CRISPR/Cas9-mediated deletion of the Wiskott-Aldrich syndrome locus causes actin cytoskeleton disorganization in murine erythroleukemia cells

Cell cultures

The murine erythroleukemia cell line MEL-DS19 (hereafter called MEL) was obtained from Dr Arthur Skoultchi (Albert Einstein College of Medicine, Bronx, New York, USA). MEL-R cells, derived from MEL cells, were previously established in our laboratory by growing MEL cells continuously in the presence of 5 mM hexamethylene bisacetamide (HMBA) (Fernandez-Nestosa et al., 2008; Fernandez-Nestosa et al., 2013). Murine 3T3-Swiss albino fibroblasts (CCL-92) were obtained from the ATTC. Cell lines were propagated in Dulbecco’s Modified Eagle’s Medium containing 10% fetal bovine serum (BSA), 100 units/ml penicillin and 100 µg/ml streptomycin (all from Gibco). MEL-R cells were routinely cultured in the presence of 5 mM HMBA (Sigma). MEL DS19 cell differentiation was induced by exposing exponentially growing cultures to 5 mM HMBA. Hemoglobinized cells were quantified by determining the proportion of benzidine-staining positive cells (B+) in cell cultures. Cell growth was assessed daily by counting samples of the cultures with a Neubauer hemocytometer chamber.

Generation of MEL/Was^−∕− cells by CRISPR/Cas9 technology

Genomic deletion of Was in MEL cells was performed by CRISPR/Cas9 technology as described (Bauer, Canver & Orkin, 2015). Two single guide RNAs (sgRNA1 and sgRNA2) were designed to separately target the entire gene (mouse X:7658591–7667617) using the online tool CRISPR (http://crispr.mit.edu/). Coupled complementary oligonucleotides (CACC was added to the 5′ end of the sense strand and AAAC was added to the 5′ end of the antisense strand) were annealed and inserted into the BbsI sites of linearized pX330 vector (Addgene plasmid ID 42230). The sequences of the sgRNA oligonucleotides are listed in Table S1. MEL cells were co-transfected with the two sgRNAs vectors and a third vector, pEFBOS-GFP, encoding green fluorescent protein (GFP) (Fig. S1), by cationic liposome-based transfection with Lipofectamine 2000 (Life Technologies). After 72 h, the top ∼3% of GFP-positive cells were individually sorted into 96-well plates. Genomic DNA was isolated from all clones and screened for biallelic deletion via PCR using non-deletion (ND) primers, whose assembly takes place internal to the sequence to be deleted, and deletion (D) primers, whose assembly is upstream and downstream of the sgRNA cleavage sites. The primers used for identifying biallelic deletion clones are listed in Table S2 and were designed with Primer3 software (http://bioinfo.ut.ee/primer3-0.4.0/) (Untergasser et al., 2012). The conditions for the PCR were as follows: pre-denaturing step of 94 °C for 7 min, followed by 35 cycles of 94 °C for 40 s, 60 °C for 1 min and 72 °C for 1 min, with a final extension at 72 °C for 7 min. PCR products were resolved on 1% agarose gels and visualized by ethidium bromide staining.

MEL-R DNA transfection

Exponentially growing MEL-R cells were transfected as described above with the pcDNA3.1 ± DYK Was expression vector (GeneScript) containing the coding region of Was (hereafter called pcDNA3.1-Was) (Fig. S1). After 6 h, cells were distributed into 96-well plates. The transfectants were selected by limited dilution and maintained in growth medium containing 1 µg/ml G418 (Sigma).

Antibodies and immunoblotting

Control 3T3 fibroblast cells, MEL, MEL-R and transfected cells (2.5 × 10⁶) were harvested, washed with phosphate buffered saline (PBS) and lysed with NP-40 buffer (20 mM Tris-HCl pH 7.5, 10% glycerol, 137 mM NaCl, 1% NP-40, 1 mM sodium orthovanadate, 10 mM sodium fluoride, and 2 mM EDTA) containing protease inhibitors (all from Sigma). Protein lysates (10–30 µg) were separated by 12% SDS-polyacrylamide gel electrophoresis and transferred to PVDF membranes (Bio-Rad). The membranes were incubated with a mouse monoclonal anti-β-actin (1:10,000; Sigma), mouse monoclonal anti-WASp (1:500, Santa Cruz), mouse monoclonal anti-Btk (1:500; Santa Cruz), and rabbit polyclonal anti-α-tubulin (1:1000; ABclonal) antibodies, then washed five times with T-TBS (20 mM Tris–HCl, 150 mM NaCl, 0.1% Tween 20). Primary antibodies were detected with horseradish peroxidase-conjugated anti-mouse (1:3,000; Santa Cruz) or anti-rabbit IgG (1:1,000, DAKO) antibodies, followed by five cycles of T-TBS washes. The analysis of filamentous (F-actin) and globular (G-actin) actin content was performed from 10⁷ cells. Samples were harvested, washed in PBS and lysed in a lysis buffer (50 mM PIPES pH 6.9 (Sigma), 5 mM MgCl₂, 5 mM EGTA (Sigma), 5% glycerol (Roche), 0.1% β-mercaptoethanol (Merck), 1 mM PMSF (Roche), 10 mM benzamidine (Sigma) and 1 mM ATP (Roche)). Protein supernatants and pellets were collected after ultracentrifugation (100,000 g, 1 h at 37 °C) and analyzed by immunoblotting as described above.

