CRISPR Gene Editing in Stem Cells

95 128

et al., 2012). Markedly, cnbpKO;bvhtdRHT mutant cells generate
percentages of Nkx2.5-GFP+ cells at day 10 (CM) comparable
to WT cells. Coinciding with our FACS analysis, the cnbpKO;
bvhtdAGIL cells also showed significant levels of Nkx2.5 and
cTnT by immunofluorescence staining (Figures 6B and 6E),
whereas levels were undetectable in bvhtdAGIL mutant alone
(Figure 3C).
We then analyzed the expression of the mesodermal marker
Brachyury and key cardiac TFs by qRT-PCR (Figures 6C and
6D). Brachyury levels in cnbpKO;bvhtdAGIL double mutants
showed comparable expression levels to WT cells. Expression
of Nkx2.5, Gata4, Gata6, Hand2, and Mef2c was also partially
restored at both day 5.3 and day 10. Moreover, we observed
that the CM-specific genes cTnT, Myh6, and Myh7 showed a
significant increase in expression (50% 70% relative to WT
cells) in cnbpKO;bvhtdAGIL double mutants compared to the
AGIL mutant alone (Figure 6E). Together, our data suggest that
CNBP and Bvhtfunction together to regulate cardiovascular line-
age commitment.
DISCUSSION
Our work establishes that RNA secondary structure determina-
tion coupled with genetic studies can reveal important functional
motifs required for lncRNA mechanisms of action. Our study
revealed several important findings regarding the role of Bvht
in cardiovascular lineage commitment. First, we show that
Bvht adopts a modular secondary structure in vitro that harbors
a 50 AGIL. Remarkably, a small 11 nt deletion in the AGIL motif
(bvhtdAGIL) within the 590 nt non-coding transcript prevents
the transition from nascent mesoderm to the CP state in our
in vitro differentiation assay. Second, we found that the zinc-
finger TF CNBP specifically interacts with Bvht. We also show
that CNBP acts as a negative regulator of the cardiac develop-
mental program and that genetic ablation of CNBP partially
rescues the differentiation defect of bvhtdAGIL mutant cells.
Collectively, these data suggest that Bvht functionally antago-
nizes CNBP to promote cardiovascular lineage commitment
(Figure 6F).
In some cases, lncRNAs such as GAS5, PANDA, NF-YA, and
NORAD have been reported to function as molecular decoys
to titrate interacting proteins away from their regulatory targets
through competitive binding (Hung et al., 2011; Kino et al.,
2010; Lee et al., 2016). However, the low abundance of Bvht
transcript makes the molecular decoy model unlikely to explain
its mode of action. Expression of Bvht from the ROSA26 locus
using its endogenous promoter largely rescues the AGIL mutant
phenotype, suggesting that low copy number is sufficient to
mediate its function in trans in a locus-specific manner (Figures
S3G–S3K). Recently, lncRNAs including Fendrr, PRNA , and
PARTICL were found to target specific genomic loci through
directly hybridizing to nascent DNA via sequence complemen-
tarity or DNA:DNA:RNA (Grote et al., 2013; O’Leary et al.,
2015; Schmitz et al., 2010). In addition, it has been proposed
that low-abundance RNAs such as the RNA component of telo-
merase (TERC), which can perform multiple turnover reactions,
could accomplish super-stoichometric functionalities (Goff and
Rinn, 2015; Mozdy and Cech, 2006; Zappulla and Cech, 2004),
providing another potential model for studying the molecular
mechanisms of low-abundance lncRNAs such as Bvht in future
studies.
Our results suggest CNBP is a critical component of Bvht’s
mode of action in cardiovascular lineage commitment. CNBP
is highly conserved among vertebrates and can bind single-
stranded G-rich DNA or RNA (Calcaterra et al., 2010). It has
been proposed that CNBP acts as a nucleic acid chaperone
and can promote the formation of G-quadruplex (G4) structures
in which four guanines are assembled in a planar arrangement by
Hoogsteen hydrogen bonding followed by intra- or inter-molec-
ular folding of the tetramers (Armas et al., 2008; Borgognone
et al., 2010; Rhodes and Lipps, 2015). For example, CNBP re-
presses the expression of heterogeneous ribonucleoprotein K
(hnRNPK) in fibrosarcoma cells and c-Myc in human HeLa cells
through its conversion of promoter G-rich sequences into G4
DNA (Chen et al., 2013; Qiu et al., 2014). We found that different
algorithms including QGRS Mapper, QGRS-H Predictor, and
TetraplexFinder all predict that the Bvht AGIL motif can form a
G4 structure (Figure S6B) (Kikin et al., 2006; Menendez et al.,
2012; Yadav et al., 2008). Notably, G4motifs have been identified
in the promoters or UTRs of cardiac genes such Nkx2.5, Gata4,
and Mef2d (Nie et al., 2015; Zhang et al., 2008). Moreover, the
specific inactivation of the G4-resolving RNA helicase RHAU in
either cardiac mesoderm or progenitors leads to abnormal heart
development (Nie et al., 2015). Thus, it is possible that Bvht and
CNBP function together to regulate cardiac gene expression
through control of G4 structures. Our probing studies indicate
that the stems that flank the AGIL motif may be important for
maintaining the G-rich loop in a single-stranded conformation,
which could be important for facilitating CNBP binding to this re-
gion. Thus, detailed mechanistic follow-up of this and other
models, as well as dissecting the function of additional Bvht
AGIL-interacting proteins, will be a focus of future investigation.
Figure 6. Loss of CNBP Partially Rescues the bvhtdAGIL Phenotype
(A) Cells at indicated time points during CM differentiation were analyzed for marker expression by flow cytometry. Numbers in plots indicate percentage of gated
populations.
(B) Immunofluorescence staining of indicated cells using anti-GFP (day 5.3) and anti-cTnT (day 10) antibodies. Nuclei were stained with DAPI. BF, bright field.
Scale bar, 100 mm.
(C and D) qRT-PCR analysis showing the relative levels of Brachyury (C) and core cardiac (D) TFs. WT value at day 4 (C) or at day 5.3 (D) was set to 1 for each gene.
(E) qRT-PCR analysis showing the relative levels of CM marker genes at day 10.
(F) Model of Bvht and CNBP regulating cardiovascular lineage commitment. Bvht functionally antagonizes the repression of CNBP on the transition from cardiac
mesoderm to progenitors. Potential additional factors working together with Bvht remained to be elucidated.
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 S6.
Molecular Cell 64 , 37–50, October 6, 2016 47