Appearance of primitive “pyrenocytes” in the fetal bloodstream. (A) Immunohistochemistry fetal blood, stained with anti–ϵγ-globin antibodies (red) and DAPI (blue) to label primitive erythroid cells and nuclei, respectively. The upper panel is a darkfield combined fluorescent image of E14.5 blood. The numerous smaller definitive erythrocytes are not evident because they are both ϵγ-globin– and DAPI-negative. The lower panel is E15.5 blood with fluorescent images layered over a Hoffman contrast image to show the abundant unstained definitive erythrocytes compared with the larger ϵγ-globin–positive primitive erythroblasts. Primitive pyrenocytes with small rim of ϵγ -globin–positive cytoplasm are indicated by arrows. Scale bar = 10 μm. (B) Morphometric analysis of E14.5 blood images indicates that pyrenocytes are significantly (P < .001) smaller than surrounding primitive orthochromatic erythroblasts (EryP). However, there is no significant difference in the size of their nuclei. Error bars are SEM.

Primitive pyrenocytes constitute a distinct cellular population in the fetal bloodstream. (A) Relative size of nucleated primitive erythroid cells as determined by ImageStream analysis reveals that pyrenocytes account for approximately 2% of nucleated blood cells in E15.5 blood (red line), less than 1% at E14.5 (blue line), and are not detected at E12.5 (green line). (B) Example of images of fetal blood cells visualized using the ImagesStream. Concurrent brightfield (BF) images and fluorescent images (anti–ϵγ-globin antibodies [ϵγ], Ter119, and the DNA stain [Draq5]) of larger primitive orthochromatic erythroblasts (rows 1 and 2) and of smaller primitive pyrenocytes (rows 3 and 4). (C) Primitive pyrenocytes and primitive orthochromatic erythroblasts (EryP) in the E15.5 circulation have similar nuclear sizes. Error bars represent SD. (D) E15.5 blood stained with annexin V (AnV), Ter119, and Draq5 (shown as merged images), and 7AAD analyzed by the ImageStream. Row 1 is a primitive pyrenocyte that is annexin V–positive and 7AAD-negative. Row 2 is a primitive orthochromatic erythroblasts that is negative for annexin V and 7AAD.

Primitive erythroblasts may associate with fetal liver macrophage cells in vivo. (A) ImageStream images of liver cells viewed in brightfield (BF) or stained with anti–ϵγ-globin (ϵγ) and Ter119 antibodies and Draq5. The right panel is a combination of images of the 3 fluorescent stains. ImageStream analysis software facilitates the identification of different erythroid populations in the liver by quantifying brightfield characteristics (contrast and gradient RMS, ie, irregular surface) as well as fluorescent stain intensities. Primitive orthochromatic erythroblasts (row 1) were contrast^hi, gradient RMS^hi, ϵγ-globin^hi, Ter119^{lo, large}, Draq5^{+, small}. Primitive enucleated erythrocytes (row 2) were contrast^hi, gradient RMS^hi, ϵγ-globin^hi, Ter119^{lo, large}, Draq5⁻. Immature and mature definitive erythroblasts (rows 3 and 4, respectively) were distinguished from primitive erythroid cells by their nuclear size as well as by being contrast^lo, gradient RMS^lo, ϵγ-globin^lo, and Ter119^hi. (B) Cytospin preparations of E14.5 liver erythroblast islands stained with Wright-Giemsa reveal the presence of small numbers of primitive erythroid cells (*) attached to macrophage cells (m). (C) Immunohistochemistry of E15.5 liver reveals the very common spatial association (arrowheads) of ϵγ-globin–positive primitive erythroid cells (red) with F4/80-positive macrophage cells (blue). Scale bar = 20 μm.

