br Experimental Procedure br Author

Experimental Procedure
Author Contributions
Introduction The human heart develops as the first functional organ in the embryo. In the lateral plate mesoderm (LPM) cardiac progenitor carnitine palmitoyltransferase are induced, which can subsequently differentiate into heart muscle cells (cardiomyocytes) (Rosenthal and Harvey, 2010; Sylva et al., 2014). Wnt signaling, a well-known key regulator of vertebrate cardiomyocyte differentiation (Hoppler et al., 2014), acts through several molecular mechanisms (Hoppler and Nakamura, 2014): a β-catenin-dependent, so-called canonical pathway, and β-catenin-independent, so-called noncanonical pathways, among which a JNK-dependent pathway is prominent during heart development (Gessert and Kühl, 2010; Gessert et al., 2008; Pandur et al., 2002a, 2002b). Several studies in mouse and other experimental models have described diverse, and often opposing, effects of canonical and noncanonical Wnt signaling on subsequent cardiomyocyte differentiation, leading to the argument that particularly the JNK-mediated noncanonical pathway may function in this context to antagonize canonical Wnt signaling (Abdul-Ghani et al., 2011; Cohen et al., 2012; reviewed by Hoppler et al., 2014). In addition, canonical Wnt signaling has been shown to play multiple and conflicting roles at different stages of heart development (Gessert and Kühl, 2010; Naito et al., 2006). However, specific roles for Wnt signaling in human cardiomyocyte development remain unclear, particularly regarding which endogenous Wnt ligands and Wnt receptors are involved.
Results
Discussion Vertebrate and invertebrate models have proved fundamental for gaining an understanding of animal heart development and the signaling and transcriptional network mechanisms governing this process. However, the molecular mechanisms underlying human heart development are still largely unclear. hESCs offer unprecedented opportunities to model and study human heart development in vitro. However, so far most of the effort in the field has been directed toward designing efficient protocols to differentiate human cardiomyocytes in vitro. Interestingly, a vast majority of these protocols relies on experimental manipulation of Wnt signaling mechanisms (e.g., Bauwens et al., 2011; Burridge et al., 2014; Chen et al., 2012; Elliott et al., 2011; Gonzalez et al., 2011; Hemmi et al., 2014; Karakikes et al., 2014; Kattman et al., 2011; Kempf et al., 2014; Lian et al., 2012; Otsuji et al., 2014; Paige et al., 2010; Phillips et al., 2008; Ting et al., 2014; Yang et al., 2008; Zhang et al., 2015), suggesting important roles for WNT signaling also in human heart development. However, while Wnt signaling clearly represents a key regulator of vertebrate heart development and particularly of cardiomyocyte differentiation (reviewed by Hoppler et al., 2014), at present there is no detailed understanding of the players and the fundamental roles of Wnt signaling at sequential developmental stages leading to human embryonic cardiomyocyte differentiation. We therefore specifically set out to test whether knowledge from animal model systems would be confirmed for human cardiomyocyte development. Using established hESC differentiation protocols, we studied the activity and requirement of Wnt signaling pathways and identified WNT signal and receptor genes during human cardiomyocyte differentiation. Evidence from model systems had previously suggested a requirement for Wnt signaling during mesoderm induction. In fact, Brachyury is a known target of β-catenin-mediated Wnt signaling (e.g., Arnold et al., 2000; Mendjan et al., 2014; Yamaguchi et al., 1999; Zhang et al., 2013), while WNT/JNK signaling had previously been associated with morphogenesis at gastrulation (Hardy et al., 2008; Tada et al., 2002; Tada and Kai, 2009). Here, we not only confirm that similarly hESCs can only efficiently differentiate in vitro into mesoderm when Wnt signaling is active, but also show that canonical Wnt signaling acts before noncanonical Wnt signaling in humans (Figure 4H). Our data suggest that WNT3 and WNT8A regulate BRY (T) expression and mesoderm induction via the canonical pathway, after which WNT5A and WNT5B activate JNK-mediated pathway activity to regulate MESP1 expression, thereby indicating a role for WNT/JNK signaling that goes beyond regulating morphogenesis during gastrulation in the intact embryo. Consistently, our results also show expression of FZD7 and the noncanonical receptor ROR2 during mesoderm induction. Interestingly FZD7 had previously been identified in pluripotent hESCs, where it plays a role in canonical WNT3-mediated self-renewal (Fernandez et al., 2014). Our results suggest that WNT3 and WNT8A and the canonical pathway regulate mesoderm induction via the FZD7 receptor (although additional roles for FZD7 in mediating noncanonical mechanisms cannot be ruled out, e.g., Medina et al., 2000), while ROR2 mediates WNT5A/B role in the commitment of the earliest cardiogenic mesoderm (i.e., MESP1-positive cells).