SEARCH

SEARCH BY CITATION

Keywords:

  • yolk sac endoderm;
  • chicken;
  • vasculature;
  • nutrition;
  • early development

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

An important function of the vascular system is nutrient delivery. In adult animals, this is mediated through a close contact of the mesoderm-derived vasculature with the endoderm-derived enterocytes and hepatocytes. During embryonic development, the yolk sac (YS) endoderm has been suggested to play a similar role. Physiological and molecular nature of the contact between the YS endoderm and the vasculature is not well-understood. To understand roles of the YS endoderm in early development, we used the avian model and carried out a gene expression profiling analysis of isolated area vasculosa YS endoderm tissues from embryonic day 2–4 chick embryos, covering the first 48 hr of postcirculation development. Genes involved in lipid metabolism are highly enriched, indicating an active modification of lipid components during their transfer from the yolk to the circulatory system. We also uncovered genes encoding major serum proteins and key regulators of vascular integrity. In particular, PTGDS, an enzyme controlling the last step of prostaglandin D2 production, shows high expression in the YS endoderm. Experimental introduction of prostaglandin D2 into embryonic circulation led to intraembryonic vessel rupture. These data suggest that the YS endoderm is the major, if not exclusive, source of lipid and protein constituents of the early embryonic serum and plays an important role in the regulation of vascular integrity in developing embryo. Developmental Dynamics 240:2002–2010, 2011. © 2011 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

During avian embryogenesis, the yolk sac (YS) vasculature supplies developing intraembryonic tissues with hematopoietic cells and nutritive substances. The hematopoietic cells, together with the YS vascular endothelial cells and smooth muscle cells, are derived from the extraembryonic mesoderm (Nakazawa et al.,2006; Shin et al.,2009); whereas nutrients are transferred indirectly from the yolk through YS endoderm cells (Mobbs and McMillan,1981; Speake et al.,1998). How the YS endoderm takes up, processes, and delivers yolk materials to the vasculature is poorly understood. Despite being of only transitory use, several interesting features of the YS endoderm underscore the importance of studying this tissue in more detail. For instance, it is capable of, when provided with proper mesoderm cues, differentiating or transdifferentiating into definitive endoderm cell types (Masui,1981,1982). The close relationship between the YS endoderm and the mesoderm-derived YS vasculature is reminiscent of that in the intestine and liver, and by analogy the YS endoderm may perform important functions, other than as a passive supply of nutrients, in the global regulation of embryonic growth.

The YS endoderm is a part of the YS membrane which consists of derivatives of all three germ layers (Sheng,2010). The YS ectoderm faces the albumen and vitelline membrane, and is continuous with the embryonic ectoderm. The YS mesoderm contains a somatopleural layer (mainly smooth muscle cells) and a splanchnopleural layer (smooth muscle, vascular endothelial, and blood cells) separated by the extraembryonic coelomic cavity. The YS endoderm is a stereotypic epithelial sheet (Fig. 1A–D) and has its apical membrane facing the yolk and its basal membrane tightly apposed to the splanchnopleural components of the YS mesoderm (Mobbs and McMillan,1979,1981; Yoshizaki et al.,2004). The nucleus of a YS endoderm cell is localized basally (Fig. 1A,D), and its cytoplasmic space is filled predominantly with yolk drops (Fig. 1A,B; Mobbs and McMillan,1979,1981). A major role of the YS endoderm is the uptake of lipids and other yolk constituents and the transfer of these components into embryonic circulation (Bellairs,1964; Mobbs and McMillan,1981; Speake et al.,1998).

thumbnail image

Figure 1. Cell biological organization of the yolk sac (YS) endoderm in the area vasculosa. A: Embryonic day (E) 2 YS endoderm in the area vasculosa, stained for E-cadherin (green) and nuclei (cyan). The vasculature is located on top. Endoderm cell boundaries are demarcated by E-cadherin expression. Endoderm cell nuclei are usually found close to the basal surface. B: E3 YS endoderm in the area vasculosa stained for E-cadherin (brown). The endoderm maintains its epithelial organization, although folds protruding into the yolk space start to form. C: A schematic diagram of the organization of the YS membrane, showing the endoderm layer and the splanchnopleural mesoderm (omitting the ectoderm, somatopleural mesoderm and blood cells inside the vasculature). EC, endothelial cells; SMC, smooth muscle cells; BM, basement membrane of the YS endoderm. D: A schematic diagram of a YS endoderm cell, with its apical side facing the yolk and basal side facing the vasculature. Its nucleus (N) is located basally, and its apical/basolateral membranes segregated by tight junctions (TJ).

