Highly efficient gene inactivation by adenoviral CRISPR/Cas9 in human primary cells

Abstract

Phenotypic assays using human primary cells are highly valuable tools for target discovery and validation in drug discovery. Expression knockdown (KD) of such targets in these assays allows the investigation of their role in models of disease processes. Therefore, efficient and fast modes of protein KD in phenotypic assays are required. The CRISPR/Cas9 system has been shown to be a versatile and efficient means of gene inactivation in immortalized cell lines. Here we describe the use of adenoviral (AdV) CRISPR/Cas9 vectors for efficient gene inactivation in two human primary cell types, normal human lung fibroblasts and human bronchial epithelial cells. The effects of gene inactivation were studied in the TGF-β-induced fibroblast to myofibroblast transition assay (FMT) and the epithelial to mesenchymal transition assay (EMT), which are SMAD3 dependent and reflect pathogenic mechanisms observed in fibrosis. Co-transduction (co-TD) of AdV Cas9 with SMAD3-targeting guide RNAs (gRNAs) resulted in fast and efficient genome editing judged by insertion/deletion (indel) formation, as well as significant reduction of SMAD3 protein expression and nuclear translocation. This led to phenotypic changes downstream of SMAD3 inhibition, including substantially decreased alpha smooth muscle actin and fibronectin 1 expression, which are markers for FMT and EMT, respectively. A direct comparison between co-TD of separate Cas9 and gRNA AdV, versus TD with a single "all-in-one" Cas9/gRNA AdV, revealed that both methods achieve similar levels of indel formation. These data demonstrate that AdV CRISPR/Cas9 is a useful and efficient tool for protein KD in human primary cell phenotypic assays. The use of AdV CRISPR/Cas9 may offer significant advantages over the current existing tools and should enhance target discovery and validation opportunities.

Fig 1. AdV CRISPR/Cas9 constructs and gRNA targeting sequences.

(A) Overview of the AdV Cas9 and gRNA constructs used in this study. CRISPR/Cas9 AdV constructs were based on human codon optimized S. pyogenes wild type Cas9 and previously published gRNA sequences [8]. Cas9 and gRNA components were either presented in separate AdVs or together in a single AdV construct (“all-in-one”). (B) Genomic structure of the human SMAD3 gene with the indicated annotated transcripts. One of the gRNA sequences targeting the SMAD3 genomic DNA is underlined with the PAM in bold. NLS: nuclear localization signal.

Fig 2. Titration of Cas9 and gRNA AdV effects on SMAD3 indel formation in NHLFs.

(A) Primary NHLFs were co-transduced at DIV 1 with varying ratios of Cas9 and gRNA AdVs at a maximal total MOI of 80. NHLFs were harvested at DIV 9 after which genomic DNA was used in PCR amplification of the SMAD3 target region. Resulting PCR products were used for SURVEYOR^® assay analysis and resolved by agarose gel electrophoresis. A single full-length band at 933 bp (highlighted with white arrowheads) indicates the uncleaved PCR product, whereas appearance of two additional bands following treatment with SURVEYOR^® nuclease indicates the presence of indels (highlighted with black arrowheads). (B) Quantification of indel frequencies as determined by densitometry of the DNA fragments depicted in panel A. (C) NHLFs were treated as in A, except that TD was performed with a single all-in-one AdV particle. (D) Quantification of indel frequencies as determined by densitometry of the DNA fragments depicted in panel C. 39: gRNA targeting SMAD3; C1: gRNA targeting control sequences.

Fig 3. Titration of Cas9 and gRNA AdV effects on SMAD3 indel formation in HBECs.

