CRISPR Gene Editing in Stem Cells

93 128

repressive complex (PRC2) (Klattenhoff et al., 2013); however,
we did not detect any change in the interaction with PRC2 in
the mutant ESCs by RNA immunoprecipitation (RIP) (data not
shown), suggesting that AGIL is not required for this interaction
and that cooperation between Bvht and PRC2 may be a later
event in regulating CM differentiation.
We next validated the interaction between Bvht and mouse
CNBP, HNRNPF, and SFRS9 by expressing mouse Flag-tagged
versions of these factors in both WT and bvhtdAGIL ESCs followed
by immunoprecipitation using an anti-Flag antibody (Figure 4D).
We found that all three candidates co-purified with WT Bvht,
but not bvhtdAGIL, as shown by qRT-PCR. Upon analysis of
ProtoArray results available for 20 distinct non-coding RNAs
(Kretz et al., 2013; Marques Howarth et al., 2014; Siprashvili
et al., 2012), CNBP and HNRNPF binding appeared to be highly
specific to Bvht, whereas SFRS9 interacted broadly with other
non-coding RNAs. HNRNPF, a member of ubiquitously ex-
pressed heterogeneous nuclear ribonucleoproteins family, is
an RNA-binding protein with roles in mRNA splicing and mRNA
metabolism and transport, and can bind G-rich sequences (Ma-
tunis et al., 1994; Reznik et al., 2014; Wang et al., 2012). CNBP
(ZNF9) is a zinc-finger TF containing seven CCHC-type zinc fin-
gers and one RNA recognition motif (RGG) (Figure 5A) that also
binds G-rich single-stranded DNA and RNA (Armas et al.,
2008; Calcaterra et al., 2010). CNBP has roles in neural crest
cell expansion, and null mice die around embryonic day 10.5
(E10.5) (Chen et al., 2003; Weiner et al., 2007, 2011); however,
its overall function is poorly characterized. Notably, CNBP is
highly expressed in heart and skeletal muscle, and heterozygous
cnbp+/ mice exhibit severe dilated cardiomyopathy (Chen et al.,
2007). Moreover, CNBP is currently the only known gene linked to
myotonic dystrophy type 2 in human, and patients often display
severe heart defects (Jones et al., 2011; Lee et al., 2012; Liquori
et al., 2001). Thus, given its binding preference for single-
stranded G-rich nucleic acids and its understudied roles in the
heart, we focused on further characterization of CNBP.
CNBP Represses CM Differentiation
To test the function of CNBP in our system, we introduced small
indels using CRISPR/Cas9 genome editing in both WT and
bvhtdAGIL ESCs, generating cnbpKO and cnbpKO;bvht dAGIL
ESCs (Figures 5A and S5A). Clones were sequenced for the
presence of the mutations and immunoblot confirmed loss of
CNBP in both cnbpKO and cnbpKO;bvhtdAGIL ESCs (Figure 5B).
Importantly, neither disruption of the Bvht AGIL motif nor cnbpKO
affected the expression of either CNBP or Bvht, respectively
(Figures 5B and S5B). Moreover, loss of CNBP did not affect
the expression of ESC pluripotency markers Oct4 and Nanog,
similar to bvhtdAGIL (Figure S5B).
We next tested two independent cnbpKO ESC clones for their
ability to differentiate into CMs. As shown in Figure S5C, cnbpKO
cells show similar morphologies to WT cells at both day 2 and
day 4 of differentiation and are fully capable of differentiating
into CPs at day 5.3 and CMs at day 10, as shown by immunoflu-
orescence analysis of Nkx2.5-GFP and cTnT, respectively. In
fact, cnbpKO cells generate significantly higher percentages of
Nkx2.5-GPF+ cells (CP) at day 5.3 and cTnT+ cells (CM) at day
10 by FACS when compared to WT cells (Figure 5C). Moreover,
qRT-PCR analysis showed that cardiac TFs (e.g., Nkx2.5, Gata4,
Gata6, Hand2, and Tbx5) at day 5.3 and CM marker genes (e.g.,
cTnT, Myh6, andMyh7) at day 10 exhibit higher expression levels
in cnbpKO cells compared to WT cells (Figures 5D and 5E).
To further test CNBP function, we constitutively overex-
pressed Flag-tagged CNBP in WT ESCs, which did not affect
the expression levels of Bvht and ESC pluripotency markers
Oct4 and Nanog (Figure S5D). In contrast to cnbpKO ESCs, cells
expressing higher levels of CNBP produced significantly lower
percentages of Nkx2.5-GPF+ cells (CP) at day 5.3 and cTnT+
cells (CM) at day 10 compared to control cells by FACS (Figures
5G and S5E). Consistent with these data, cardiac TFs and CM
marker genes showed decreased expression levels by qRT-
PCR upon CNBP overexpression (Figures 5H and 5I). Together,
our data suggest that CNBP functions, in part, as a negative
regulator of cardiovascular lineage commitment.
Loss of CNBP Partially Rescues the bvhtdAGIL Phenotype
Based on the above results, we hypothesized that Bvht may
functionally antagonize CNBP to promote cardiovascular lineage
commitment, predicting that loss of CNBP would rescue the
bvhtdAGIL mutant phenotype. To test this idea, we first performed
EB differentiation of cnbpKO;bvht dAGIL ESCs compared to WT
ESCs. At day 12 of EB differentiation, the expression levels
of CM marker genes including cTnT, Myh6, and Myh7 were
significantly restored in the cnbp KO;bvhtdAGIL double mutant
cells compared to bvhtdAGIL single mutant (Figure S6A). We
then performed the CM differentiation assay and found that the
cnbpKO;bvht dAGIL double mutants produced significantly
increased percentages of CP and CM cells compared to the
bvhtdAGIL mutant alone (Figure 6A). Nkx2.5 is expressed
throughout the CP-to-CM stages (Ma et al., 2008; Wamstad
Figure 5. CNBP Represses CM Differentiation
(A) Diagram of CNBP (Uniprot: P53996-2) functional domains, including seven CCHC zinc fingers (aa 4–21, 45–62, 65–82, 89–106, 110–127, 128–145, and
149–166) and RGG box of RNA binding (aa 22–35). The target sequence of CNBP_sgRNA-1 is labeled on the bottom.
(B) Immunoblot analysis with anti-CNBP antibody showing the protein levels of CNBP in indicated ESC lines. GAPDH was used as loading control.
(C) Cells at indicated time points were analyzed for marker expression by flow cytometry. Numbers in plots indicate percentage of gated populations.
(D and E) qRT-PCR analysis showing the relative levels of cardiac marker genes at day 5.3 (D) and day 10 (E) of CM differentiation.
(F) Immunoblot analysis with anti-CNBP antibody showing the protein levels of endogenous CNBP and recombinant CNBP-FLAG in ESCs. GAPDH was used as
loading control.
(G) Cells at indicated time points were analyzed for marker expression by flow cytometry. Numbers in plots indicate percentage of gated populations.
(H and I) qRT-PCR analysis showing the relative levels of cardiac marker genes at day 5.3 (H) and day 10 (I) of CM differentiation.
All experiments were performed in triplicate and data are represented as mean values ± SD. *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001 (two-tailed Student’s
t test).
See also Figure S5.
Molecular Cell 64 , 37–50, October 6, 2016 45