Molecular cloning method and vector thereof

Abstract

The present invention relates to a method for providing a multiplex RNA expression plasmid, comprising the steps of: providing a plasmid backbone fragment comprising a 5'backbone terminal and a 3 'backbone terminal; a plurality of n-insert fragments are provided, where each insert comprises a sequence encoding a transcribed RNA sequence, and one of the inserts additionally encodes an insert selection marker under the control of a promoter operable in a host cell. These fragments are inserted into the backbone by Gibson assembly. And transforming a Gibson assembled product into a bacterial host cell, and separating a connected plasmid product. The invention also relates to a library prepared by the method of the invention.

Description

Technical Field

[0001] This application claims the benefit of priority of European application No. EP21209377 filed on November 19, 2021, the contents of which are incorporated herein by reference.

[0002] The present invention relates to a method for providing a multiplex RNA expression plasmid library encoding single guide (sg) RNA molecules with full genome coverage. Another aspect of the present invention relates to a method for preparing a transfected cell library, each member of which comprises a member of a plasmid library, and to a transfected cell library obtained by the method. The present invention also relates to an expression plasmid library. Background Art

[0003] RNA interference or mutagen-mediated screens have greatly improved our understanding of biology and human health and transformed drug development. Recently developed CRISPR-mediated technologies have now greatly expanded the toolkit for genetic screens and now allow for gene knockout (CRISPRo) and activation (CRISPRa). Furthermore, studies have shown that CRISPR-based screens produce overlapping and distinct hits compared to RNA interference-based screens and that CRISPR-mediated gene perturbations are relatively more specific than RNA interference approaches. Therefore, genome-wide CRISPR-based gene perturbation libraries are essential for the identification of genes involved in biological processes on a global scale, especially in understanding disease.

[0004] In fact, many genome-wide CRISPR-based pooled libraries, including CRISPRo and CRISPRa, have been generated and have been very successful in screening cell-based phenotypes, including cell survival, reproduction or sensitivity to damage, cell function, and gene expression. However, these pooled libraries are not easily used for the study of cell non-autonomous phenotypes, in which the phenotype observed in one cell is caused by a genetic perturbation in another cell. Typical examples are protein secretion, mitochondrial unfolded protein response (UPRmt, a conserved transcriptional stress response in mitochondria), and glial-neuronal interactions, which are widely present and essential for brain function in normal and diseased conditions. In addition, the pooled libraries are of limited use for high-content screening methods, such as image-based neurite extension / contraction observations, regardless of whether the underlying mechanism is cell-autonomous or non-cell-autonomous.

[0005] The limited utility of pooled libraries can be circumvented by array libraries. However, compared with a variety of available and rapidly developing pooled CRISPR libraries, limited resources (including commercial and academic) of human CRISPR array libraries are available, and their effectiveness remains to be determined. In addition, certain shortcomings limit the usefulness of these CRISPR array libraries. First, libraries based on synthetic RNA are expensive and difficult to use in non-transfected cells. Secondly, these libraries, especially those based on lentivirus, rely mainly on single gRNA, which is relatively inefficient and unpredictable. Third, the knockout induced by a single gRNA can be homozygous or heterozygous, and in most cases, heterozygous knockout cannot cause functional consequences in protein-coding genes, resulting in false negative determinations. Fourth, a single gRNA is under the control of a determined promoter, and any cell type-specific reduction or silencing expression of the promoter will greatly affect the effectiveness of these libraries. Fifth, existing libraries are obtained using gRNA design algorithms based on reference genomes. However, the presence of frequent polymorphisms in the human genome can largely reduce the effectiveness of gRNAs in manipulating proprietary cells, such as proprietary induced pluripotent stem cells (iPSCs). Therefore, a new generation of highly active, robust, universal, and versatile CRISPR array libraries is urgently needed, especially in the study of phenotypes that cannot be addressed by empirically active pooled libraries.

[0006] CRISPR activation (CRISPRa) utilizes an endonuclease-inactive form of the CRISPR effector (dCas9), in which an added transcriptional activator (e.g., VP64, VPR, etc.) is fused to dCas9 or recruited to a single guide RNA (sgRNA) to activate the transcriptional expression of a gene of interest by targeting a specific sequence in the gene via sgRNA. The 4sgRNA / gene of the present invention synergizes the efficacy of a single sgRNA and provides an efficient and robust method for gene activation, and in principle, the present invention is compatible with all CRISPR activation systems in which a transcriptional activator is fused or recruited to dCas9, such as dCas9-VP64, dCas9-VPR, and dCas9-SunTag.

[0007] Based on the above prior art, the object of the present invention is to provide means and methods to provide a multiplex sgRNA expression library that allows targeting a large number of genes in a cell. This object is achieved by the subject matter of the independent claims of the present specification, and further advantageous embodiments are described in the dependent claims, examples, drawings and general description of the present specification. Summary of the invention

[0008] A first aspect of the present invention relates to a method for providing multiple RNA expression plasmids, comprising the following steps:

[0009] a. Providing a plasmid backbone fragment comprising a 5′ backbone end and a 3′ backbone end;

[0010] b. providing a plurality of n insertion fragments comprising a first insertion, a second insertion, ... an nth insertion, wherein

[0011] i. each insertion comprises a sequence encoding a transcribed RNA sequence, and each transcribed RNA sequence in an insertion is different from the transcribed RNA sequence in any other insertion,

[0012] ii. Each insert comprises a 5' insertion end and a 3' insertion end;

[0013] iii. Among them

[0014] The 5' first insertion end overlaps the 3' backbone end, and

[0015] · The 3′ first insertion end overlaps the 5′ second insertion end, and

[0016] The 3′ second insertion end overlaps with the 5′ nth insertion end, and the 3′ nth insertion end overlaps with the 5′ backbone end;

[0017] and wherein each overlap has a length of 15 to 40 bp, in particular a length of 17 to 22 bp,

[0018] iv. one of the inserts further encodes an insert selectable marker under the control of a promoter operable in a bacterial host cell;

[0019] c. In the Gibson assembly step, the plasmid backbone fragment and each of the plurality of insert fragments are contacted in an appropriate buffer, at a concentration of dNTPs and divalent cations that is favorable for the enzyme activity, in the presence of the following enzyme activities:

[0020] -5′ exonuclease activity,

[0021] -DNA-dependent DNA polymerase activity

[0022] -DNA ligase activity;

[0023] d. transforming the product of the Gibson assembly step into a bacterial host cell and growing the host cell in bacterial culture;

[0024] e. isolating the ligated plasmid product from the bacterial culture.

[0025] Another aspect of the present invention relates to a method for providing a multiplex RNA expression plasmid library, comprising performing the method according to the preceding aspect in a plurality of parallel reactions, wherein in each of said reactions, said transcribed RNA sequence is different.

[0026] A further aspect of the invention relates to libraries provided by the above methods, in particular libraries having the sequences described herein and in the sequence schemes.

[0027] The inventors experimentally confirmed the superiority and robustness of 4gRNA / vector in the efficiency of CRISPR-mediated gene activation and ablation. They developed an automated liquid phase plasmid assembly and cloning scheme, which assembled 4 gRNAs (driven by 4 different promoters) into a single vector, and utilized antibiotic resistance switching in the 4gRNA-vector backbone to eliminate the necessity of colony selection, enabling automation of high-throughput liquid phase cloning of libraries in a cost-effective manner. The updated gRNA design algorithm allows the selection of specific and universal gRNAs without targeting "hotspot" polymorphic regions in the human genome to maximize the synergistic effect of 4gRNA. The CRISPRa and CRISPRo libraries prepared by the inventors consist of 22,326 and 19,820 polyclonal plasmids, respectively, suitable for whole genome screening of human protein-coding genes. Therefore, the library provided according to the method of the present invention represents the next generation of improvements and provides the community with a powerful whole genome protein-coding gene perturbation resource. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 Shown are (A) increased gene activation by 4-sgRNA strategy. (B) Gene activation of test genes. (C) to (D) Gene ablation using 4-sgRNA strategy. (E) Digestion of pYJA5 vector. (F) Detection of recombination after 4-sgRNA transfection into HEK293 cells.

[0029] Figure 2 Shown are (A) high-throughput cloning protocol. (B) to (E) accuracy of sgRNA and tracrRNA region reads. (F) Viral titer of 4-sgRNA packaged in lentivirus. (G) sgRNA delivery rate. (H) Gene-dependent activation.

[0030] Figure 3 Shown are (A) general scheme for library generation. (B) Overlapping and non-overlapping combinations of sgRNAs.

[0031] Figure 4 Shown are (A) sgRNAs selected by the update algorithm. (B) Coverage of human protein-coding genes. (C) to (D) Selected sgRNAs selected for the library. (E) to (F) Targeting of common regions. (G) Ratio of spacer sgRNAs. (H) Targeting effect of CRISPRa sgRNAs on surrounding genes. (I) Off-site target impact.

[0032] Figure 5Shown are (A) 4-sgRNA mediated gene knockout.

[0033] Figure 6 The contribution of mutations and deletions in the plasmid and contamination is shown.

[0034] Figure 7 It is shown that the prevalence of correct plasmids decreases due to sequence homology between sgRNAs.

[0035] Figure 8 (A) to (B) 4-sgRNA combinations that avoid sharing any identical subsequences of 8 base pairs or more in length are shown. (C) to (F) Characteristics of the 4-sgRNA library as a whole.

[0036] Fig. 9 Characteristics of the 4-sgRNA library as a whole are shown.

[0037] Fig.10 Shown are (A) cloning scheme for generating high purity plasmids. (B) Scheme of antibiotic selection switch cloning method.

[0038] FIG. 11 shows (A) the empty backbone vector pYJA5 and (B) the 4sg-RNA vector.

[0039] Fig.12 A magnified scheme of the 4 sgRNA expression cassettes and selectable markers is shown.

[0040] Fig.13 A scheme of insertion and precursor vector assembly is shown.

[0041] Fig.14 Data illustrating the viral production capacity of 24 HEK293T monoclonal lines are shown. Clone 5 showed the most robust viral titer and was approximately 4 to 5 times higher than most other lines.

[0042] Fig.15 Shown are exemplary titers of lentivirus produced in 384-well plates by the methods described herein.

[0043] Fig.16 The workflow for plasmid transfection for lentiviral production in 384-well plates is shown. DETAILED DESCRIPTION

[0044] Terms and Definitions

[0045] For the purpose of interpreting this specification, the following definitions will apply and, where appropriate, terms used in the singular will also include the plural and vice versa. In the event of any conflict between any definition set forth below and any document incorporated herein by reference, the set forth definition shall prevail.

[0046] As used herein, the terms "comprising," "having," "containing," and "including," and other similar forms and grammatical equivalents thereof, are intended to have equivalent meanings and are open ended in that the item or items following any of these words are not intended to be an exhaustive list of such item or items, nor are they intended to be limited to the listed item or items. For example, an article "comprising" components A, B, and C may consist of components A, B, and C (i.e., contain only components A, B, and C), or may contain not only components A, B, and C, but also one or more other components. Thus, it is intended and understood that "comprising" and its similar forms and grammatical equivalents thereof include disclosure of embodiments that "consist essentially of" or "consist of."

[0047] Where a range of values ​​is provided, it is understood that, unless the context clearly dictates otherwise, each intervening value, to the tenth of the unit of the lower limit, between the upper and lower limits of that range and any other stated or intervening value in that stated range is encompassed within the disclosure, subject to any specifically excluded limit in the stated range. Where the stated range includes one or both limits, ranges excluding either or both of those included limits are also included in the disclosure.

[0048] Reference herein to "about" a value or parameter includes (and describes) variations with respect to that value or parameter itself. For example, description referring to "about X" includes description of "X".

[0049] As used herein, including in the appended claims, the singular forms "a," "or," and "the" include plural referents unless the context clearly dictates otherwise.

[0050] 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 (e.g., in cell culture, molecular genetics, nucleic acid chemistry, hybridization techniques and biochemistry). Standard techniques are used for molecular, genetic and biochemical methods (see generally Sambrook et al., Molecular Cloning: A Laboratory Manual, 4th ed. (2012) Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY and Ausubel et al., Short Protocols in Molecular Biology (2002) 5th Ed, John Wiley & Sons, Inc.) and chemical methods.

[0051] sequence

[0052] Sequences similar or homologous to the sequences disclosed herein (e.g., at least about 70% sequence identity) are also part of the invention. In some embodiments, the sequence identity at the amino acid level may be about 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or more. At the nucleic acid level, the sequence identity may be about 70%, 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or more. Alternatively, substantial identity exists when the nucleic acid fragment will hybridize to the complement of the chain under selective hybridization conditions (e.g., very high stringency hybridization conditions). The nucleic acid may be present in whole cells, cell lysates, or in partially purified or substantially pure form.

[0053] In the context of this specification, the terms sequence identity and sequence identity percentage refer to a single quantitative parameter representing a sequence comparison result determined by comparing two aligned sequences position by position. Sequence alignment methods for comparison are well known in the art. Sequence alignment for comparison can be performed by the local homology algorithm in Smith and Waterman's Advances in Applied Mathematics 2:482 (1981), by the global alignment algorithm in Needleman and Wunsch's J. Mol. Biol. 48:443 (1970), by Pearson and Lipman's Proc. Nat. Acad. Sci. 85:2444 (1988) similarity search method or by computerized implementation of these algorithms (including but not limited to: CLUSTAL, GAP, BESTFIT, BLAST, FASTA and TFASTA). Software for performing BLAST analyses is publicly available, for example, through the National Center for Biotechnology-Information (http: / / blast.ncbi.nlm.nih.gov / ).

