convalescing from the original ancestral WT strain. Reduced binding to 20H (beta), 20J (gamma), CoV-1 and negative control cell lines (PD-1, CTLA-4 and BFP) was also observed.

[0184] In this study, when generating CoV-2 antigen cell lines, antigen sequences were transfected individually into cells and then pooled for PolyMap runs. Alternatively, multiple antigen sequences can be transfected into cells in bulk. To examine the feasibility of this approach, 22 antigen sequences were transfected together, including 18 unique CoV-2 S sequences representing phylogenetically distinct clades (16 sequences previously used and shown in Figure 19, as well as 21M omicron (SEQ ID NO: 139) and 22C omicron (SEQ ID NO: 141)) and four control antigens (CoV-1 S, CTLA-4, PD-1, and BFP). These antigen sequences were cloned into an antigen backbone with unique barcodes and then transfected en bloc into a CHOZN® landing pad line. Sequencing of genomic DNA (gDNA) from the bulk-transfected strains showed the representation of all variants with only moderate variation in frequency (Figure 24). This method of bulk transfection allows for the efficient generation and testing of large antigen libraries.

[0185] Validating PolyMap results Using individual cell lines for each antigen available, each line was stained separately with the ARM complex library. RNA was isolated and sequenced as before, and the percentage of reads from each antibody was calculated. Due to the skewed distribution of the antibody library, it was important to examine enrichment relative to the input sample rather than raw antibody read percentages. The log2 fold change in this enrichment value was used to better distinguish between enriched and de-enriched sequences, and the top 40 most enriched clones are shown (Figure 22A). Again, many clones were observed that were de-enriched due to binding to 20H (beta), 20J (gamma), and omicron variants.

[0186] Eleven antibody clones (termed M1-M11) found within the top 100 sequences and with diverse antigen-binding patterns were selected for validation. Some had previously been isolated as individual CHOZN cell lines, for which small-scale production runs were performed. The remaining antibodies were validated using a second sequencing run on the input library. L The sequence and CDR3H were determined, and then this was compared with the previous V H The data were combined to assemble full-length sequences for expression. These variable regions were cloned into separate expression vectors and transiently expressed in HEK293 cells.

[0187] The binding of these antibodies was verified by staining each spike variant cell line with antibody-containing supernatant diluted to a concentration expected to be in the linear sensitivity range. Antibody binding was detected with a secondary antibody specific to human Fc and then analyzed by flow cytometry. The median fluorescence intensity (MFI) of the data was compared with both Drop-seq (Figure 20B) and single-antigen staining (Figure 22A) PolyMap data. In general, the characteristic binding patterns of each clone derived from Drop-seq were closely matched by the single-antigen and flow staining data (Figure 22B). One interesting note is that since single-antigen staining is normalized within a single cell line, enrichment values ​​may provide some insight into the relative binding affinities between antibodies (Figure 23). For example, clone M1 shows good alignment across all datasets (Figure 22B), but may have lower affinity due to lower enrichment in the competitive binding scenario of the WT antigen. This may also explain some false positives from single-antigen staining of the Omicron variant, as there were fewer competing sequences. Clone M10 may be an example of a sequence with very high translation or PCR efficiency that resulted in strong enrichment using a single antigen staining approach.

[0188] Finally, we further investigated the sequence diversity of the recovered antibodies and their binding patterns. As expected, few clones retained binding to the highly mutated omicron spike in these early patient samples. Several recurrent shedding patterns were observed, which correlated with shared mutations in the RBD. For example, the K417 mutation was found in beta, gamma, and omicron (21K) and later (Figure 19), and is a potential epitope for M1, M2, M3, M6, M7, M9, and M10. Variants beta, gamma, kappa, eta, iota, and mu later contain the E484 mutation, which is a potential epitope for clones such as M5. Clones such as M8 and M11 showed binding restricted to only the earliest clones and may recognize epitopes containing the above amino acids and other amino acids such as L452 or N501. Finally, clone M4 showed broad binding and some reactivity to omicron, suggesting it may bind to a more conserved epitope.

[0189] Consideration A solution for pairwise interaction mapping between antigen and antibody libraries was sought without some of the limitations of existing technologies. To this end, PolyMap, a platform for probing libraries of mammalian cell surface-expressed proteins with soluble antibody libraries, was developed. This approach was developed and validated using a small matrix of clinical SARS-CoV-2 antibodies with known spike mutation susceptibility. PolyMap was then applied to an uncharacterized immune library from a convalescent COVID-19 patient, discovering antibodies with several distinct binding patterns across circulating SARS-CoV-2 spike variants.

[0190] A key strength of the PolyMap platform is its flexible antigen expression system, tailored for membrane protein expression. Unlike other published multiplex mapping methods, which require chimeras for purification, in vitro translation, or yeast display, PolyMap uses full-length antigens displayed on the surface of mammalian cells. Robust quality control mechanisms in the cell ensure the native structure and glycosylation of complex membrane-bound proteins, which are important for many therapeutic targets, such as GPCRs or oligomeric viral antigens. A key component of the antigen expression system is the parent CHOZN cell line, which features engineered FRT sites in highly expressed loci for recombinase-mediated integration. This platform strain has been widely used for one-pot transfection and the production of thousands of diverse antibody libraries. Other methods for generating large antigen or antibody libraries in mammalian cells are also compatible with PolyMap, but uniform clonal distribution and even expression levels are beneficial.

