VEGF-C alters barrier function of cultured lymphatic endothelial cells through a VEGFR-3-dependent mechanism

Abstract

Background: The lymphatic endothelium is an important semi-permeable barrier separating lymph from the interstitial space. However, there is currently a limited understanding of the lymphatic endothelial barrier and the mechanisms of lymph formation. The objectives of this study were to investigate the potential active role of lymphatic endothelial cells in barrier regulation, and to test whether the endothelial cell agonists VEGF-A and VEGF-C can alter lymphatic endothelial barrier function.

Methods and results: Cultured adult human dermal microlymphatic endothelial cells (HMLEC-d) and human umbilical vein endothelial cells (HUVEC) were respectively used as models of lymphatic and vascular endothelium. Transendothelial electrical resistance (TER) of endothelial monolayers served as an index of barrier function. Cells were treated with VEGF-A, VEGF-C, or the VEGFR-3 selective mutant VEGF-C156S. MAZ51 was used to inhibit VEGFR-3 signaling. The results show that while VEGF-A causes a time-dependent decrease in TER in HUVEC, there is no response in HMLEC-d. In contrast, VEGF-C and VEGF-C156S cause a similar decrease in TER in HMLEC-d that is not observed in HUVEC. These results corresponded to the protein expression of VEGFR-2 and VEGFR-3 in these cell types, determined by Western blotting. In addition, the VEGF-C- and VEGF-C156S-induced TER changes were inhibited by MAZ51.

Conclusions: The results indicate differential responses of the lymphatic and vascular endothelial barriers to VEGF-A and VEGF-C. Furthermore, our data suggest that VEGF-C alters lymphatic endothelial function through a mechanism involving VEGFR-3.

Fig. 1

Expression of lymphatic endothelial markers in cultured HMLEC-d. The green labeling in panels A and B respectively show localization of prox-1 and LYVE-1. Panel C shows VEGFR-3 labeling in red. Podoplanin is labeled green in panel D. Panels E and F respectively show VE-cadherin and PECAM-1 labeled in red. Blue reflects nuclear labeling by Hoechst 33342. All images are representative of at least three experiments. Bar = 20 μm.

Fig. 2

VEGF-A alters barrier function in HUVEC, but not in HMLEC-d. A. Tracings of the time-courses of HUVEC and HMLEC-d TER before and during exposure to 1 nM VEGF-A are shown. The tracings represent the averages of N=4 wells each. Panel B shows a comparison of the mean fractional change in TER 1 min. after VEGF-A was added. **P<0.01 HUVEC vs. HMLEC-d.

Fig. 3

VEGF-C causes differential changes in TER in HMLEC-d and HUVEC. A. The tracings show average TER of HMLEC-d and HUVEC before and during treatment with 10 nM VEGF-C (N=4 each). Panel B shows the HMLEC-d and HUVEC TER before and during treatment with 10 nM VEGF-C156S (N=4 each). C. The mean fractional changes in TER 30 minutes after the addition of VEGF-C or VEGF-C156S are shown. *P<0.05 between the indicated groups.

Fig. 4

HMLEC-d and HUVEC differentially express VEGFR-2 and VEGFR-3. The upper blot shows a strong band for VEGFR-2 in HUVEC and a very weak band in HMLEC-d (arrowhead). The lower blot shows strong bands for VEGFR-3 (arrows) in the HMLEC-d lane and one weak VEGFR-3 band in the HUVEC lane (middle arrow). The bands at ~195, ~175, and ~125 kDa reflect the uncleaved, mature form, a unglycosylated precuror, and cleaved form of VEGFR-3, respectively, . The blots are representative of three separate experiments.

Fig. 5

The time-course and dose-response characteristics of VEGF-C- and VEGF-C156S-induced TER changes in HMLEC-d are similar A. No significant differences in the time courses of VEGF-C and VEGF-C156S were apparent. N=4 for both treatments. B. VEGF-C and VEGF-C156s also showed very similar dose response characteristics, with both curves in the same order of magnitude and with VEGF-C156S having a similar E_max value and VEGF-C. The estimated EC50s for VEGF-C and VEGF-C156S are 4.13 and 2.25 nM, respectively. There were no significant differences between the responses to VEGF-C and VEGF-C156S at each concentration tested. N=4 for each concentration tested.

Fig. 6

Inhibition of VEGFR-3 blocks VEGF-C156S-induced changes in TER in HMLEC-d. MAZ51 (5 μM) was added 30 minutes prior to VEGF-C156S (10 nM), and itself did not alter TER. However, MAZ51-treated HMLEC-d did not display decreased TER after treatment with VEGF-C156S, unlike cells that were not treated with MAZ51. The tracings show the average TER from both groups. N=4 for both groups.