Laboratory of Morphometry, Metabolism and Cardiovascular Disease, Biomedical Center, Institute of Biology, State University of Rio de Janeiro, Brazil
Laboratorio de Morfometria, Metabolismo e Doença Cardiovascular, Centro Biomedico, Instituto de Biologia, Universidade do Estado do Rio de Janeiro, Av 28 de Setembro 87 fds, Rio de Janeiro 20551-030, RJ, Brazil
Previously, researchers believed that coronary arteries (CAs) budded from the aorta during development in humans and other mammals (Hutchins et al., 1988; Wada et al., 2003; Tomanek, 2005). However, this theory of proximal CA development is inadequate, as it does not explain the known possible congenital abnormalities of the CA (Bogers et al., 1988). Evidence from quail (Bogers et al., 1989) and humans (Mandarim-de-Lacerda, 1990) has demonstrated that CAs do not grow out of the aorta but instead grow into the aorta from the peritruncal ring of the coronary arterial vasculature. In addition, it seems that CAs arise from the proepicardium, a transitory structure in the embryo that contacts and spreads over the developing heart to form its epithelial covering (epicardium) and several internal tissues (Mikawa and Gourdie, 1996; Perez-Pomares et al., 2002).
Nutritional programming is the process through which variation in the quality or quantity of nutrients consumed during pregnancy exerts permanent effects on the developing fetus. Programming during fetal development is considered an important risk factor for noncommunicable diseases that occur in adulthood, including coronary heart disease (CHD) and other disorders related to insulin resistance (Langley-Evans, 2009). This is a significant concern, as CHD is the single largest cause of death in developed countries and is one of the leading causes of disease burden in developing countries (Gupta, 2008; Gaziano et al., 2010).
This study was designed to investigate how maternal nutrition could influence the development of the CA in staged embryos of mice. We hypothesized that maternal protein restriction might affect the origin and the structure of the CA in the early stages of development. Thus, this influence could offer a critical explanation of the high incidence of CHD in adulthood.
MATERIALS AND METHODS
Samples and Procedures
All procedures were carried out in accordance with the Conventional Guidelines for Experimentation with Animals (NIH Publication No. 85-23, revised 1996). The Ethics Committee for Animal Experimentation of the State University of Rio de Janeiro (Rio de Janeiro, Brazil) approved the experimental procedures used in the study (protocol number CEA/249/2008).
C57BL/6 virgin females were mated with males according to a schedule designed for the purpose of knowing the moment of fecundation with an accuracy of 1 hr. The presence of a vaginal plug the next morning after mating was used for pregnancy diagnosis and was considered as Stage 1 [embryonic day 0.5 (E0.5) for mice]. Mothers were separated into two groups according to the chow fed: normal protein (NP) chow (containing 19% casein) or low protein (LP) chow (containing 5% casein).The diets are made by PragSolucoes (Jau, Sao Paulo, Brazil) in accordance with the American Institute of Nutrition's recommendations during pregnancy (AIN 93G; Reeves et al., 1993). The NP and LP diets were isocaloric and their difference occurred only in the amount of protein. To compensate for the difference in the energy mix and isocaloric diets maintaining, complex carbohydrates were only added in the form of cornstarch. Because the addition of fat and simple carbohydrate would represent other models of fetal programming, it was avoided (Table 1). The experimental diets were administered to the females from 2 weeks before mating until euthanasia.
Table 1. Composition of experimental diets
Normal protein (NP) chow and low-protein chow (LP) are prepared according to the AIN-93G.
Casein (>85% protein)
When the time came to obtain the staged embryos, the mothers were sacrificed, and the embryos were collected and measured immediately for the following: body mass and crown-rump length (from the top of the head to the bottom of the buttocks). Then, the heart was quickly dissected under stereomicroscopy with the incision of the aorta, pulmonary artery, cranial and caudal cava veins, and pulmonary veins, and the heart mass was measured for each embryo. Embryos were studied from Stage 16 to Stage 23, a period that corresponded to the postsomitic stages. In humans, this period corresponds to the second month of development when organogenesis occurs (staged according to Carnegie Institution of Washington; Butler and Juurlink, 1987). For light microscopy, hearts were fixed (4% phosphate buffered paraformaldehyde, pH 7.2), embedded in paraplast plus (Sigma-Aldrich, St. Louis, MO), serially sectioned (5-μm thick), and stained with hematoxylin and eosin. Sections were systematically scanned under microscopy for the identification and location of the coronary vasculature relative to the aorticopulmonary trunk and aortic sinus.