Download figure to PowerPoint

The composition of the chicken yolk, especially its lipid constituents, has been described previously (Bellairs,1964; Noble,1991; Speake et al.,1998). This is briefly summarized here. Lipids comprise two thirds of the yolk dry weight. Over 90% of all lipids are located in the very low density lipoprotein (VLDL) fraction of the yolk, with the rest being in the more soluble lipovitellin-phosvitin protein-rich fraction. Two thirds of all lipids are in the form of triacylglycerides, one fourth as phospholipids, and the remaining as free cholesterol and cholesterol esters. The fatty acid composition of yolk lipids is similar to that found in the serum VLDL of laying hen, but very different from that in the embryonic serum or liver. Free cholesterol of the yolk is rapidly esterified to cholesterol ester following uptake by the YS endoderm. Levels of long chain polyunsaturated fatty acids, arachidonic acid (20:4n-6) and docosahexaenoic acid (22:6n-3), increase dramatically after the transfer. These data suggest an active and extensive remodeling during lipid transfer. In addition to its role as a digestive organ similar to that of the small intestine, the YS endoderm has also been proposed to function as a glycogen store (Bellairs,1964; Willier,1968). Although the extent is unclear, the YS endoderm has been shown in addition to contribute to the production of the embryonic serum proteins (Amin,1961; Kram and Klein,1976; Young et al.,1980; Young and Klein,1983). These observations suggest that the YS endoderm can also perform the functions of an adult liver.

In contrast to its physiological importance, virtually no molecular study has been reported for the YS endoderm. To gain a molecular understanding of the physiological roles played by this important tissue layer, especially during early embryonic development, we performed an expression profiling analysis of the YS endoderm isolated from the area vasculosa region of embryonic day (E) 2–E4 chick embryos, corresponding to the first 2 days of postcirculation development.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Tissue Preparation for Microarray Analysis

YS endoderm tissues were prepared as shown in Figure 2. Chicken eggs were incubated in ovo to embryonic day (E) 2, E3, or E4. Embryonic circulation starts at E2 (Hamburger and Hamilton stage [HH] 13–HH14). Growth reaches the early limb-bud stage (HH18–HH19) at E3 and mid limb-bud stage (HH24–HH25) at E4. YS tissues from E2–E4 embryos (Fig. 2A,C) were washed in phosphate buffered saline (PBS), and treated with dispase for 5 min at room temperature at a concentration range of 0.8–2.4 U/ml depending on the stage. Separation of layers was performed in PBS after dispase treatment. Starting from the sinus terminalis, the lateral border of the area vasculosa (Fig. 2B–D), the ectoderm and mesoderm layers of the area vasculosa were peeled off, leaving the YS endoderm as a fragile but still intact layer (Fig. 2E,F). Care was taken to exclude the endoderm from the more proximal area pellucida and more lateral area vitellina. Duplicate samples were prepared for each stage. In addition, tissues from the area opaca of E1 (HH7–HH8) embryos (Fig. 2A) were used as a control without layer separation. After RNA extraction, 5 μg of total RNA from E1–E4 samples were used for gene expression profiling on Affymetrix Chicken Genome Arrays. Assuming that nutritional needs for embryonic growth would increase after the establishment of circulation, we expected that genes involved in YS endoderm functions should be enriched at E2 compared with E1, and that these genes should exhibit a further increase or stay highly expressed at E3 and E4.

thumbnail image

Figure 2. Isolation of the yolk sac (YS) endoderm. A: Region (orange) of embryonic day (E) 2–E4 embryos used for endoderm isolation. The area vitellina is omitted in this drawing. For control E1 tissues, the entire area opaca (orange) is used. B: A schematic diagram of how the ectoderm and mesoderm tissues are peeled off during endoderm isolation. C: A piece of the YS with all three germ layers. The sinus terminalis, the lateral border of the area vasculosa, is indicated by arrows. D: A magnified view of a region in C. E,F: Endoderm remains as a sheet after the ectoderm and mesoderm being peeled off.

