Endogenous Wnt/beta-catenin signaling is required for cardiac differentiation in human embryonic stem cells

Abstract

Background: Wnt/beta-catenin signaling is an important regulator of differentiation and morphogenesis that can also control stem cell fates. Our group has developed an efficient protocol to generate cardiomyocytes from human embryonic stem (ES) cells via induction with activin A and BMP4.

Methodology/principal findings: We tested the hypothesis that Wnt/beta-catenin signals control both early mesoderm induction and later cardiac differentiation in this system. Addition of exogenous Wnt3a at the time of induction enhanced cardiac differentiation, while early inhibition of endogenous Wnt/beta-catenin signaling with Dkk1 inhibited cardiac differentiation, as indicated by quantitative RT-PCR analysis for beta-myosin heavy chain (beta-MHC), cardiac troponin T (cTnT), Nkx2.5, and flow cytometry analysis for sarcomeric myosin heavy chain (sMHC). Conversely, late antagonism of endogenously produced Wnts enhanced cardiogenesis, indicating a biphasic role for the pathway in human cardiac differentiation. Using quantitative RT-PCR, we show that canonical Wnt ligand expression is induced by activin A/BMP4 treatment, and the extent of early Wnt ligand expression can predict the subsequent efficiency of cardiogenesis. Measurement of Brachyury expression showed that addition of Wnt3a enhances mesoderm induction, whereas blockade of endogenously produced Wnts markedly inhibits mesoderm formation. Finally, we show that Wnt/beta-catenin signaling is required for Smad1 activation by BMP4.

Conclusions/significance: Our data indicate that induction of mesoderm and subsequent cardiac differentiation from human ES cells requires fine-tuned cross talk between activin A/BMP4 and Wnt/beta-catenin pathways. Controlling these pathways permits efficient generation of cardiomyocytes for basic studies or cardiac repair applications.

Figure 1. Early Wnt/β-catenin signaling enhances cardiac differentiation.

(A) Quantitative RT-PCR on day 14 of differentiation shows induction of β-MHC, cTnT and Nkx2.5 expression by activin A/BMP4 is attenuated with Dkk1. Data are presented as mean fold expression normalized to HPRT relative to control cells. (B) Flow cytometry analysis for sarcomeric myosin heavy chain (sMHC) on day 21 of differentiation shows enhanced cardiac differentiation with addition of exogenous Wnt3a and decreased cardiac differentiation with Dkk1. (C) Late addition of Dkk1 enhanced day 14 β-MHC expression relative to samples treated with activin A/BMP4 only. Data are presented as mean fold expression normalized to HPRT relative to control cells. *p<0.05 relative to control, **p<0.01 relative to control, †p<0.05 relative to activin A/BMP4, ‡p<0.01 relative to activin A/BMP4.

Figure 2. Canonical Wnt expression during mesoderm specification correlates with the efficiency of cardiac differentiation.

Quantitative RT-PCR for canonical Wnt ligands Wnt1, Wnt3a and Wnt8a. Data are presented as mean fold expression normalized to HPRT relative to undifferentiated cells. (A) Representative of cultures of H7 hES cells that showed efficient and inefficient directed differentiation to cardiomyocytes (>25% and <5% sMHC+, respectively), as well as H1 hES cells that are poorly cardiogenic. (B) Cultures that routinely generated <5% cardiomyocytes were treated with Wnt3a or Dkk1 in addition to activin A and BMP4. After 17 days of differentiation following activin A treatment, cells were stained for sMHC. Exogenous Wnt3a partially rescued inefficient cardiac differentiation.

Figure 3. Wnt/β-catenin signaling enhances mesoderm induction.

Quantitative RT-PCR analysis on day 2 of differentiation shows induction of Brachyury (mesoderm), FoxA2 (endoderm), and Sox1 (ectoderm). Brachyury expression induced by activin A/BMP4 was enhanced by Wnt3a and reduced by Dkk1. Data are presented as mean fold expression normalized to HPRT relative to undifferentiated cells + SEM. *p<0.05 relative to control, **p<0.01 relative to control, †p<0.05 relative to activin A/BMP4, ‡p<0.01 relative to activin A/BMP4.

Figure 4. Wnt/β-catenin signaling regulates Smad1 activation.

High-density undifferentiated human ES cells were treated with activin A for 24 hours followed by BMP4 or BMP4 plus Dkk1. Lane marked as “pre” indicates cells collected prior to cytokine treatment. Western blot shows probing for C-terminal phosphorylated Smad1 (p-Smad1) and total Smad1 at 30 m, 1 h, 2 h and 3 h following BMP4. Densitometry data are presented as phosphorylated Smad1 normalized to total Smad1 relative to pretreated cultures. Treatment with BMP4 increased phosphorylation of Smad1, which was reduced by inhibition of canonical Wnt signaling with Dkk1.