Molecular Indexing of Proteins by Self-Assembly (MIPSA) for Efficient Proteomics Research

Abstract

The present disclosure relates to the field of proteomics. More specifically, the present disclosure provides compositions and methods for molecular indexing of proteins by self-assembly. In one aspect, the present disclosure provides a library of self-assembling protein-DNA conjugates. In certain embodiments, each protein-DNA conjugate comprises: (a) a cDNA comprising a barcode, the cDNA being conjugated with a ligand that specifically binds to a polypeptide tag; and (b) a fusion protein comprising the polypeptide tag and a protein of interest, the ligand being covalently bound to the polypeptide tag.

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of U.S. Provisional Patent Application No. 63 / 155,086, filed March 1, 2021, which is incorporated by reference herein in its entirety. Government Assistance Clause GOVERNMENT SUPPORT This invention was made with Government support under Grant No. GM127353 awarded by the National Institutes of Health. The Government has certain rights in this invention.

[0002] Field The present disclosure relates to the field of proteomics. More specifically, the present disclosure provides compositions and methods for molecular indexing of proteins by self-assembly.

[0003] Incorporation by Reference of Electronically Submitted Materials This application contains a Sequence Listing. The Sequence Listing has been submitted electronically via EFS-Web as an ASCII text file entitled "P16720_01_ST25.txt". The Sequence Listing is 10,148 bytes in size and was created on March 1, 2021, which is incorporated by reference in its entirety. [Background technology]

[0004] Unbiased analysis of antibody binding specificity can provide important insights into healthy and disease states. (1-4) Although phage display is useful for these and many other applications, most protein-protein, protein-antibody and protein-small molecule interactions require some degree of conformation that is not captured using programmable phage display. Profiling conformational protein interactions at the proteome scale has traditionally relied on protein microarray technology. However, protein microarrays tend to suffer from high per-assay costs and a myriad of technical artifacts, including those associated with high-throughput expression and purification of proteins, spotting of proteins onto solid supports, drying and rehydration of arrayed proteins, and slide-scanning fluorescent imaging-based readout. (5, 6) Although alternative approaches to protein microarray production and storage have been developed (e.g., nucleic acid programmable protein arrays, NAPPA (7) or SIMPLEX (8)), robust, scalable and cost-effective technologies are lacking.

[0005] To overcome the limitations associated with array-based profiling of full-length proteins, we previously established a methodology called Parallel Analysis of Translated Open reading frames (PLATO), which utilizes ribosome display of ORFeome libraries (9). Ribosome display relies on the in vitro translation of mRNAs lacking stop codons, stalling ribosomes at the ends of mRNA molecules in complex with the nascent proteins they encode. PLATO suffers from several important limitations that limit its usefulness. An ideal alternative would be to covalently link proteins to short, amplifiable DNA barcodes. Indeed, individually prepared DNA barcoded antibodies and proteins have been employed in a myriad of applications, as recently reviewed by Liszczak and Muir. (10) One particularly attractive protein-DNA conjugation method involves the Halo-tag system, which utilizes a bacterial enzyme that forms an irreversible covalent bond with a halogen-terminal alkane moiety. (11) Individual DNA-barcoded Halo tag fusion proteins have been shown to significantly improve the sensitivity and dynamic range of autoantibody detection compared to traditional ELISA. (12) Scaling individual protein barcoding to whole ORFeome libraries would be highly valuable but challenging due to high cost and low throughput. Thus, a self-assembly approach may provide a much more efficient route to library generation. Summary of the Invention [Means for solving the problem]

[0006] The present disclosure is based, at least in part, on the development of a novel molecular display technology, Molecular Indexing of Proteins by Self Assembly (MIPSA), which overcomes major drawbacks of PLATO and other full-length protein array technologies. In certain embodiments, MIPSA produces a library of soluble full-length proteins, each of which is uniquely identifiable via covalent attachment to a DNA barcode adjacent to a universal PCR primer binding sequence (FIGS. 1A-1C). The barcode is introduced near the 5' end of the transcribed mRNA sequence, upstream of the ribosome binding site (RBS). Reverse transcription (RT) of the 5' end of the in vitro transcribed mRNA creates a cDNA barcode, which in some embodiments is linked to a haloalkane-labeled RT primer. An N-terminal Halo tag fusion protein is encoded downstream of the RBS, and in vitro translation results in covalent attachment of the cDNA barcode to the protein product encoded by the Halo tag and its downstream open reading frame (ORF) in a complex. The resulting library of uniquely indexed full-length proteins can be used for inexpensive proteome-wide interaction studies such as unbiased autoantibody profiling. As described below, in one embodiment, we demonstrate the utility of the platform by uncovering known and novel autoantibodies in the plasma of patients with severe COVID-19.

[0007] In one aspect, the present disclosure provides a method for efficient proteomic interrogation via molecular indexing of proteins by self-assembly (MIPSA). In one embodiment, the method comprises the steps of: (a) transcribing a vector library into messenger ribonucleic acid (mRNA), the vector library encoding a plurality of proteins, each vector of the vector library being comprised in a 5' to 3' direction of: (i) a polymerase transcription initiation site; (ii) a barcode; (iii) a reverse transcription primer binding site; (iv) a ribosome binding site (RBS); and (v) a nucleotide sequence encoding a fusion protein comprising (1) a polypeptide tag and (2) a protein, wherein the polypeptide tag specifically binds to a ligand; (b) reverse transcribing the 5' end of the mRNA using a primer that binds upstream of the RBS, the primer being conjugated to the ligand that specifically binds to the polypeptide tag of the fusion protein, forming a complementary deoxyribonucleic acid (cDNA) comprising the ligand, primer and barcode; and (c) translating the mRNA, wherein the ligand of the cDNA binds to the polypeptide tag of the fusion protein. In certain embodiments, the vector library is nicked prior to step (a). In other certain embodiments, the vector further comprises (vi) an endonuclease site for vector linearization, and the vector library is linearized prior to step (a).

[0008] In another aspect, the present disclosure provides a self-assembling protein-DNA conjugate composition.In certain embodiments, the present disclosure provides a library of self-assembling protein-DNA conjugates.In certain embodiments, each protein-DNA conjugate comprises:(a) a cDNA comprising a barcode, the cDNA being conjugated with a ligand that specifically binds to a polypeptide tag; and (b) a fusion protein comprising the polypeptide tag and a protein of interest, the ligand being covalently bound to the polypeptide tag.

[0009] In certain embodiments, the polypeptide tag is a haloalkane dehalogenase or 6 -alkylguanine-DNA-alkyltransferase. In certain embodiments, the polypeptide tag comprises a HALO tag and the ligand comprises a HALO ligand. In more specific embodiments, the HALO tag comprises the amino acid sequence set forth in SEQ ID NO: 22. In other embodiments, the HALO-ligand comprises any one of the following: [ka]

[0010] In another embodiment, the polypeptide tag comprises a SNAP tag and the ligand comprises a SNAP ligand. In a more specific embodiment, the SNAP tag comprises the amino acid sequence set forth in SEQ ID NO: 23. In other embodiments, the SNAP-ligand comprises benzylguanine or a derivative thereof.

[0011] In further embodiments, the polypeptide tag comprises a CLIP tag and the ligand comprises a CLIP ligand. In more specific embodiments, the CLIP tag comprises the amino acid sequence set forth in SEQ ID NO: 24. In other aspects, the CLIP-ligand comprises benzylcytosine or a derivative thereof.

[0012] The present disclosure also provides a method for using a library of self-assembled protein-DNA conjugates. In one embodiment, the method for studying protein-protein interactions comprises performing a pull-down assay of a library of protein-DNA conjugates with a protein of interest. In another embodiment, the method for studying protein-small molecule interactions comprises performing a pull-down assay of a library of protein-DNA conjugates with a small molecule. In yet another embodiment, the method comprises performing immunoprecipitation of a library of protein-DNA conjugates with an antibody obtained from a biological sample. In a further embodiment, the method for identifying a target of a first small molecule comprises (a) incubating a library of protein-DNA conjugates with a first small molecule that binds to the target, and (b) performing a pull-down assay of the library of step (a) with a second small molecule, where the first small molecule bound to the target blocks the binding of the second small molecule. In a more specific embodiment, two or more small molecules are used in the pull-down assay of step (b).

[0013] In yet another aspect, the present disclosure provides a self-assembly protein display library comprising a vector and a plurality of vectors, each vector comprising a nucleic acid sequence encoding a protein of interest.In one embodiment, the vector comprises, along the 5' to 3' direction, (a) a polymerase transcription initiation site; (b) a barcode; (c) a reverse transcription primer binding site; (d) an RBS; and (e) a nucleotide sequence encoding a fusion protein comprising (i) a polypeptide tag and (ii) a protein of interest, wherein the polypeptide tag specifically binds to a ligand.

[0014] In certain embodiments, the vector further comprises an endonuclease site for vector linearization, hi other embodiments, the vector further comprises (vii) a stop codon.

[0015] In certain embodiments, the barcode is adjacent to a polymerase chain reaction (PCR) primer binding site, hi another embodiment, the barcode comprises a PCR primer binding site.

[0016] In another embodiment, the RBS comprises an internal ribosome entry site. In certain embodiments, the polypeptide tag is fused to the N-terminus of the protein of interest. In other embodiments, the polypeptide tag is fused to the C-terminus of the protein of interest.

[0017] The present disclosure also provides a method for using self-assembling protein display library.In certain embodiments, the method comprises: (a) transcribing a plurality of linearized or nicked vectors comprising self-assembling protein display library to produce mRNA; (b) reverse transcribing the 5' end of mRNA using a primer conjugated to a ligand to produce a cDNA comprising a barcode; and (c) translating the mRNA, where the polypeptide tag of the fusion protein is covalently linked to the ligand conjugated to the cDNA comprising the barcode.

[0018] In another aspect, the disclosure provides a method for treating COVID-19. In one embodiment, a method for treating a patient with severe COVID-19 comprises administering an effective amount of interferon therapy to the patient, and an autoantibody that neutralizes IFN-λ3 is detected in a biological sample obtained from the patient. In another embodiment, a method for treating a patient with severe COVID-19 comprises (a) detecting an autoantibody that neutralizes IFN-λ3 in a biological sample obtained from the patient, and (b) treating the patient with an effective amount of interferon therapy. In a further embodiment, a method for identifying a COVID-19 patient who would benefit from interferon therapy comprises detecting an autoantibody that neutralizes IFN-λ3 in a biological sample obtained from the patient. In certain embodiments, the interferon therapy comprises interferon lambda (IFN-λ) or interferon beta (IFN-β). In certain embodiments, the interferon lambda (IFN-λ) or interferon beta (IFN-β) is pegylated. [Brief description of the drawings]

[0019] [Fig. 1A-1G]The MIPSA method is shown. Figure 1A: Schematic of the recombinant pDEST-MIPSA vector highlighting key components: unique clone identifier (UCI, blue), ribosome binding site (RBS, yellow), N-terminal Halo tag (purple), FLAG epitope (orange), open reading frame (ORF, green), and I-SceI restriction endonuclease site for vector linearization (black). Figure 1B: Schematic showing in vitro transcribed (IVT) RNA from the vector template shown in (Figure 1A). Isothermal base-balanced UCI sequence: (SW)18-AGGGA-(SW)18. Figure 1C: Cell-free translation of the RNA-cDNA shown in (Figure 1B). The Halo tag protein forms a covalent bond with the UCI-containing cDNA conjugated in cis with the Halo ligand during translation. Figure 1D: RT primer positions tested for their effect on translation. Figure 1E: α-FLAG Western blot analysis of translation in the presence of RT primers shown in (Figure 1D) (NC, negative control, no RT primer). Figure 1F: Western blot analysis of TRIM21 protein translated from RNA with UCI-cDNA primed from position -32, conjugated (+) or not (-) with Halo ligand. Sjögren's syndrome, SS; healthy control, HC. Figure 1G: qPCR analysis of IPed TRIM21 UCI. Fold difference is by comparison to Halo ligand (-) HC IP. [Figures 2A-2D]Cis- vs. trans-UCI conjugation is shown. Figure 2A: IVT-RNA encoding TRIM21 or GAPDH with distinct UCI barcodes were translated before or after mixing in a 1:1 ratio. qPCR analysis of IPs using UCI-specific primers reported as fold change relative to IP with HC plasma when IVT-RNA was mixed post-translation. Figure 2B: IVT-RNA encoding TRIM21 (black UCI) and GAPDH (grey UCI) were mixed 1:1 in a background of 100-fold excess of GAPDH (white UCI) and then translated as mock libraries. Sequencing analysis of IPs reported as fold change relative to HC IP of 100x GAPDH. Figure 2C: hORFeome MIPSA library containing spike-in TRIM21 IPed with SS plasma and compared to the average of 8 mock IPs (no plasma input). TRIM21 UCI is shown in red. FIG. 2D: Relative fold difference in TRIM21 UCI in SS vs. HC IPs as determined by sequencing. [Figure 3A-3D]Construction of the UCI-ORF dictionary is shown. Figure 3A: (i) Tagmentation randomly inserts adapters into the MIPSA vector library. (ii) Using the PCR1 forward primer and the reverse primer of the adapter inserted by tagmentation, DNA fragments are amplified and size-selected to be approximately 1.5 kb capturing the 5' end of the ORF. (iii) These fragments are amplified using the PCR2 forward primer containing P5 and the P7 reverse primer. (iv) Illumina sequencing is used to read the UCIs and ORFs from the same fragments, allowing their association in the dictionary. Figure 3B: For each member of the pDEST-MIPSA hORFeome library, the number of monospecific UCIs is shown overlaid on the length of the ORF. Figure 3C: Histogram of ORF representation in the library by their aggregated UCI-associated read counts. Red vertical lines indicate + / - 10-fold of the median UCI-associated read counts. Figure 3D: IP of the hORFeome MIPSA library with Sjögren's syndrome (SS) plasma compared to the average of eight mock IPs. Sequencing read counts for each UCI are plotted. UCIs associated with two GAPDH isoforms (filled black) and spiked-in TRIM21 (red) are shown. [Figure 4A-4C]MIPSA analysis of autoantibodies in severe COVID-19. Figure 4A: Box plot showing total number of autoreactive proteins in plasma from healthy controls, mild-moderate COVID-19 patients, or severe COVID-19 patients. * indicates p<0.05 by one-tailed t-test to compare means. Figure 4B: Hierarchical cluster map of all proteins represented by at least two reactive UCIs in at least one severe COVID-19 plasma, but not in one or less control (healthy or mild-moderate COVID-19 plasma). Figure 4C: MIPSA analysis of autoantibodies in 10 inclusion body myositis (IBM) patients and 10 healthy controls (HC) using hORFeome libraries. Fold change of IPed 5'-nucleotidase, cytoplasmic 1A (NT5C1A) measured as both UCI-qPCR fold change (relative to the average of 10 HCs) and sequencing fold change (relative to mock IP). [Figure 5A-5H]Showing that MIPSA detects known and novel neutralizing interferon autoantibodies. Figure 5A-5C: Scatter plots highlighting reactive interferon UCIs for three severe COVID-19 patients. Figure 5D: Summary of interferon reactivity detected in 5 of 55 individuals with severe COVID-19. Hit fold change values ​​(cell color) and number of reactive UCIs (number in cells) are provided. Figure 5E-5F: Recombinant interferon alpha 2 (IFN-α2) or interferon lambda 3 (IFN-λ3) neutralizing activity of the same patient shown in Figure 5D. Prior to incubation with A549 cells, plasma was preincubated with 100 U / ml IFN-α2 or 1 ng / ml IFN-λ3. Fold change of the interferon stimulated gene MX1 was calculated by RT-qPCR relative to unstimulated cells. GAPDH was used as a housekeeping control gene for normalization. Red bars indicate which samples are predicted by MIPSA to have neutralizing activity for each interferon. Figure 5G: PhIP-Seq analysis of interferon autoantibodies in the five patients from Figure 5D (row and column order maintained). Hit fold change values ​​(cell color) and number of reactive peptides (number in cell) are provided. Figure 5H: Epitopefindr analysis of PhIP-Seq reactive type I interferon 90-aa peptides. [Figure 6A-6C] Halo ligand conjugation to a reverse transcription primer. Figure 6A: Top is an oligonucleotide reverse transcription (RT) primer sequence modified with a 5' primary amine. Bottom is a Halo ligand with a reactive succinimidyl ester group separated by one ethylene glycol moiety (O2). The succinimidyl ester reacts with the primary amine to form an amide bond between the RT primer and the Halo ligand, resulting in a Halo ligand-conjugated RT primer. Figure 6B: HPLC chromatography of the RT primer without the Halo ligand modification. Figure 6C: HPLC chromatography of the RT primer with the Halo ligand modification after purification. The conjugated product elutes later due to the increased hydrophobicity imparted by the modification. [Figure 7A-7C]Cis vs. trans UCI-ORF associations are shown. Schematic of cis FIG. 7A vs. trans FIG. 7B of UCI-ORF conjugation during translation of MIPSA IVT-RNA library. FIG. 7C: Left panel: 50% cis conjugates ("C") consisting of correct protein-UCI associations (e.g., blue UCI and blue protein). Center panel: Unconjugated protein then randomly associates in trans ("T") with unconjugated UCI. Right panel: The ratio of correctly IPed UCI to incorrectly IPed UCI in this two-part experiment is 3:1 (75%:25%), similar to the experimental observations (FIG. 2A). [Figure 8] Dual translation and IP of TRIM21 and GAPDH are shown. TRIM21(T) and GAPDH(G) IVT-RNA-cDNA were translated separately or together, and then IP was performed with healthy control (HC) or Sjögren's syndrome (SS) plasma. Analysis was performed by immunoblotting with M2 antibody, which recognizes the common FLAG epitope tag that links the Halo tag to the protein. [Figure 9] Interferon sequence homology is shown. The pairwise blastp alignment bit score matrix for all interferon (IFN) proteins is shown in Figure 5D. [Figure 10A-10C] The reproducibility and linearity of MIPSA detection of autoantibodies in patient P2 are shown. Figure 10A: Mean and standard deviation of 100 ORF fold changes for all consistently reactive monospecific UCIs (fold change >3 in all three replicates). Values ​​to the right of the error bars are the coefficient of variation. Figure 10B: Number of overlapping reactive monospecific UCIs across three independent MIPSA analyses of P2 plasma. Area is proportional to number of hits. Figure 10C: Mean ORF fold change for P2 plasma compared to P2 plasma diluted 10-fold into background of healthy control plasma. Dot size indicates the number of reactive UCIs corresponding to each ORF. [Figure 11A-11B]Figure 11 shows titration-based estimates of interferon autoantibody levels for patient P2. Mouse monoclonal blocking antibodies were used at different concentrations in a cell-based IFN neutralization assay: Figure 11A: IFN-α2 and Figure 11B IFN-λ3. Neutralization curves were fitted and used to estimate the corresponding interferon autoantibody levels for patient P2. The plasma dilutions shown were chosen to be within the dynamic range of the assay; neutralizing activity of P2 plasma at the indicated dilutions was assayed in triplicate. [Figures 12A-12C] MIPSA analysis of interferon antibodies in serial dilutions. Summary of interferon reactivity detected by MIPSA in serially diluted P2 plasma (FIG. 12A), IFN-α2 mAb (FIG. 12B), and IFN-λ3 mAb (FIG. 12C). Hit fold change values ​​(color within cell) and number of reactive UCIs (number within cell) are provided as in FIG. 5D. [Figure 13] IFN-λ3 autoantibodies do not efficiently neutralize IFN-λ1. The IFN-λ3 neutralizing activity of plasma from patient P2 was compared with its IFN-λ1 neutralizing activity. Neutralization of IFN-λ3 was complete and partial at 1:10 and 1:100 dilutions, respectively. Neutralization of IFN-λ1 was partial and not detected at 1:10 and 1:100 dilutions, respectively (ND). DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0020] It is understood that the disclosure is not limited to the specific methods and components described herein, which may vary. It should also be understood that the terms used herein are used for the purpose of describing specific embodiments only, and are not intended to limit the scope of the disclosure. It should be noted that, as used in this specification and the appended claims, the singular forms "a", "an", and "the" include plural references unless the context clearly indicates otherwise. Thus, for example, a reference to a "protein" is a reference to one or more proteins, including equivalents thereof known to those skilled in the art, and so forth.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. Specific methods, devices, and materials are described, although any methods and materials similar or equivalent to those described herein can be used in the practice or testing of this disclosure.

[0022] All publications cited herein, including all journal articles, books, manuals, published patent applications, and issued patents, are incorporated herein by reference.In addition, the meanings of certain terms and phrases used in this specification, examples, and the appended claims are provided.The definitions are not meant to be limiting in nature, but serve to provide a clearer understanding of certain aspects of the present disclosure.

