Wnt3a signal pathways activate MyoD expression by targeting cis-elements inside and outside its distal enhancer

Abstract

Wnt proteins are secreted cytokines and several Wnts are expressed in the developing somites and surrounding tissues. Without proper Wnt stimulation, the organization of the dermomyotome and myotome can become defective. These Wnt signals received by somitic cells can lead to activation of Pax3/Pax7 and myogenic regulatory factors (MRFs), especially Myf5 and MyoD. However, it is currently unknown whether Wnts activate Myf5 and MyoD through direct targeting of their cis-regulatory elements or via indirect pathways. To clarify this issue, in the present study, we tested the regulation of MyoD cis-regulatory elements by Wnt3a secreted from human embryonic kidney (HEK)-293T cells. We found that Wnt3a activated the MyoD proximal 6.0k promoter (P6P) only marginally, but highly enhanced the activity of the composite P6P plus distal enhancer (DE) reporter through canonical and non-canonical pathways. Further screening of the intervening fragments between the DE and the P6P identified a strong Wnt-response element (WRE) in the upstream -8 to -9k region (L fragment) that acted independently of the DE, but was dependent on the P6P. Deletion of a Pax3/Pax7-targeted site in the L fragment significantly reduced its response to Wnt3a, implying that Wnt3a activates the L fragment partially through Pax3/Pax7 action. Binding of β-catenin and Pax7 to their target sites in the DE and the L fragment respectively was also demonstrated by ChIP. These observations demonstrated the first time that Wnt3a can directly activate MyoD expression through targeting cis-elements in the DE and the L fragment.

Figure 1. Establishment of Wnt3a-secreting HEK-293T stable clones

(A) The Wnt3a levels (in 20 μl of medium) accumulated in the last 24 h of HEK-293T-Py and -Wnt3a stable clones cultured in serum-free medium or regular medium (with 10% FBS) was determined by Western blotting. (B) Growth medium (containing 10% FBS) from HEK-293T-Wnt3a cells was collected on day 1 and day 2 after cells became confluent. Different volumes of this medium were then subjected to Western blotting using anti-Wnt3a antibody. Total lysate (50 μg) from cells kept in medium containing 10% FBS was used as positive control. (C) Either TOPflash or FOPflash was transiently transfected into C2C12 myoblasts and then treated with medium from HEK-293T-Py or -Wnt3a cells for 24 h before being harvested for determination of luciferase activity. The activity in cells treated with Py medium was set as 1-fold activation. **P<0.01 compared with Py medium. (D and E) Wnt3a activates MyoD expression in C2C12 cells (D) and in somite explants (E). The expression level of MyoD in proliferating (PMB) and confluent (CMB) C2C12 cells and in cultured somite explants treated with or without Wnt3a medium was determined by qRT-PCR. The MyoD level in cells or somites treated with Py medium was set as 1-fold. **P<0.01 compared with Py medium.

Figure 2. Wnt3a activates MyoD expression via distal enhancer

(A) Schematic diagrams of the construction of MyoD promoter-driven reporters. The genomic organization of the MyoD gene is shown on the top. Reporters driven by the proximal 6.0k promoter (P6P) or P6P plus the distal enhancer were named MyoD-P6P-luc and MyoD-PE-luc respectively. An MluI site between the promoter and enhancer allows insertion of other genomic fragments. The activation of both reporters by Wnt3a is shown in (B). (C) Diagrams showing various MyoD-PE-luc-derived reporters in which the DE was replaced by various individual DE fragments. Multiple putative TCF/LEF target sites (T1–T6, blue boxes) in the DE and the conserved core enhancer (258 bp, orange box) are indicated. The responses of these reporters to Wnt3a are shown in (D). *P<0.05 and **P<0.01 compared with full-length (FL) DE. All reporters used were stably carried by C2C12 cells.