[0054] One example for comparing amino acid sequences is the BLASTP algorithm using default settings: Expected value threshold: 10; Word length: 3; Maximum number of matches within the query region: 0; Matrix: BLOSUM62; Gap cost: Existence 11, Extension 1; Composition correction: Conditional composition score matrix adjustment. One such example for comparing nucleic acid sequences is the BLASTN algorithm using default settings: Expected value threshold: 10; Word length: 28; Maximum number of matches within the query region: 0; Match / mismatch score: 1.-2; Gap cost: linear. Unless otherwise indicated, the sequence identity values ​​provided herein refer to the values ​​obtained using the BLAST suite of programs (Altschul et al., Journal of Mol. Biol. 215: 403-410 (1990)) using the default parameters for the above-identified protein and nucleic acid comparisons, respectively.

[0055] Reference to identical sequences without specifying a percentage value means 100% identical sequences (ie, identical sequences).

[0056] General Molecular Biology: Nucleic Acid Sequencing, Expression

[0057] The term gene refers to a polynucleotide containing at least one open reading frame (ORF) that can encode a specific polypeptide or protein after being transcribed and translated. The polynucleotide sequence can be used to identify larger fragments or full-length coding sequences of genes related thereto. Methods for separating larger fragment sequences are known to those skilled in the art.

[0058] The term gene expression or expression, or alternatively the term gene product, can refer to either or both of the process of generating a nucleic acid (RNA) or generating a peptide or polypeptide or its product, and also refers to transcription and translation, respectively, or any intermediate process of regulating the processing of genetic information to produce a polypeptide product. The term gene expression can also be applied to transcription and processing of RNA gene products, such as regulatory RNA or structural (e.g., ribosomal) RNA. If the expressed polynucleotide is derived from genomic DNA, expression can include splicing of mRNA in eukaryotic cells. Expression can be measured at the transcription and translation levels, in other words, at mRNA and / or protein products.

[0059] In the context of the present specification, the term sgRNA (single guide RNA) relates to an RNA molecule capable of sequence-specific inhibition (or regulation in the case of CRISPRa-mediated activation) of gene expression via the CRISPR (clustered regularly interspaced short palindromic repeats) mechanism.

[0060] In the context of this specification, the term nucleic acid expression vector relates to a plasmid, a viral genome or an RNA, which is used to transfect (in the case of a plasmid or RNA) or transduce (in the case of a viral genome) a target cell with a specific gene of interest, or - in the case of an RNA construct being transfected - to translate the corresponding protein of interest from the transfected mRNA. For vectors operating at the transcription and subsequent translation levels, the gene of interest is under the control of a promoter sequence, and the promoter sequence is operable in the target cell, so that the gene of interest is transcribed constitutively or in response to a stimulus or depending on the state of the cell. In certain embodiments, the viral genome is packaged into a capsid to become a viral vector capable of transducing a target cell.

[0061] Detailed description of the invention

[0062] The first aspect of the present invention relates to a method for providing multiple RNA expression plasmids. The method comprises the following steps.

[0063] First, a plasmid backbone fragment comprising a 5' backbone end and a 3' backbone end is provided. In addition, a plurality of n insert fragments are provided. These insert fragments consist of a first insert, a second insert, and subsequent numbered inserts up to the nth insert.

[0064] Each insertion comprises a sequence encoding a transcribed RNA sequence, and each transcribed RNA sequence in the insertion is different from the transcribed RNA sequence in any other insertion. Each insertion comprises a 5' insertion end and a 3' insertion end, based on the numbering of the insertion (5' first insertion end, 5' second insertion end, 5' third insertion end, ... 5' nth insertion end). In this article, it should be understood that the plasmid backbone fragment is composed of double-stranded DNA, and therefore any linear segment of dsDNA has two 5'OH ends (which can be phosphorylated). However, in order to specify the end and direction in the DNA chain, only one chain is considered, and the term 5' insertion end and 3' insertion end refer to the end of the chain.

[0065] Inserts are synthesized for use in a subsequent Gibson assembly step (see Gibson et al., Nature Methods, Vol. 6, pp. 343-345 (2009)). Thus, the 5' first insertion end overlaps the 3' plasmid backbone end, and the 3' first insertion end overlaps the 5' second insertion end, and the 3' second insertion end overlaps the 5' third overlap (if present), and each subsequent 3' overlap of a lower named fragment overlaps the 5' overlap of the next higher dominating fragment (the nth insertion end overlaps), until the last (for n fragments, the nth) 3' nth insertion end overlaps the 5' backbone end. Each overlap has a length of 15 to 40 bp. In a specific embodiment, the overlap has a length of 17 to 22 bp.

[0066] One of the inserts additionally encodes an inserted selectable marker under the control of a promoter operable in a bacterial host cell.

[0067] In a subsequent Gibson assembly step, the plasmid backbone fragment is contacted with each of the plurality of insert fragments in an appropriate buffer, at concentrations of dNTPs and divalent cations that are favorable for the enzyme activity, in the presence of the following enzyme activities:

[0068] -5′ exonuclease activity,

[0069] -DNA-dependent DNA polymerase activity

[0070] -DNA ligase activity.

[0071] The product of the Gibson assembly step is then transformed (transferred) into bacterial host cells, the host cells are propagated in bacterial culture, and the ligated plasmid product is isolated from the bacterial culture.

[0072] Each transcribed RNA sequence is under the control of an RNA polymerase promoter sequence operable in a target cell, particularly a mammalian target cell. In certain embodiments, the promoter sequence is an RNA polymerase III specific promoter sequence. In certain embodiments, the promoter sequences controlling the expression of the transcribed RNA sequence in the plasmid are all different in the same plasmid (and each plasmid library member carries the same promoter). In certain specific embodiments, the promoter sequences controlling the expression of different transcribed RNA sequences in the plasmid are selected from hU6, mU6, hH1 and h7SK.

[0073] In certain embodiments, the promoter controlling the expression of the first transcribed sequence in the first fragment is included on the backbone adjacent to the attachment / insertion site of the first fragment. In certain specific embodiments thereof, the promoter controlling the expression of the first transcribed sequence is selected from the hU6 and mU6 promoters. This facilitates the use of the mU6 and hU6 promoters to control different transcribed sequences, and during the PCR generation of the fragments, there is no possibility of interference between the two based on their sequence homology.

[0074] This design makes the homology parts of Gibson assembly always occur on different sequences of the promoter, rather than on very similar tracrRNA sequences, thus avoiding recombination between the four tracrRNAs. This design avoids an important limitation of Gibson assembly, which prevents the assembly of very similar sequences using this method.

[0075] The selectable marker is part of one of the inserts. In the specific examples given herein, the selectable marker expression cassette is synthesized and integrated into insert1, please refer to the workflow diagram in the figure.

[0076] In certain embodiments, three fragments are inserted, in other words, the plurality of n inserted fragments consists of a first fragment, a second fragment, and a third fragment.

[0077] In certain embodiments, the plurality of insertions encode four transcribed RNA sequences. In those embodiments in which three fragments are inserted, one insertion comprises two transcribed RNA sequences.

[0078] In certain embodiments, the transcribed RNA sequences each comprise an sgRNA and a trcrRNA.

[0079] In certain embodiments, each of the sgRNAs is specific for a sequence contained in the same gene of the target cell.

[0080] In certain embodiments, a plasmid backbone fragment is obtained by propagating a precursor plasmid comprising a backbone selectable marker under backbone marker selection conditions, isolating the precursor plasmid and excising the backbone selectable marker to produce a plasmid backbone fragment.

[0081] Insertion selectable markers are distinct from backbone selectable markers.

[0082] In certain embodiments, the insert is provided by first synthesizing the insert sequence followed by PCR using a primer pair to introduce the sgRNA and create an overlap.

[0083] In certain embodiments, the plasmid backbone and ligated plasmid products can be packaged into retroviral vectors.

[0084] library

[0085] Another aspect of the present invention relates to a method for providing a multiplex RNA expression plasmid library, comprising performing the method according to any one of the above aspects and embodiments of the present invention in multiple parallel reactions, wherein in each of the reactions, the transcribed RNA sequence is different.

[0086] The inventors prepared two libraries based on the method of the present invention. The first library was designed for activating genome-wide human protein-coding genes with dCas9-VPR or any other CRISPR activator compatible with canonical tracrRNA (Chavez, A et al., Nature Methods 12, Pages 326-328.). The second library was designed for genome-wide knockout of these genes. The activation library can also be used to turn off gene expression with CRISPRoff ( et al., Cell 184, Issue 9, 29 April 2021, Pages 2503-2519).

[0087] In certain embodiments, each of said transcribed RNA sequences comprises a sgRNA, and for any one reaction, each of said sgRNAs is specific for a sequence in the same gene contained in said target cell, and for each of said reactions, said gene is different.

[0088] For CRISPRa, the inventors in some cases constructed more than one set of RNA transcripts to generate more than one plasmid targeting a specific gene, because the design according to the invention targets the major transcription start site (TSS) and some genes have more than one TSS.

[0089] In certain embodiments, at least 100, more particularly at least 1000 sgRNA sequences selected from SEQ ID NOs 64 to 170943 are expressed.

[0090] Another aspect of the invention relates to a multiplex RNA expression plasmid library obtained or obtainable by a method according to any one of the preceding aspects or embodiments of the invention.

[0091] Pooling equimolar amounts of array library plasmids can generate evenly distributed pooled libraries, which was not possible before but can now be achieved by the method of the present invention. Statistically, evenly distributed libraries can easily achieve true whole genome coverage, which cannot be achieved by other existing pooled libraries.

[0092] Pooled libraries are generated by cloning sgRNA libraries into backbone vectors in a single tube. Due to variations in ligation efficiency or / and the variable abundance of synthesized sgRNA oligos, the prevalence of ligation products for each sgRNA plasmid can vary across the genome. Preparation of pooled libraries from arrayed libraries (such as the 4sgRNA array libraries provided herein) can easily address this problem, since each plasmid is cloned individually and equal amounts of plasmids can be pooled to prepare pooled libraries with the same prevalence for each gene.

[0093] skeleton

[0094] The plasmid backbone fragment is obtained by digestion of the backbone plasmid with type IIS restriction enzymes and is characterized by carrying a first (backbone) antibiotic resistance selection gene, allowing the plasmid to be propagated and amplified in bacterial culture.

[0095] In certain embodiments, the backbone comprises an LTR (long terminal repeat) sequence element as a eukaryotic expression element, a CMV immediate early enhancer promoter, Fig.10 A and Fig.10 B One sgRNA promoter in the arrangement shown.

[0096] After separation and IIS type restriction enzyme digestion of the backbone plasmid, the antibiotic A (in an embodiment of the present invention: ampicillin) selection fragment is removed. The linear backbone is then Gibson assembled using multiple (n) insertions encoding different sgRNA-trcrRNA constructs, which target the same genetic sequence in the target cell within the same plasmid. In the applications envisioned by the inventors, the target cells are eukaryotic cells, particularly mammalian cells. However, the method is equally applicable to other cells; those skilled in the art recognize that this will require changing the promoter of the transcribed sequence from a eukaryotic promoter (RNA Pol III) to an appropriate prokaryotic promoter.

[0097] The first sgRNA promoter will be located at the 3′ backbone end, and the last sgRNA tracrRNA will be located at the 5′ backbone end, and they will be available for subsequent Gibson assembly with the insert.

[0098] Type II restriction endonucleases recognize asymmetric sequences and cleave DNA carrying these sequences, not at the recognition site as Type I endonucleases do, but at a distance from the recognition site. DNA ends can be designed to flank Type II restriction sites so that digestion of the fragments removes the enzyme recognition sites and generates complementary overhangs. Such ends can be joined seamlessly, producing a junction that lacks the original site or scar.

[0099] In certain embodiments, the Type II restriction enzyme is BbsI.

[0100] One of the inserts contains an inserted selectable marker, in the case illustrated in the examples, an antibiotic resistance marker, such as trimethoprim resistance. After Gibson assembly, the resulting plasmid carries the inserted selectable marker, and only bacteria containing the correctly assembled plasmid grow and amplify the desired plasmid. Incorrectly ligated products or any remaining backbone plasmid will not propagate under conditions favorable to the inserted selectable marker.

[0101] In certain embodiments, the ligated plasmid product is subsequently isolated following a Gibson assembly step.

[0102] In certain specific embodiments, the ligated plasmid product is not isolated after the Gibson assembly step. The product of the Gibson assembly step can be directly used to transform bacterial cells and then produce plasmid library members.

[0103] The ability to transform the ligation products without further preparation of the plasmids is one of the major advantages of the method according to the invention. Isolation or purification of the ligation products is not necessary in the method according to the invention.