[0191] The antibody library was previously validated and expressed as scFv in a ribosome display format, which possesses several advantageous features. First, the soluble antibody library enabled the staining of suspension cells in bulk, a feasibility previously described only for adherent cells. Second, although the ARM complex could potentially contain several translating ribosomes, a short 19-amino acid C-terminal linker was used to promote the binding of a single full-length antibody to each ribosome. This functionally monovalent format was important for preventing cell aggregation and allowed the stained cells to be processed through a microfluidic system. Monovalent binding was also crucial for eliminating avidity effects and allowing small changes in binding affinity to be distinguished. Other advantages of ribosome display include the ease of library generation, compatibility with very large library sizes, and commercially available reagents.

[0192] The PolyMap screening process is simple and scalable, allowing both antigen and antibody libraries to be generated once and reused indefinitely. Unique to PolyMap, each cell supports hundreds to thousands of interactions, meaning that available screening capacity is largely defined by the number of antigen-expressing cells that can be individually processed. When only a few antigen variants are to be assayed, individual cell lines can be generated, stained, and directly subjected to CDR3H sequencing without single-cell manipulation (Figures 17A and 22A). For studies involving tens, hundreds, or more antigens, bulk transfection of plasmid libraries into cells is convenient and presents an option not available with some alternative technologies. Drop-seq, used here, can typically process up to 10,000 single cells per hour, making it suitable for studies involving tens to potentially hundreds of unique antigens. Other single-cell and microfluidic sequencing technologies are also compatible with PolyMap and may be considered in future iterations.

[0193] While the data processing method is relatively straightforward, there are some caveats that are important to understand when interpreting the data. For Drop-seq analysis, we assessed the read distribution of a single antibody across all antigens. With this approach, it is easy to see the binding patterns within each antibody, but it is more difficult to understand how different antibodies may bind to each other. In the single-antigen staining approach, reads are normalized within each antigen, which can allow for unambiguous identification of the top binders for each variant. The consideration here is that if the antibody distribution is skewed, normalizing to the input distribution is important; as the total number of antibodies binding to a particular antigen decreases, the proportion of reads attributable to the remaining antibodies will appear higher. In this case, rare binders to an antigen may not be of particularly high affinity, but may be good candidates for further manipulation.

[0194] When PolyMap was applied to an enriched antibody library derived from a convalescent COVID donor and a set of known spike variants, several notable findings were obtained. Several distinctive clonal binding patterns were observed, which were shared across several antibodies (some of these interaction patterns were confirmed using full-length antibodies). Although confirmation experiments were not performed, examination of the spike amino acid sequence (Figure 19) revealed distinct sequence variants associated with different binding patterns, potentially indicating important amino acids in the antibody epitope. Interestingly, despite the antibodies being derived from donor samples collected in April 2020, strong binding was found to all subsequent SARS-CoV-2 variants, including lambda, which did not become a WHO variant of interest until June 2021. Several monoclonals were found with weak binding to the highly mutated omicron variant, which may target conserved epitopes of potential value for future therapeutic development. Finally, since one strategy to provide broad coverage is to target several non-overlapping epitopes, PolyMap data can help select antibodies with possible functional synergies.

[0195] Careful selection of antibody and antigen libraries may enable broad applications of the PolyMap platform. Pairing PolyMap with antibody repertoire capture technology may enable profiling of immune responses across donors, across treatments, or over time. Libraries of human surface proteins or other target protein populations can be amplified from cDNA and generated in bulk for antibody specificity profiling. Both the antigen and antibody platforms are compatible with any library generation technology, allowing for the generation of synthetic libraries for protein engineering studies. For example, deep mutational scanning libraries of antigens can be generated and screened with antibody libraries in massively parallel epitope mapping platforms. Targeted CDR mutagenesis and panels of antigen-expressing cells can be used to evolve antibodies with broad or highly targeted specificities. An interesting extension of this method could include engineered cells with a ligand-activatable phenotype. In this case, ARM-stained and activated cells can be sorted based on marker activation and subsequently subjected to PolyMap for identification of specific binding partners. Furthermore, in principle, the technology can be adapted to any library of soluble proteins and their cognate binding partners expressed on the cell surface, such as cytokines and their cell-associated receptors, or immune checkpoint proteins.

[0196] method Antigen constructs and cloning The amino acid sequence of the wild-type (WT) SARS-CoV-2 S protein (UniProt P0DTC2) was modified to improve expression: six stabilizing proline mutations were added, the furin site was removed (RRAR (SEQ ID NO: 120) → AGAG (SEQ ID NO: 121)), and 19 amino acids were truncated from the C-terminus. The sequence was codon-optimized, synthesized, and cloned into an antigen expression vector (Figure 14A) along with an 18-mer barcode sequence. The expression vector contains a CMV promoter with a 2G-UNic® translation-enhancing element (ProteoNic) and an FRT site followed by a 2A ribosomal skipping motif and a glutamine synthetase gene. Point mutations were introduced into the WT spike sequence by Gibson assembly and cloned into the same expression vector backbone along with the unique barcode.

[0197] To generate the circulating spike variant library, protein sequences of all available spike variants were retrieved from GISAID.org as of August 2023. A phylogenetic analysis of 4,217 SARS-CoV-2 genomes sampled between December 2019 and July 2023 from nextstrain.org was also performed. The most common sequences within each phylogenetic clade were selected and filtered for unique sequences. Sixteen representative spike variants covering a diverse range of mutations were selected (Figure 19). These sequences, including proline and furin mutations and C-terminal truncations, were synthesized with unique barcodes (HHSWNNHHCTGGNNHHSWHH, SEQ ID NO: 114) and cloned into antigen expression vectors. Wild-type sequences of CoV-1 S (P59594), human CTLA-4 (P16410), human PD-1 (Q15116), and mTagBFP2 were included as controls.

