Autophagy is involved in ethanol-induced cardia bifida during chick cardiogenesis

Abstract

Excess alcohol consumption during pregnancy has been acknowledged to increase the incidence of congenital disorders, especially the cardiovascular system. However, the mechanism involved in ethanol-induced cardiac malformation in prenatal fetus is still unknown. We demonstrated that ethanol exposure during gastrulation in the chick embryo increased the incidence of cardia bifida. Previously, we reported that autophagy was involved in heart tube formation. In this context, we demonstrated that ethanol exposure increased ATG7 and LC3 expression. mTOR was found to be inhibited by ethanol exposure. We activated autophagy using exogenous rapamycin (RAPA) and observed that it induced cardiac bifida and increased GATA5 expression. RAPA beads implantation experiments revealed that RAPA restricted ventricular myosin heavy chain (VMHC) expression. In vitro explant cultures of anterior primitive streak demonstrated that both ethanol and RAPA treatments could reduce cell differentiation and the spontaneous beating of cardiac precursor cells. In addition, the bead experiments showed that RAPA inhibited GATA5 expression during heart tube formation. Semiquantitative RT-PCR analysis indicated that BMP2 expression was increased while GATA4 expression was suppressed. In the embryos exposed to excess ethanol, BMP2, GATA4 and FGF8 expression was repressed. These genes are associated with cardiomyocyte differentiation, while heart tube fusion is associated with increased Wnt3a but reduced VEGF and Slit2 expression. Furthermore, the ethanol exposure also caused the production of excess ROS, which might damage the cardiac precursor cells of developing embryos. In sum, our results revealed that disrupting autophagy and excess ROS generation are responsible for inducing abnormal cardiogenesis in ethanol-treated chick embryos.

Abbreviations

Atg

Autophagy-related genes

BMP2

Bone morphogenetic protein 2

_EC culture

Early chick culture

FASD

Fetal Alcohol Spectrum Disorder

FGF

fibroblast growth factor

HH

Hamburger and Hamilton

RAPA

Rapamycin

ROS

reactive oxygen species

VMHC

ventricular myosin heavy chain

Introduction

Excess alcohol consumption during pregnancy has become a major social concern because it is associated with increased risks of growth and mental retardation, as well as congenital malformations.¹ Clinical studies have also suggested that maternal ethanol exposure could lead to immune deficiencies in neonates – implying that ethanol does not only affect the nervous system but also other systems. Bake et al. reported that ethanol disrupted angiogenesis and neurogenesis in the developing brain.² The collective damage induced by maternal alcohol consumption is called Fetal Alcohol Spectrum Disorders (FASD), which includes defects in the brain, craniofacial structures, cardiovascular system and the limbs.^3-5 However, little is known about the underlying mechanism of FASD pathogenesis induced by maternal ethanol exposure. In this study, we investigated the effects of ethanol on the early stages of cardiogenesis in the chick embryo to elucidate the mechanisms involved in ethanol-induced cardiac malformation. The negative impact of embryonic exposure to ethanol during cardiogenesis has been widely studied since congenital heart diseases account for one-third of all birth defects. Daft et al. demonstrated that intraperitoneal injection of ethanol into pregnant C57 mice induced abnormal fetal heart development – notably, formation of abnormal contours of the heart tube at 9 day gestation.⁶ However, the underlying cause is unknown, especially how ethanol affects cardiac precursor cell differentiation during early heart tube morphogenesis. Hence, we attempted to address this questions using the early chick embryo as an experimental model.⁷

The major components of the embryonic heart tube are cardiac precursor cells (from primary and secondary heart fields), cardiac neural crest and the pro-epicardium. Cardiogenesis involves primary heart tube fusion, cardiac looping and accretion, cardiac septation and coronary vasculogenesis in a set chronological order.⁸ The primary heart field contributes to the major structures of the heart, including the atria and ventricles, while the secondary heart field gives rise to the cardiac outflow tracts.⁹ During gastrulation, the precardiac cells that migrate out of the anterior primitive streak and are then directed to move laterally and anteriorly to form the bilateral primary heart fields.^10-12 Borgave et al. reported that there were mesenchymal stem cells present in the primary heart field of HH 4 chick embryo and that these stem cells could differentiate into various cell lineages, such as cardiomyocyte in response to inductive signals from the Hensen's node.¹³ In this context, any disruption to cardiac precursor cell migration and differentiation during cardiogenesis may lead to congenital heart malformations. Ste´phane Zaffran, et al. demonstrated that altering the embryonic origin of right and left ventricular myocardium has important implications in congenital heart disease.¹⁴

Heart development is a complex process that is tightly regulated through spatio-temporal gene expressions and cell-cell interaction. Specifically, fibroblast growth factor (FGF) signaling is required for regulating pro-cardiac mesoderm cell migration.^11,15 and cardiac precursor cell specification occurs when the cells reach to the anterior lateral plate mesoderm. Bone morphogenetic protein 2 (BMP2) released from the anterior endoderm plays an important role as an inducer of myocardium differentiation. Moreover, transcription factors Nkx-2.5, GATA4, myocardin and Tbx20 play crucial roles in dictating morphogenesis and differentiation of the heart.¹⁶