Download figure to PowerPoint

Overview of the Microarray Data

The raw datasets have been deposited in NCBI GEO with the accession number GSE27958. The Affymetrix Chicken Genome Array contains 38.5K probe sets, of which 26.3K probe sets have 14.9K unique GeneIDs and 12.2K probe sets have no GeneIDs. Approximately 43–52% of all probe sets in each of the analyzed samples had a present call. This range of total expressed genes judged by present calls is similar to what was observed in our transcriptome analysis of streak tissues reported previously (Alev et al.,2010). Several criteria were applied successively to narrow down genes enriched in the YS endoderm samples. First, we selected in total 985 probe sets with present calls in all E2–E4 samples and with an average expression value in the E2, E3, or E4 duplicate samples being greater than 2.5-fold of that in the E1 samples. We then removed from this initial list those with no geneIDs (200 probe sets) and with duplicate geneIDs (176 probe sets). An additional 20 genes were removed after comparison with the transcriptome profile of purified E4 circulating blood cells (McIntyre et al.,2008). This was likely due to a minor contamination of blood cells during YS tissue preparation. The remaining 589 genes were subject to intra-probeset correlation analysis using the microarray data analysis software xIntegrator (www.cdb.riken.go.jp/scb/MartinG/html/exintegrator.html; and our previous publications; McIntyre et al.,2008; Nakazawa et al.,2009; Alev et al.,2010). This removed further 211 genes showing large variations and poor correlation among different probes of a given probe set. Genes removed from the list after xIntegrator analysis may still contain ones important for YS endoderm function, but those were not followed up in this study. A final list of 378 remaining genes is shown in Supp. Table S1, which is available online. As an example, the xIntegrator analysis profiles of 20 genes are shown in Supp. Figure S1.

Categories of YS Endoderm Enriched Genes

To gain an understanding of what biological activities are significantly associated with these enriched genes, we performed a gene ontology (GO) analysis using the g:Profiler software (Reimand et al.,2007). GO annotation of chicken genes is relatively poor, we therefore used the human GO annotation and analyzed biological processes that are significantly associated with human orthologues of the enriched chicken genes. In total, 323 of the 378 genes could be annotated this way (Supp. Table S2 for all significantly associated biological processes). Not surprisingly, over 60% of the annotated genes were associated with metabolic processes. Constituent genes associated with lipid (54), carbohydrate (32), and amino acid (35) metabolisms are shown in Supp. Table S2. A few other biological processes were also highly represented. These include body fluid regulation, blood coagulation, nutrient sensing, and liver function (Supp. Table S2), suggesting that the YS endoderm also plays important roles in the global regulation of embryonic development.