[0023] We describe herein a novel molecular display technology for full-length proteins that offers significant advantages over protein microarrays, PLATO, and alternative technologies. In certain embodiments, MIPSA utilizes self-assembly to generate libraries of proteins linked to relatively short (e.g., 158 nt) single-stranded barcodes, for example, via the 25 kDa Halo tag domain. This compact barcoding approach is likely to find numerous applications unavailable to alternative display formats with bulky linked cargo (e.g., yeast, phage, ribosome, mRNA). Indeed, conjugating minimal DNA barcodes individually to proteins, particularly antibodies and antigens, has already proven useful in several contexts, including CITE-Seq, (25) LIBRA-seq, (26) and related methodologies (22, 27). On a proteome scale, MIPSA allows for unbiased analysis of protein-antibody, protein-protein, and protein-small molecule interactions, as well as studies of post-translational modifications, such as, for example, hapten modification studies or protease activity profiling. Key advantages of MIPSA include its high throughput, low cost, simple sequencing library preparation, and stability of protein-DNA complexes (important for both manipulating and storing display libraries). Importantly, MIPSA does not require specialized training or equipment, and may be readily adopted by low-complexity laboratories, as they need only simple access to high-throughput sequencing equipment or facilities.

[0024] Complementarity of MIPSA and PhIP-Seq. Display technologies often complement each other, but may not be routinely used in concert. MIPSA is more likely to detect antibodies to conformational epitopes on proteins that are well expressed in vitro than PhIP-Seq. This was exemplified by the robust detection of interferon-alpha autoantibodies via MIPSA described below that were not detected via PhIP-Seq. On the other hand, PhIP-Seq is more likely to detect antibodies to less abundant conformational epitopes contained within proteins that are either not present in the ORFeome library or may not be well expressed in bacterial lysates. Because MIPSA and PhIP-Seq naturally complement each other in these methods, we designed the MIPSA UCI amplification primers to be the same as those we used for PhIP-Seq. Because MIPSA and PhIP-Seq are stable in bacterial phage lysates, they can be easily performed together in a single reaction using a single set of amplification and sequencing primers. Thus, the natural compatibility of these two display modalities lowers the barrier to exploiting their synergies.

[0025] Variations of the MIPSA system. A key aspect of MIPSA involves the binding of a protein to its associated UCI in cis, compared to the binding of another library member's UCI in trans. Here, we utilized covalent binding via the Halo tag / Halo ligand system, but there are others that may work similarly. For example, the SNAP tag (a 20 kDa mutant of the DNA repair protein O6-alkylguanine-DNA alkyltransferase) forms a covalent bond with a benzylguanine (BG) derivative (28), so the RT primer can be labeled with BG instead of the Halo ligand. The CLIP tag, a mutant derivative of the SNAP tag, binds to an O2-benzylcytosine (BC) derivative, and this can also be adapted for MIPSA (29).

[0026] The rates of fusion tag maturation and ligand binding are important for the relative yield of cis vs. trans binding. According to a study by Samelson et al., the rate of production of Halo tag proteins is approximately four times faster than the rate of functional maturation of the Halo tag (30). Considering that typical protein sizes are less than 1,000 amino acids in ORFeome libraries, these data predict that most proteins will be released from the ribosome before Halo tag maturation, i.e., before cisHalo ligand binding can occur, thereby favoring undesirable trans barcoding. During optimization experiments, we found that the rate of cis barcoding was slightly improved by excluding a release factor from the translation mixture, which causes ribosomes to stall on their native ORF stop codons. That is, Halo tag maturation continues while remaining in close proximity to the cisHalo ligand-bound primer. Alternative approaches to promote controlled ribosome stalling may also include stop codon removal / suppression or the use of dominant-negative terminators. Ribosome release can then be achieved through the addition of the chain terminator puromycin.

[0027] Since UCI is formed on the 5'UTR of mRNA, eukaryotic ribosomes would not be able to scan from the 5'cap to the initiating Kozak sequence. If cap-dependent translation is required, two alternative methods can be used. First, the current 5'UCI system can be used if an internal ribosome entry site (IRES) is placed between the RT primer and the Kozak sequence. Second, it is possible to place the UCI at the 3'end of the mRNA, preventing the RT from extending into the ORF. If any of these approaches are developed beyond cell-free translation, the mRNA-cDNA hybrids can be transfected into living cells or tissues, where UCI protein formation can be performed in situ.

[0028] ORF-associated UCIs can be embodied in a variety of ways. In a particular embodiment, as described in the Examples section, we probabilistically assigned indexes to the human ORFeome with approximately 10-fold representation. This approach has two main advantages: first, the low cost of a synthetic oligonucleotide library (a single degenerate oligonucleotide pool), and second, the multiple independent pieces of evidence reported by the set of UCIs associated with each ORF. In a particular embodiment, the library of probabilistic barcodes is designed to feature sequences with uniform melting temperatures and therefore uniform PCR amplification efficiency. For simplicity, we chose not to incorporate unique molecular identifiers (UMIs) into the primers, but this approach is compatible with MIPSA UMIs and could potentially enhance quantification. One drawback of probabilistic indexing is the possibility of ORF dropout and thus the need for relatively high UCI representation, which increases the sequencing depth required to quantify each UCI and therefore the overall per-sample cost. A second drawback is the need to build a UCI-ORFeome matching dictionary. Using short-read sequencing, we were unable to clarify the fraction of the library that was mostly composed of alternative isoforms. Using long-read sequencing technologies such as PacBio or Oxford Nanopore Technologies instead of or in addition to short-read technologies can overcome the incomplete disambiguation during UCI-ORF matching. In contrast to probabilistic barcoding, individual ORF-UCI cloning is possible but costly and laborious. However, smaller UCI sets offer the advantage of lower sequencing costs per assay. We previously developed a method to clone the ORFeome using Long Adapter Single Stranded Oligonucleotide (LASSO) probes (31).Incorporating target-specific indexes into capture probe libraries yields uniquely indexed ORFs without dramatically increasing the cost of LASSO probe libraries. Thus, LASSO cloning of ORFeome libraries may synergize with MIPSA-based applications.

[0029] MIPSA readout by qPCR. A useful feature of properly designed UCIs is that they can also function as qPCR readout probes. The degenerate UCIs we designed and used herein (Figure 1B) also contain 18 nt Tm-balanced forward and reverse primer binding sites. Thus, the low cost and rapid turnaround time of qPCR assays can be exploited in combination with MIPSA. For example, incorporating assay quality control measures such as TRIM21 IP can be used to qualify a set of samples prior to the more costly sequencing run. Troubleshooting and optimization can similarly be facilitated by using qPCR rather than NGS as a readout. qPCR testing of specific UCIs could also, in theory, offer enhanced sensitivity compared to sequencing and may be more suitable for analysis in a clinical setting. I. Definition

[0030] As used herein, the term "amino acid" refers to an organic compound that contains an amine group, a carboxylic acid group, and a side chain specific to each amino acid, which serves as a monomer subunit of peptides. Amino acids include the 20 standard amino acids, natural or standard amino acids, and non-standard amino acids. Standard natural amino acids include alanine (A or Ala), cysteine ​​(C or Cys), aspartic acid (D or Asp), glutamic acid (E or Glu), phenylalanine (F or Phe), glycine (G or Gly), histidine (H or His), isoleucine (I or Ile), lysine (K or Lys), leucine (L or Leu), methionine (M or Met), asparagine (N or Asn), proline (P or Pro), glutamine (Q or Gln), arginine (R or Arg), serine (S or Ser), threonine (T or Thr), valine (V or Val), tryptophan (W or Trp), and tyrosine (Y or Tyr). Amino acids may be L- or D-amino acids. Non-standard amino acids may be modified amino acids, amino acid analogs, amino acid mimetics, non-standard proteinogenic amino acids, or naturally occurring or chemically synthesized non-proteinogenic amino acids. Examples of non-standard amino acids include, but are not limited to, selenocysteine, pyrrolysine, and N-formylmethionine, β-amino acids, homoamino acids, proline and pyruvic acid derivatives, 3-substituted alanine derivatives, glycine derivatives, ring-substituted phenylalanine and tyrosine derivatives, linear core amino acids, and N-methyl amino acids.

[0031] As used herein, the term "polypeptide" encompasses peptides and proteins and refers to a molecule comprising a chain of two or more amino acids linked by peptide bonds. In some embodiments, a polypeptide comprises 2-50 amino acids, e.g., more than 20-30 amino acids. In some embodiments, a peptide does not comprise secondary structure, domain structure, or higher order structure. In some embodiments, a protein comprises 30 or more amino acids, e.g., more than 50 amino acids. In some embodiments, a protein comprises secondary structure, domain structure, or higher order structure in addition to the primary structure. The amino acids of a polypeptide are most typically L-amino acids, but may be D-amino acids, unnatural amino acids, modified amino acids, amino acid analogs, amino acid mimetics, or any combination thereof. Polypeptides may be naturally occurring, synthetically produced, or recombinantly expressed. Polypeptides may also comprise additional groups that modify the amino acid chain, e.g., functional groups added via post-translational modifications. A polymer may be linear or branched, may comprise modified amino acids, and may be interrupted by non-amino acids. The term also embraces amino acid polymers that are modified naturally or by intervention; for example, disulfide bond formation, glycosylation, lipidation, acetylation, phosphorylation, or any other manipulation or modification, such as conjugation with a labeling component.

[0032] As used herein, the term "proteome" may include the entire set of proteins, polypeptides, or peptides (including their conjugates or complexes) expressed by any biological target, such as a genome, cell, tissue, or organism, at a particular time. In one aspect, it is the set of proteins expressed in a given type of cell or organism at a given time under defined conditions. Proteomics is the study of the proteome. For example, a "cellular proteome" may include the collection of proteins found in a particular cell type under a particular set of environmental conditions, such as exposure to a hormone stimulus. The complete proteome of an organism may include the complete set of proteins from all of the various cellular proteomes. A proteome may also include the collection of proteins in a particular intracellular biological system. For example, all the proteins in a virus may be referred to as a viral proteome. As used herein, the term "proteome" includes subsets of the proteome, including, but not limited to, the kinome; the secretome; the receptosome (e.g., the GPCRome); the immunoproteome; the nutriproteome; proteome subsets defined by post-translational modifications (e.g., phosphorylation, ubiquitination, methylation, acetylation, glycosylation, oxidation, lipidation, and / or nitrosylation), such as the phosphoproteome (e.g., the phosphotyrosine-proteome, the tyrosine-kinome, and the tyrosine-phosphatome), the glycoproteome, etc.; proteome subsets associated with a tissue or organ, a developmental stage, or a physiological or pathological state; proteome subsets associated with a cellular process, such as the cell cycle, differentiation (or dedifferentiation), cell death, senescence, cell migration, transformation, or metastasis; or any combination thereof.

[0033] As used herein, the term "nucleic acid molecule" or "polynucleotide" refers to single- or double-stranded polynucleotides containing deoxyribonucleotides or ribonucleotides linked by 3'-5' phosphodiester bonds, as well as polynucleotide analogs. Nucleic acid molecules include, but are not limited to, DNA, RNA, and cDNA. Polynucleotide analogs may have backbones other than the standard phosphodiester bonds found in natural polynucleotides, and optionally modified sugar moieties or moieties other than ribose or deoxyribose. Polynucleotide analogs contain bases that can hydrogen bond to standard polynucleotide bases by Watson-Crick base pairing, where the analog backbone presents the bases in a manner that allows such hydrogen bonding in a sequence-specific manner between the oligonucleotide analog molecule and the bases in the standard polynucleotide.

[0034] As used herein, the term "barcode" refers to a sequence of about 2 to about 10 bases (e.g., 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, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, Barcode refers to a nucleic acid molecule of 35, 36, 37, 38, 39, 40, 41, 42, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, or 100 bases. Barcodes can be artificial or naturally occurring sequences. The concept of barcodes is that, prior to any amplification, each original target molecule is "tagged" with a unique barcode sequence. In some embodiments, the DNA sequence must be long enough to provide sufficient permutations to assign a unique barcode to each founder molecule.

[0035] As used herein, the term "universal priming site" or "universal primer" or "universal priming sequence" refers to a nucleic acid molecule that can be used for library amplification and / or sequencing reactions. Universal priming sites can include, but are not limited to, priming sites (primer sequences) for PCR amplification, flow cell adapter sequences that anneal to complementary oligonucleotides on the flow cell surface, sequencing priming sites that allow bridge amplification in some next-generation sequencing platforms, or combinations thereof. The term "forward" when used in the context of "universal priming site" or "universal primer" can also be referred to as "5'" or "sense". The term "reverse" when used in the context of "universal priming site" or "universal primer" can also be referred to as "3'" or "antisense".

[0036] As used herein, "next-generation sequencing" refers to a high-throughput sequencing method that allows the sequencing of millions to billions of molecules in parallel. Examples of next-generation sequencing methods include sequencing by synthesis, sequencing by ligation, sequencing by hybridization, polony sequencing, ion semiconductor sequencing, and pyrosequencing. Nucleic acid molecules can be hybridized to solid substrates via primers by attaching primers to solid substrates and attaching complementary sequences to nucleic acid molecules, and then can be amplified with polymerase to generate multiple copies in separate regions on solid substrates (these groupings are sometimes called polymerase colonies or polony). As a result, in the sequencing process, nucleotides at specific positions can be sequenced multiple times (e.g., hundreds or thousands of times). Examples of high-throughput nucleic acid sequencing technologies include platforms offered by Illumina, BGI, Qiagen, Thermo-Fisher, and Roche, and include formats such as parallel bead arrays, sequencing-by-synthesis, sequencing-by-ligation, capillary electrophoresis, electronic microchips, "biochips," microarrays, parallel microchips, and single molecule arrays.

[0037] The terms "specifically binds to," "specific for," and related grammatical variations refer to binding that occurs between paired species such as ligand / tag, antibody / antigen, aptamer / target, enzyme / substrate, receptor / agonist, and lectin / carbohydrate, which may be mediated by covalent or non-covalent interactions or a combination of covalent and non-covalent interactions. When the interaction of the two species produces a non-covalent complex, the resulting binding is typically the result of electrostatic, hydrogen, or lipophilic interactions. Thus, in certain embodiments, "specific binding" occurs between paired species where there is an interaction between the two that produces a binding complex having the characteristics of, for example, an antibody / antigen or enzyme / substrate interaction. In particular, specific binding is characterized by one member of the pair binding to a particular species and not binding to other species within the compound family to which the corresponding member of the binding member belongs. Thus, for example, an antibody typically binds to a single epitope and not other epitopes within the protein family. In certain embodiments, specific binding between an antigen and an antibody is characterized by a specific binding affinity of at least 10 -6 In other embodiments, the antigen and antibody have a binding affinity of at least 10 -7 M, 10 -8 M~10 -9 M, 10 -10 M, 10 -11 M, or 10 -12 It binds with an affinity of M. In certain embodiments, this term refers to a molecule (e.g., an aptamer) that binds to a target (e.g., a protein) with an affinity that is at least 5 times greater than any non-target, for example, at least 10, 20, 50, or 100 times greater. In certain embodiments, the polypeptide tag specifically binds to its ligand. In certain embodiments, the polypeptide tag is covalently bound to the ligand.

[0038] A "biological sample", as used herein, is generally a sample from an individual or subject. Non-limiting examples of biological samples include blood, serum, plasma, or cerebrospinal fluid. In addition, solid tissues (e.g., spinal cord or brain biopsies) can be used. II. VECTORS, LIBRARIES THEREOF AND METHODS OF USING SAME

[0039] The present disclosure provides a self-assembly protein display library comprising vector and a plurality of vectors.In certain embodiments, the vector comprises a nucleic acid sequence encoding a protein of interest.In one embodiment, the vector comprises, along the 5' to 3' direction, (a) a polymerase transcription initiation site; (b) a barcode; (c) a reverse transcription primer binding site; (d) an RBS; and (e) a nucleotide sequence encoding a fusion protein comprising (i) a polypeptide tag and (ii) a protein of interest, wherein the polypeptide tag specifically binds to a ligand.

[0040] In certain embodiments, the vector further comprises an endonuclease site for vector linearization, hi other embodiments, the vector further comprises (vii) a stop codon.

[0041] In certain embodiments, the barcode is adjacent to a polymerase chain reaction (PCR) primer binding site, hi another embodiment, the barcode comprises a PCR primer binding site.

[0042] In another embodiment, the RBS comprises an internal ribosome entry site.

[0043] In certain embodiments, each barcode in the population of barcodes is different. In other embodiments, a portion of the barcodes in the population of barcodes are different, for example, at least about 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 97%, or 99% of the barcodes in the population of barcodes are different.

[0044] The population of barcodes may be randomly or non-randomly generated. In some embodiments, the barcodes contain randomized nucleotides and are incorporated into the nucleic acid. For example, a 12-base random sequence is generated 4 times for each target molecule in a sample. 12 Or provide 16,777,216 UMIs.

[0045] In certain embodiments, barcodes can be used to computationally deconvolute multiplexed sequencing data to identify sequences originating from individual macromolecules, samples, libraries, etc.

[0046] The present disclosure also provides a method for using self-assembling protein display library.In certain embodiments, the method comprises: (a) transcribing a plurality of linearized or nicked vectors comprising self-assembling protein display library to produce mRNA; (b) reverse transcribing the 5' end of mRNA using a primer conjugated to a ligand to produce a cDNA comprising a barcode; and (c) translating the mRNA, where the polypeptide tag of the fusion protein is covalently linked to the ligand conjugated to the cDNA comprising the barcode.

[0047] A more specific embodiment is a method comprising the steps of: (a) transcribing a vector library into messenger ribonucleic acid (mRNA), the vector library encoding a plurality of proteins, each vector of the vector library being comprised in a 5' to 3' direction of: (i) a polymerase transcription initiation site; (ii) a barcode; (iii) a reverse transcription primer binding site; (iv) a ribosome binding site (RBS); and (v) a nucleotide sequence encoding a fusion protein comprising (1) a polypeptide tag and (2) a protein, wherein the polypeptide tag specifically binds to a ligand; (b) reverse transcribing the 5' end of the mRNA using a primer that binds upstream of the RBS, the primer being conjugated to the ligand that specifically binds to the polypeptide tag of the fusion protein, forming a complementary deoxyribonucleic acid (cDNA) comprising the ligand, primer and barcode; and (c) translating the mRNA, wherein the ligand of the cDNA binds to the polypeptide tag of the fusion protein. In certain embodiments, the vector library is nicked prior to step (a). In other certain embodiments, the vector further comprises (vi) an endonuclease site for vector linearization, and the vector library is linearized prior to step (a). III. Self-assembling protein-DNA conjugates, libraries thereof and methods of using same

[0048] The present disclosure also provides self-assembling protein-DNA conjugate compositions and libraries comprising the same. In certain embodiments, each protein-DNA conjugate comprises: (a) a cDNA comprising a barcode, the cDNA being conjugated to a ligand that specifically binds to a polypeptide tag; and (b) a fusion protein comprising the polypeptide tag and a protein of interest, the ligand being covalently bound to the polypeptide tag.

[0049] In certain embodiments, two or more copies of a protein of interest can be present as protein-DNA conjugates in a library of protein-DNA conjugates, and each copy of the protein of interest can include a unique barcode.

[0050] In certain embodiments, the polypeptide tag is fused to the N-terminus of the protein of interest, hi other embodiments, the polypeptide tag is fused to the C-terminus of the protein of interest.

[0051] In certain embodiments, the polypeptide tag is a haloalkane dehalogenase or 6 -alkylguanine-DNA-alkyltransferase. In certain embodiments, the polypeptide tag comprises a HALO tag and the ligand comprises a HALO ligand. In more specific embodiments, the HALO tag comprises the amino acid sequence set forth in SEQ ID NO: 22. In other embodiments, the HALO-ligand comprises any one of the following: [ka]

[0052] HALOTAG® tags and ligands are commercially available from Promega (Madison, WI) and are conjugated to nucleic acids according to the manufacturer's instructions. In certain embodiments, to conjugate HALOTAG® ligands to DNA sequences (e.g., reverse transcription primers), the DNA sequences are modified with alkyne groups. The azidohalo ligands are then reacted with alkyne-terminated DNA sequences using Cu-catalyzed cycloaddition ("click" chemistry). See, for example, Duckworth et al. 46 ANGEW CHEM. INT. 8819-22 (2007).