Physiologically, autophagy is the process where energy is provided for embryonic development through the lysosomal degradation of cellular contents.¹⁷ Autophagy also acts as a housekeeper by preventing protein accumulation within cells through removing dead or damaged organelles.¹⁷ Autophagy-related genes (Atg), Atg5, Atg7, and Atg8 (LC3) participate at various stages of the autophagy process.¹⁸ In addition, Atg5 and Atg7 mutant mice could survive and develop during the embryonic period but die soon after birth.^19,20 Recently, there are more attention being paid to the role of autophagy in embryonic development and RAPA has been used as a pharmacological inducer of autophagy.

In this study, we investigated the effects of ethanol exposure on the pathogenesis of cardiac bifida. We also examined the role of autophagy and excess ROS generation in ethanol-induced cardiac bifida.

Materials and Methods

Chick embryos and treatment

Fertilized chick eggs were obtained from the Avian Farm of the South China Agriculture University. The eggs were incubated until the required Hamburger and Hamilton (HH) stage.²¹ inside a humidified incubator (Yiheng Instrument, Shanghai, China) at 38°C and 70% humidity. _EC (early chick) culture.²² was employed to culture gastrula chick embryos which make them amenable to experimentation. The 0HH chick embryos in _EC culture were treated with either 2% (342.5mM) ethanol or simple saline (control) until the experimentally required developmental stage. The HH0 chick embryos were also incubated with 40 nM RAPA (LC Labs, USA), 1 μg/ml tunicamycin (Sigma- aldrich, USA), 150 ng/ml thapsigargin (Sigma-aldrich, USA).^23,24 or 13.8mM AAPH (Sigma-aldrich, USA),²⁵ while the control embryos were exposed to 1‰ DMSO or PBS as a control, respectively. Briefly, these compounds were directly applied to either the _EC cultures or in vitro culture medium to reach the final concentrations 40 nM (RAPA), 1 μg/ml (tunicamycin), 150 ng/ml (thapsigargin) and 13.8 mM AAPH respectively. The treated embryos were then incubated for either 18 or 45 hours before they were fixed in 4% paraformaldehyde for histological, morphological and molecular analysis.

Explant culture

We divided the streak-stage chick embryos (HH3) into 6 equal segments, with the second segment from the cranial side treated as the anterior primitive streak (Fig. S1). The anterior primitive streak explants were cultured in DMEM-F12 culture medium (Life Technologies) at 37°C and 5% CO₂.²⁶ The incubation time varies depending on the experimental requirement. Each treatment was performed in triplicates. Approximately, 95% of the cells in the control explant culture were determined to be cardiomyocyte, as validated by the presence of myosin heavy chain using MF20 immunofluorescent staining (Fig. S1).

In situ hybridization

Whole-mount in situ hybridization of chick embryos was performed according to standard in situ hybridization protocol.²⁷ Briefly, Digoxigenin-labeled probes were synthesized for VMHC.²⁸ and GATA5.²⁹ Dr. Thomas M. Schultheiss kindly provide the GATA5 plasmid for generating GATA5 riboprobe. The whole-mount stained embryos were photographed and then frozen sections were prepared from the embryos on a cryostat microtome (Leica CM1900) at a thickness of 14-18 μm.

Immunofluorescent staining

The chick embryos were harvested at the end of experimentation and fixed in 4% paraformaldehyde overnight at 4°C. Whole-mount embryos were immunofluorescently stained using MF20 (1:500, DSHB, USA), mTOR (1:50, Bioworld, USA) and Atg7 (1:200, Sigma-aldrich, USA) antibodies. Briefly, the fixed embryos were incubated with these primary antibodies at 4°C overnight on a rocker. Following extensive washing, the embryos were incubated with the appropriate anti-mouse IgM conjugated to Alexa Fluor 555 or anti-rabbit IgG conjugated to Alexa Fluor 488, overnight at 4°C on a rocker. All the embryos were finally counterstained with DAPI (1:1000, Invitrogen, USA) at room temperature for 1 hour.

RNA isolation and semiquantitative RT-PCR

Total RNA was isolated from HH10 chick embryos using a Trizol kit (Invitrogen, USA) according to the manufacturer's instructions. First-strand cDNA was synthesized to a final volume of 25 μl using SuperScript RIII first-strand (Invitrogen, USA). Following reverse transcription, PCR amplification of the cDNA was performed as described previously.^30,31 The primers used for RT-PCR are provided in the Fig. S2. The PCR reactions were performed on a Bio-Rad S1000TM Thermal cycler (Bio-Rad, USA). The final reaction volume was 50 μl composed of 1 μl of first-strand cDNA, 25 μM forward primer, 25 μM reverse primer, 10 μl PrimeSTARTM Buffer (Mg2+ plus), 4μl dNTPs Mixture (TaKaRa, Japan), 0.5 μl PrimeSTARTM HS DNA Polymerase (2.5U/μl TaKaRa, Japan) and RNase-free water. The cDNA was amplified for 30 cycles. One round of amplification was performed at 94°C for 30 s, 30 s at 58°C, and 30 s at 72°C. The PCR products (20 μl) were resolved using 1% agarose gels (Biowest, Spain) in 1× TAE buffer (0.04 M Trisacetate and 0.001 M EDTA) and 10,000x GeneGreen Nucleic Acid Dye (Tiangen, China) solution. The resolved products were visualized using a transilluminator (Syngene, UK), and photographs were captured using a computer-assisted gel documentation system (Syngene).