The GO profile of highly associated biological processes was confirmed by manual analysis of constituent genes. One example is provided here. Among genes involved in lipid metabolism, we found many key proteins required for binding (e.g., fatty acid binding protein FABP2, vitamin D binding protein GC, and riboflavin binding protein RBP), breakdown (e.g., phospholipases PLA1A, PLA2G12B, and FAAH), biosynthesis (e.g., monoacylglycerol acyltransferase MOGAT1, long chain acyl-CoA synthetase ACSL5, and lecithin-cholesterol acyltransferase LCAT), and export (e.g., apolipoproteins ApoA4, A5, B, and H) of lipids. In addition, of particular interest to us, many enriched genes appear to encode serum proteins, suggesting a non–cell-autonomous role of the YS endoderm. The protein composition of the earliest serum is not known. The entire YS membrane has been shown to be capable of synthesizing some serum proteins (Kram and Klein,1976) at peri-circulation stages; and later during embryogenesis, it has been suggested that the YS endoderm participates in the synthesis and secretion of transferrin, alpha globulins and albumin into the circulation (Amin,1961; Kram and Klein,1976; Young et al.,1980; Young and Klein,1983). In our screen, we detected many enriched genes with a clear functional implication as constituent serum proteins, and many minor ones which may be involved in growth or hormonal regulations. Among the major proteins, prealbumin, alpha-2-macroglobulin, and transferrin were highly expressed, confirming these earlier studies. The apolipoproteins, ApoA4, ApoA5, ApoB, and ApoH, constitute another group of genes with high and enriched expression in E2–E4 YS endoderm. ApoA1 is highly expressed in both E1 and E2–E4 from our gene chip and in situ data, confirming the data from a previous report (Bertocchini and Stern,2008). A large group of additional enriched genes appear to play a role in the embryonic vasculature. These include complement proteins CF1, C1R, C5, C8A, and C8G; coagulation regulators F2, F5, F7, TFPI, PROC, and PROZ; as well as several antitrypsins, antithrombin, COL3A1, COL6A1, FGA, FGG, PLG, and VTN. These data suggest that YS endoderm produces a whole range of major and minor serum proteins similar to those produced in adult liver. A third group of genes of interest to us is the transcription factors. They include SOX5, ARH, NR3C2, TFCP2L1, PROX1, RXRG, and IRF7.

Combining both the GO and manual annotation approaches, these enriched YS endoderm genes can be grouped in five broad categories (Fig. 3): enzymes, serum proteins, transcription factors, and others (annotated and nonannotated). A large proportion, 42% (159), of the 378 enriched genes belong to the group with catalytic activities, 8% (30) to the serum proteins, and 2% (7) to the transcription factors. The remaining 48% (182), including both annotated (126) and nonannotated (56), contain genes with various other functions.

thumbnail image

Figure 3. Enriched genes can be broadly grouped into five categories. Number of genes and percentage of total enriched genes are also shown. In the others-annotated category, genes of a variety of known functions (ligands, receptors, intracellular signaling, cell adhesion, etc) are grouped together.

Download figure to PowerPoint

Verification by Whole-Mount In Situ Hybridization

The stringent criteria we have used for the selection gave us a high level of confidence for the enrichment of these genes in the YS endoderm. This was also supported by an across-the-board comparison of the YS endoderm data sets with our inhouse transcriptome data sets from over one hundred chicken embryonic tissue samples. These enriched genes therefore can serve as a molecular resource for understanding the physiological roles the YS endoderm plays during early development. As an alternative method of verification for the usefulness of our screen, we performed RNA in situ gene expression analyses for eight genes. Seven of them are genes that have met all the criteria in our selection, and one (PCBD1) represents genes that are putatively involved in YS endoderm differentiation, but have been dropped during the selection process. These seven genes represented four molecular categories: metabolic enzymes (MOGAT1, HMGCS2, and PLA2G12B), serum proteins (VTN and TRF), transcription factors (TFCP2L1) and others with annotated function (PSAP). All seven genes showed an endoderm specific expression (Fig. 4). One of them, PSAP, could be detected in the endoderm as early as at HH4 (Fig. 4I). In the rest, we saw weak expression at HH7 and robust expression at around HH10, followed by a further increase afterward (exemplified by TRF shown in Fig. 4J). For PCBD1, although expression was detected readily at HH7, strong and specific expression in the YS endoderm was seen at E2. Exclusion of genes like PCBD1 in our final list suggested that our criteria could lead to false negatives, but this also gave us a high level of confidence in those selected ones.

thumbnail image

Figure 4. RNA in situ hybridization verification of eight genes. A: MOGAT1. B: HMGCS2. C: PLA2G12B. D: VTN. E: PCBD1. F: PSAP. G: TRF. H: TFCP2L1. All section views. A–F: Embryonic day (E) 2 yolk sac. G,H: E3 yolk sac. Four categories are represented here: enzymes (A–C), serum proteins (D,G), transcription factors (H), and others (PSAP). PCBD1 (E) was chosen as an example of putative relevant genes that have been dropped during the selection process. I: Whole-mount views of PSAP in situ hybridization at HH4, HH10, and HH13. J: Whole-mount views of TRF in situ hybridization at HH7, HH10, and HH13.

