Triple combinatorial CRISPR screening system and method for analyzing target interactions
Through the CombiGEM-CRISPR v2.0 platform, the cost and efficiency of high-order drug combination screening is solved using CRISPR technology and barcoded RNA library, and the rapid screening of multigene combinations and drug combination verification is achieved, and the combination therapy is promoted. develop.
Patent Information
- Application Number
- CN202180028845.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2020-04-16
- Filing Date
- 2021-04-14
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2041-04-14
AI Technical Summary
The prior art is difficult to effectively screen and verify combinations containing three or more drugs, especially in the prediction of synergistic or antagonistic effects, and the experimental cost of higher-order genes and drug combinations is limited by the combination of combination therapy. develop.
The CombiGEM-CRISPR v2.0 platform was developed, using a CRISPR-based multigene knockout screening system and a barcoded advanced combinatorial guide RNA library to quickly screen the disease-relieving gene combinations through lentiviral vectors and next-generation sequencing technology to identify potential Therapy target combination and matching drug regimen.
It has achieved systematic characterization of multiple genetic interactions, efficient screening of binary or ternary synergistic drug combinations that alleviate diseases, reducing experimental costs and time, and verified a drug combination scheme that inhibits cancer cell growth and protects against toxicity.
Smart Images

Figure BDA0003891470420000151 
Figure BDA0003891470420000161 
Figure BDA0003891470420000171
Abstract
Description
[0001] This international patent application claims the benefit of U.S. Provisional Patent Application No. 63 / 010,877, filed on April 16, 2020, which is incorporated by reference in its entirety for all purposes. Technical Field
[0002] Systems for multiplex genome editing or binary or ternary combinatorial CRISPR screening are provided herein. High-throughput screening of combinations of genes that mitigate disease to identify binary and ternary synergistic drug combinations as potential therapeutic options is also provided. Lentiviral ternary combinatorial guide RNA expression cassettes and combinatorial guide RNA libraries are also provided. 2. Background of the invention
[0004] Although combination therapy holds promise for improving treatment outcomes for a variety of diseases (Al-Lazikani et al., 2012), only a limited number of effective combinations have been discovered to date, particularly those involving three or more drugs (Table S1). The effects of drug combinations are difficult to predict due to unexpected synergies or antagonisms and are not simply the sum of the effects of each drug alone (Borisy et al., 2003). Microplate arrays have been coupled to robotic systems to screen large-scale drug combinations. However, this approach can become prohibitively expensive, as the number of experiments increases exponentially with the number of drugs and the order of complexity of the combinations studied. RNA interference and clustered regularly interspaced short palindromic repeats (CRISPR)-based genetic perturbation systems have been applied to facilitate the screening of effective drug-target pairs (Doench, 2018; Wong et al., 2016b), but no library assembly and screening methods have been validated to evaluate more than two targets simultaneously. This can be attributed to the relatively low and variable cleavage efficiency of polycistronic systems expressing multiple RNAs (Han et al., 2017; Xu et al., 2017), and / or the large-scale oligonucleotide synthesis and high sequencing costs required for the characterization of higher-order combinations (specified in the design). Mathematical models for predicting ternary and higher-order drug interactions have been developed (Cokol et al., 2017; Wood et al., 2012; Zimmer et al., 2016), but high-throughput methods are needed to experimentally validate sets of potential combinations. To overcome this bottleneck, we established and validated a scalable platform (named CombiGEM-CRISPR v2.0) for rapid screening of disease-modifying gene knockouts to study higher-order genetic interactions, identify potential therapeutic target combinations, and deploy their matching drug regimens for further testing. 3. Summary of the Invention
[0006] A CRISPR-based multi-gene knockout screening system and a new toolkit are provided for universal and scalable assembly of barcoded high-order combinatorial guide RNA libraries. The system was applied to systematically identify not only pairwise synergistic therapeutic target combinations but also ternary synergistic therapeutic target combinations, and successfully validated dual and triple combination schemes for inhibiting cancer cell growth and protecting against Parkinson's disease-related toxicity. The system overcomes the practical challenges of experimenting with large numbers of high-order gene and drug combinations and is suitable for revealing rare synergistic interactions between druggable targets.
[0007] Systematically characterize the genetic interactions between multiple (i.e., more than two) elements. One application of this system is the ability to perform high-throughput screening of disease-mitigating gene combinations to identify binary or even ternary synergistic drug combinations as potential treatment options. The effect of drug combinations is difficult to predict due to unexpected synergies or antagonisms and is not simply the sum of the effects of each drug. Due to technical difficulties in systematically screening a large number of possible combinations, it has always been a major challenge to find effective drug combinations for diseases. For example, microplate arrays are coupled to robotic systems to screen large-scale drug combinations. However, since the number of experiments increases exponentially with the number of drugs and the order of the complexity of the combination studied, this method that requires a large number of drugs may become very expensive. Therefore, although combination therapy is expected to improve the therapeutic effect of various diseases, only a limited number of effective combinations have been found so far, especially combinations containing more than two drugs.
[0008] Provided herein are systems for multiplex genome editing or binary or ternary combinatorial CRISPR screening, comprising: a lentiviral vector comprising a human U6 (hU6) promoter, a mouse U6 (mU6) promoter, and a human H1 (hH1) promoter, wherein the lentiviral vector expresses an array of three or more barcoded guide RNA ("gRNA") oligonucleotide pairs, wherein the promoter has a 3' end comprising modified hU6, mU6, and hH1 promoter sequences for paired annealing of the barcoded gRNA oligonucleotide pairs.
[0009] Provided herein is a system for multiplex genome editing or binary or ternary combinatorial CRISPR screening, comprising: a lentiviral vector comprising a human U6 (hU6) promoter, a mouse U6 (mU6) promoter, and a human H1 (hH1) promoter, wherein the lentiviral vector expresses an array of three or more barcoded guide RNA ("gRNA") oligonucleotide pairs, the hU6 promoter having an unmodified promoter sequence at the 3' end, and the mU6 and hH1 promoters having modified promoter sequences at the 3' end for paired annealing of the barcoded gRNA oligonucleotide pairs.
[0010] In one embodiment, paired pairs of barcoded gRNA oligonucleotides anneal to form an RNA scaffold.
[0011] In one embodiment, the combinatorial gRNA library is assembled by CombiGEM-CRISPR v2.0.
[0012] In one embodiment, a lentiviral vector is transfected into a human cell, and the barcoded gRNA is delivered to the human cell.
[0013] In one embodiment, the system further comprises quantifying the barcoded gRNA using next generation sequencing at a time point after transfection.
[0014] In one embodiment, the triple combinatorial CRISPR screen is a high throughput screen.
[0015] In one embodiment, the gRNA forms an RNA scaffold sequence comprising the same modified 3' end of the hU6, mU6, and hH1 promoter sequences as the combinatorial gRNA library.
[0016] In one embodiment, the gRNA forms an RNA scaffold sequence that comprises the 3' end of the same hU6, mU6, and hH1 promoter sequences as the combinatorial gRNA library.
[0017] Provided herein is a method for screening at least a ternary drug target combination; the method comprising: (i) providing a gRNA library targeting druggable genes of HGSOC, wherein each gene comprises an array of 3 gRNAs; (ii) transfecting human cells; and (iii) quantifying the barcoded gRNAs using next generation sequencing.
[0018] The present invention provides a system for screening at least a three-way drug target combination, which comprises: (i) providing a three-way combination gRNA expression construct of a lentiviral vector expressing gRNA; (ii) transfecting human cells with a fluorescent reporter gene; (iii) measuring the percentage of the cell population that is fluorescent positive within a period of time after transfection.
[0019] In one embodiment, fluorescence is measured using flow cytometry and wherein the fluorescence is GFP, RFP and BFP fluorescence.
[0020] In one embodiment, the gRNA targets exonic regions of green (GFP), red (RFP), and blue (BFP) fluorescent protein reporter genes.
[0021] In one embodiment, the human cell is an ovarian cancer cell.
[0022] In one embodiment, the ovarian cancer cells are high-grade serous ovarian cancer ("HGSOC") cells.
[0023] In one embodiment, the ovarian cancer cells are OVCAR8-ADR and OVCAR8-ADR-Cas9.
[0024] The present invention provides a method for screening at least a three-way drug target combination, which comprises the following steps: (i) providing a three-way combination gRNA expression construct of a lentiviral vector expressing gRNA; (ii) transfecting human cells with a fluorescent reporter gene; (iii) measuring the percentage of the cell population that is fluorescently positive within a period of time after transfection.
[0025] In one embodiment, the method further comprises validating the ternary drug-target combination by matching the drug to the drug target.
[0026] In one embodiment, the triple drug target combination provides a three-drug regimen for the disease.
[0027] In one embodiment, the fluorescent protein is GFP, RFP or BFP.
[0028] In one embodiment, the at least three-way drug target combination is a synergistic combination.
[0029] In one embodiment, the disease is cancer or Parkinson's disease.
[0030] Provided herein are methods for treating HGSOC comprising administering a drug targeting PARP1, DNMT1, CDK2, FKBP1A, or a combination thereof.
[0031] In one embodiment, the medicament comprises olaparib (OLA), azacitidine (AZA), celecoxib (SEL), sirolimus (SIR), or a combination thereof.
[0032] In one embodiment, the medicament comprises OLA and AZA.
[0033] Provided herein is a system for CRISPR-based multi-gene knockout screening, the system comprising a barcoded gRNA expression cassette comprising: (i) a first promoter operably linked to a first gRNA; (ii) a second promoter operably linked to a second gRNA; (iii) a third promoter operably linked to a third gRNA; and (iv) three barcoded gRNA sequencing regions, wherein the gRNA expression cassette is in a single vector.
[0034] In one embodiment, the promoters are human U6, mouse U6, and human H1 promoters, and the three barcoded gRNAs are modified gRNA scaffold variants.
[0035] In one embodiment, the promoter comprises a modified 3' terminal sequence that is complementary to the modified gRNA scaffold variant, and the 3' terminal sequence anneals to the modified gRNA scaffold variant.
[0036] In one embodiment, the system further comprises: (i) pooled digestion and ligation of the annealed 3' end sequences and the gRNA scaffold variants to form an assembly of a pooled barcoded combinatorial gRNA library.
[0037] In one embodiment, the expression cassette knocks out the target GFP gene in OVACR8-ADR-Cas9 cells.
[0038] In one embodiment, the gRNA scaffold variant comprises: (i) higher on-target activity than the wild-type scaffold; (ii) low off-target activity; (iii) high on-to-off-target activity.
[0039] Provided herein is a system for CRISPR-based multi-gene knockout screening, comprising a barcoded gRNA expression cassette comprising v3.11, v.3.12 or v.3.13. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication containing color drawings will be provided by the Office upon request and payment of the necessary fee.
[0041] Figure 1 . Scalability assessment of existing methods and possible toolkits for assembling ternary gRNA panels for screening
[0042] Figure 2 .CombiGEM-CRISPR v2.0 Overview
[0043] The workflow begins with the synthesis of barcoded gRNA oligonucleotide pairs, which are annealed and cloned in pooled form into gRNA expression vectors. Only one set of oligonucleotides is needed to construct higher-order gRNA combinatorial libraries for multiplex CRISPR screening, as the 3' end of the promoter is a sequence that accommodates the sticky ends of the annealed oligonucleotides. Figure 6 The preassembled libraries are assembled multiplexed using a one-pot reaction as described in and delivered into human cells via lentivirus. The barcoded representation within the cell pool is quantified using next-generation sequencing. This is then validated by using matched drugs. The preassembled libraries can also be flexibly expanded to higher-order libraries, or more focused libraries can be constructed using subsets of the same oligonucleotides for secondary screening.