Western blot

Chick embryos (HH10) were collected and lysed with CytoBuster™ Protein Extraction Reagent (#71009, Novagen). The total protein concentration was established using a BCA quantification kit (BCA01, DingGuo BioTECH, CHN). Samples containing equal amounts of protein were resolved by SDS-PAGE and then transferred to PVDF membranes (Bio-Rad). The membranes were blocked with 5% Difco™ skim milk (BD) and then incubated with primary and secondary antibodies. The antibodies used were LC3B (Cell Signaling Technology, USA); mTOR (Bioworld, USA); HRP-conjugated anti-mouse IgG and anti-rabbit IgG (Cell Signaling Technology, USA). All primary and secondary antibodies used were diluted to 1:1000 and 1:2000 in 5% skim milk, respectively. The protein bands of interest were visualized using an ECL kit (#34079, Thermo Fischer Scientific Inc..) and GeneGnome5 (Syngene). The staining intensity of the bands was determined and analyzed using Quantity One software (Bio-Rad).

Transmission electron microscopy

Control and ethanol-treated HH4 chick embryos were fixed with 2.5% glutaraldehyde in 0.1 M PBS for 2 hours. The anterior primitive streaks were then dissected and stained with Osmium tetroxide. The specimens were embedded in resin, ultrathin sectioned and examined using a Tecnai G2 Spirit Twin (FEI, USA).

Beads transplantation experiments

Heparin beads were soaked in 40 nM RAPA (LC Labs, USA).²³ for 5 hours. The beads were then were inserted into the anterolateral site of the Hensen's node of HH4 chick embryos in _EC culture. Beads soaked in 1‰ DMSO were implanted in another anterolateral site of Hensen's node as a control. The embryos were incubated continually until they reached HH9.(Fig. 4G)

Figure 4.

RAPA-induced inhibition of autophagy induces cardia bifida during cardiogenesis. (A-B) Representative bright-field image of whole mount control chick embryos (A) and MF20 immunofluorescent staining of their C-loop heart tubes (B). (B1-B1′) Transverse sections of the heart at the level indicated by white dotted line in B (B1) and high magnification of B1 (B1′). Sections counterstained with DAPI. (C-D) Representative bright-field image of whole mount RAPA-treated chick embryos (A) and MF20 immunofluorescent staining of bifida heart tube (D). (D1-D1′) Transverse sections at level indicated by dotted white line in D (D1) and higher magnification of D1 (D1′). (E-F) In situ hybridization showing GATA5 expression in control (E) and RAPA-treated (F) embryos. (E1-F1) High magnification of the framed heart regions E and F. (E2-F2) Transverse sections (levels indicated in E1 and F1) showing GATA5 expression in stronger in RAPA-treated hearts than control hearts. (G) In situ hybridization showing VMHC expression after implantation of beads, impregnated with DMSO (left hand side, control) and 40nM RAPA (right hand side), in the anterolateral sites of HH4 Hensen's node for 18 hours. (G1) High magnification of the implant site (square frame) in G. (G2) Transverse section of G1 showing VMHC expression (arrows) was unaffected by DMSO but inhibited by RAPA. (H) Bar chart showing incidence of cardia bifida after RAPA-treatment compare with control. (I) Bar chart showing the developmental stages that embryos managed to reach (including mortality) after RAPA-treatment. Abbreviation: RAPA, rapamycin. Scale bars = 600 µm in A, C; 200 µm in B, D, E1,F1,G1; 200 µm in B1, D1, G2, E2-F2; 100 µm in B1′, D1′and 600 µm in E-F, G.

Photography

Following immunofluorescent staining or in situ hybridization, the whole-mount embryos were photographed using a stereo-fluorescent microscope (Olympus MVX10) and associated Olympus software package Image-Pro Plus 7.0. The embryos were sectioned into 14 µm-thick slices using a cryostat microtome (Leica CM1900) and then the sections were photographed using an epi-fluorescent microscope (Olympus LX51, Leica DM 4000B) with the CN4000 FISH Olympus software package.

We conducted time-lapse photography on anterior primitive streak explants inside an incubation chamber. The chamber was heated to 37°C by humidified warm air (World Precision Instruments) and mounted on the stage of an inverted microscope (Nikon Eclipse TI-U). Bright-field images were captured every 1 min, continuously over several hours, using a cooled CCD camera (Nikon DS-Fi1c). We processed the resulting images to generate movies using the VideoMach program.