Download figure to PowerPoint

PTGDS in Vascular Tone Control

As a proof of principle and because of our primary interest in the circulatory system, we asked whether any of the enriched genes can offer us novel insight on the function of the YS endoderm in vascular development. In addition to serum proteins and growth factors, two genes, the angiotensin converting enzyme (ACE) and prostaglandin D2 synthase (PTGDS), known to be critical for vascular tone regulation in mammals, are highly enriched (Supp. Fig. S1). Both genes were confirmed to be expressed in the YS endoderm by in situ analysis, and PTGDS will be described in more detail here. Prostaglandin D2 (PGD2) is an acidic lipid mediator derived from arachidonic acid by the sequential action of cyclooxygenase and PTGDS. PTGDS converts prostaglandin H2 into biologically active PGD2 (Fig. 5A; Giles and Leff,1988; Narumiya et al.,1999). PTGDS appeared on the very top of our list based on the fold change of expression levels (Supp. Fig. S1; raw data sets in GSE27958). PTGDS expression was initiated at HH10 (Fig. 5B) and its abundant expression was seen in the YS endoderm after E2 (Fig. 5C,D). Supporting an important role of PGD2 in vascular tone regulation, we found that injection of exogenous PGD2 into E2–E3 vasculature caused vessel rupture around the cardiac area (Fig. 5E,F for control and PGD2 postinjection video snapshots; and compare control pre- and postinjection, Supp. Movies S1 and S2, with PGD2 pre- and postinjection, Supp. Movies S3 and S4; each of the Supp. Movie represents a 5 min video taken with one frame per 5 sec). The vascular lesion by PGD2 injection was highly efficient, with 89% (31/35) showing the phenotype in comparison to 5.5% (1/18) for control injections. Co-injection of PGD2 with receptor antagonists led to a partial rescue of this phenotype, with 29% (4/14) showing vascular lesion with AH6809 co-injection and 21% (3/14) with GW627368X co-injection. The possibility of lesions caused by injection itself was ruled out as the rupture appeared to happen with a 1- to 2-min delay after injection, and a similar phenotype was seen when PGD2 was injected from the vitelline artery. Administration of PGD2 at E3–E4 pericardially, instead of intracardially, also caused similar rupture (Supp. Movie S5; approximately 2-min video taken with 24 frames per sec). We were unable to pinpoint, however, the exact location of the rupture or the exact affected tissue that led to the rupture.

thumbnail image

Figure 5. PTGDS is involved in the regulation of vascular integrity. A: Schematic diagram of how PGD2 is produced in cells. PTGDS regulates the last step in its production. B–D: Expression of PTGDS in the YS endoderm. B: Whole-mount view of embryonic day (E) 1.5 embryo. C: Section view of E2 yolk sac. D: Section view of E3 yolk sac. E: Snapshots of an E2.5 embryo, 0–5 min after control injection. F: Snapshots of an E2.5 embryo, 0–5 min after PGD2 injection. Vessel rupture is seen 1–2 min after injection.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Our results support the hypothesis that YS endoderm functions as a digestive system in early development (Speake et al.,1998; Fig. 6). Our data also suggest that this function starts as soon as the embryonic circulation is established at E2. For this to take place, the differentiation of YS endoderm cells may precede E2, similar to the differentiation of blood and endothelial cells in the YS hematopoietic system (Nakazawa et al.,2006; Sheng,2010). Most of the enriched genes exhibit high expression levels at E2, with a modest increase at E3 and E4. When compared with our transcriptome data of the primitive streak and early extraembryonic tissues at E0.5–E1 (Nakazawa et al.,2009; Alev et al.,2010), many of the enriched YS endoderm genes are already weakly expressed at E1. This coincides with the time when hematopoietic cells start their terminal differentiation by initiating hemoglobin gene expression. Whether these two differentiation processes influence each other awaits future investigation.

thumbnail image

Figure 6. Model for how yolk sac (YS) endoderm cells function in early postcirculation development. The sub-cellular structures are based on electron microscopy studies (Mobbs and McMillan,1979,1981). Functional compartmentalization is based on molecular profiling from this work and on lipid biochemistry reviewed in (Speake et al.,1998).