[0044] Figure 3 .Multiple gene function disruption using CRISPR-Cas9-based multi-gene knockout system
[0045] Lentiviral delivery of combined gRNA expression constructs effectively disrupts multiple target genes. Flow cytometry was used to measure the percentage of cell populations positive for GFP, RFP, and BFP fluorescence in OVCAR8-ADR and OVCAR8-ADR-Cas9 cells at days 11-14 post-infection. Data are mean ± SD, n = 3 biological replicates.
[0046] Figure 4 .CRISPR-based triple gene knockout screen identifies synergistic ternary combinations that inhibit ovarian cancer cell growth
[0047] (A) Distribution of barcode reads in plasmid and infected OVCAR8-ADR-Cas9 cell pools. High coverage of the three-way combinatorial gRNA library was obtained in both plasmid and cell pools (99.8% of all expected gRNA combinations; 32,705 out of 32,768). Most barcoded gRNA combinations were detected within 5-fold of the average barcode reads for each combination (highlighted by the shaded area).
[0048] (B) The high correlation between barcode representation (normalized barcode counts) in the plasmid pool and infected cell pool indicates efficient lentiviral delivery of the ternary combinatorial gRNA library into cells. The horizontal dashed line in the Bland-Altman plot indicates the 95% identity limit.
[0049] (CD) High reproducibility of barcode representation between two biological replicates in cells cultured 15 days (C) and 26 days (D) post-infection using a ternary combinatorial gRNA library. The horizontal dashed lines in the Bland-Altman plots indicate the 95% identity limit. The vertical dashed lines indicate the threshold of 100 raw barcode counts.
[0050] (E) The coefficient of variation (CV; defined as SD / mean of the fold change of normalized barcode counts for 26-day versus 15-day cultured cells) was determined for two biological replicates. More than 94.8% of the paired gRNA combinations in the screen had a CV <1.
[0051] (F) OVCAR8-ADR-Cas9 cells infected with barcoded ternary combinatorial gRNA libraries were cultured for 15 and 26 days. Barcode representation within the cell pool was quantified using Illumina HiSeq. The barcoded library vectors used the hH1-gRNA-WT scaffold, hU6-gRNA-v1 scaffold, and mU6-gRNA-v2 scaffold in the first, second, and third expression cassettes, respectively.
[0052] (G) Screening data graph showing the change in abundance (average log2(fold change); x-axis) and its genetic interaction (GI3) score (y-axis) for each barcoded gRNA combination at day 26 vs. day 15 post infection. Hit combinations DNMT1+POLA1+EGFR, DNMT1+POLA1+ERBB2, and CDK4+MAP2K1+POLA1 are highlighted in red. Data were collected from two biological replicates.
[0053] (H) Comparison of the mean log2 (fold change) of the ternary gRNA hit combinations with the single and paired gRNA combinations that comprise them (see STAR Methods for details). Statistical significance was analyzed by one-way ANOVA and Dunnett's post hoc test. * Indicates P<0.05.
[0054] (I) Survival of OVCAR8-ADR-Cas9 cells infected with the indicated triple knockouts and controls as determined by MTT assay. Data are shown as mean ± SD from biological replicates (n = 4). Statistical significance was analyzed by one-way ANOVA and Tukey post hoc test. * Indicates P < 0.05; # P < 0.05 indicates comparison with triple knockout of the safe harbor locus.
[0055] Figure 5 . Validation of screening hits using a matched three-drug regimen
[0056] (A) Survival of OVCAR8-ADR cells treated with AZA-, FLU- and / or ERL as determined by MTT assay. Data are shown as mean ± SD from biological replicates (n = 3). Statistical significance was analyzed by one-way ANOVA and Dunnett's post hoc test. * Indicates P<0.05.
[0057] (B) Colony formation assay of AZA-, FLU- and / or ERL-treated OVCAR8-ADR cells. The number and area of colonies were quantified. Data are shown as mean ± SD from biological replicates (n = 3). Statistical significance was analyzed by one-way ANOVA and Dunnett's post hoc test. * Indicates P<0.05.
[0058] (C) Surface plots depicting drug synergy for AZA+FLU+ERL (orange) and AZA+FLU+LAP (cyan). The circles on the transparent triangular planes represent the expected IC for each two-drug combination. 50 , the expected IC of triple drug treatment 50is located in the center of the triangle. The gray dots are the IC observed for single-drug and dual-drug treatments. 50 The red dots are the observed IC for the three-drug treatment 50 Concave or convex color plots indicate synergistic or antagonistic drug interactions, respectively. FIC3 is the fractional inhibitory concentration. Views from two angles are shown (left and right).
[0059] (D) Survival of OVCAR8-ADR cells treated with RIB-, TRA- and / or FLU as determined by MTT assay (left panel). Data are shown as mean ± SD from biological replicates (n = 3). Statistical significance was analyzed by one-way ANOVA and Dunnett's post hoc test. * Indicates P < 0.05. Surface plot (right) depicts the drug synergistic effect of RIB+TRA+FLU as shown in (C).
[0060] Figure 6 .Strategies for assembling barcoded gRNA library pools
[0061] (A) gRNA expression cassettes with different promoters and scaffold variants resulted in efficient knockout of the target GFP gene in OVCAR8-ADR-Cas9 cells. GFPsg1 was used. Flow cytometry was performed to measure the percentage of cell populations positive for GFP fluorescence. Data are mean ± SD, n = 3 biological replicates. Statistical significance was analyzed by one-way ANOVA and Tukey post hoc test. Modifications in gRNA scaffolds v1 and v2 are shown.
[0062] (B) Library assembly procedure. Oligonucleotide pairs (Oligo-F and Oligo-R, respectively) were synthesized and annealed to generate double-stranded inserts with a 20 bp gRNA target sequence, two BbsI sites, an 8 bp barcode, and a 5' overhang at its end. The inserts were mixed in equimolar ratios and cloned into three BbsI and Mfe-digested storage vectors containing hH1, hU6, or mU6 promoters, and a one-pot ligation reaction was performed through their compatible ends to generate a merged storage vector library. The 3' ends of the promoter sequences of hH1 and mU6 were modified so that the same gRNA oligonucleotide pool could be used to construct the library. The subsequent one-pot ligation reaction was performed using the BbsI-digested merged storage vector library and an insert containing a gRNA scaffold sequence (WT, v1, or v2), BamHI and EcoRI sites, and a 5' overhang at its end to assemble a barcoded gRNA library pool.
[0063] Figure 7 .gRNA activity of targeted genes in OVCAR8-ADR-Cas9 cells
[0064] (A) The efficiency of each gRNA was plotted against the on-target efficiency score predicted by Azimuth 2.0. The efficiency of gRNAs targeting fluorescent reporter genes was determined using flow cytometry, and data for endogenous genes were obtained from published data (Wong et al., 2016a).
[0065] (BC) Targeted alleles of cells infected with gRNA expression constructs were sequenced. OVCAR8-ADR-Cas9 cells were infected with lentivirus encoding the indicated paired (B) and triple combination (C) gRNA expression constructs. Targeted alleles were amplified from genomic DNA by PCR and cloned into storage vectors for Sanger sequencing. The editing efficiency of each targeted locus and the number of individual clones being sequenced are shown.
[0066] Figure 8 . On-target and frameshift score predictions for gRNAs used in screening libraries
[0067] The gRNAs selected for the ternary combinatorial CRISPR screen showed higher average on-target scores and higher average frameshift scores predicted by inDelphi and FORECasT compared to those used in the pairwise CRISPR screens. The combinatorial effects brought by on-target and frameshift predictions were also evaluated.
[0068] Fig. 9 Growth inhibition induced by simultaneous targeting of DNMT1+POLA1+EGFR / ERBB2 and CDK4+MAP2K1+POLA1 genes but not the safe harbor locus
[0069] (AB) OVCAR8-ADR-Cas9 cells were infected with triple gRNA combinations targeting the specified genes (A), and single and / or double gRNA combinations (B). The three safe harbor loci targeted were PPP1R12C, THUMPD3-AS1, and CCR5. The gRNAs used in each combination are listed in Table S5. Cell viability was determined by MTT assay. Data are mean ± SD from biological replicates (n = 4).
[0070] Fig.10 RNA-seq analysis of OVCAR8-ADR cells treated with drug combinations
[0071] (A) Heat map showing the log2 fold changes in expression of 3,834 genes identified with FDR < 0.05 as differentially expressed in OVCAR8-ADR cells treated with the three-drug regimen (AZA + FLU + ERL) compared to untreated controls. The log2 fold changes in expression of these genes after treatment of cells with the corresponding drug pairs are included. Hierarchical clustering of genes and samples was performed based on Pearson correlation.
[0072] (B) Matrix showing mapped pathways of differentially expressed genes that were at least 20% upregulated (orange) or downregulated (blue) in OVCAR8-ADR cells treated with the three-drug regimen (AZA+FLU+ERL) compared to untreated controls, and also had a fold change of >10%, >5%, or <5% in cells treated with the three-drug regimen compared to all cells treated with the corresponding two-drug regimen. The bars at the top of the x-axis represent the fold change levels detected with different shades of green. Each dark dot indicates that a gene is present in a specific pathway.
[0073] Fig.11 Synergistic drug interactions in OVCAR-ADR cells evaluated by AZA-FLU-ERL / LAP at different inhibitory concentrations
[0074] Surface plots depicting drug interactions and fractional inhibitory concentration (FIC3) scores for AZA+FLU+ERL (orange) and AZA+FLU+LAP (cyan) are shown. IC 50 Figure Figure 5 As shown in C.
[0075] Fig.12 Synergistic drug interactions in OVSAHO cells evaluated by AZA-FLU-ERL / LAP at different inhibitory concentrations
[0076] Surface plots depicting drug interactions and fractional inhibitory concentration (FIC3) scores for AZA+FLU+ERL (orange) and AZA+FLU+LAP (cyan) are shown.
[0077] Fig.13 CRISPR-based double knockout screen identifies drug pairs that inhibit ovarian cancer cell growth
[0078] (A) Distribution of barcode reads in plasmids and OVCAR8-ADR-Cas9 cell pools 9 days after infection. 99.7% (25,201 out of 25,281) and 99.0% (25,027 out of 25,281) of all expected gRNA combinations were obtained in plasmids and cell pools, respectively. Most barcoded gRNA combinations were detected within 5-fold of the average barcode reads for each combination (highlighted by the shaded area).
[0079] (B) The high correlation between barcode representation (normalized barcode counts) in the plasmid pool and infected cell pool indicates efficient lentiviral delivery of paired gRNA libraries into cells. The horizontal dashed lines in the Bland-Altman plot represent the 95% identity limit.
[0080] (CD) Highly reproducible barcode representation between two biological replicates in cells cultured 15 days (C) and 21 days (D) after infection with paired gRNA libraries. The horizontal dashed lines in the Bland-Altman plots represent the 95% identity limits. The vertical dashed lines represent the threshold of 100 raw barcode counts.
[0081] (E) The coefficient of variation (CV; defined as SD / mean of the fold change of normalized barcode counts for 21-day versus 15-day cultured cells) was determined for two biological replicates. More than 99.2% of the paired gRNA combinations in the screen had a CV <1.
[0082] (F) OVCAR8-ADR-Cas9 cells infected with barcoded paired gRNA libraries were cultured for 15 and 21 days. Barcode representation within the cell pool was quantified using Illumina HiSeq. The barcoded library vectors used the hH1-gRNA-WT scaffold and the hU6-gRNA-v1 scaffold in the first and second expression cassettes, respectively.
[0083] (G) Volcano plot comparing the abundance changes of each barcoded gRNA combination at day 21 versus day 15 post-infection. Hit combinations DNMT1+PARP1 and FKBP1A+CDK2 are highlighted in blue and red, respectively. Data were collected from two biological replicates.
