Ventricular myocardium development and the role of connexins in the human fetal heart

Abstract

The developmental timeline of the human heart remains elusive. The heart takes on its characteristic four chambered appearance by ~56 days gestational age (DGA). However, owing to the complexities (both technical and logistical) of exploring development in utero, we understand little of how the ventricular walls develop. To address this, we employed diffusion tensor magnetic resonance imaging to explore the architecture and tissue organization of the developing heart aged 95-143 DGA. We show that fractional anisotropy increases (from ~0.1 to ~0.5), diffusion coefficients decrease (from ~1 × 10^-3mm²/sec to ~0.4 × 10^-3mm²/sec), and fiber paths, extracted by tractography, increase linearly with gestation, indicative of the increasing organization of the ventricular myocytes. By 143 DGA, the developing heart has the classical helical organization observed in mature mammalian tissue. This was accompanied by an increase in connexin 43 and connexin 40 expression levels, suggesting their role in the development of the ventricular conduction system and that electrical propagation across the heart is facilitated in later gestation. Our findings highlight a key developmental window for the structural organization of the fetal heart.

Figure 1

Development of ventricular wall organization during gestation. Fractional anisotropy (A and B) and apparent diffusion coefficient (C and D) for the left (A and C) and right (B and D) ventricles for human fetal hearts aged 95–143 DGA. Fractional anisotropy (FA) is illustrated as a scalar value from 0 to 1. FA significantly increases within the left ventricle (A: p = 0.0007, R² = 0.43) and the right ventricle (B: p = 0.0002, R² = 0.49) with gestational age. Apparent diffusion coefficient (ADC) quantifies the average diffusion within the wall of each ventricle. It was found to significantly decrease in the left (C: p = 0.009, R² = 0.29) and right (D: p = 0.008, R² = 0.29) ventricle and with gestational age.

Figure 2

Tractography of ventricular wall fibers during gestation. (A) Cardiac fibers computed from DT-MRI derived ventricular fiber architecture, for fetal hearts aged 105, 110, 124, 128 and 143DGA, are interpreted visually with a coloring scheme based on the curve features of local tangent directions. The length (B) and volume (C) of the computed ventricular cardiac fibers for human fetal hearts were found to significantly correlate with gestation (p < 0.0001 for both length and volume; R² = 0.87 for length and R² = 0.94 for volume).

Figure 3

Quantification of ventricular wall cardiac fibers during gestation. (A) The fractional anisotropy (FA) of the computed cardiac fibers, derived from tractography performed on DT-MRI datasets, was found to significantly increase with gestational age (p = 0.0009, R² = 0.41). (B) Apparent diffusion coefficient (ADC) was also found to be significantly correlated with age (p = 0.0196, R² = 0.23). To further identify the extend of these changes, we investigated diffusions in the primary (C) and tertiary (D) orientation. We found that only the radial diffusion (RD; D) was significantly correlated with gestation (p = 0.011, R² = 0.27).

Figure 4

Connexin 43 expression, relative to GAPDH, in human fetal hearts aged 67–137 DGA. (A) Cx43 expression is significantly increased in the mid-second trimester (107–136 DGA) when compared to 67–73 DGA and 96 DGA. (B) Example Western blot of Cx43 expression across gestation, and the GAPDH loading control, which did not demonstrate developmental regulation. (**p = 0.001 and ***p = 0.0004).

Figure 5

Connexin 40 expression, relative to GAPDH, in human fetal hearts aged 67 to 137 DGA. (A) Cx40 expression is significantly increased in the mid-second trimester (107–136 DGA) compared to the first trimer (67–73 DGA). (B) Example Western blot of Cx40 expression across development, and the GAPDH loading control, which did not demonstrate developmental regulation. ***p = 0.001.

Figure 6

The absence of Connexin 43 expression in one human fetal heart aged 97 DGA. In the process of testing Cx43 expression in the fetal heart samples, we discovered that one heart aged 97 DGA lacked any detectable Cx43 expression (left, top, sample 3). We confirmed this across three separate gels. Despite the lack of Cx43, the GAPDH loading control produced a clear band (left, bottom). At comparative ages (e.g. 73–107 DGA) Cx43 expression was clearly detectable in all other samples (right, top; blots from Fig. 4).