[0104] In certain embodiments, after Gibson assembly, the Gibson assembly product is transformed, and bacteria transformed with the correctly assembled product grow due to antibiotic switching. Any bacteria transformed with unassembled or incorrectly assembled transformation products will not grow. Therefore, switching antibiotics to select bacteria is a key feature of the methods disclosed herein because it avoids colony picking, which is a common method for isolating the correct plasmid in classical cloning procedures.

[0105] Promoters particularly useful in the context of the present invention include the human and mouse U6 promoters (hU6, mU6, respectively), the human H1 promoter, and the human 7SK promoter, all of which are RNA polymerase III promoters. Pol III genes and sequences under the transcriptional control of such promoters are not polyadenylated and are therefore ideal for making small nuclear RNAs such as shRNA hairpins.

[0106] The selectable markers in the examples listed in the examples encode a trimethoprim antibiotic resistance gene and are incorporated between two sgRNA expression cassettes (each cassette includes a promoter, sgRNA, tracrRNA, and transcription end sequences). The selectable marker gene cassette includes a bacterial promoter and a marker gene coding sequence.

[0107] In some embodiments, the following architecture is selected for insertion;

[0108] - the first insert comprises a sequence encoding a first transcribed RNA sequence, a first tracrRNA sequence and a transcription termination sequence, followed by a promoter for a second RNA;

[0109] - the second insert comprises a sequence encoding a second transcribed RNA sequence, a second tracrRNA sequence and a transcription termination sequence, followed by a promoter for a third RNA;

[0110] - the (n-1)th (where n>=2, the total number of insertions n is N RNA sequences minus 1, i.e. n=N-1) insertion comprising a sequence encoding the (n-1)th transcribed RNA sequence, the (n-1)th tracrRNA sequence and a transcription termination sequence, followed by the promoter of the nth RNA;

[0111] The nth insertion comprises a sequence encoding an nth transcribed RNA sequence, an nth tracrRNA sequence, and a transcription termination sequence, followed by a promoter for the (n+1)th (or N)th RNA and a sequence for the (n+1)th (or N)th RNA;

[0112] Example

[0113] Example 1: Superiority and robustness of 4sgRNA in gene activation and ablation

[0114] To test the efficiency and robustness of the 4-sgRNA approach in CRISPR-mediated gene perturbations, the inventors cloned several 4-sgRNA expression plasmids and validated their efficiency for CRISPRa and CRISPRo in the human embryonic kidney cell line HEK293. For CRISPRa, three genes ASCL1, NEUROD1, and CXCR4 were tested, which showed low, medium, and high expression at baseline, respectively. The inventors found that single sgRNAs showed quite variable efficiency in the activation of these genes. On the other hand, the 4-sgRNA strategy significantly increased the degree of gene activation in a robust manner ( Figure 1 A). This is consistent with previous findings that three sgRNAs work more effectively than a single sgRNA. To further test the robustness and efficacy of the 4-sgRNA approach for gene activation, the inventors tested genes that were found to be difficult to activate in previous studies. The inventors found strong and robust gene activation for all genes tested ( Figure 1 B). The inventors also observed the superiority of the 4-sgRNA strategy in terms of gene ablation efficiency ( Figure 1 C and 1D). In addition, 4-sgRNA-mediated gene knockout resulted in the deletion of the genomic region between the cleavage sites ( Figure 5 A), which increases the likelihood of loss of target gene function. In summary, these results demonstrate the high efficiency and robustness of the inventors' 4-sgRNA / gene strategy for gene activation and ablation.

[0115] Example 2: Design of APPEAL cloning method

[0116] In traditional cloning methods, bacterial colony picking and verification are required because the growth of unwanted plasmid recombinants and the original vector backbone greatly reduces the abundance of the desired plasmid. This feature hinders the automation of large-scale array plasmid library generation and reduces throughput. To circumvent the shortcomings of traditional cloning strategies, the inventors developed APPEAL (Automated Liquid Phase Plasmid Assembly and Cloning) by utilizing antibiotic selection switching between the starting vector and the final desired vector to generate high-purity plasmids ( Figure 1 E and 5B). This eliminates the requirement for colony picking and validation of individual clones. The inventors' final vector includes four sgRNAs driven by four ubiquitously active type III RNA polymerase promoters (i.e., human U6, mouse U6, human H1, and human 7SK promoters). In addition, the inventors' 4-sgRNA vector includes puromycin and TagBFP selection, lentiviral packaging, and the PiggyBac (PB) transposon element ( Figure 1 E and 5B). These features contribute to the high efficiency and low recombination propensity of inventor libraries and make them very versatile resources.

[0117] These four sgRNAs were synthesized separately in the form of oligonucleotide primers containing 20 nucleotides of the prototype spacer sequence and the constant region. In three different PCR reactions, a single constant fragment amplification primer was used. The three polymerase chain reaction (PCR) amplicons obtained shared homologous overlapping ends (about 20 nucleotides in length) with each other and with the digested empty vector (pYJA5). The digested vector and the three amplicons were assembled together using the Gibson assembly cloning method to form the final 4-sgRNA vector (Gibson et al, 2009, Nat Methods 6, 343-345). During the cloning process, the antibiotics used for bacterial selection were switched from ampicillin (AmpR) of the starting vector (pYJA5) to trimethoprim (TmpR) of the final 4-sgRNA / vector (4sgRNA-pYJA5). This was achieved by removing the ampicillin resistance gene from the pYJA5 backbone (after restriction enzyme digestion at two BbsI recognition sites) and incorporating the trimethoprim resistance gene dihydrofolate reductase between the sgRNA1 and sgRNA2 cassettes in the 4-sgRNA vector ( Figure 1 E and 5B). Unlike previously established multiplex sgRNA cloning methods, our cloning method starts with BbsI digestion and purification of the starting vector, and all following steps do not require gel or column purification. This makes APPEAL an ideal solution for cost-effective large-scale liquid phase library generation.

[0118] The inventors performed a proof-of-concept test of the APPEAL method. Digestion of the pYJA5 vector and three fragment PCR amplifications generated the correct starting material for Gibson assembly ( Figure 1 F). After transformation, in order to verify the method, the inventors spread the transformants on agar plates and performed single colony PCR using primers flanking the 4-sgRNA region in the vector ( Figure 1 F). Virtually all colonies tested showed the correct size of 2.2 kilobases for the 4-sgRNA expression cassette ( Figure 1 F). In addition, no detectable recombination was observed after transfection of the 4-sgRNA plasmid into HEK293 cells ( Figure 1 G). This demonstrates the feasibility of applying the APPEAL method to liquid phase colony-free plasmid generation.

[0119] Example 3: Scaling APPEAL to 384-well cloning

[0120] The inventors have successfully expanded the APPEAL method from a single tube to a high-throughput process ( Figure 2A). Primer synthesis, PCR and Gibson assembly are performed in 384-well plates. The products of Gibson assembly are then used for transformation, bacterial culture and plasmid miniprep in deep-well 96-well plates. Therefore, the final products can be stored as bacterial glycerol stocks, plasmids, lentiviral particles and transposons for downstream applications ( Figure 2 A).

[0121] To test the quality of the plasmids generated by the APPEAL method, the inventors cloned a 384-well plate of 4-sgRNA plasmids with APPEAL and subjected the plasmid library to single-molecule long-read sequencing (Pacific Biosystems). The inventors amplified the microprepared plasmid with a pair of primers flanking the 4-sgRNA expression cassette and performed single-molecule real-time sequencing on amplicons with a length of 2.2 kilobases. Because this sequencing method has long reading capabilities, it is able to analyze plasmids at single-copy resolution. Because the mutation rate in the promoter region of the 4-sgRNA cassette is low (<5%), and isolated mutations in these regions can generally be expected to have no functional consequences, the inventors focused on the protospacer and tracrRNA regions, and found that most wells of the 384-well plate had more than 85% of reads with completely correct sgRNA and tracrRNA regions. Others showed that at least 2 to 3 sgRNA regions were 100% correct ( Figure 2 B to 2E). The inventors further analyzed plasmids that were not 100% correct to understand the contribution of mutations and deletions in these plasmids and found that both appeared in similar proportions ( Figure 6 The inventors also examined the frequency of cross-contamination of plasmids between wells of the same plate and found that less than 1% of the reads were contaminants ( Figure 6 ). In conclusion, scaling up the APPEAL method to a high-throughput format is feasible and the quality is acceptable for generating whole-genome CRISPR array libraries.

[0122] Example 4: Versatility of 4-sgRNA Vectors

[0123] The inventors' 4-sgRNA vectors are suitable for lentiviral packaging; however, the insertion of the 4-sgRNA expression cassette greatly increases the packaging size of the lentiviral particles. The inventors packaged the inventors' 4-sgRNA vectors into lentiviruses and robustly obtained >10 7 TU / ml titer ( Figure 2F). These viruses greatly increased the rate of sgRNA delivery, as measured by the proportion of tagBFP-positive cells by flow cytometry in non-transfectable cells, including the human lymphocyte-related cell lines THP-1 and ARH-77, the human neuroblastoma cell line GIMEN, the human glioblastoma cell line U251-MG, and patient-derived iPSCs ( Figure 2 G). The inventors further examined the gene activation efficiency in non-transfectable iPSC-derived neurons (iNeurons) using lentiviral-mediated co-delivery of dCas9-VPR and 4-sgRNA vectors, as the inventors encountered difficulties in performing FACS on iNeurons due to morphological changes after trypsinization. Figure 2 As shown in Figure 3, gene activation was quite efficient, although its extent was gene-dependent. Taken together, these results further demonstrate the versatility and usefulness of the inventors' 4-sgRNA vectors.

[0124] Example 5: Updated algorithms for universal, specific, and collaborative sgRNA selection

[0125] To enable arrayed CRISPR screening for gain-of-function and loss-of-function, the inventors decided to generate CRISPRa and CRISPRo array libraries for human protein-coding genes. Recently developed sgRNA design algorithms greatly increase the likelihood of obtaining active sgRNAs. The inventors decided to generate array libraries using sgRNAs characterized in existing widely used pooled libraries, i.e., for CRISPRa libraries, sgRNAs from Calabrese and hCRISPRa-v2, and for CRISPRo libraries, sgRNAs from Brunello and TKOv3 ( Figure 3 A). Common DNA polymorphisms in the human genome, such as single nucleotide polymorphisms (SNPs), may reduce the effectiveness of sgRNAs if the sgRNA affects the protospacer sequence or the protospacer adjacent motif (PAM). However, with the exception of TKOv3, most existing libraries do not consider DNA polymorphisms when selecting sgRNAs. This shortcoming may hinder the efficiency of using these libraries to screen patient-derived cells, such as patient-derived iPSCs. In addition, the recently developed GuideScan algorithm (http: / / www.guidescan.com) has been shown to be a good predictor of off-target effects of sgRNAs, showing a strong correlation with the unbiased whole-genome off-target assay GUIDE-Seq. Based on this recent data, an sgRNA is considered relatively specific if its GuideScan score exceeds 0.2.

[0126] In order to improve the above two parameters and select the top 4 sgRNAs for library construction, the inventors designed a customized sgRNA selection algorithm ( Figure 3 A). The efficacy of CRISPRa-mediated gene activation relies on sgRNAs targeting a relatively narrow window around a gene's transcription start site (TSS), and some genes have more than one TSS. To account for multiple TSSs, the hCRISPRa-v2 library was designed to target major TSSs separately if they are significantly spaced apart (>1 kilobase), and 10 sgRNAs were designed per TSS. For the human CRISPRa library, the inventors employed the same strategy to isolate major TSSs and included additional TSSs defined by the hCRISPRa-v2 library. The inventors then employed two filters to select sgRNAs: 1) Based on aggregated data from over 10'000 whole human genomes, any common polymorphisms (with a frequency exceeding 0.1%) should not affect the 20-nucleotide protospacer sequence, or the two G nucleotides of the PAM sequence ( http: / / db.systemsbiology.net / kaviar / ); 2) GuideScan score should be more than 0.2. Because some genes or TSSs do not have 4 sgRNAs that meet these requirements, the inventors used the CRISPick portal ( https: / / portals.broadinstitute.org / gppx / crispick / public ) complemented the above libraries using the same algorithm used to design sgRNAs for the Calabrese and Brunello libraries for CRISPRa and CRISPRo, respectively. Finally, all possible combinations of the four sgRNAs were ranked by their aggregated specificity scores, and the best combination was selected to minimize potential off-target effects ( Figure 3 A).