[0084] (HJ, MN) Cell viability of OVCAR8-ADR-Cas9 cells (H) infected with the indicated gRNA and OVCAR8-ADR cells (I, J, M, N) treated with drugs as determined by MTT assay. In (J and N), the percentage of growth inhibition was calculated by comparing each drug-treated group with the untreated control group. Drug synergy was measured by the Bliss Independent model and the HSA model. The growth inhibition effect of 8 μM OLA and 1.2 μM AZA (J) and 5 μM SEL and 0.5 μM SIR (N) treatments were illustrated as examples, and the dotted line represents the expected drug combination effect based on the Bliss Independent model. Data are shown as mean ± SD from biological replicates (n = 3 in H; n = 6 in I and J; n = 8 in M and N).
[0085] (K, O) Colony formation assay of OLA- and / or AZA- (K) and SEL- and / or SIR- (O) treated OVCAR8-ADR cells. The number and area of colonies were quantified. Data are shown as mean ± SD from biological replicates (n = 3).
[0086] (L, P) Cell cycle analysis of OLA- and / or AZA- (K) and SEL- and / or SIR- (O) treated OVCAR8-ADR cells. The percentage of cells in each cell cycle phase was quantified. Data are shown as mean ± SD from biological replicates (n = 3).
[0087] The statistical significance of HP was analyzed by one-way ANOVA and Tukey's post hoc test. * P<0.05, ** P < 0.01 and **** P < 0.0001; JL and NO # P < 0.05 indicates comparison with untreated control.
[0088] Fig.14 .Inhibitory effects of AZA and OLA, as well as SEL and SIR, on the growth of OVSAHO and KURAMOCHI cells
[0089] (AB) The growth inhibition of OVSAHO (A) and KURAMOCHI (B) cells by the indicated drug treatments was measured by MTT assay and calculated by comparing each drug treatment to the untreated control. Drug synergy was measured by the Bliss Independent model and the HSA model. The dashed lines in the figure in the right figure represent the expected drug combination effects based on the Bliss Independent model. Data are mean ± SD from biological replicates (n = 4). Statistical significance was analyzed by one-way ANOVA and Tukey post hoc test. * P < 0.05; # P < 0.05 indicates comparison with untreated control.
[0090] Fig.15 CRISPR-based double knockout screen identifies drug combinations that enhance protection against Parkinson's disease
[0091] (A) Distribution of barcode reads in plasmid and infected SK-N-MC-Cas9 cell pools. 99.1% (7,499 out of 7,569) and 98.7% (7,467 out of 7,569) of all expected gRNA combinations were obtained in the plasmid and cell pools, respectively. Most barcoded gRNA combinations were detected within 5-fold of the mean barcode reads for each combination (highlighted by the shaded area).
[0092] (B) The high correlation between barcode representation (normalized barcode counts) in the plasmid pool and the infected cell pool indicates efficient lentiviral delivery of the paired gRNA library into cells.
[0093] (C) The high correlation between barcode representation in cell pools cultured for 10 and 22 days indicates that CRISPR-mediated gene knockout does not result in severe cell death.
[0094] (D-E) Barcode representation between two biological replicates in untreated (D) and rotenone-treated (E) cells is highly reproducible. The dashed line in (D) indicates the threshold of 20 raw barcode counts. R is the Pearson correlation coefficient.
[0095] (F) SK-N-MC-Cas9 cells infected with barcoded paired gRNA libraries were treated with rotenone or remained untreated. Barcode representation within the cell pool was quantified using Illumina HiSeq.
[0096] (G) Lentiviral delivery of dual gRNA expression constructs effectively disrupts multiple target genes in SK-N-MC-Cas9 cells. Flow cytometry was used to measure the percentage of cell populations positive for GFP and RFP fluorescence on day 6 post-infection. Data are mean ± SD, n = 3 biological replicates.
[0097] (H) Volcano plot comparing the abundance changes of each barcoded gRNA combination in rotenone-treated versus untreated cell pools. The hit combination HSP90B1+HDAC2 is highlighted in red. Data were collected from two biological replicates.
[0098] (I) Survival rate of SK-N-MC-Cas9 cells infected with the indicated sgRNAs determined by MTT assay in the presence of rotenone. Data are shown as mean ± SD from biological replicates (n = 4), and the data for the left and right panels were obtained from the same experiment.
[0099] (JK) Cell viability of indicated drug-treated SK-N-MC cells (J) and iPSC-derived dopaminergic neurons (K) determined by MTT assay and DAPI uptake assay, respectively, in the presence of rotenone. Data are shown as mean ± SD from biological replicates (n = 9 in (J); n = 3 in (K)).
[0100] (L) In MPP + Cell viability of SK-N-MC cells treated with the indicated drugs determined by MTT assay in the presence of 5-nitro-2-nitro-4-acetyl-2-yl-4-acetyl-2 ...
[0101] (M) Quantification of the number of rods per ommatidium in wild-type and α-synuclein-expressing flies fed with the indicated drugs. The combination of 17-DMAG and vorinostat restored the number of rods per ommatidium in the eyes of flies expressing α-synuclein. Representative images show rods of wild-type and α-synuclein-expressing flies fed with the indicated drugs. Data are mean ± SD of at least 100 ommatidia from 5-10 flies.
[0102] Statistical significance in (IM) was analyzed by one-way ANOVA and Tukey post hoc test. * P<0.05, ** P < 0.01 and **** P<0.0001 indicates that there is a significant difference between the indicated samples. # P<0.05, ## P<0.01, ### P < 0.001 and #### P < 0.0001 indicates comparison with the corresponding no-gRNA control. The dashed line indicates the expected drug combination effect based on the Bliss Independent model.
[0103] Fig.16 .Engineered gRNA scaffold variants exhibited improved on-target activity and low off-target activity.
[0104] (A) Sequences of the gRNA scaffold sequences used.
[0105] (BC) OVCAR8-ADR cells containing reporter constructs with on-target (B) and off-target (C) sites were infected with lentivirus encoding wild-type or Opti-SpCas9. Editing efficiency of gRNA scaffold variants was measured as the percentage of cells with depleted RFP fluorescence.
[0106] (D) Evaluation of gRNA scaffold variants for efficient on-target editing using gRNA targeting endogenous loci. The percentage of sites with indels was measured using the T7 endonuclease I (T7E1) assay. The ratio of on-target activity to scaffold activity of the gRNA scaffold variants was determined, and the median and interquartile range of the normalized percentage of indel formation were shown for the 5 loci tested. Each locus was measured 3 times.
[0107] (E) GUIDE-seq genome-wide specificity profiles of gRNA scaffold variants paired with the indicated gRNAs. The positions of mismatches at off-target sites are colored, and GUIDE-seq read counts are used as a measure of cutting efficiency at a given site. 5. Detailed description of the invention
[0109] Combination drug treatments targeting multiple pathways could limit the development of drug-resistant phenotypes in cancer cells because it would be more difficult for cells to activate multiple compensatory survival mechanisms.
[0110] The present disclosure relates to systems and compositions that enable efficient, multiplexed genome editing and CRISPR screening. The engineered guide RNA scaffolds and promoters developed in this study enable ternary combinatorial CRISPR screening based on an assembly method previously known as CombiGEM-CRISPR, and can be used to improve genome editing efficiency in routine experiments and applications. High-throughput CRISPR screening from prior art can only study pairwise interactions between guide RNAs, but not those with higher complexity (such as interactions between three or more guide RNAs). Systematic screening of complex genetic interactions is achieved through the generation of engineered scaffolds and promoter sequences that minimize lentiviral vector recombination due to long homologous sequences and allow the construction of high-quality combinatorial guide RNA libraries using the same pool of guide RNA oligonucleotides.
[0111] Technically, the guide RNA scaffold sequences were engineered to minimize lentiviral vector recombination due to long homologous sequences, and the 3' ends of the promoter sequences of hH1 and mU6 were modified so that the same gRNA oligonucleotide pool could be used to construct a combinatorial guide RNA library. The engineered guide RNA scaffolds and modified promoter sequences showed higher or comparable activity in driving guide RNA expression compared to their wild-type counterparts.
[0112] 5.1 Design of CombiGEM-CRISPR v2.0
[0113] The CombiGEM-CRISPR v2.0 toolkit includes an additional design of library vectors so that only one set of reusable oligonucleotides to be synthesized can be used to perform high-order combinatorial CRISPR screens. Applicants and others have shown that CRISPR screens can be performed by simultaneously targeting two genes using dual guide RNA (gRNA) expression cassettes (Chow et al., 2019; Du et al., 2017; Han et al., 2017; Najm et al., 2018; Shen et al., 2017; Wong et al., 2016a). This paper evaluates the scalability of existing methods and other possible toolkits for assembling ternary combinatorial gRNA libraries for screening ( Figure 1 ). Consider using oligonucleotide synthesis-based methods to construct m a xm b xm cTo create a combined library, one must first design and synthesize an equal number of oligonucleotides containing the gRNA protospacer sequence (i.e., m a xm b xm c ) to include all combinations, followed by the insertion of promoter and scaffold sequences. Disadvantages are that (1) all intermediate cloning steps require large-scale bacterial transformation to maintain high library coverage, which is technically more demanding than the CombiGEM-CRISPR v2.0 method, which only requires large-scale transformation at the final assembly step, and (2) the rigid cloning framework does not allow the insertion of additional gRNAs after assembly to construct higher-order libraries. The inflexible workflow also limits the reusability of oligonucleotides in constructing more compact secondary screening libraries. For example, if paired gRNA libraries are constructed for identifying core effectors to be included in a ternary gRNA library, construction of a higher-order library requires the synthesis of a new set of oligonucleotides to complete the reassembly of the library. Alternative library assembly strategies are by methods based on Gibson (Gibson et al., 2009) or Golden-gate (Engler et al., 2008), which require the design of overlapping regions or complementary overhangs of adjacent gRNA expression units to fuse multiple parts together. However, these methods also do not allow the introduction of additional parts into pre-existing barcoded libraries. The CombiGEM-CRISPR v2.0 strategy provides a truly scalable protocol that enables multiplicative assembly of additional barcoded gRNA expression units to expand from n to n+1 combinatorial CRISPR libraries ( Figure 2 ). Libraries can also be customized and reassembled from selected gRNAs by selecting from a set of gRNAs stored in vectors for another screen or for a higher resolution secondary screen. In addition, creating a unique barcode for each assembly combination allows for its rapid characterization by short sequencing reads, thereby reducing sequencing costs and potential errors generated by long reads. Therefore, it was decided to adopt this approach to construct a triple combinatorial gRNA library for screening applications.
[0114] Since the previous CombiGEM toolkit cannot be directly applied to assemble ternary or even higher-order combinatorial gRNA libraries, this paper further created a "one-size-fits-all" design to construct an n-ary combinatorial CRISPR screening library of m guide RNAs using a multi-gRNA expression cassette that always only requires the synthesis of m (rather than nxm) pairs of oligonucleotides. To ensure the expression of three gRNAs in a single cell, multiple gRNA expression cassettes were assembled in one vector. Multiple promoters (including human U6, mouse U6, and human H1) (Adamson et al., 2016; Ma et al., 2014; Vidigal and Ventura, 2015) and modified gRNA scaffolds (Adamson et al., 2016; Briner et al., 2014; Dang et al., 2015; Grevet et al., 2018) were used in the expression cassettes. Figure 6 ), which minimizes lentiviral vector recombination due to long homologous sequences. However, the use of multiple promoters requires multiple sets of oligonucleotides, and / or additional PCR and restriction enzyme digestion reactions to construct different expression cassettes for library assembly. To allow researchers to synthesize and anneal only one set of oligonucleotides as building blocks to construct all possible high-order combinations of gRNAs for multiplex CRISPR screening, sequence adaptation was performed by modifying the 3' end sequence of the promoter to a sequence complementary to the sticky end of the annealed oligonucleotides, and these promoters expressed gRNAs and produced efficient gene knockout ( Figure 6 These new standardized vector components can be used for flexible assembly of high-order combinatorial gRNA libraries and scalable combinatorial CRISPR screening.