[0053] Alternatively, other polypeptide tag-ligand capture subsystems can be used. For example, O6-alkylguanine-DNA alkyltransferase reacts specifically and rapidly with benzylguanine (BG) and its derivatives. In certain embodiments, the polypeptide tag comprises SNAP-TAG® (New England Biolabs, Ipwich, Mass.). SNAP-TAG® is a ligand for the human O 6 SNAP-TAG® is a self-labeling protein derived from O-alkylguanine-DNA-alkyltransferase. 6 -benzylguanine derivatives. In one embodiment, the polypeptide tag comprises the amino acid sequence set forth in SEQ ID NO: 23. In another specific embodiment, the polypeptide tag comprises CLIP-TAG (New England Biolabs), a modified version of SNAP-TAG®, which also binds to human O 6 -alkylguanine-DNA-alkyltransferase is a self-labeling protein derived from. Instead of benzylguanine derivatives, the CLIP tag is engineered to react with benzylcytosine derivatives. In certain embodiments, the polypeptide tag comprises the amino acid sequence set forth in SEQ ID NO: 24. See Keppler et al. 1 NAT BIOTECHNOL. 86-99 (2003); and Gautier et al. 15(2) CHEM. BIOL. 128-36 (2008).

[0054] The present disclosure also provides a method for using a library of self-assembled protein-DNA conjugates. In one embodiment, the method for studying protein-protein interactions comprises performing a pull-down assay of a library of protein-DNA conjugates with a protein of interest. In another embodiment, the method for studying protein-small molecule interactions comprises performing a pull-down assay of a library of protein-DNA conjugates with a small molecule. In yet another embodiment, the method comprises performing immunoprecipitation of a library of protein-DNA conjugates with an antibody obtained from a biological sample. In a further embodiment, the method for identifying a target of a first small molecule comprises (a) incubating a library of protein-DNA conjugates with a first small molecule that binds to the target, and (b) performing a pull-down assay of the library of step (a) with a second small molecule, where the first small molecule bound to the target blocks the binding of the second small molecule. In a more specific embodiment, two or more small molecules are used in the pull-down assay of step (b). IV. Treatment of COVID-19

[0055] The present disclosure also provides a method for treating COVID-19. In one embodiment, a method for treating a patient with severe COVID-19 comprises administering an effective amount of interferon therapy to the patient, and an autoantibody that neutralizes IFN-λ3 is detected in a biological sample obtained from the patient. In another embodiment, a method for treating a patient with severe COVID-19 comprises (a) detecting an autoantibody that neutralizes IFN-λ3 in a biological sample obtained from the patient, and (b) treating the patient with an effective amount of interferon therapy. In a further embodiment, a method for identifying a COVID-19 patient who would benefit from interferon therapy comprises detecting an autoantibody that neutralizes IFN-λ3 in a biological sample obtained from the patient. In certain embodiments, the interferon therapy comprises interferon lambda (IFN-λ) or interferon beta (IFN-β). In certain embodiments, the interferon lambda (IFN-λ) or interferon beta (IFN-β) is pegylated. In a further embodiment, the interferon therapy comprises interferon omega (IFN-ω).

[0056] The terms "interferon," "IFN," and "interferon molecule" are used interchangeably herein. They refer to any interferon or interferon derivative (e.g., pegylated interferon) that can be used to treat COVID-19.

[0057] Interferon is a family of cytokines produced by eukaryotic cells in response to viral infection and other antigenic stimuli, and exhibits broad-spectrum antiviral, antiproliferative and immunomodulatory effects. Recombinant forms of interferon have been widely applied in the treatment of various conditions and diseases, such as viral infections (e.g., HCV, HBV and HIV), inflammatory disorders and diseases (e.g., multiple sclerosis, arthritis, cystic fibrosis), and tumors (e.g., liver cancer, lymphoma, myeloma, etc.).

[0058] Interferons are classified into types I, II and III, depending on the cellular receptors to which they bind. Type I interferons bind to a specific cell surface receptor complex known as the IFN-alpha (IFN-α) receptor (IFNAR), which consists of two chains (IFNAR1 and IFNAR2). The type I interferons present in humans are interferon-alpha (IFN-α), interferon-beta (IFN-β) and interferon-omega (IFN-ω).

[0059] Type III interferons signal through a receptor complex consisting of the interferon-lambda receptor (IFNLR1 or CRF2-12) and interleukin 10 receptor 2 (IL10R2 or CRF2-4). In humans, type III interferons comprise three interferon lambda (IFN-λ) proteins, termed IFN-λ1, IFN-λ2 and IFN-λ3, also known as interleukin 29 (IL-29), interleukin 28A (IL-28A) and interleukin 28B (IL-28B), respectively.

[0060] Thus, in certain embodiments, the interferon therapy includes one or more of IFN-α, IFN-β, IFN-ω, IFN-γ, IFN-λ, analogs thereof, and derivatives thereof. In certain embodiments, the interferon therapy includes IFN-λ, analogs thereof, and derivatives thereof. In other embodiments, the interferon therapy includes IFN-β, analogs thereof, and derivatives thereof.

[0061] As used herein, the terms "interferon," "IFN," and "IFN molecule" more specifically refer to a peptide or protein having amino acids substantially identical (e.g., at least 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99% or even 100% identical) to all or a portion of the sequence of an interferon (e.g., a human interferon) (e.g., IFN-α, IFN-β, IFN-ω, IFN-γ, and IFN-λ) known in the art. Interferons suitable for use in the present disclosure include, but are not limited to, natural human interferons produced using human cells, recombinant human interferons produced from mammalian cells, E. coli-produced recombinant human interferons, synthetic versions of human interferons, and equivalents thereof. Other suitable interferons include consensus interferons, which are a type of synthetic interferon that has an amino acid sequence that is approximately the average of the sequences of all known human IFN subtypes (e.g., all known IFN-β subtypes, or all known IFN-λ subtypes).

[0062] The terms "interferon", "IFN", and "IFN molecule" also include interferon derivatives, i.e., modified or transformed molecules of interferon (as described above). A suitable transformation can be any modification that confers desired properties to the interferon molecule. Examples of desirable properties include, but are not limited to, extended in vivo half-life, improved therapeutic efficacy, reduced frequency of administration, increased solubility / water solubility, increased resistance to proteolysis, facilitated controlled release, and the like. As described above, pegylated interferons have been produced (e.g., pegylated IFN-λ) and are currently used to treat hepatitis. Pegylated interferons exhibit longer half-lives, which allows for less frequent administration of the drug. Pegylation of an interferon molecule involves covalently attaching interferon to polyethylene glycol (PEG), an inert, non-toxic and biodegradable organic polymer. Thus, in certain embodiments, interferon therapy includes pegylated interferon. Interferon has also been produced as a fusion protein with human albumin (e.g., albumin-IFN-λ). The albumin fusion platform takes advantage of the long half-life of human albumin to provide a therapy that can reduce the frequency of IFN administration. Thus, in certain embodiments, interferon therapy includes an albumin-interferon fusion protein.

[0063] The present disclosure provides a method for detecting autoantibodies against IFN-λ3. In a more specific embodiment, autoantibodies that neutralize IFN-λ3 are detected. The presence of autoantibodies that neutralize IFN-λ3 can be used to identify COVID-19 patients who will benefit from interferon therapy. In a specific embodiment, the patient has severe COVID-10. Interferon therapy can be administered to COVID-19 patients in which autoantibodies that neutralize IFN-λ3 have been detected in biological samples obtained from the patient.

[0064] IFN-λ3 polypeptide can be used in immunoassay to detect IFN-λ3 specific autoantibody in biological sample. IFNλ-3 polypeptide used in immunoassay can be present in cell lysate (e.g., whole cell lysate or cell fraction), or purified IFNλ-3 polypeptide or its fragment can be used, provided that at least one antigen site recognized by IFNλ-3 specific autoantibody remains available for binding. Depending on the nature of the sample, either or both of immunoassay and immunocytochemical staining techniques can be used. Enzyme-linked immunosorbent assay (ELISA), Western blot, and radioimmunoassay can be used as described herein to detect the presence of IFNλ-3 specific autoantibody in biological sample.

[0065] IFNλ-3 polypeptide or its fragment can be used with or without modification for detecting IFNλ-3 specific autoantibody.Polypeptide can be labeled by covalently or non-covalently binding the polypeptide with a second substance that provides detectable signal.A wide variety of labeling and conjugation techniques can be used.Some examples of labels that can be used include radioisotopes, enzymes, substrates, cofactors, inhibitors, fluorescent agents, chemiluminescent agents, magnetic particles, etc.

[0066] Without further elaboration, it is believed that one skilled in the art can, using the preceding description, utilize the present invention to its fullest extent. The following examples are illustrative only, and not limiting of the remainder of the disclosure in any way. EXAMPLES

[0067] The following examples are presented to provide those skilled in the art with a complete disclosure and description of how the compounds, compositions, articles, devices, and / or methods described and claimed herein are made and evaluated, are intended to be purely illustrative, and are not intended to limit the scope of what the inventors regard as their disclosure. Efforts have been made to ensure accuracy with respect to numbers (e.g., amounts, temperatures, etc.), but some errors and deviations should be accounted for herein. Unless otherwise indicated, parts are parts by weight, temperatures are in degrees Celsius or at ambient temperature, and pressures are at or near atmospheric pressure. There are numerous variations and combinations of reaction conditions, such as component concentrations, desired solvents, solvent mixtures, temperatures, pressures, and other reaction ranges and conditions that can be used to optimize product purity and yields obtained from the described processes. Only reasonable and routine experimentation is required to optimize such process conditions.

[0068] Example 1: Molecular Indexing of Proteins by Self-Assembly (MIPSA) for efficient proteomic investigations. material and method

[0069] MIPSA destination vector construction and UCI barcode library construction. MIPSA vectors were constructed using pDEST15 vector as backbone. A gBlock fragment (Integrated DNA Technologies) encoding the RBS, Kozak sequence, N-terminal Halo tag fusion protein, FLAG tag, and attR1 sequence was cloned into the parent plasmid. A 150 bp poly(A) sequence was also added after attR2 and the stop codon. A 41 nt barcode oligo was generated within the gBlock Gene Fragment (Integrated DNA Technologies) with alternating mixed bases (S:G / C;W:A / T) and the following sequence: (SW) 18 -AGGGA-(SW) 18The degenerate barcode was generated using standard PhIP-Seq PCR1 and PCR2 primer binding sites (43). 18 ng of the starting UCI library was used for 40 cycles of PCR to amplify the library and incorporate BglII and PspxI restriction sites. The MIPSA vector and amplified UCI library were then digested overnight with restriction enzymes, column purified, and ligated at a vector-to-insert ratio of 1:5. The ligated MIPSA vector was used to transfect electrocompetent One Shot ccdB 2 T1 R Cells (Thermo Fisher Scientific) were transformed, and approximately 800,000 colonies were obtained from six transformation reactions to generate the pDEST-MIPSA UCI library.

[0070] Human ORFeome recombination into barcoded MIPSA vector. 150ng of pENTR-hORFeome-(L1-L5) vector was combined with 150ng of pDEST-MIPSA vector and 2μL of Gateway LR Clonase II mix (Life Technologies) in a total reaction volume of 10uL. The reaction was incubated overnight at 25°C. The entire reaction was incubated in 50μL of One Shot OmniMAX 2 T1 R The plasmids were transformed into chemically competent E. coli (Life Technologies). The transformation yielded approximately 120,000 colonies, which is approximately 10 times the size of each human subpool library. The colonies were collected and pooled by scraping, and the barcoded pDEST-MIPSA-hsORFeome plasmid MIPSA (human ORFeome DNA library) was purified using the Qiagen Plasmid Midi Kit (Qiagen).

[0071] Halo Ligand Conjugation to RT Oligos and HPLC Purification. 100 μg of 5′ amine modified oligos (Table 1) were incubated with 75 μL (17.85 μg / μL) of succinimidyl ester (O2) Halo Ligand (Promega Corporation) in 0.1 M sodium borate buffer according to Gu et al. (14) for 6 h at room temperature. 3 M NaCl and ice-cold ethanol were added to the labeling reaction at 10% (v / v) and 250% (v / v), respectively, and incubated overnight at −80°C. The reaction was centrifuged at 12,000 × g for 30 min. The pellet was rinsed once in ice-cold 70% ethanol and air-dried for 10 min.

[0072] Halo-liganded RT primers were HPLC-purified using a Brownlee Aquapore RP-300 7u, 100x4.6mm column (Perkin Elmer) using a 2-buffer gradient of 0-70% CH3CN / MeCN (100mM trimethylamine acetate in acetonitrile) over 70 minutes. Fractions corresponding to the labeled oligos were collected and lyophilized (Figure 6). Oligos were resuspended at 1μM and stored at -80°C.

[0073] MIPSA RNA library preparation. The pDEST-MIPSA vector containing the human ORFeome library (4 μg) was linearized overnight with I-SceI restriction endonuclease (New England Biolabs). The product was column purified using NucleoSpin Gel and PCR Clean Up kit (Macherey-Nagel GmbH & Co. KG). 1 μg of purified linearized product was transcribed utilizing 40 μL of HiScribe T7 High Yield RNA Synthesis Kit (New England Biolabs). The product was diluted with 60 μL of molecular biology grade water and 1 μL of DNAseI was added. The reaction was incubated at 37°C for an additional 15 minutes. Then, 50 μL of 1M LiCl was added to the solution and incubated at -80°C overnight. The centrifuge was cooled to 4°C and the RNA was spun at maximum speed for 30 minutes. The supernatant was removed and the RNA pellet was washed with 70% ethanol. The samples were spun down for an additional 10 min at 4°C and the 70% ethanol was removed. The pellet was dried at room temperature for 15 min and then resuspended in 100 μL of water. To preserve the samples, 1 μL of 40 U / μL RNAseOUT Recombinant Ribonuclease Inhibitor (Life Technologies, Carlsbad CA) was added.

[0074] MIPSA RNA Library Reverse Transcription and Translation. Reverse transcription reactions were prepared using the SuperScript IV First-Strand Synthesis System (Life Technologies). First, 1 μL of 10 mM dNTPs, 1 μL of RNAseOUT (40 U / μL), 4.17 μL of RNA library (1.5 μM), and 7.83 μL of Halo ligand-conjugated RT primer (1 μM, Table 1) were combined into a single 14 μL reaction and incubated at 65 °C for 5 min, followed by incubation on ice for 2 min. 4 μL of 5x RT buffer, 1 μL of 0.1 M DTT, and 1 μL of SuperScript IV RT enzyme (200 U / μL) were added to the 14 μL reaction on ice and incubated at 42 °C for 20 min. 36 μL of RNAClean XP beads (Beckman Coulter) were added to the single 20 μL RT reaction and incubated at room temperature for 10 min. Beads were collected with a magnet and washed five times with 70% ethanol. Beads were air-dried at room temperature for 10 min and resuspended in 7 μL of 5 mM Tris-HCl, pH 8.5. The product (2 μL) was analyzed spectrophotometrically to measure RNA yield. Translation reactions were set up on ice using the PURExpress ΔRibosome Kit (New England Biolabs) (44). Reactions were modified to a final concentration of ribosomes of 0.3 μM. 4.57 μL of RT reaction was added to 4 μL of solution A, 1.2 μL of factor mix, and 0.23 μL of ribosomes (13.3 μM). The reaction was incubated at 37 °C for 2 h, diluted with 35 μL of 1× PBS to a total volume of 45 μL, and used immediately or stored at −80 °C after the addition of 25% glycerol. In optimization experiments utilizing the PURExpressΔRF123 Kit (New England Biolabs), solution B was replaced with NEB custom-made Factor Mix (-RF123, -ribosome). After a 2-h incubation step at 37°C, RNase A was added or release factors 1, 2, and 3 were added and the reaction was allowed to proceed for 30 min on ice.

[0075] Immunoprecipitation with MIPSA library. 5 μL of serum is mixed with 45 μL of diluted MIPSA library (see above) and incubated overnight at 4 °C with gentle agitation. For each IP, a mixture of 5 μL of Protein A Dynabeads and 5 μL of Protein G Dynabeads (Life Technologies) was washed three times with 1 × PBS in twice their original volume. The beads were then resuspended in 1 × PBS in the original volume and added to each IP. Binding was allowed to proceed for 4 hours at 4 °C. The beads were collected on a magnet and the beads were washed three times with 1 × PBS, changing tubes or plates between washes. The beads were then collected and resuspended in 20 μL of PCR master mix containing T7-Pep 2 PCR1 F forward primer and T7-Peps PCR1 R + ad min reverse primer (Table 1) and Herculase-II (Agilent). The PCR cycle was as follows: an initial denaturation step at 95°C for 2 min, followed by 30 cycles of 95°C for 20 s, 58°C for 30 s, 72°C for 30 s, and a final extension at 72°C for 3 min. Two microliters of the amplified product was used as input to a 20 μL dual index PCR reaction with PhIP PCR2 F forward primer and Ad min BCX P7 reverse primer for 10 cycles. The PCR cycle was as follows: an initial denaturation step at 95°C for 2 min, followed by 10 cycles of 95°C for 20 s, 58°C for 30 s, 72°C for 30 s, and a final extension at 72°C for 3 min. The i5 / i7 indexed libraries were pooled and column purified. The libraries were sequenced on an Illumina NextSeq 500 using the 1x75nt protocol. Plato2_i5_NextSeq_SP and Standard_i7_SP primers were used for i5 / i7 identification (Table 1). The outputs were demultiplexed using i5 and i7 without allowing any mismatch.

[0076] Phage immunoprecipitation sequencing. The design and cloning of a 90 amino acid human peptidome library was described previously (24). Phage immunoprecipitation and sequencing were performed according to our published protocol (45). Briefly, 0.2 μl of each plasma was individually mixed with the human phage library and then immunoprecipitated using protein A and protein G coated magnetic beads. A set of eight mock IPs was performed on each 96-well plate. Amplicons were sequenced on an Illumina NextSeq 500 instrument.

[0077] For quantification of MIPSA experiments by qPCR, PCR1 products were analyzed as follows. 4.6 μL of a 1 / 1000 dilution of PCR1 reaction was resuspended in 10 μL of qPCR master mix containing 5 μL of Brilliant III Ultra Fast 2X SYBR Green Mix (Agilent), 0.2 μL of 2 μM reference dye and 0.2 μL of 10 μM forward and reverse primer mix (specific for the target UCI). The PCR cycle was as follows: an initial denaturation step at 95°C for 2 min, followed by 30 cycles of 95°C for 20 s and 60°C for 30 s, 45 cycles. After completion of thermocycling, the amplified products were subjected to melting curve analysis. The qPCR primers for MIPSA immunoprecipitation experiments are as follows: BT2_F and BT2_R for TRIM21, BG4_F and BG4_R for GAPDH, and NT5C1A_F and NT5C1A_R for NT5C1A (Table 1).

[0078] Plasma Samples. All samples were collected by studies in which subjects met protocol eligibility criteria, as described below. All of the studies protected the rights and privacy of study participants and were approved by their respective institutional review boards for the original sample collection and subsequent analysis.

[0079] Pre-pandemic plasma samples. All human samples were collected prior to 2017 following NIAID IRB-approved procedures at the National Institutes of Health (NIH), Vaccine Research Center (VRC) / National Institutes of Allergy and Infectious Diseases (NIAID) / NIH protocol “VRC 000: Screening Subjects for HIV Vaccine Research Studies” (NCT00031304).

[0080] COVID-19 convalescent plasma (CCP) from non-hospitalized patients. Eligible CCP donors were contacted by study personnel as previously described (46,47). All donors were at least 18 years of age and had a confirmed diagnosis of SARS-CoV-2 by detection of RNA in a nasopharyngeal swab sample. Basic demographic information (age, sex, hospitalization with COVID-19) was obtained from each donor; first diagnosis and date of diagnosis of SARS-CoV-2 were confirmed by medical record review. Samples were separated into plasma and peripheral blood mononuclear cells within 12 hours of collection, and plasma samples were immediately frozen at -80°C.

[0081] Severe COVID-19 plasma samples. The study cohort was comprised of hospitalized patients who met the following criteria: 1) confirmed diagnosis of COVID-19; 2) death or survival to discharge; and 3) had remnant specimens in the Johns Hopkins COVID-19 Remnant Specimen Biorepository (opportunity samples including 59% of Johns Hopkins COVID-19 patients and 66% of patients with hospital stay ≥3 days) (48). Patient outcomes were defined by the World Health Organization (WHO) COVID-19 disease severity scale. Samples from severe COVID-19 patients included in this study were obtained from 17 patients who died, 13 patients who recovered after mechanical ventilation, 22 patients who required supplemental oxygen to recover, and 3 patients who recovered without supplemental oxygen. This study was approved by the JHU Institutional Review Board (IRB00248332, IRB00273516) and a waiver of consent was granted as all specimens and clinical data were de-identified by the Core for Clinical Research Data Acquisition at the Johns Hopkins Institute for Clinical and Translational Research; the study team had no access to identifiable patient data.

