Paired guide RNA CRISPR-Cas9 screening for protein-coding genes and lncRNAs involved in transdifferentiation of human B-cells to macrophages

Abstract

CRISPR-Cas9 screening libraries have arisen as a powerful tool to identify protein-coding (pc) and non-coding genes playing a role along different processes. In particular, the usage of a nuclease active Cas9 coupled to a single gRNA has proven to efficiently impair the expression of pc-genes by generating deleterious frameshifts. Here, we first demonstrate that targeting the same gene simultaneously with two guide RNAs (paired guide RNAs, pgRNAs) synergistically enhances the capacity of the CRISPR-Cas9 system to knock out pc-genes. We next design a library to target, in parallel, pc-genes and lncRNAs known to change expression during the transdifferentiation from pre-B cells to macrophages. We show that this system is able to identify known players in this process, and also predicts 26 potential novel ones, of which we select four (two pc-genes and two lncRNAs) for deeper characterization. Our results suggest that in the case of the candidate lncRNAs, their impact in transdifferentiation may be actually mediated by enhancer regions at the targeted loci, rather than by the lncRNA transcripts themselves. The CRISPR-Cas9 coupled to a pgRNAs system is, therefore, a suitable tool to simultaneously target pc-genes and lncRNAs for genomic perturbation assays.

Keywords: CRISPR-Cas9 screening; Cellular transdifferentiation; Genome editing; Human B-cells; Long non-coding RNA (lncRNA); Macrophages; Paired guide RNA (pgRNA).

Fig. 1

Cellular model and targets selection. A Transdifferentiation of BLaER1 pre-B cells into macrophages is accompanied by a dynamic transcriptomic remodeling of the cells. BLaER1 lymphocytes transdifferentiate into functional macrophages in the presence of Interleukin 3 (IL-3) and Macrophage colony-stimulating factor (M-CSF) upon β-estradiol induced release of CEBPaER to the nucleus. B Flow cytometry analysis of cell surface markers at T0, T3 (3 days) and T6 (6 days) after induced transdifferentiation in the BLaER1-Cas9 cell line. During the process, BLaER1 cells progressively lose the CD19 (B-cell marker staining -X-axis-) and gain the Mac1 (macrophage marker staining -Y-axis-). C Merged k-means clustered expression profiles (color code) of peaking and upregulated genes during transdifferentiation: 16 initial clusters of lncRNA (n = 174) and 36 initial clusters of protein coding genes (n = 939). FPKM values were log10 transformed before the normalization to z-score. Each line shows the expression pattern of a gene along transdifferentiation. The color corresponds to the k-means cluster to which the gene belongs (see also Supplementary Fig. S1 and S2)

Fig. 2

pgRNA CRISPR library for lncRNA and pc-genes. A (Upper panel) Diagram of the CD19 gene indicating the target sequence of CD19 pgRNAs (sgRNA1 and sgRNA2, from left to right). (Lower panel) Flow cytometry analysis of fluorescence intensity of the CD19 protein in BLaER1-Cas9 cells infected with sgRNAs and pgRNAs. The relative Stain Index of the different infected cells compared to the maximum expression level of CD19 in control cells (BLaER1-Cas9 cells infected with pDECKO-GFP [11]) is represented. CD19 expression is reduced between 30 and 95% upon infection of sgRNAs. The infection of pgRNAs induces a consistent reduction of CD19 signal up to 95% with all pgRNAs tested. B Schematic diagram showing the position of pgRNAs targeting lncRNAs (targeting the promoter and the transcription start site) and pc-genes (targeting coding exons). C CRISPR library composition (number of targets of each biotype and pgRNA pairs designed per target)

Fig. 3

CRISPR-Cas9 screening in BLaER cells. A Workflow of the CRISPR screening experiment. The pDECKO plasmid library was transfected into HeK293T cells to obtain a library of lentivirus. BLaER1-Cas9 cells were infected at a low multiplicity of infection and double selected with antibiotics (Blasticidin and Puromycin) for 20 days. The infected cells were induced for transdifferentiation into macrophages for 3 days (T3) and 6 days (T6). Cells were labeled with antibodies against cell surface markers: CD19 (for B-lymphocytes) and Mac1 (for macrophages). Transdifferentiation status was assessed by flow cytometry. Transdifferentiated and delayed populations were isolated by Fluorescence-Activated Cell Sorting (FACS). B Flow cytometry analysis of BLaER1-Cas9 cells infected with the pDECKO_non-targeting control (left panels) and with the pDECKO_CRISPR-library (right panels) at T0, T3 and T6 of transdifferentiation. CD19 antibody, conjugated with BV510 fluorophore, was used to identify B-cells and Mac1 antibody, conjugated with PE-Cy7 fluorophore, was used to identify macrophages. Quadrants are as follows: Q1 (macrophage-like cells with presence of Mac1 and absence of CD19 surface markers); Q2 (transition cells with the presence of Mac1 and CD19); Q3 (background and not stained cells, negative for Mac1 and CD19); Q4 (lymphocyte B-like cells with the presence of CD19 and absence of Mac1 surface markers). The percentage of cells in each of the 4 quadrants is shown. The fraction of sorted cells showing a delay of transdifferentiation (“delayed” fraction) is marked in blue (gate P4), and sorted cells that differentiate at a normal pace (“differentiated” fraction) are marked in orange (gate P5). See also Supplementary Fig. S5. C Workflow for processing the sorted cell populations for deep sequencing. Genomic DNA of sorted cells was extracted and PCR amplified in two steps. For the first PCR, specific staggered primers were used to amplify the integrated fragment which contains the pgRNAs. For the second PCR, Illumina barcoded primers were used to pool different samples (see also Supplementary Fig. S4). Samples were sequenced by 150 bp paired-end Illumina sequencing. DDE (differentiation delayed effect) was calculated as the ratio of pgRNA counts in the delayed population versus the counts in the transdifferentiated population