[0115] In some embodiments, the promoter has a sequence as shown below.
[0116] >Wild-type human U6 promoter:
[0117] AGAGGGCCTATTTCCCATGATTCCTTCATATTTGCATATACGATACAAGGCTGTTAGAGAGATAATTAGAATTAATTTGACTGTAAACACAAAGATATTAGTACAAAATACGTGACGTAGAAAGTAATAATTTCTTGGGTAGTTTGCAGTTTTAAAATTATGTTTTAAAATGGACTATCATATGCTTACCGTAACTTGAAAGTATTTCGATTTCTTGGCTTTATATATCTTGTGGAAAGGACGAAACACCG
[0118] >Modified mouse U6 promoter:
[0119] GATCCGACGCCGCCATCTCTAGGCCCGCGCCGGCCCCCTCGCACAGACTTGTGGGAGAAGCTCGGCTACTCCCCTGCCCCGGTTAATTTGCATATAATATTTCCTAGTAACTATAGAGGCTTAATGTGCGATAAAAGACAGATAATCTGTTCTTTTTAATACTAGCTACATTTTACATGATAGGCTTGGATTTCTATAAGAGATACAAATACTAAATTATTATTTTAAAAAACAGCACAAAAGGAAACTCACCCTAACTGTAAAGTAATTGTGTGTTTTGAGACTATAAATATCCCTTGGAGAAAAGCCTT CACCG
[0120] >Wild-type mouse U6 promoter (GeneBank: X06980.1)
[0121] GATCCGACGCCGCCATCTCTAGGCCCGCGCCGGCCCCCTCGCACAGACTTGTGGGAGAAGCTCGGCTACTCCCCTGCCCCGGTTAATTTGCATATAATATTTCCTAGTAACTATAGAGGCTTAATGTGCGATAAAAGACAGATAATCTGTTCTTTTTAATACTAGCTACATTTTACATGATAGGCTTGGATTTCTATAAGAGATACAAATACTAAATTATTATTTTAAAAAACAGCACAAAAGGAAACTCACCCTAACTGTAAAGTAATTGTGTGTTTTGAGACTATAAATATCCCTTGGAGAAAAGCCTT
[0122] >Modified human H1 promoter:
[0123] TTCGAACGCTGACGTCATCAACCCGCTCCAAGGAATCGCGGGCCCAGTGTCACTAGGCGGGAACACCCAGCGCGCGTGCGCCCTGGCAGGAAGATGGCTGTGAGGGACAGGGGAGTGGCGCCCTGCAATATTTGCATGTCGCTATGTGTTCTGGGAAATCACCATAAACGTGAAATGTCTTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGACCACTCTT C ACCG
[0124] >Wild-type human H1 promoter (GeneBank: X16612.1)
[0125] TTCGAACGCTGACGTCATCAACCCGCTCCAAGGAATCGCGGGCCCAGTGTCACTAGGCGGGAACACCCAGCGCGCGTGCGCCCTGGCAGGAAGATGGCTGTGAGGGACAGGGGAGTGGCGCCCTGCAATATTTGCATGTCGCTATGTGTTCTGGGAAATCACCATAAACGTGAAATGTCTTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGACCACTCTT
[0126] 5.2 Establishment of a multi-gene knockout system based on CRISPR-Cas9
[0127] A lentiviral combination gRNA expression vector containing multiple gRNA expression cassettes was constructed to effectively knock out three target genes simultaneously ( Figure 3 ), and their function in human cells (OVCAR8-ADR) was evaluated using gRNAs targeting exonic regions of green fluorescent protein (GFP), red fluorescent protein (RFP), and blue fluorescent protein (BFP) reporter genes. Lentivirus carrying the BFP reporter gene and combined gRNA units were generated to infect OVCAR8-ADR and OVCAR8-ADR-Cas9 cells stably expressing RFP and GFP reporter genes. Flow cytometric analysis was performed to confirm the efficient generation of double and triple knockouts of the reporter proteins (>81% and >74%, respectively) ( Figure 3 ). Similar knockout efficiencies were observed when wild-type (WT) or modified gRNA scaffolds were used. The results suggest that lentiviral vectors can be used to deliver gRNAs to generate multi-gene knockouts and that this vector design was used for the combinatorial CRISPR-Cas9 screen in this study.
[0128] 5.3 Ternary Combination CRISPR-Cas9 Screening for Synergistic Anticancer Gene Combinations
[0129] Combination drug treatments targeting multiple pathways can limit the development of drug-resistant phenotypes in cancer cells because it is more difficult for cells to activate multiple compensatory survival mechanisms (Bozic et al., 2013). A high-throughput study was conducted to search for effective therapeutic combinations for high-grade serous ovarian cancer (HGSOC), the most prevalent subtype that contributes to two-thirds of all ovarian cancer deaths (Bowtell, 2010). Using the CRISPR-based multi-gene knockout system described above, CombiGEM-CRISPRv2.0 was used to assemble a high-coverage (99.8%) ternary combinatorial gRNA library (with 32x32x32 gRNAs = 32,768 total combinations) ( Figure 4 AE). The library includes 15 druggable protein-coding genes (2 gRNAs per gene) that are commonly targeted in anticancer therapy and whose expression is reported in OVCAR8-ADR, an established HGSOC cell model (Baratta et al., 2015) and other ovarian cancer cells based on the NCI-60 proteome database (Gholami et al., 2013) to demonstrate the feasibility of this approach (Tables S2 and S3). In order to generate combinatorial gene knockouts at a high rate, gRNAs with predicted high on-target (and low off-target) activity were selected based on the Azimuth 2.0 model (Doench et al., 2016). The on-target efficacy scores and indel generation efficiencies of gRNAs were compared, and it was observed that the on-target efficacy scores largely predicted the efficiency of gRNAs in OVCAR8-ADR cells ( Figure 7 A). gRNAs with on-target scores > 0.64 were selected based on the Azimuth 2.0 model, and most of them also had high frameshift generation rates when analyzed under the newly developed inDelphi and FORECasT models ( Figure 8 )(Allen et al., 2018; Shen et al., 2018). Two non-targeting control gRNAs from the GeCKOv2 library (Shalem et al., 2014) that do not target loci in the human genome were also included in the library as references.
[0130] Table S2
[0131] List of druggable genes included in ovarian cancer research
[0132]
[0133]
[0134]
[0135] Table S3
[0136] List of gRNAs used in this study
[0137]
[0138]
[0139]
[0140] A pooled screen was then performed to isolate ternary gRNA combinations that modulated the growth of OVCAR8-ADR ( Figure 4 F). Barcode abundance and yield log2 values were compared between day 15 and day 26 groups as a measure of cell growth. Based on the genetic screen data, 3 (out of 455) triplet combinations had an average log2 ratio <-1 (based on data obtained from two biological replicates with at least a 50% reduction in barcode counts in cells cultured on day 26 compared to day 15) and a GI3 score <-0.2 ( Figure 4 G and Table S4). These triple combinations (targeting DNMT1+POLA1+EGFR or ERBB2, and CDK4+MAP2K1+POLA1) also showed growth regulatory effects that were significantly different from their respective single and pairwise combinations ( Figure 4 H and Table S4). Because they showed strong growth inhibition and strong synergistic genetic interactions, these three combinations were selected for further characterization. Their growth inhibition (51.4% to 88.4% reduction for the six tested gRNA combinations for DNMT1+POLA1+EGFR or ERBB2, and 50.5% and 54.9% reduction for the two tested gRNA combinations for CDK4+MAP2K1+POLA1) was further confirmed by single non-pooled assays and was not a false positive caused by excessive DNA double-strand breaks generated by CRISPR-Cas9 (Aguirre et al., 2016; Munoz et al., 2016), because simultaneous targeting of three safe harbor loci did not result in strong growth arrest ( Figure 4 I, 7C and 9). In addition, based on analysis using CellMiner, the DNA copy numbers of the DNMT1, POLA1, EGFR, ERBB2, CDK4 and MAP2K1 loci were not significantly amplified in the genome of OVCAR8-ADR (Reinhold et al., 2012).
[0141] Table S4
[0142]
[0143]
[0144]
[0145]
[0146]
[0147]
[0148]
[0149] 5.4 Validation of screening hits and matching drug combinations
[0150] Azacitidine (AZA), fludarabine (FLU), and erlotinib (ERL) are used to target DNMT1, POLA1, and EGFR, respectively. Lapatinib (LAP) is used to inhibit ERBB2, and it also acts on EGFR, which belongs to the same Erb protein family. The triple therapy of AZA, FLU, and ERL / LAP showed significantly stronger growth inhibition than single-drug and double-drug therapy ( Figure 5 AB), and also resulted in a different set of perturbed genes when cells treated with the three-drug regimen were compared with all corresponding two-drug combinations, including genes involved in cell cycle regulation ( Fig.10 The DiaMOND scoring method was used to confirm the synergistic effect of the three drugs in inhibiting the growth of ovarian cancer cells ( Figure 5 C) (Cokol et al., 2017). In OVCAR8-ADR cells, the fractional inhibitory concentration (FIC) scores of AZA+FLU+ERL and AZA+FLU+LAP were 0.64 and 0.8, respectively. By comparing the half-maximal inhibitory concentration (IC 50 ) required, it was found that the three-drug formulation of AZA+FLU+ERL required approximately 4.5-fold less of each component (for AZA+FLU+LAP, the reduction was approximately 3.7-fold). 30 、IC 40 and IC 60 ) observed similar results ( Fig.11 The synergistic effect between the three drugs (AZA+FLU+ERL or LAP) was also detected in OVSAHO cells ( Fig.12 The synergistic effect of three matching drugs (ribociclib (RIB) + trametinib (TRA) + FLU) of the third screening hit (i.e., CDK4 + MAP2K1 + POLA1) in inhibiting the growth of ovarian cancer cells was also confirmed ( Figure 5D) These results demonstrate that this platform enables high-throughput screening and identification of synergistic ternary therapeutic combinations.
[0151] 5.5 CRISPR-Cas9 screening of drug pairs for ovarian cancer
[0152] By using CombiGEM-CRISPR v2.0 ( Figure 6 B) one-pot reaction, a paired gRNA library targeting 52 druggable genes (3 gRNAs per gene; Tables S2 and S3) was constructed, and its expression was shown in OVCAR8-ADR and other ovarian cancer cells based on the NCI-60 proteome database (Gholami et al., 2013). GRNAs with a target score score> 0.63 (with a predicted efficiency> about 80%) were selected, except for one with a score of 0.60. Three control gRNAs from the GeCKOv2 library (Shalem et al., 2014) that do not target loci in the human genome were included as references. The paired gRNA library pool (with 159x 159gRNA=25,281 total combinations) was then delivered to OVCAR8-ADR cells via lentivirus. Using the experimental pipeline established by the applicants (Wong et al., 2015; Wong et al., 2016), Illumina HiSeq was performed to confirm the high coverage of paired libraries (>99.0%) and the high correlation of barcode representation between plasmid and infected cell pools ( Fig.13 AE). Using a similar time window used in our previous study, barcode abundance was compared between day 15 and day 21 groups to log2 values of production to measure cell growth ( Fig.13 F). Based on the selection criteria requiring an average log2 ratio < -1 (at least 50% reduction in barcode counts on day 21 compared to cells cultured on day 15 based on data obtained from two biological replicates) and the detection of multiple gRNAs targeting the same gene pair with at least P < 0.1, two combinations (PARP1+DNMT1 and CDK2+FKBP1A in blue and red, respectively) were Fig.13 The growth inhibition brought about by these two double gene knockouts was verified using separate non-combined assays and was not caused by the knockout of either gene ( Fig.13 H; 7B).