To examine the subepicardial plexuses observed in the embryos better, the paraplast was removed, heart sections were incubated overnight with antisera against von Willebrand factor (DakoCytomation, code A0082, Glostrup, Denmark) at 4°C. The von Willebrand factor is a method of immunostaining microvessel and is responsible for highlighting small, immature microvessels or single endothelial cells and to investigate the angiogenic endothelial precursors in the subepicardium. Then, sections were rinsed with phosphate-buffered saline (PBS). A biotinylated antibody (K0679; Universal DakoCytomation LSAB+Kit, Peroxidase, Glostrup, Denmark) was used as the secondary antibody; this secondary antibody was detected by a reaction with a horseradish peroxidase–streptavidin–biotin complex. A positive immunoreaction was identified after incubation with 3,3′-diaminobenzidine tetrachloride (K3466, DAB, DakoCytomation, Glostrup, Denmark) and counterstaining with Mayer hematoxylin.
Confocal Laser Scanning Microscopy
For immunofluorescence, antigen retrieval was accomplished using citrate buffer, pH 6.0, at 60°C for 20 min and blocked with ammonium chloride, glycine 2%, and PBS, pH 7.4. Heart sections were incubated with the primary monoclonal antibody anti-FLK1 (Santa Cruz Biotechnology, code sc-579, CA). Primary antibody was diluted 1:100 in blocking buffer [PBS/1% (bovine serum albumin (BSA)] and incubated overnight at 4°C. The vascular endothelial growth factor receptor 2 (VEGF-r2 or FLK1) informs about the major transduction signals for angiogenesis. Samples were incubated for 1 hr at room temperature with fluorochrome-conjugated secondary antibody and goat anti-mouse IgG 546 nm Alexa-labeled secondary antibody (Invitrogen, Molecular Probes, Carlsbad, CA), all diluted 1:100 in PBS, pH 7.4. After rinsing in PBS, the slides were mounted with 4′,6-diamidino-2-phenylindole (DAPI) nucleic acid stain and the SlowFade Antifade Kit (Invitrogen, Molecular Probes). Indirect immunofluorescence was viewed, and an image was acquired using confocal microscopy (Zeiss Confocal Laser Scanning Microscope—LSM 510 Meta, Germany). Negative controls were subjected to identical procedures with both fluorescent secondary antibodies, but there was no incubation with primary antibody.
Apoptosis was detected in situ with the Apoptag plus peroxidase kit (Apoptosis Detection Kit, Chemicon Int, Millipore, Billerica, MA) on paraplast-embedded sections. This detection kit is based on the TUNEL method, that is, modification of genomic DNA using deoxynucleotidyl transferase for the detection of positive cells via specific staining. Positive nuclei were visualized by 3,3′-diaminobenzidine tetrachloride (Gavrieli et al., 1992).
The numerical density per area of the apoptotic nuclei (QA[ap-n]) was estimated in the region adjacent to the aortic sinus of the embryos at Stage 18. The number of marked apoptotic nuclei into a frame of a known area, AT, was counted on five different nonconsecutive fields per embryo, five embryos per developmental stage. Then, QA[ap-n] was calculated as the ratio between the number of apoptotic nuclei (N[ap-n]) and AT (Mandarim-de-Lacerda et al., 2010).