Primitive erythroid cells enucleate in association with macrophages. (A) Late-stage primitive erythroblasts derived from fetal blood do not autonomously enucleate when cultured in vitro. In control cultures, immature definitive erythroblasts derived from the fetal liver (E14.5 liver, solid blue line) continue to divide as they mature. This maturation includes autonomous enucleation as evidenced by the increasing accumulation of erythrocytes (dashed blue line). The nucleated cells in E14.5 blood are late-stage primitive erythroblasts that have ceased dividing. Unlike definitive cells, late-stage primitive erythroblasts did not complete their maturation by enucleating in vitro as evidenced by the persistence of nucleated cells (solid red line) and the lack of increasing numbers of enucleated erythrocytes (dashed red line). (B) E14.5 primitive erythroblasts enucleate in vitro when cocultured with macrophage cells. The enucleation of primitive blood cells attached to macrophages was significantly higher than that of the same blood cells cultured in parallel without macrophages. The mean plus or minus SEM of 7 independent experiments is plotted. (C) Nuclear engulfment by macrophages is evident after reconstitution of erythroid islands with E14.5 primitive erythroid cells. E14.5 blood was prelabeled with cytotracker dyes to determine the source of nuclei. The left panels are Hoffman contrast images overlayed with HO staining of nuclei. The right panels are fluorescent images revealing the prestained cytotracker green (top) or cytotracker red (bottom) cells. After 48 hours of reconstitution, most of the attached cells from E14.5 blood (top panels) are primitive orthochromatic erythroblasts with occasional enucleated primitive erythrocytes (*). Additional small nuclei, independent of orthochromatic erythroblasts, are also evident (arrowheads). Stripping of erythroblast islands (lower panels) reveals the persistence of labeled nuclei engulfed by macrophage cells.

Primitive erythroblasts interact with fetal liver macrophage cells in vitro. (A) Immunohistochemical analysis of 2 erythroblast islands reconstituted with E14.5 blood and E14.5 liver macrophages. (Left panel) ϵγ-globin–positive primitive erythroid cells (red) attached to F4/80-positive macrophage cells (blue). (Right panel) The same erythroblast islands with nuclei stained by HO. Small bright nuclei of the primitive erythroblasts are seen as well as larger fainter macrophage nuclei (m). An enucleated primitive erythrocyte is indicated by an arrowhead. (B) Primitive blood cells preferentially bind F4/80-positive macrophages compared with F4/80-negative adherent cells. The percentage of F4/80-positive macrophages (left panel) and F4/80-negative adherent cells (right panel) with 0, 1, 2, 3, or 4 or more ϵγ-globin–positive cells attached is plotted. These data represent the mean plus or minus SEM of 3 independent experiments. (C) The presence of α4-integrin blocking antibody during the reconstitution of erythroblast islands resulted in both a significant decrease in the percentage of macrophages in erythroblast islands, defined as containing 4 or more erythroblasts (4+, red bars) and a significant increase in the percentage of macrophages lacking any attached primitive blood cells (0, black bars) compared with isotype controls (P < .02). These data represent the mean plus or minus SEM of 3 independent experiments. (D) The addition of α4-integrin blocking antibody also caused a decrease in the average number of primitive erythroid cells per reconstituted erythroblast island (defined as 4 or more primitive erythroid cells per macrophage) compared with isotype controls (P ≤ .005). The results of 4 independent experiments are shown.

Primitive erythroblasts do not enucleate by splenic pitting or by karyolysis. (A) The estimated number of primitive erythroblasts that undergo enucleation per day between E11.5 and E18.5 of gestation in outbred mice. Enucleation events were estimated using the percentage of enucleated primitive erythroid cells previously determined for each embryonic day, assuming that there are 6.7 × 10⁶ primitive erythroid cells per fetus and that total primitive erythroid cell numbers do not decrease between E12.5 and E17.5 of gestation.¹² (B) Immunohistochemistry of transverse section of E15.5 mouse embryo with F4/80 (blue) and anti–ϵγ-globin (red) antibodies reveals the presence of mature macrophage cells and primitive erythroid cells, respectively. At lower power (i), most of the macrophages can be seen in the liver (L) with few in the surrounding body wall and spleen (Sp) or stomach (St). At higher power (ii), infrequent large primitive erythrocytes are seen in vessels in the spleen (arrowhead), but they are not associated with the few splenic macrophages present at this time. Labeled primitive erythroblasts are observed as a subset of blood cells in the aorta (Ao) (iii). Scale bars = 0.1 mm. (C) Analysis of fragmentation of DNA from primitive erythroid cells in E15.5 peripheral blood. Lanes 1 and 2 show no fragmentation of E15.5 peripheral blood DNA in either untreated (lane 1) or after 24 hours of treatment with 0.5 μM staurosporine (lane 2). In contrast, bone marrow DNA shows significant laddering after 6 hours (lane 3) or 24 hours (lane 4) of treatment. Vertical lines have been inserted to indicate repositioned gel lanes.