[0153] Then, the growth inhibitory effect brought about by the combination of these two hits was evaluated by treating OVCAR8-ADR cells with drug pairs. Olaparib (OLA), azacitidine (AZA), celecoxib (SEL), and sirolimus (SIR) were used as drugs targeting PARP1, DNMT1, CDK2, and FKBP1A, respectively. These drug molecules have been reported to have potent effects on their targets (McClue et al., 2002; Muvarak et al., 2016; Sabers et al., 1995; Wishart et al., 2006; Yang et al., 2017). Our results showed that OLA and AZA synergistically inhibited the growth of OVCAR8-ADR cells ( Fig.13 IK) and induces G2 cell cycle arrest ( Fig.13 L), while combined treatment with SEL and SIR exerted an additive effect, inhibiting cell growth ( Fig.13 MP). Similar growth inhibitory effects were observed when these drugs were used in combination to treat OVSAHO and KURAMOCHI (two other characterized cell models of HGSOC, Coscia et al., 2016; Domcke et al., 2013) Fig.14 Our results also confirm the observation that co-administration of the PARP1 inhibitor talazoparib and the DNMT1 inhibitor guadecitabine synergistically inhibited tumor growth in PEO1 and PEO4 cells as well as in the OVCAR4 xenograft model (Pulliam et al., 2018), further suggesting the PARP1+DNMT1 inhibitor combination as an effective treatment option for ovarian cancer.
[0154] 5.6 CRISPR-Cas9 screening of drug pairs targeting Parkinson's disease
[0155] This screening approach can also be used to find effective therapeutic combinations to enhance protection against toxicities associated with other disease phenotypes such as Parkinson's disease (PD). Another high-coverage (99.1%) paired gRNA library targeting 28 druggable genes was assembled, whose ablation or matched pharmacological inhibitors have been reported to suppress neuronal toxicity ( Fig.15 AE; Table S9). The library was delivered into SK-N-MC-Cas9 cells via lentivirus to generate double gene knockouts, and the cells were then treated with rotenone to induce PD-related toxicity ( Fig.15FG). Barcode abundance was compared between rotenone-treated and untreated groups to identify enriched gRNA combinations that protected cells from rotenone-induced toxicity. Based on the selection criteria requiring an average log2 fold change > 0.378 (at least 30% more barcode counts in rotenone-treated cells compared to untreated cells based on data obtained from two biological replicates) and detection of multiple gRNAs targeting the same gene pair with a P < 0.05, our genetic screen identified HSP90B1+HDAC2 as the top hit (where six gRNA combinations were identified with an average 51.6% increase in barcode counts) that improved cell survival after rotenone treatment ( Fig.15 H; Table S8). The protective effect of simultaneous knockout of HSP90B1+HDAC2 was confirmed in a non-combined assay and was greater than that of knockout of either gene ( Fig.15 I). The protective effect of this identified combination was further confirmed by matching drugs (17-(dimethylaminoethylamino)-17-demethoxygeldanamycin (17-DMAG) + vorinostat). Fig.15 J) and iPSC-derived dopaminergic neurons ( Fig.15 K), combined drug treatment enhanced the survival of cells against rotenone-induced toxicity compared with single-drug treatment. In addition, this drug combination was observed to reduce the toxicity induced by MPP+ treatment in cultured cells ( Fig.15 L) and α-synuclein expression in transgenic flies ( Fig.15 M) (two other well-characterized PD models). Our results demonstrate the versatility of our platform for screening therapeutic combinations targeting different disease phenotypes, including those for alleviating neurodegenerative disease phenotypes.
[0156] Table S8
[0157] List of paired gRNA combinations with average log2 ratio > 0.378 and P < 0.05 identified in Parkinson's disease research
[0158]
[0159] Table S9
[0160] List of druggable genes included in Parkinson's disease research
[0161]
[0162] 5.7 Discussion
[0163] In summary, a CRISPR-based multi-gene knockout screening platform has been established to address the unmet need for rapid identification of effective ternary therapeutic combinations. By pairing drug mechanisms of action with specific genes, it helps to accelerate the identification of effective combinations for guiding secondary screening and narrow down the large number of possible combinations to a few best-performing hits for further testing. It has been demonstrated that systematic characterization of ternary combinations using CRISPR-based screening discovered rare combinations with synergistic interactions, as the majority of the combinations showed buffering interactions, and was able to verify that all three screening hits had strong growth inhibitory effects and ternary interactions. The CombiGEM-CRISPR v2.0 platform has a wide range of uses as it can also be used to identify inhibitors of cancer cell growth ( Figure 13-14 ) and enhanced protection against other disease phenotypes such as Parkinson's disease (PD)-related toxicity ( Fig.15 ) and through the use of additional engineered promoters and scaffolds (including the newly engineered v3.11, v.3.12, and v3.13 scaffolds (see Fig.16 )) was extended to analyze interactions between >3 genetic components for CRISPR-based multiplex editing (Reis et al., 2019).
[0164] like Fig.16 As described in, the engineered gRNA scaffold variants exhibit improved on-target activity and low off-target activity. A, sequence of the gRNA scaffold sequence used. B, C, OVCAR8-ADR cells containing reporter constructs with on-target (B) and off-target (C) sites were infected with lentivirus encoding wild-type or Opti-SpCas9. The editing efficiency of the gRNA scaffold variants was measured as the percentage of cells with depleted RFP fluorescence. D, gRNA scaffold variants were evaluated for effective targeted editing using gRNA targeting endogenous loci. The percentage of sites with indels was measured using the T7 endonuclease I (T7E1) assay. The ratio of on-target activity to scaffold activity of the gRNA scaffold variants was determined, and the median and interquartile range of the normalized percentage of indel formation were shown for the five loci tested. Each locus was measured 3 times. E, GUIDE-seq genome-wide specific spectrum of the gRNA scaffold variant group paired with the indicated gRNA. The mismatch position of the off-target site is colored, and the GUIDE-seq read count is used as a measure of the efficiency of cutting at a given site.
[0165] The platform is also versatile and can be used with the dCas9-based CRISPR interference system (Qi et al., 2013) to partially reduce target gene expression to mimic the effects of drug inhibitors. The platform can be combined with other technologies such as single-cell RNA-seq to explore different cellular features and contribute to the generation of druggable gene interaction networks using existing knowledge (Adamson et al., 2016; Bassik et al., 2013; Chow et al., 2019; Du et al., 2017; Han et al., 2017; Shen et al., 2017). The platform proposed in this study is easy to implement and will be of great value for perturbing multi-layer genetic networks to understand complex biological systems and design new combination therapies. Example
[0166] Experimental model and subject details
[0167] Cell culture and generation of cell lines
[0168] HEK293T (female) and SK-N-MC (female) cells were obtained from the American Type Culture Collection (ATCC). OVCAR8-ADR (female) cells were gifted by T. Ochiya (National Cancer Center Research Institute, Japan). (Honma et al., 2008) The identity of OVCAR8-ADR cells was confirmed by cell line identification test (Genetica DNA Laboratories). KURAMOCHI (female) and OVSAHO (female) cells were obtained from the Japan Research Bioresource (JCRB) cell bank. iPSC-derived dopaminergic neurons were obtained from TGD Life Company Limited. OVCAR8-ADR-Cas9 and SK-N-MC-Cas9 cells were generated by transducing pAWp30 (Addgene, 73857) into OVCAR8-ADR and SK-N-MC cells, respectively, and then selected using zeocin (Life Technologies) to obtain stable Cas9-integrated cells. Streptococcus pyogenes Cas9 was used in this study. OVCAR8-ADR reporter cells stably expressing RFP and GFP were generated by transducing cells with pAWp9 and then classified according to GFP and RFP signals. Reporter cells were then infected with pAWp30 to stably integrate Cas9 after zeocin selection. HEK 293T and SN-MC cells were cultured at 37°C and 5% CO2 in DMEM supplemented with 10% FBS and 1X antibiotic-antimycotic (Life Technologies). KURAMOCHI, OVSAHO and OVCAR8-ADR cells were cultured at 37°C and 5% CO2 in RPMI 1640 supplemented with 10% FBS and 1X antibiotic-antimycotic. Cells were checked for mycoplasma contamination every three or four months and were never tested positive.
[0169] Method details
[0170] Plasmid construction
[0171] The vectors used in this study (Table S5) were generated by a standard molecular cloning strategy including PCR, oligonucleotide annealing, restriction enzyme digestion, ligation, and Gibson assembly. Customized oligonucleotides were purchased from Genewiz. The vectors were transformed into competent cells of Escherichia coli strain DH5α and selected with ampicillin (100 μg / ml, USB) or carbenicillin (50 μg / ml, Teknova). DNA was extracted and purified by Plasmid Mini (Takara and Tiangen) or Midi preparation (Qiagen) kits. The vector sequence was verified by Sanger sequencing.
[0172] Table S5
[0173]
[0174] To construct storage vectors with mouse U6 (mU6) and human H1 (hH1) promoter-gRNA WT scaffold sequences, promoter sequences were amplified from mouse and human genomic DNA, respectively, and cloned into the vector backbone of pAWp28 (Addgene, 73850). pAWp28 is a storage vector with human U6 (hU6) promoter-gRNA WT scaffold sequences. Storage vectors with hU6-gRNA v1 scaffolds and mU6-gRNA v2 scaffolds were generated by PCR-based mutagenesis. To drive gRNA expression to target genes of interest, oligonucleotide pairs with gRNA target sequences were synthesized, annealed, and cloned into BbsI-digested storage vectors by T4 DNA ligase (New England Biolabs). To generate a lentiviral vector for expressing a single gRNA targeting GFP, RFP, and BFP genes, the gRNA expression cassette released from the storage vector was digested with BglII and EcoRI enzymes (Thermo Fisher Scientific) and cloned into the pAWp9 vector (Addgene, 73851) using compatible sticky ends generated by digesting the vector with BamHI and EcoRI enzymes (Thermo Fisher Scientific). To construct a lentiviral vector to express multiple gRNAs targeting fluorescent proteins, the second gRNA expression cassette with the hU6-gRNA-v1 (or WT) scaffold was released from the storage vector by digestion with BglII and EcoRI enzymes and connected to the BamHI and EcoRI digested storage vector containing the first gRNA expression cassette with the hH1-gRNA-WT scaffold. Similarly, the third gRNA expression cassette with the mU6-gRNA-v2 (or WT) scaffold was released from the storage vector by digestion and connected to the storage vector containing the first and second gRNA expression cassettes. Lentiviral vectors were then generated by PCR amplification of paired or ternary gRNA expression cassettes from storage vectors and cloned into SbfI-digested pFUGW vector backbone (pAWp40) by Gibson assembly.
[0175] Design and assembly of guide RNA libraries
[0176] The gRNAs used in this study were designed based on GPP sgRNA Designer (Table S3). For paired gRNA libraries, three gRNAs were selected for each target gene based on the following criteria: 1) on-target efficacy score>0.6; 2) off-target grade<100; and 3) target site was within 5-65% of the protein-coding sequence. gRNA sequences containing BamHI, EcoRI, BglII, and MfeI digestion sites were excluded to avoid incompatibility with CombiGEM. For the ternary combinatorial gRNA library, two gRNAs were selected for each target gene using the same criteria, but their on-target efficacy scores were all>0.64. inDelphi and FORECasT were used to predict the frameshift rate of gRNA. The gRNA sequence was entered into BLAST to extract the 70-nucleotide context sequence of the gRNA. The PAM sequence index is located in the 70-nt sequence and is entered into inDelphi and FORECasT downloaded from GitHub together with the context sequence. The K562 cell line was the prediction model used in Delphi, and the output frameshift score was extracted from the "Frameshift Frequency" option. The output summary file of FORECasT was input into the Python code, and the predicted frameshift frequency was calculated by summing the percentage of target frameshift categories that were not multiples of three and then dividing the sum by 10.