[0127] In addition to considering common DNA polymorphisms and predicted off-target effects, three additional points were considered when selecting sgRNAs for the library. First, because the inventors used the Gibson assembly method to generate their library, if two or more sgRNAs shared the same subsequence of 8 nucleotides or more, the prevalence of correct plasmids decreased due to recombination between the same sequences in the four sgRNAs ( Figure 7). Secondly, previous libraries were based on single sgRNAs that were selected based on their ranking in predicted targeting efficacy and off-target effects. However, many of the sgRNAs selected by these libraries for targeting specific genes are located in the same genomic region, differing by only a few nucleotides. This is particularly common in CRISPRa libraries because the target window of sgRNAs around the TSS is relatively narrow. The inventors aimed to investigate whether the proximity between the binding sites of the four sgRNAs might affect their activity, and whether overlapping or spaced sgRNAs should be preferred. The inventors tested six genes and compared 4-sgRNA combinations spaced at least 50 nucleotides apart with combinations that did not meet this criterion. Interestingly, the inventors observed that the use of four non-overlapping sgRNAs (spaced at least 50 nucleotides apart) resulted in significantly higher gene activation, suggesting that unhindered binding of the sgRNAs leads to further synergistic effects ( Figure 3 B). Third, there are some sgRNAs in previously published libraries that target other genes besides the intended target gene. For example, when paralogous genes share fragments of the same sequence, this can occur with additional perfect matching binding sites elsewhere in the genome (non-site non-intended target gene). It can also occur at a single intended binding site if more than one gene may be affected. In the case of CRISPRa, this can happen if two genes located on opposite strands of the genome share a promoter region (bidirectional promoter). Typically, sgRNAs are annotated only with the intended gene, so many users of the library may not be aware of this problem, and this affects the interpretation of the data. In the sgRNA selection process, the inventors avoided sgRNAs with multiple perfect matching binding sites as much as possible ( Figure 3 A) In the case of closely related or nearly identical gene families, for simplicity, the inventors still created a separate 4-sgRNA plasmid for each gene with its own unique Entrez gene identifier.

[0128] Example 6: Overview of CRISPRa and CRISPRo Libraries

[0129] like Figure 4 As shown in A, the sgRNAs of the two libraries selected by the updated algorithm were mainly from previously published libraries. Compared with a single pooled library with most sgRNAs, the whole genome library showed the broadest coverage of human protein-coding genes ( Figure 4 B). Since the inventors preferred highly specific sgRNAs (and avoided those with GuideScan scores below 0.2), the sgRNAs selected for the library showed higher predicted specificity ( Figure 4 C) without sacrificing its predictive power ( Figure 4D). Importantly, CRISPRa and CRISPRo libraries showed significant improvement in targeting universal regions of the genome by avoiding genetic polymorphisms ( Figure 4 E and 4F). Notably, the algorithm used for sgRNA selection greatly increased the proportion of sgRNAs that were at least 50 nucleotides apart (compared to simply selecting the first four sgRNAs) ( Figure 4 G), and thus potentially further increased the activity of both libraries, e.g. Figure 3 B. In addition, in the library, the inventors were able to completely avoid 4-sgRNA combinations that shared any identical subsequences of 8 base pairs or longer ( Figure 8 A and 8B), thereby ensuring minimal spacing for all genes. The inventors divided the library into mutually exclusive sub-libraries (see below) mainly based on the categories used by the pooled library hCRISPRa-v2.

[0130]

[0131]

[0132] As described in the sgRNA design section above, the inventors attempted to exclude sgRNAs with potential pleiotropic off-target effects. In some cases, targeting additional unintended genes is inevitable due to the structure of the human genome. In fact, the inventors observed that within a kilobase window around the TSS, about 20% of CRISPRa sgRNAs targeted additional protein-coding or non-coding genes; this proportion was almost the same in all libraries examined by the inventors (including CRISPRa libraries) ( Figure 4 H). If an sgRNA has more than one perfect match to its target sequence in the genome, it can lead to unintended gene targeting to locations far from the intended gene (off-site targets). In the CRISPRo library, such nonspecific sgRNAs were largely excluded, whereas in the CRISPRa library, the proportion of sgRNAs with off-site targets was comparable to that of the reference pooled library (likely due to the narrow target window around the TSS, which limits the pooling of eligible sgRNAs) ( Figure 4 I). For comparisons between libraries, only genes (and TSSs) common to all libraries were included to ensure a fair comparison; Figure 8 C to Figure 8 F and Fig. 9 Characteristics of the 4-sgRNA library as a whole are shown.

[0133] discuss

[0134] In this study, the inventors circumvented the relatively low and unpredictable efficiency of a single sgRNA in gene perturbation by combining four sgRNAs into a single vector. The newly developed APPEAL method for cloning 4-sgRNA vectors enables the generation of human genome-wide CRISPRa and CRISPRo libraries in a short time and in a cost-effective manner. A customized sgRNA selection algorithm ensures that the libraries target specific and universal regions of the human genome. Thus, the inventors have generated a next-generation highly active and robust whole-genome array library that can be easily adapted to many systems and should help meet needs that are not adequately covered by existing resources.

[0135] Efficacy of CRISPR-based genetic perturbations

[0136] Efficacy is one of the most important parameters of any gene perturbation method. In CRISPR-based methods, most perturbations rely on a single sgRNA. Although algorithms for predicting sgRNA activity have been improved, the efficacy of single sgRNA-based methods is still relatively low and unpredictable. Previous CRISPR libraries, especially pooled libraries, usually include several sgRNAs per gene in separate vectors, which can circumvent the problem of unpredictable efficiency of gene perturbations to a certain extent. However, the relatively low efficacy of a single sgRNA cannot be solved in this way. More importantly, the relatively low efficacy of a single sgRNA may often fail to cause functional consequences, and this will lead to false negative determinations when a hit gene is determined. These shortcomings hinder the goal of achieving a comprehensive understanding of the phenotype of interest and may result in the omission of important druggable targets using existing CRISPR libraries in pooled and array methods. In addition, multiple perturbation strategies based on single sgRNAs increase the workload and cost of CRISPR screening because the number of vectors far exceeds the number of targeted genes. The 4-sgRNA / vector approach greatly improves the efficiency and robustness of CRISPR-mediated gene activation and ablation. Therefore, the 4-sgRNA-based efficient and robust human array library is a next-generation resource for genome-wide CRISPR screening.

[0137] 4-sgRNA vector strategy

[0138] In the traditional cloning method, colony picking and complex procedures are required, which makes it unsuitable for the construction of 4-sgRNA vectors. In addition, the cloning strategies previously developed for multiple sgRNA expression in a single vector use complex steps, especially the need to gel extract the required DNA fragments between the cloning procedures, which makes these methods impractical for large-scale projects. The newly developed APPEAL method avoids colony picking and simplifies the cloning procedure, making the generation of array libraries based on 4-sgRNA / vectors feasible in an automatic and highly cost-effective manner. As far as the inventors know, the antibiotic switching-based cloning method APPEAL makes plasmid construction possible without the need for colony picking for the first time. In addition, in principle, this antibiotic switching-based cloning method can be applied to the generation of any large-scale library, even general molecular cloning. Therefore, the inventors' method can provide information for future methods of generating large-scale plasmid libraries.

[0139] Delivering multiple sgRNAs to the same cell can increase the possibility of off-target effects. Here, the inventors have developed their own updated sgRNA selection algorithm to adopt the most specific combination of four sgRNAs from existing, fully validated libraries and resources to minimize the possibility of off-target effects. However, further genome-wide systematic empirical validation of the off-target effects of 4-sgRNA vectors in the library will be valuable. However, although the inventors do not have a complete description of the precise off-target effects of the library, this does not reduce the effectiveness and usefulness of the library. In any phenotypic screening method, subsequent validation of the hit genes is required. Compared with existing libraries, the superiority and robustness of the library in gene perturbations should allow them to identify significantly more hits. In addition, there are many orthogonal methods for performing a second round of validation on hits identified with the library. Therefore, the combination of the inventor's library and orthogonal resources reduces the impact of possible off-target effects and enables genetic screening in a powerful and effective manner.

[0140] Versatility of 4-sgRNA-based libraries

[0141] The 4-sgRNA vector includes basic elements such as puromycin resistance and TagBFP, and is compatible with lentiviral packaging and PiggyBac-mediated transposon integration. Therefore, the inventor's library can be used as a plasmid in transfectable cells and lentiviral particles in non-transfectable cells (such as primary cells and patient-derived iPSCs and derivatives). In addition, the inventor's vector uses four different promoters-human U6, mouse U6, human H1 and human 7SK promoters, which are active in many cell types to drive the expression of four different sgRNAs. In addition to minimizing the possible recombination tendency between sgRNA expression cassettes, the design maximizes the activity of sgRNA in different cell lines, thereby minimizing the risk of reduced or missing expression of sgRNA driven by a single promoter (usually human U6 promoter). In addition, the inventor's internal sgRNA selection algorithm ensures that the sgRNA used in the inventor's library targets the most common region of the human genome by avoiding genetic polymorphism. This will contribute to the high efficiency of our libraries and make them a particularly useful resource for identifying genetic modifiers (especially correctors) of disease-associated phenotypes in patient-derived cells and organoids.

[0142] Validation of hit genes from pooled CRISPR screens and orthogonal validation of hits from RNA interference-based screens involve perturbations of individual genes, and typically use methods based on single sgRNAs. However, due to their unpredictability, the efficiency of individual sgRNAs needs to be verified before they can be used for hit validation. The inventors' libraries have high activity and redundancy, and therefore also provide very valuable resources for manipulating a selected set of genes. Although multiplexing of sgRNA expression vectors from pooled libraries is very challenging, it is relatively easy to multiplex array libraries into pooled libraries. Due to the high efficacy of 4-sgRNA vectors, it is also very attractive to pool array libraries into pooled CRISPRa and CRISPRo libraries.

[0143] In summary, the inventors aimed to maximize the versatility and potential usefulness of the library, and in principle, the library will benefit all areas of research, including academic and industry-based research.

[0144] Mutation and recombination in 4-sgRNA vectors

[0145] Cloning methods based on antibiotic switching eliminate the necessity for colony picking and enable high-throughput liquid-phase cloning of libraries in an automated fashion. Therefore, the composition of plasmids in each well of these libraries is polyclonal. Figure 2As shown in Figures B to 2E, the proportion of completely correct 4-sgRNA plasmids in defined wells is typically about 80%, and each of the four sgRNA regions is correct in about 90% of the plasmids. The remaining clones are 4-sgRNA plasmids, but show some mutations, mainly in the regions of sgRNA and tracrRNA. These mutations have two origins: a) mutations can be introduced during oligonucleotide synthesis; b) due to the fidelity limitations of Taq DNA ligase, some mismatches in double-stranded DNA overhangs are tolerated, which allows ligation of mismatched DNA during Gibson assembly cloning. The mutation rate of the sgRNA cassette is consistent with the findings of the Gibson assembly method, with mutations occurring in about 10% of the plasmids.

[0146] The inventors use four different promoters to drive the expression of four sgRNAs and four tracrRNA variants to minimize the tendency of recombination between sgRNA expression cassettes. In fact, the inventors observed a low prevalence of recombination between sgRNA expression cassettes (<10%). Importantly, most recombinant plasmids retain two or three complete sgRNA sequences, and should therefore remain functional. In addition, if multiple copies of the vector are delivered to cells (ie, by transfection or when high multiplicity infection is used), the remaining correct clones will compensate for a small portion of a single plasmid by recombination or mutation. In fact, the average percentage of completely correct prototype spacer sequences for each of the four sgRNAs is greater than 90%.

[0147] General cloning methods

[0148] In addition to the above, the present specification also covers any method that employs antibiotic switching or any way of selective switching between a precursor plasmid and the final desired plasmid from the method, which enables the avoidance of colony picking in molecular cloning, which can be applied to any large-scale library construction / generation and even small-scale normal molecular cloning.

[0149] The inventors emphasize that the methods of the present invention may be generalized to avoid any commercial adaptation of any method derived from the methods described herein.

[0150] Another aspect of the invention relates to a method for molecular cloning of a plasmid containing an insert from a parent plasmid, wherein

[0151] a. The parent plasmid comprises a first selection marker flanked by a recognition sequence sensitive to the endonuclease action (digestion, cleavage) of a type II restriction enzyme;

[0152] b. Cut (digest) the parent plasmid with a type II restriction enzyme to generate two fragments:

[0153] i. A parental plasmid backbone (single-stranded overlap) fragment having a 5′ backbone end and a 3′ backbone end (single-stranded end);

[0154] ii. the first resistance marker fragment;

[0155] c. ligating an insert fragment containing a second selection marker and comprising ends annealing (matching) to the parental plasmid backbone fragment, wherein the insert fragment is flanked by sequences of non-reconstructed type II recognition sites adjacent to the ends;

[0156] d. Transferring the ligation product into cells under conditions where only cells containing the second selection marker can propagate, and processing (preparing) the resulting cell suspension to obtain the plasmid containing the insert.

[0157] There are several problems that prevent the extension of this approach to high-throughput cloning by digesting the insert with restriction enzymes to generate complementary single-stranded ends that match the vector backbone: 1) If the insert comes from PCR, this will require an additional step of purification of the digestion product after digestion, otherwise DNA polymerases generally have exonuclease activity and will degrade the single-stranded ends; 2) If the insert comes from digestion of an existing vector, purification is also required, otherwise the insert can be subsequently ligated back to the vector from which it came, and this will hinder the antibiotic switching principle during the ligation step, because not only will the desired plasmid be grown under selection conditions, but the insert also has a selection marker; 3) In theory, there is a third option to obtain the insert, which is whole genome synthesis of these fragments followed by digestion with restriction enzymes, however, the cost will be significant.

[0158] The selectable marker switching method presented here has not been published and has never been used before for large-scale library generation. In addition, the concept of eliminating colony picking in molecular cloning procedures has not been published.

