Department of Obstetrics and Gynecology, Division of Obstetrics and Prenatal Medicine, Erasmus MC, University Medical Center, Rotterdam, The Netherlands
Department of Obstetrics and Gynecology, Division of Obstetrics and Prenatal Medicine, University Medical Center, Dr. Molewaterplein 50, Room Hs-523, 3015 GE Rotterdam, The Netherlands. Fax: +31-10-704-3532
Congenital heart defects (CHD) are among the most common congenital malformations in newborns (Botto and Correa, 2003). Epidemiological studies have shown that the periconceptional use of folic acid reduces the risk of CHD in the offspring (Botto et al., 2003) and reduces the birth prevalence rate with ∼20% (Hoffman and Kaplan, 2002; van Beynum et al., 2009). Low folate decreases the remethylation of homocysteine into methionine, resulting in a mild to moderate hyperhomocysteinemia (Kang et al., 1987). A meta-analysis has shown that maternal hyperhomocysteinemia is associated with an approximately three-fold increased risk of newborns with a congenital outflow tract defect (Verkleij-Hagoort et al., 2007). This association, as well as the association of homocysteine exposure and a wide range of other congenital malformations among which neural tube defects, have been substantiated in animal experimental models (van Mil et al., 2010). In the chicken embryo ventricular septal defects and conotruncal heart defects have been reported (Rosenquist et al., 1996; Boot et al., 2004). The malformations observed in the chicken embryo share the mutual involvement of neural crest cells. Cardiac neural crest cells are important in heart and pharyngeal arch artery development. A previous chicken embryo study has demonstrated that homocysteine treatment results in failure of the cardiac neural crest cells to undergo apoptosis with subsequent decreased myocardialization of the outflow tract (OFT), subarterial ventricular septal defects and increased volume of the OFT cushions at Hamburger and Hamilton (HH) (Hamburger and Hamilton, 1951) stage 35 (9 days incubation; ED 9) (Boot et al., 2004).
Above described defects, seen in the chicken embryo after homocysteine treatment, are comparable to those observed in chicken embryo studies in which the extra embryonic blood flow has been altered by venous clipping. The venous clip model consists of ligation of the right lateral vitelline vein with a microclip, and is used to alter hemodynamic parameters such as the intracardiac blood flow in the chicken embryo (Hogers et al., 1997). The short-term functional effects of venous clipping are acute reductions of all hemodynamic parameters, which are observed for up to 5 hr after clipping (Stekelenburg-de Vos et al., 2003). Moreover, at stage HH 24 (ED 4), venous clipping leads to altered heart function presented by changes in the “active” and “passive” filling component of the ventricle (Ursem et al., 2004). Eventually, these chicken embryos display OFT defects, in particular ventricular septal defects, valve anomalies and pharyngeal arch artery malformations (Hogers et al., 1999), comparable defects as seen after homocysteine treatment (Boot et al., 2004).
Thus far morphological effects of homocysteine have been reported when the heart is already septated and the valves are functional. Possible hemodynamic alterations induced by homocysteine affecting heart development of the chicken embryo, however, are unknown. Interestingly, the embryonic heart is functioning well before cardiogenesis is complete. Therefore, it is conceivable that early functional changes may influence cardiac morphogenesis. Because of the overlap in CHD seen after elevated homocysteine exposure in human and in chicken embryos, and in CHD reported in the venous clipping model, we hypothesized that homocysteine may cause alterations in early embryonic heart function, preceding structural malformations. Consequently, we are most interested in the effects of homocysteine on the hemodynamic parameters heart rate, peak velocity, velocity integrated over time and wave times, as a reflection of early embryonic chicken heart function.
To assess early heart function in animal experimental models, ultrasound biomicroscopy can be applied (Phoon, 2006). This high frequency ultrasound technique has already provided more insight into the early heart function of chicken embryos (Butcher et al., 2007; McQuinn et al., 2007). Previously, we have shown that it is possible to acquire specific Doppler flow velocity waveforms at different anatomic locations in the embryonic chicken heart and that these velocity waveforms are specific for different embryonic stages of development. Blood flow profiles were collected at HH stages 18, 21, and 23, comparable to humans at 5–8 weeks of gestation (Oosterbaan et al., 2009). The aim of this study was to investigate whether L-homocysteine thiolactone hydrochloride, administered in concentrations comparable to the mild to moderate hyperhomocysteinemia associated with CHD in human, induces hemodynamic alterations in the ventricle and OFT of the embryonic chicken heart at HH 21 (ED 3.5).