[0177] To assemble the gRNA library ( Figure 6B), oligonucleotide pairs were annealed to the gRNA target sequence, two BbsI restriction digestion sites, and a unique 8-bp barcode, combined in equal molar ratios, and cloned into storage vector backbones containing hH1, hU6, and mU6 promoter sequences (pAWp92, AWp28, and pAWp100, respectively). The gRNA scaffold sequences (WT, v1, and v2) were then inserted into the vector to generate three pooled libraries of barcoded single gRNA expression cassettes. The combined gRNA library was assembled using the CombiGEM-CRISPR method (Wong et al., 2016a). The first set of inserts was released from the storage vector containing the hH1-gRNA-WT scaffold by digestion with BglII and MfeI enzymes and ligated into pAWp12 (Addgene, 72732) digested with BamHI and EcoRI to generate a barcoded single gRNA library in a lentiviral vector. Then, the second set of inserts was released from the storage vector containing the hU6-gRNA-v1 scaffold by digestion and connected to a lentiviral vector containing the first gRNA expression cassette to generate a barcoded paired gRNA library. Similarly, the third set of inserts was released from the storage vector containing the mU6-gRNA-v2 scaffold and inserted into a lentiviral vector containing two gRNA expression cassettes to generate a barcoded ternary combination gRNA library. The ternary combination gRNA library was delivered to OVCAR8-ADR cells using lentivirus, and genomic DNA extracted from single cell-derived clones was subjected to Sanger sequencing analysis, and most of the assembled barcoded gRNA constructs (7 of 8 colonies) were confirmed to contain the expected gRNA target sequence. In order to construct the single combination gRNA vector used in the validation experiment, the same assembly strategy was used except that the annealed oligonucleotide pairs were not merged.
[0178] Lentiviral vector generation and transduction
[0179] This study used a second-generation lentiviral vector system. HEK293T cells were transfected with FuGene HD transfection reagent (Promega) in 6-well plates according to the manufacturer's instructions, with 0.5 μg pCMV-VSV-G, 1 μg pCMV-dR8.2-dvpr and 0.5 μg of each lentiviral vector per well. The supernatant containing lentivirus was collected 48 and 72 hours after transfection, and then combined and filtered through a 0.45 μm polyethersulfone membrane (Pall). For conventional transduction, 300 μL of filtered supernatant was added to one well of a 12-well plate in the presence of 8 μg / ml polybrene (Sigma), where the cell confluency was about 30%. For library transduction, Cas9-expressing cells were seeded at approximately 50% confluence in 150-mm dishes with a cell number roughly equal to 400 times the library size and transduced with virus at a multiplicity of infection (MOI) of approximately 0.3 to ensure that most cells were infected with only one type of viral particle.
[0180] Flow cytometry, cell cycle analysis, and cell sorting
[0181] To prepare samples for flow cytometry, cells were trypsinized and resuspended in FACS buffer (PBS containing 2% FBS). BD LSR Fortessa analyzer (Becton Dickinson) was used to detect the signals of TurboRFP, EGFP and mTagBFP by 561nm yellow-green laser (610 / 20nm), 488nm blue laser (530 / 30nm) and 405nm violet laser (450 / 50nm), respectively. For cell cycle analysis, cells were fixed at 4°C for 1 hour with ice-cold 70% ethanol and then rehydrated at room temperature by replacing ethanol with PBS for 15 minutes. In order to remove RNA, RNase A (10mg / ml) was added to the cells and incubated at 37°C for 15 minutes. Genomic DNA content was stained in the dark for 1 hour at room temperature with propidium iodide (PI; Invitrogen). Signals were detected by 561nm yellow-green laser (586 / 15nm) using BD LSR Fortessa analyzer. FlowJo software (v10.5.3, Becton Dickinson) was used for data analysis. For cell sorting, samples were prepared similarly to FACS analysis except that 2X antibiotic-antimycotic was added to the FACS buffer. A BD Influx cell sorter (Becton Dickinson) equipped with a 100-μm nozzle (24psi, frequency of 39.2kHz) was used. GFP-positive cells were detected by a 488nm blue laser (530 / 40nm) and sorted using the 1.0Drop Pure mode. For cells infected by the screening library, 1-2% of cells with the strongest GFP signal were not collected to minimize the chance of obtaining cells infected by multiple virions. Cells at least 100 times more than the library size were collected.
[0182] Sample preparation for barcode reading
[0183] For the library-transduced cell pools, genomic DNA was extracted from the cells using the DNeasy Blood and Tissue Kit (Qiagen) and quantified by the Quant-iT PicoGreen dsDNA Assay Kit (Life Technologies). To extract fragments containing the 298 bp barcode, 0.5 ng of library plasmid DNA and 800 ng of genomic DNA were used in each 50 μl PCR reaction and PCR amplification was performed using Kapa HiFi Hotstart Ready-mix (Kapa Biosystems). The forward and reverse primers used were 5'-GGATCCGCAACGGAATTC-3' and 5'-GGTTGCGTCAGCAAACACAG-3'. PCR amplification was maintained in the exponential phase to minimize PCR bias. To ensure adequate library coverage was amplified from genomic DNA, 20 and 10 PCR reactions were performed for paired libraries used to study ovarian cancer and Parkinson's disease, respectively, and 30 PCR reactions were performed for the triple combinatorial library. Illumina adapters and sequencing indexes were then added to the amplicons by PCR using Kapa HiFi Hotstart Ready-mix. The forward and reverse primers used were 5'-CAAGCAGAAGACGGCATACGAGATGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGGTTGCGTCAGCAAACACAG-3' and 5'-AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT NNNNNNNN(N 1-4 ) GGATCCGCAACGGAATTC-3', NNNNNNNN represents the specific index barcode assigned to each experimental sample and (N 1-4 ) represents 1 to 4 nucleotides added to increase the diversity of the sequencing library. The final amplicon was purified by two rounds of size selection using Agencourt AMPure XP beads (Beckman Coulter Genomics) at a ratio of 1:0.5 and 1:0.95. The quantity and quality of the samples were measured by real-time PCR using Kapa SYBR Fast qPCR Master Mix (Kapa Biosystems) with primer pairs 5'-AATGATACGGCGACCACCGA-3' and 5'-CAAGCAGAAGACGGCATACGA-3' and analyzed using a high-sensitivity DNA chip (Agilent) on an Agilent 2100 bioanalyzer.
[0184] Barcode sequencing data analysis
[0185] Barcode reads were processed from sequencing data and normalized to read counts per million to facilitate comparison between samples. Normalized barcode counts for each gRNA combination in the cell pool were compared to barcode counts for a dummy control gRNA combination in each sample to generate log2-transformed fold changes. To improve data reliability, combinations with raw barcode reads <100 in samples from early time points in the ovarian cancer study were excluded ( Figure 4 C and 13C). In the untreated samples of the Parkinson's disease study, combinations with <20 raw barcode reads were filtered out because of the low total read count ( Fig.15 D). In addition, only combinations containing gRNAs targeting different genes and with a coefficient of variation (CV) < 1 were included in the analysis ( Figure 4 E and 13E). Overall, the ovarian cancer study included 97.3% (22,921 of 23,556 combinations) and 79.1% (21,172 of 26,782 combinations) of pairwise and triadic combinations, respectively. 96.7% (6,878 of 7,108 combinations) of pairwise combinations were included in the Parkinson's disease study. In the ovarian cancer study, gRNA combinations with a log2 fold change <-1 are listed in Tables S6 and S7. In the Parkinson's disease study, gRNA combinations with a log2 fold change >0.378 (increase of approximately 30%) and P<0.05 are listed in Table S8. To identify synergistic triadic combinations, gRNA combinations targeting three non-redundant genes with a coefficient of variation <1 were grouped into a total of 455 unique triadic targeting combinations. R pachage DescTools was used to perform Dunnett's test, with the triadic combination as a control and compared with the di- and single-gene targeting combinations (a total of 6 comparisons). To measure genetic interactions, a scoring system similar to that described previously was applied (Wong et al., 2015). Genetic interaction (GI) scores were calculated by subtracting the expected fold change of the triple knockout, which was estimated by the sum of the fold changes of the individual double and single gene knockouts, from the observed fold change of the triple knockout, where negative GI scores represent genetic synergy in this study. The GI score calculated for a given triple combination represents the interaction between the third gene and the remaining two gene combinations. The GI3 score for the triple combination was calculated based on the geometric mean of the three GI scores. Potential synergistic gene combinations were selected by adjusting the P value threshold <0.05, the average log2 fold change <-1, and the GI3 score <-0.2 in all 6 comparisons of the Dunnett test (the GI score was also <-0.14 for the same triple combination in all 3 possible permutations ("A, B" + "C", "A, C" + "B", "B, C" + "A")).
[0186] Table S6
[0187] List of ternary gRNA combinations with average log2 ratio < -1 and P < 0.05 identified in ovarian cancer studies
[0188]
[0189]
[0190]
[0191]
[0192]
[0193]
[0194]
[0195]
[0196]
[0197]
[0198]
[0199]
[0200]
[0201]
[0202]
[0203]
[0204]
[0205]
[0206] Table S7
[0207] List of gRNA combinations for paired targeted genes with average log2 ratios < -1 identified in ovarian cancer studies
[0208]
[0209]
[0210] Cell viability assay and drug interaction analysis
[0211] One day before drug treatment, 1,500 OVCAR8-ADR cells were seeded into one well of a 96-well plate. 4,800 SK-N-MC cells were seeded into one well of a 96-well plate and pretreated with drugs for 72 hours before the addition of rotenone (Abcam, ab143145) or MPP. + (Abcam, ab144783) to induce toxicity. Drugs were administered at the indicated doses. Azacitidine (A-5959), olaparib (O-9201), sirolimus (R-5000), celecoxib (R-1234), lapatinib (L-4899), erlotinib (E4007), and vorinostat (V-8477) were purchased from LC Laboratories. Fludarabine (#14128) was purchased from Cayman Chemical Company. 17-DMAG (A2213) was purchased from ApexBio. 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was performed as previously described (Wong et al., 2015) to assess cell growth at different time points. Briefly, the medium in the cell growth wells was replaced with 100 μl of 1X MTT solution in RPMI 1640 without phenol red and incubated at 37 ° C and 5% CO2 for 3 hours. Then 100 μl of solubilization buffer (10% Triton X-100, 0.1N HCl in isopropanol) was added to each well to dissolve the blue formazan crystals. The absorbance was measured at 570 nm and 650 nm by a VARIOSKAN FLASH microplate reader (Thermo Scientific). The Bliss independence model (Bliss, 1939) and the HSA model (Borisy et al., 2003) were used to evaluate the interaction between drug pairs, and the DiaMOND model (Cokol et al., 2017; Cokol-Cakmak et al., 2018) was used to measure ternary drug interactions. The excess over the Bliss independence model was calculated as g12-(g1+g2-g1 x g2 / 100), where g represents the percentage of growth inhibition and the numbers represent the drug components; the excess over the HSA model was calculated by subtracting the highest growth inhibition of a single agent from the combination; the DiaMOND model was used to calculate the fractional inhibitory concentration (FIC3) equal to (o1+o2+o3) / ((e1+e2+e3) / 3), where o represents the observed concentration of each component in the combination and e represents the expected concentration of a single drug at a certain level of inhibition as determined by the drug response curves. To generate the drug response curves, the three drugs were combined at their respective IC 50The fixed ratio of 1:1:1 is combined and scaled. If FIC3 < 1, the interaction is synergistic; if FIC3 = 1, the interaction is additive; if FIC3 > 1, the interaction is antagonistic.
[0212] Colony formation assay
[0213] One day before drug treatment, 1,000 OVCAR8-ADR cells were seeded into one well of a 6-well plate at the indicated dose. Colonies were fixed with ice-cold methanol at -20°C for 30 min and stained with crystal violet. The number and area of colonies were determined by ImageJ software.