Measurement of intracellular reactive oxygen species (ROS)

Intracellular ROS was determined using a non-fluorescent dye DCF-DA( 2′7-dichlorodihydrofluorescein diacetate) (Sigma-aldrich, USA) which could be oxidized by ROS to the fluorescentdye DCF (2,7-dichlorofluorescin ). H9c2 cells were incubated for 15 h at 37°C and 5% CO2. The cells were then exposed to simple saline or 2% ethanol (342.5mM) for 30 min. At the end of treatment, the cells were incubated in 10μM DCF-DA for 20 min. The intensity of the fluorescence was measured using a BD FACSAria (USA).

Data analysis

Statistical analysis for all the experimental data generated was performed using a SPSS 13.0 statistical package program for Window. The data were presented as mean ± SD. Statistical significance were determined using paired T-test, independent samples T-test or one-way analysis of variance (ANOVA). P < 0.05 was considered to be significant.

Results

Ethanol exposure increases cardia bifida during chick cardiogenesis

We investigated the effects of excess ethanol exposure on heart tube formation in the early chick embryos. Firstly, we found that 342.5mM of ethanol retarded the growth of gastrulating chick embryos (Fig. S3). Normally, fertilized eggs require 38-hours incubation for embryos to reach stage HH11 (Fig. S3A-C). However, some of ethanol-treated eggs require 43-hours incubation to reach stage HH11 (Figs. S3D-G). After 38-hours incubation, only 26% of embryos reached to HH11, 6% held at HH7-8, 58% held at HH4-6.death and 10% embryos had died (Fig. S3H).

We did not find any obviously heart tube malformation between control and ethanol-treated embryos before heart tube fusion at stage HH8-9 (Fig. 1A-B and D-E). At stage HH11, the C-shape loop of the heart tube has formed in control embryos (Fig. 1C and G) as revealed by MF20 immunofluorescent staining (Fig. 1J). The single cavity of the heart tube was also evident in corresponding transverse section in these control embryos (Fig. 1J). In contrast, the heart tubes of some of the stage HH11 chick embryos treated with ethanol presented the characteristic features of cardia bifida, as shown in whole-mount of the heart tube labeled with MF20 immunofluorescence (Fig. 1L) and corresponding transverse section of the heart tube (Fig. 1L’). There were 2 cavities present in the heart tube of these transverse sections and the rate of cardia bifida development was 22% (n = 11/50, Fig. 1M).

Figure 1.

Development of cardia bifida following ethanol exposure in gastrulating chick embryos. (A-C) Representative appearance of control HH8 (A), HH9 (B) and HH11 (C) chick embryos exposed to simple saline. The dotted white lines outline the developing heart tube. (D-F) Representative appearance of HH8 (D), HH9 (E) and HH11 (F) chick embryos treated with 2% ethanol (342.5mM). (G-H) Illustrations showing the overall shape of control and ethanol-treated hearts. (I-J’) Representative images of control embryo under bright-field (I), heart tube immunofluorescently stained with MF20 antibody (J) and transverse section of the heart tube (J’) at the level indicated by dotted lines in J. (K-L’) Representative images of ethanol-treated embryo under bright-field (K), heart tube immunofluorescently stained with MF20 antibody (L) and transverse section of the heart tube (L’) at the level indicated by dotted white lines in L. (M) Bar chart showing the incidence of cardia bifida (%) in control and ethanol-treated embryos. Scale bars = 200 µm in A-F; 600 µm in I, K; 200 µm in J, L and 200 µm in J’, L’.

Ethanol enhances autophagy in gastrulating chick embryos

The heart tube is derived from the bilateral primary heart fields of the anterior primitive streak. The mesodermal cells in the heart field migrate bilaterally out of the primitive streak.^10,11 In our previously study, we found that the autophagy was involved in EMT.³² and heart tube formation.³³ in early chick embryo. Hence, we examined whether autophagy was involved in the pathogenesis of ethanol-induced cardia bifida. It has been reported that ATG7 was central to the activation of cell autophagy.²⁰ Using immunofluorescent staining, we showed that ATG7 was normally expressed in the apical side of the epiblast and preferentially in the hypoblast layer of HH4 chick gastrulating embryos (Fig. 2A, B). However, exposing the gastrulating chick embryos to ethanol, dramatically increased ATG7 expression in all the 3 germ layers (epiblast, mesoderm and hypoblast), which was discernable in whole-mount embryos and transverse sections of the embryos (Fig. 2C, D). Furthermore, transmission electron microscopy showed the formation of autophagosomes (Fig. 1E, arrow) and membrane accumulation of vacuoles with remanent cytoplasmic contents in primitive streak embryos (Fig. 1F, arrow). We also examined the expression of another important autophagy gene, LC3, in the gastrulating embryos by RT-PCR and western blot following ethanol exposure. The results showed ethanol exposure significantly increased ATG7 and LC3 expression at mRNA level (P < 0.05, n = 3; Fig. 2G). Western blot revealed the ratio of LC3ІІ/LC3І protein level was also increased in ethanol-treated embryo (P < 0.001, n = 3; Fig. 2H).

