Equine Endothelial Cells Show Pro-Angiogenic Behaviours in Response to Fibroblast Growth Factor 2 but Not Vascular Endothelial Growth Factor A

Abstract

Understanding the factors which control endothelial cell (EC) function and angiogenesis is crucial for developing the horse as a disease model, but equine ECs remain poorly studied. In this study, we have optimised methods for the isolation and culture of equine aortic endothelial cells (EAoECs) and characterised their angiogenic functions in vitro. Mechanical dissociation, followed by magnetic purification using an anti-VE-cadherin antibody, resulted in EC-enriched cultures suitable for further study. Fibroblast growth factor 2 (FGF2) increased the EAoEC proliferation rate and stimulated scratch wound closure and tube formation by EAoECs on the extracellular matrix. Pharmacological inhibitors of FGF receptor 1 (FGFR1) (SU5402) or mitogen-activated protein kinase (MEK) (PD184352) blocked FGF2-induced extracellular signal-regulated kinase 1/2 (ERK1/2) phosphorylation and functional responses, suggesting that these are dependent on FGFR1/MEK-ERK signalling. In marked contrast, vascular endothelial growth factor-A (VEGF-A) had no effect on EAoEC proliferation, migration, or tubulogenesis and did not promote ERK1/2 phosphorylation, indicating a lack of sensitivity to this classical pro-angiogenic growth factor. Gene expression analysis showed that unlike human ECs, FGFR1 is expressed by EAoECs at a much higher level than both VEGF receptor (VEGFR)1 and VEGFR2. These results suggest a predominant role for FGF2 versus VEGF-A in controlling the angiogenic functions of equine ECs. Collectively, our novel data provide a sound basis for studying angiogenic processes in horses and lay the foundations for comparative studies of EC biology in horses versus humans.

Keywords: angiogenesis; comparative; endothelial; equine.

Figure 1

Optimisation of EAoEC isolation and culture. (a) Left: mechanical scraping yielded a higher proportion of EC-rich cultures (defined as >65% VWF-positive cells) than collagenase digestion (**** p < 0.0001; Fisher’s exact test). Data are from a total of 63 isolations. Right: representative images of EAoEC isolates fixed and processed for VWF immunofluorescence (green). (b) Left: EAoECs proliferated at a faster rate when cultured in complete EGM2 than in M199 (* p = 0.03; t-test; mean ± S.E.M for n = two biological replicates with three technical repeats). Right: EAoECs were fixed, and nuclei were stained (Hoechst) for cell counting after 72 h. (c) Left: EAoECs were cultured in complete EGM2 or EGM2 without additives (EBM), and cell morphology was evaluated via light microscopy (* p = 0.02; t-test; mean ± S.E.M, n = three isolates). Right upper: representative phase contrast images of EAoECs cultured to confluence in complete EGM2 or EBM. Right lower: representative images of EAoEC cultured in complete EGM2 or EBM, fixed and processed for VWF immunofluorescence (green). (d) Final EAoEC isolation strategy: equine aortas were cleaned of connective and adipose tissue via blunt dissection and incised longitudinally between the paired intercostal artery openings. The luminal surface was scraped with the back of a sterile scalpel blade. The accumulated material on the scalpel blade was transferred to a sterile 15 mL centrifuge tube and incubated with collagenase solution to dissociate the material into individual cells. The cell pellet was transferred to a gelatin-coated tissue culture flask (×10 magnification) and cultured to confluence. Figure created with BioRender.com (https://app.biorender.com/user/signin, accessed 14 September 2023).

Figure 2

Purification of EAoEC populations. EAoECs were purified using magnetic cell sorting. (a) High-power confocal (×63; left) and low-power widefield (×20; right) images of EAoECs immunostained for cytoplasmic VWF (green) and membranous VE-cadherin (red). (b) Maximum projection of a z-stack confocal image of equine intercostal artery en face showing immunodetection of cytoplasmic VWF (green) and membranous VE-cadherin (red). (c) Live EAoECs were incubated with a PE-conjugated anti-VE-cadherin antibody and analysed via flow cytometry for surface VE-cadherin expression. VE-cadherin-stained cells (green frequency distribution) exhibited a brighter fluorescent signal than cells incubated with the isotype control antibody (white frequency distribution). (d) Unsorted cells were subjected to magnetic cell sorting using magnetic beads conjugated to an anti-VE-cadherin antibody, resulting in the positive population showing the classic cobblestone endothelial morphology and the negative population showing morphology consistent with vascular smooth muscle cells (created with BioRender.com (https://app.biorender.com/user/signin, accessed 14 September 2023)). (e) The proportion of VWF-positive cells was greater in the sorted population than the unsorted population, as assessed by immunofluorescence (** p = 0.008; t-test; mean ± S.E.M, n = 3).