(A) Primary HBECs were co-transduced at DIV 1 with varying ratios of Cas9 and gRNA AdV at a maximal total MOI of 40. HBECs were harvested at DIV 9 after which genomic DNA was analyzed for the presence of indels by SURVEYOR^® assay as described earlier. (B) Quantification of indel frequencies as determined by densitometry of the DNA fragments depicted in panel A. (C) HBECs were treated as in A, except that TD was performed with a single all-in-one AdV particle. (D) Quantification of indel frequencies as determined by densitometry of the DNA fragments depicted in panel C. White arrowheads indicate the positions of parental PCR fragments whereas black arrowheads mark the SURVEYOR^® nuclease digested fragments containing indels. 39: gRNA targeting SMAD3; C1: gRNA targeting control sequences.

Fig 4. CRISPR/Cas9-mediated depletion of SMAD3 protein in primary NHLFs and HBECs.

(A) SMAD3 genomic editing in NHLFs and HBECs following co-TD with Cas9 and SMAD3_v39 gRNA AdV. Primary NHLFs (left) and HBECs (right) were co-transduced with Cas9 and either SMAD3 or control gRNA AdV at total MOI 30 (Cas9:gRNA ratio 1:2) after which genomic DNA was analyzed for the presence of indels at DIV 12 by SURVEYOR^® assay as described earlier. Indel frequencies are shown below each lane and were determined by densitometry of the full-length and cleaved PCR fragments. 39: gRNA targeting SMAD3; C1: gRNA targeting control sequence; SA Ctrl: SURVEYOR® assay control. (B) AdV CRISPR/Cas9-mediated SMAD3 protein KD in NHLFs and HBECs. Cells were harvested at DIV 12 for protein analysis by immunoblotting. A total of 35 μg or 28 μg protein per lane was loaded for the NHLFs and HBECs, respectively. Immunoblots were stained with the indicated primary antibodies. (C) Quantification of SMAD3 protein levels in NHLFs and HBECs treated with Cas9 and SMAD3-targeting AdV. SMAD3 protein expression was quantified by densitometry of the bands shown in panel B. Expression is blotted relative to the control (co-TD with Cas9 and control gRNA C1). Data points are from biological duplicates (n = 2) and error bars represent standard deviation.

Fig 5. Application of CRISPR/Cas9 to human primary cell phenotypic assays.

(A) Overview of the FMT assay with indicated time points of treatment in days. Upregulation of markers specific for myofibroblasts is indicated with upward arrows. At the end of the assay HCA is performed on ACTA2 and SMAD3. (B-D) FMT phenotypic assay with HCA on ACTA2 and SMAD3. Primary NHLFs were co-transduced at DIV 1 with Cas9 and either SMAD3, ACTA2 or control gRNA AdV at total MOI 30 (Cas9:gRNA ratio 1:2), followed by addition of TGF-β1 at DIV 6. Cells were fixed for immunofluorescent labelling of SMAD3 and ACTA2 followed by HCA at DIV 9. Graphs are representing HCA quantification of ACTA2 and SMAD3 immunofluorescent intensities in optimized segmentation masks and are blotted relative to the control (C1). Data points are from triplicate wells (n = 3) and error bars represent standard deviation. Exemplifying images of target protein and nuclear DAPI labelling are shown. (E) Overview of the EMT assay with indicated time points of treatment in days. Change in expression of certain markers is shown, with arrows demonstrating up- or downregulation (upward and downward arrows, respectively). At the end of the assay HCA is performed on FN1 and SMAD3. (F-G) EMT phenotypic assay with HCA on FN1 and SMAD3. Primary HBECs were co-transduced at DIV 1 with Cas9 and either SMAD3 or control gRNA AdV at total MOI 12 (Cas9:gRNA ratio 1:1), followed by addition of a cocktail containing TGF-β1 and TNFα at DIV 6. Cells were fixed and stained as described in panel A, except that staining was performed for FN1 and SMAD3 in this case. Graphs are blotted relative to the control (C1). Scale bars indicate 100 μm. 39, 40, 41: gRNAs targeting SMAD3; 7, 8, 9: gRNAs targeting ACTA2; C1: gRNA targeting control sequence.