Download figure to PowerPoint

The morphology of YS endoderm cells is very different from that of the junctional endoblast and definitive endoderm (Mobbs and McMillan,1979,1981). The difference in morphology between the YS endoderm in the area vasculosa and that in the area vitellina, however, is minor. This brings up two relevant questions. The first is the difference among the three endoderm components as mentioned above in early embryogenesis. The second is how functional differentiation of the YS endoderm is reflected in its morphology. Contrary to what has been stated in some publications, the YS endoderm does not arise from the hypoblast. Although the exact contributions are unclear, the YS endoderm is likely derived from the lower layer cells of the early germ wall and the cellularization of new endoderm cells from the periblastic “syncytium” underneath the periphery of the expanding epiblast sheet, as well as the proliferation of these two cell populations (Bellairs,1964; Yoshizaki et al.,2004). Despite its extremely distinct morphology, however, the YS endoderm can differentiate into definitive endoderm cell types when provided with proper mesoderm cues (Masui,1981,1982). Our transcriptome data presented in this work also indicate a close relationship of the YS endoderm to the definitive endoderm derived tissues and organs. These functional commonalities may be due to the fact that a physiologically shared feature of all endoderm cells is their juxtaposition, either by direct contact or by topological continuity, to the nutritive substance at the apical surface. These observations suggest that the YS endoderm may serve as a good experimental model for definitive endoderm-related organogenesis and stem cell research.

From electron microscopy studies, the similarity in the morphology of YS endoderm cells in the area vasculosa and area vitellina has been attributed to the inclusion of yolk drops from the yolk during YS endoderm cellularization process (Mobbs and McMillan,1979,1981). These studies also showed that major differences for the YS endoderm cells in the area vasculosa, in comparison with those in the area vitellina, are the presence of coated pits and canaliculi in the apical surface and vacuoles in the apical cytoplasm, indicating an active uptake of new yolk materials and their breakdown in this region. This suggests that mesoderm components of the YS have an instructive role in initiating functional differentiation of the YS endoderm. Mobbs et al. proposed a “topping-up” hypothesis whereby endoderm cells of the area vasculosa continuously supply yolk materials to be incorporated into digestive/biosynthetic intracellular yolk drops with their end products being transported across the vascular side of the endoderm to the extraembryonic circulation and then to the embryo. Without this “topping-up,” embryonic growth will cease after using up the initial intracellular supply of yolk materials. Although we did not find Clathrin among the enriched genes (both heavy chain CLTC and light chain CLTA genes are, however, highly expressed in the endoderm, as are their known interaction partners PICALM, CLINT1, and ARCN1), the presence in the enriched list of those encoding several putative adaptor genes (AP3S2, DAB2, and ARRDC2), many lipid binding and transporting proteins and a large number of lipid metabolic and catabolic enzymes is consistent with the model. This is also supported by our observation that embryos cultured in a modified “Cornish pasty” setting (devoid of yolk) can grow normally for approximately 24 hr after circulation, followed by severe growth retardation (Nagai et al.,2011).

The complexity of genes encoding serum proteins suggests that the YS endoderm is a major site for their synthesis. This is supported by in situ analysis data from several selected genes. A likely scenario is that functional serum proteins and lipoproteins in early embryonic circulation are exclusively produced by the YS endoderm, with little or no contribution from extraembryonic mesoderm cells or intraembryonic tissues. The YS endoderm and primitive red blood cells, both derived extraembryonically, thus are among the earliest functional cell types in avian development. The heavy burden of usage and its rapid expansion in surface area suggest that the YS endoderm, like the hematopoietic cells and adult endoderm organs, may also contain a stem cell/progenitor cell population for renewal and repair. Although the screen presented in this work was not intended to find markers for these rare cells, identifying markers/niches for such cell population would complement similar studies on regeneration and stem cell maintenance in definitive endoderm derived tissues.