[0159] The method according to this last aspect can be considered as part of the overall invention provided herein. The process can be entered with a precursor plasmid or backbone fragment from which the first selection marker has been excised. For de novo cloning, one usually starts with a precursor plasmid and prepares a large number of digested backbones, and then for subsequent cloning, one can start with the excised backbone.

[0160] Examples - Materials and Methods

[0161] DNA constructs

[0162] DNA constructs used in the study, in addition to the 4-gRNA expression plasmid (whose construction is described separately below), included hCas9 (addgene #41815) and lentiCas9-Blast (addgene #52962), SP-dCas9-VPR (addgene #63798) and pXPR_120 (lenti-dCas9-VPR-Blast, addgene #96917), psPAX2 (addgene #12260), VSV-G (addgene #8454), and pYJA5.

[0163] The pYJA5 construct was generated by modifying the lenti-PB vector (Metzakopian et al., 2017, Sci Rep 7, 2244) (gift from Allan Bradley) in three steps. First, a synthetic DNA fragment including a fourth variant of the human U6 promoter and tracrRNA and an ampicillin resistance gene (β-lactamase expression cassette) was used to replace the DNA fragment flanked by the recognition sites of restriction enzymes MluI and AgeI in the lenti-PB vector. The two BbsI (Type II restriction enzyme) recognition sites flanking the β-lactamase expression cassette were introduced into the new fragment to facilitate the removal of the β-lactamase expression cassette. In the second step, the original ampicillin resistance (β-lactamase) expression cassette in the lenti-PB vector was removed between the two BspHI restriction enzyme recognition sites. After removal, the insertion of the 4-gRNA expression cassette containing the trimethoprim resistance gene (dihydrofolate reductase) achieved cloning based on antibiotic switching. In addition, all BsmBI recognition sites were mutated. The detailed sequences of pYJA5 and 4-gRNA-pYJA5 constructs are shown below:

[0164] Sequence of backbone pYJA5 vector (SEQ ID NO.001)

[0165] (See sequence scheme)

[0166] 4-gRNA-pYJA5 sequence (N 20 Represents sgRNA sequence) (SEQ ID NO.002):

[0167] (See sequence scheme)

[0168] The four sgRNA primer sequences 5′-3′ (N in sg1 primer Fwd, sg2 primer Fwd, and sg3 primer Fwd) 20 It is the sgRNA sequence, however, N 20 In the sg4 primer Rev should be the reverse complement sequence of the sgRNA):

[0169] sg1 primer Fwd (SEQ ID NO.003): ttgtggaaaggacgaaacaccGN 20 GTTTAAGAGCTAAGCTG

[0170] sg2 primer Fwd (SEQ ID NO.004): cttggagaaaagccttgtttGN 20 GTTTGAGAGCTAAGCAGA

[0171] sg3 primer Fwd (SEQ ID NO.005): gtatgagaccactctttcccGN 20 GTTTCAGAGCTAAGCACA

[0172] sg4 Primer Rev (SEQ ID NO.006): ATTTCTGCTGTAGCTCTGAAACN 20 Cgaggtacccaagcggc

[0173] Universal primer sequence 5′-3′:

[0174] mU6 Rev (SEQ ID NO.007): CAAACAAGGCTTTTCTCCAAGGGM Rev (SEQ ID NO.008): Cgggaaagagtggtctcataca

[0175] Constant template sequence

[0176] C1 sequence 5′-3′ (SEQ ID NO.009):

[0177] (See sequence scheme)

[0178] M5′-3′ (SEQ ID NO.010):

[0179] (See sequence scheme)

[0180] C2s 5′-3′ (SEQ ID NO.011):

[0181] (See sequence scheme)

[0182] Individual gRNAs were individually cloned into homemade pYJA4 vectors by a previously established method (Koike-Yusa et al., 2014, Nat Biotechnol 32, 267-273).

[0183] In silico 4gRNA library design

[0184] Bringing together existing libraries

[0185] To provide a starting point for guide RNA selection, the inventors collected gRNAs from previously published and validated libraries and tools, each of which uses its own algorithm to select gRNAs with high predicted on-target efficacy. The inventors included Calabrese (Sanson et al., 2018, Nat Commun 9, 5416) and hCRISPRa v2 (Horlbeck et al., 2016, Elife 5) libraries for CRISPRa, and TKOv3 (Hart et al., 2017, G3 (Bethesda) 7, 2719-2727) and Brunello (Doench et al., 2016, Nat Biotechnol 34, 184-191; Sanson et al., ibid.) libraries for CRISPRo. The inventors used gRNAs from the CRISPick tool (formerly known as GPP sgRNA Designer) (Doench et al., 2016, ibid; Hanna and Doench, 2020, Nat Biotechnol 38, 813-823; Sanson et al., ibid) to ensure optimal coverage of difficult-to-target and newly annotated genes (the website was accessed in April 2020 after being updated on March 20, 2020).

[0186] Gene definition

[0187] Common gene definitions for gRNAs from all sources were provided using Entrez gene identifiers. If the source library did not provide Entrez identifiers, the official gene symbols were mapped to Entrez IDs, and the genomic position was used to disambiguate the gene symbols when necessary. Genes not defined as protein coding by NCBI or Ensembl were excluded (according to annotation files ftp: / / ftp.ncbi.nih.gov / gene / DATA / GENE_INFO / Mammalia / Homo_sapiens.gene_info.gz and ftp: / / ftp.ensembl.org / pub / release-99 / tsv / homo_sapiens / Homo_sapiens.GRCh38.99.entrez.tsv.gz, both downloaded on March 25, 2020). The final library included 19,839 protein-coding genes for CRISPRa and 19,819 for CRISPRo; the difference in gene counts stems from the small number of genes that were present in only one pattern in the source libraries (e.g., highly polymorphic genes associated with adaptive immunity, such as the T cell receptor alpha locus (TRA) gene, which can be used for CRISPRa but not CRISPRo).

[0188] TSS Definition

[0189] To ensure good coverage of alternative transcripts and the wide applicability of CRISPRa libraries in a variety of cell lines, the inventors adopted the definition of alternative transcription start sites (TSSs) from the hCRISPRa-v2 library (Horlbeck et al., 2016, supra). The authors of the library used the FANTOM5 CAGE-Seq data set (FANTOM Consortium and the RIKEN PMI and CLST (DGT) et al., 2014, Nature 507, 462-470), supplemented by the Ensembl (Yates et al., 2020, Nucleic Acids Res. 48, D682-D688.) transcript model to define TSS positions; if the FANTOM5 score indicates significant transcriptional activity, and if they are more than one thousand bases apart from the primary TSS, additional TSSs are targeted by their own gRNA set. The inventors selected a separate set of four gRNAs for each TSS, treating multiple TSSs as separate genes. To group gRNAs by TSS, the inventors mapped gRNAs from all sources (including the first five gRNAs from GPP sgRNA Designer) to their genomic locations and iterated each gRNA starting from the lowest genomic coordinate; if the distance from one guide to the next exceeded 1000 base pairs, a new TSS group was defined. Additional TSSs were targeted only if valid combinations of 4 guides were available. 2311 genes included multiple TSSs (using 4803 four-guide combinations), while the remaining 17528 genes targeted a single TSS.

[0190] Avoid genetic polymorphism

[0191] For each gRNA, the inventors checked the overlap with frequent genetic polymorphism regions in two guanosine nucleotides of the prototype spacer sequence of 20 nucleotides or the adjacent motif (NGG) of the prototype spacer in the human population. The inventors avoided the gRNA whose target region contained any genetic polymorphism with a frequency greater than 0.1%. The variation frequency is from the Kaviar database (Glusman et al., 2011, Bioinforma. Oxf. Engl. 27, 3216-3217), which includes correction genome data on single nucleotide variants, indels and complex variants from more than 77,000 individuals (including more than 13,000 whole genomes). This data set (variants with only 3 or more occurrences, version 160204-hg38) was downloaded on August 7, 2019. The polymorphism frequencies in the Kaviar database are generally similar to those from TOPMED, gnomAD and 1000 genome projects.

[0192] Specificity score

[0193] To select a four-guide combination with minimal off-target effects, the inventors calculated the specificity score of each gRNA from the source library. The inventors used the method introduced by the authors of the GuideScan (Perez et al., 2017, Nat. Biotechnol. 35, 347-349) tool: for each guide, potential off-target sites were weighted by their CFD (cutting frequency determination) score (Doench et al., 2016, supra), and the CFD scores were summarized into a single score using the following formula: 1 / (1+sum of CFD scores from all off-target sites) (Hsu et al., 2013, Nat. Biotechnol. 31, 827-832). Because the pre-calculated GuideScan Cas9 database does not contain all gRNAs (it excludes those gRNAs with perfect matching or mismatching off-target sites in the reference genome), the inventors used GuideScan and CRISPOR (Concordet and Haeussler, 2018, Nucleic Acids Res. 46, W242-W245; Haeussler et al., 2018, Nucleic Acids Res. 46, W242-W245; Haeussler et al., 2018, Nucleic Acids Res. 46, W242-W245). al., 2016, Genome Biol. 17, 148) tools to annotate gRNAs. Local installations of these tools were used, and source code was downloaded on December 2020 (GuideScan version 2018-05-16 and CRISPOR version 4.97); the output of the local installation was confirmed to be identical to that of the web-based tools. When available, the GuideScan specificity score was used (considering up to 3 mismatches); otherwise, the CRISPOR specificity score was used (considering up to 4 mismatches). Using a detailed output file listing each off-target site, CRISPOR triple-mismatch (3MM) and quadruple-mismatch (4MM) specificity scores similar to those from GuideScan were calculated. GuideScan and CRISPOR specificity scores are highly correlated, but not identical, because The number of off-target sites identified for the same sequence was slightly different. When selecting gRNAs, the inventors avoided low specificity guides with 3MM scores below 0.2; this cutoff has recently been shown to have good predictive power for identifying gRNAs with significant off-target activity. However, in the case where all qualified gRNAs have a specificity score below 0.2, for example, when targeting genes that exist in multiple copies in the genome or belong to large gene families with many closely related paralogs and pseudogenes, this criterion must be relaxed. Finally, in order to select among all qualified four-guide combinations, the inventors calculated the total specificity score using the following formula: 1 / (1+sum of CFD scores from all four guides), and selected the combination with the highest score, indicating high predicted specificity.

[0194] Guide RNA spacing

[0195] To allow unhindered multiple binding to produce synergistic effects, the inventors aimed to select four gRNAs whose "cutting" positions were spaced at least 50 base pairs apart. However, for CRISPRa, the target sequence should be located within a window of about 400 base pairs upstream of the TSS for optimal activity, which is reflected in the selection of gRNAs in the source library; therefore, overlap is inevitable for some genes. On the other hand, for CRISPRo, overlap is often inevitable when targeting genes with very short coding sequences. In these cases, the inventors still aim to minimize the total number of overlaps between adjacent guides. In addition, all four guide combinations strictly follow another criterion: no two gRNAs are allowed to share the same subsequence of more than 7 base pairs. This is done primarily to minimize recombination events between identical regions during Gibson assembly of the plasmid. However, this also enforces a minimum spacing between the four selected guides.

[0196] Selection of four gRNAs

[0197] After integrating and annotating the gRNAs from the source libraries, the inventors selected a final combination of four gRNAs for each gene or TSS. First, gRNAs containing a stretch of four or more T nucleotides were excluded because this sequence can induce transcription termination. Next, all possible four-guide combinations for each gene were generated, and combinations that shared the same subsequence of more than 7 base pairs in length were excluded. Potential combinations were then ranked using a list of criteria applied in sequence; if multiple combinations were tied first, the next criterion in the list was used to make a decision. The criteria were as follows: 1) maximizing the number of gRNAs (from 0 to 4) that met certain minimum requirements—the gRNA could be mapped to a defined genomic position in the reference genome with an N(GG) PAM; there was no overlap with common genetic polymorphisms (>0.1%); the 3MM specificity score was at least 0.2; and only for CRISPRo, the guide met the criteria of Graf et al. (Graf et al., 2019, .Cell Rep. 26, 1098-1103.e3); 2) maximizing the number of gRNAs with exactly one perfect match position in the reference genome, 3) minimizing the number of overlaps between two adjacent gRNAs separated by less than 50 base pairs, 4) maximizing the number of sgRNAs derived from GPPs. Designer tool instead of previously published libraries, 5) for CRISPRa, minimize the number of gRNAs derived from the "supplemental 5" instead of the "top 5" gRNAs of the hCRISPRa-v2 library, and for CRISPRo, minimize the number of gRNAs derived from GPPs outside the top 10, and 6) maximize the total specificity score from all 4 guides. The highest ranked four-guide combination was selected. Since the total specificity score was the only quantitative criterion, it served as a tiebreaker and had the greatest impact on the selection of guides.

