Lane MC and Sheets MD (2002), Primitive and definitive blood share a common o...

XB-ART-34680

Dev Biol 2002 Aug 01;2481:52-67. doi: 10.1006/dbio.2002.0717.

Show Gene links Show Anatomy links

Primitive and definitive blood share a common origin in Xenopus: a comparison of lineage techniques used to construct fate maps.

Lane MC , Sheets MD .

???displayArticle.abstract???
Primitive blood constitutes the ventralmost mesoderm in amphibians, and its cleavage-stage origin reveals important clues about the orientation of the dorsal/ventral axis in the embryo. In recent years, investigators employing various lineage-labeling strategies have reported disparate results for the origin of primitive blood in Xenopus [W. D. Tracey, Jr., M. E. Pepling, G. H. Thomsen, and J. P. Gergen (1998). Development 125, 1371-1380; M. C. Lane W. C. Smith (1999). Development 126, 423-434; K. R. Mills, D. Kruep, and M. S. Saha (1999). Dev. Biol. 209, 352-368; A. Ciau-Uitz, M. Walmsley, and R. Patient (2000). Cell 102, 787-796]. These discrepancies must be resolved in order to elucidate early embryonic patterning mechanisms in vivo. We directly compared two of the techniques used to determine the origin of the ventral blood islands and primitive blood, injection of either beta-galactosidase mRNA or conjugated dextrans, by coinjecting both tracers simultaneously into individual blastomeres in cleavage-stage embryos. We find that dextrans label progeny efficiently, while beta-galactosidase activity is not present in many of the progeny of an injected blastomere, suggesting that mRNA fails to diffuse throughout a blastomere. This result demonstrates that beta-galactosidase mRNA fails to meet the criterion for a true lineage label, namely efficient detection of the progeny of a blastomere, and raises questions about interpretations based on mapping the ventral blood islands using Lac Z mRNA as a tracer. We examined the origins of the ventral blood islands and primitive blood from the vegetal region of the marginal zone in regularly cleaving embryos by coinjecting both reporters into C-tier blastomeres. Our results demonstrate that both the ventral blood islands and primitive blood routinely arise from all C-tier blastomeres. Our data, in combination with published mapping results for the dorsal aorta, demonstrate that primitive and definitive blood do not have separate origins at the 32-cell stage in Xenopus. In addition, these results support a proposal to align the dorsal/ventral axis of the mesendoderm with the animal/vegetal axis in pregastrula Xenopus.

???displayArticle.pubmedLink??? 12142020
???displayArticle.link??? Dev Biol

Species referenced: Xenopus
Genes referenced: hba3 kcnt1 tbx2

???attribute.lit??? ???displayArticles.show???