[0082] Sjögren's syndrome and inclusion body myositis (IBM) plasma samples. Sjögren's syndrome samples were collected under protocol NA_00013201. All patients were >18 years old and gave informed consent. IBM patient samples were collected under protocol IRB00235256. All patients fulfilled the ENMC 2011 diagnostic criteria (49) and gave informed consent.

[0083] Immunoblot analysis. Laemmli buffer containing 5% β-ME was added to post-translation samples, boiled for 5 min, and analyzed on NuPAGE 4–12% Bis-Tris polyacrylamide gels (Life Technologies). After transfer to PVDF membranes, blots were blocked for >1 h at room temperature in 20 mM Tris-buffered saline (pH 7.6) containing 0.1% Tween 20 (TBST) and 5% (wt / vol) nonfat dry milk. Blots were subsequently incubated with primary antibodies overnight at 4 °C, followed by incubation in secondary antibodies for 4 h at room temperature.

[0084] Construction of the UCI-ORF dictionary. The Nextera XT DNA Library Preparation kit (Illumina) was used to tagment 150 ng of each library to obtain an optimal size distribution centered around 1.5 kb. The tagmented MIPSA human ORFeome libraries were amplified using Herculase-II (Agilent) with T7-Pep2 PCR1 F forward and Nextera Index 1 Read primers. The PCR cycle was as follows: an initial denaturation step at 95°C for 2 min, followed by 30 cycles of 95°C for 20 s, 53.5°C for 30 s, 72°C for 30 s, and a final extension at 72°C for 3 min. The PCR reactions were run on a 1% agarose gel, followed by excision of the approximately 1.5 kb product and purification using NucleoSpin Gel&PCR Clean-up columns (Mackery Nagel). The purified products were then amplified for an additional 10 cycles using PhIP PCR2 F forward primer and P7.2 reverse primer (see Table 1 for a list of primer sequences). Products were gel purified and sequenced on a MiSeq (Illumina) using T7-Pep2.2 SP subA primer for read 1 and MISEQPLATOR2 primer for read 2. Read 1 was 60 bp long to capture UCI. The first index read I1 was replaced with a 50 bp read to the ORF. I2 was used to identify the i5 index for sample demultiplexing.

[0085] Human ORFeome V8.1 DNA sequences were truncated to the first 50 nt and ORF names corresponding to non-unique sequences were concatenated. The demultiplexed output of 50 nt R2 (ORF) reads from Illumina MiSeq was aligned to the truncated Human ORFeome V8.1 library using the Rbowtie2 package (50) with the following parameters: options = "-a --very-sensitive-local". The unique FASTQ identifiers were then used to extract the corresponding sequences from the 60 bp R1 (UCI) reads. These sequences were then truncated using the 3' anchor ACGATA and sequences without the anchor were removed. Additionally, any truncated R1 sequences with less than 18 nucleotides were removed. ORF sequences that still had corresponding UCIs after filtering were retained using the FASTQ identifiers. The names of ORFs with the same UCI were then concatenated and this final dictionary was used to generate a FASTA alignment file with ORF names and UCI sequences.

[0086] Informatics analysis of MIPSA data. Illumina output FASTQ files were truncated using a constant ACGAT anchor sequence after all UCI sequences. The truncated sequences were then mapped to their concatenated ORFs via the UCI-ORF lookup dictionary using exact match alignment. A count matrix is ​​constructed in which rows correspond to individual UCIs and columns correspond to samples. We then used the edgeR software package (51), which uses a negative binomial model to compare the signal detected in each sample to a set of negative control ("mock") IPs performed without serum, returning fold change values ​​and test statistics for each UCI in each sample, and created a fold change and significance matrix. Significantly enriched UCIs ("hits") required a read count of at least 15, a p-value of less than 0.001, and a fold change of at least 3. The hit fold change matrix reports fold change values ​​for "hits" and "1" for UCIs that are not hits.

[0087] Protein sequence similarity. To assess sequence homology between proteins in the hORFeome v8.1 library, each protein sequence was compared to all other library members using blastp alignment (parameters: "-outfmt 6 -evalue 100 -max_hsps 1 -soft_masking false -word_size 7 -max_target_seqs 100000").

[0088] Phage immunoprecipitation sequencing (PhIP-Seq) analysis. PhIP-Seq was performed according to a previously published protocol (45). Briefly, 0.2 μl of each plasma was individually mixed with a 90-aa human phage library and immunoprecipitated using protein A and protein G coated magnetic beads. A set of 6–8 mock immunoprecipitations (no plasma input) were performed on each 96-well plate. Magnetic beads were suspended in PCR master mix and subjected to thermocycling. A second PCR reaction was used for the sample bars. Amplicons were pooled and sequenced on an Illumina NextSeq 500 instrument. PhIP-Seq with human libraries was used to characterize autoantibodies in a collection of plasma from healthy donors. For a fair comparison with the severe COVID-19 cohort, we first determined the minimum sequencing depth that would have been required to detect IFN-λ3 reactivity in both positive individuals. We then considered only 423 datasets from the healthy cohort that had a sequencing depth greater than this minimum threshold, and none of these 423 individuals were found to be reactive to any peptide derived from IFN-λ3.

[0089] Type I / III Interferon Neutralization Assay. IFN-α2 (cat. no. 11100-1), IFN-λ1 (cat. no. 1598-IL-025), and IFN-λ3 (cat. no. 5259-IL-025) were purchased from R&D Systems. 20 μL of crude patient serum was incubated with either 100 U / mL IFN-α2 or 1 ng / mL IFN-λ3 in a total volume of 200 μL complete DMEM medium for 1 h at room temperature, followed by 7.5 × 10 4 A549 cells were added with 1x PBS. After 4 h of incubation, cells were washed with 1x PBS, and cellular mRNA was extracted and purified using RNeasy Plus Mini Kit (Qiagen). 600 ng of extracted mRNA was reverse transcribed using SuperScript III First-Strand Synthesis System (Life Technologies) and diluted 10-fold on a QuantStudio 6 Flex System (Applied Biosystems) for qPCR run. PCR was performed with a two-step cycling protocol, consisting of a 3 min cycle at 95°C, followed by 45 cycles of 15 s at 95°C and 30 s at 60°C. MX1 expression was chosen as a measure of cell stimulation by interferon, and relative mRNA expression was normalized by GAPDH expression. qPCR primers GAPDH and MX1 were obtained from Integrated DNA Technologies (Table 1). Table 1 Primer sequences [Table 1]

[0090] result Development of the MIPSA system. The MIPSA Gateway destination vector contains the following key elements: a T7 RNA polymerase transcription initiation site, an isothermal unique clone identifier ("UCI" barcode) flanked by constant primer binding sequences, a ribosome binding site (RBS), an N-terminal Halo tag fusion protein (891 nt), a recombination sequence for ORF insertion, a stop codon, and a homing endonuclease site for plasmid linearization. The recombinant ORF-containing pDEST-MIPSA plasmid is shown in Figure 1A.

[0091] We first sought to establish a library of pDEST-MIPSA plasmids containing stochastic isothermal UCIs located between the transcription start site and the ribosome binding site. Melting temperature (Tm) balanced sequences: (SW) 18 -AGGGA-(SW) 18 We synthesized a degenerate oligonucleotide pool containing 100 nucleotides (S, G, A, T) (Figure 1B), where S represents an equal mixture of C and G, and W represents an equal mixture of A and T. We determined that this inexpensive sequence pool was (i) of sufficient complexity for unique ORF labeling (2 36 ~7×10 10 ), (ii) amplify without distortion, and (iii) serve as ORF-specific forward and reverse qPCR primer binding sites for measurement of individual UCIs of interest. The degenerate oligonucleotide pool was amplified by PCR, restriction cloned into the MIPSA destination vector, and transformed into E. coli (Methods). Approximately 800,000 transformants were scraped from the selection plates to obtain the pDEST-MIPSA UCI plasmid library. ORFs encoding the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and a known autoantigen, tripartite motif containing-21 (TRIM21, commonly known as Ro52), were separately recombined into the pDEST-MIPSA UCI plasmid library and used for the following experiments. Single barcoded GAPDH and TRIM21 clones were isolated and sequenced.

[0092] The MIPSA procedure involves reverse transcription of stochastic barcodes using succinimidyl ester (O2)-haloalkane (Halo Ligand)-conjugated reverse transcription (RT) primers. The conjugated RT primer should not interfere with E. coli ribosome assembly and initiation of translation, but should be proximal enough that coupling of the Halo Ligand-Halo Tag-protein complex may disrupt further rounds of translation. We tested a series of RT primers annealing at distances ranging from -30 nucleotides to +7 nucleotides (5' to 3') from the 3' end of the RBS (Figure 1D). Based on the yield of protein product from mRNAs saturated with primers at these various positions, we selected the -20 position because it did not interfere with translation efficiency (Figure 1E). In contrast, RT from primers located within 20 nucleotides of the RBS reduced or abolished protein translation. This result is consistent with the predicted footprint of the assembled 70S E. coli ribosome, which has been shown to protect mRNAs of a minimum of 15 nucleotides ( 13 ).

[0093] Next, we evaluated the ability of SuperScript IV to reverse transcribe from a primer labeled with a Halo ligand at the 5' end and to form a covalent bond with the Halo ligand-conjugated primer in the translation reaction of Halo-tag-TRIM21 protein. Halo ligand conjugation and purification were performed according to Gu et al. (Materials and Methods, Figure 6) (14). Either unconjugated RT primer or aHalo ligand-conjugated RT primer was used for RT of barcoded Halo-tag-TRIM21 mRNA. The translation product was then immunoprecipitated (IPed) with serum from a healthy donor or serum from a TRIM21 (Ro52) autoantibody-positive patient with Sjögren's syndrome (SS). SS serum efficiently IPed TRIM21 protein regardless of RT primer binding, but pulled down only TRIM21 cDNA UCI when Halo ligand-conjugated primer was used in the RT reaction (Figure 1F-G).

[0094] Assessment of cis vs. trans UCI barcoding. Previous experiments indeed showed that Halo ligand does not interfere with RT priming and that Halo tag can form a covalent bond with Halo ligand during translation reaction, but the amount of cis (intra-complex) and trans (inter-complex) Halo tag-UCI conjugation was not elucidated (Figure 7). To measure the amount of cis and trans Halo tag-UCI-conjugation. GAPDH and TRIM21 mRNA were reverse transcribed separately (using Halo ligand primer) and then mixed 1:1 or kept separately for in vitro translation. As expected, translation of the mixture produced approximately equal amounts of each protein compared to individual translation (Figure 8). Regardless of the translation conditions, SS plasma specifically IPed TRIM21 protein (Figure 8, IPed fraction). However, we noted that although the SS IP contained high levels of TRIM21 UCI, more GAPDH UCI was pulled down by SS serum compared to that by HC serum when the mRNAs were mixed pretranslation, as intended. This indicates that some trans-barcoding does indeed occur (Figure 2A). We estimate that approximately 50% of the proteins are cis-barcoded, and the remaining 50% of trans-barcoded proteins are equally represented by both proteins. Thus, in this two-compartment system, 25% of TRIM21 proteins are conjugated to GAPDH-UCI.

[0095] In a complex library setting, even when approximately 50% of the proteins are trans-barcoded, this undesired by-product is uniformly distributed across all members of the library. We tested this using a model MIPSA library composed of a 100-fold excess of a second GAPDH clone combined with a 1:1 mixture of the first GAPDH and TRIM21 clones (Figure 2B). We further developed a sequencing workflow that utilizes PCR spike-in sequences for absolute quantification of each UCI. IP with SS serum using an optimized protocol resulted in specific IP of TRIM21-UCI, with negligible detection of trans-bound GAPDH-UCI IP (Figure 2B). Using the spike-in sequence for absolute quantification and assuming 100% pull-down of TRIM21 protein, we calculated a cis-coupling efficiency of approximately 0.2% (i.e., 0.2% of the input TRIM21 RNA molecules were converted into the intended UCI-coupled TRIM21 protein).

[0096] Establish and deconvolute a stochastically barcoded human ORFeome MIPSA library. The sequence-verified human ORFeome v8.1 is composed of 12,680 cloned ORFs that map to 11,437 genes in pDONR223 (15). Five subpools of libraries were generated, each consisting of approximately 2,500 similarly sized ORFs. Each of the five subpools was separately recombined and transformed into the pDEST-MIPSA UCI plasmid library to obtain approximately 10-fold ORF coverage (approximately 30,000 clones per subpool). Each subpool was assessed by Bioanalyzer electrophoresis, sequencing of approximately 20 colonies, and Illumina sequencing of the superpool. TRIM21 plasmid was spiked into the superpooled hORFeome library at 1:10,000, comparable to a typical library member. We then performed SS IP experiments on the hORFeome MIPSA library using sequencing as readout. Reads from all barcodes in the library, including spiked-in TRIM21, are shown in Figure 2C. SS autoantibody-dependent enrichment of TRIM21 (17-fold) was similar to the simple system (Figure 2D). Given the previously derived coupling efficiencies, we estimated approximately 6 × 10 6 correctly cis-coupled TRIM21 molecules. 5 molecules (and therefore, on average, each library member) input into the IP reaction.

[0097] Next, we established a system to create a UCI-ORF lookup dictionary using tagmentation and sequencing (Figure 3A). Sequencing of the 5'50nt ORF inserts detected 11,300 of 11,887 unique 5'50nt sequences. Of the 153,161 unique barcodes detected, 82.9% (126,975) were found to be associated with a single ORF. Each ORF was uniquely associated with a median of 9 UCI, ranging from 0 to 123 UCI (Figure 3B). When the reads corresponding to each ORF were aggregated, more than 99% of the represented ORFs were present within a 10-fold difference of the median ORF abundance (Figure 3C). Taken together, these data demonstrated that we established a uniform library of 11,300 probabilistically indexed human ORFs and a well-defined dictionary for downstream analysis. FIG. 3D shows the scatter plot of FIG. 2C, but with 47 dictionary-deciphered GAPDH UCIs (corresponding to the two GAPDH isoforms present in the hORFeome library) that appear along the y=x diagonal as expected.

[0098] Unbiased MIPSA analysis of autoantibodies associated with severe COVID-19. Several recent reports have described elevated autoantibody reactivity in patients with severe COVID-19 (16-20). We therefore used MIPSA with a human ORFeome library for unbiased identification of autoreactivity in plasma from 55 severe COVID-19 patients. For comparison, we used MIPSA to detect autoreactivity in plasma from 10 healthy donors and 10 non-hospitalized COVID-19 convalescent plasma donors (Table 2). Each sample was compared to a set of eight "mock IPs" containing all reaction components except serum. This comparison to mock IPs accounts for bias in the library and background binding. Importantly, the informatics pipeline used to detect antibody-dependent reactivity resulted in a median of 5 false-positive UCI hits per mock IP (range 2-9). However, IP using serum from severe COVID-19 patients yielded an average of 132 reactive UCIs, which was significantly higher than the average of 93 reactive UCIs among controls (p=0.018, t-test). Collapsing the UCIs into their corresponding proteins yielded an average of 83 reactive proteins among severe COVID-19 patients, which was significantly higher than the average of 63 reactive proteins among controls (Figure 4A, p=0.019, t-test). Table 2. Study subjects [Table 2]

[0099] We next examined proteins in severe COVID-19 IPs that had at least two reactive UCIs, were reactive in at least one severe patient, and were not reactive in more than one control (in healthy or mild / moderate convalescent plasma). Proteins that were reactive in one severe patient and one control were excluded. The 115 proteins that met these criteria are shown in the clustering heatmap in Figure 4B. Fifty-two of the 55 severe COVID-19 patients showed reactivity to at least one of these proteins. We focused on protein reactivity that occurred simultaneously in multiple individuals, the majority of which lacked homology by protein sequence alignment.

[0100] One notable autoreactive cluster (Figure 4B) contains 5'-nucleotidase, cytosolic 1A (NT5C1A), which is highly expressed in skeletal muscle and is the best-characterized autoantibody target in inclusion body myositis (IBM). Multiple UCIs linked to NT5C1A were significantly increased in 3 (5.5%) of 55 severe COVID-19 patients. NT5C1A autoantibodies have been reported in up to 70% of IBM patients (1), in about 20% of SS patients, and in up to about 5% of healthy donors (21). The frequency of NT5C1A reactivity in severe COVID-19 cohorts has not necessarily been elevated. However, we wondered whether MIPSA could reliably distinguish healthy donor from IBM plasma based on NT5C1A reactivity. We tested plasma from 10 healthy donors and 10 IBM patients, the latter selected based on NT5C1A seropositivity determined by PhIP-Seq (1). The clear separation of patients from controls in this independent cohort suggests that MIPSA may indeed have utility in clinical diagnostic testing using either qPCR or sequencing, which were closely correlated readouts (Figure 4C).

[0101] Neutralizing autoantibodies of type I and type III interferons in severe COVID-19. Neutralizing autoantibodies targeting type I interferon alpha (IFN-α) and omega (IFN-ω) have been associated with severe COVID-19 (17, 22, 23). All type I interferons except IFN-α16 are represented in the human MIPSA library and dictionary. However, IFN-α4, IFN-α17, and IFN-α21 cannot be distinguished by sequencing the first 50 nucleotides of the ORF sequences that encode them. Two of the severe COVID-19 patients in this cohort (3.6%) showed dramatic IFN-α autoreactivity (43 and 41 UCIs across 10 distinct IFN-α ORFs, and 5 and 2 IFN-ω UCIs, Figures 5A-5B). The extensive co-reactivity of these proteins is likely due to their sequence homology (Figure 9). By requiring at least two IFN UCIs to be considered positive, we identified three additional severe COVID-19 patients with lower levels of reactivity, each with only two reactive IFN-α UCIs. Interestingly, one plasma (P5) precipitated five UCIs from the type III interferon IFN-λ3, but no UCIs from any type I or type II interferons (Figures 5C-5D). None of the healthy or non-hospitalized COVID-19 controls were positive for more than two interferon UCIs.

[0102] Incubation of A549 human adenocarcinoma lung epithelial cells with 100 U / ml IFN-α2 or 1 ng / ml IFN-λ3 in serum-free medium for 4 h strongly upregulated the IFN-responsive gene MX1 by approximately 1,000-fold and approximately 100-fold, respectively. Preincubation of IFN-α2 with plasma P1, P2 or P3 completely abolished the A549 interferon response (Figure 5E). The plasma with the weakest IFN-α reactivity by MIPSA (P4) partially neutralized the cytokine. Neither HC nor P5 plasma had any effect on the response of A549 cells to IFN-α. However, preincubation of IFN-λ3 with MIPSA-reactive plasmas P2 and P5 neutralized the cytokine (Figure 5F). None of the other plasmas (HC, P1, P3 or P4) had any effect on the response of A549 cells to IFN-λ. In summary, antibody testing in this severe COVID-19 cohort identified strongly neutralizing IFN-α autoantibodies in 5.5% of patients and strongly neutralizing IFN-λ3 autoantibodies in 3.6% of patients, with one patient (1.8%) harboring both autoreactivities.

[0103] We reasoned that PhIP-Seq with a 90-aa human peptidome library (24) might also detect interferon antibodies in this cohort. PhIP-Seq detected IFN-α reactivity in plasma from P1 and P2, but to a much lower extent (Figure 5G). Two weaker IFN-α reactivities detected by MIPSA in plasma from P3 and P4 were both lost by PhIP-Seq. PhIP-Seq identified a single additional weakly IFN-α reactive sample that was negative by MIPSA (not shown). Detection of type III interferon autoreactivity (directed only to IFN-λ3) was fully concordant between the two techniques. PhIP-Seq data was used to narrow down the location of the dominant epitopes in type I and type III autoantigens (Figure 5H-5I).

[0104] We next considered the prevalence of IFN-λ3 autoreactivity in the general population and whether it may be increased among patients with severe COVID-19. PhIP-Seq was used to profile the plasma of 423 healthy controls, none of whom were found to have detectable IFN-λ3 autoreactivity. These data suggest that IFN-λ3 autoreactivity may be more frequent among individuals with severe COVID-19. This is the first report to describe neutralizing anti-IFNλ autoantibodies, thus proposing a potentially novel pathogenic mechanism that contributes to life-threatening COVID-19 in some patients.

[0105] Example 2: Neutralizing IFNL3 autoantibodies in severe COVID-19 identified by protein display technology.

[0106] Autoantibodies detected in severe COVID-19 patients using MIPSA. The association between autoimmunity and severe COVID-19 disease is increasingly recognized. In a cohort of 55 hospitalized individuals, we detected multiple established autoantibodies, including one that we previously associated with inclusion body myositis (1). We then tested the ability of MIPSA to detect NT5C1A autoantibodies in a separate cohort of seropositive IBM patients and healthy controls. The results support future efforts in evaluating the clinical utility of MIPSA for standardized comprehensive autoantibody testing. Such testing could utilize either singleplex qPCR or unbiased sequencing as readouts.