[0198] Sublibrary allocation

[0199] To facilitate the centralized screening of genomic subsets, the inventors divided the entire set of protein-coding genes into mutually exclusive sub-libraries. The inventors' two sub-libraries - transcription factors and secretory proteomes - are based on a recent publication that combines bioinformatics analysis with expert collation to obtain a comprehensive list of genes in these categories (Lambert et al., 2018, Cell 172, 650-665; Uhlén et al., 2019, supra). These lists are obtained from the supplementary data of the publication (for secretory proteomes) or the author's website (for transcription factors; hhumantfs.ccbr.utoronto.ca, database version 1.01). Ensembl gene IDs were translated into Entrez gene IDs, using HUGO gene symbols to eliminate ambiguity in one-to-many mappings of some genes. The third sublibrary was based on a list of G-protein coupled receptors, curated by the HUGO Gene Nomenclature Committee (HGNC) (Braschi et al., 2019, Nucleic Acids Res. 47, D786-D792) (https: / / www.genenames.org / cgi-bin / genegroup / download?id=139&type=branch, accessed on March 11, 2020). Seven additional thematic sublibraries were adopted from the hCRISPRa-v2 library (Horlbeck et al., 2016, supra): membrane proteins, kinases / phosphatases / drug targets, mitochondria / transport / movement, stress / protein homeostasis, cancer / apoptosis, gene expression, and unspecified. The first two of these thematic sublibraries were updated to include a small number of additional transmembrane receptors, transporters, kinases, and phosphatase sublibraries using Gene Ontology terms (exported on 25 March 2020 from BioMart (Smedley et al., 2009, BMC Genomics 10, 22)) and a list of membrane proteins provided by the Human Protein Atlas Project (Uhlén et al., 2015, Science 347, 1260419) (https: / / www.proteinatlas.org / search / protein_class:Predicted+membrane+proteins, accessed on 11 March 2020). If a gene belonged to more than one class, it was assigned to the first sublibrary (in the order in which they are listed in this section), and all remaining genes were added to the unassigned sublibrary.

[0200] Classification of unintended genetic perturbations

[0201] In addition to the intended target gene, some gRNAs are expected to disrupt other genes. In some cases, this occurs at loci that are different from the intended target site: for example, a gene family of very close paralogs can usually only be targeted with gRNAs that have multiple fully matched binding sites in the genome. However, in most cases, this involves a single locus - the intended binding site - where the gRNA can disrupt more than one gene. In the case of CRISPRa, the same promoter region is often shared by two genes located on opposite strands of the chromosome, so that their transcription start sites (TSS) are only a few hundred base pairs apart. In this case, a guide RNA that effectively activates one gene will inevitably also activate another gene. As a guide for library users, and to help interpret hit genes, the inventors annotated gRNAs with a complete list of all genes they target. To summarize this phenomenon across the entire library, the inventors classified each gRNA as 1) targeting only the intended gene, 2) targeting an unexpected gene, but at a single location, or 3) targeting an unexpected gene at another location. For this analysis, two perfectly matched gRNA binding sites were considered to target an unintended gene at the same location if they had any target genes in common (this is particularly relevant for gRNAs targeting pseudoautosomal regions of chromosomes X and Y).

[0202] Annotation of unexpected target genes

[0203] To annotate each gRNA with all its potential target genes, a TSS position database was constructed by merging the FANTOM5 dataset (upgraded to the hg38 genome (Abugessaisa et al., 2017, Sci. Data 4, 170107), version 3) with data from BioMart (Smedley et al., 2009, supra) (exported on March 25, 2020), using Entrez gene IDs as common identifiers. Similarly, data on coding sequences (CDS) and exon positions were compiled from BioMart, the "TxDb.Hsapiens.UCSC.hg38.knownGene" Bioconductor package (version 3.10.0) and GENCODE (Frankish et al., 2019, Nucleic Acids Res. 47, D766-D773) annotation data (version 33), and the position data were merged using Entrez gene identifiers (if available) or Ensembl gene identifiers. Genes annotated as pseudogenes or with unclear classification were excluded from further analysis. For CRISPRa, perfect matching sgRNA binding sites within a 1000 base pair window around the TSS were considered. For CRISPRo, the sgRNA cutting site must be within the coding sequence (CDS) of protein-coding genes, or within the exons of noncoding RNAs.

[0204] Predicting missing annotations

[0205] In the case of CRISPRo, when four gRNAs are active in the same cell, multiple closely spaced double-strand breaks typically result in the loss of DNA segments between the gRNA cutting sites. Therefore, in addition to annotating a single gRNA, the inventors also determined which genes were affected by the expected deletions—the segments between the first and last cutting sites. The inventors also considered deletions induced by (perfectly matched) off-target binding sites. Because deletions may be less likely to occur if the cutting sites are far apart, the inventors imposed a maximum distance of one megabase between the cutting sites, making multiple expected deletions (or separated cutting sites) on the same chromosome possible.

[0206] In silico comparison of CRISPR libraries

[0207] To compare the in silico properties of existing libraries and the 4sg library, the top four guides for each gene were selected. While the Brunello (Doench et al., 2016, supra; Sanson et al., 2018, supra) and TKOv3 (Hart et al., 2017, supra) libraries were designed to contain four sgRNAs per gene, the authors divided the Calabrese library (Sanson et al., 2018, supra) into group A and group B, each containing three sgRNAs per gene. To define the top four sgRNAs, the sgRNAs from group A were supplemented with randomly selected sgRNAs from group B. For hCRISPRa v2 (Horlbeck et al., 2016, supra) and GPP libraries, the four highest-ranked gRNAs were selected (using the "selection order" column in the output of the GPP sgRNA design tool). Because the genes covered by the libraries are different, and because different genes differ in the availability of potential gRNAs with high predicted activity and specificity, only genes present in all libraries were used for benchmarking. In addition, for genes in which the 4sg and hCRISPRa v2 libraries included more than one transcription start site (TSS), only gRNAs targeting the primary TSS were included, defined as the TSS with the highest score in the FANTOM5 dataset, or - if data for the gene were not available - the most upstream TSS. To compare the expected number of gRNA binding sites affected by genetic polymorphisms, the frequency of the most common polymorphisms overlapping with each sgRNA was summarized. This is a conservative estimate because SNPs with a frequency below 0.1% were excluded. In addition, in the case of multiple single nucleotide polymorphisms (SNPs) overlapping with a gRNA, only the most frequent SNP was considered. Because linkage disequilibrium between SNPs affecting the same sgRNA is very likely, an accurate estimate of the total probability of all overlapping polymorphisms will require access to the individual sequencing data that underlies the SNP database.

[0208] Software and Code

[0209] Annotation and selection of gRNAs for library design were performed using the R statistical programming environment (R Core Team, 2020), version 3.6.3 and the Bioconductor suite (Huber et al., 2015, Nat. Methods 12, 115-121), version 3.10. The source code is available upon request and will be made publicly available upon final publication of the manuscript.

[0210] PacBio Next Generation Sequencing

[0211] Barcoding, amplification, and long-read sequencing

[0212] In order to evaluate the frequency of mutation, recombination and deletion in the polyclonal population of 4sg plasmid, single molecule long read sequencing was performed. The plasmid was amplified using a barcoded primer that uniquely identifies each target gene using polymerase chain reaction (PCR). In the pilot sequencing run of the inventors, this was achieved using a combination of 16 different forward primers and 24 different reverse primers (distinguishing the rows and columns of 384-well plates, respectively). The length of the amplified region was 2225 base pairs, covering the entire 4sg expression cassette (containing all 4 promoters, guide RNA and tracrRNA sequences, and trimethoprim resistance elements), and flanked by two 10-bp paired barcode sequences. Amplicons from all wells were pooled and single molecule real-time (SMRT) sequencing data was generated using a PacBio Sequel instrument.

[0213] Processing of long-read sequencing data

[0214] Using default parameters, PacBio SmrtLink software was used to perform circular consensus read (CCS) consensus calls, and only consensus reads with at least 5 full-pass subreads and estimated read accuracy ≥ 99.9% were retained. Barcode demultiplexing was also performed using SmrtLink software, and in order to minimize incorrect well assignments, reads with barcode scores below 60 or scores leading below 30 were omitted. Reads were further filtered using custom scripts to ensure that each barcode met the following criteria: the barcode sequence must be complete and completely correct, or the sequence length must be at least 8bp and flanked by a completely correct 20-bp flanking constant region (representing the constant region of the primer used for PCR amplification). These additional steps proved to be necessary to ensure that only complete reads (containing forward and reverse primer sequences) were retained, and truncated reads whose terminal sequences were incorrectly interpreted as truncated barcodes were excluded. Finally, consensus reads with an average Field quality score of less than 85 per base were excluded (the highest average Field quality score achievable was 93). In the pilot sequencing run, 78,351 consensus reads were retained after filtering, with an average of 204 reads per well (range 63 to 1098).

[0215] Consensus read analysis

[0216] In order to quantify the percentage of correct guide RNA sequences and identify contamination from other wells, the gRNA+tracrRNA sequence of each read was searched in the forward and reverse directions, and all complete matches were counted. To further characterize incorrect sequences, each consensus read was aligned with the corresponding barcode reference sequence of the well using the "pairwise alignment" function of the Biostrings R / Bioconducto package (version 2.54.0) using default parameters. The region corresponding to the reference gRNA+tracrRNA sequence was then extracted from the aligned reads, and each sequence was classified as a) completely correct, b) contamination (if it is a complete match of the gRNA sequence from another well), c) large deletions (if >50% of the aligned sequence is composed of gaps), or d) some other mutations.

[0217] APPEAL high-throughput library generation

[0218] Oligo synthesis

[0219] The 20-nucleotide sgRNA sequence was incorporated into an oligonucleotide sequence with an additional constant sequence and synthesized in a 384-well plate using the high affinity purification (HAP) purification method of Sangon Biotech (China). The oligonucleotide sequence of sgRNA1 (gRNA1 sequence, N20sg1) is: 5′-ttgtggaaaggacgaaacaccGN20sg1GTTTAAGAGCTAAGCTG-3′ (SEQ ID NO.003); the oligonucleotide sequence of sgRNA2 (gRNA2 sequence, N20sg2) is: 5′-cttggagaaaagccttgtttGN20sg2GTTTGAGAGCTAAGCAGA-3′ (SEQ ID NO.004); the oligonucleotide sequence of sgRNA3 (gRNA3 sequence, N20sg3) is: 5′-gtatgagaccactctttcccGN20sg3GTTTCAGAGCTAAGCACA-3′ (SEQ ID NO.005); and the sgRNA4 (the reverse complementary sequence of gRNA4, N20crsg4) oligomer sequence is 5′-ATTTCTGCTGTAGCTCTGAAACN20crsg4Cgaggtacccaagcggc-3′ (SEQ ID NO.006). The oligonucleotides were diluted with ultrapure water to a working concentration of 4 μM.

[0220] Three-fragment polymerase chain reaction (PCR)

[0221] A total of 10 μL PCR reactions per well were performed in 384-well plates.

[0222] The C1 fragment (amplicon size 761 bp) PCR mixture was prepared as follows:

[0223]

[0224]

[0225] 9.5 μL of the mixture was aliquoted into each well of a 384-well plate, and 0.5 μL of sgRNA1 primer (at a concentration of 4 μM) was added to each well and mixed.

[0226] The M fragment (amplicon size 360 ​​bp) PCR mixture was prepared as follows:

[0227]

[0228] 9.5 μL of the mixture was aliquoted into each well of a 384-well plate, and then 0.5 μl of sgRNA2 primer (at a concentration of 4 μM) was added to each well and mixed.

[0229] The C2s fragment (amplicon size 422 bp) PCR mixture was prepared as follows:

[0230]

[0231] 9 μL of the mixture was aliquoted into each well of a 384-well plate, and then 0.5 μL of sgRNA3 primer (at a concentration of 4 μM) and 0.5 μL of sgRNA4 primer (also at a concentration of 4 μM) were added to each well and mixed.

[0232] The Integra ViaFlo 384-well pipetting system was used for all 384-well liquid handling. All PCR plates were sealed and centrifuged at 2000 rpm for 2 minutes and placed in a thermal cycler with the following program: preheat lid at 99°C; initial denaturation at 98°C for 30 seconds, 36 cycles including 10 seconds at 98°C, 30 seconds at 60°C and 25 seconds at 72°C, and a final extension at 72°C for 5 minutes, followed by cooling to 20°C. All PCR products were then diluted with 9 μL of ultrapure water for subsequent Gibson assembly. The success of PCR on each plate was confirmed by DNA agarose gel electrophoresis of several random samples on the plate.

[0233] Gibson Assembly

[0234] The three fragment PCR products were assembled into the pYJA5 vector by Gibson assembly in a 384-well plate. The reaction mixture was as follows:

[0235] Volume Reagents

[0236] 2μL C1 amplified fragment (estimated to be about 16ng / μL)

[0237] 1μL M amplified fragment (estimated to be about 16ng / μL)

[0238] 1μL C2s amplified fragment (estimated to be about 20ng / μL)

[0239] 1 μL pYJA5 BbsI-digested purified vector, diluted to 120 ng / μL

[0240] 5 μL 2× Homemade HiFi Gibson Master Mix

[0241] The mixture was incubated at 50°C for 1 hour in a thermal cycler and then used to transform competent cells or immediately stored at -20°C.