	FIG. 1. Conflicting reports on the origins of primitive blood in Xenopus. (A) Lane and Smith (1999) labeled each of the 16 different blastomeres at st. 6 with either fluorescent dextran or mRNA GFP and scored primitive blood as fluorescent cells in circulation in st. 41 tadpoles (B). Primitive blood cells arose principally from all vegetal hemisphere blastomeres, C1–C4 and D1–D4. This region, shown in orange, corresponds to the vegetal region of the marginal zone in gastrula-stage embryos and gives rise to the lateral plate mesoderm, LP. (C) Ciau-Uitz et al. (2000) labeled six of the eight vegetal hemisphere blastomeres at st. 6 with -galactosidase mRNA and scored the presence of progeny at st. 26 in the region of the VBI (D), in some cases by overlap with marker genes such as globin. The expression pattern for globin at st. 26 is indicated in purple. They concluded that primitive blood arises from only three blastomeres, C1, D1, and D4 (red), and that definitive blood arises only from blastomere C3 (blue). The nomenclature for blastomeres (Nakamura and Kishiyama, 1971) is shown in (B). Abbreviations: cbi, caudal blood islands; DA, dorsal aorta; DLP, dorsolateral plate; LP, lateral plate, also known as leading edge mesoderm; N, notochord; rbi, rostral blood islands; S, somites; orange asterisk indicates two blastomeres observed to give a small but reproducible contribution to primitive blood in some embryos found by Lane and Smith (1999).
	FIG. 2. Messenger RNA is stable in DEPC-treated dextran. Synthesized Lac Z mRNA was analyzed by agarose gel electrophoresis (lane 2). Mini-ruby dextran (“ruby,” lane 3) migrates slightly toward the anode( ). When Lac Z mRNA is mixed with ruby, ruby migrates normally, but Lac Z mRNA migrates abnormally, and is either digested or shifted into a low-mobility smear (white asterisks, lane 4). The same low-mobility smear was seen in three trials. Following DEPC treatment, ruby migrates less toward the anode (lane 5), but is still fluorescent. Lac Z mRNA mixed with DEPC treated ruby migrates normally.
	FIG. 3. An assay of diffusion by two putative lineage tracers coinjected at the four-cell stage. (A). The animal pole region of a rostral blastomere at the four-cell stage gives rise to the cement gland, the brain, and the retina. The equatorial region of a rostral blastomere gives rise to rostral notochord and pharyngeal endoderm. (B). In the coinjection assay, a mixture of mini-ruby dextran and Lac Z mRNA was coinjected at either the animal pole site or subequatorial site of one rostral blastomere. (C). At st. 30, the embryos were fixed, stained for Lac Z activity, cleared in benzyl benzoate/benzyl alcohol (2:1), and scored for the presence of both tracers in the five tissues shown on the embryo.
	FIG. 4. Results of the coinjection assay. (A–F). St. 30 embryos after coinjection of Lac Z mRNA and ruby dextran at a rostral, subequatorial position at the four-cell stage. The distributions of Lac Z activity (red nuclear staining in A, C, E) or ruby dextran (nuclear and cytoplasmic fluorescence in B, D, F) in tissues distant from the injection site are shown in whole-mount or sectioned specimens. At low magnification, the distributions of both tracers (A, B) appear similar, but at higher magnification (C, D), the two labels are differentially distributed in the head. Lac Z activity is not detected in the brain, retina, or cement gland, while ruby labels all three tissues. Sectioned specimens (E, F) reveal internal structures, such as the pharyngeal endoderm, PE, which is labeled by both tracers. (G–L) St. 30 embryos after coinjection into the animal pole region of a rostral blastomere, at the four-cell stage. The distribution of Lac Z activity (G, I, K) or ruby dextran (H, J, L) in tissues that arise from the injection site are shown. At low magnification (G, H) and at higher magnification (I, J and K, L), both tracers label the same tissues. These results demonstrate that mini-ruby dextran diffuses farther and labels more progeny of an injected blastomere than Lac Z mRNA.
	FIG. 5. Injection of blastomere C3 with Lac Z mRNA results in variable labeling. Embryos injected solely with Lac Z mRNA into blastomere C3 at st. 6 were cultured to st. 32, stained for Lac Z activity (red), and probed by in situ hybridization for globin expression (turquoise). (A, C, D) The range of results observed. While globin expression is constant from embryo to embryo, Lac Z activity is variable and sometimes overlaps with globin expression in the VBI (A–C). Although Lac Z detected somites in all C3-injected embryos, in some specimens (A, C), it labeled both the dorsal (DS) and ventral (VS) aspect of the somites, while in others (D), it labeled the dorsal, but not the ventral aspect. Other tissues showing variable Lac Z expression include the pronephros (P), the lateral plate (LP), and the branchial arches (BA), all tissues identified in previous maps as normal descendants of C3 (Dale and Slack, 1987; Moody, 1987). These results suggest that Lac Z mRNA does not diffuse well and generally labels only subsets of C3 progeny.
	FIG. 6. Ruby dextran detects more progeny than Lac Z mRNA when coinjected into blastomere C3. St. 32 embryos were stained for Lac Z activity (red), probed by in situ hybridization for globin expression (turquoise), and stained for ruby dextran (brown). Lac Z staining varies between embryos (A), while globin expression and ruby dextran staining appear fairly consistent. Most noticeable is the detection of many progeny of C3 in the lateral plate mesoderm. In (B), both red nuclei and brown cytoplasmic staining detect C3 progeny overlapping with globin expression in the VBI. In (C), C3-derived, Lac Z-positive cells are among the globin-positive cells in the VBI. Numerous dextran-positive, Lac Z-negative (brown cells without red nuclei) are visible in the sectioned embryo (C, D), and are present in the VBI. The contralateral, uninjected side of the embryo serves as a staining control.
	FIG. 7. Blastomeres C2 and C4 contribute progeny to the VBI. (A, B) Mini-ruby dextran and Lac Z mRNA were coinjected into blastomere C2 at st. 6. Overlap of dextran-labeled (brown), Lac Z-expressing (red) cells with globin expression in the VBI (blue) demonstrates that C2 contributes to the VBI (arrow in B). (C, D) Coinjection of blastomere C4 demonstrates this cell also contributes progeny to the VBI.
	FIG. 8. Fluorescent primitive blood cells from blastomeres C3 and C4 circulate through the gills in tadpoles. Embryos were injected at st. 6 and filmed at st. 40. Neither blastomere was identified as a source of the VBI by Ciau-Uitz et al. (2000). The retina is to the upper left in all images. (A–D) C3 progeny. (E–H) C4 progeny. The gill arches of embryo are shown at three successive time points in (A–C) and (E–G). In each set of images, stationary labeled cells are indicated by white asterisks, while a gill arch is indicated by a white arrow. Blood cells flowing through the gill arch appear as ever-changing traces. (D, H) Low-magnification images of the two embryos reveal typical labeling patterns for blastomeres C3 and C4, respectively. Caudal lateral plate and caudal somites are labeled in both embryos.
	FIG. 9. Schematic diagram illustrating marginal zone morphogenesis and the migration of the lateral plate/leading edge mesoderm into the region of the ventral blood islands. Ectodermal and endodermal derivatives are ignored throughout this figure. See the text for discussion. All tissues in (A–E) are prospective as no histological differentiation has occurred. (C–E) Modified from Keller (1991) and Lane and Sheets (2000), with E further modified based on the recent results of Davidson et al. (2002) on closure of the mesodermal mantle. These authors showed that the mesodermal mantle does not enclose the blastocoel concentrically, as originally depicted by Keller (1991), but that the mantle closes primarily by advancing rostral and caudal fronts. Abbreviations: CBI, caudal blood islands; CC, circumblastoporal collar; CLEM, caudal leading edge mesoderm, which gives rise to the intermediate and caudal lateral plate mesoderms; DLP, dorsolateral plate; hd, head mesoderm; ht, heart primordia; N, notochord; RBI, rostral blood islands; RLEM, rostral leading edge mesoderm, which includes head, heart, liver and RBI; S, somite; , indicates the animal pole, which in anurans forms epidermis covering the heart; X, marks the site where the blastopore closes at st. 13, which becomes the anus.