Immunofluorescence staining and confocal microscopy

MEL, MEL-R and transfected cells were plated on poly-L-lysine-coated slides and incubated at 37 °C for 30 min. Cells were fixed with 4% paraformaldehyde for 30 min, permeabilized with 0.1% Triton-X 100 in PBS for 30 min and blocked with 1% BSA in PBS/0.1% Triton-X 100 for 1 h, all at room temperature (RT, ∼22 °C). Cells were stained with anti-β-actin (1:3,000; Sigma) or anti-Btk (1:200; Santa Cruz) antibodies for 1 h at RT followed by washing twice with PBS. Primary antibodies were detected with an Alexa Fluor 568 secondary antibody (Molecular Probes) and 1 µg/ml DAPI (4,6-diamidino-2-phenylindole; Sigma) was added to stain nuclei, for 1 h at RT, followed by two washes with PBS. Finally, cells were mounted on cover slips with Prolong Diamond Antifade Mountant reagent (Invitrogen). Fluorescence images were acquired on a Leica TCS SP2 confocal microscope using a 100 × objective with zoom.

Statistical analysis

Data are presented as means ± standard deviation of the densitometric analysis. Differences were tested by the Student t-test. The values P < 0.05 were considered statistically significant. Statistical analysis of western blot data and immunofluorescence images are presented in Figs. S2 and S3, respectively.

Effects of ectopic WASp expression on the actin cytoskeleton in MEL-R cells

It is generally recognized that WASp plays an important role in the maintenance of cytoskeletal organization in hematopoietic cells (Alekhina, Burstein & Billadeau, 2017). In our previous study, we found that Was is expressed in erythroleukemia MEL cells, whereas it was barely detectable in MEL-R cell lines (Fernandez-Calleja et al., 2017). To further evaluate the status of Was at the protein level, we compared proteins extracts from MEL cells, both undifferentiated and differentiated in the presence of the inducer-mediated differentiation HMBA, with those from a representative MEL-R resistant line. We also included the 3T3 fibroblast cell line as a negative control as WASp is not expressed in non-hematopoietic lineages. Immunoblotting showed robust WASp expression in MEL cells, which decreased during cell differentiation (Fig. 1A). By contrast, the levels of WASp in MEL-R cells were much lower than in undifferentiated MEL cells and were comparable with that observed in 3T3 fibroblasts. Because our recent study (Fernandez-Calleja et al., 2017) suggested that low levels of several actin-cytoskeletal proteins, including WASp, result in anomalous cytoskeleton organization, we sought to determine whether forced expression of WASp might reverse this effect. To do this, we established MEL-R cell lines constitutively expressing WASp and screened for clones with robust WASp expression, which identified clones 9, 10 and 11 for further analysis (Fig. 1B).

Confocal immunofluorescence analysis of MEL cells stained with an anti-actin antibody revealed a clear rim of actin fluorescence surrounding the nuclear periphery (Fig. 2, column 1). A similar pattern was observed in the WASp-overexpressing MEL-R clones 9, 10 and 11 (Fig. 2, columns 2–4), whereas no detectable signal was evident in non-Wasp-expressing MEL-R cells (Fig. 2, column 5). These results confirm that WASp expression has a positive impact on the organization of the actin cytoskeleton in erythroleukemia cells.

We next used immunoblotting to evaluate total actin protein levels in MEL-R and WASp-overexpressing MEL-R transfectants, finding no differences between the two (Fig. 3A). Given this result, we next asked whether the changes observed by confocal microscopy might be due to an altered ratio of monomeric G-actin and polymeric F-actin. To address this, we used high-speed centrifugation to separate G- and F-actin pools from cell lysates, followed by immunoblotting with an anti-β actin antibody. Results showed that the F-actin pool in WASp-overexpressing MEL-R clones 9, 10 and 11 was considerably greater than in MEL-R cells, whereas the free G-actin pool remained low (Fig. 3B). By contrast, the G- and F-actin pools in control MEL-R cells were similar. These results demonstrate that the proportion of F-actin increases after WASp overexpression and suggest that it might contribute to nuclear actin polymerization.