Scanning Electron Microscopy
The embryonic hearts were also investigated under scanning electron microscopy (SEM) to verify the modifications in the aorticopulmonary trunk region. The hearts were fixed for 30 min at room temperature in 2.5% glutaraldehyde diluted in 0.1 M Na-cacodylate buffer, pH 7.2, and then postfixed for 30 min at room temperature in 1% osmium tetroxide. The hearts, after being washed in buffer, were dehydrated using an ascending acetone series, dried by the critical point method with CO2, mounted with silver cellotape on aluminum stubs, and coated with a 20-nm-thick layer of gold. The samples were examined with a Zeiss DSM 940 microscope.
Differences between the NP and LP groups were analyzed with an unpaired t test. The frequency differences of the CA origins were evaluated with a χ2 test. The growth of the heart and the body mass were plotted against the age of the embryos using a second order polynomial function. In addition, the regression analysis plotted heart mass (as dependent variable, y) against both the body mass and crown-rump length (as independent variable, x). This bivariate study used log-transformed data and the allometric model log y = log a + b log x (Jolicoeur and Heusner, 1986). Linear regressions were established for each analysis, and the regressions were compared (comparison of slopes). Pearson's coefficient of correlation was used to determine the significance of each regression. A P-value <0.05 was considered statistically significant (Prism version 5.04 for Windows, GraphPad Software, San Diego, CA).
Biometry and Allometry
The LP embryos had lower masses, and they had a lower cardiac mass than the NP embryos in all stages studied (Figs. 1 and 2). The heart mass in the LP embryos was 58% lower at Stage 16 (P < 0.001), 21% lower at Stage 20 (P < 0.001), and 29% lower at Stage 23 (P < 0.001). Consequently, the LP embryos showed greater allometric growth rates for the heart than the NP embryos for the period analyzed (P < 0.0001, t test; Table 2 and Fig. 3). Therefore, the LP embryos showed positive allometry for heart growth rates, whereas the NP embryos showed negative allometry for heart growth rates.
Table 2. Linear regressions with comparative growth rates (slopes) using the allometric model (log y = log a + b log x)
Comparison of the slopes
The condition of the isometry for each analysis is indicated and serves to compare with the slopes of the regressions.
Abbreviations: CRL, crow-rump length; BM, body mass; HM, heart mass; NP, normal protein offspring; LP, low protein offspring; P, probability; SE, standard error of the mean; and R, Pearson's coefficient of correlation.
CRL (mm, log)
HM (mg, log)
BM (mg, log)
Development of Coronary Arteries
Table 3 summarizes the findings in the heart of the LP and NP groups of embryos. Whereas subepicardial plexuses were observed in Stages 16 and 17 in the NP embryos, they were observed in Stages 17 and 18 in LP embryos. No CA was observed on both sides of the peritruncal region at these stages.
Table 3. Frequency of the CA developmental findings in staged (S) C57BL/6 mice
n.o.: not observed.
*P < 0.05 when compared with embryos from normal protein group at same stage.
The right and left CA were simultaneously observed in more than half of the specimens starting at Stage 18 in the NP embryos and in less than half of the specimens in the LP embryos starting at Stage 19. From Stage 20 and on, both CAs were present in the NP embryos; by Stage 22 and later, they were present in the LP embryos (Fig. 4). The differences in the frequency of appearance of the CA were significant (at Stage 18, left CA: χ2 = 6.32, P < 0.01; right CA: χ2 = 7.00, P < 0.01). Beginning with Stage 20 in the NP embryos and Stage 22 in the LP embryos, both the right and left CAs were seen in all specimens. No more than one coronary orifice was observed on the aortic sinus in all groups studied. In addition, at the end of the postsomitic period, the CA was identified with its characteristic relief using SEM (Fig. 4k).
Detection of Apoptosis and FLK1+ Cells
Positive apoptotic nuclei were observed in all groups at Stage 18 (Fig. 4). However, the QA[ap-n] was more than 130% greater in the NP embryos (402.2 ± 18.6 nuclei/mm2) than in the LP embryos (169.2 ± 12.8 nuclei/mm2; t test, P < 0.001). FLK1+ cells were observed at Stage 16, usually with a stronger reaction in the NP embryos than in the LP embryos. Even at later stages, FLK1+ cells showed weaker reactions in the LP embryos than in NP embryos of the same age. In addition, FLK1+ cells were homogeneously distributed in the ventricles of the NP embryos as early as Stage 18 and later, whereas the FLK1+ cells were better expressed only at Stages 22 and 23 in the LP embryos (Fig. 5).