Figure 2.

Ethanol treatment alters ATG7 and LC3 expression in gastrulating chick embryos. (A-B) Representative bright-field (A) and ATG7 immunofluorescent stained (B) images of control HH4 chick embryo. (B1-B1′ and B2-B2′) Representative transverse sections of the control embryo at the levels indicated by dotted white lines in A and B. DAPI staining is used as a counterstain in B1′-B2′. (C-D) Representative bright-field (C) and ATG7 immunofluorescent stained (D) images of ethanol-treated HH4 chick embryo. (D1-D1′ and D2-D2′) Representative transverse sections at the levels indicated by dotted white lines in C and D. DAPI staining is used as a counterstain in D1′-D2′. (E-F) Transmission electron microscopy was performed in primitive streak chick embryo (HH4). An omegasome (arrow in E), a double-membrane protrusion from cytoplasmic organelles that serves as a platform for autophagosome biogenesis. F showed an autophagic vesicle containing cytoplasmic remnants (arrow). (E) RT-PCR and corresponding bar chart showing ATG7 and LC3 expressions were upregulated in gastrulating embryos after ethanol exposure. (H) Western blot and corresponding bar chart showing LC3-І expression was down-regulated while LC-3ІІ was increased in ethanol-treated embryos. *p < 0.05 and ***p < 0.001 indicating significant differences between ethanol-treated and control groups. Scale bars = 600 µm in A-B, C-D, 200 nm in E-F and 50 µm in B1-B2, B1′-B2′, D1-D2, D1′-D2′

Ethanol inhibits mTOR expression in gastrulating chick embryos

In the chick gastrula (HH4), mTOR is weakly expressed in the epiblast, mesoderm and hypoblast (Fig. 3A-A1). In older HH10 embryos, mTOR is preferentially expressed in the apical layer of neural epithelium (Fig. 3B1-B1) and diffusely expressed in the extraembryonic area opaca (Fig. 3B). However, ethanol exposure repressed mTOR expression in both stage HH4 and HH10 embryos. Specifically, mTOR was mainly expressed in the apical side of the epiblast in these HH4 ethanol-treated embryos (Fig. 3C-C1). In contrast, it was mainly expressed in the endoderm of HH10 ethanol-treated embryos (Fig. 3D-D1). We performed protein gel blot to confirm that ethanol treatment significantly reduced mTOR expression at the protein level (Fig. 3E). Furthermore, RAPA treatment also similarly suppressed mTOR expression in the gastrula embryo (Fig. 3E).

Figure 3.

Ethanol exposure increases mTOR expression in gastrulating chick embryos. (A-B) Representative images of control HH4 (A) and HH10 (B) embryos immunofluorescently stained for mTOR. (A1-A1′ and B1-B1′) The transverse sections at levels indicated by dotted white line in B. The sections were counterstained with DAPI. (C-D) Representative images of ethanol-treated HH4 (C) and HH10 (D) embryos immunofluorescently stained for mTOR. (C1-C1′ and D1-D1′) Transverse sections at the levels indicated by dotted white line in C and D. The sections were counterstained with DAP. (E) Western blot analysis and corresponding bar chart showing that mTOR expression was repressed following ethanol- and RAPA-treatment. **p < 0.01 and ***p < 0.001 indicate significant differences between treated and untreated groups. Abbreviation: RAPA, rapamycin. Scale bars = 600 µm in A-D and 50 µm in A1-D1, A1′-D1′.

Rapamycin-induced autophagy increases risk of cardia bifida

RAPA has been demonstrated to activate autophagy in epithelial cells.³⁴ In previous study,³² we discovered that RAPA treatment promoted the expression of ATG7, and Atg7-mediated autophagy did appear in the gastrulation stage of early embryonic development. We incubated chick embryos (HH0) in the presence of 40nM of RAPA until they reached stage HH11, when the heart tube has formed. We found that RAPA-treatment induced the development of cardia bifida, with the heart tube containing 2 cavities (Fig. 4D-D1), while in control group the C-looping heart tube contains a single cavity (Fig. 4A-B1). The incidence rate of cardia bifida was 28% (n = 14/50) compared to 0% in the control group (Fig. 4H). We also examined whether RAPA exposure retarded the growth of the gastrulating chick embryos. In the control, 100% of the embryos have developed to stages HH9 - 11. However, for the RAPA-treated embryos, they were found at different stages of development 78% (n = 39/50) at HH4 - 6, 6% (n = 3/50) at HH7 – 8, 10% (n = 5/50) HH9 - 11 and 6% (n=3/50) of embryos had died (Fig. 4I). GATA5 is normally expressed in the primary heart tube.^35,36 Presently, GATA5 in situ hybridization also indicated the cardia bifida malformation in RAPA-treated embyo (Fig. 4E-F2). Thapsigargin and Tunicamycin could also induce autophagy because they are able to increase ER stress.²³ However, both of these 2 compounds did not induce cardia bifida even though ER stress might be negatively regulates by the AKT/TSC/mTOR pathway.³⁷

To determine whether if RAPA could affect myocardial differentiation, we applied RAPA to the site where the heart tube forms by transplanting beads impregnated with RAPA or DMSO (control) in HH4 chick embryos. These experimental embryos were allowed to develop until the heart tube stage (Fig. 4G). In situ hybridization with ventricular myosin heavy chain (VMHC) riboprobes revealed that VMHC expression was stronger around the implanted DMSO beads than RAPA beads (Fig. 4G1 and G2). VMHC is a marker for mature myocardium. This indicates that RAPA-induced autophagy negatively affects myocardial differentiation.