MATERIALS AND METHODS
Application of Homocysteine
Fertilized White Leghorn chicken eggs (Gallus gallus (L.)) (Drost Loosdrecht BV, Loosdrecht, Netherlands) were incubated blunt-end up at 37.5°C. A hole of 2 mm was made in the outer shell membrane at the site of the air space at 24 hr of incubation. At 24, 48, and 72 hr of incubation, a homocysteine solution or a sham solution was applied through the hole, on the inner shell membrane, using a micropipette. The homocysteine treated group (n = 36) received 50 μL 0.05 M of the more stable metabolite of homocysteine, L-homocysteine thiolactione hydrochloride (Sigma–Aldrich, Steinheim, Germany) in Locke's solution (0.94% NaCl, 0.045% KCl, 0.033% CaCl2 w/v in Milli-Q) per application. Thus per application the embryos received 2.5 μmol homocysteine. The sham (control) group (n = 37) received 50 μL Locke's solution only. After each application the hole was sealed with Scotch tape and the incubation was continued. Thirty minutes after the last application the eggs were placed horizontally in the incubator for further development until ED 3.5. This mode of application of L-homocysteine thiolactone hydrochloride has been described previously and resulted in structural CHD, in particular ventricular septal defects (Rosenquist et al., 1996).
At day 3.5 the eggs were taken out of the incubator, the embryo was localized and a window of ∼1 × 1 cm2 was made in the egg shell to allow ultrasound imaging. The embryo was staged by microscopic examination according to Hamburger and Hamilton (Hamburger and Hamilton, 1951) and HH 21 embryos were included. Embryos that showed visible bleeding were excluded, as well as embryos floating on their right side. In total 13 out of 36 homocysteine treated embryos and 13 out of 37 control embryos were suitable for further analysis. The window in the eggshell was covered by a thin piece of polyethylene foil on which warmed ultrasonic gel was deposited. Embryo temperature was measured with a thermometer inside the egg and maintained at 37°C ± 1°C with the help of a heated egg holder on a heated plate and a heat lamp.
Ultrasound biomicroscopy measurements were performed with a high frequency ultrasound system: the Vevo 770 (VisualSonics, Toronto, Ontario, Canada) equipped with a 55-MHz transducer (RMV 708) exhibiting an axial and lateral resolution of ∼30 and 75 μm, respectively (Zhou et al., 2002; McQuinn et al., 2007). The embryonic heart was imaged at a focal depth of 4.5 mm. The pulsed Doppler sample volume varied between 0.15 and 0.17 mm and the scanning depth ranged between 4.15 and 5.15 mm. The measurements were performed and described in detail previously by our research group (Oosterbaan et al., 2009). From each embryo velocity signals were recorded at three cardiac sites, in the primitive left ventricle (PLV), the primitive right ventricle (PRV), and the OFT (Fig. 1a). To acquire signals from these sites the transducer was outlined along the cross section from the apex to the OFT (Fig. 1b). Because chicken embryos are generally floating on their left side, these embryos were included to increase group homogeneity and limit the variation in the angle of insonation. To ensure accurate probe positioning B-mode ultrasound was used (Fig. 1c). Power Doppler mode was used to visualize the blood flow in the cross section through the heart (Fig. 1d). After identifying these flows, the Doppler sample volume was placed in the centre of the blood stream to obtain flow velocity waveform recordings. From the obtained velocity waveforms we determined the heart rate (bpm; beats per minute), peak velocities (cm s−1), velocities integrated over time (VTI; cm) and wave times (ms) (Fig. 2a). The Doppler angle was arbitrarily set to zero as flow direction could not be assessed accurately. Therefore, the Doppler velocities should be interpreted as relative velocities, which more precisely are the vectorial velocities in the direction of the ultrasound beam. From the cardiac cycle the “Passive” (P) wave, Phase I, and the “Active” (A) wave, Phase II of ventricular filling, and the ejection wave, Phase III of the cardiac cycle, were determined (Fig. 2a). Wave times were additionally calculated in percentages of total cycle time to find proportional changes of the three phases of the cardiac cycle between homocysteine and control embryos. For each embryo we analyzed three consecutive cardiac cycles. Embryo temperature was monitored and collected simultaneously with the Doppler data.
The data are presented as mean (standard deviation; SD). To determine heart rate and temperature variance within embryos during the experimental time span, analysis of variance (One-Way ANOVA) was performed. All other statistics were carried out using the independent-samples t test. Statistical significance was defined as a P value of <0.05. Calculations were performed with SPSS 15.0 (SPSS, Chicago, IL).
