Promotion of corneal angiogenesis by sensory neuron-derived calcitonin gene-related peptide

Abstract

The normal cornea has no blood vessels but has abundant innervation. There is emerging evidence that sensory nerves, originated from the trigeminal ganglion (TG) neurons, play a key role in corneal angiogenesis. In the current study, we examined the role of TG sensory neuron-derived calcitonin gene-related peptide (CGRP) in promoting corneal neovascularization (CNV). We found that CGRP was expressed in the TG and cultured TG neurons. In the cornea, minimal CGRP mRNA was detected and CGRP immunohistochemical staining was exclusively co-localized with corneal nerves, suggesting corneal nerves are likely the source of CGRP in the cornea. In response to intrastromal suture placement and neovascularization in the cornea, CGRP expression was increased in the TG. In addition, we showed that CGRP was potently pro-angiogenic, leading to vascular endothelial cell (VEC) proliferation, migration, and tube formation in vitro and corneal hemangiogenesis and lymphangiogenesis in vivo. In a co-culture system of TG neurons and VEC, blocking CGRP signaling in the conditioned media of TG neurons led to decreased VEC migration and tube formation. More importantly, subconjunctival injection of a CGRP antagonist CGRP8-37 reduced suture-induced corneal hemangiogenesis and lymphangiogenesis in vivo. Taken together, our data suggest that TG sensory neuron and corneal nerve-derived CGRP promotes corneal angiogenesis.

Keywords: Calcitonin gene-related peptide; Corneal angiogenesis; Corneal innervation; Neovascularization; Neuropeptide; Trigeminal ganglion neuron.

Fig. 1.. Expression of CGRP and its receptor.

(A–C) Representative images of CGRP (red) and β-tubulin III (nerve marker, green) double immunofluorescence showing the distribution of CGRP in the trigeminal ganglion (TG, A), cornea (B) and cultured TG neurons (C); arrows showing representative double-positive neurons/nerves. (D) CGRP positivity was calculated as percentage of CGRP-positive versus total β-tubulin III-positive neurons or nerves (n = 3). (E) CGRP mRNA expression levels in TG and cornea, (n = 3). (F) Corneal intrastromal suture was placed in the mouse cornea to induce corneal neovascularization (CNV) for 1 week. The expression of CGRP in TG and cornea and the expression of CGRP receptor CLR and RAMP1 in the cornea were determined using RT-PCR (n = 6) and expressed as number of copies in log scale per 10⁶ GAPDH (internal control). *p < 0.05 unpaired t-test. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 2.. CGRP promotes vascular endothelial cell activities in vitro.

VEC were treated with or without exogenous CGRP (100 nM, n = 5). (A) VEC proliferation was determined after 24hr incubation. (B, C) Migration of VEC was quantified and photographed at 0hr and 24hr. (D-F) Tube formation in the Matrigel (F), was quantified as the number of junctions (D), and length of the tubes formed (E); black arrowheads indicating non-tube VEC, while the white arrows showing VEC tube formation. **p < 0.01, ****p < 0.0001, unpaired t-test.

Fig. 3.. CGRP promotes corneal hemangiogenesis and lymphangiogenesis in vivo.

A single figure-eight intrastromal suture was placed on mouse cornea and CGRP (50μΜ) or PBS control was given topically three times daily for 14 days (n = 5). (A) Representative images of slit lamp photography. (B) Representative images of corneal whole mount stained with CD31 (Blood vessels, green) and LYVE-1 (Lymph vessels, red), × 100 magnification, dashed line indicates the limbus. (C, D) Quantification of corneal neovascularization (CNV) area (CD31, C) and Lymph vessel area (LYVE-1, D) at day 14 after. (E) mRNA level of VEGF-A in the cornea was assessed. *p < 0.05, **p < 0.01, ****p < 0.0001, unpaired t-test. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4.. Blocking CGRP signaling in the conditioned media (CM) of trigeminal ganglion (TG) neurons suppresses VEC activities in vitro.

(A) Representative image of TG neurons cultured for 3 days then stained with β-tubulin III. (B) Schematic of the co-culture system, in which CM from cultured TG neurons were collected and applied to VEC culture. CGRP antagonist CGRP8-37 (100 nM) was added to VEC culture (n = 4). (C, D) Migration of VEC was quantified and photographed at 0hr and 24hr. (E–G) Tube formation in the Matrigel was photographed (G), and quantified as the number of junctions (E), and length of the tube formed (F); black arrowheads indicating non-tube VEC, while the white arrows showing VEC tube formation. *p < 0.05, one-way ANOVA followed by Tukey’s multiple comparisons test.

Fig. 5.. Blocking CGRP signaling inhibits corneal hemangiogenesis and lymphangiogenesis in vivo.

A single figure-eight intrastromal suture was placed on mouse cornea and CGRP8-37 (200 μg/ml) or PBS control was administered via subconjunctival injection three times/week for two weeks (n = 5). (A) Representative images of corneal neovascularization (CNV) with slit-lamp photography. (B) Representative images of corneal whole mount stained with CD31 (Blood vessels, green) and LYVE-1 (Lymph vessels, red), × 100 magnification, dashed line indicates the limbus. (C, D) Quantification of CNV area (CD31, C) and Lymph vessel area (LYVE-1, D) at day 14. (E) mRNA level of VEGF-A in cornea was assessed. *p < 0.05, **p < 0.01, unpaired t-test. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)