[0107] Although clusters of autoreactivity were observed in multiple individuals, it is unclear what function, if any, they may play in severe COVID-19. In larger studies, we expect that patterns of co-occurring reactivity, or reactivity to proteins with related biological functions, may ultimately define new autoimmune syndromes associated with severe COVID-19. Neutralizing IFN-α and IFN-ω autoantibodies have been described in patients with severe COVID-19 and are presumed to be pathogenic (17). These presumably pre-existing autoantibodies are very rarely present in the general population and likely block the restriction of viral replication in cell culture, thus preventing disease resolution. This finding paves the way for identifying a subset of individuals at risk for life-threatening COVID-19 pneumonia and suggested potential therapeutic avenues utilizing interferon beta that is not neutralized by these autoantibodies. In this study, MIPSA identified two individuals with broad reactivity to the entire family of IFN-α cytokines. Indeed, plasma from both individuals, plus one individual with weaker IFN-α reactivity detected by MIPSA, strongly neutralized recombinant IFN-α2 in a lung adenocarcinoma cell culture model. Unexpectedly, one individual in the cohort with no IFN-α reactivity pulled down five IFN-λ3 UCIs. A second IFN-α autoreactive individual also pulled down a single IFN-λ3 UCI. The same autoreactivity was also detected using PhIP-Seq. Interestingly, neither MIPSA nor PhIP-Seq detected reactivity to IFN-λ2, despite their high degree of sequence homology (Figure 9). We tested the IFN-λ3 neutralizing capacity in the plasma of these patients and observed a near complete loss of cellular response to the recombinant cytokine (Figure 5F). These data propose that IFN-λ3 autoreactivity is a novel potential pathogenic mechanism contributing to severe COVID-19 disease.

[0108] Type III IFNs (IFN-λ, also known as IL-28 / 29) are cytokines with potent antiviral activity that act primarily at barrier sites. The IFN-λR1 / IL-10RB heterodimeric receptor for IFN-λ is expressed on lung epithelial cells and is important for the innate response to viral infection. Mordstein et al. found that IFN-λ reduced the virulence and suppressed the replication of influenza virus, respiratory syncytial virus, human metapneumovirus, and SARS coronavirus (SARS-CoV-1) in mice (32). IFN-λ has been proposed to exert much of its antiviral activity in vivo through stimulatory interactions with immune cells rather than through the induction of an antiviral cellular state (33). Importantly, IFN-λ strongly restricts SARS-CoV-2 replication in primary human bronchial epithelial cells (34), primary human airway epithelial cultures (35), and primary human intestinal epithelial cells (36). Taken together, these studies suggest multifaceted mechanisms by which neutralizing IFN-λ autoantibodies may exacerbate SARS-CoV-2 infection.

[0109] did not detect type III IFN neutralizing antibodies among 101 individuals with type I IFN autoantibodies tested (17). In our study, one of the three IFN-α autoreactive individuals (P2, a 22-year-old male) also carried autoantibodies neutralizing IFN-λ3. This co-reactivity is very rare and therefore may not be shown in the Casanova cohort. Alternatively, it is possible that different assay conditions show different detection sensitivities. Casanova et al. cultured A549 cells with 50 ng / ml IFN-λ3 and did not preincubate in plasma, whereas we preincubated A549 cells in plasma for 1 h before culturing them with 1 ng / ml IFN-λ3. Those readouts of STAT3 phosphorylation may also offer different detection sensitivities compared to MX1 upregulation. Larger studies are needed to determine the true frequency of these reactivities in severe COVID-19 patients and matched controls. Here, we report neutralizing IFN-α and IFN-λ3 autoantibodies in 3 (5.5%) and 2 (3.6%) of 55 individuals with severe COVID-19, respectively. IFN-λ3 autoantibodies were not detected via PhIP-Seq in a larger cohort of 541 healthy controls collected pre-pandemic.

[0110] Type III interferons have been proposed as a therapeutic modality for SARS-CoV-2 infection (35, 37–41), and there are currently three ongoing clinical trials to test pegylated IFN-λ1 for efficacy in reducing morbidity and mortality associated with COVID-19 (ClinicalTrials.gov Identifiers: NCT04343976, NCT04534673, NCT04344600). One recently completed double-blind, placebo-controlled trial, NCT04354259, reported a significant reduction of 2·42 log copies per mL of SARS-CoV-2 at day 7 among patients with mild to moderate COVID-19 in an outpatient setting (p=0·0041) (42). Future studies will determine whether anti-IFN-λ3 autoantibodies are pre-existing or arise in response to SARS-CoV-2 infection and how frequently they also cross-neutralize IFN-λ1. However, based on sequence alignment of IFN-λ1 and IFN-λ3 (approximately 29% homology, Figure 9 ), cross-neutralization is expected to be rare, raising the possibility that patients with neutralizing IFN-λ3 autoantibodies may be particularly benefited from PEGylated IFN-λ1 treatment.

[0111] conclusion MIPSA is a novel self-assembling protein display technology that has important advantages over alternative approaches. It has the property that complementary technologies such as PhIP-Seq and MIPSA libraries can be conveniently screened in the same reaction using a programmable phage display library. The MIPSA protocol presented herein requires cap-independent cell-free translation, but future adaptations may overcome this limitation. Applications for MIPSA-based research include protein-protein, protein-antibody, and protein-small molecule interaction studies, including unbiased analysis of post-translational modifications. Here, we used MIPSA to discover neutralizing IFN-λ3 autoantibodies, which may contribute to life-threatening COVID-19 pneumonia in a subset of at-risk individuals, among many other potentially pathogenic autoreactivities. References 1. H. B. Larman et al., Cytosolic 5’-nucleotidase 1A autoimmunity in sporadic inclusion body myositis. Annals of neurology 73, 408-418 (2013). 2. G. J. Xu et al., Viral immunology. Comprehensive serological profiling of human populations using a synthetic human virome. Science 348, aaa0698 (2015). 3. E. Shrock et al., Viral epitope profiling of COVID-19 patients reveals cross-reactivity and correlates of severity. Science 370, (2020). 4. D. R. Monaco et al., Profiling serum antibodies with a pan allergen phage library identifies key wheat allergy epitopes. Nat Commun 12, 379 (2021). 5. S. F. Kingsmore, Multiplexed protein measurement: technologies and applications of protein and antibody arrays. Nat Rev Drug Discov 5, 310-320 (2006). 6. T. Kodadek, Protein microarrays: prospects and problems. Chem Biol 8, 105-115 (2001). 7. N. Ramachandran, E. Hainsworth, G. Demirkan, J. LaBaer, On-chip protein synthesis for making microarrays. Methods Mol Biol 328, 1-14 (2006). 8. S. Rungpragayphan, T. Yamane, H. Nakano, SIMPLEX: single-molecule PCR-linked in vitro expression: a novel method for high-throughput construction and screening of protein libraries. Methods Mol Biol 375, 79-94 (2007). 9. J. Zhu et al., Protein interaction discovery using parallel analysis of translated ORF 10. G. Liszczak, T. W. Muir, Nucleic Acid-Barcoding Technologies: Converting DNA Sequencing into a Broad-Spectrum Molecular Counter. Angew Chem Int Ed Engl 58, 4144-4162 (2019). 11. G. V. Los et al., HaloTag: a novel protein labeling technology for cell imaging and protein analysis. ACS Chem Biol 3, 373-382 (2008). 12. J. Yazaki et al., HaloTag -based conjugation of proteins to barcoding-oligonucleotides. Nucleic Acids Res 48, e8 (2020). 13. F. Mohammad, R. Green, A. R. Buskirk, A systematically-revised ribosome profiling method for bacteria reveals pauses at single-codon resolution. Elife 8, (2019). 14. L. Gu et al., Multiplex single-molecule interaction profiling of DNA-barcoded proteins. Nature 515, 554-557 (2014).

[0001] 15. X. Yang et al., A public genome-scale lentiviral expression library of human ORFs. Nat Methods 8, 659-661 (2011). 16. C. R. Consiglio et al., The Immunology of Multisystem Inflammatory Syndrome in Children with COVID-19. Cell 183, 968-981 e967 (2020). 17. P. Bastard et al., Autoantibodies against type I IFNs in patients with life-threatening COVID-19. Science 370, (2020). 18. Y. Zuo et al., Prothrombotic autoantibodies in serum from patients hospitalized with COVID-19. Sci Transl Med 12, (2020). 19. L. Casciola-Rosen et al., IgM autoantibodies recognizing ACE2 are associated with severe COVID-19. medRxiv, (2020). 20. M. C. Woodruff, R. P. Ramonell, F. E. Lee, I. Sanz, Broadly-targeted autoreactivity is common in severe SARS-CoV-2 Infection. medRxiv, (2020). 21. T. E. Lloyd et al., Cytosolic 5’-Nucleotidase 1A As a Target of Circulating Autoantibodies in Autoimmune Diseases. Arthritis Care Res (Hoboken) 68, 66-71 (2016). 22. E. Y. Wang et al., Diverse Functional Autoantibodies in Patients with COVID-19. medRxiv, (2020). 23. S. Gupta, S. Nakabo, J. Chu, S. Hasni, M. J. Kaplan, Association between anti-interferon-alpha autoantibodies and COVID-19 in systemic lupus erythematosus. medRxiv, (2020). 24. G. J. Xu et al., Systematic autoantigen analysis identifies a distinct subtype of scleroderma with coincident cancer. Proc Natl Acad Sci U S A, (2016). 25. M. Stoeckius et al., Simultaneous epitope and transcriptome measurement in single cells. Nat Methods 14, 865-868 (2017). 26. I. Setliff et al., High-Throughput Mapping of B Cell Receptor Sequences to Antigen Specificity. Cell 179, 1636-1646 e1615 (2019). 27. S. K. Saka et al., Immuno-SABER enables highly multiplexed and amplified protein. 8. M. A. Jongsma, R. H. Litjens, Self-assembling protein arrays on DNA chips by auto-labeling fusion proteins with a single DNA address. Proteomics 6, 2650-2655 (2006). 29. A. Gautier et al., An engineered protein tag for multiprotein labeling in living cells. Chem Biol 15, 128-136 (2008). 30. A. J. Samelson et al., Kinetic and structural comparison of a protein’s cotranslational folding and refolding pathways. Sci Adv 4, eaas9098 (2018). 31. L. Tosi et al., Long-adapter single-strand oligonucleotide probes for the massively multiplexed cloning of kilobase genome regions. Nat Biomed Eng 1, (2017). 32. M. Mordstein et al., Lambda interferon renders epithelial cells of the respiratory and gastrointestinal tracts resistant to viral infections. J Virol 84, 5670-5677 (2010). 33. N. Ank et al., Lambda interferon (IFN-lambda), a type III IFN, is induced by viruses and IFNs and displays potent antiviral activity against select virus infections in vivo. J Virol 80, 4501-4509 (2006). 34. I. Busnadiego et al., Antiviral Activity of Type I, II, and III Interferons Counterbalances ACE2 Inducibility and Restricts SARS-CoV-2. mBio 11, (2020). 35. A. Vanderheiden et al., Type I and Type III Interferons Restrict SARS-CoV-2 Infection of Human Airway Epithelial Cultures. J Virol 94, (2020). 36. M. L. Stanifer et al., Critical Role of Type III Interferon in Controlling SARS-CoV-2 Infection in Human Intestinal Epithelial Cells. Cell Rep 32, 107863 (2020). 37. I. E. Galani et al., Untuned antiviral immunity in COVID-19 revealed by temporal type I / III interferon patterns and flu comparison. Nat Immunol 22, 32-40 (2021). 38. U. Felgenhauer et al., Inhibition of SARS-CoV-2 by type I and type III interferons. J Biol Chem 295, 13958-13964 (2020). 39. T. R. O’Brien et al., Weak Induction of Interferon Expression by Severe Acute Respiratory Syndrome Coronavirus 2 Supports Clinical Trials of Interferon-lambda to Treat Early Coronavirus Disease 2019. Clin Infect Dis 71, 1410-1412 (2020). 40. E. Andreakos, S. Tsiodras, COVID-19: lambda interferon against viral load and hyperinflammation. EMBO Mol Med 12, e12465 (2020). 41. L. Prokunina-Olsson et al., COVID-19 and emerging viral infections: The case for interferon lambda. J Exp Med 217, (2020). 42. J. J. Feld et al., Peginterferon lambda for the treatment of outpatients with COVID-19: a phase 2, placebo-controlled randomised trial. Lancet Respir Med, (2021). 43. D. Mohan et al., Publisher Correction: PhIP-Seq characterization of serum antibodies using oligonucleotide-encoded peptidomes. Nature protocols 14, 2596 (2019). 44. C. Tuckey, H. Asahara, Y. Zhou, S. Chong, Protein synthesis using a reconstituted cell-free system. Curr Protoc Mol Biol 108, 16 31 11-22 (2014). 45. D. Mohan et al., PhIP-Seq characterization of serum antibodies using oligonucleotide-encoded peptidomes. Nature protocols 13, 1958-1978 (2018). 46. S. L. Klein et al., Sex, age, and hospitalization drive antibody responses in a COVID-19 convalescent plasma donor population. J Clin Invest 130, 6141-6150 (2020). 47. R. A. Zyskind I, Zimmerman J, Naiditch H, Glatt AE, Pinter A, Theel ES, Joyner MJ, Hill DA, Lieberman MR, Bigajer E, Stok D, Frank E, Silverberg JI, SARS-CoV-2 Seroprevalence and Symptom Onset in Culturally-Linked Orthodox Jewish Communities Across Multiple Regions in the United States. JAMA Open Network In Press, (2021). 48. Correction: Patient Trajectories Among Persons Hospitalized for COVID-19. Ann Intern Med 174, 144 (2021). 49. M. R. Rose, E. I. W. Group, 188th ENMC International Workshop: Inclusion Body Myositis, 2-4 December 2011, Naarden, The Netherlands. Neuromuscul Disord 23, 1044-1055 (2013). 50. Z. Wei, W. Zhang, H. Fang, Y. Li, X. Wang, esATAC: an easy-to-use systematic pipeline for ATAC-seq data analysis. Bioinformatics 34, 2664-2665 (2018). 51. MD Robinson, DJ McCarthy, GK Smyth, edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139-140 (2010). The remainder of this page has been intentionally left blank.

[0112] Example 3: Molecular Indexing of Proteins by Self-Assembly (MIPSA) identifies neutralizing type I and type III interferon autoantibodies in severe COVID-19.

[0113] Unbiased analysis of antibody binding specificity can provide important insights into healthy and disease states. We and others have utilized programmable phage display libraries to identify novel autoantibodies, characterize antiviral immunity, and profile allergen-specific IgE antibodies. 1~4 Although phage display is useful for these and many other applications, most protein-protein, protein-antibody and protein-small molecule interactions require a degree of conformation that is not captured by bacteriophage-displayed peptide libraries. Profiling conformational protein interactions on a proteome scale has traditionally relied on protein microarray technology. However, protein microarrays tend to suffer from a high per-assay cost and a myriad of technical artifacts, including those associated with high-throughput expression and purification of proteins, spotting of proteins onto solid supports, drying and rehydration of arrayed proteins, and slide-scanning fluorescent imaging-based readouts. 5,6 Although alternative approaches to protein microarray production and storage have been developed (e.g., Nucleic Acid Programmable Protein Arrays, NAPPA).7 , or single molecule PCR-associated in vitro expression, SIMPLEX 8 ), robust, scalable, and cost-effective alternatives are lacking.

[0114] To overcome the limitations associated with array-based profiling of full-length proteins, we previously established a methodology called Parallel Analysis of Translated Open reading frames (PLATO), which utilizes ribosome display of open reading frame (ORF) libraries. 9 Ribosome display relies on the in vitro translation of mRNAs lacking stop codons, stalling ribosomes at the ends of mRNA molecules in complex with the nascent proteins they encode. PLATO suffers from several important limitations that have hindered its adoption. An ideal alternative would be the covalent attachment of proteins to short, amplifiable DNA barcodes. Indeed, individually prepared DNA-barcoded antibodies and proteins have been used successfully in a variety of applications. 10 One particularly attractive protein-DNA conjugation method is the HaloTag system, which utilizes a bacterial enzyme that forms an irreversible covalent bond with a halogen-terminated alkane moiety. 11 Individual DNA-barcoded Halo tag fusion proteins have been shown to significantly enhance the sensitivity and dynamic range of autoantibody detection compared to conventional ELISA. 12 Scaling individual protein barcoding to whole ORFeome libraries, while highly useful, is prohibitive due to high cost and low throughput. Thus, self-assembly approaches may offer a much more efficient route to library generation.

[0115] Here, a novel molecular display technology, Molecular Indexing of Proteins by Self Assembly (MIPSA), is described, which overcomes major drawbacks of PLATO and other full-length protein array technologies. MIPSA produces a library of soluble full-length proteins, each uniquely identifiable via covalent attachment to an amplifiable barcode. The barcode is introduced upstream of a ribosome binding site (RBS). Partial reverse transcription (RT) of in vitro transcribed RNA (IVT-RNA) creates a cDNA barcode that is ligated to a haloalkane-labeled RT primer. An N-terminal Halo tag fusion protein is encoded downstream of the RBS such that in vitro translation results in covalent attachment in-complex ("cis") of the Halo tag and the cDNA barcode to its downstream open reading frame (ORF)-encoded protein product. The resulting library of uniquely indexed full-length proteins can be used for inexpensive proteome-wide interaction studies such as unbiased autoantibody profiling.

[0116] Coronavirus disease 2019 (COVID-19), caused by SARS-CoV-2 infection, ranges from an asymptomatic course to life-threatening pneumonia and death. A causal relationship between autoimmunity and severe COVID-19 is supported by multiple studies. 13,14 A wide variety of autoantibodies have been documented. 15 , neutralizing type I interferon autoantibodies appear to play a particularly important role. 16,17 Here, we investigate the utility of the MIPSA platform by searching for novel autoantibodies in the plasma of patients with severe COVID-19.

[0117] method

[0118] MIPSA destination vector construction

[0119] The MIPSA vector was constructed using the Gateway pDEST15 vector as the backbone. A gBlock fragment (Integrated DNA Technologies) encoding the RBS, Kozak sequence, N-terminal Halo tag fusion protein, and FLAG tag followed by the attR1 sequence was cloned into the parent plasmid. A 150 bp poly(A) sequence was also added after the attR2 site. The TRIM21 and GAPDH ORF sequences used to characterize and optimize the two-component system contained the native stop codons that were retained in the final MIPSA construct.

[0120] UCI barcode library construction

[0121] A 41 nt barcode oligo was generated within the gBlock Gene Fragment (Integrated DNA Technologies) with alternating mixed bases (S:G / C; W:A / T) and the following sequence: (SW) 18 -AGGGA-(SW) 18 The sequences flanking the degenerate barcodes incorporated standard PhIP-Seq PCR1 and PCR2 primer binding sites. 51 Eighteen nanograms of the starting UCI library was used to perform 40 cycles of PCR to amplify the library and incorporate BglII and PspxI restriction sites. The MIPSA vector and amplified UCI library were then digested overnight with restriction enzymes, column purified, and ligated at a vector to insert ratio of 1:5. The ligated MIPSA vector was used to transform electrocompetent One Shot ccdB2 T1 R Cells (Thermo Fisher Scientific) were transformed with 6 rounds of transformation to obtain approximately 800,000 colonies, and the pDEST-MIPSA UCI library was generated.

[0122] Human ORFeome recombination into the pDEST-MIPSA UCI plasmid library

[0123] 150ng of each pENTR-hORFeome subpool (L1-L5) from hORFeome v8.1 was individually combined with 150ng of pDEST-MIPSA UCI library plasmid and 2μl of Gateway LR Clonase II mix (Life Technologies) in a total reaction volume of 10μl. Reactions were incubated overnight at 25°C. All reactions were aliquoted in 50μl of One Shot OmniMAX 2 T1 R The plasmids were transformed into chemically competent E. coli (Life Technologies). Overall, the transformation yielded approximately 120,000 colonies, which is approximately 10 times the complexity of the hORFeome v8.1. After collecting and pooling the colonies by scraping, the barcoded pDEST-MIPSA-hORFeome plasmid MIPSA (human ORFeome DNA library) was purified using the Qiagen Plasmid Midi Kit (Qiagen). The human hORFeome v8.1 collection was cloned without a stop codon; therefore, the displayed proteins may contain a polylysine C-terminus resulting from translation of the polyA tail. More recent versions of the MIPSA destination vector contain a stop codon in frame with the recombinant ORF.

