SCL/Tal-1 is essential for hematopoietic commitment of the hemangioblast but not for its development

Abstract

In this report, we have defined the stage at which Scl functions in the establishment of the hematopoietic system and provide evidence that its primary role is in the generation of the hematopoietic lineages from a progenitor called the blast colony-forming cell (BL-CFC), a cell considered to be the in vitro equivalent of the hemangioblast. Using an embryonic stem (ES) cell line in which lacZ cDNA has been targeted to the Scl locus, we show that most of the BL-CFCs are detected in the SCL/lacZ- population, indicating that this progenitor does not express Scl. In the blast colony assay, Scl-/- cells initiate colony growth but are unable to generate endothelial and hematopoietic progeny and thus form colonies consisting of vascular smooth muscle cells only. The capacity to give rise to blast colonies can be rescued by retroviral transduction of a wild-type Scl gene into Scl-/- FLK-1+ cells, suggesting that the BL-CFC is generated in this population. Finally, we show that Scl-/- endothelial cells display a growth deficiency in monolayer cultures that can be partially overcome by maintaining this population as 3-dimensional aggregates indicating that specific cellular interactions are required for maintenance of the Scl-/- endothelial lineage in vitro.

Figure 1.

BL-CFC potential of Scl-expressing EB populations and Scl expression during blast colony development. (A) Percentage of cells expressing SCL in day-3 EBs. Boxes indicate the populations isolated by cell sorting. (B) Frequency (colonies/10⁵ cells) of blast colonies in isolated populations; PS indicates presort; SCL^-, SCL/lacZ^-; SCL⁺, SCL/lacZ⁺ fractions. Blast colonies (BLASTS; ▪) and secondary EBs (2° EBs; ) were scored after 4 days in culture. Data are presented as mean number of blast colonies from 3 dishes. Bars, where visible, represent SEM. (C) Expression analyses of the PS, SCL^-, and SCL⁺ fractions. Each lane represents the amplified products from 1000 single cells deposited directly into lysis buffer. The 3′ cDNA was prepared and analyzed as described., (D) Total blast colonies in each fraction. Data are presented as mean number of blast colonies from 3 dishes. Bars, where visible, represent SEM. (E) FLK-1⁺ SCL/lacZ^- cells were isolated from day 3 EBs and replated into hemangioblast cultures. Shown is colony morphology at day 1, the core stage; day 2, the early hematopoietic stage; and day 3, the blast colony stage. Arrow indicates the developing vascular core at the day 2 early hematopoietic stage of blast colony development. Original magnification, × 400. Objective × 40, numerical aperture 0.50. MagnaFire software (Optronics, Goleta, CA) was used to analyze the images. (F) Expression analyses of the starting population and of colonies at the 3 different developmental stages; 10⁵ FLK-1⁺ SCL/lacZ^- (F⁺ S^-) cells, 1000 day 1 and day 2 pooled colonies and 500 day 3 pooled colonies were used for this analysis. Hybridization was carried out with probes designed against the 3′ UTR of the indicated genes. Hybridization to the ribosomal gene L-32 was included to control for the level of cDNA in each lane.

Figure 2.

BL-CFC and hematopoietic progenitor content of FLK-1/SCL populations from day 4 EBs. (A) FLK-1^-SCL/lacZ^- (F^-S^-), FLK-1⁺ SCL/lacZ^- (F⁺ S^-), FLK-1⁺ SCL/lacZ⁺ (F⁺ S⁺) and FLK-1^-SCL/lacZ⁺ (F^-S⁺) populations were isolated by cell sorting. Boxes indicate the percentages of different fractions. (B) Total number of blast colonies (BLASTS; ▪) and secondary EBs (2° EBs; ) were scored from the unfractionated (PS indicates presort), F^-S^-, F⁺ S^-, F⁺ S⁺, and F^-S⁺ populations. Data are presented as mean number of blast colonies from 3 dishes. Bars, where visible, represent SEM. (C) Expression analyses of isolated fractions. Each lane represents the amplified products from 1000 single cells. Hybridization was carried out with probes designed against the 3′ UTR of the indicated genes. (D) Total number of hematopoietic colonies in each of the indicated fractions. Colonies were categorized as follows: ERY/P (), primitive erythroid; MAC (□), macrophages; E/MAC (), erythroid-macrophage; and MIX (), multilineage. Data are presented as mean number of colonies from 3 dishes. Bars, where visible, represent SEM.