A surprise finding from this work is the strong association of regulators for vascular integrity with the YS endoderm. In addition to major constituent serum proteins, complement proteins, coagulation factors, and potent regulators such as angiotensin and prostaglandin are produced by the YS endoderm. PGD2 has been reported to induce endothelium dependent arterial relaxation (Braun and Schror,1992). In the present study, we observed that PGD2 tends to induce constriction of capillaries followed by dilatation of the atrial inflow component and major vitelline vein, eventually leading to a rupture. The precise nature of the rupture is unclear. It is likely caused by a localized weakening of vascular integrity coupled with an increase of systolic pressure and cardiac arrhythmia. These phenomena are most likely mediated by receptor activation. However, no PGD2 receptor (DP) has been identified in the chicken genome so far (Lagerstrom et al.,2006). Molecular phylogenetic analysis by us and from published work (Brink et al.,2004; Kwok et al.,2008) indicated that mammalian prostaglandin receptors DP, EP2 and EP4 form a closely related group, separated from the other two prostaglandin E receptors (EP1 and EP3). Three chicken prostaglandin E receptors are present in the chicken genome (EP2, EP3, and EP4). It is therefore possible that PGD2 made in the YS endoderm can regulate the vascular tone through the EP2 or EP4 receptor. Indeed, biochemical studies indicated that both PGD2 and PGE2 can bind with high affinity to both DP and EP2 receptors (Matsuo and Cynader,1993). In our work, administration of prostaglandin E2 caused a similar vascular rupture. Future studies, both on the etiology of vascular rupture (dilation/constriction; endothelial/smooth muscle) and on the molecular nature of this effect (PGD2/PGE2; EP2/EP4), may shed light on how the embryo can elicit tight regulation of vascular integrity while at the same time undergoing active tissue growth and vascular remodeling.

In summary, our results indicate that the YS endoderm is a primary source of de novo-synthesized serum proteins and lipids and is involved in regulating vascular integrity in early chick development. During mammalian evolution, the nutritional source for embryonic growth has shifted from the YS to the chorioallantoic placenta. However, YS endoderm-mediated nutrient uptake has been shown to be critical in monotremes and marsupials for a large part of their embryonic development and in eutherian mammals (including mice and humans) during early phases of their post-circulation development (Lambson,1966; King and Enders,1970; Jollie,1990; Freyer and Renfree,2009; Zohn and Sarkar,2010). It will be worthwhile to investigate in the future whether mammalian YS endoderm can play similar roles in regulating the integrity of newly formed vasculature and in supplying the earliest serum proteins and lipids to the embryonic circulation.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Egg Incubation, Tissue Preparation, Chemicals, and Array Analysis

Fertilized hens' eggs were purchased from Shiroyama Farm (Kanagawa, Japan) and incubated at 38.5°C to desired stages. Dispase (#17105) was purchased from Invitrogen. PGD2 (#12010), AH6809 (#14050), and GW627368X (#10009162) were purchased from Cayman Chemical. The 500 kDa fluorescein dextran (#D7136) was obtained from Molecular Probes. For gene array analysis, tissues were dissected as duplicate samples for the area opaca of E1 or the YS endoderm of E2–E4 (1–2 embryos per sample). Total RNA was extracted with the RNeasy mini kit (Qiagen). Five micrograms of each total RNA sample were used without an amplification step for Affymetrix Chicken Genome Array analysis (No. 900592, Affymetrix). Probeset correlation analysis and visual plot of array data was based on the eXintegrator program (www.cdb.riken.go.jp/scb/MartinG/html/exintegrator.html). GO analysis was based on the g:Profiler program (http://biit.cs.ut.ee/gprofiler/) (Reimand et al.,2007).