[0198] Generation and analysis of stable cell lines Stable cell lines were generated by transfecting each individual antigen construct along with a plasmid encoding Flp recombinase into CHOZN® cells containing compatible FRT sites and landing pads. Cells were cultured at 10°C for 10 min using the default CHO protocol. 6 Cells were electroporated with 5 μg of a 4:1 mixture of antigen and recombinase plasmid (MaxCyte). After 4 days of recovery in EX-CELL® CD CHO Fusion medium (MilliporeSigma, 14365C) ​​supplemented with GlutaMax (ThermoFisher, 35050061), cells were selected in the absence of GlutaMax for 18 days or until viability reached >95%. After selection, gDNA was collected and the barcode region was amplified using flanking primers containing sequencing adapters (Antigen BC P5 FWD, Antigen BC P7 REV; primer sequences are listed in Table 4). DNA samples were quantified using qPCR and normalized to 1.8 nM. Libraries were diluted to a final concentration of 9 pM, supplemented with 5% PhiX DNA, and sequenced on an Illumina MiSeq instrument using custom primers (M13 SEQ, BGH INDEX SEQ, BGH SEQ). Sequencing results were analyzed using a Python script that calculated the number of times each unique antigen barcode appeared per library and its frequency relative to the total number of barcodes.

[0199] Flow cytometry To assess spike protein expression, cells were stained with 10 μg / mL of an anti-COVID-19 & SARS-CoV S glycoprotein antibody (clone CR3022, Absolute Antibody, Ab01680-10.0), followed by PE or APC anti-human IgG (H+L) antibody (Jackson ImmunoResearch, 709-606-149). To validate binding of specific antibodies identified by PolyMap, spike variant-expressing cell lines were stained with supernatant from CHO cells expressing the indicated antibody clones at 0.5 μg / mL, followed by APC anti-human IgG Fc antibody (BioLegend, 409306). After secondary antibody staining, cells were stained with 4',6-diamidino-2-phenylindole (DAPI; BioLegend, 422801). Samples were analyzed on a CytoFLEX LX flow cytometer (Beckman Coulter), and the mean fluorescence intensity (MFI) of the APC signal on live cells (DAPI negative) was determined with FlowJo (v10.6.1, BD Biosciences).

[0200] Antibody construction and cloning Antibody scFv sequence (V L -Linker-V H ) was assembled, codon-optimized, and synthesized into a ribosome display vector using overlap for Gibson assembly. Instead of a stop codon, the vector encodes a strep tag II, a tolA-derived unstructured sequence (GGQKQAEE, SEQ ID NO: 122), and a secM stall sequence (FSTPVWISQAQGIRAGP, SEQ ID NO: 123). Transcription from the plasmid is controlled by flanking conventional T7 promoter and terminator elements.

[0201] A primer pool designed to capture all germline sequences (scFv library FWD 1-5, REV 1-3; Table 4) was used to amplify an antibody library from a naturally pairing antibody repertoire previously generated from a convalescent COVID patient. These sequences were bulk cloned into a linearized ribosome display plasmid using Gibson assembly. For sequencing, V sequences were generated directly from the plasmid library or from cDNA using universal primers that bind to the linker and secM regions (scFv linker1 P5 FWD, scFv P7 REV). H Samples were prepared by amplifying the region.

[0202] ARM complex generation and cell staining The ribosome display plasmid was subjected to PCR to generate a fragment spanning the T7 promoter, scFv cassette, and T7 terminator sequences (primers ARM complex FWD, ARM complex REV; Table 4). The PCR product served as a template in an in vitro transcription reaction (HiScribe® T7 Quick High Yield RNA Synthesis Kit, NEB, E2050S) to generate RNA, which was subsequently DNase treated and purified. To generate the ARM complex, 150–450 nM of RNA template (based on 1,100 bp length) was added to a release factor-free in vitro transcription-translation kit (PureExpress® ΔRF123, NEB, E6850S) and incubated at 37 °C for 10 min. The reaction mixture was immediately chilled on ice and diluted 1:5 or greater with RBTH buffer (20 mM HEPES, 50 mM MgCl, 50 mM NaCl, pH 7.4, 0.05% BSA, 2.5 mg / mL heparin, 0.01% Tween®-20). 6 Antigen-expressing cells were stained with 50 μL of diluted ARM complex for 30 min on ice and then washed three times with 1 mL of ice-cold RB (20 mM HEPES, 50 mM MgCl, 50 mM NaCl, pH 7.4 + 0.05% BSA) before further analysis.

[0203] RNA isolation and qPCR For qPCR analysis of ARM complex binding after staining of antigen-expressing cells (Figure 14D), the lysis and stop solutions from the Cells-to-CT™ 1-Step Power SYBR™ Green Kit (ThermoFisher, A25599) were used. In a 96-well plate, 10,000 stained cells per well were lysed in 20 μL of lysis solution for 30 minutes at room temperature, followed by the addition of 2 μL of stop solution and incubation at room temperature for 2 minutes. The lysates were probed for secM in an RT-qPCR assay (using primers ARM qPCR FWD, ARM qPCR REV, and ARM qPCR probe; Table 4). The TaqMan™ RNA-to-CT™ 1-Step Kit (ThermoFisher, 4392938) was used according to the manufacturer's instructions. 2 μL of each lysate was distributed into 30 μL reaction mixtures (15-fold dilution). Each reaction was aliquoted into quadruplicates at 6 μL per well in a 384-well plate. A 7-point standard curve using known concentrations of RNA was included, starting from 1 nM to 1 fM (10-fold serial dilutions).