Figure 3.

Rescue of blast and hematopoietic colonies from Scl^-/- FLK-1⁺ precursors. (A) Kinetics of FLK-1 expression in Scl⁺/⁺ and Scl^-/- EBs. Numbers above the graphs indicate the day of differentiation. Percentages of FLK-1⁺ cells are indicated above the gates. (B) Blast colony potential of day 4 Scl^-/- FLK-1⁺ EB-derived cells following 24 hours of coculture with either SCL or control (MSCV) virus-producing cells. Colony numbers are presented per 10⁵ FLK-1⁺ cells seeded onto virus-producing cells. Data are presented as mean number of blast colonies from 3 dishes. Bars, where visible, represent SEM. (C) Hematopoietic progenitor potential of the infected populations described in panel B. □ indicates ERY/P; ▪, MAC; and , MIX. Data are presented as mean number of colonies from 3 dishes. Bars, where visible, represent SEM. (D) Expression of the endogenous and viral Scl transcripts in day 3.5 Scl^+/+ FLK-1⁺ (sort, +/+), day 4 Scl^-/- FLK-1⁺ (sort, -/-), and 2 pools of rescued Scl^-/- FLK-1⁺ cells (rescue, -/-) harvested following 3 days of coculture with SCL virus-producing cells (SCLVPCs) and 5 days of expansion in KL, IL-3, and EPO. To facilitate the detection of endogenous Scl, primers were designed against the 3′ UTR of the Scl gene. Detection of viral Scl was carried out using a 5′ primer designed against the 3′ translated region of the Scl gene and a 3′ primer designed against the PGK promoter present only in the retroviral transcript. (E) Expression analyses of the populations described in panel D were performed using probes designed against the 3′ UTR of the indicated genes.

Figure 4.

Kinetics of blast and hematopoietic colony rescue from Scl^-/- derived populations. (A) Kinetics of blast colony development in Scl^+/+ EBs ( , 2° EBs; ▪, BLASTS). Numbers below the graph indicate the day of differentiation. Data are presented as mean number of colonies from 3 dishes. Bars, where visible, represent SEM. (B) Blast colony potential of FLK-1⁺ cells from day 3.5 Scl^+/+ EBs (+/+, 3.5) and FLK-1⁺ cells from day 3.5, 4, 5, and 6 Scl^-/- EBs (-/-, 3.5/4/5/6). Colony-forming potential was measured in the various FLK-1⁺ populations prior to (- virus) and following infection with SCLVPCs (+ virus). Colony numbers are presented per 10⁵ input cells. Data are presented as mean number of colonies from 3 dishes. Bars, where visible, represent SEM. (C) Hematopoietic progenitor potential of FLK-1⁺ cells from day 3.5 Scl^+/+ EBs (+/+, 3.5) and FLK-1⁺ cells from day 4, 5, and 6 Scl^-/- EBs (-/-, 4/5/6). Colony-forming potential was measured prior to (- virus) and following infection with SCLVPCs (+ virus). □ indicates ERY/P; ▪, MAC; , E/MAC; and , MIX. Colony numbers are represented per 10⁵ input cells. Data are presented as mean number of colonies from 3 dishes. Bars, where visible, represent SEM. (D) Expression analyses of the presorted (P), FLK-1^- (-) and FLK-1⁺ (+) fractions from day 3.5 Scl^+/+ EBs (+/+) and from the different stage Scl^-/- EBs (-/-). Boxes highlight Brachyury expression in FLK-1⁺ fractions from Scl^+/+ and Scl^-/- EBs at the indicated time points (in days) shown below the figure.