Whole-Mount In Situ Analysis

Antisense digoxigenin-labeled in situ probes were generated against the following regions: MOGAT1 (1302-2115 of NCBI No. XR_027084); HMGCS2 (14-572 of NCBI No. XM_422225); PLA2G12B (939-1658 of NCBI No. XM_421584.2); VTN (768-1200 of NCBI No. NM_205061); TRF (1203-1693 of NCBI No. NM_205304); PCBD1 (56-905 of NCBI No. NM_204905); TFCP2L1 (7-586 of NCBI No. XM_422087); PSAP (2452-3007 of NCBI No. NM_204811.2) and PTGDS (16-573 of NCBI No. XM_001234014). In situ hybridization analysis was carried out as previously described (Stern,1998).

Microinjection

PGD2 or control solution was microinjected together with the tracer dye fluorescein isothiocyanate (FITC) -dextran into the cardiac cavity of HH13–HH17 embryos in ovo with a glass capillary. For the preparation of working solution, 10 mM of PGD2 in dimethyl sulfoxide, 1 mg/ml of FITC-dextran and 1% of Fast Green injection dye were diluted 10× in PBS. The Fast Green dye was used to aid the injection, and the FITC-dextran was used to assess possible leakage and the efficiency of the spread of injected material in the vasculature. Injection capillary needles were made from 50-μl capillary tubes (Drummond Scientific #2-000-050, Broomall, PA) using a vertical puller. Approximately 0.1 μl of working solution or control solution was injected intracardially. Successful injection was judged under fluorescent microscope by minimal leakage of FITC-dextran, no leakage of blood cells, and an even distribution of FITC-dextran throughout the vasculature. Injected embryos were then subjected to filming (5-sec frame interval for a total duration of 5 min). Rescue experiment was performed by co-injection of PGD2 (see above) with 1 mM of AH6809 or 1 mM of GW627368X. AH6809 is an antagonist for EP and DP receptors with nearly equal affinity for human EP1, EP2, EP3, and DP1 receptors. GW627368X is a potent and selective competitive antagonist of the EP4 receptor with additional human TP receptor affinity.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

We thank Dr. Takeya Kasukawa in the Functional Genomics Unit for help with the gene chip analysis and members of the Functional Genomics Unit for gene chip runs, Dr. Yukiko Nakaya in the Lab for Early Embryogenesis for help with obtaining the confocal image shown in Figure 1A, and Dr. Wei Weng in the Lab for Early Embryogenesis for sharing data from ongoing serum proteome project. This work was support by RIKEN CDB.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
DVDY_22690_sm_suppfig1.tif1942KSupporting Figure 1 Correlation analysis profiles of 20 of the final 378 genes using xIntegrator. The rest of the genes have similar profiles. Twenty genes are selected to represent two general trends of expression increase: those with very low or no expression at day 1 and those with some expression at day 1. The primitive streak (PS) data from our previous publication are also included in this analysis for comparison. See the Experimental Procedures section for details of the xIntegrator analysis.
DVDY_22690_sm_suppmovie1.mov2115KSupporting Movie 1 Movies S1–S5. Vascular rupture caused by PGD2 injection. Movie S1 and S2: Before and after intracardiac control injection, respectively; Supp. Movie S3 and S4: Before and after intracardiac PGD2 injection, respectively. Each one of the Supp. Movies S1–S4 is a 5-min movie taken with 5-sec interval per frame. Supp. Movie S5: After pericardial PGD2 injection. Two-min movie in real time, taken with 24 frames per sec.
DVDY_22690_sm_suppmovie2.mov2210KSupporting Movie 2.
DVDY_22690_sm_suppmovie3.mov3006KSupporting Movie 3.
DVDY_22690_sm_suppmovie4.mov2885KSupporting Movie 4.
DVDY_22690_sm_suppmovie5.mov9388KSupporting Movie 5.
DVDY_22690_sm_supptable2.pdf262KSupporting Table 2 A list of GO terms (biological processes only) significantly associated with 323 annotated genes out of the 378 enriched genes. A few of them are highlighted (see text). Constituent genes of these highlighted terms (except for the general GO term metabolic process) are shown following the GO list.
DVDY_22690_sm_supptable2.xls72KSupporting Table 2.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.