[0204] Single antigen staining ARM complexes were generated as described above using an equimolar mixture of ribosome-displayed RNA for bamlanivimab, casirivimab, imdevimab, ipilimumab, and pembrolizumab, then diluted to 15 nM input RNA and added to four individual antigen strains (WT, K444T, E484K, F486K; Figure 17B). For the per-library experiment (Figure 20B), ARM complexes were generated using the COVID-19 antibody library and used to stain cells with 4.5 nM input RNA. After washing, total RNA was collected and purified using universal primers (scFv linker 1 P5 FWD, scFv linker 2 P5 FWD, scFv P7 REV; Table 4) that add sequencing adapters and Illumina barcodes for demultiplexing. HThe resulting library was loaded into the Reagent Nano Kit v2 (Illumina, MS-103-1003) at 1.8 nM for 500 cycles and run on a MiSeq with a 255 x 255 read length scheme using custom primers (M13 SEQ, scFv REV SEQ, scFv INDEX SEQ).

[0205] Single cell sorting Equal mixtures of the four antigen strains were stained and washed as described above, and then single live (DAPI-negative) cells were sorted into individual wells of a 96-well plate containing RNAse inhibitor and 2x RT-PCR master mix (SuperScript IV One-Step RT-PCR System, ThermoFisher). After a freeze-thaw cycle, additional reagents were added: master mix, RT, and antigen barcode or antibody V. H Two pairs of primers (SCS scFv FWD, SCS scFv REV, SCS Antigen FWD, SCS Antigen REV; Table 4) were used to amplify the RT-PCR product and add a universal PCR adapter. Three µL of the RT-PCR product was used in a second PCR (NebNext Q5 Ultra II, NEB) using a forward primer that binds to the adapter and adds one of the eight row barcodes, and an antigen- or antibody-specific reverse primer (SCS P5 FWD, SCS scFv P7 REV, SCS Antigen P7 REV) bearing one of the 12 column barcodes. Because primers are arrayed across the plate, each well obtains a unique combination of column and row barcodes, allowing for the identification of a single well. The resulting library was loaded at 1.8 nM into a Reagent Nano Kit v2 (Illumina, MS-103-1003) for 500 cycles and run on a MiSeq with a 255 x 255 read length scheme.

[0206] Drop-Seq Drop-seq encapsulation, cDNA generation, and amplification steps were performed as previously described (https: / / mccarrolllab.org / dropseq / ) with some modifications. Because cell barcodes allowed for differentiation of droplets containing multiple cells, higher cell and bead densities were used to maximize the possibility of co-encapsulation. Cells were resuspended in RB buffer containing 12% OptiPrep™ (Sigma-Aldrich, D1556-250ML) at 500 cells / µL (Figure 17D) or 400 cells / µL (Figure 20B), filtered through a 35 µm cell strainer and then a 20 µm cell strainer, and centrifuged at 30 x g for 1 min to remove any remaining large cell clumps. The supernatant was carefully aspirated without disturbing the bottom approximately 30 µL and loaded into a 1 mL syringe. Drop-seq beads (produced and purchased by ChemGenes Corporation, Macosko-2011-10(V+)) were washed according to the manufacturer's instructions and resuspended in 10 mM Tris pH 8.0, 1 mM EDTA, 0.01% Tween® at 120 or 220 beads / μL. Prior to encapsulation, beads were resuspended in Drop-seq Lysis Buffer (DLB, 200 mM Tris pH 7.5, 6% Ficoll PM-400, 0.2% sarkosyl, 20 mM EDTA, 50 mM DTT) at 240 beads / μL (FIG. 17D) or 440 beads / μL (FIG. 20B).

[0207] Microfluidic chips for Drop-seq were purchased from FlowJEM. Pico-Surf 1 (Sphere Fluidics, C023) was used as the droplet generation oil. During encapsulation, syringe pumps were operated at flow rates of 900 μL / h for each aqueous suspension and 1600 μL / h for the oil. Droplet breakage was performed as described above, except that a 2 mL Pico-Break (Sphere Fluidics, C082) was used to break the droplets.

[0208] Reverse transcription and exonuclease treatment were performed as described above, except that a modified template switch oligo (Table 4) was used. After RT and exonuclease treatment, bead concentration was determined using a Fuchs-Rosenthal cell counter (Bulldog Bio, DHC-N01). Aliquots of 8,000 beads were PCR-amplified in a 50 μL volume using 1× HiFi HotStart ReadyMix (Kapa Biosystems, KK2602) and 0.4 μM of each cDNA PCR FWD and REV primer. In Figure 17D, three aliquots were amplified. In Figure 20B, 10 aliquots were amplified.

[0209] To prepare samples for sequencing, 5 ng (Figure 17D) or 20 ng (Figure 20B) of the cDNA library was PCR-amplified in a 50 μL volume using 1x Q5 High-Fidelity Master Mix (NEB, M0494S), 0.5 μM forward primer, and 0.5 μM reverse primer. The forward and reverse primers were designed to selectively bind to and amplify antibody and antigen sequences in the cDNA library. The primers used for antibody amplification have reverse P5 / P7 Illumina adapters (Table 4): the reverse primer (Drop-seq P5 REV) contains the P5 Illumina adapter and a unique Illumina index for demultiplexing. The forward primers (scFv linker 1 P7 FWD in Figure 17D and scFv linker 2 P7 FWD in Figure 20B) contain the P7 Illumina adapter. The primers used for antigen amplification were CoV-2 S FWD in Figure 17D and antigen BC FWD in Figure 20B. The reverse primer was Drop-seq P7 REV with different Illumina indices for demultiplexing.

