Spontaneous corneal hem- and lymphangiogenesis in mice with destrin-mutation depend on VEGFR3 signaling

Abstract

Lymphangiogenesis, the formation of new lymphatic vessels, is important for tumor metastasis and induction of immunity to peripheral antigens including organ transplants. We herein describe a novel mouse model of spontaneous, secondary lymphangiogenesis in the normally avascular cornea. corn1 mice, which suffer from a deletion in the gene encoding the cytoskeletal protein destrin, develop hemangiogenesis as well as spontaneous outgrowth of LYVE-1+++/CD31+ lymphatic vessels into the cornea starting at age 4 weeks. Corneal lymphangiogenesis is delayed in onset, is less intense, and regresses earlier compared with hemangiogenesis. Moreover, the lymphangiogenesis is preceded only by a mild recruitment of CD45+ inflammatory cells into the cornea. In contrast to mice with inflammation-induced hem- and lymphangiogenesis, corn1 mice do not develop breakdown of the blood-aqueous barrier. Finally, in this novel mouse model, a blocking anti-VEGFR3 antibody significantly inhibited not only lymph- but also hemangiogenesis. In summary, destrin deletion has differential effects on spontaneous hem- and lymphangiogenesis in the normally avascular cornea and represents a novel mouse model to study the mechanisms of lymphangiogenesis and to test the antihem- and antilymphangiogenic properties of known or new antiangiogenic agents.

Figure 1

Spontaneous hem- and lymphangiogenesis in mice with a destrin gene deletion (corn1 mice). A: Representative slit-lamp aspect of the cornea of an 8-week-old corn1 mouse with blood vessels growing into the normally avascular cornea (hemangiogenesis). B: Double immunostaining of the cornea using LYVE-1 (red; arrow: L) as a specific lymphatic endothelial marker and CD31 (green; arrowhead: B) as a blood vascular marker demonstrates both hem- and lymphangiogenesis into the cornea (magnification, ×40). C: Detail form B with abundant blood (B) and some lymphatic (L) vessels (magnification, ×100; limbus at the left: transition zone from physiologically vascularized conjunctiva to normally avascular cornea).

Figure 2

Delayed onset of spontaneous lymphangiogenesis compared with hemangiogenesis in corn1 mice (representative segments from corneal flat mounts stained with LYVE-1 [red] and CD31 [green]; magnification, ×100): A: At postnatal day 7, neither blood nor lymphatic vessels grow out from the limbal arcade (Li, limbus). B: At 2 weeks of age, very small sprouts of blood vessels (arrowhead: B) grow into the cornea. C: At 3 weeks, a robust hemangiogenesis (green) can be observed with lymphangiogenesis showing initial minor outgrowth. D: At 6 weeks, the whole cornea is occupied by blood vessels and partly covered by lymphatic vessels (arrow: L).

Figure 3

Spontaneous, secondary lymphangiogenesis in corn1 mice is less extensive and regresses earlier compared with hemangiogenesis. Depicted are the relative proportions of the normally avascular cornea covered by blood (black bars) versus lymphatic vessels (white bars; mo., months; P < 0.001 at all time points).

Figure 4

Absence of the massive corneal inflammation (normally associated with inflammatory corneal neovascularization) in corn1 mice developing spontaneous hem- and lymphangiogenesis compared with wild-type controls (and “positive” inflamed controls). A: Numbers of CD45⁺ cells per corneal section comparing wild-type controls (age, 4 weeks), corn1 mice (age, 4 weeks), and experimentally vascularized corneas (positive control; inflammation-associated vascularization in corneas 14 days after mouse corneal transplantation). B through D: Representative segments of central cornea from inflamed vascularized corneas (B; positive control with numerous inflammatory cells; ×100), 4-week-old corn1 mice (C) and same-aged wild-type control (D) both showing absence of significant stromal inflammatory cell influx (H&E; magnification, ×100). E and F: CD45 immunohistochemistry (red, with DAPI blue counterstaining of nuclei) demonstrates very few CD45⁺ cells in a wild-type cornea (F) compared with significantly more, yet still few, CD45⁺ cells in a representative corn1 mouse cornea (E; magnification, ×100).

Figure 5

Absent destrin immunoreactivity of limbal blood and lymphatic vessels in wild-type control mice. A: Positive immunoreactivity of limbal blood and lymphatic vessels with CD31/PECAM1 (magnification, ×100; green). B: Absent immunoreactivity with destrin antibody at the same location (magnification, ×100; red). C: Positive immunoreactivity of central corneal epithelium with the destrin antibody (magnification, ×100; red; DSTN/CFN).

Figure 6

Absent differences in corneal mRNA levels of angiogenic growth factors of the VEGF family and VEGF receptor 3 (VEGFR3) between wild-type controls (WT; lane 1) and corn1 mice (lane 2; A through F). G and H: Representative sections from central cornea demonstrate equal expression of VEGFR3 protein in corneas from wild-type controls (G; WT) and corn1 mice (H). VEGFR3 is primarily expressed on corneal epithelial cells (arrows) and in the vascularized corn1 cornea in addition to new stromal blood and lymphatic vessels (arrowhead; magnification, ×400). Immunohistochemistry for VEGF-C, VEGF-D, and VEGF-A revealed no significant difference in expression between wild-type controls and corn1 mice.

Figure 7

Systemic application of a neutralizing anti-VEGFR3 antibody (mF4-31C1) significantly inhibits spontaneous hemangiogenesis as well as lymphangiogenesis in the corn1 mouse model. Whereas corn1 mice receiving control IgG injections intraperitoneally at day 14 and 21 of age display a robust hem- and lymphangiogenesis at age 5 weeks (A: whole mount of cornea stained with LYVE-1 [red] and CD31 [green], magnification, ×40; C: detail from A, magnification, ×100), there is significant inhibition of hemangiogenesis and nearly complete inhibition of lymphangiogenesis in corn1 mice that received anti-VEGFR3 antibodies at the same time points (B, D, and E: morphometrical analysis; Li, limbal vascular arcade).