Fig. 4

Identification of lncRNAs and protein coding genes involved in transdifferentiation. A Correlation between replicates of the differentiation delaying effect (DDE, ratio of reads from delayed versus transdifferentiated fraction) observed per pgRNA of ratCEBPa (left panel) and intergenic negative controls (right panel) after 6 days (T6) of transdifferentiation. Each dot represents a different pgRNA. Spearman correlation values are stated above. The DDE values of CEBPa pgRNAs are very large and show a positive correlation between replicates, whereas intergenic pgRNAs do not show reproducible DDE values between replicates. B Scatterplot of log10 transformed counts in delayed versus differentiated fractions at T3 and T6 after induction of transdifferentiation. Each dot represents a different pgRNA. pgRNAs targeting positive controls are depicted in blue, intergenic pgRNAs in red, screened candidates in black, and pgRNAs of selected candidates showing high average DDE score in green (merged counts of both replicates). C Decision tree followed to identify candidate genes, from the CRISPR-Cas9 screening, involved in the transdifferentiation process. From the original list of 1,040 pc-genes and lncRNAs, we ended up with a set of seven candidates to undergo further validation

Fig. 5

Individual target validation by flow cytometry. Flow cytometry analysis of control and candidate pgRNAs at T0, T3 and T6 after induction of transdifferentiation. A Flow cytometry plots of intergenic negative control, two positive controls targeting ratCEBPa and SPI1, and two protein coding targets FURIN and NFE2. CD19 B-cell marker is represented on the X-axis and Mac1 macrophage marker is represented on the Y-axis. Cells that do not undergo transdifferentiation remain in the Q4 quadrant (positive for CD19 -X-axis- and negative for Mac1 -Y-axis) (the percentages of cells in this quadrant are shown). B Percentage of cells with delayed transdifferentiation (Q4 quadrant) observed in controls and individually validated candidates for two biological replicates (R1 and R2) at T3 and T6 after induction of transdifferentiation. For lncRNAs LINC02432 and MIR3945HG we only have data for one biological replicate at T6. Average (Avg) and standard deviation (SD) between replicates, and two-tailed p-values (comparing the delayed population from each individual target and the intergenic negative control) are also shown. Most of the selected candidates show significant delay compared to the intergenic negative control at T3. Although the value observed for FURIN is not statistically significant, the magnitude of the delay indicates that it is a strong candidate to perform further validations

Fig. 6

FURIN and NFE2 expression after CRISPR edition. A FURIN RNA and protein expression. Cells were collected at T0 (before induction) and T3 (3 days after transdifferentiation induction). (CT0) and (CT3) negative control pDECKO-Intergenic at T0 and T3 respectively, (FUT0) and (FUT3) pDECKO-FURIN at T0 and T3, (FUT3s) pDECKO-FURIN at T3 and sorted from gate P4 (delayed population). Upper panel, qRT-PCR to check the expression of FURIN using two different sets of primers. Results are normalized to GAPDH and the fold change is calculated relative to the expression of cells infected with pDECKO-intergenic pgRNA at T3. The expression of FURIN decreases in cells infected with FURIN pgRNAs, especially in the delayed subpopulation (FUT3s). Bottom panel, western blot to assess the levels of the FURIN protein in BLaER1-Cas9 infected cells. Anti-FURIN antibodies recognize a band (marked with an arrowhead), the signal of which increases at T3, in line with RNA-Seq data (Supplementary Table S4). The FURIN band is not detectable in the pDECKO-FURIN infected cells (FUT3 and FUT3s). Uncropped blots are shown in Supplementary Fig. S16A. B NFE2 RNA and protein expression. (CT0) and (CT2) negative control pDECKO-Intergenic at T0 (before induction) and T2 (2 days after transdifferentiation induction) respectively, (NFT0) and (NFT2) pDECKO-NFE2 at T0 and T2, (NFT2s) pDECKO-NFE2 at T2 and sorted from gate P4 (delayed population). Upper panel, qRT-PCR to check the expression of NFE2 using 2 different sets of primers. Results are normalized to GAPDH and the fold change is calculated relative to the expression of cells infected with pDECKO-intergenic T2. NFE2 expression in NFE2 pgRNA targeted cells is higher than in intergenic control cells (NFT2 and NFT2s compared to CT2). Bottom panel, western blot to check the protein levels of NFE2 in BLaER1-Cas9 infected cells. Anti-NFE2 antibodies detect two bands, the signal of which increases at T2 (CT2 compared to CT0). These two bands are strongly reduced in NFE2 targeted populations (NFT2 and NFT2s compared to CT2). Uncropped blots are shown in Supplementary Fig. S16B