[0210] The antibody and antigen samples were sequenced separately. The antibody samples were sequenced on an Illumina MiSeq using 600 cycles of the MiSeq Reagent Kit v3 (Illumina, MS-102-3003), and the antigen samples were sequenced using 500 cycles of the MiSeq Reagent Nano Kit v2 (Illumina, MS-103-1003). For antibody sequencing, read 1 was 26 bp (Cell BC SEQ), which was the minimum requirement for good cluster formation on the MiSeq (bases 1–12 cell barcode, bases 13–20 UMI); index 1 was 150 bp to capture the antibody CDR3 (Strep tag II SEQ); and read 2 was 6 bp to read the Illumina barcode (Illumina BC SEQ). For antigen sequencing, read 1 was 270 bp (M13 SEQ primer) for Figure 17D and 76 bp for Figure 20B; index 1 was 6 bp to capture the Illumina barcode (Illumina BC SEQ); read 2 was 20 bp (bases 1-12 cell barcode, bases 13-20 UMI; Cell BC SEQ).

[0211] Sequencing analysis of antibody repertoire Sequencing analysis was performed as previously described. Briefly, the expected number of errors (E) of a read was calculated from its Phred score, and reads with E > 2 were discarded. IMGT immunoglobulin sequences were processed to generate position-specific sequence matrices (PSSMs) and identify framework / CDR junctions for each nucleotide sequence. Reads were required to have valid predicted CDR3H sequences. Antibody clones were then conservatively defined as having a single amino acid difference for CDR3Hs 5–6 amino acids long, and unique sequences were combined if they had a single amino acid difference for CDR3Hs > 6 amino acids long. Only clones with at least two sequencing reads were included in the analysis.

[0212] Single-cell sorting sequence analysis A custom Perl script was used to identify plate, row, and column barcodes, along with the antigen or antibody identity of each sequencing read from the Fastq file. To assign the antigen identity of each well, only wells with 10 or more total antigen reads were considered. To exclude wells with multiple cells, only wells where the antigen with the most reads represented 90% or more of the total antigen reads in a given well were considered. Further filtering was performed for wells with 600 or more total antibody reads. The percentage of antibody reads for each antigen was visualized as a heatmap / scatterplot using ggplot2 version 3.4.2 in R.

[0213] Single antigen staining sequencing analysis A read count table of antibody clones across all antigen lines was generated. The percent reads for each antibody within each antigen line were calculated. The percent differential in antibody reads between antibody compositions was then calculated for each antigen line compared to the input antibody library, followed by the log2 of the fold change (log2fc). Within each antigen, if an antibody's log2fc value was less than the log2fc of any of the negative control antibodies, the antibody's log2fc was set to NA. For visualization, the top 40 antibodies by sum of log2fc across all antigens were used. The percent antibody reads and log2fc values ​​relative to the input antibody library were visualized as a heatmap using ggplot2 version 3.4.2. NA values ​​are in white.

[0214] Drop-Seq sequencing analysis A custom Python script was used to identify Drop-seq cell barcodes, unique molecular identifiers (UMIs), and antigen barcodes or antibody CDR3Hs for each sequencing read. Only reads with unique UMIs were considered. To determine antigen identity, droplets containing multiple cells were filtered out, requiring each cell to contain 10 or more antigen reads and 90% or more single antigens. The total number of antibody reads was determined for each cell. The percentage of antibody reads for each antigen was then normalized by the number of cells. In the mini-library experiment (Figure 17D), 243 unique antigen cells were identified, and 473 cells with associated antibodies were identified. Merging these datasets resulted in 91 overlapping cell barcodes, enabling mapping of antibody-antigen interactions. In the per-library experiment (Figure 20B), a total of 2,783 cells with antigen identity were identified. The total reads for each antibody clone across all antigen cell lines were counted, and the percentage of reads for each antibody was then normalized by the number of associated cells. Normalized percent reads were visualized as a heatmap using ggplot2 version 3.4.2.

[0215] Protein expression The receptor-binding domain of spike harboring the K444T, E484K, or F486K mutation was cloned in frame with the hinge and Fc of mouse IgG2a in a mammalian expression construct. Each plasmid was transiently expressed using the Expi293™ system (ThermoFisher) and purified using Protein A chromatography (PrismA, Cytiva). Size and purity were verified using SDS-PAGE and SE-HPLC.

[0216] V of selected monoclonal antibodies H and V LThe sequences were cloned into separate vectors containing the human IgG1 or human kappa constant region. Plasmids were transiently transfected into HEK293 cells, and antibodies were provided as harvested cell culture fluid (Twist Bioscience).

[0217] Several monoclonal antibodies were cloned into stable expression vectors containing FRT and GS genes compatible with landing pad cell lines using human IgG1 or kappa constant regions. These were transfected and selected as described above to generate stable antibody-expressing lines. Production runs were performed as described above, and supernatants were collected for binding studies.

[0218] Affinity measurement Biolayer interferometry (BLI) kinetic affinity measurements were performed using a GatorPrime Instrument (GatorBio). ScFvs were generated using an in vitro transcription-translation kit (PURExpress, NEB, E6800). 5 μg / mL biotinylated strep tag II capture antibody (GenScript, A01737) was bound to a Gator streptavidin-coated probe (GatorBio, 160002) at 50% saturation for 30 seconds, followed by loading of the 10x diluted scFv reaction mixture for 600 seconds. Association and dissociation data were collected for 180 and 300 seconds for each scFv against each RBD variant. RBD antigen concentrations were titrated to 100 nM, 33 nM, 11 nM, and 0 nM, and multi-cycle kinetic assays were performed. Kinetics were analyzed using a double-referencing strategy as previously described and fitted to an overall 1:1 kinetic model of Rmax Unlinked. A 30 s baseline measurement followed the capture probe, scFv loading and association steps.