[0124] Halo Ligand Binding to RT Oligos and HPLC Purification

[0125] 100 μg of 5' amine modified oligo HL-32_ad (Table 1) was co-transfected with 75 μl (17.85 μg / μl) of Halo-tag Succinimidyl Ester (O2) (Promega Corporation) as a Halo ligand in 0.1 M sodium borate buffer as described by Gu et al. 19 The reaction was incubated at room temperature for 6 h according to the protocol. 3 M NaCl and ice-cold ethanol were added to the labeling reaction at 10% (v / v) and 250% (v / v), respectively, and incubated at -80°C overnight. The reaction was centrifuged at 12,000 x g for 30 min. The pellet was rinsed once with ice-cold 70% ethanol and air-dried for 10 min.

[0126] Halo-liganded RT primers were HPLC-purified using a Brownlee Aquapore RP-300 7u, 100x4.6mm column (Perkin Elmer) using a 2-buffer gradient of 0-70% CH3CN / MeCN (100mM trimethylamine acetate in acetonitrile) over 70 min. Fractions corresponding to the labeled oligos were collected and lyophilized (Figures 15A-15C). Oligos were resuspended at 1μM (15.4ng / μl) and stored at -80°C.

[0127] MIPSA library IVT-RNA preparation

[0128] Human ORFeome MIPSA library plasmids (4 μg) were linearized overnight with I-SceI restriction endonuclease (New England Biolabs). The products were column purified using NucleoSpin Gel and PCR Clean Up kit (Macherey-Nagel). 40 μl of in vitro transcription reaction was performed using HiScribe T7 High Yield RNA Synthesis Kit (New England Biolabs) to transcribe 1 μg of purified linearized pDEST-MIPSA plasmid library. The products were diluted with 60 μl of molecular biology grade water and 1 μl of DNAseI was added. The reaction was incubated at 37 °C for an additional 15 minutes. 50 μl of 1 M LiCl was then added to the solution and incubated at -80 °C overnight. The centrifuge was cooled to 4 °C and the RNA was spun at maximum speed for 30 minutes. The supernatant was removed and the RNA pellet was washed with 70% ethanol. The samples were spun down for an additional 10 minutes at 4 °C and the 70% ethanol was removed. The pellet was dried at room temperature for 15 min and then resuspended in 100 μl of water. To preserve the sample, 1 μl of 40 U / μl RNAseOUT Recombinant Ribonuclease Inhibitor (Life Technologies) was added.

[0129] MIPSA library IVT-RNA reverse transcription and translation

[0130] Reverse transcription reactions were prepared using the SuperScript IV First-Strand Synthesis System (Life Technologies). First, 1 μl of 10 mM dNTPs, 1 μl of RNAseOUT (40 U / μl), 4.17 μl of RNA library (1.5 μM), and 7.83 μl of Halo ligand-conjugated RT primer (1 μM, Table 1) were combined in a single 14 μl reaction and incubated at 65° C. for 5 minutes, followed by incubation on ice for 2 minutes. Four microliters of 5× RT buffer, 1 μl of 0.1 M DTT, and 1 μl of SuperScript IV RT enzyme (200 U / μl) were added to the 14 μl reaction on ice and incubated at 42° C. for 20 minutes. 36 μl of RNAClean XP beads (Beckman Coulter) were added to the single 20 μl RT reaction and incubated at room temperature for 10 minutes. The beads were collected on a magnet and washed five times with 70% ethanol. The beads were air-dried at room temperature for 10 min and resuspended in 7 μl of 5 mM Tris-HCl, pH 8.5. The products were analyzed spectrophotometrically to measure RNA yield. Translation reactions were set up on ice using the PURExpress ΔRibosome Kit (New England Biolabs). 52 Reactions were modified to a final concentration of 0.3 μM ribosomes. For each 10 μl translation reaction, 4.57 μl of RT reaction was added to 4 μl of solution A, 1.2 μl of Factor Mix, and 0.23 μl of ribosomes (13.3 μM). The reactions were incubated at 37°C for 2 h, diluted to a total volume of 45 μl with 35 μl of 1× PBS, and either used immediately or stored at −80°C after adding glycerol to a final concentration of 25% (v / v).

[0131] Immunoprecipitation of translated MIPSA hORFeome libraries.

[0132] 5 μl of plasma (diluted 1:100 in PBS) is mixed with 45 μl of diluted MIPSA library translation reaction (see above) and incubated overnight at 4 °C with gentle agitation. For each IP, a mixture of 5 μl of Protein A Dynabeads and 5 μl of Protein G Dynabeads (Life Technologies) was washed three times with 1 × PBS in twice their original volume. The beads were then resuspended in 1 × PBS in the original volume and added to each IP. Antibody capture was allowed to proceed for 4 h at 4 °C. The beads were collected on a magnet and washed three times with 1 × PBS, changing tubes or plates between washes. The beads were then collected and resuspended in 20 μl of PCR master mix containing T7-Pep2_PCR1_F forward primer and T7-Pep2_PCR1_R+ad_min reverse primer (Table 1) and Herculase-II (Agilent). The PCR cycle was as follows: an initial denaturation and enzyme activation step at 95°C for 2 min, followed by 20 cycles of 95°C for 20 s, 58°C for 30 s, and 72°C for 30 s. A final extension step was performed at 72°C for 3 min. Two microliters of PCR1 amplified product was used as input to a 20 μl dual-index PCR reaction with PhIP_PCR2_F forward and PhIP_PCR2_R reverse primers, each containing a 10 nt barcode (i5 and i7, respectively). The PCR cycle was as follows: an initial denaturation step at 95°C for 2 min, followed by 20 cycles of 95°C for 20 s, 58°C for 30 s, and 72°C for 30 s. A final extension step was performed at 72°C for 3 min. The i5 / i7 indexed libraries were pooled and column purified (NucleoSpin columns, Takara). Libraries were sequenced on an Illumina NextSeq 500 using 1x50nt SE or 1x75nt SE protocols. MIPSA_i5_NextSeq_SP and Standard_i7_SP primers were used for i5 / i7 sequencing (Table 1). Output was demultiplexed using i5 and i7 without allowing mismatches.

[0133] For quantification of MIPSA experiments by qPCR, PCR1 products (above) were analyzed as follows: 4.6 μl of a 1:1,000 dilution of PCR1 reaction was added to 5 μl of Brilliant III Ultra Fast 2X SYBR Green Mix (Agilent), 0.2 μl of 2 μM reference dye and 0.2 μl of 10 μM forward and reverse primer mix (specific for target UCI). The PCR cycle was as follows: an initial denaturation step at 95°C for 2 min, followed by 45 cycles of 95°C for 20 s and 60°C for 30 s. After completion of thermocycling, the amplified products were subjected to melting curve analysis. The qPCR primers for MIPSA immunoprecipitation experiments were BG 2_F and BG 2_R for TRIM21, BG4_F and BG4_R for GAPDH, and NT5C1A_F and NT5C1A_R for NT5C1A (Table 1).

[0134] Oligonucleotides

[0135] Table 5 provides a list of probes, primers and gRNAs.

[0136] Plasma samples

[0137] All samples were collected from subjects who met the protocol eligibility criteria, as described below. All studies protected the rights and privacy of study participants and were approved by the respective Institutional Review Boards for the original sample collection and subsequent analysis.

[0138] Pre-pandemic and healthy control plasma samples. All human samples were collected prior to 2017 according to NIAID IRB-approved procedures at the National Institutes of Health (NIH), Vaccine Research Center (VRC) / National Institutes of Allergy and Infectious Diseases (NIAID) / NIH protocol “VRC 000: Screening Subjects for HIV Vaccine Research Studies” (NCT00031304).

[0139] COVID-19 convalescent plasma (CCP) from non-hospitalized patients. Eligible CCP donors were contacted by study personnel as previously described. 53 All donors were at least 18 years of age and had a confirmed diagnosis of SARS-CoV-2 by detection of RNA in a nasopharyngeal swab sample. Basic demographic information (age, sex, hospitalization with COVID-19) was obtained for each donor; first diagnosis and date of diagnosis of SARS-CoV-2 were confirmed by medical record review.

[0140] Severe COVID-19 plasma samples. The study cohort was comprised of hospitalized patients with: 1) a confirmed diagnosis of COVID-19; 2) death or survival to discharge; and 3) a remnant specimen in the Johns Hopkins COVID-19 Remnant Specimen Biorepository (a chance sample that includes 59% of Johns Hopkins COVID-19 patients and 66% of patients with length of stay ≥ 3 days). 54,55Patient outcomes were defined by the World Health Organization (WHO) COVID-19 Disease Severity Scale. Samples from Johns Hopkins with severe COVID-19 included in this study were obtained from 17 patients who died, 13 patients who recovered after mechanical ventilation, 22 patients who required supplemental oxygen for recovery, and 3 patients who recovered without supplemental oxygen. The study was approved by the JHU Institutional Review Board (IRB00248332, IRB00273516), and a waiver of consent was granted as all specimens and clinical data were de-identified by the Core for Clinical Research Data Acquisition at the Johns Hopkins Institute for Clinical and Translational Research; the study team had no access to identifiable patient data.

[0141] Sjögren's syndrome and inclusion body myositis (IBM) plasma samples. Sjögren's syndrome samples were collected under protocol NA_00013201. All patients were >18 years old and gave informed consent. IBM patient samples were collected under protocol IRB00235256. All patients met the ENMC 2011 diagnostic criteria. 56 and provided informed consent.

[0142] Immunoblot analysis

[0143] Laemmli buffer containing 5% β-ME was added to the samples, boiled for 5 min, and analyzed on NuPAGE 4–12% Bis-Tris polyacrylamide gels (Life Technologies). After transfer to PVDF membranes, blots were blocked for 30 min at room temperature in 20 mM Tris-buffered saline (pH 7.6) containing 0.1% Tween 20 (TBST) and 5% (wt / vol) nonfat dry milk. Blots were subsequently incubated overnight at 4°C with a primary anti-FLAG antibody (#F3165, MilliporeSigma) at 1:4,000 (v / v), followed by incubation in a 1:2,000 (v / v) anti-mouse IgG, HRP-conjugated secondary antibody (#7076, Cell Signaling) at room temperature for 4 h.

[0144] Construction of the UCI-ORF dictionary.

[0145] Nextera XT DNA Library Preparation Kit (Illumina) was used for tagmentation of 150 ng of pDEST-MIPSA hORFeome plasmid library to obtain optimal size distribution centered at approximately 1.5 kb. The tagmented library was amplified using Herculase-II (Agilent) with T7-Pep2_PCR1_F forward and Nextera Index 1 Read primers. PCR cycles were as follows: an initial denaturation step at 95°C for 2 min, followed by 30 cycles of 95°C for 20 s, 53.5°C for 30 s, and 72°C for 30 s. A final extension step was performed at 72°C for 3 min. PCR reactions were run on a 1% agarose gel, followed by excision of the approximately 1.5 kb product and purification using NucleoSpin Gel and PCR Clean-up columns (Macherey-Nagel). The purified products were then amplified for an additional 10 cycles using PhIP_PCR2_F forward primer and P7.2 reverse primer (see Table 1 for a list of primer sequences). Products were gel purified and sequenced on a MiSeq (Illumina) using T7-Pep2.2_SP_subA primer for read 1 and MISEQ_MIPSA_R2 primer for read 2. Read 1 was 60 bp long to capture UCI. The first index read I1 was replaced with a 50 bp read to the ORF. I2 was used to identify the i5 index for sample demultiplexing.

[0146] The hORFeome v8.1 sequences were truncated to the first 50 nt, and ORF names corresponding to non-unique sequences were concatenated with a "|" delimiter. The demultiplexed output of 50 nt R2 (ORF) reads from Illumina MiSeq was aligned to the truncated human ORFeome v8.1 library using the Rbowtie2 package with the following parameters: options = "-a --very-sensitive-local" (57). The unique FASTQ identifiers were then used to extract corresponding sequences from the 60 bp R1 (UCI) reads. These sequences were truncated using the 3' anchor ACGATA, and sequences without an anchor were removed. Additionally, any truncated R1 sequences with fewer than 18 nucleotides were removed. ORF sequences that still had corresponding UCIs after filtering were retained using the FASTQ identifier. The names of ORFs with the same UCI were concatenated with a "&" delimiter, and this final dictionary was used to generate a FASTA alignment file consisting of ORF names and UCI sequences.

[0147] Informatics analysis of MIPSA sequencing data

[0148] Illumina output FASTQ files were truncated using a constant ACGAT anchor sequence after all UCI sequences. The truncated sequences were then mapped to their concatenated ORFs via the UCI-ORF lookup dictionary using exact match alignment. A read count matrix was constructed where rows correspond to individual UCIs and columns correspond to samples. The edgeR software package 58We used EdgeR, which uses a negative binomial model to compare the signal detected in each sample to a set of negative control ("mock") IPs performed without plasma to return maximum likelihood fold change estimates and test statistics for each UCI in all samples, thus creating fold change and -log10(p-value) matrices. Comparison of EdgeR output data from replicate IPs established that significantly enriched UCIs ("hits") should require a read count of at least 15, a p-value of less than 0.001, and a fold change of at least 3. The hit fold change matrix reports fold change values ​​for "hits" and "1" for UCIs that are not hits.

[0149] Protein sequence similarity

[0150] To assess sequence homology between proteins in the hORFeome v8.1 library, each protein sequence was compared to all other library members using blastp alignment (parameters: "-outfmt 6 -evalue 100 -max_hsps 1 -soft_masking false -word_size 7 -max_target_seqs 100000"). To assess sequence homology between reactive peptides in the human 90-aa phage display library, epitopefindr 59 Software was used.

[0151] Phage immunoprecipitation sequencing (PhIP-Seq) analysis

[0152] hIP-Seq was performed according to a previously published protocol. 51Briefly, 0.2 μl of each plasma was individually mixed with a 90-aa human phage library and immunoprecipitated using protein A and protein G coated magnetic beads. A set of 6–8 mock immunoprecipitations (no plasma input) was performed on each 96-well plate. The magnetic beads were resuspended in PCR master mix and subjected to thermocycling. A second PCR reaction was used for the sample bar. Amplicons were pooled and sequenced on an Illumina NextSeq 500 instrument using a 1 × 50nt SE or 1 × 75nt SE protocol. PhIP-Seq with human libraries was used to characterize autoantibodies in a collection of plasma from healthy controls. For a fair comparison with the severe COVID-19 cohort, we first determined the minimum sequencing depth that would have been required to detect IFN-λ3 reactivity in both positive and negative individuals. Only then were 423 datasets from the healthy cohort considered with a sequencing depth greater than this minimum threshold. None of these 423 individuals were found to be reactive to any peptide derived from IFN-λ3.

[0153] Type I / III interferon neutralization assay

[0154] IFN-α2 (catalog no. 11100-1), IFN-λ1 (catalog no. 1598-IL-025), and IFN-λ3 (catalog no. 5259-IL-025) were purchased from R&D Systems. Twenty microliters of plasma was incubated with either 100 U / ml IFN-α2 or 1 ng / ml IFN-λ3 and 180 μl of DMEM in a total volume of 200 μl for 1 h at room temperature, followed by dissociation of 7.5 × 10 cells in 48-well tissue culture plates. 4A549 cells were added to the MX1-binding domain. After 4 h of incubation, cells were washed with 1x PBS, and cellular mRNA was extracted and purified using the RNeasy Plus Mini Kit (Qiagen). 600 ng of extracted mRNA was reverse transcribed using the SuperScript III First-Strand Synthesis System (Life Technologies) and diluted 10-fold for qPCR run on a QuantStudio 6 Flex System (Applied Biosystems). PCR was run using a two-step cycling protocol, consisting of a 3 min cycle at 95°C, followed by 45 cycles of 15 s at 95°C and 30 s at 60°C. MX1 expression was chosen as a measure of cell stimulation by interferon, and relative mRNA expression was normalized to GAPDH expression. qPCR primers for GAPDH and MX1 were obtained from Integrated DNA Technologies (Table 1). Anti-hIFN-α2-IgG (catalog number mabg-hifna-3) and anti-hIL-28b-IgG (catalog number mabg-hil28b-3) were purchased from InvivoGen. Manufacturer's note for mabg-hifna-3: "This antibody reacts with hIFN-α1, hIFN-α2, hIFN-α5, hIFN-α8, hIFN-α14, hIFN-α16, hIFN-α17 and hIFN-α21; reacts very weakly with hIFN-α4 and IFN-α10; does not react with hIFN-α6 or hIFN-α7." Manufacturer's note for mabg-hil28b-3: "Reacts with human IL-28A and human IL-28B."

[0155] result

[0156] Development of the MIPSA system

[0157] The MIPSA Gateway destination vector for E. coli cell-free translation contains the following key elements: a T7 RNA polymerase transcription initiation site, an isothermal unique clone identifier ("UCI") barcode sequence, an E. coli ribosome binding site (RBS), an N-terminal Halo tag fusion protein (891 nt), a recombination sequence for ORF insertion, and a homing endonuclease (I-SceI) site for plasmid linearization. The recombinant ORF-containing pDEST-MIPSA plasmid is shown in Figure 1A.

[0158] It was first sought to establish a library of pDEST-MIPSA plasmids containing stochastic isothermal UCIs located between the transcription initiation site and the ribosome binding site. Melting temperature (Tm) balanced sequences: (SW) 18 -AGGGA-(SW) 18 We synthesized a degenerate oligonucleotide pool containing 100 kDa of 100 kDa (where S represents an equal mixture of C and G, and W represents an equal mixture of A and T) (Figure 1B). This inexpensive pool of sequences was designed to be (i) of sufficient complexity for unique ORF labeling (2 36 ~7×10 10 ), (ii) amplify without distortion, and (iii) serve as ORF-specific forward and reverse qPCR primer binding sites for the measurement of individual UCI of subjects. The degenerate oligonucleotide pool was amplified by PCR, restriction cloned into the MIPSA destination vector, and transformed into E. coli (Methods). Approximately 800,000 transformants were scraped from the selection plates to obtain the pDEST-MIPSA UCI plasmid library. ORFs encoding the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and a known autoantigen, tripartite motif-containing-21 (TRIM21, commonly known as Ro52), were separately recombined into the pDEST-MIPSA UCI plasmid library. Individually barcoded GAPDH and TRIM21 clones were isolated, sequenced, and used in the following experiments.

[0159] The MIPSA procedure involves RT of UCI using succinimidyl ester (O2)-haloalkane (Halo Ligand)-conjugated RT primers (Figure 6A-6C). The conjugated RT primer should not interfere with E. coli ribosome assembly and initiation of translation, but should be proximal enough that coupling of the Halo Ligand-Halo Tag-protein complex may disrupt further rounds of translation. A series of RT primers that anneal at distances ranging from -42 nucleotides to -7 nucleotides relative to the 3' end of the ribosome binding site were evaluated (Figure 1D). Based on the yield of protein product from mRNAs saturated with primers at these different positions, we selected the -32 position, which did not disrupt translation efficiency (Figure 1E). In contrast, RT from primers located within 20 nucleotides of the RBS reduced or abolished protein translation, consistent with the predicted footprint of assembled 70S E. coli ribosomes shown to protect a minimum of 15 nucleotides of mRNA. 18 .

[0160] We assessed whether SuperScript IV can perform RT from primers labeled with Halo ligand at the 5' end and whether Halo-tagged TRIM21 protein can form a covalent bond with the Halo ligand-bound primer during the translation reaction. Halo ligand conjugation and purification were performed according to Gu et al. (Methods, Figures A-15C). 19Either unconjugated RT primers or Halo ligand-conjugated RT primers were used for RT of barcoded Halo tag-TRIM21 mRNA. The translation products were then immunocaptured (i.e., immunoprecipitated, "IPed") with plasma from healthy donors or plasma from TRIM21 (Ro52) autoantibody positive patients with Sjögren's syndrome (SS) using protein A and protein G coated magnetic beads. SS plasma efficiently IP'd TRIM21 protein regardless of RT primer binding, but only pulled down TRIM21 UCI when Halo ligand-conjugated primers were used in the RT reaction (Figure 10F-10G).

[0161] Assessment of the level of cis vs. trans UCI barcoding

[0162] Previous experiments indeed showed that Halo ligand does not interfere with RT priming and that Halo tag can form a covalent bond with Halo ligand during translation reaction, but did not elucidate the amount of cis (intra-complex, desirable) versus trans (inter-complex, undesirable) Halo tag-UCI conjugation (Figure 16A-16C). Here, "intra-complex" is defined as conjugation to UCIs that associate with the same RNA molecule that encodes a protein. To measure the amount of cis and trans Halo tag-UCI conjugation, GAPDH and TRIM21 mRNAs were reverse transcribed separately (using Halo ligand-conjugated primers) and then mixed 1:1 for in vitro translation or kept separate. As expected, translation of the mixture produced approximately equal amounts of each protein compared to individual translations (Figure 8). Regardless of translation conditions, SS plasma specifically IP'd TRIM21 protein (Figure 8, IPed fraction). However, while SS IP contained high levels of TRIM21 UCI, it was noted that more GAPDH UCI was pulled down by SS plasma compared to that by HC plasma when the mRNAs were mixed pretranslationally. This indicates that some amount of trans-barcoding does indeed occur (Figure 11A). We estimate that approximately 50% of the protein is cis-barcoded, and the remaining 50% of the trans-barcoded protein is equally conjugated to both UCIs. Thus, in this two-component system, 25% of the TRIM21 protein is conjugated to the GAPDH UCI (Figures 16A-16C).