[0242] Transformation and bacterial storage

[0243] Transform in 96-well deep well plates (2.3mL, Axygene P-DW-20-C) in a cold room. 5 μL (per well) Gibson mixture from 384-well plates is transferred to four 96-well plates and rotated to the bottom of each well. 50 μL (per well) homemade competent cells (NEB stable competent cells) are distributed and mixed twice with Gibson mixture. The plate is then kept immersed in ice for 30 minutes. Heat shock is carried out at 42 ° C for 30 seconds by placing the plate in a water bath. The plate is put back on ice for 5 minutes. Then 300 μL homemade SOC culture medium (0.5% yeast extract, 2% tryptone, 10mM NaCl, 2.5mM KCl, 10mM MgCl , 10mM MgSO , 20mM glucose) is added to the plate, and incubated for 1 hour at 37 ° C using a thermal oscillator at 900 rpm. Then, 900 μL (per well) of TB medium (TerrificBroth medium) (https: / / openwetware.org / wiki / Terrific_Broth) containing 15 μg / mL trimethoprim and 15 μg / mL tetracycline was added to the transformation mixture and incubated at 30°C with shaking at 900 rpm for 40 to 48 hours.

[0244] Bacteria were then stored at -80°C in 96-well plates (300 μL final stock volume) and 384-well plates (150 μL final stock volume) at a final concentration of 16.7% (v / v) glycerol.

[0245] Magnetic bead-based 96-well plasmid microprep

[0246] 50 μl of bacteria transformed with the Gibson assembly product were transferred to 1.2 mL of TB medium (with 15 μg / mL trimethoprim and 15 μg / mL tetracycline in a 96-well deep-well plate), and the bacteria were immediately stored and grown at 900 rpm for 40 to 48 hours at 30° C. The bacteria were then subjected to an in-house magnetic bead-based plasmid miniprep procedure that employed a standard plasmid miniprep protocol (Bimboim and Doly, 1979). Briefly, the bacteria were centrifuged at 4000 rpm for 10 minutes and resuspended in 200 μl of P1 buffer [50 mM glucose, 10 mM EDTA, 25 mM Tris (pH 8.0)], then lysed in 200 μl of P2 buffer [0.2 M NaOH, 1% SDS (w / v)], and the lysate mixture was neutralized in 200 μl of P3 buffer (3 M KOAc, pH 6.0) and centrifuged at 4000 rpm for 10 minutes at 4°C. Then 400 μl of supernatant was transferred to a new deep well plate, and 1000 μl of cold anhydrous ethanol was added and mixed, and then centrifuged at 4000 rpm for 10 minutes at 4°C. The supernatant was discarded, and 50 μl of ddH2O was added to the plasmid pellet and mixed to dissolve the plasmid. Then 75 μl of bead buffer [2.5 M NaCl, 10 mM Tris base, 1 mM EDTA, 3.36 mM HCl, 20% (w / v) PEG8000, 0.05% (w / v) Tween 20] and 50 μl of SpeedBeads TM Magnetic carboxylate modified particles (GE Healthcare 65152105050250, diluted 1:50 in bead buffer) were added to the plasmid, mixed and incubated on a magnetic stand for 5 minutes to separate the beads from the supernatant. The beads were then washed twice with 70% ethanol and dried in a water bath (65°C). The plasmid DNA was then eluted from the beads at 65°C with 150 μl of sterile tris-EDTA buffer [1 mM EDTA, 10 mM Tris-HCl (pH 8.0)] for 10 minutes and transferred to a new low-profile 96-well plate. To ensure that the complete cloning procedure was correct, plasmids from two wells of each 96-well plate were subjected to Sanger sequencing.

[0247] Cell culture, transfection, transduction, and flow cytometry

[0248] All cells were cultured at 37 ° C with appropriate growth medium containing 5% CO2. Transfection was performed in 24-well plates via Lipofectamine3000 (Thermo Fisher Scientific) at a cell density of 80 to 90% and with 0.25 μg of gRNA plasmid and 0.25 μg of Cas9 or dCas9-VPR plasmid. For lentiviral transduction, an infection multiplicity of about 1 to 2 was used, and 3 days after infection, cells were subjected to flow cytometry or RNA extraction for real-time quantitative PCR. To confirm the gene knockout efficiency, cells were cultured under puromycin selection for about 7 days, followed by genomic DNA extraction. Flow cytometric analysis was performed by BD Canto II or LSRFortessa at the core institution center of the University of Zurich. TM Cell analyzer.

[0249] Real-time quantitative PCR

[0250] Total RNA from HEK293 cells or iNeuron was isolated by TRIzol reagent (Thermo Fisher Scientific) according to the manual. 600 ng of RNA was reverse transcribed into cDNA by quantitative reverse transcription kit (Qiagen). Real-time quantitative PCR was performed with SYBR green (Roche) according to the manual, and the primer set for each gene was as follows. GAPDH, ACTB and HMBS were used as internal controls.

[0251]

[0252]

[0253] Surveyor detection and quantification of gene editing efficiency

[0254] HEK293 cells were cultured at 4.0 × 10 5The density of cells / well was seeded in a 24-well plate. After 24 hours, cells grown at about 90% confluence were co-transfected with lentiCas9-Blast (Addgene, 52962, 250ng / well) and sgRNA plasmid (250ng / well) using Lipofectamine 3000 transfection reagent (Thermo Fisher Scientific, L3000015) according to the manufacturer's introduction. 72 hours after transfection, cells were selected for 3 days by puromycin (1μg / ml). Then, cells were harvested for genomic DNA isolation using the DNeasy blood and tissue kit (Qiagen, 69506). Genomic DNA was used as a template for PCR amplification of the targeted region in the genome using Phusion high-fidelity DNA polymerase (New England Biolabs, M0530S). For each PCR reaction in a volume of 50 μl, approximately 150 ng of genomic DNA, 0.5 μl of Phusion DNA polymerase, 5 μM forward / reverse primers, 10 mM dNTPs and 10 μl of 5× Phusion HF buffer were included, followed by temperature conditions: initial denaturation at 98°C for 30 s, 37 cycles including 98°C for 10 s, 60°C for 30 s and 72°C for 30 s / kb, and finally extension at 72°C for 10 min.

[0255] For the genome editing samples transfected with 4-sgRNA, a large number of mutant DNA bands in the PCR amplicon were visualized on the gel image after electrophoresis, and the intensity of the DNA bands was determined using ImageJ software (NIH). The gene editing efficiency was defined as the percentage of the integrated intensity of the mutant DNA band to the integrated intensity of the total DNA band.

[0256] For genome-edited samples transfected with 1-sgRNA, genetic editing efficiency was determined using the Surveyor Mutation Detection Kit (Integrated DNA Technologies, 706020). Briefly, to generate heteroduplex DNA fragments, the above PCR amplicons of 1-sgRNA sample and corresponding non-targeting control sgRNA sample were mixed in equal proportions with 4 μl / sample in a 10 μl reaction volume, followed by incubation under annealing temperature conditions: 95°C for 10 min, ramping from 95°C to 85°C at 3°C / sec, 85°C for 1 min, ramping from 85°C to 75°C at 0.3°C / sec, 75°C for 1 min, ramping from 75°C to 65°C at 0.3°C / sec, 65°C for 1 min, ramping from 65°C to 55°C at 0.3°C / sec, 55°C for 1 min, ramping from 55°C to 45°C at 0.3°C / sec, 45°C for 1 min, ramping from 45°C to 35°C at 0.3°C / sec, 35°C for 1 min, and ramping from 35°C to 25°C at 0.3°C / sec, 25°C for 1 min. The heteroduplex DNA was then treated with MgCl2 (1 μl), Surveyor enhancer (1 μl) and Surveyor nuclease (4 μl) provided in the kit at 42°C for 30 minutes and separated on a 2% agarose gel. The percentage of gene editing was calculated using the following formula: (1-(1-b / (a+b))0.5)×100, where "a" represents the intensity of the uncleaved band and "b" represents the combined intensity of all cleaved bands.

[0257] In order to further quantify the gene editing efficiency more accurately, the PCR amplicons of 1-sgRNA and 4-sgRNA gene perturbations were further subjected to next generation sequencing (NGS). First, the barcodes were attached to the common primers used for the above-mentioned PCR amplification. Four forward barcodes were designed to correspond to four biological replicates, and six reverse barcodes were designed to correspond to six genetic perturbations, including non-targeting sgRNA, sgRNA1, sgRNA2, sgRNA3, sgRNA4 and 4-sgRNA. Secondly, the barcoded primers and the above-mentioned genomic DNA were used for PCR amplification. The PCR products were then purified by gel extraction using NucleoSpin gel and PCR cleanup kit (MachereyNagel, 40609.250). Third, the PCR amplicons were pooled into 614ng, with equal molar concentrations in non-targeting sgRNA samples for all genes detected, and equal amounts in all genetic perturbations for each gene.

[0258] Lentiviral Packaging - Overview

[0259] Briefly, HEK293T cells were grown to 80 to 90% confluence in DMEM + 10% FBS on poly-D-lysine coated 24-well plates and lipofectamine 3000 was used to produce lentivirus with 3 different plasmids (transfer plasmid, PAX-2 and VSV-G; ratio: 5:3:2). After 6 hours or overnight incubation, the medium was changed to virus harvest medium (DMEM + 10% FBS + 1% BSA). The supernatant containing lentiviral particles was then harvested 48 to 72 hours after changing to virus harvest medium. Suspended cells or cell debris were precipitated by centrifugation at 1500 rpm for 5 minutes. The clarified supernatant was then titrated and stored at -80°C.

[0260] To titrate the lentiviral particles, the same number of HEK293T cells were grown in 24-well plates and infected by adding a small volume (V) of the above-mentioned viral supernatant (e.g., 3 μL). Representative batches of cells were used to determine the number of cells (N) at the time of infection. After 72 hours of infection, cells were harvested and analyzed by flow cytometry to quantify the proportion of infected cells (BFP positive). The titer (T) of the virus was then calculated using the percentage of positive cells (P) according to the following formula: T = (P * N) / V.

[0261] Lentiviral packaging - details

[0262] Method for generating HEK293T monoclonal cell lines: HEK293T cells at low passage number (before passage 8) were serially diluted to obtain approximately 10 single cells per 10 cm diameter culture dish, and 6 plates were prepared. After 10 days, 24 monoclonal lines with clear and separated boundaries from nearby clones were picked and further grown in 6-well plates respectively. After reaching 80% confluence, each clonal cell line was stored in seed form at -80°C. All 24 lines were then tested for their ability to produce lentivirus in 24-well plates with 400 ng of 4sgRNA Nt plasmid, 225 ng of PAX-2 (addgene #12260) and 150 ng of VSV-G (addgene #8454). After several rounds of validation, clone 5 among these clones showed the best performance compared to the other clones (see Results Fig.14 ).

[0263] Protocol for high titer lentiviral production in 384-well format (see also Fig.16 ):

[0264] - Maintenance of cell culture | HEK293T clone 5 cells were maintained at a maximum of 80% confluence until a maximum of passage 20. Cell seeding Two flasks (T150) were required for lentiviral production.

[0265] - Seeding of cells in 384-well cell culture plates | 7.5-8×10 3The cells were inoculated into 384-well cell culture plates with a volume of 50 μl of DMEM + 10% FBS medium using a 384-well channel 125 μl ntegra Viaflo tip (ViaFlo setting 32 mm, acceleration: 8; deceleration: 3). Considering the dead volume of the experiment, 280 ml of cell suspension of this density (mixed in a container) was required to inoculate 10 plates.

[0266] - Viral Packaging | 24 hours after cell seeding, clone 5 of HEK293T cells will reach 80 to 90% confluence. The cells are then ready for transfection, the core of the protocol. Each plasmid set for a 384-well plate will be prepared in a larger volume for triplicate transfections.

[0267] Prepare reagents for 3× plates of 384-well plates of the plasmid set as follows:

[0268]

[0269] The transfection step is time sensitive (should be completed within 1 hour) and reagents should be measured accurately.

[0270] - Transfection: Prepare Tube 1 and use a multichannel pipette to dispense 38 μl / well into a 384-deep well plate ( Fig.16 : 1a; plate with mix 1); dispense 10.1 μl / well from the plate with mix 1 to each of the three 384-PCR plates using Viaflo (12.5 μl tip); prepare tube 2 and dispense 46 μl / well to the 384 deep-well plate ( Fig.16 :2A-b);

[0271] In parallel: add 72 ng (2.4 μl of 30 μg / μl dilution) of transfer plasmid (from 384 format library) to the plate with the mixture from tube 1 - mix for 10 cycles ( Figure 1 : 3; acceleration / deceleration: 8); incubate at room temperature for 5 minutes;

[0272] - Add 12.5 μl / well of the plate with tube 2 mixture to the plate with tube 1 mixture to obtain 25 μl / well transfection mixture - Mix for 20 cycles ( Figure 1 : 4; Speed ​​up / down: 8) Reuse tips in transfection steps - so keep them in their own box; Incubate at room temperature for 20 minutes;

[0273] - During the incubation period, remove 36 μl of culture medium from the cell culture plate (ViaFlo setting: 31.5 mm; acceleration 3); centrifuge the plate containing the transfection mixture before use (1500 rpm, 2 minutes); transfect the cells by adding 8 μl of transfection mixture to each cell replica (ViaFlo setting: 16 mm; deceleration 3), with a total of 3 replicates for each plasmid group; incubate at 37°C in a virus incubator for 6 to 12 hours.