[0219] Lentiviral transfection and stable cell line generation The amino acid sequence of CoV-2 S WT (UniProt P0DTC2) was cloned into a plasmid containing the EF1α promoter and 5' and 3' lentiviral LTRs, along with the human IgK signal peptide, c-myc tag, and PDGFR transmembrane region. This plasmid was transiently transfected into HEKTa cells along with Rev, Gag / Pol, and VSV-G envelope plasmids for lentiviral packaging. The lentivirus-containing supernatant was collected at 24 and 48 hours, filtered, and added to Expi293™ cells (ThermoFisher) at a final dilution of 1:4. Transduced cells were stained and sorted for c-myc expression.

[0220] Checking spike surface expression Cells were stained with serial dilutions of bamlanivimab and anti-hIgG1-Fc-PE, washed, and analyzed by flow cytometry (CytoFLEX LX flow cytometer, Beckman Coulter). Median fluorescence intensity of the PE signal is reported.

[0221] In vitro translation optimization A 1:1 mixture of RNA encoding the positive (casirivimab) scFv and control (ipilimumab) scFv was added at 450 nM to an in vitro transcription / translation kit with or without ribosome release factor (PURExpress, NEB, E6800) or without (PURExpress ΔRF123, NEB E6850) and incubated at 37 °C for 10 or 30 minutes to generate ARM complexes. These mixtures were diluted to 90 nM and used to stain CoV-2 S WT CHOZN cells. Total RNA from washed cells was collected and quantified using RT-qPCR (One-Step RNA-to-Ct Kit, ThermoFisher). A TaqMan probe against the CHO housekeeping gene Fkpb1a was used to normalize cell numbers, while probes specific for the unique CDRH3 sequences of casirivimab or ipilimumab, along with a standard curve, were used to quantify cell-bound ARM RNA. The ratio of casirivimab RNA to ipilimumab RNA is reported (Figure 15).

[0222] Competitive ARM staining An equimolar mixture of RNAs encoding 14 scFvs in a ribosome display format was diluted 1:1 with RNA from an unrelated library and used to generate ARM complexes. The mixture was diluted to 45 nM or 2.25 nM positive ARM and used to stain CoV-2 S WT CHOZN cells. Total RNA was collected, and the VH region was amplified and sequenced. Bound ARMs were quantified by identifying reads with the correct CDRH3 match and quantifying the percentage of each clone. The enrichment of each clone was calculated by dividing the clonal frequency by the input staining sample (Figure 18).

[0223] Affinity measurement Clones selected from competitive staining were expressed as scFvs as previously described. Single-cycle BLI kinetic measurements were performed using a GatorPrime Instrument (GatorBio). ScFvs were generated using an in vitro transcription-translation kit (PURExpress, NEB, E6800) and then loaded onto a Strep-Tactin XT probe (GatorBio, 160033) for 60 or 120 seconds, followed by a 30 or 60 second association and a 300 second dissociation step. RBD antigen concentrations were titrated to 100 nM, 33 nM, 11 nM, and 0 nM and analyzed using a single-reference strategy and fitted to an overall 1:1 kinetic model of Rmax Unlinked. A 60 second baseline measurement was performed before scFv loading, and a 300 or 700 second measurement was performed before the association step. [Table 4-1] [Table 4-2] [Table 4-3] List of oligo sequences used in this study. The 6-base Illumina indexing barcode is represented by [INDEX]. For single cell sorting (SCS) primers, the 5-base plate, row, and column barcodes are represented by [PLATE], [ROW], and [COLUMN], respectively.

[0224] 8. Equivalents and Incorporation by Reference While the present invention has been particularly shown and described with reference to preferred and various alternative embodiments, it will be understood by those skilled in the art that various changes in form and detail can be made therein without departing from the spirit and scope of the invention.

[0225] All references, issued patents, and patent applications cited within the body of this specification are hereby incorporated by reference in their entirety for all purposes. Table 5-1 Table 5-2 Table 5-3 Table 5-4 Table 5-5 Table 5-6 Table 5-7 Table 5-8 Table 5-9 Table 5-10 Table 5-11 Table 5-12 Table 5-13 Table 5-14 Table 5-15 Table 5-16 Table 5-17 Table 5-18 Table 5-19 Table 5-20 Table 5-21 Table 5-22

Claims

1. 1. A method for high-throughput analysis of target binding proteins (TBPs), comprising: providing a library of target-decorated cells, wherein each of said target-decorated cells displays a target of interest on its membrane; contacting the library of target decorated cells with a plurality of TBP-ribosome-mRNA (TRM) complexes, thereby inducing binding between the target decorated cells and the TRM complexes; generating a plurality of monodisperse or polydisperse emulsion microdroplets, each microdroplet containing a single one of the target decorated cells, one or more TRM complexes bound to the single cell, and a lysis reagent that induces lysis of the single cell; capturing the RNA released from the single cells on a solid surface or within a semipermeable shell; and generating a library of hybrid polynucleic acids comprising sequences from the transcripts of said single cell and / or sequences from said mRNA of said TRM complex; A method comprising:

2. 10. The method of claim 1, further comprising sequencing said library of hybrid polynucleic acids.

3. 3. The method of claim 2, further comprising identifying target-TBP pairs based on said sequencing of said library of hybrid polynucleic acids.