[0214] Prosthetic eye measurement
[0215] Wild-type and transgenic Drosophila strains carrying gmr-GAL4 and UAS-α-syn were used (Auluck et al., 2002). gmr-GAL4 female flies were crossed with w 1118 (control) or UAS-α-syn male fruit flies were crossed and cultured at 21.5°C on cornmeal medium supplemented with drugs or vehicle controls. Drug treatment was performed by adding vorinostat (LC Laboratories, V-8477) and / or 17-DMAG (InvivoGen, ant-dgl-25) to 2ml culture medium at a final concentration of 0.5μM and 96μg / ml, respectively. After eclosion, the offspring were transferred to a new vial, in which fresh drugs were added to the culture medium, and the drugs were replaced every 3-4 days before the assay. As previously described (Wong et al., 2008), the eyes of 7- to 11-day-old fruit flies were examined using a 60X oil objective under an optical microscope (Olympus CX31). At least 100 ommatidia from 5-10 fruit flies were examined and the number of rods was recorded.
[0216] RNA-seq
[0217] Total RNA was isolated from drug-treated OVCAR8-ADR cells by MiniBEST Universal RNA Extraction Kit (Takara). RNA samples were quantified and analyzed using Qubit assay and high sensitivity DNA chips (Agilent) on Agilent 2100 Bioanalyzer, respectively. RNA-seq experiments were performed at the Center for Genome Sciences (LKS Faculty of Medicine, The University of Hong Kong). Illumina adapters of paired-end raw sequence reads were adjusted by Trimmomatic 0.39. STARaligner version 2.7 was used to align sequence reads to the human genome, where the genome index was constructed using the primary assembly of the human genome released by Gencode version 30. Raw count reads were extracted using the R package Rsubread. RpackagesEdgeR, limma, and HTSFilter were used for differential expression analysis comparing each pairwise and three-way drug combination to untreated samples. An FDR < 0.05 filter was applied to the three-way combination compared to untreated samples, while an FDR < 1 filter was applied to each two-way combination compared to untreated samples. Combinations and genes were clustered by complete linkage clustering, with distance defined as 1-Pearson correlation. Genes that were at least 20% upregulated or downregulated in cells treated with the three-drug regimen compared to untreated controls were entered into the DAVID web tool for pathway analysis and used the Reactome pathway database. Pathway mapping used a P = 0.05 threshold.
[0218] Quantification and statistical analysis
[0219] Data analysis was performed using GraphPad Prism 7 software (GraphPad software). Data presented are mean ± SD, and biological replicates are specified for each experiment in the figure legends. Statistical comparisons between two groups were performed by Student's t-test, while one-way ANOVA followed by Tukey or Dunnett's post hoc test was used for comparisons of more than two groups.
[0220] Data and code availability
[0221] All sequencing data generated or analyzed in this study are available.
[0222] Key Resources Table
[0223]
[0224]
[0225] References
[0226] Adamson,B.,Norman,T.M.,Jost,M.,Cho,M.Y.,Nunez,J.K.,Chen,Y.,Villalta,J.E.,Gilbert,L.A.,Horlbeck,M.A.,Hein,M.Y.,et al.(2016).A Multiplexed Single-Cell CRISPR Screening Platform Enables Systematic Dissection of the UnfoldedProtein Response.Cell 167,1867-1882e1821.
[0227] Aguirre,A.J.,Meyers,R.M.,Weir,B.A.,Vazquez,F.,Zhang,C.Z.,Ben-David,U.,Cook,A.,Ha,G.,Harrington,W.F.,Doshi,M.B.,et al.(2016).Genomic Copy NumberDictates a Gene-Independent Cell Response to CRISPR / Cas9 Targeting.CancerDiscov 6,914-929.
[0228] Al-Lazikani,B.,Banerji,U.,and Workman,P.(2012).Combinatorial drugtherapy for cancer in the post-genomic era.Nat Biotechnol 30,679-692.
[0229] Allen,F.,Crepaldi,L.,Alsinet,C.,Strong,A.J.,Kleshchevnikov,V.,DeAngeli,P.,Palenikova,P.,Khodak,A.,Kiselev,V.,Kosicki,M.,et al.(2018).Predicting the mutations generated by repair of Cas9-induced double-strandbreaks.Nat Biotechnol.
[0230] Auluck,P.K.,Chan,H.Y.,Trojanowski,J.Q.,Lee,V.M.,and Bonini,N.M.(2002).Chaperone suppression of alpha-synuclein toxicity in a Drosophilamodel for Parkinson's disease.Science295,865-868.
[0231] Baratta,M.G.,Schinzel,A.C.,Zwang,Y.,Bandopadhayay,P.,Bowman-Colin,C.,Kutt,J.,Curtis,J.,Piao,H.,Wong,L.C.,Kung,A.L.,et al.(2015).An in-tumorgenetic screen reveals that the BET bromodomain protein,BRD4,is a potentialtherapeutic target in ovariancarcinoma.Proc Natl Acad Sci U S A 112,232-237.
[0232] Bassik,M.C.,Kampmann,M.,Lebbink,R.J.,Wang,S.,Hein,M.Y.,Poser,I.,Weibezahn,J.,Horlbeck,M.A.,Chen,S.,Mann,M.,et al.(2013).A systematicmammalian genetic interaction map reveals pathways underlying ricinsusceptibility.Cell 152,909-922.
[0233] Bliss,C.(1939).The toxicity of poisons applied jointly.Annals ofapplied biology 26,585-615.
[0234] Borisy,A.A.,Elliott,P.J.,Hurst,N.W.,Lee,M.S.,Lehar,J.,Price,E.R.,Serbedzija,G.,Zimmermann,G.R.,Foley,M.A.,Stockwell,B.R.,et al.(2003).Systematic discovery of multicomponent therapeutics.Proc Natl Acad Sci U S A100,7977-7982.
[0235] Bowtell,D.D.(2010).The genesis and evolution of high-grade serousovarian cancer.Nat Rev Cancer 10,803-808.
[0236] Bozic,I.,Reiter,J.G.,Allen,B.,Antal,T.,Chatterjee,K.,Shah,P.,Moon,Y.S.,Yaqubie,A.,Kelly,N.,Le,D.T.,et al.(2013).Evolutionary dynamics of cancerin response to targeted combination therapy.Elife 2,e00747.
[0237] Briner,A.E.,Donohoue,P.D.,Gomaa,A.A.,Selle,K.,Slorach,E.M.,Nye,C.H.,Haurwitz,R.E.,Beisel,C.L.,May,A.P.,and Barrangou,R.(2014).Guide RNAfunctional modules direct Cas9 activity and orthogonality.Mol Cell 56,333-339.
[0238] Chow,R.D.,Wang,G.,Ye,L.,Codina,A.,Kim,H.R.,Shen,L.,Dong,M.B.,Errami,Y.,and Chen,S.(2019).In vivo profiling of metastatic double knockouts throughCRISPR-Cpf1 screens.Nat Methods 16,405-408.
[0239] Cokol,M.,Kuru,N.,Bicak,E.,Larkins-Ford,J.,and Aldridge,B.B.(2017).Efficient measurement and factorization of high-order drug interactions inMycobacterium tuberculosis.Sci Adv 3,e1701881.
[0240] Cokol-Cakmak,M.,Bakan,F.,Cetiner,S.,and Cokol,M.(2018).DiagonalMethod to Measure Synergy Among Any Number of Drugs.J Vis Exp.
[0241] Dang,Y.,Jia,G.,Choi,J.,Ma,H.,Anaya,E.,Ye,C.,Shankar,P.,and Wu,H.(2015).Optimizing sgRNA structure to improve CRISPR-Cas9 knockoutefficiency.Genome Biol 16,280.
[0242] Doench,J.G.(2018).Am I ready for CRISPR?A user's guide to geneticscreens.Nat Rev Genet 19,67-80.
[0243] 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.NatBiotechnol 34,184-191.
[0244] Du,D.,Roguev,A.,Gordon,D.E.,Chen,M.,Chen,S.H.,Shales,M.,Shen,J.P.,Ideker,T.,Mali,P.,Qi,L.S.,et al.(2017).Genetic interaction mapping inmammalian cells using CRISPR interference.Nat Methods 14,577-580.
[0245] Engler,C.,Kandzia,R.,and Marillonnet,S.(2008).A one pot,one step,precision cloning method with high throughput capability.PLoS One 3,e3647.
[0246] Gholami,A.M.,Hahne,H.,Wu,Z.,Auer,F.J.,Meng,C.,Wilhelm,M.,and Kuster,B.(2013).Global proteome analysis of the NCI-60 cell line panel.Cell Rep 4,609-620.
[0247] Gibson,D.G.,Young,L.,Chuang,R.Y.,Venter,J.C.,Hutchison,C.A.,3rd,andSmith,H.O.(2009).Enzymatic assembly of DNA molecules up to several hundredkilobases.Nat Methods6,343-345.
[0248] Grevet,J.D.,Lan,X.,Hamagami,N.,Edwards,C.R.,Sankaranarayanan,L.,Ji,X.,Bhardwaj,S.K.,Face,C.J.,Posocco,D.F.,Abdulmalik,O.,et al.(2018).Domain-focused CRISPR screen identifies HRI as a fetal hemoglobin regulator in humanerythroid cells.Science 361,285-290.
[0249] Han,K.,Jeng,E.E.,Hess,G.T.,Morgens,D.W.,Li,A.,and Bassik,M.C.(2017).Synergistic drug combinations for cancer identified in a CRISPR screen forpairwise genetic interactions.Nat Biotechnol 35,463-474.
[0250] Honma,K.,Iwao-Koizumi,K.,Takeshita,F.,Yamamoto,Y.,Yoshida,T.,Nishio,K.,Nagahara,S.,Kato,K.,and Ochiya,T.(2008).RPN2 gene confers docetaxelresistance in breast cancer.Nat Med 14,939-948.
[0251] Ma,H.,Wu,Y.,Dang,Y.,Choi,J.G.,Zhang,J.,and Wu,H.(2014).Pol IIIPromoters to Express Small RNAs:Delineation of Transcription Initiation.MolTher Nucleic Acids 3,e161.
[0252] Munoz,D.M.,Cassiani,P.J.,Li,L.,Billy,E.,Korn,J.M.,Jones,M.D.,Golji,J.,Ruddy,D.A.,Yu,K.,McAllister,G.,et al.(2016).CRISPR Screens Provide aComprehensive Assessment of Cancer Vulnerabilities but Generate False-Positive Hits for Highly Amplified Genomic Regions.Cancer Discov 6,900-913.
[0253] Najm,F.J.,Strand,C.,Donovan,K.F.,Hegde,M.,Sanson,K.R.,Vaimberg,E.W.,Sullender,M.E.,Hartenian,E.,Kalani,Z.,Fusi,N.,et al.(2018).OrthologousCRISPR-Cas9 enzymes for combinatorial genetic screens.Nat Biotechnol 36,179-189.
[0254] Qi,L.S.,Larson,M.H.,Gilbert,L.A.,Doudna,J.A.,Weissman,J.S.,Arkin,A.P.,and Lim,W.A.(2013).Repurposing CRISPR as an RNA-guided platform forsequence-specific control of gene expression.Cell 152,1173-1183.
[0255] Reinhold,W.C.,Sunshine,M.,Liu,H.,Varma,S.,Kohn,K.W.,Morris,J.,Doroshow,J.,and Pommier,Y.(2012).CellMiner:a web-based suite of genomic andpharmacologic tools to explore transcript and drug patterns in the NCI-60cell line set.Cancer Res 72,3499-3511.
[0256] Reis,A.C.,Halper,S.M.,Vezeau,G.E.,Cetnar,D.P.,Hossain,A.,Clauer,P.R.,and Salis,H.M.(2019).Simultaneous repression of multiple bacterial genesusing nonrepetitive extra-long sgRNA arrays.Nat Biotechnol.
