Embryonic mouse blood flow and oxygen correlate with early pancreatic differentiation

Abstract

The mammalian embryo represents a fundamental paradox in biology. Its location within the uterus, especially early during development when embryonic cardiovascular development and placental blood flow are not well-established, leads to an obligate hypoxic environment. Despite this hypoxia, the embryonic cells are able to undergo remarkable growth, morphogenesis, and differentiation. Recent evidence suggests that embryonic organ differentiation, including pancreatic β-cells, is tightly regulated by oxygen levels. Since a major determinant of oxygen tension in mammalian embryos after implantation is embryonic blood flow, here we used a novel survivable in utero intracardiac injection technique to deliver a vascular tracer to living mouse embryos. Once injected, the embryonic heart could be visualized to continue contracting normally, thereby distributing the tracer specifically only to those regions where embryonic blood was flowing. We found that the embryonic pancreas early in development shows a remarkable paucity of blood flow and that the presence of blood flow correlates with the differentiation state of the developing pancreatic epithelial cells in the region of the blood flow.

Figure 1. Embryonic pancreas is vascularized early during development

Whole-mount immunohistochemistry analyses of E12.5 (A), E14.5 (B), and E17.5 (C) pancreas using antibodies against Pecam reveals that it is highly vascularized. (D) Ultrasound backscatter image of an E12.5 embryo with the echo-bright needle tip in the heart. (E) After in utero injection of the embryonic heart (at E12.5) with the endothelial tracer FITC-tomato lectin, a whole embryo “angiogram” of perfused vessels is demonstrated. (F,G) Histologic sections of an E10.5 embryo demonstrate that there is scant perfusion with tomato lectin (TL) in the region of the dorsal pancreas (dp) and gut (g), but there are numerous unperfused vessels (Pecam⁺). The arrow in (G) indicates a glucagon-positive area in the dorsal pancreas that has significant perfusion. (H–M) At E12.5 there is a regionalization within the pancreas, with areas that receive detectable blood flow and that also show glucagon⁺ differentiation (K–M), and then other unperfused areas (H–J) that lack glucagon⁺ differentiation. (N–P) Perfusion of an early E14 pancreas is more diffuse compared with the regional perfusion at E12.5, but still only a subset of the Pecam⁺ vessels are perfused (O,P). Again, glucagon-positive cells are only in the vicinity of the tomato-lectin perfused vessels (P), whereas numerous E-cadherin⁺ (Ecad) cells are in the vicinity of unperfused vessels (N,O).

Figure 2. Pancreatic cell differentiation coincides with embryonic blood flow perfusion

(A–H) A transition appears to occur between early E14 (A–D) and late E14 (E–H) wherein the percentage of vessels receiving blood flow increases substantially such that nearly all vessels become tomato lectin⁺ (compare A and B with E and F, quantified in K). (I,J) Confocal images of an early E14 pancreas demonstrate the specific regionalization of endocrine cells (both insulin⁺ and glucagon⁺ cells) to the vicinity of perfused vessels (tomato lectin⁺, TL⁺). Those epithelial regions in which FITC-tomato lectin⁻/Pecam⁺ (non-perfused) vessels predominate were notably endocrine-negative. E-cadherin staining in I and J demonstrates the large area of unperfused pancreatic epithelium. I and J represent adjacent sections of the same tissue sample. (K) Comparative linear quantification of the Pecam and tomato-lectin areas demonstrating the increase that occurs in the percentage of vessels that are perfused over this period of transition from early to late e14 (error bars SEM, N=10 pancreases, P<.001). (L) Morphometric analysis of the distance between differentiating endocrine cells (α-cells or β cells) and the nearest perfused vessel versus undifferentiated E-cadherin⁺ cells (error bars SEM, N=100 cells, P<.001 between either alpha or beta cells and non-endocrine epithelial cells). TL – FITC-tomato lectin, E-cad – E-cadherin.

Figure 3. Oxygenation and blood flow in the developing pancreas

(A–C) Oxidized thiols were used to identify areas of higher tissue oxygenation. Regions with perfused FITC-tomato lectin⁺ (TL) vessels (A) showed higher levels of oxidized thiols (B,C). (D,E) Pancreatic epithelium, including glucagon-positive differentiated cells, adjacent to FITC-tomato lectin⁺ vessels also showed higher levels of oxidized thiols as compared to the surrounding embryonic pancreas. (F) An intensity “heat” map for oxidized thiol staining is a more sensitive method for assessing the gradient of tissue oxygenation, with decreasing oxygen tension seen with increasing distance from perfused vessels. TL – FITC-tomato lectin, Gluc – Glucagon, Thiols – Oxidized thiols.

Figure 4. Hypoxia allows maintained proliferation and an arrest of pancreatic differentiation

A comparison of pancreas explants grown for 7 days in 21% oxygen culture conditions (A,E,I) or in hypoxia (1% oxygen) (D,H,L). Proliferation (BrdU staining) is maintained in hypoxia (A compared with D), but differentiation is blunted (E compared with H, L). In order to mimic the in utero increase in pancreatic blood flow that we observed at E14.5, E11.5 pancreases were cultured in 1% oxygen for 3 days, followed by 4 days of 21% oxygen (B,F), or with the reverse sequence as a control (C,G). By reproducing the in vivo blood flow transition and culturing the pancreas explants in three days of hypoxia followed by four days of 21% oxygen (B,F) endocrine and exocrine differentiation were rescued to the same or higher levels as in 21% oxygen-cultured explants (I). Quantification of these data is shown in (I) (error bars are SEM, N=5 explants for each group, no difference between any glucagon-positive cell groups, including the number of glucagon cells present at harvest (data not shown), all insulin-positive groups differ with a p value of less than 0.05). Comparable numbers of neurogenin3⁺ cells in either hypoxic or 21% oxygen-grown explants harvested after 3 days (J,L) suggest that simple toxic effects of hypoxia on endocrine progenitors is not the reason for less endocrine cells. Cell differentiation in explants cultured under hypoxic conditions was impaired despite the presence of equivalent amounts of Pecam⁺ vascular endothelium (K,M). Ecad: E-cadherin, Ins: insulin, Glu: glucagon, Amyl: amylase, Ngn3: neurogenin3.