CRISPR/Cas9-mediated Was deletion in MEL cells

Proteins associated with the actin polymerization machinery, such as WASp, are silenced or minimally expressed in MEL-R cells (Fernandez-Calleja et al., 2017). We have shown that ectopic expression of WASp in MEL-R cells can restore the wild-type phenotype observed for the actin cytoskeleton (Fig. 2). Assuming that WASp is essential for actin polymerization in erythroleukemia cells, we hypothesized that the knockout of the gene would alter actin cytoskeleton organization in MEL cells. Thus, we used the CRISPR/Cas9 gene-editing platform to delete the genomic region encompassing the Was gene in the X chromosome of MEL cells. A schematic representation of the genomic target sites is shown in Fig. 4A. Cells were transfected with three plasmids: pX330Was 1 and pX330Was 2, expressing the sgRNAs, and pEFBOS-GFP (Fig. S1). Cells were then double selected for GFP-positivity and G418 resistance. PCR analysis of bi-allelic GFP-positive cells (Fig. 4B) and immunoblotting (Fig. 4C) confirmed three knockout clones: 1, 4 and 73. Cell clone-1 was used in most of the following experiments.

We first assessed whether Was knockout (MEL/Was^−∕−) cells had any noticeable phenotypic differences compared with wild-type MEL cells. Cell proliferation analysis showed no significant alterations in growth rate between MEL and MEL/Was^−∕− cells (Fig. S4A ). Cell differentiation was also evaluated after treatment with 5 mM HMBA, a potent inducer of cell differentiation in erythroleukemia cells (Fig. S4B ). Results showed no changes in the percentage of cell differentiation between the two cultures, indicating that CRISPR/Cas9-mediated deletion had no deleterious effects on cell transfectants.

We then questioned whether MEL/Was^−∕− cells presented a defect in the actin cytoskeleton organization. By confocal analysis, loss of Was function in MEL/Was^−∕− cells led to a considerably reduced actin fluorescence signal, which was similar in intensity to that observed in MEL-R cells (Fig. 5 compare columns 2 and 4). Since the ectopic overexpression of Was in MEL-R cells could rescue the wild-type phenotype (Fig. 2), we performed a similar analysis in MEL/Was^−∕− cells. We transiently transfected the pcDNA3.1-Was vector into MEL/Was^−∕− cells and collected cells 48 h later for immunofluorescence analysis. As expected, a rim of actin staining was clearly visible around the nuclei of MEL/Was^−∕− cells overexpressing Was, and was comparable with that observed in MEL cells (Fig. 5, compare columns 3 and 1). Overall, the knockout and rescue experiments confirm the important effect of WASp for the actin cytoskeleton organization in erythroleukemia cells.

To examine whether changes in actin expression or an altered ratio of monomeric G-actin and polymeric F-actin was the origin of the actin phenotype in MEL-Was^−∕− cells, we first measured the expression of actin in MEL-Was^−∕− cells by immunoblotting We found that the total amount of actin was similar to that in MEL and MEL-R cell lines (Fig. 6A). We then used ultracentrifugation to fractionate monomeric G-actin to the supernatant and polymeric F-actin to the pellet from MEL-Was^−∕− whole extracts and immunoblotted these fractions against actin. Results showed that the balance between G- and F-actin shifted towards G-actin, with markedly less F-actin detected in the pellet of the Was mutant (Fig. 6B).

Ectopic expression of WASp promotes Btk activation in MEL-R and MEL/Was^−∕− cells

There is increasing evidence that actin is also present in the cell nucleus, which is referred to as “nuclear actin”, where it participates in transcriptional activation and chromatin remodeling (for a recent review see Misu, Takebayashi & Miyamoto, 2017). Sadhukan and co-workers previously demonstrated a nuclear role for WASp in transcriptional activation of several master genes during Th1 cell differentiation, a role that is independent of actin polymerization (Sadhukhan et al., 2014). In a different study, WASp was reported to be present both in the cytoplasm and the nucleus, and regulated gene transcription in K562 myeloid cells (Looi et al., 2014). Because WASp can physically associate with Bruton’s tyrosine kinase (Btk) (Baba et al., 1999; Sakuma et al., 2012), and since Btk was included among the silenced actin-associated proteins in MEL-R cells (Fernandez-Calleja et al., 2017), we finally asked whether the ectopic expression of Was affects Btk expression. Figure 7 shows the results of immunoblotting of whole-cell lysates from the stable WASp-overexpressing MEL-R transfectants 9, 10 and 11 against an anti-Btk antibody. We observed that Btk was expressed at higher levels in all three transfectants than in MEL-R cells, although it was more evident in clones 10 and 11. We also analyzed MEL-Was^−∕− cells (clones 1 and 73), before and after enforced expression of WASp. In all cases, Btk was expressed at a level similar to that of the MEL progenitor line irrespective of whether Wasp was overexpressed or not (Fig. 7). These results reveal that Was overexpression can mediate Btk activation in MEL-R cells, suggesting that Btk is downstream of Was in the signaling cascade. By contrast, an alternative signaling pathway might be operative in MEL cells with targeted deletion of Was. Indeed, Btk expression can be modulated by different upstream activators, such as PU.1 (Christie et al., 2015; Himmelmann et al., 1996). We confirmed these results by confocal microscopy, observing a marked increase in the levels of Btk antibody staining in all three MEL-R clones overexpressing Was, and also Btk positivity in wild-type MEL and MEL/Was^−∕− cells (Fig. 8).