Figure 3

FGF2 is a potent pro-angiogenic factor for EAoECs. (a) EAoECs were exposed to FGF2 (10 ng/mL), epithelial growth factor (EGF; 10 ng/mL), insulin-like growth factor (IGF; 50 ng/mL), or VEGF-A (25 ng/mL) for 16 h, and tube formation was analysed by manually counting the number of branches (* p = 0.02; repeated measures one-way ANOVA with Dunnett’s multiple comparisons test compared to control; mean ± S.E.M, n = 4). Images are representative of tubulogenesis induced by FGF-2 at 16 h (×10 magnification). (b) Effects of growth factors on EAoEC proliferation rate over 72 h (* p = 0.04, ** p = 0.009, *** p = 0.0007; one-way ANOVA with Dunnett’s multiple comparisons test compared to control; mean ± S.E.M for n = three technical replicates). (c) EAoECs were exposed to growth factors for 10 min and ERK phosphorylation assessed by immunoblotting using a phospho-specific ERK1/2 antibody. Blots were stripped and re-probed for total ERK1/2, and all blots analysed by densitometry. Data are mean ± S.E.M (n = four independent experiments on separate EAoEC isolates). ** p = 0.007; repeated measures one-way ANOVA with Dunnett’s multiple comparisons test compared to control. Representative cropped immunoblots are shown. Concentrations of growth factors used were consistent across experiments.

Figure 4

Concentration-response characteristics of EAoECs to equine recombinant FGF2. Tubulogenesis (panel (a); n = 5), proliferation (panel (b); n = 3), scratch wound closure (panel (c); n = 3), and ERK1/2 phosphorylation at 10 min (panel (d); n = 3) were evaluated in response to the indicated concentrations of FGF2. The concentration ranges used were designed to identify the maximal response. In the scratch wound closure experiments, the FGF2 concentration was increased to 100 ng/mL after preliminary experiments, indicating that concentrations below 50 ng/mL did not lead to a maximal response. Representative cropped immunoblots are shown. The images in panel c are representative and show control and FGF2-stimulated wounds after 18 h (×5 magnification). Data are mean ± S.E.M and were analysed via repeated measures one-way ANOVA with Dunnett’s multiple comparisons test (**** p < 0.0001, *** p < 0.001, ** p < 0.01, * p < 0.05 versus unstimulated control).

Figure 5

The pro-angiogenic effects of FGF2 are mediated by FGFR1 and MEK-ERK signalling. Tube formation ((a); n = two biological replicates with six technical repeats; FGF2, 20 ng/mL), scratch wound closure ((b); n = three biological replicates; FGF2, 100 ng/mL), proliferation ((c); n = four biological replicates; FGF2, 5 ng/mL), and ERK/2 phosphorylation ((d); n = three biological replicates; FGF2, 5 ng/mL) were measured in control and FGF2-stimulated (at a concentration to give maximal response (see Figure 4)) EAoECs in the absence or presence of SU5402 (FGFR1 inhibitor, 10 µM) or PD184352 (MEK1/2 inhibitor, 10 µM). Representative images are shown for tubulogenesis (×10 magnification) (a) and scratch wound closure (×5 magnification) (b), as well as representative immunoblots for pERK, total ERK expression, and β actin (d). All data are given as mean ± S.E.M and were analysed via repeated measures one-way ANOVA with Sidak’s multiple comparisons test. (* p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001 for inhibitor + FGF2 compared to FGF2-stimulated).

Figure 6

VEGF-A does not exert significant pro-angiogenic effects on EAoECs. EAoECs were exposed to FGF2 or VEGF-A at the indicated concentrations and (a) tube formation (16 h), (b) scratch wound closure (18 h), and (c) proliferation (72 h), assessed as described in the Materials and Methods Section 4. (a) Mean ± S.E.M., n = 4 (** p = 0.008 versus unstimulated control); (b) mean ± S.E.M. for n = 2; (c) mean ± S.E.M., n = 3 (*** p = 0.0003 versus unstimulated control). (d) ERK phosphorylation (10 min) was assessed via immunoblotting using a phospho-specific ERK1/2 antibody and β-actin as a loading control; blots were analysed via densitometry. Data are mean ± S.E.M. for n = 3 (*** p = 0.0006 versus unstimulated control). All statistical analysis used repeated measures one-way ANOVA with Dunnett’s multiple comparisons test.

Figure 7

Growth factor receptor expression in equine versus human endothelial cells. Receptor expression was measured in unstimulated EAoECs, HAoECs, and HUVECs using qPCR and quantified relative to β-actin expression in the same cell type. (a) Relative FGFR1 expression in EAoECs is significantly greater than FGFR2 (* p = 0.01), VEGFR1 (* p = 0.01), and VEGFR2 (* p = 0.01). Mean ± S.E.M., n = 3). (b,c) Relative FGFR1 expression is significantly lower than NRP1 in HAoECs (*** p = 0.0002) and HUVECs (** p = 0.006). Data given are mean ± S.E.M of three and four biological repeats, respectively. All statistical analyses used repeated measures one-way ANOVA with Dunnett’s multiple comparisons test. (NS = not significant versus FGFR1, * p < 0.05, ** p < 0.01; *** p < 0.001).