Figure 5.

Colony-forming potential of Scl^-/- FLK-1⁺ EB-derived cells. (A) Blast colony (□) and core colony () development from FLK-1⁺ day 3 Scl^+/+ and day 3.5 Scl^-/- EBs. Data are presented as mean number of blast colonies and cores per 10⁵ input cells from 3 dishes. (B) Photographs of Scl^+/+ blast colonies and Scl^-/- cores at day 2 and 3 of development. Original magnification, × 200. Objective × 20, numerical aperture 0.30. MagnaFire software was used to analyze the images. (C) RT-PCR analysis of Scl^+/+ (+/+) and Scl^-/- (-/-) colonies at days 2 and 3 of growth. Each lane represents a pool of 500 colonies. Hybridization was carried out with probes designed against the 3′ UTR of the indicated genes. (D) Immunohistochemistry of adhesive cells generated from Scl^+/+ blast colonies (Scl^+/+) and Scl^-/- cores (Scl^-/-). CD31⁺ (red) and SMA⁺ (green) populations are indicated. Original magnification, × 200. Objective × 20, numerical aperture 0.70. OpenLab software was used to analyze the images.

Figure 6.

Vascular potential of FLK-1⁺ cells from Scl^+/+ and Scl^-/- EBs. (A) FLK-1⁺ cells from day 3.0 Scl^+/+ or day 3.5 Scl^-/- EBs were cultured on gelatin-coated wells in the presence of VEGF and bFGF for 3 days. Following culture the cells were harvested and stained for CD31, FLK-1, VECAD, and SMA. Numbers indicate the percentage of the respective populations. (B) Immunohistochemistry of day 3 Scl^+/+ FLK-1⁺ cells (Scl^+/+) and day 3.5 Scl^-/- FLK-1⁺ cells (Scl^-/-) cultured on fibronectin-coated glass coverslips for 3 days. CD31⁺ cells are shown in red, whereas the SMA⁺ cells are indicated in green. Original magnification, × 200. Objective × 20, numerical aperture 0.70. OpenLab software was used to analyze the images. (C) FLK-1⁺ cells from day 3 Scl^+/+ or day 3.5 Scl^-/- EBs were allowed to aggregate by culturing at high density in low-adherence plates. The aggregates were cultured for 3 days in VEGF and bFGF and then processed for CD31, FLK-1, VECAD, and SMA as described in panel A. (D) RT-PCR analyses of day 3.0 Scl^+/+ FLK-1⁺ sorted cells (+/+, S), Scl^+/+ adherent cells (+/+, Ad), Scl^+/+ aggregated cells (+/+, Agg), Scl^-/- day 3.5 FLK-1⁺ sorted cells (-/-, S), Scl^-/- adherent cells (-/-, Ad) and Scl^-/- aggregated cells (-/-, Agg). Hybridization was carried out with probes designed against the 3′ UTR of the indicated genes. (E) RT-PCR analysis of day 3.0 Scl^+/+ FLK-1⁺ sorted cells (+/+, S), Scl^+/+ adherent cells (+/+, Ad), and Scl^+/+ aggregated cells (+/+, Agg) for the expression of the macrophage-specific genes c-fms, f4/80, CD11b, for the erythroid-specific gene, β-major, and for the endothelial-specific gene, VWF. β-Actin was used to normalize the content of cDNA in these populations.

Figure 7.

Model of early hemangioblast commitment. The SCL-independent and -dependent phases are indicated. The BL-CFC is represented as a Flk-1⁺/Brachyury⁺/Scl^- progenitor. In appropriate culture conditions, wild-type (WT) progenitors generate blast colonies composed of the hematopoietic (Hemat), vascular smooth muscle (VSM), and endothelial (endo) lineages. In the absence of SCL, these progenitors fail to generate the hematopoietic and endothelial lineages and give rise to colonies of VSM cells only.