Figure 3. Wnt3a activates MyoD expression through multiple pathways

(A) MyoD-PE-Luc was transiently co-transfected with vectors expressing dominant-negative (DN) mutants of Wnt pathway downstream effectors into C2C12 cells treated with or without Wnt3a. Luciferase activity was determined 24 h after transfection and their relative activation by Wnt3a is shown. (B) C2C12-MyoD-PE-luc stable clones were treated with Wnt3a in the presence or absence of MAPK inhibitors or A23187 for 48 h, and their relative activation by Wnt3a is shown. (C) Increasing amount of engrail–β-catenin (EN-β-catenin) was co-transfected with MyoD-PE-Luc into C2C12 cells as described in (A). (D and E) C2C12-MyoD-PE-luc stable clones were treated with Wnt3a or LiCl or neither (D) or increasing amounts of LiCl alone (E) for 48 h, and their activity relative to Py or vehicle control is shown. *P<0.05 and **P<0.01 compared with vector or vehicle control.

Figure 4. Constitutively active β-catenin enhances MyoD expression

(A) Transient transfection of MyoD-PE-luc and β-catenin expression vectors into C2C12 cells and the luciferase activity was assayed 48 h after transfection. Δ90 and Δ151 represent N-terminally truncated β-catenin without the first 90 and 151 amino acids respectively. (B) C2C12 stable clones carrying either Py or Py-β-catenin Δ151–FLAG were established and their β-catenin Δ151–FLAG levels were examined by Western blotting with anti-FLAG antibody. Gapdh served as a control. (C) Myotube morphology and fusion index (%) after 4 days in differentiation medium. (D) Relative MyoD mRNA levels in these stable clones of various stages [proliferating myoblast (PMB), confluent myoblast (CMB) and myotube (MT)] were determined by qRT-PCR. (E) Binding of β-catenin Δ151–FLAG to the TCF/LEF sites in the MyoD gene DE region was examined by ChIP assay. Chromatin of the C2C12-β-catenin Δ151–FLAG cells at the CMB stage was precipitated with anti-FLAG antibody or control IgG and then the DNA was amplified with the indicated primer sets (P1–P3). The signals of Dickkopf-related protein 1 (DKK1) and M-cadherin serve as positive and negative controls respectively. The amplicon of each primer set is shown to the right. *P<0.05 and **P<0.01 compared with vector or vehicle control.

Figure 5. Cis-elements between the DE and the P6P also mediate the Wnt3a response

(A) Genomic fragments between the DE and the P6P were inserted into the MluI site in MyoD-PE-luc and their response to Wnt3a is shown in (B). SP, a random vector sequence (1 kb) without known transcription factor-binding sites serves as a non-specific spacer. All reporters used were stably carried by C2C12 cells. (C) C2C12 stable clones carrying either MyoD-PE-luc or MyoD-6.0-L-luc were treated with Wnt3a and their luciferase activity was determined 48 h after treatment. *P<0.05 and **P<0.01 compared with vehicle (Py) control. (D) Transient transfection of reporters into C2C12 cells to test their response to Wnt3a as described in Figure 1(C). TOP/FOP, TOPflash/FOPflash; TKE: thymidine kinase promoter and enhancer driven reporter; TKE-L: reporter driven by L fragment plus TKE. (E and F) C2C12 stable clones carrying various MyoD cis-elements were treated with Wnt3a and/or its downstream pathways inhibitors or agonists to identify pathways mediating the activation of the L fragment.

Figure 6. The L fragment mediates the Wnt3a response via canonical and Pax3/Pax7-mediated pathways

(A) Schematic diagrams of the locations of the Pax3/Pax7-binding site in the L fragment. (B) C2C12 stable clones carrying P6P-L or P6P-L-ΔPax3/7 were treated with vehicle or Wnt3a. **P<0.01 compared with Py; ##P<0.01 compared with Wnt3a-treated P6P-L-luc. (C) The relative mRNA levels of MyoD, M-cadherin (M-cad), Pax3 and Pax7 in CMB C2C12 cells treated with or without Wnt3a were examined with qRT-PCR. The mRNA level of each gene in cells treated with Py medium was set as 1-fold. **P<0.01 compared with Py. (D) ChIP assay of vehicle or Wnt3a-treated C2C12 cells with IgG or anti-Pax7 antibody and the primer set (L1) shown in (A). The signals of the Id3 gene serve as a positive control of the Pax7-targeted site. NTC, no template control.