Liver Organogenesis Promoted by Endothelial Cells Prior to Vascular Function K. Matsumoto, H. Yoshitomi, J. Rossant, K. S. Zaret

Supplemental Figure 1. RNA was isolated from liver bud and control wild type tissues from E9.5 embryos of the designated flk-1 genotypes and subjected to reverse transcription with an oligo dT primer, followed by PCR with gene-specific primers for distinct exon sequences. Products of three different cycle steps were run on a polyacrylamide gel. Data from the same representative, non-saturated cycle step is shown for all samples. PCR reactions with control RNA samples that were not treated with reverse transcriptase gave blank lanes (data not shown).

Medium version | Full size version

The data show that fkl-1 -/- embryos undergo normal hepatic gene induction, including that for albumin, transthyretin, and the homeobox factor Hex. Due to a lack of endothelial cells, the same embryos' liver buds are greatly diminished for the expression of PE-CAM. Note that the expression of liver genes in the flk-1 -/- liver buds appears less than that in wild type because of impaired liver bud growth (i.e., fewer cells). Finally, these results are fully consistent with Figures 3-5, demonstrating induction of albumin in flk-1 -/- liver buds cultured in vitro.

Supplemental Figure 2. In development, distinct morphological budding of the liver, at E9.0-9.5, precedes that of other endoderm-derived organs by nearly a day. Thus, to investigate the extent to which endothelial cells intermingle with tissue buds for other organs, we performed PE-CAM staining of embryo sections at E10.5.

Medium version | Full size version

As seen in the data below, isolated groups of PE-CAM positive cells, prior to vascular formation, are interspersed within tissue buds for the lung and the pancreas, and surround the glandular epithelium for the stomach. These data support the hypothesis that endothelial cells are critical for organogenesis of endoderm-derived organs besides the liver.

II. Methods:

Methodology for PE-CAM and -galactosidase staining of mouse embryos:

Staged C3H mouse embryos were rinsed with PBS and fixed in PBS containing 4% paraformaldehyde at 4°C overnight. Noon of the day of appearance of a vaginal plug was taken to be 0.5 days gestation (E0.5). The fixed embryos were rinsed in PBS, dehydrated in ethanol, and embedded in paraffin. Tissue sections (5 µm thickness) were rehydrated and treated with 0.3% H₂O₂ in PBS for 15 min at room temperature. Sections were incubated with PBS with 0.25% (w/v) trypsin for 30 min, washed in PBS alone and subsequently blocked in PBS containing 1.5% normal rabbit serum. Sections were incubated for 2 hr with rat anti-mouse PECAM (10 µg/ml) from PharMingen (San Diego, CA). Sections were washed with PBS and incubated in rabbit biotinylated anti-rat IgG (Vector Laboratories, Inc., Burlingame, CA) for 1 hr. After washing with PBS, sections were incubated with peroxidase-conjugated streptoavidine-biotin complex (Elite ABC reagent, Vector Laboratories) for 1 hr. Sections were washed with PBS and stained with peroxidase by incubating sections in 50 mM Tris-HCl buffer (pH 7.6) containing 0.05% H₂O₂ and 0.2% diaminobenzidine (Sigma Chemicals). Reactions were stopped by rinsing in H₂O and counter-staining with hematoxylin. galactosidase staining of flk-1^LacZ embryos was as described by Shalaby, et al., Nature 376, 62 (1995).

Methodology for combined -galactosidase staining and in situ hybridization

After 72 hr in culture, liver bud explants were fixed in PBS containing 4% paraformaldehyde on ice for 1 hr. Following two washes with PBS, cells were incubated in PBS containing 5 mM potassium ferricyanide, 5 mM potassium ferrocyanide, 1 ng/ml X-gal, 2 mM MgCl₂, and 0.02% NP-40 at 37°C overnight. After two washes with ice cold PBS, cells were dehydrated into methanol from PBS containing 0.1% Tween-20 (PBST). Dehydrated explants were stored at -20°C in methanol until use. For in situ hybridization, dehydrated cells were rehydrated into PBST, treated with proteinase K (10 g/ml in PBST) for 3 min, washed twice with glycine (2 mg/ml) in PBST, and postfixed in 0.2% glutaraldehyde / 4% paraformaldehyde in PBST for 20 min at room temp. Following PBST washes, embryos were rinsed with 1:1 prehybridization buffer (50% formamide, 1.3X SSC, 5 mM EDTA, 50 g / ml yeast transfer RNA, 0.2% Tween-20, 0.5% CHAPS, and 100 g/ml heparin): PBST and once with prehybridization buffer. Following prehybridization at 70°C with prehybridization buffer for 2 hr, cells were hybridized overnight with a digoxigenin-labeled antisense RNA probe for serum albumin mRNA (Cascio and Zaret, 1991) (approx. 1 g/ml) in prehybridization buffer at 70°C. Hybridized cells were rinsed with pre-warmed (70°C) prehybridization buffer, washed three times in prehybridization buffer, and washed in 1:1 prehybridization buffer / Tris buffered saline containing 0.1% Tween-20 (TBST) at 70°C for 30 min each. Following washes with TBST, cells were blocked with 2% blocking reagent (BBR) (Roche) in TBST for 1 hr and with 10% heat-inactivated normal goat serum (NGS) and 2% BBR in TBST for 2 hr. To prevent non-specific binding of antibody, the anti-digoxigenin Fab alkaline phosphatase conjugate (Roche) was preabsorbed as previously described (Sasaki and Hogan, 1993) and diluted 1:2000 with 10% NGS and 2% BBR in TBST. The cells were incubated overnight with preabsorbed antibody solution at 4°C. After rinsing with TBST, cells were washed six times in TBST for 1 hr each and then overnight. Following four 10 min washes in NTMTL (100 mM NaCl, 100 mM Tris HCl (PH 9.5), 50 mM MgCl₂, 1% Tween-20, and 2 mM Levamisole), color reactions were performed in BCIP/NBT solution (3.38 g/ml NBT (Roche), 1.75 g/ml BCIP (Roche) in NTMTL). The color development reactions were continued 2-24 hr with occasional changes of BCIP/NBT solution. For FastRed staining (Figure 4), after TBST washes the cells were washed in NTMTL (pH 8.0) and color reactions were performed in a FastRed solution (Sigma). The color development reactions went 1-6 hr and were stopped by washes with PBS.

Cell areas were measured with NIH Image 1.62 computer software (http://rsb.info.nih.gov/nih-image/). Data for each experiment were presented as mean ± SD of indicated number of experiments, and subjected to a paired Student's t-test.

Methodology for NK4 angiogenesis inhibitor

Human NK4 (or HGF/NK4) was purified from elastase-treated human hepatocyte growth factor (HGF) as described (Date et al., FEBS Lett. 420, 1 (1997). Anti-rat HGF antibody was raised in rabbit by immunizing with rat recombinant HGF and IgG was purified from the antisera using protein A-Sepharose. One µg/ml anti-rat HGF IgG almost completely neutralizes the biological activity of 1 ng/ml rat and mouse HGF (Sato et al., Nature 376, 70 (1995); Barros et al., Proc. Nat. Acad. Sci. USA 92, 4412(1995)). Human recombinant HGF was kindly provided by Dr. Toshikazu Nakamura (Osaka University).