Standardization of the Measurements
Within both the homocysteine treated and the control group the variance as expressed by the SD of the three consecutively measured heart rates per embryo was small, 1.5 and 1.0 bpm, respectively. The mean heart rate of homocysteine treated embryos was 134 (22) bpm, which was significantly lower than the 150 (14) bpm in control embryos (Table 1). The SD of the temperature within the embryos was also very small, 0.04°C within homocysteine and 0.07°C within control embryos. The mean temperature between the homocysteine and the control group was not significantly different (Table 1).
Table 1. Comparison of general and specific hemodynamic parameters
Homocysteine treated N = 13; Mean (SD)
Control; N = 13; Mean (SD)
Comparison of general and specific hemodynamic parameters measured in the primitive left ventricle (PLV), the primitive right ventricle (PRV) and the outflow tract (OFT) in homocysteine treated and control chicken embryos. Includes velocity-time-integral (VTI; cm), peak velocity (cm s−1) and time (ms) measurements, as well as wave times in percentage of total cycle time.
Figure 2b shows flow velocity waveforms derived from the PLV and the PRV in homocysteine treated and control embryos. The blood flow in both study groups and primitive ventricles was unidirectional during all phases of the cardiac cycle.
The PLV showed a significantly smaller ejection VTI in homocysteine-treated embryos compared to control embryos (Table 1). In the PRV both A wave and ejection wave VTI were significantly decreased in homocysteine-treated embryos compared to controls. In the PLV, peak velocities were not significantly different between both study groups. In the PRV, P and A wave peak velocity were significantly lower in homocysteine-treated embryos than in control embryos. “Passive” wave time was significantly prolonged in the PLV of homocysteine treated embryos compared to controls. Although the proportions of “active” and ejection wave times were significantly decreased in both primitive ventricles in homocysteine-treated embryos compared to controls, the increase in “passive” wave times in percentages of total cycle time were not significantly different.
The outflow tract waveform consisted of one dominant forward peak in both homocysteine treated and control embryos (Fig. 2b). The outflow VTI, peak velocity and ejection time were similar in both study groups. Diastolic time, consisting of the wave times of Phase I plus Phase II, was significantly prolonged in homocysteine treated embryos compared to controls (Table 1). This resulted in a significantly increased proportion of diastolic time and a significantly decreased proportion of ejection time, as shown by the changes in percentages of total cycle time. Backflow in the OFT during ventricular filling was seen in 5 out of 13 homocysteine treated embryos, compared to 1 out of 13 control embryos.
Within homocysteine-treated and control embryos the same differences between the PLV and the PRV waveforms were observed. There was a significant decrease of peak velocities and VTIs of Phase I and II of the cardiac cycle, from PLV to PRV. The hemodynamic parameters of Phase III significantly increased from PLV to PRV.
In this chicken embryo study we have shown that application of low dose L-homocysteine thiolactone hydrochloride during cardiogenesis significantly affects hemodynamic parameters of early embryonic chicken heart function. In general, an inhibiting effect of homocysteine on hemodynamic parameters is described. Homocysteine also significantly reduces mean heart rate. These functional changes may precede the development of structural heart defects associated with homocysteine.
We have shown that homocysteine significantly reduces mean heart rate to 134 (22) bpm compared with 150 (14) bpm in controls. The mean heart rate in controls corresponds with the heart rate, 155 bpm, as has previously been described in stage HH 21 chicken embryos (Hu and Clark, 1989) and is lower than the 170 (25) bpm reported in our previous study (Oosterbaan et al., 2009). This can be explained as an effect of sham treatment, executed in this study. During the experiments temperature was controlled and temperature differences within and between embryos were extremely small, through which temperature can be excluded as a confounder of heart rates. We suggest that the reduction of heart rate in homocysteine embryos reflects a delay in development, as heart rate increases with developmental stage (Hu and Clark, 1989). Homocysteine has been shown to reduce the synthesis of several proteins (Malanovic et al., 2008) implicated in growth processes. Reduced somite number, embryonic size (Greene et al., 2003; Han et al., 2009) and mass (Miller et al., 2003) have been observed after homocysteine treatment in chicken and mouse embryo studies.