The present study was interested in analyzing the chronology and nature of the structure related to CA development in C57BL/6 mice, as well as how severe maternal protein restriction would affect this development. Indeed, protein restriction during pregnancy significantly delayed the appearance of both the CAs. The LP offspring had a smaller heart than the NP offspring during the postsomitic period; however, the LP offspring had greater growth rates.
As widely discussed in the literature, programming by protein restriction entails a vast series of biochemical, endocrine, and molecular changes that together can produce adverse effects in adult life (Barker, 2002; Langley-Evans, 2009; Simmons, 2009). When pups from maternal protein restriction have normal nutrition or overnutrition after birth they usually show a compensatory growth in weight and length. This growth restores the animal's body size but may have long-term effects, which include a reduced lifespan (Metcalfe and Monaghan, 2001; Ozanne and Hales, 2004).
Maternal protein restriction impairs various organ functions due to an important reduction of functional cells such as beta cells in pancreas (Boujendar et al., 2002), glomeruli in kidney (Villar-Martini et al., 2009), and cardiomyocytes in heart (Bezerra et al., 2008). Therefore, it programs to Type 2 diabetes, hypertension, and heart failure later in life. In the present work, the postnatal life of the pups was not studied to observe the CA structure and function, as that opens the opportunity of further studies.
We also investigated the role of the FLK1+ cells in the CA development in this mice model of programming. The FLK1+ cells derived from embryonic stem cells can differentiate into both endothelial and mural cells and can reproduce the vascular organization process. It is through the FLK1 receptor that VEGF induces the reprogramming of epithelial progenitor cells (Kang et al., 2010), making the FLK1+ cells markers of latent vasculature during development (Yamashita et al., 2000). Although the mechanisms underlying the role of FLK1 morphogenesis in the CA are still unknown, at the CA origins, the epicardial and subepicardial cells stain more intensely for VEGF (Tomanek et al., 2002). This suggests a role for VEGF and its receptors in the formation of the CA stems, as interaction between endothelial cells and mural cells (pericytes and vascular smooth muscle) is essential for vascular development and maintenance (Munoz-Chapuli et al., 2002; Tomanek et al., 2006). This evidence that VEGF receptors contribute in some way to the CA formation and consequently to a better heart function has already been described in other animal models such as quail (Yue and Tomanek, 2001), rats (Zheng et al., 2001), rabbits (Wright, 2002; Nash et al., 2006; Wafai et al., 2009), and bovine (Pepper et al., 1998). In the present study, the FLK1+ cells were seen in earlier stages in the NP embryos, but only in later stages in the LP embryos, supporting the hypothesis that the FLK1+ cells are involved in the CA development in this animal model.
Another question considered in the present study was the significance of apoptosis around the region of origin of CA. Apoptosis usually plays a role in the process of structural remodeling (Velkey and Bernanke, 2001), for example, transforming the endocardial cushions into mature cardiac structures in the embryo (Person et al., 2005). In the present study, apoptotic nuclei were detected near the future site of the aortic coronary orifice in earlier stages in the NP embryos than in the LP embryos. This allows the supposition that the delay of peritruncal area remodeling in the LP embryos could be relevant to the manifestation of CHD later in life. A relationship between apoptosis and coronary ingrowth has been reported (Velkey and Bernanke, 2001), but any specific regulatory mechanism acting during CA development was described leaving avenues for future research. Factors involved in regulating apoptosis should be included in future models of CA development (Bernanke and Velkey, 2002).
In summary, maternal protein restriction in mice leads to a delay in the growth of the heart in the embryonic period modifying the development of the subepicardial peritruncal plexus and the apoptosis in the future coronary orifice region.
The authors thank Mrs. Thatiany Marinho, Patricia Martins, and Angelica Figueiredo for their technical assistance.