[0257] Shalem,O.,Sanjana,N.E.,Hartenian,E.,Shi,X.,Scott,D.A.,Mikkelsen,T.S.,Heckl,D.,Ebert,B.L.,Root,D.E.,Doench,J.G.,et al.(2014).Genome-scale CRISPR-Cas9 knockout screening in human cells.Science 343,84-87.
[0258] Shen,J.P.,Zhao,D.,Sasik,R.,Luebeck,J.,Birmingham,A.,Bojorquez-Gomez,A.,Licon,K.,Klepper,K.,Pekin,D.,Beckett,A.N.,et al.(2017).CombinatorialCRISPR-Cas9 screens for de novo mapping of genetic interactions.Nat Methods14,573-576.
[0259] Shen,M.W.,Arbab,M.,Hsu,J.Y.,Worstell,D.,Culbertson,S.J.,Krabbe,O.,Cassa,C.A.,Liu,D.R.,Gifford,D.K.,and Sherwood,R.I.(2018).Predictable andprecise template-free CRISPR editing of pathogenic variants.Nature 563,646-651.
[0260] Vidigal,J.A.,and Ventura,A.(2015).Rapid and efficient one-stepgeneration of paired gRNA CRISPR-Cas9 libraries.Nat Commun 6,8083.
[0261] Wong,A.S.,Choi,G.C.,Cheng,A.A.,Purcell,O.,and Lu,T.K.(2015).Massivelyparallel high-order combinatorial genetics in human cells.Nat Biotechnol 33,952-961.
[0262] Wong,A.S.,Choi,G.C.,Cui,C.H.,Pregernig,G.,Milani,P.,Adam,M.,Perli,S.D.,Kazer,S.W.,Gaillard,A.,Hermann,M.,et al.(2016a).Multiplexed barcodedCRISPR-Cas9screening enabled by CombiGEM.Proc Natl Acad Sci U S A 113,2544-2549.
[0263] Wong,A.S.,Choi,G.C.,and Lu,T.K.(2016b).Deciphering CombinatorialGenetics.Annu Rev Genet 50,515-538.
[0264] Wong,S.L.,Chan,W.M.,and Chan,H.Y.(2008).Sodium dodecyl sulfate-insoluble oligomers are involved in polyglutamine degeneration.FASEB J 22,3348-3357.
[0265] Wood,K.,Nishida,S.,Sontag,E.D.,and Cluzel,P.(2012).Mechanism-independent method for predicting response to multidrug combinations inbacteria.Proc Natl Acad Sci U S A 109,12254-12259.
[0266] Xu,L.,Zhao,L.,Gao,Y.,Xu,J.,and Han,R.(2017).Empower multiplex celland tissue-specific CRISPR-mediated gene manipulation with self-cleavingribozymes and tRNA.Nucleic Acids Res 45,e28.
[0267] Zimmer,A.,Katzir,I.,Dekel,E.,Mayo,A.E.,and Alon,U.(2016).Predictionof multidimensional drug dose responses based on measurements of drugpairs.Proc Natl Acad Sci U S A 113,10442-10447.
[0268] Coscia,F.,Watters,K.M.,Curtis,M.,Eckert,M.A.,Chiang,C.Y.,Tyanova,S.,Montag,A.,Lastra,R.R.,Lengyel,E.,and Mann,M.(2016).Integrative proteomicprofiling of ovarian cancer cell lines reveals precursor cell associatedproteins and functional status.Nat Commun 7,12645.
[0269] Domcke,S.,Sinha,R.,Levine,D.A.,Sander,C.,and Schultz,N.(2013).Evaluating cell lines as tumour models by comparison of genomic profiles.NatCommun 4,2126.
[0270] Gholami,A.M.,Hahne,H.,Wu,Z.,Auer,F.J.,Meng,C.,Wilhelm,M.,and Kuster,B.(2013).Global proteome analysis of the NCI-60cell line panel.Cell Rep 4,609-620.
[0271] McClue,S.J.,Blake,D.,Clarke,R.,Cowan,A.,Cummings,L.,Fischer,P.M.,MacKenzie,M.,Melville,J.,Stewart,K.,Wang,S.,et al.(2002).In vitro and in vivoantitumor properties of the cyclin dependent kinase inhibitor CYC202(R-roscovitine).Int J Cancer 102,463-468.
[0272] Muvarak,N.E.,Chowdhury,K.,Xia,L.,Robert,C.,Choi,E.Y.,Cai,Y.,Bellani,M.,Zou,Y.,Singh,Z.N.,Duong,V.H.,et al.(2016).Enhancing the Cytotoxic Effectsof PARP Inhibitors with DNA Demethylating Agents-A Potential Therapy forCancer.Cancer Cell 30,637-650.
[0273] Pulliam,N.,Fang,F.,Ozes,A.R.,Tang,J.,Adewuyi,A.,Keer,H.,Lyons,J.,Baylin,S.B.,Matei,D.,Nakshatri,H.,et al.(2018).An Effective Epigenetic-PARPInhibitor Combination Therapy for Breast and Ovarian Cancers Independent ofBRCA Mutations.Clin Cancer Res 24,3163-3175.
[0274] Sabers,C.J.,Martin,M.M.,Brunn,G.J.,Williams,J.M.,Dumont,F.J.,Wiederrecht,G.,and Abraham,R.T.(1995).Isolation of a protein target of theFKBP12-rapamycin complex in mammalian cells.J Biol Chem 270,815-822.
[0275] Shalem,O.,Sanjana,N.E.,Hartenian,E.,Shi,X.,Scott,D.A.,Mikkelsen,T.S.,Heckl,D.,Ebert,B.L.,Root,D.E.,Doench,J.G.,et al.(2014).Genome-scale CRISPR-Cas9 knockout screening in human cells.Science 343,84-87.
[0276] Wishart,D.S.,Knox,C.,Guo,A.C.,Shrivastava,S.,Hassanali,M.,Stothard,P.,Chang,Z.,and Woolsey,J.(2006).DrugBank:a comprehensive resource for insilico drug discovery and exploration.Nucleic Acids Res 34,D668-672.
[0277] Wong,A.S.,Choi,G.C.,Cheng,A.A.,Purcell,O.,and Lu,T.K.(2015).Massivelyparallel high-order combinatorial genetics in human cells.Nat Biotechnol 33,952-961.
[0278] Wong,A.S.,Choi,G.C.,Cui,C.H.,Pregernig,G.,Milani,P.,Adam,M.,Perli,S.D.,Kazer,S.W.,Gaillard,A.,Hermann,M.,et al.(2016).Multiplexed barcodedCRISPR-Cas9 screening enabled by CombiGEM.Proc Natl Acad Sci U S A 113,2544-2549.
[0279] Yang,L.,Zhang,Y.,Shan,W.,Hu,Z.,Yuan,J.,Pi,J.,Wang,Y.,Fan,L.,Tang,Z.,Li,C.,et al.(2017).Repression of BET activity sensitizes homologousrecombination-proficient cancers to PARP inhibition.Sci Transl Med 9.
[0280] The above description of specific embodiments will fully reveal the general nature of the present disclosure, so that others can easily modify and / or adapt various applications such as specific embodiments without departing from the general concept of the present disclosure by applying the knowledge in the technology of the relevant field (including the content of the documents mentioned and incorporated herein by reference) without excessive experimentation. Therefore, based on the teachings and guidance presented herein, such modifications and modifications are intended to fall within the meaning and scope of the equivalents of the disclosed embodiments. It should be understood that the wording or terminology herein is for the purpose of description rather than limitation, so that the terms or wording of this specification will be interpreted by those skilled in the art according to the teachings and guidance proposed herein in combination with the knowledge of those skilled in the relevant field.
[0281] Although various embodiments of the present disclosure have been described above, it should be understood that they are presented by way of example rather than limitation. It will be apparent to those skilled in the relevant art that various changes may be made to the form and details thereof without departing from the spirit and scope of the present disclosure. Therefore, the present disclosure should not be limited to any of the above exemplary embodiments, but should only be limited according to the appended claims and their equivalents.
[0282] All references cited herein are incorporated by reference in their entirety and for all purposes to the same extent as if each individual publication or patent or patent application was specifically and individually indicated to be incorporated by reference in its entirety for all purposes.
Claims
1. A system for CRISPR-based multi-gene knockout screening, the system comprising a barcoded gRNA expression cassette, the barcoded gRNA expression cassette comprising: (i) a first promoter operably linked to a first gRNA and a first barcoded sequencing region; (ii) a second promoter operably linked to a second gRNA and a second barcoded sequencing region; (iii) a third promoter operably linked to a third gRNA and a third barcoded sequencing region; wherein the gRNA expression cassette is in a single vector, the first promoter, the second promoter and the third promoter are human U6, mouse U6 and human H1 promoters, respectively, and the human U6, mouse U6 and human H1 promoters have the same modified 3' terminal sequence, the modified 3' terminal sequence is 5'-CACCG-3'; The first, second and third gRNAs all contain modified gRNA scaffolds; the promoter contains a modified 3' terminal sequence complementary to the sticky end of the gRNA target sequence of the gRNA, and the 3' terminal sequence anneals with the sticky end of the gRNA target sequence of the gRNA; the annealed sequence and the gRNA scaffold are combined, digested and connected to form an assembly of a combined barcoded combination gRNA array.
2. The system of claim 1, wherein the vector is a lentiviral vector.
3. The system of claim 1 or 2, wherein the combinatorial gRNA array is assembled by CombiGEM-CRISPR v2.
0.
4. The system of claim 2, wherein the lentiviral vector transfects human cells and the barcoded gRNA is delivered to the human cells.
5. The system of claim 1, 2 or 4, further comprising quantifying the barcoded gRNA using next generation sequencing at a time point after transfection.
6. The system of claim 1, wherein the triple combinatorial CRISPR screen is a high throughput screen.
7. The system of claim 1, wherein the expression cassette knocks out a target GFP gene in OVACR8-ADR-Cas9 cells.
8. The system of claim 1, wherein the gRNA scaffold comprises: (i) higher on-target activity than a wild-type scaffold; (ii) low off-target activity; and (iii) high on-to-off-target activity.
9. The system of claim 1, wherein the barcoded gRNA expression cassette comprises a v3.11, v.3.12, or v.3.13 scaffold.
10. A method for screening at least a ternary drug target combination, comprising the steps of: (i) providing a lentiviral triple combination gRNA expression construct expressing gRNA; (ii) transfecting human cells with the lentiviral triple combination gRNA expression construct; and (iii) measuring cell growth over a period of time after transfection; Wherein, the lentiviral ternary combination gRNA expression construct comprises a barcoded gRNA expression cassette, and the barcoded gRNA expression cassette comprises: (i) a first promoter operably connected to the first gRNA and a first barcoded sequencing region; (ii) a second promoter operably connected to the second gRNA and a second barcoded sequencing region; (iii) a third promoter operably connected to the third gRNA and a third barcoded sequencing region; wherein the gRNA expression cassette is in a single vector, and the first promoter, the second promoter and the third promoter are human U6, mouse U6 and human H1 promoters, respectively, and the human U6, mouse U6 and human H1 promoters have the same modified 3' terminal sequence, and the modified 3' terminal sequence is 5'-CACCG-3'; The first, second and third gRNAs all include a modified gRNA scaffold; the promoter comprises a modified 3' terminal sequence complementary to the sticky end of the gRNA target sequence of the gRNA, and the 3' terminal sequence anneals with the sticky end of the gRNA target sequence of the gRNA; the annealed sequence and the gRNA scaffold are combined, digested and connected to form an assembly of a combined barcoded combination gRNA array.
11. The method of claim 10, further comprising validating the ternary drug target combination by matching a drug to the drug target.
12. The method of claim 10, wherein the ternary drug target combination provides a three-drug regimen for the disease.
13. The method of claim 10, wherein the at least three-way drug target combination is a synergistic combination.
14. The method of claim 12, wherein the disease is cancer or Parkinson's disease.