[0163] In a complex library setting, even when ∼50% of each protein is trans-barcoded, this by-product should be associated with low levels of randomly sampled UCIs. We tested this using a mock MIPSA library composed of a 100-fold excess of a second GAPDH clone, combined with a 1:1 mixture of the first GAPDH and TRIM21 clones (Figure 2B).

[0164] Establishment and deconvolution of a stochastically barcoded human ORFeome MIPSA library

[0165] The sequence-verified human ORFeome (hORFeome) v8.1 consists of 12,680 cloned ORFs that map to 11,437 genes in the Gateway Entry plasmid (pDONR223). 20 Five subpools of the library were generated, each consisting of approximately 2,500 similarly sized ORFs. Each of the five subpools was separately recombined and transformed into a pDEST-MIPSA UCI plasmid library to obtain approximately 10-fold ORF coverage (approximately 25,000 clones per subpool). Each subpool was assessed by Bioanalyzer electrophoresis, sequencing of approximately 20 colonies, and Illumina sequencing of the combined superpool. TRIM21 plasmid was spiked into the superpooled hORFeome library at 1:10,000, a typical library member. SS IP experiments were then performed on the hORFeome MIPSA library using sequencing as the readout. Read counts from all UCIs in the library with spiked-in TRIM21 are shown for SS IP relative to the average of eight mock IPs in Figure 11C. Indeed, SS autoantibody-dependent enrichment of TRIM21 (17-fold) was similar to the model system (Figure S11D). For a description of the analysis pipeline for the sequencing data, see the informatics analysis of MIPSA sequencing data in the Methods section.

[0166] Next, we established a system to create a UCI-ORF lookup dictionary using tagmentation and sequencing (Figure 3A). Sequencing of the 5'50 nt ORF inserts detected 11,076 of the 11,887 unique 5'50 nt library sequences. Of the 153,161 UCIs detected, 82.9% (126,975) were found to be associated with a single ORF (termed "monospecific UCIs"). Each ORF was uniquely associated with a median of 9 (range 0-123) monospecific UCIs (Figure 3B). Importantly, an ensemble of monospecific UCIs with consistent behavior can provide further strong support for the reactivity of their associated ORFs. A weak inverse correlation between UCI number and ORF size was observed, which most likely reflects lower efficiency of recombination of larger ORF-containing plasmids in pooled recombination reactions. After tabulation of read counts corresponding to each ORF, >99% of the represented ORFs were present within a 10-fold difference of the median ORF abundance (Figure 3C). Collectively, these data indicate that a uniform library of 11,076 probabilistically indexed human ORFs was established to define the lookup dictionary for downstream analysis. Figure 3D shows the UCI read counts of the SS IP relative to the average of the eight mock hORFeomes, as well as the 47 dictionary-decoded GAPDH monospecific UCIs (corresponding to the two GAPDH isoforms present in the IPS library) that appear along the y=x diagonal as expected. To avoid ambiguity, any UCIs associated with more than one ORF were excluded from further analysis.

[0167] Unbiased MIPSA analysis of autoantibodies associated with severe COVID-19 Several recent reports have documented increased autoantibody reactivity in severe COVID-19 patients. 21-25Because clinical metadata availability was incomplete, we used MIPSA with the human ORFeome library to unbiasedly identify autoreactivity in plasma from 55 severe COVID-19 patients, defined here only on the basis of hospitalization. For comparison, we used MIPSA to detect autoreactivity in plasma from 10 healthy donors and 10 non-hospitalized COVID-19 convalescent plasma donors (Table 2). As previously performed for phage immunoprecipitation sequencing (PhIP-Seq) analysis, each sample was compared to a set of eight "mock IPs" containing all reaction components except plasma, run on the same plate. Comparison to mock IPs accounts for bias in the library and background binding. The informatics pipeline (Methods) used to detect antibody-dependent reactivity yielded a median of 5 (range 2-9) false-positive UCI hits per mock IP. However, IP using plasma from severe COVID-19 patients yielded an average of 83 reactive proteins among severe COVID-19 patients, which was significantly more than the average of 64 reactive proteins among healthy pre-pandemic controls and significantly more than the average of 62 reactive proteins among individuals who had recovered after mild to moderate COVID-19 (p=0.02 and p=0.05, respectively, one-tailed t-test; Figure 4A).

[0168] Proteins were examined in severe COVID-19 IPs that had at least two reactive UCIs (within the same IP), were reactive in at least one severe patient, and were not reactive in two or more controls (in healthy or mild / moderate convalescent plasma). Proteins were excluded if they were reactive in one severe patient and one control. The 103 proteins that met these criteria are shown in the cluster map in Figure 4B. Fifty-one of the 55 severe COVID-19 patients showed reactivity to at least one of these proteins. Co-occurring protein reactivity in multiple individuals was noted, the majority of which lacked homology by protein sequence alignment. Table 4 shows that they were identified in the human autoantigen database AAgAtlas 1.0. 26Summary statistics are provided for these reactive proteins, including whether they are autoantigens as previously defined according to. Proteins were included if they had at least two reactive UCIs in at least one severe patient and were not reactive in more than one control (healthy or mild / moderate convalescent plasma). Proteins were not included if they were reactive in one severe patient and one control. Each row corresponds to a single UCI organized by protein in alphabetical order (gene symbol is provided to the left of the underline). Each column is an individual COVID-19 patient. If the UCI read count was not significantly enriched versus mock IP, it is reported as "1". If the UCI read count was significantly enriched versus mock IP, a fold change estimate (from EdgeR) is provided.

[0169] One notable autoreactive cluster (Table 4, cluster #5) contains 5'-nucleotidase, cytosolic 1A (NT5C1A), which is highly expressed in skeletal muscle and is the best-characterized autoantibody target in inclusion body myositis (IBM). Multiple UCIs linked to NT5C1A were significantly increased in 3 of 55 (5.5%) severe COVID-19 patients. NT5C1A autoantibodies were also significantly increased in IBM patients. 1 reported in up to 70% of patients with glaucoma, approximately 20% of patients with Sjögren's syndrome (SS), and up to approximately 5% of healthy donors. 27 Therefore, the prevalence of NT5C1A reactivity in severe COVID-19 cohorts is not necessarily elevated. However, we wondered whether MIPSA could reliably distinguish healthy donor and IBM plasma based on NT5C1A reactivity. We used plasma from 10 healthy donors and 10 IBM patients, the latter of which were analyzed by PhIP-Seq. 1The clear separation of patients from controls in this independent cohort suggests that UCI may indeed have utility in clinical diagnostic testing using either MIPSA-specific qPCR or library sequencing, which were closely correlated readouts (Figure 4C).

[0170] Neutralizing autoantibodies to type I and III interferon in patients with severe COVID-19

[0171] Neutralizing autoantibodies targeting type I interferon alpha (IFN-α) and omega (IFN-ω) are associated with severe COVID-19. 15,22,28All type I interferons except IFN-α16 are represented in the human MIPSA ORFeome library and annotated in the lookup dictionary. IFN-α4, IFN-α17, and IFN-α21 are analyzed as single ORFs because they are indistinguishable by the first 50 nucleotides of the ORF sequences that encode them. Two of the severe COVID-19 patients in this cohort (P1 and P2) (3.6%) showed dramatic type I IFN autoreactivity (49 and 46 type I interferon UCIs, across 11 separate ORFs corresponding to many IFN-α and IFN-ω; Figure 5A and Figure 5B). The extensive coreactivity of these proteins is likely due to their sequence homology (Figure 9). By requiring at least two reactive IFN UCIs to be considered positive, two further severe COVID-19 plasmas (P3 and P4) with detectable levels of IFN-α reactivity were identified, each with only two reactive IFN-α UCIs. Fifty percent of these four IFN-α autoreactive patients died, compared with approximately 30% in the remaining cohort. Interestingly, one further plasma (P5) did not precipitate any type I or type II interferon-derived UCIs, but precipitated five UCIs derived from the type III interferon IFN-λ3 (Figures 5C, 5D). This patient also died of COVID-19. No further interferon autoreactivity was detected among the severe COVID-19 patients. None of the healthy or non-hospitalized COVID-19 controls were positive for more than two interferon UCIs.

[0172] The ability of MIPSA with P2 plasma to neutralize both type I and type III interferons was evaluated. MIPSA was performed in triplicate in P2 plasma, yielding a high level of assay reproducibility, both in consistent detection of hits and a low coefficient of variation (average CV=22%) (Figures 19A, 19B). The linearity of the assay was evaluated by diluting P2 plasma 10-fold into healthy plasma and then performing MISPA again. The results demonstrate a consistent decrease in signal between reactivities (average 5.4-fold for reactive interferons) and loss of detection of some hits, especially ORFs with monoreactive UCIs.

[0173] Incubation of A549 human adenocarcinoma lung epithelial cells with 100 U / ml IFN-α or 1 ng / ml IFN-λ3 for 4 h in serum-free medium strongly upregulated the IFN-responsive gene MX1 by approximately 1,000-fold and approximately 100-fold, respectively. Preincubation of IFN-α2 with plasma P1, P2, or P3 completely abolished MX1 upregulation (Figure 5E). Plasma P4, which had the weakest IFN-α reactivity by MIPSA, only partially neutralized the cytokine. Neither HC nor P5 plasma had any effect on the response of A549 cells to IFN-α2 treatment. However, preincubation of IFN-λ3 cytokines with MIPSA-positive plasma, P2 and P5, eliminated the interferon response (Figure 5F). None of the other plasmas (HC, P1, P3, or P4) had any effect on the response of A549 cells to IFN-λ3. In comparison with titration curves using IFN-α2 and IFN-λ3 monoclonal antibodies, serial titrations using three replicates of patient P2 plasma showed circulating levels of these autoantibodies of ∼20 μg / ml and ∼100 ng / ml, respectively (Figures 11A, 11B). MIPSA analysis of serially diluted IFN-α2 mAbs demonstrated broad IFN-α cross-recognition but mutually exclusive binding of the mAbs to the appropriate type I or type III interferons (Figures 12A, 12B). Importantly, we noted that the loss of MIPSA detection sensitivity corresponded to equal or higher plasma dilutions at which the neutralizing activity of IFN-α2 and IFN-λ3 was also lost. Finally, titers of P2's autoantibodies showed at least a 10-fold preference for IFN-γ3 neutralization over IFN-γ1 neutralization (Figure 13). In summary, MIPSA-based autoantibody profiling of this severe COVID-19 cohort identified potently neutralizing IFN-α autoantibodies in 7.3% of patients and potently neutralizing IFN-λ3 autoantibodies in 3.6% of patients, with a single patient (1.8%) harboring both autoreactivities.

[0174] Then, a 90 aa human peptidome library 29We determined whether phage immunoprecipitation sequencing with MIPSA (PhIP-Seq) could also detect interferon antibodies in this cohort. PhIP-Seq detected IFN-α reactivity in plasma from P1 and P2, but to a much lower extent (Figure 5G). Two weaker IFN-α reactivity detected by MIPSA in plasma from P3 and P4 were both lost by PhIP-Seq. PhIP-Seq identified a single additional weakly IFN-α reactive sample that was negative by MIPSA (not shown). Both techniques detected type III interferon autoreactivity (directed only to IFN-λ3). We used PhIP-Seq data to narrow down the location of the dominant epitopes in these type I and type III interferon autoantigens (Figure 5H for IFN-α; amino acid positions 45-135 for IFN-λ3).

[0175] We assessed the prevalence of previously unreported IFN-λ3 autoreactivity in the general population and whether it may be increased among patients with severe COVID-19. PhIP-Seq was previously used to profile the plasma of 423 healthy controls, none of whom were found to have detectable IFN-λ3 autoreactivity. 30 These data suggest that IFN-λ3 autoreactivity is likely more frequent among individuals with severe COVID-19. This is the first report to describe neutralizing IFN-λ3 autoantibodies, thus providing a potentially novel pathogenic mechanism contributing to life-threatening COVID-19 in a subset of patients.

[0176] Consideration

[0177] Here, a novel molecular display technology is presented for full-length proteins that offers significant advantages over protein microarrays, PLATO, and alternative technologies. MIPSA utilizes self-assembly to produce libraries of proteins linked to relatively short (158 nt) single-stranded barcodes via a 25 kDa Halo tag domain. This compact barcoding approach is likely to have numerous applications not available to alternative display formats with bulky linked cargo (e.g., yeast, bacteria, viruses, phages, ribosomes, mRNA, cDNA). Indeed, conjugating minimal DNA barcodes individually to proteins, particularly antibodies and antigens, is a promising strategy for CITE-Seq. 31 ,LIBRA-seq 32 , and related methodologies have already proven useful in several settings. 33 On a proteome scale, MIPSA allows for unbiased analysis of protein-antibody, protein-protein, and protein-small molecule interactions, as well as for example hapten modification studies. 34 or protease activity profiling 35 It enables the study of post-translational modifications such as ribosomal transcription, transcriptional regulation, and transcriptional regulation. Key advantages of MIPSA include its high throughput, low cost, simple sequencing library preparation, inherent compatibility with PhIP-Seq, and stability of protein-DNA complexes (important for manipulating and storing display libraries). Importantly, MIPSA does not require specialized training or equipment, merely access to high-throughput sequencing equipment or facilities, and thus can be readily adopted by standard molecular biology laboratories.

[0178] Autoantibodies detected in severe COVID-19 patients using MIPSA

[0179] Neutralizing IFN-α / ω autoantibodies have been described in patients with severe COVID-19 disease and are presumed to be pathogenic 22These pre-existing autoantibodies, very rarely present in the general population, likely block the restriction of viral replication in cell culture and thus prevent resolution of the disease. This finding paves the way for identifying a subset of individuals at risk for life-threatening COVID-19 and suggests a therapeutic use of interferon-β in this patient population. In this study, MIPSA identified two individuals with broad reactivity to the entire family of IFN-α cytokines. Indeed, plasma from both individuals, plus two individuals with weaker IFN-α reactivity detected by MIPSA, strongly neutralized recombinant IFN-α2 in a lung adenocarcinoma cell culture model.

[0180] Type III IFNs (IFN-lambda, also known as IL-28 / 29) are cytokines with potent antiviral activity that act primarily at barrier sites. The IFN-lambdaR1 / IL-10RB heterodimeric receptor for IFN-lambda is expressed on lung epithelial cells and is important for the innate response to viral infection. Mordstein et al. found that IFN-lambda reduced the virulence and suppressed the replication of influenza virus, respiratory syncytial virus, human metapneumovirus, and Sars coronavirus (SARS-CoV-1) in mice. 36 IFN-λ has been proposed to exert much of its antiviral activity in vivo through stimulatory interactions with immune cells, rather than through the induction of an antiviral cellular state. 37 However, IFN-λ was not found to be effective in primary human bronchial epithelial cells. 38 , primary human airway epithelial cultures 39 , and primary human intestinal epithelial cells 40 In humans, IFN-λ autoantibodies have been found to strongly restrict SARS-CoV-2 replication. Together, these studies suggest multifaceted mechanisms by which neutralizing IFN-λ autoantibodies may exacerbate SARS-CoV-2 infection.

[0181] Among 55 severe COVID-19 patients, MIPSA detected two individuals with IFN-λ3-reactive autoantibodies. We tested the IFN-λ3 neutralizing capacity in the plasma of these patients and observed a near complete loss of cellular responses to the recombinant cytokine (Figure 5F). These data propose that IFN-λ3 autoreactivity is a novel potential pathogenic mechanism contributing to severe COVID-19 disease.

[0182] Casanova et al. did not detect any type III IFN neutralizing antibodies among 101 individuals with type I IFN autoantibodies tested. 22 In this study, one of the four IFN-α autoreactive individuals (P2, a 22-year-old male) also had autoantibodies neutralizing IFN-λ3. This co-reactivity is very rare and may therefore not be shown in the Casanova cohort. Alternatively, it is possible that different assay conditions show different detection sensitivities. Casanova et al. cultured A549 cells with 50 ng / ml IFN-λ3 without preincubation in plasma, whereas here A549 cells were cultured with 1 ng / ml IFN-λ3 after 1 h preincubation in plasma. Those readouts of STAT3 phosphorylation may also provide different detection sensitivities compared to upregulation of MX1 expression. Larger studies should determine the true frequency of these reactivities in severe COVID-19 patients and matched controls. Here, in a cohort of 55 patients with severe COVID-19, we report the detection of strongly neutralizing IFN-α and IFN-λ3 autoantibodies in 4 (7.3%) and 2 (3.6%) individuals, respectively. IFN-λ3 autoantibodies were not detected via PhIP-Seq in a larger cohort of 423 healthy controls collected pre-pandemic.

[0183] Exogenously administered type III interferon has been proposed as a treatment for SARS-CoV-2 infection. 39,41-45There are currently three ongoing clinical trials to test pegylated IFN-λ1 for efficacy in reducing morbidity and mortality associated with COVID-19 (ClinicalTrials.gov Identifiers: NCT04343976, NCT04534673, NCT04344600). One recently completed double-blind placebo-controlled trial, NCT04354259, reported a significant reduction of SARS-CoV-2 by 2.42 log copies / ml at day 7 (p=0.0041) among patients with mild to moderate COVID-19 in an outpatient setting. 46 Future studies will determine whether anti-IFN-λ3 autoantibodies are pre-existing or arise in response to SARS-CoV-2 infection and how frequently they cross-neutralize IFN-λ1. Based on the neutralization data from P2 (Figure 13) and sequence alignment of IFN-λ1 and IFN-λ3 (approximately 29% homology, Figure 9), cross-neutralization is expected to be rare, raising the possibility that patients with neutralizing IFN-λ3 autoantibodies may benefit from PEGylated IFN-λ1 treatment.

[0184] Clusters of uncharacterized autoreactivity were observed in multiple individuals, but it is unclear what function, if any, they may play in severe COVID-19. In larger studies, we predict that patterns of co-occurring reactivity, or reactivity to proteins with related biological functions, may ultimately define a new autoimmune syndrome associated with severe COVID-19.

[0185] Complementation of MIPSA and PhIP-Seq

[0186] Display technologies often complement each other, but may not be suitable for routine simultaneous use. MIPSA is more likely to detect antibodies against conformational epitopes on proteins that are well expressed in vitro than PhIP-Seq. This was exemplified by the robust detection of interferon-alpha autoantibodies via MIPSA that were less sensitively detected via PhIP-Seq. On the other hand, PhIP-Seq is more likely to detect antibodies against less abundant conformational epitopes contained within proteins that are either not present in the ORFeome library or may not be well expressed in cell-free lysates. Because MIPSA and PhIP-Seq naturally complement each other in these methods, we designed the MIPSA UCI amplification primers to be the same as those we used for PhIP-Seq. Because MIPSA and PhIP-Seq are stable in phage preparations, they can be easily performed together in a single reaction using a single set of amplification and sequencing primers. The compatibility of these two display modalities lowers the barrier to exploiting their synergies.

[0187] Variations of the MIPSA system

[0188] An important aspect of MIPSA involves the binding of a protein to its associated UCI in cis, compared to the binding of another library member's UCI in trans. Here, we utilized covalent binding via the Halo tag / Halo ligand system, but others may work as well. For example, the SNAP tag (a 20 kDa mutant of DNA repair protein O6-alkylguanine-DNA alkyltransferase) forms a covalent bond with a benzylguanine (BG) derivative. 47 Therefore, BG can be used to label the RT primer instead of the Halo ligand. The CLIP tag, a mutant derivative of the SNAP tag, binds an O2-benzylcytosine derivative and can also be adapted for MIPSA. 48 .

[0189] The rates of Halo tag maturation and ligand binding are important for the relative yields of cis vs. trans UCI conjugation. Studies by Samelson et al. have shown that the rate of Halo tag protein production is approximately four times faster than the rate of functional maturation of the Halo tag. 49 Considering that typical protein sizes are <1,000 amino acids in ORFeome libraries, these data predict that most proteins should be released from the ribosome before Halo tag maturation and therefore before cisHalo ligand conjugation can occur, thereby favoring undesired trans-barcoding. Here, however, it was observed that approximately 50% of protein-UCI conjugates were formed in cis, thereby enabling superior assay performance in a complex library setting. During optimization experiments, it was found that the rate of cis-barcoding was slightly improved by excluding a release factor from the translation mixture, which stalls ribosomes on their stop codons and allows Halo tag maturation to continue in close proximity to their UCI. Alternative approaches to promote controlled ribosome stalling may include stop codon removal / suppression or the use of dominant-negative termination factors. Ribosome release can then be induced via the addition of the chain terminator puromycin.