4. 3. The method of claim 2, further comprising identifying a target binding protein specific to the target of interest.

5. 3. The method of claim 2, further comprising identifying the binding affinity or specificity of a target binding protein specific for the target of interest.

6. 6. The method of any one of claims 2 to 5, further comprising determining a lead distribution of multiple TBPs in said multiple TRM complexes across two or more targets of interest.

7. The method of claim 6 , further comprising normalizing the read distribution based on an input distribution of the plurality of TBPs.

8. 8. The method of any one of claims 1 to 7, wherein at least one target of interest comprises or is conjugated to a domain that is capable of inducing expression of an activation marker in target-decorated cells when said target of interest binds to TBP.

9. 9. The method of claim 8, wherein generating a plurality of monodisperse or polydisperse emulsion microdroplets comprises sorting the target decorated cells based on the presence or absence of the activation marker.

10. The method of any one of claims 1 to 9, wherein the target-decorated cells express a fusion protein comprising a target of interest and a transmembrane domain.

11. The method of claim 10, wherein the target decorated cells comprise a construct encoding a fusion protein comprising a target of interest and a transmembrane domain.

12. The method of claim 11 , wherein the construct further comprises a barcode sequence.

13. The method of claims 10 to 12, wherein the construct further comprises a sequence encoding a fluorescent protein.

14. 14. The method of claim 13, further comprising isolating the plurality of monodisperse or polydisperse emulsion microdroplets containing the target decorated cells by detecting expression of the fluorescent protein.

15. The method of any one of claims 1 to 14, wherein said library of target-decorated cells comprises one cell clone displaying one target of interest.

16. 15. The method of any one of claims 1 to 14, wherein the library of target-decorated cells comprises two, three, or four cell clones, each of which displays a unique target of interest that is different from the other cell clones.

17. 15. The method of any one of claims 1 to 14, wherein the library of target-decorated cells comprises at least five cell clones, wherein each of the cell clones displays a unique target of interest that is different from the other cell clones.

18. 18. The method of claim 17, wherein the library of target-decorated cells comprises at least 10 cell clones, wherein each of the cell clones displays a unique target of interest that is distinct from the other cell clones.

19. 19. The method of claim 18, wherein the library of target-decorated cells comprises at least 100 cell clones or at least 1000 cell clones, wherein each of the cell clones displays a unique target of interest that is distinct from other cell clones.

20. 20. The method of any one of claims 1 to 19, wherein the transcript of the isolated single cell comprises the coding sequence of the target of interest.

21. 20. The method of any one of claims 1 to 19, wherein the transcript of the isolated single cell comprises a barcode sequence.

22. The method of any one of claims 1 to 21, wherein the target is an antigen, a ligand, a receptor or a modification thereof.

23. The method of any one of claims 1 to 22, wherein the target binding protein is an antibody, a ligand, a receptor or a modification thereof.

24. The method of any one of claims 1 to 23, wherein each of the TRM complexes comprises a target binding protein.

25. 25. The method of claim 24, wherein each of the TRM complexes comprises a TBP comprising an scFv.

26. The method of any one of claims 1 to 23, wherein each of the TRM complexes comprises a TBP comprising a heavy chain variable region.

27. The method of any one of claims 1 to 23, wherein each of the TRM complexes comprises a TBP comprising a light chain variable region.

28. 28. The method of any one of claims 1-27, wherein the plurality of TRM complexes comprises 1 to 5 unique TRM complexes, wherein each of the unique TRM complexes comprises a unique target binding protein that is distinct from the remainder of the unique TRMS.

29. 28. The method of any one of claims 1-27, wherein the plurality of TRM complexes comprises 6-10 unique TRM complexes, wherein each of the unique TRM complexes comprises a unique target binding protein that is different from the remainder of the unique TRM complexes.

30. 28. The method of any one of claims 1-27, wherein the plurality of TRM complexes comprises at least 10 unique TRM complexes, wherein each of the unique TRM complexes comprises a unique target binding protein that is distinct from other unique TRM complexes.

31. 31. The method of claim 30, wherein the plurality of TRM complexes comprises at least 100 unique TRM complexes, wherein each of the unique TRM complexes comprises a unique target binding protein that is distinct from other unique TRM complexes.

32. 32. The method of claim 31 , wherein the plurality of TRM complexes comprises at least 1000 unique TRM complexes, wherein each of the unique TRM complexes comprises a unique target binding protein that is distinct from other unique TRM complexes.

33. 33. The method of any one of claims 1 to 32, wherein each of the TRM complexes comprises a target binding protein and an mRNA encoding the target binding protein.

34. 34. The method of claim 33, wherein the mRNA comprises a coding sequence for a complementarity determining region (CDR) of the target binding protein.

35. 35. The method of claim 34, wherein the mRNA comprises a coding sequence for CDR3 of the target binding protein.

36. The method of any one of claims 1 to 35, wherein the mRNA comprises a barcode sequence.

37. The method of any one of claims 1 to 36, wherein RNA capture is performed using oligonucleotides immobilized on beads.

38. 38. The method of claim 37, wherein RNA capture is performed using oligonucleotides immobilized on beads, wherein each bead has a diameter greater than 10 μm, between 0.5 and 10 μm, less than 1 μm, or about 1 μm.

39. 39. The method of claim 37 or 38, wherein the oligonucleotides immobilized on beads comprise a barcode sequence.

40. 40. The method of claim 39, wherein the barcode sequence in the oligonucleotide immobilized on a bead is unique for a given bead.

41. 41. The method of claim 39 or 40, wherein the hybrid polynucleic acid further comprises a barcode sequence derived from the oligonucleotide immobilized on a bead.