Ethanol and RAPA exposure interrupts cardiac precursor cell differentiation

It has been reported that interference with cardiomyocyte differentiation could induce cardiac malformation.³⁸ Hence, we examined whether ethanol and RAPA could affect cardiac precursor cell differentiation. Anterior primitive streak explants were cultured and photographed at 0-, 24-, 36- and 48-hour incubation. The results showed that both exposure of ethanol (Fig. 5B-B3) and RAPA (Fig. 5C-C3) could inhibit the extension area of cardiac precursor cells from anterior primitive streak explants compared with the control explants (Fig. 5A-A3, D; n = 6 for each group). We observed the cardiomyocyte that developed from the anterior primitive streak explants contracted spontaneously. We also found that exposures to both ethanol and RAPA decreased the rate of cardiomyocyte contraction in the explants compared with the control (n = 6 for each group; Fig. 5E). This difference in the cardiomyocyte contraction rates can be clearly seen in our real-time recording (supplementary movie).

Figure 5.

Ethanol and RAPA-treatment represses cell migration and contraction of cardiac progenitor cells in vitro. Anterior primitive streak tissues (excluding Hensen's node) were isolated from HH3 chick embryos and maintained for 24, 36 and 48 hours. (A-C3) RAPA- and ethanol-treatment repressed cardiac progenitor cell migration at all 3 time points analyzed compared with the control. (D) Graph showing the extent of cardiac progenitor cell migration in all 3 groups. (E and F) Graph and bar chart showing the rate of cardiac progenitor cell spontaneous contraction in all 3 groups. ***p<0.001 indicates significantly different between treated and untreated groups. Abbreviation: RAPA, rapamycin. Scale bars = 800 µm in A-C, A1-A3, B1-B3 C1-C3.

Since, ethanol and RAPA treatment affected the contractility of the cardiomyocytes, we decided to examine whether cardiomyocyte differentiation was also correspondingly affected. We immunofluorescently stained the anterior primitive streak explants with MF20 antibodies, a marker for myosin heavy chain, 3-days, 6-days and 12-days after culture (n = 6 for each group; Fig. 6). We determined that ethanol- and RAPA-treatment repressed cardiomyocyte differentiation and this may explain why these cardiomyocytes contracted less frequent than control cardiomyocytes.

Figure 6.

Ethanol and RAPA treatment inhibits cardiac progenitor cell differentiation. (A-C2) Anterior primitive streak tissues were cultured for 3 - 12 days and the cardiomyocytes that formed were immunofluorescently stained with MF20 antibody. The immunofluorescent staining revealed that ethanol- and RAPA-treatment inhibited cardiac progenitor cell differentiation (as indicated by the extent of myosin II expression). Abbreviation: RAPA, rapamycin. Scale bars = 100 µm in A-C, A1-A2, B1-B2 C1-C2.

RAPA-induced autophagy alters expression of genes associated with cardiac differentiation

We implanted beads loaded with DMSO (control) or RAPA into the gastrulating chick embryo during the formation of the heart tube. In situ hybridization revealed that the implanted RAPA beads inhibited GATA5 expression while the DMSO beads, implanted on the contralateral side, did not disrupt GATA5 expression (n = 6/8; Fig. 7A and B). This inhibitory effect is distinctly discernable in transverse sections of these embryos which show GATA5 expression was weaker in tissues near that RAPA beads than near DMSO control beads (Fig. 7B1-B2). We also examined the expression BMP2, BMP4 and GATA4, which are related to cardiac precursor cell dedifferentiation, by Semi-quantitative RT-PCR. RAPA exposure was determined to increase BMP2 expression but did not affect BMP4 expression. RAPA dramatically inhibited GATA4 expression (Fig. 7C and D). The results imply that RAPA-induced autophagy could significantly inhibit GATA4 and GATA5 expression during heart tube formation.

Figure 7.

