To Marianne Bronner-Fraser, who has been insightful and thoroughly supported me throughout my years here. To Mory Gharib, who always made me feel that my research was exciting and important. To Melanie Martin, who has now left the Fraser lab, for singing Canadian children's songs in the lab with me and sharing the Seder with me every year.
I would also like to thank the technical staff at the lab and at Caltech in general. To Chris Waters, who put up with my mood swings and ignored me when I stole the lab equipment. To Gary Belford, who has calmed me down when everything around me was crashing, or at least all the computers.
To James, who has been my shoulder to cry on when things have been bad and who has forgivingly accepted all the bullshit I've thrown his way. Many genes known to be regulated by shear stress are important for vascular remodeling in the embryo.
Introduction
Hematopoietic and endothelial precursors are formed in the mesodermal tissue of the embryo that receives signals from the endoderm [4]. One of the open questions in the field is how endothelial cells can sense the presence of shear stress. This calculation requires knowledge of the velocity profile of the blood vessels and the hematocrit or packing density of the red blood cells.
Blood flow patterns in the heart and dorsal aorta of the developing mouse embryo were also investigated using ultrasound biomicroscopy. Eaves, Properties of the earliest clonogenic hemopoietic precursors that appear in the developing mouse yolk sac. Ackerman, A phase and electron microscopic study of vasculogenesis and erythropoiesis in the yolk sac of the mouse.
Chapman, W.B., The effect of the heart rate on the development of the vascular system in the chick. Both endothelial and hematopoietic cells begin as a tight cluster in the proximal portion of the yolk sac.
Dynamic In Vivo Imaging of Post-Implantation Mammalian Embryos
For 8.5-dpc embryos, the ability of the embryos to rotate was recorded and compared with in vivo and roller culture results. Embryos cultured in the presence of 10 µg/mL transferrin had significantly lower heart rates (mean of 40 beats per minute or bpm, . Chapter 2 - Imaging Mammalian Embryos Using Embryo Culture 43 n=12 for 8.5-dpc, n =6 for 9.5-dpc) towards the end of the culture period. Therefore, the yolk sac was removed for 9.5-dpc embryos, while younger embryos with intact yolk sacs were cultured.
Evaporation causes the yolk sacs of 8.5-dpc embryos to wrinkle and stop circulation in the yolk sac. Embryos grown for 24 h in culture (Figure 2, a–b) were compared with freshly hatched 9.5-dpc embryos (Figure 2, c). Another major event in the development of 8.5-dpc embryos is axial rotation or rotation (Figure 6, shown in Supplemental Video 2).
The yolk sac of an 8.5-dpc embryo was removed for neural tube closure imaging. Somite numbers of 9.5-dpc embryos in culture were counted (solid line) at t=0, 6, and 12 h and compared to in vivo rates (dashed line).
Measuring Hemodynamic Changes During Mammalian
We focused our studies on yolk sac blood flow in the developing early mouse embryo. To help visualize yolk sac blood flow, whole mouse embryos expressing GFP in primitive erythroblasts [24] were grown in culture on a confocal laser scanning microscope stage [25] . Line scanning perpendicular to the vessel was used to obtain flow profiles from yolk sac vessels.
Embryos lacking VEGFR2/Flk-1 do not form vessels in the yolk sac [38], but VEGF signaling may be required at later stages of hematopoietic and vascular development, as it remains expressed in endothelial cells throughout embryonic development [39]. Our data show that levels of shear stress in developing vessels in the yolk sac are similar to those shown in vitro to regulate key factors required for remodeling. In the first method, the average of the top 2% of all measured speeds for the vessel was taken as Vmax.
The scan through the center of the vessel was used as representative of the flow present throughout the vessel, since unperturbed laminar flow is parabolic (Figure 4). Turnbull, 40 MHz Doppler characterization of umbilical and dorsal aortic blood flow in the early mouse embryo. We have shown that both morphological changes in the vasculature and the appearance of molecular markers of remodeling correlate with periods of high shear stress.
An important step in the development of the embryo is the remodeling of the yolk sac capillary plexus. By measuring the levels of shear stress in the early yolk sac, we show that endothelial cells are exposed to a transient increase in shear stress shortly after erythroblasts begin to circulate. We therefore measured shear stress in the embryonic vessels using confocal line scanning as described in Jones et al.