42. The method of any one of claims 1 to 41, wherein the hybrid polynucleic acid is generated by overlap extension polymerase chain reaction (OE-PCR).

43. 43. The method of any one of claims 1 to 42, wherein the generation of said hybrid polynucleic acid is preceded by first strand cDNA synthesis.

44. 44. The method of any one of claims 37 to 43, further comprising generating a second set of emulsion microdroplets comprising the bead-captured RNA released from the single cells prior to generating the library of hybrid polynucleic acids.

45. 45. The method of claim 44, wherein said library of hybrid polynucleic acids is generated in said second set of emulsion microdroplets.

46. The step of contacting the library of target decorated cells with the plurality of TBP-ribosome-mRNA (TRM) complexes comprises contacting the library with 25 to 100 mM Mg ２＋ 46. ​​The method of any one of claims 1 to 45, carried out in a buffer comprising:

47. The buffer contains 50 mM Mg ２＋ 47. The method of claim 46, comprising:

48. The buffer contains 50 mM MgCl ２ 48. The method of claim 47, comprising:

49. The buffer solution was HEPES, NaCl, 50mM MgCl ２ , polysorbate 20, heparin, and BSA.

50. The buffer solution was 20mM HEPES, 50mM NaCl, 50mM MgCl. ２ 50. The method of claim 49, comprising 0.01% polysorbate 20, 2.5 mg / ml heparin, and 0.5% BSA.

51. The buffer solution was 20mM HEPES, 50mM NaCl, 50mM MgCl. ２ 50. The method of claim 49, comprising 0.01% polysorbate 20, 2.5 mg / ml heparin, and 0.05% BSA.

52. 1. A kit for high-throughput analysis of target binding proteins (TBPs), comprising: multiple constructs, each construct encoding a fusion protein containing a unique target of interest and a transmembrane domain; a plurality of TBP-ribosome-mRNA (TRM) complexes, each TRM complex containing a unique target binding protein; and a buffer solution.

53. 53. The kit of claim 52, further comprising a host cell.

54. 54. The kit of claim 52 or 53, wherein the kit comprises a single construct encoding one target of interest.

55. 54. The kit of claim 52 or 53, wherein the kit comprises 2 to 10 unique constructs, wherein each unique construct encodes a unique target of interest.

56. 54. The kit of claim 52 or 53, wherein the kit comprises at least 10 unique constructs, wherein each of the unique constructs encodes a unique target of interest.

57. 55. The kit of claim 54, wherein the kit comprises at least 100 unique constructs.

58. 58. The kit of claim 57, wherein the kit comprises at least 1000 unique constructs.

59. 59. The kit of any one of claims 52-58, comprising one unique TRM complex, wherein said one unique TRM complex comprises one target binding protein.

60. 59. The kit of any one of claims 52-58, comprising at least 10 unique TRM complexes, wherein each of the unique TRM complexes comprises a unique target binding protein.

61. 60. The kit of claim 59, comprising at least 100 unique TRM complexes.

62. 62. The kit of claim 61, comprising at least 1000 unique TRM complexes.

63. 63. The kit of any one of claims 52 to 62, further comprising reagents for overlap extension polymerase chain reaction (OE-PCR).

64. The buffer contains 50 mM Mg ２＋ 64. The kit of any one of claims 52 to 63, comprising:

65. The buffer contains 50 mM MgCl ２ 65. The kit of claim 64, comprising:

66. The buffer solution was HEPES, NaCl, 50mM MgCl ２ 66. The kit of any one of claims 52 to 65, comprising: , polysorbate 20, heparin, and BSA.

67. The buffer solution was 20mM HEPES, 50mM NaCl, 50mM MgCl. ２ , 0.01% polysorbate 20, 2.5 mg / ml heparin, and 0.5% BSA.

68. The buffer solution was 20mM HEPES, 50mM NaCl, 50mM MgCl. ２ , 0.01% polysorbate 20, 2.5 mg / ml heparin, and 0.05% BSA.

69. A library of hybrid polynucleic acids produced by the method of any one of claims 1 to 51.

70. 1. A method for high throughput analysis of receptors, comprising: providing a library of ligand-decorated cells, wherein each of said ligand-decorated cells displays a ligand of interest on its membrane; contacting the library of ligand-decorated cells with a plurality of receptor-ribosome-mRNA (RRM) complexes, thereby inducing binding between the ligand-decorated cells and the RRM complexes; generating a plurality of emulsion microdroplets, each containing a single one of the ligand-decorated cells, one or more RRM complexes bound to the single cell, and a lysis reagent that induces lysis of the single cell; capturing the RNA released from the single cells on a solid surface or within a semipermeable shell; and generating a library of hybrid polynucleic acids comprising sequences from the transcripts of said single cell and / or sequences from said mRNA of said RRM complex; A method comprising:

71. 1. A method for high throughput analysis of receptors, comprising: providing a library of receptor-decorated cells, wherein each of said receptor-decorated cells displays a receptor of interest on its membrane; contacting the library of receptor-decorated cells with a plurality of ligand-ribosome-mRNA (LRM) complexes, thereby inducing binding between the receptor-decorated cells and the LRM complexes; generating a plurality of emulsion microdroplets, each microdroplet containing a single one of the receptor-decorated cells, one or more LRM complexes bound to the single cell, and a lysis reagent that induces lysis of the single cell; capturing the RNA released from the single cells on a solid surface or within a semipermeable shell; and generating a library of hybrid polynucleic acids comprising sequences from the transcripts of said single cell and / or sequences from said mRNA of said LRM complex.

Images