[0274] -Add Virus Harvest Medium | Add Virus Harvest Medium to each well of transfected cells 6 hours after transfection or the next morning without removing the transfection mix. Briefly, warm the Virus Harvest Medium at 37°C for 30 minutes before use and then pour into the container immediately before use. Add 44 μl of Virus Harvest Medium to the transfected cells; keep close to the well wall when releasing the medium from the pipette tip to avoid detaching any cells in the well (ViaFlo settings: 37 mm, acceleration: 8, deceleration: 2).

[0275] -Virus collection | 48 to 60 hours after adding virus harvest medium, take 45μl of virus from each replicate of the transfection plate, place three replicates of the same plasmid in a deep-well 384-well plate, and mix for 20 cycles (Viaflo setting acceleration / deceleration 8); centrifuge at 1500rpm for 2 minutes to pellet the cells; aliquot 15μl of viral supernatant into 6× PCR plates and aliquot 20μl of virus into another PCR plate to prepare a pooled library. The virus in the PCR plate was tightly sealed with an aluminum lid and stored at -80°C. The remaining virus in the deep-well plate was stored and used to transduce wild-type HEK293T cells, and then lentiviral titration was performed in a 384-well plate using a high-throughput 384-well flow cytometer ( Fig.15 ).

[0276] -Appendix: Culture medium chapter

[0277] -Cell culture medium: DMEM + 10% FBS

[0278] - Virus harvest medium: DMEM (66 ml from a 500 ml bottle), supplemented with 59 ml of FBS, 6.82 g of BSA powder and 6.8 ml of penicillin / streptomycin.

[0279] References (Materials and Methods Section)

[0280] Birnboim,H.C.,and Doly,J.(1979).A rapid alkaline extraction procedurefor screening recombinant plasmid DNA.Nucleic Acids Res 7,1513-1523.

[0281] Koike-Yusa,H.,Li,Y.,Tan,E.P.,Velasco-Herrera Mdel,C.,and Yusa,K.(2014).Genome-wide recessive genetic screening in mammalian cells with alentiviral CRISPR-guide RNA library.Nat Biotechnol 32,267-273.

[0282] Metzakopian,E.,Strong,A.,Iyer,V.,Hodgkins,A.,Tzelepis,K.,Antunes,L.,Friedrich,M.J.,Kang,Q.,Davidson,T.,Lamberth,J.,et al.(2017).Enhancing thegenome editing toolbox:genome wide CRISPR arrayed libraries.Sci Rep 7,2244.

[0283] Abugessaisa,I.,Noguchi,S.,Hasegawa,A.,Harshbarger,J.,Kondo,A.,Lizio,M.,Severin,J.,Carninci,P.,Kawaji,H.,and Kasukawa,T.(2017).FANTOM5 CAGEprofiles of human and mouse reprocessed for GRCh38 and GRCm38 genomeassemblies.Sci.Data 4,170107.

[0284] Braschi,B.,Denny,P.,Gray,K.,Jones,T.,Seal,R.,Tweedie,S.,Yates,B.,andBruford,E.(2019).Genenames.org:the HGNC and VGNC resources in 2019.NucleicAcids Res.47,D786-D792.

[0285] Concordet,J.-P.,and Haeussler,M.(2018).CRISPOR:intuitive guideselection for CRISPR / Cas9 genome editing experiments and screens.NucleicAcids Res.46,W242-W245.

[0286] Doench,J.G.,Fusi,N.,Sullender,M.,Hegde,M.,Vaimberg,E.W.,Donovan,K.F.,Smith,I.,Tothova,Z.,Wilen,C.,Orchard,R.,et al.(2016).Optimized sgRNA designto maximize activity and minimize off-target effects of CRISPR-Cas9.Nat.Biotechnol.34,184-191.

[0287] FANTOM Consortium and the RIKEN PMI and CLST(DGT),Forrest,A.R.R.,Kawaji,H.,Rehli,M.,Baillie,J.K.,de Hoon,M.J.L.,Haberle,V.,Lassmann,T.,Kulakovskiy,I.V.,Lizio,M.,et al.(2014).A promoter-level mammalian expressionatlas.Nature 507,462-470.

[0288] Frankish,A.,Diekhans,M.,Ferreira,A.-M.,Johnson,R.,Jungreis,I.,Loveland,J.,Mudge,J.M.,Sisu,C.,Wright,J.,Armstrong,J.,et al.(2019).GENCODEreference annotation for the human and mouse genomes.Nucleic Acids Res.47,D766-D773.

[0289] Gilbert,L.A.,Horlbeck,M.A.,Adamson,B.,Villalta,J.E.,Chen,Y.,Whitehead,E.H.,Guimaraes,C.,Panning,B.,Ploegh,H.L.,Bassik,M.C.,et al.(2014).Genome-Scale CRISPR-Mediated Control of Gene Repression and Activation.Cell159,647-661.

[0290] Glusman,G.,Caballero,J.,Mauldin,D.E.,Hood,L.,and Roach,J.C.(2011).Kaviar:an accessible system for testing SNV novelty.Bioinforma.Oxf.Engl.27,3216-3217.

[0291] Graf,R.,Li,X.,Chu,V.T.,and Rajewsky,K.(2019).sgRNA Sequence MotifsBlocking Efficient CRISPR / Cas9-Mediated Gene Editing.Cell Rep.26,1098-1103.e3.

[0292] Haeussler,M., K.,Eckert,H.,Eschstruth,A.,Mianné,J.,Renaud,J.-B.,Schneider-Maunoury,S.,Shkumatava,A.,Teboul,L.,Kent,J.,et al.(2016).Evaluation of off-target and on-target scoring algorithms and integrationinto the guide RNA selection tool CRISPOR.Genome Biol.17,148.

[0293] Hanna,R.E.,and Doench,J.G.(2020).Design and analysis of CRISPR-Casexperiments.Nat.Biotechnol.

[0294] Hart,T.,Tong,A.H.Y.,Chan,K.,Van Leeuwen,J.,Seetharaman,A.,Aregger,M.,Chandrashekhar,M.,Hustedt,N.,Seth,S.,Noonan,A.,et al.(2017).Evaluation andDesign of Genome-Wide CRISPR / SpCas9 Knockout Screens.G3 Bethesda Md 7,2719-2727.

[0295] Horlbeck,M.A.,Gilbert,L.A.,Villalta,J.E.,Adamson,B.,Pak,R.A.,Chen,Y.,Fields,A.P.,Park,C.Y.,Corn,J.E.,Kampmann,M.,et al.(2016).Compact and highlyactive next-generation libraries for CRISPR-mediated gene repression andactivation.ELife 5.

[0296] Hsu,P.D.,Scott,D.A.,Weinstein,J.A.,Ran,F.A.,Konermann,S.,Agarwala,V.,Li,Y.,Fine,E.J.,Wu,X.,Shalem,O.,et al.(2013).DNA targeting specificity ofRNA-guided Cas9 nucleases.Nat.Biotechnol.31,827-832.

[0297] Huber,W.,Carey,V.J.,Gentleman,R.,Anders,S.,Carlson,M.,Carvalho,B.S.,Bravo,H.C.,Davis,S.,Gatto,L.,Girke,T.,et al.(2015).Orchestrating high-throughput genomic analysis with Bioconductor.Nat.Methods 12,115-121.

[0298] Lambert,S.A.,Jolma,A.,Campitelli,L.F.,Das,P.K.,Yin,Y.,Albu,M.,Chen,X.,Taipale,J.,Hughes,T.R.,and Weirauch,M.T.(2018).The Human TranscriptionFactors.Cell 172,650-665.

[0299] Perez,A.R.,Pritykin,Y.,Vidigal,J.A.,Chhangawala,S.,Zamparo,L.,Leslie,C.S.,and Ventura,A.(2017).GuideScan software for improved single and pairedCRISPR guide RNA design.Nat.Biotechnol.35,347-349.

[0300] R Core Team(2020).R:A Language and Environment for StatisticalComputing(Vienna,Austria:R Foundation for Statistical Computing).

[0301] Sanson,KR,Hanna,RE,Hegde,M.,Donovan,KF,Strand,C.,Sullender,ME,Vaimberg,EW,Goodale,A.,Root,DE,Piccioni,F.,et al.(2018).Optimizedlibraries for CRISPR-Cas9 genetic screens with multiplemodalities.Nat.Commun.9,5416.

[0302] Smedley,D.,Haider,S.,Ballester,B.,Holland,R.,London,D.,Thorisson,G.,and Kasprzyk,A.(2009).BioMart--biological queries made easy.BMC Genomics 10,22.

[0303] Uhlén,M.,Fagerberg,L., BM,Lindskog,C.,Oksvold,P.,Mardinoglu,A.,Sivertsson, Campaign, C., E.,Asplund,A.,et al.(2015).Proteomics.Tissue-based map of the human proteome.Science 347,1260419.

[0304] Uhlén,M.,Karlsson,MJ,Hober,A.,Svensson,A.-S.,Scheffel,J.,Kotol,D.,Zhong,W.,Tebani,A.,Strandberg,L.,Edfors,F.,et al.(2019).The humansecretome.Sci.Signal.12.

[0305] Yates,AD,Achuthan,P.,Akanni,W.,Allen,J.,Allen,J.,Alvarez-Jarreta,J.,Amode,MR,Armean,IM,Azov,AG,Bennett,R.,et al.(2020).Ensembl2020.Nucleic Acids Res.48,D682-D688.

Claims

1. A method for providing multiple RNA expression plasmids, comprising the following steps: a. Providing a plasmid backbone fragment comprising a 5′ backbone end and a 3′ backbone end; b. providing a plurality of n insertion fragments comprising a first insertion, a second insertion, ..... an nth insertion, wherein i. each insert comprises a sequence encoding a transcribed RNA sequence, ii. Each insert comprises a 5' insertion end and a 3' insertion end; iii. Among them The 5' first insertion end overlaps the 3' backbone end, and · The 3′ first insertion end overlaps the 5′ second insertion end, and The 3′ second insertion end overlaps with the 5′ nth insertion end, and the 3′ nth insertion end overlaps with the 5′ backbone end; and wherein each overlap has a length of 15 to 40 bp, in particular a length of 17 to 22 bp, iv. one of the inserts further encodes an insert selectable marker under the control of a promoter operable in a host cell; c. In a Gibson assembly step, the plasmid backbone fragment and each of the plurality of insert fragments are contacted in the presence of the following enzyme activities: -5′ exonuclease activity, -DNA-dependent DNA polymerase activity -DNA ligase activity; d. transforming the product of the Gibson assembly step into a bacterial host cell and growing the host cell in bacterial culture; e. isolating the ligated plasmid product from the bacterial culture. 2 . The method according to claim 1 , wherein the plurality of n insert fragments consists of a first fragment, a second fragment and a third fragment.

3. The method of claim 1 or 2, wherein the plurality of insertions encode four transcribed RNA sequences.

4. A method according to any one of the preceding claims, wherein the transcribed RNA sequence comprises sgRNA and trcrRNA.

5. The method of claim 4, wherein each of the sgRNAs is specific for a sequence contained in the same gene of the target cell.

6. The method according to any one of the preceding claims, wherein the plasmid backbone fragment is obtained by propagating a precursor plasmid comprising a backbone selection marker, isolating the precursor plasmid and excising the backbone selection marker to produce the plasmid backbone fragment.

7. A method according to any one of the preceding claims, wherein the insert is provided by first synthesizing the insert sequence followed by PCR using a primer pair to introduce the sgRNA and create an overlap.

8. The method according to any one of the preceding claims, wherein the plasmid backbone and the ligated plasmid product can be packaged into a retroviral vector.

9. A method for providing a multiplex RNA expression plasmid library, comprising carrying out the method according to any one of claims 1 to 8 in a plurality of parallel reactions, wherein in each of the reactions, the transcribed RNA sequence is different.

10. The method of claim 9, wherein each of the transcribed RNA sequences comprises an sgRNA, and for any one reaction, each of the sgRNAs is specific for a sequence in the same gene contained in the target cell, and for each of the reactions, the gene is different.

11. The method according to claim 10, wherein at least 100, more particularly at least 1000 sgRNA sequences selected from SEQ ID NOs 64 to 170943 are expressed.

12. A multiplex RNA expression plasmid library obtained or obtainable by the method according to claims 9 to 11.

13. A method for molecular cloning of a plasmid containing an insert from a parent plasmid, wherein a. The parent plasmid comprises a first selection marker flanked by a recognition sequence sensitive to the endonuclease action (cleavage) of a type II restriction enzyme; b. Cut (digest) the parent plasmid with a type II restriction enzyme to generate two fragments: i. a parental plasmid backbone fragment having a 5′ backbone (single-stranded overlapping) end and a 3′ backbone (single-stranded overlapping) end; ii. the first resistance marker fragment; c. connecting an insert fragment containing a second selection marker and comprising ends annealing (matching) to the parental plasmid backbone fragment, wherein the insert fragment has a sequence of a non-reconstructed type II recognition site adjacent to the end; d. Transferring the ligation product into cells under conditions where only cells containing the second selection marker can proliferate, and processing (preparing) the resulting cell suspension to obtain the plasmid containing the insert.

Images