Before this time, ZO-1 staining can be seen in cuboidal endoderm cells in the yolk sac (Figure 5A). Our data are consistent with the presence of hemogenic endothelium in the early yolk sac and embryo. Shear stress levels are highest when the capillary plexus initiates remodeling and correlates strongly with morphological change in the plexus.
Thus, it is possible that the proliferation of Flk-1+ cells in the yolk sac is dependent on shear stress signals. Here we have shown that α-SMA-positive cells appear in the yolk sac with 11 somites, at times when shear stress levels reach a transient peak.
Hemodynamic Analysis of MLC2a Knockout Phenotype
The presence of increased oscillatory flow was not found to be the main cause of the phenotype of MLC2a-/- embryos since the heart was weakened. Although erythroblasts move with blood flow in mutant embryos, net displacement within a cardiac cycle is limited due to the large amount of retrograde flow. These results show that the phenotype after Ncx inhibitor treatment is similar to that of MLC2a-/- embryos.
Blood island polymerization results in most erythroblasts remaining fixed at the proximal end of the yolk sac even at the 14-somite stage (Figure 6D), when most erythroblasts are circulating in wild-type embryos (Figure 6A). The results of focal polymerization experiments show that some aspects of vascular remodeling are present without. In MLC2a-/- mice, this process does not occur and the vessels of the plexus remain dilated.
Erythroblasts carry oxygen to tissues and play a role in the hemodynamics of flowing blood. In MLC2a-/- embryos, there does not appear to be a net forward movement of circulating erythroblasts. Embryos were collected on the morning of day eight and cultured as previously described [ 32 ].
After culture, heart rate was measured and the expansion of the blood islands was imaged on a fluorescent dissecting microscope. Embryos were then injected with TEMED directly into the blood islands of the yolk sac. Several focal injections were performed to polymerize the entire circumference of the blood islands.
After culture, the expansion of blood islets was visualized on a fluorescence dissecting microscope using the GFP marker. The expansion of erythroblasts from blood islets was followed using a transgenic mouse expressing GFP in its erythroblasts [21]. Erythroblasts enter the field of view within individual frames, indicating the presence of flow.
Conclusions
The work described here focuses on relating quantitative differences in shear stress to the response of neighboring endothelial cells. Biologists have many tools available to understand changes in gene and protein expression within cells. The first step to understanding the role of mechanical forces in biology is therefore to find methods to measure these forces.
The ability to measure blood flow velocity and shear stress in mammalian embryos provides one such technique ([1], Chapter 3). This work allowed velocity profiles and shear stresses to be calculated for the first time within developing mammalian embryos and established that shear stress levels were within the range known to activate cell signaling. This work then went on to look at changes in vascular morphology during embryonic development in the context of the changes in blood flow that occur (Chapter 4).
This work proved that the onset of vascular remodeling coincides with the entry of erythroblasts into the circulation; an event that is preceded by a significant period of plasma flow. The presence of this plasma flow limited the possible models whose events could be dependent on fluid flow, specifically ruling out the possibility that paracrine signaling factors were blood-borne to initiate remodeling events. This was done using MLC2a mutant mice, which lack an atrial contractility in the heart, and specific chemical techniques to reduce shear stress (Chapter 5).
We found that when erythroblasts were prevented from entering the circulation, proper vascular remodeling was inhibited, even if the heart was normal. Other problems associated with the MLC2a-/- embryos, such as increased oscillatory flow, were secondary to problems associated with low shear stress flow. This work showed that changes in viscosity due to the entry of erythroblasts into circulation are essential for the formation of large vessels.
Thus, by dissecting abnormal flow patterns, we begin to understand how blood flow, and the mechanical forces it imparts, are involved in normal cardiovascular development. The shear stress exerted on the blood vessels during development is dependent on several factors, including the velocity profile of the blood, the viscosity of the blood and the pulsatility of the flow, all of which change during development. Although this makes it difficult to understand the role of fluid forces during development, it also gives us several parameters to adjust to change the shear force on the early blood vessels.