[0190] Since UCI cDNA is formed on the 5'UTR of IVT-RNA, eukaryotic ribosomes cannot scan from the 5'cap to the initiating Kozak sequence. Therefore, the MIPSA system described herein is incompatible with cap-dependent eukaryotic cell-free translation systems. However, if cap-dependent translation is desired, two alternative methods can be developed. First, the current 5'UCI system can be used if an internal ribosome entry site (IRES) is placed between the RT primer and the Kozak sequence. Second, UCI can be introduced instead at the 3' end of the RNA, provided that it prevents RT from extending into the ORF. In an extension of eukaryotic culture MIPSA, RNA-cDNA hybrids can potentially be transfected into living cells or tissues, where UCI protein formation can occur in situ, allowing for many additional applications.

[0191] ORF-associated UCIs can be materialized in a variety of ways. Here, we assigned probabilistically assigned indexes to the human ORFeome with approximately 10-fold representation. This approach has two main advantages: first, a single degenerate oligonucleotide pool is low cost; second, multiple independent measurements are reported by the ensemble of UCIs associated with each ORF. Here, libraries were designed to have UCIs with uniform GC content and therefore uniform PCR amplification efficiency. For simplicity, we chose not to incorporate unique molecular identifiers (UMIs) into the RT primers, but this approach is compatible with MIPSA UMIs and could potentially enhance quantification. One drawback of probabilistic indexing is the possibility of ORF dropout and thus the need for relatively high UCI representation, which increases the sequencing depth required to quantify each UCI and thus the overall per-sample cost. A second drawback is the need to build a UCI-ORFeome matching dictionary. Using short-read sequencing, we were unable to clarify the fraction of the library that is mostly composed of alternative isoforms. Using long-read sequencing technologies, such as PacBio or Oxford Nanopore Technologies, instead of or in addition to short-read sequencing technologies, may overcome incomplete disambiguation. In contrast to probabilistic barcoding, individual UCI-ORF cloning is possible, but costly and laborious. However, smaller UCI sets offer the advantage of lower sequencing costs per assay. Methodologies for cloning ORFeomes using Long Adapter Single Stranded Oligonucleotide (LASSO) probes have been developed previously. 50 LASSO cloning of ORFeome libraries naturally synergizes with MIPSA-based applications.

[0192] MIPSA readout by qPCR

[0193] A useful feature of well-designed UCIs is that they can also function as qPCR readout probes. The degenerate UCI designed and used here (Figure 1B) contains 18-nt base-balanced forward and reverse primer binding sites. Thus, the low cost and rapid turnaround time of qPCR assays can be exploited in combination with MIPSA. For example, incorporating assay quality control measures such as TRIM21 IP could be used to qualify a set of samples prior to a more costly sequencing run. Troubleshooting and optimization can be similarly facilitated by using qPCR as a readout rather than relying solely on NGS. qPCR testing for specific UCIs could also, in theory, offer enhanced sensitivity compared to sequencing, making it more suitable for analysis in a clinical setting.

[0194] conclusion

[0195] MIPSA is a self-assembling protein display technology that has important advantages over alternative approaches. It has the property that complementary technologies such as PhIP-Seq and MIPSA ORFeome libraries can be conveniently screened in the same reaction with a phage display library. The MIPSA protocol presented herein requires cap-independent cell-free translation, but future adaptations may overcome this limitation. Applications of MIPSA-based research include protein-protein, protein-antibody, and protein-small molecule interaction studies, as well as analysis of post-translational modifications. Here, we used MIPSA to detect known autoantibodies and discovered neutralizing IFN-λ3 autoantibodies, which, among many other potentially pathogenic autoreactivities (Table 4), may contribute to life-threatening COVID-19 in a subset of at-risk individuals.

[0196] Table 4: Reactive proteins in severe COVID-19 patients (continued on next page). Symbol, gene symbol; AAgAtlas, protein listed in AAgAtlas 1.0; #Severe, number of severe COVID-19 patients who showed reactivity to at least one UCI; #Controls, number of control donors (healthy or mild-moderate COVID-19 patients) who showed reactivity to at least one UCI; #Reactive_UCIs, number of reactive UCIs associated with a given ORF; hits_FCs, mean and range (min-max) of maximum hit fold changes per ORF observed among patients who showed reactivity; Cluster_ID, antigen cluster defined in Figure 4B. [Table 4-1] [Table 4-2]

[0197] References 1 Larman, HB et al. Cytosolic 5'-nucleotidase 1A autoimmunity in sporadic inclusion body myositis. Annals of neurology 73, 408-418, doi:10.1002 / ana.23840 (2013). 2 Xu, GJ et al. Viral immunology. Comprehensive serological profiling of human populations using a synthetic human virome. Science 348, aaa0698, doi:10.1126 / science.aaa0698 (2015). 3 Shrock, E. et al. Viral epitope profiling of COVID-19 patients reveals cross-reactivity and correlates of severity. Science 370, doi:10.1126 / science.abd4250 (2020). 4 Monaco, D. R. et al. Profiling serum antibodies with a pan allergen phage library identifies key wheat allergy epitopes. Nat Commun 12, 379, doi:10.1038 / s41467-020-20622-1 (2021). 5 Kingsmore, S. F. Multiplexed protein measurement: technologies and applications of protein and antibody arrays. Nat Rev Drug Discov 5, 310-320, doi:10.1038 / nrd2006 (2006). 6 Kodadek, T. Protein microarrays: prospects and problems. Chem Biol 8, 105-115, doi:10.1016 / s1074-5521(00)90067-x (2001). 7 Ramachandran, N., Hainsworth, E., Demirkan, G. & LaBaer, J. On-chip protein synthesis for making microarrays. Methods Mol Biol 328, 1-14, doi:10.1385 / 1-59745-026-X:1 (2006). 8 Rungpragayphan, S., Yamane, T. & Nakano, H. SIMPLEX: single-molecule PCR-linked in vitro expression: a novel method for high-throughput construction and screening of protein libraries. Methods Mol Biol 375, 79-94, doi:10.1007 / 978-1-59745-388-2_4 (2007). 9 Zhu, J. et al. Protein interaction discovery using parallel analysis of translated ORFs (PLATO). Nat Biotechnol 31, 331-334, doi:10.1038 / nbt.2539 (2013). 10 Liszczak, G. & Muir, T. W. Nucleic Acid-Barcoding Technologies: Converting DNA Sequencing into a Broad-Spectrum Molecular Counter. Angew Chem Int Ed Engl 58, 4144-4162, doi:10.1002 / anie.201808956 (2019). 11 Los, G. V. et al. Haloタグ: a novel protein labeling technology for cell imaging and protein analysis. ACS Chem Biol 3, 373-382, doi:10.1021 / cb800025k (2008). 12 Yazaki, J. et al. Haloタグ-based conjugation of proteins to barcoding-oligonucleotides. Nucleic Acids Res 48, e8, doi:10.1093 / nar / gkz1086 (2020). 13 Liu, Y., Sawalha, A. H. & Lu, Q. COVID-19 and autoimmune diseases. Curr Opin Rheumatol 33, 155-162, doi:10.1097 (2021). 14 Knight, J. S. et al. The intersection of COVID-19 and autoimmunity. J Clin Invest 131, doi:10.1172 / JCI154886 (2021). 15 Wang, E. Y. et al. Diverse functional autoantibodies in patients with COVID-19. Nature 595, 283-288, doi:10.1038 / s41586-021-03631-y (2021). 16 Bastard, P. et al. Autoantibodies neutralizing type I IFNs are present in ~4% of uninfected individuals over 70 years old and account for ~20% of COVID-19 deaths. Sci Immunol 6, doi:10.1126 / sciimmunol.abl4340 (2021). 17 Abers, M. S. et al. Neutralizing type-I interferon autoantibodies are associated with delayed viral clearance and intensive care unit admission in patients with COVID-19. Immunol Cell Biol 99, 917-921, doi:10.1111 / imcb.12495 (2021). 18 Mohammad, F., Green, R. & Buskirk, A. R. A systematically-revised ribosome profiling method for bacteria reveals pauses at single-codon resolution. Elife 8, doi:10.7554 / eLife.42591 (2019). 19 Gu, L. et al. Multiplex single-molecule interaction profiling of DNA-barcoded proteins. Nature 515, 554-557, doi:10.1038 / nature13761 (2014). 20 Yang, X. et al. A public genome-scale lentiviral expression library of human ORFs. Nat Methods 8, 659-661, doi:10.1038 / nmeth.1638 (2011). 21 Consiglio, C. R. et al. The Immunology of Multisystem Inflammatory Syndrome in Children with COVID-19. Cell 183, 968-981 e967, doi:10.1016 / j.cell.2020.09.016 (2020). 22 Bastard, P. et al. Autoantibodies against type I IFNs in patients with life-threatening COVID-19. Science 370, doi:10.1126 / science.abd4585 (2020). 23 Zuo, Y. et al. Prothrombotic autoantibodies in serum from patients hospitalized with COVID-19. Sci Transl Med 12, doi:10.1126 / scitranslmed.abd3876 (2020). 24 Casciola-Rosen, L. et al. IgM autoantibodies recognizing ACE2 are associated with severe COVID-19. medRxiv, doi:10.1101 / 2020.10.13.20211664 (2020). 25 Woodruff, M. C., Ramonell, R. P., Lee, F. E. & Sanz, I. Broadly-targeted autoreactivity is common in severe SARS-CoV-2 Infection. medRxiv, doi:10.1101 / 2020.10.21.20216192 (2020). 26 Wang, D. et al. AAgAtlas 1.0: a human autoantigen database. Nucleic Acids Res 45, D769-D776, doi:10.1093 / nar / gkw946 (2017). 27 Lloyd, T. E. et al. Cytosolic 5'-Nucleotidase 1A As a Target of Circulating Autoantibodies in Autoimmune Diseases. Arthritis Care Res (Hoboken) 68, 66-71, doi:10.1002 / acr.22600 (2016). 28 Gupta, S., Nakabo, S., Chu, J., Hasni, S. & Kaplan, M. J. Association between anti-interferon-alpha autoantibodies and COVID-19 in systemic lupus erythematosus. medRxiv, doi:10.1101 / 2020.10.29.20222000 (2020). 29 Xu, G. J. et al. Systematic autoantigen analysis identifies a distinct subtype of scleroderma with coincident cancer. Proc Natl Acad Sci U S A, doi:10.1073 / pnas.1615990113 (2016). 30 Venkataraman, T. et al. Analysis of antibody binding specificities in twin and SNP-genotyped cohorts reveals that antiviral antibody epitope selection is a heritable trait. Immunity 55, 174-184 e175, doi:10.1016 / j.immuni.2021.12.004 (2022). 31 Stoeckius, M. et al. Simultaneous epitope and transcriptome measurement in single cells. Nat Methods 14, 865-868, doi:10.1038 / nmeth.4380 (2017). 32 Setliff, I. et al. High-Throughput Mapping of B Cell Receptor Sequences to Antigen Specificity. Cell 179, 1636-1646 e1615, doi:10.1016 / j.cell.2019.11.003 (2019). 33 Saka, S. K. et al. Immuno-SABER enables highly multiplexed and amplified protein imaging in tissues. Nat Biotechnol 37, 1080-1090, doi:10.1038 / s41587-019-0207-y (2019). 34 Roman-Melendez, G. D. et al. Citrullination of a phage-displayed human peptidome library reveals the fine specificities of rheumatoid arthritis-associated autoantibodies. EBioMedicine 71, 103506, doi:10.1016 / j.ebiom.2021.103506 (2021). 35 Roman-Melendez, G. D., Venkataraman, T., Monaco, D. R. & Larman, H. B. Protease Activity Profiling via Programmable Phage Display of Comprehensive Proteome-Scale Peptide Libraries. Cell Syst 11, 375-381 e374, doi:10.1016 / j.cels.2020.08.013 (2020). 36 Mordstein, M. et al. Lambda interferon renders epithelial cells of the respiratory and gastrointestinal tracts resistant to viral infections. J Virol 84, 5670-5677, doi:10.1128 / JVI.00272-10 (2010). 37 Ank, N. et al. Lambda interferon (IFN-lambda), a type III IFN, is induced by viruses and IFNs and displays potent antiviral activity against select virus infections in vivo. J Virol 80, 4501-4509, doi:10.1128 / JVI.80.9.4501-4509.2006 (2006). 38 Busnadiego, I. et al. Antiviral Activity of Type I, II, and III Interferons Counterbalances ACE2 Inducibility and Restricts SARS-CoV-2. mBio 11, doi:10.1128 / mBio.01928-20 (2020). 39 Vanderheiden, A. et al. Type I and Type III Interferons Restrict SARS-CoV-2 Infection of Human Airway Epithelial Cultures. J Virol 94, doi:10.1128 / JVI.00985-20 (2020). 40 Stanifer, M. L. et al. Critical Role of Type III Interferon in Controlling SARS-CoV-2 Infection in Human Intestinal Epithelial Cells. Cell Rep 32, 107863, doi:10.1016 / j.celrep.2020.107863 (2020). 41 Galani, I. E. et al. Untuned antiviral immunity in COVID-19 revealed by temporal type I / III interferon patterns and flu comparison. Nat Immunol 22, 32-40, doi:10.1038 / s41590-020-00840-x (2021). 42 Felgenhauer, U. et al. Inhibition of SARS-CoV-2 by type I and type III interferons. J Biol Chem 295, 13958-13964, doi:10.1074 / jbc.AC120.013788 (2020). 43 O'Brien, T. R. et al. Weak Induction of Interferon Expression by Severe Acute Respiratory Syndrome Coronavirus 2 Supports Clinical Trials of Interferon-lambda to Treat Early Coronavirus Disease 2019. Clin Infect Dis 71, 1410-1412, doi:10.1093 / cid / ciaa453 (2020). 44 Andreakos, E. & Tsiodras, S. COVID-19: lambda interferon against viral load and hyperinflammation. EMBO Mol Med 12, e12465, doi:10.15252 / emmm.202012465 (2020). 45 Prokunina-Olsson, L. et al. COVID-19 and emerging viral infections: The case for interferon lambda. J Exp Med 217, doi:10.1084 / jem.20200653 (2020). 46 Feld, J. J. et al. Peginterferon lambda for the treatment of outpatients with COVID-19: a phase 2, placebo-controlled randomised trial. Lancet Respir Med, doi:10.1016 / S2213-2600(20)30566-X (2021). 47 Jongsma, M. A. & Litjens, R. H. Self-assembling protein arrays on DNA chips by auto-labeling fusion proteins with a single DNA address. Proteomics 6, 2650-2655, doi:10.1002 / pmic.200500654 (2006). 48 Gautier, A. et al. An engineered protein tag for multiprotein labeling in living cells. Chem Biol 15, 128-136, doi:10.1016 / j.chembiol.2008.01.007 (2008). 49 Samelson, A. J. et al. Kinetic and structural comparison of a protein's cotranslational folding and refolding pathways. Sci Adv 4, eaas9098, doi:10.1126 / sciadv.aas9098 (2018). 50 Tosi, L. et al. Long-adapter single-strand oligonucleotide probes for the massively multiplexed cloning of kilobase genome regions. Nat Biomed Eng 1, doi:10.1038 / s41551-017-0092 (2017). 51 Mohan, D. et al. Publisher Correction: PhIP-Seq characterization of serum antibodies using oligonucleotide-encoded peptidomes. Nature protocols 14, 2596, doi:10.1038 / s41596-018-0088-4 (2019). 52 Tuckey, C., Asahara, H., Zhou, Y. & Chong, S. Protein synthesis using a reconstituted cell-free system. Curr Protoc Mol Biol 108, 16 31 11-22, doi:10.1002 / 0471142727.mb1631s108 (2014). 53 Klein, S. L. et al. Sex, age, and hospitalization drive antibody responses in a COVID-19 convalescent plasma donor population. J Clin Invest 130, 6141-6150, doi:10.1172 / JCI142004 (2020). 54 Correction: Patient Trajectories Among Persons Hospitalized for COVID-19. Ann Intern Med 174, 144, doi:10.7326 / L20-1322 (2021). 55 Zyskind I, R. A., Zimmerman J, Naiditch H, Glatt AE, Pinter A, Theel ES, Joyner MJ, Hill DA, Lieberman MR, Bigajer E, Stok D, Frank E, Silverberg JI. SARS-CoV-2 Seroprevalence and Symptom Onset in Culturally-Linked Orthodox Jewish Communities Across Multiple Regions in the United States. JAMA Open Network In Press (2021). 56 Rose, M. R. & Group, E. I. W. 188th ENMC International Workshop: Inclusion Body Myositis, 2-4 December 2011, Naarden, The Netherlands. Neuromuscul Disord 23, 1044-1055, doi:10.1016 / j.nmd.2013.08.007 (2013). 57 Wei, Z., Zhang, W., Fang, H., Li, Y. & Wang, X. esATAC: an easy-to-use systematic pipeline for ATAC-seq data analysis. Bioinformatics 34, 2664-2665, doi:10.1093 / bioinformatics / bty141 (2018). 58 Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139-140, doi:10.1093 / bioinformatics / btp616 (2010). 59 brandonsie.github.io / epitopefindr / . [Table 5-1] [Table 5-2] [Table 5-3] [Table 5-4] [Table 5-5]

Claims

1. A method comprising the steps of: (a) transcribing a vector library into messenger ribonucleic acid (mRNA), said vector library encoding a plurality of proteins, each vector of said vector library being configured in a 5' to 3' direction: (i) a polymerase transcription start site; (ii) a barcode; (iii) a reverse transcription primer binding site; (iv) a ribosome binding site (RBS); and (v) a nucleotide sequence encoding a fusion protein comprising (1) a polypeptide tag and (2) a protein, wherein the polypeptide tag specifically binds to a ligand; (b) reverse transcribing the 5' end of the mRNA using a primer that binds upstream of the RBS, wherein the primer is conjugated to the ligand that specifically binds to the polypeptide tag of the fusion protein, forming a complementary deoxyribonucleic acid (cDNA) that includes the ligand, primer and a barcode; (c) translating the mRNA, wherein the ligand of the cDNA binds to the polypeptide tag of the fusion protein;

2. 2. The method of claim 1, wherein the vector library is nicked prior to step (a).

3. 2. The method of claim 1, wherein the vector further comprises (vi) an endonuclease site for vector linearization, and the vector library is linearized prior to step (a).

4. The method of any one of claims 1 to 3, wherein the barcode of the vector is flanked by binding sites for polymerase chain reaction (PCR) primers.

5. The method of any one of claims 1 to 4, wherein the barcode comprises binding sites for PCR primers.

6. The method of any one of claims 1 to 5, wherein the RBS comprises an internal ribosome entry site.

7. The polypeptide tag is a haloalkane dehalogenase or ６ - alkylguanine-DNA-alkyltransferase.

8. The method of any one of claims 1 to 7, wherein the polypeptide tag comprises a HALO tag and the ligand comprises a HALO ligand.

9. 1. A library of self-assembling protein-DNA conjugates, each protein-DNA conjugate comprising: (a) a cDNA comprising a barcode, the cDNA being conjugated to a ligand that specifically binds to a polypeptide tag; and (b) a fusion protein comprising the polypeptide tag and a protein of interest, the ligand being covalently bound to the polypeptide tag.

10. A method for studying protein-protein interactions, comprising the step of performing a pull-down assay of the library of claim 9 with a protein of interest.

11. A method for studying protein-small molecule interactions, comprising the step of performing a pull-down assay of the library of claim 9 with a small molecule.

12. A method comprising the step of immunoprecipitation of the library described in claim 9 using an antibody obtained from a biological sample.

13. 10. A method for identifying a target of a first small molecule, comprising the steps of: (a) incubating the library of claim 9 with a first small molecule that binds to the target; and (b) performing a pull-down assay of the library of step (a) with a second small molecule, wherein the first small molecule bound to its target blocks binding of the second small molecule.

14. A self-assembling protein-DNA composition comprising: (a) a cDNA comprising a barcode, the cDNA being conjugated to a ligand that specifically binds to the polypeptide tag; and (b) a fusion protein comprising the polypeptide tag and a protein of interest, the ligand being covalently bound to the polypeptide tag.

15. A composition for interferon therapy for treating a patient with severe COVID-19, comprising an autoantibody that neutralizes IFN-λ3 in a biological sample obtained from the patient.

Images