RAPA bead implants downregulated GATA5 expression during cardiogenesis. The GATA5 In situ hybridization for GATA5 expression in embryos implanted with DMSO (control) and RAPA (40nM) impregnated beads. The beads were transplanted in the anterolateral regions of HH4 Hensen's node for 18 hours. (A-B) Representative appearance of GATA5 expression in the heart (A) and its higher magnification as indicated by square frame in A (B). (B1-B2:“) Transverse sections produced from dotted lines in B. The sections show DMSO did not affect GATA5 expression but RAPA repressed GATA5 expression. (C) Semi-quantative RT-PCR and corresponding bar chart showing the effects of RAPA beads on BMP2, BMP4 and GATA4 expression compared with control beads. Abbreviation: RAPA, rapamycin. Scale bars = 400 µm in A; 200 µm in B; 200 µm in B1-B2 and 100 µm in B1′-B1,“ B2′-B2.“

Ethanol disrupts expressions of key cardiogenic-associated genes

Semi-quantitative RT-PCR analysis revealed that genes associated with cardiomyocyte differentiation, BMP2, GATA4 and FGF8, were normally expressed in the embryonic heart tissues. However, following ethanol treatment, expression of these genes were significantly inhibited (Fig. 8A and B). We also examined the expression of heart tube fusion-related genes Wnt3A, VEGF and Slit2.³⁹ Here, we show ethanol exposure increased Wnt3A expression but repressed VEGF and Slit2 expressions.(Fig. 8C and D)

Figure 8.

Effects of ethanol on cardiomyocyte differentiation-related and heart tube fusion-related gene expression. (A and B) Semiquantative RT-PCR and corresponding bar chart showing BMP2, GATA4 and FGF8 (cardiomyocyte differentiation-related genes) expression in control and ethanol-treated embryos. (C and D) Semiquantative RT-PCR and corresponding bar chart showing Wnt3A, VEGF and Slit2 (heart tube fusion-related gene) in control and ethanol-treated embryos.

Ethanol causes excess ROS generation during heart tube formation

We have found that ROS generation was significantly increased after 30 minutes of ethanol exposure in H9c2 cells (control = 3088, ethanol = 4010, n = 3, p < 0.01; Fig. 9A-B). We used 2,2-azobis (2-amidinopropane) dihydrochloride (AAPH, a free radicals generator ^25,40) to establish whether excess ROS could also induce cardia bifida. The result showed that after 45 hours of AAPH exposure, some of the AAPH-treated embryos developed cardia bifida with 2 cavities ( 13.3%, n = 4/30, Fig. 9E-F1 and G), while control embryos developed C-looping heart tube with one single cavity (100%,n = 30/30, Fig. 9 C-D1′ and G).

Figure 9.

Ethanol exposure increases ROS generation in H9c2 cells. (A) H9c2 cells were detached, and intracellular ROS levels were then measured by flow cytometry analysis using 2′,7′-dichlorofluorescein diacetate (DCF-DA) after 30 minutes ethanol exposure. (B) The quantitative analysis for intracellular ROS levels in control and ethanol exposure groups. (C-D) The representive bright-field image of whole chick embryo (C) and MF20 immunofluorescent image of C-looping heart tube (D) respectively from control embryo. (D1-D1′) The representive transverse sections at the levels indicated by white dotted lines in D (D1) and high magnification from D1 (D1′) respectively. (E-F) The representive bright-field image of whole chick embryo (E) and MF20 immunofluorescent image of bifida heart tube (F) respectively from 13.8mM AAPH-exposed embryo. (F1-F1′) The representive transverse sections at the levels indicated by white dotted lines in F (F1) and high magnification from F1 (F1′) respectively. (G) Bar chart showing the incidence of cardia bifida (%) in control and ethanol-treated groups. Scale bars = 600 µm in C,E; 200 µm in D,F; 200 µm in D1,F1 and 100 µm in D1′,F1′.

Discussion

Exposure to harmful environmental factor is known to increase the risk of congenital heart defects from developing.⁴¹ However, most human pregnancies are unplanned and the pregnant woman has been subjected to potential teratogenic factors before she could take precautionary actions. The high risk period for birth defects in the nervous and cardiovascular systems is during the first few weeks of development when these systems are actively being developed. Ethanol is widely consumed in excess and presents a potential teratogen in pregnant women. Hence, it is important that we understand the mechanisms underlying ethanol-induced cardiac birth defects so that we could develop and implement precautionary measures.

Chick and mouse vertebrate models have been employed to investigate the effect of alcohol on embryo development. Using the early gastrulating chick model, we found ethanol exposure significantly increased the production of cardiac bifida in embryos. It has been proposed that excess ethanol inhibits essential nutrients from being absorbed by the embryo and also excess generation of ROS that are possible causes of ethanol-induced cardiac defects.⁴² However, we have reported that a normal autophagy process was dispensable for early heart tube formation.³³ Consequently, we examined whether or not autophagy was affected by ethanol exposure in the gastrulating embryo (developmental stage when heart tube formation is initiated). ATG7 is an important autophagy gene which we showed is normally expressed in the apical side of epiblast, hypoblast and some mesodermal cells of the chick gastrula. Ethanol treatment dramatically alters ATG7 expression in the gastrulating embryo which is similar with the RAPA-treated embryos,³² this result indicates that ethanol disrupts autophagy.