The differences in intracardiac blood flow patterns and in proportions of wave times in percentages of total cycle time between homocysteine-treated and control embryos demonstrate that there are functional differences between both study groups. The proportions of ventricular ejection as well as outflow ejection time are decreased in homocysteine embryos in contrast with the linear relationship between heart rate and electrical and mechanical systole described in normal mature human subjects (Boudoulas et al., 1981), implying that ejection time should increase with reduced mean heart rate. These effects exerted by homocysteine do not only reflect the impact of a delay in development. When comparing the hemodynamic parameters of homocysteine embryos in this study to parameters of sham-treated embryos of an earlier Hamburger and Hamilton stage, HH 18, the differences reported in this study remain to exist (A.M. Oosterbaan, unpublished data). Therefore we conclude that homocysteine causes direct functional changes of early embryonic chicken heart function that can not only be explained by developmental delay.
So how can homocysteine be linked to altered hemodynamics and eventually to the development of congenital heart defects? Several mechanisms of homocysteine have been described in literature that may potentially inhibit myocardial function and cause disturbed heart development. These mechanisms may occur either in a linked or independent way. For example, the effects homocysteine exerts on endothelial cells, causing endothelial cell damage and dysfunction (Wall et al., 1980; Lee et al., 2005). Also, homocysteine has been shown to derange the synthesis of nitric oxide (NO) (Upchurch et al., 1997; Welch et al., 1998) and induce oxidative stress (Tyagi et al., 2005), which may cause disturbed angiogenesis as shown by our research group and by Latacha and Rosenquist (Latacha and Rosenquist, 2005; Oosterbaan et al., 2011). Homocysteine exposure increases connexin43 levels in mouse neural crest cells, affecting neural crest cell migration (Boot et al., 2006). This increased expression of connexin43 has been shown to result in conotruncal heart defects in mice in vivo (Ewart et al., 1997). Boot et al. have also described a 60% reduction of apoptosis of the myocardial cells at the site of the OFT and a 25% increased volume of the OFT cushion tissue in chicken embryos treated with homocysteine (Boot et al., 2004). This may affect myocardial function.
The reported changes in intracardiac blood flow patterns after homocysteine exposure may lead to local changes in shear stress (Hogers et al., 1997). Changes in shear stress have been previously shown to result in changes in the expression of shear-responsive genes in the endocardial cells of the developing heart (Groenendijk et al., 2005). We suggest that the reduction in blood flow brought about by homocysteine in this study may affect genetic cascades, causing a differential regulation of genes implicated in hemodynamics, eventually resulting in altered cardiac morphogenesis.
Because we have approached the embryonic chicken heart in a standardized way, interpretation of the velocity measurements is warranted. Because we were not able to exactly determine the angle of insonation in a 2D image, angle variation was minimized by selecting embryos from the same HH stage, floating on the same side, and by using Power Doppler to obtain the highest velocity signal. The same amount of homocysteine and control embryos floated on the right side, therefore selection bias is limited. This selection, combined with the experimental procedure, may explain the 33% viable embryos in this study, compared to the 64 % survival rate reported by Afman et al. (2003) and the >85% survival in the study performed by Rosenquist et al. (1996). Additionally, survival rates may differ because of variations in nutritional factors between the used chicken strains, possibly altering the yolk sac composition (Afman et al., 2003).
The concentration of 0.05 M L-homocysteine thiolactone hydrochloride applicated to the embryos in ovo was half the concentration applicated by Rosenquist et al., causing neural tube defects and congenital heart defects in the chicken embryo (Rosenquist et al., 1996). Using this two-fold higher concentration in our study resulted in only 8% viable embryos. Rosenquist et al. determined blood homocysteine levels in the chicken embryo after application of 50 μL 0.1 M homocysteine solution (Rosenquist et al., 1996). Blood levels of 15–150 μmol L−1 were reported, comparable to the mild to moderate hyperhomocysteinemia in humans associated with neural tube defects and congenital heart defects (Steegers-Theunissen et al., 1991; Steegers-Theunissen et al., 1994; Verkleij-Hagoort et al., 2007). Although blood homocysteine levels were not measured in this study, we considered the administration of 50 μL 0.05 M L-homocysteine thiolactone hydrochloride adequate, as it is plausible that with this halved concentration homocysteine levels of >15 μmol per liter are achieved.
In conclusion, homocysteine exposure leads to altered hemodynamic parameters of early embryonic chicken heart function. Several mechanisms have been proposed through which homocysteine may cause these hemodynamic changes. We suggest that the described functional effects may contribute to the development of CHD defects through changes in shear stress and shear stress related genes, as seen before in venous clipping studies (Hogers et al., 1997). Further research should identify the differential effects of homocysteine on the proteins, genes and pathways involved in the hemodynamics, implicated in heart development.