RAPA, which acts on the mTOR signaling pathway, has been shown to regulate the autophagy process through inhibiting the initiation of autophagy.⁴³ In this study, we revealed that was mTOR expressed in gastrulating HH4 and HH11 chick embryos and that mTOR expression was inhibited by rapamycin. To establish whether if mTOR was activated by ethanol through the ER stress pathway, we used tunicamycin and thapsigargin to activate ER stress, thereby influencing the mTOR/autophagy process. However, we did not find any cardia bifida being produced in both tunicamycin- and thapsigargin-treated embryos (Fig. S4). This suggests that mTOR activation induced by ethanol exposure was not attained through the ER stress pathway. These findings imply that ethanol exposure alters autophagic activities in the gastrulating chick embryos, which in turn affects the formation of the heart tube during this crucial period of cardiogenesis. It is obvious that merely altering ATG and mTOR expression is not sufficient to support our hypothesis that autophagy is involved in the ethanol-induced heart tube malformation. Hence we treated gastrulating chick with RAPA, an inducer of autophagy, and discovered that RAPA also can induce the cardiac bifida. Furthermore beads soaked with RAPA could inhibit VMHC, a marker for mature myocyte, expression in tissues surrounding the implanted bead. These findings suggest that perturbing normal autophagic functions might be one of the mechanisms that cause cardiac bifida to develop following ethanol exposure.

The cardiac precursor cells in the first heart field are derived from the anterior primitive streak of the gastrulating chick embryo. These precursor cells migrate anterior primitive streak toward to the heart field and initiate heart tube formation.¹¹ We cultured explants isolated from the anterior primitive streaks and maintained them in the presence of ethanol or RAPA. We demonstrated that both of these reagents inhibited the migration of cardiogenic precursor cells from the anterior primitive streak and also repressed the differentiation of these cells. This may explain why the cardiomyocytes that formed in the presence of ethanol or RAPA contracted less frequently than untreated cardiomyocytes.

Morphogenesis of the heart tube during embryo development relies on a precisely coordinated expression of cardiac-associated genes, such as GATA5.⁴⁴ We transplanted beads impregnated with RAPA to the heart tube forming site and revealed that enhanced autophagy inhibited GATA5 expression. GATA4 expression was also down-regulated by RAPA in heart tube tissue. GATA4 and GATA5 are zinc-finger transcription factors that dictate cardiogenesis. It has been reported that GATA4 knock-out mice produced bilateral heart tubes (cardia bifida) and the heart contained fewer cardiomyocytes. Likewise, GATA5 mutants also developed cardia bifida and over-expression of GATA5 induces ectopic Nkx-2.5 expression.¹⁶

Now, what we need to address is if ethanol exposure affected the expressions of those cardiomyocyte differentiation and heart tube fusion-related genes. We have determined that BMP2, GATA4 and FGF8 which are differentiation-related genes, were inhibited by ethanol. VEGF and Wnt3A (fusion-related genes) expression were downregulated and up-regulated, respectively. We proposed that it is the abnormal expression of these cardiac differentiation and fusion genes which are partly responsible for the ethanol-induced cardia bifida. Nevertheless, there are other important genes, such as Ercc6l, which is crucial in the induction of fetal alcohol syndrome including heart defect.⁴⁵ There is growing evidence that indicate the genetic background determine whether the embryo could adapt to the teratogenic effects of ethanol exposure.⁴⁶

We have also examined whether over production of ROS induced by ethanol-exposure was another causative factor of cardia bifid - since excessive O2 free radicals could induce apoptosis to impair organ development.^47-49 There are many reports in the literatures that show ethanol could stimulates ROS generation.^50-52 In this study, we also confirmed that ethanol could increase ROS which might in turn induce autophagy.⁵³ Furthermore, we demonstrated that exposure to AAPH (ROS generator) could induce cardia bifida to develop - similar to RAPA and ethanol treatment. Thapsigargin and Tunicamycin treatment could cause an increase of ROS in gastrulating chick embryos,⁵⁴ but the 2 compounds did not induce cardia bifida. There is no doubt that further experimentation is required to explore the precise molecular biological mechanism. In this context, autophagy may also play an important role in the pathogenesis of cardia bifida in AAPH-treated embryos. We have schematically summarized our findings in Figure. 10 and propose a model for how ethanol is able to induce the formation of cardia bifida.

Figure 10.

Hypothetic model depicting how ethanol exposure induces cardia bifida during heart tube formation.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Funding

This study was supported by NSFC grant (31401230); Science and Technology Planning Project of Guangdong Province (2014A020213008); Science and Technology Program of Guangzhou (201510010073); The Fundamental Research Funds for the Central Universities (21615421, 21614319); Chinese Medicine Science and Technology Program of Guangdong Province (20142035); China Postdoctoral Science Foundation (2015T80940, 2014M560694); Students Research Training Program Fund (201510559047, 141212010) and The Funds for Young Creative Talents of Higher Education in Guangdong Province (2014KQNCX026)

Author Contributions

S.L., G.W., L.G., W.L. and X.W. performed the experiments and collected the data; S.L., G.W., L.C., and X. Y. designed the study and analyzed the data; M.C., K.L. and X.Y. wrote manuscript.

Supplemental Material

Supplemental data for this article can be accessed on the publisher's website.