Free
Review  |   November 2014
Transgenic Mouse Models of Corneal Neovascularization: New Perspectives for Angiogenesis Research
Author Notes
  • Department of Vascular Biology and Tumor Angiogenesis, Center for Biomedicine and Medical Technology Mannheim (CBTM), Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany and Division of Vascular Oncology and Metastasis, German Cancer Research Center (DKFZ-ZMBH Alliance), Heidelberg, Germany 
  • Correspondence: Jens Kroll, Center for Biomedicine and Medical Technology Mannheim (CBTM), Department of Vascular Biology and Tumor Angiogenesis, Medical Faculty Mannheim, Heidelberg University, Ludolf-Krehl-Str. 13-17, 68167 Mannheim, Germany; jens.kroll@medma.uni-heidelberg.de
Investigative Ophthalmology & Visual Science November 2014, Vol.55, 7637-7651. doi:10.1167/iovs.14-15430
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Jakob Nikolas Kather, Jens Kroll; Transgenic Mouse Models of Corneal Neovascularization: New Perspectives for Angiogenesis Research. Invest. Ophthalmol. Vis. Sci. 2014;55(11):7637-7651. doi: 10.1167/iovs.14-15430.

      Download citation file:


      © 2016 Association for Research in Vision and Ophthalmology.

      ×
  • Supplements
Abstract

Corneal neovascularization (NV) refers to the growth of blood vessels and/or lymphatics into the physiologically avascular cornea, which occurs in several pathological processes. In mouse models, corneal NV can be artificially induced to investigate mechanisms of corneal pathologies. However, mouse models of corneal NV are not restricted to cornea-specific research, but also are widely used to investigate general mechanisms of angiogenesis. Because the cornea is transparent and easily accessible, corneal NV models are among the most useful in vivo models in angiogenesis research.

The three different approaches that are used to study corneal NV in mice are based on direct application of proangiogenic or antiangiogenic transmitters, external injury to the cornea, or genetically engineered mice, which spontaneously develop corneal NV. The aim of this review is to compare the scope and limitations of the different approaches for corneal NV in mice. Our main focus is to highlight the potential of transgenic spontaneous models of corneal NV. Transgenic models do not require any experimental interference and make it possible to investigate different interconnected proangiogenic signaling cascades. As a result, transgenic models are highly useful for disease-centered angiogenesis research.

In summary, transgenic models of corneal NV will complement and advance existing ocular NV assays, and help to discover new angiogenesis-related treatment strategies for ocular and extraocular diseases.

THE MOUSE EYE AS A MODEL FOR STUDYING ANGIOGENESIS
Many human diseases are directly or indirectly correlated to pathological angiogenesis, for example cancer, ischemic heart diseases, or inflammatory diseases.1 An intricate cascade of many different steps regulates angiogenesis. Some parts of the angiogenic cascades can be efficiently studied in vitro, for example by endothelial cell sprouting assays. However, a plethora of organ-specific or tissue-specific interactions are not adequately modeled by in vitro approaches.2 Therefore, in vivo models, especially mouse models, are an essential part of the toolbox of angiogenesis researchers. 
Mechanisms of Corneal Avascularity
The cornea is attractive for vascular research because it is one of very few tissues that are completely devoid of blood vessels and lymphatics under physiological conditions.3 Corneal cells are supplied with nutrients and oxygen from the tear film, aqueous humor in the anterior chamber, and blood vessels surrounding the cornea at the limbus. Functionally, corneal avascularity is important for clarity of vision and pathological corneal neovascularization (NV) is directly associated with loss of visual acuity.4,5 Corneal NV comprises two different processes: hemangiogenesis (HA) and lymphangiogenesis (LA). Corneal HA and LA are detectable and are mechanistically involved in pathological cascades in several sight-threatening diseases, like corneal infections, corneal allograft rejection, or hereditary corneal dystrophies.57 Molecularly, corneal avascularity is maintained by a number of mechanisms. Using in vivo experiments, several mediators have presented antiangiogenic properties in the cornea. Among them are pigment epithelium derived factor (PEDF), angiostatin, a plasminogen fragment, and endostatin.5,8,9 Others have systematically characterized and reviewed several other corneal antiangiogenic factors, such as restin, arresten, tumstatin, neostatin, and canstatin.3,5,8,9 Expression of most of these molecules and/or their receptors has been demonstrated in physiological avascular corneas, suggesting an involvement of these molecules in maintenance of corneal avascularity.9 Surprisingly, contrary to prior expectations, many experimentally tested inhibitors of corneal NV do not seem to be strictly necessary for corneal avascularity in vivo.10,11 Among the few remaining avascularity-related mechanisms are expression of soluble VEGF receptor 1 (VEGFR-1/sFlt1) and ectopic expression of VEGFR-3 (Flt-4) by the epithelium, which inhibit the growth of blood vessels into the corneal stroma.10,11 Additionally, VEGF expression in the cornea is transcriptionally suppressed by inhibitory PAS (IPAS), a negative regulator of angiogenesis which inhibits hypoxia-induced factor-1α (HIF-1α) signaling (Fig. 1A).12 
Figure 1
 
Regulation of physiological avascularity, induction of corneal NV and therapeutic intervention in the cornea. Whenever the balance of pro- and antiangiogenic transmitters in the cornea is disturbed, corneal NV occurs. In this Figure, the progression from physiological avascularity to complete vascularization is shown. Blue represents DAPI-positive epithelial cell nuclei, red lines represent lymphatics, green lines represent blood vessels. I, epithelium; II, Bowman's layer; III, stroma; IV, Descemet's membrane; V, endothelium. (A) A number of mediators (orange) actively maintain corneal avascularity. Green margin indicates the inhibition of blood vessel growth, red margin indicates the inhibition of lymphatic vessel growth. (B) The most relevant proangiogenic factors are shown in yellow. Various studies have shown that if the concentration of any one of these transmitters is experimentally elevated to a certain degree, blood vessels and/or lymphatics sprout from the corneal limbal plexus toward the corneal center. Also, macrophages (Mϕ) are involved in corneal NV. (C) Several externally applied substances can inhibit angiogenesis or drive vessels into regression. Some prominent examples are shown. TKI, tyrosine-kinase inhibitors; NSAID, nonsteroidal anti-inflammatory drugs.
Figure 1
 
Regulation of physiological avascularity, induction of corneal NV and therapeutic intervention in the cornea. Whenever the balance of pro- and antiangiogenic transmitters in the cornea is disturbed, corneal NV occurs. In this Figure, the progression from physiological avascularity to complete vascularization is shown. Blue represents DAPI-positive epithelial cell nuclei, red lines represent lymphatics, green lines represent blood vessels. I, epithelium; II, Bowman's layer; III, stroma; IV, Descemet's membrane; V, endothelium. (A) A number of mediators (orange) actively maintain corneal avascularity. Green margin indicates the inhibition of blood vessel growth, red margin indicates the inhibition of lymphatic vessel growth. (B) The most relevant proangiogenic factors are shown in yellow. Various studies have shown that if the concentration of any one of these transmitters is experimentally elevated to a certain degree, blood vessels and/or lymphatics sprout from the corneal limbal plexus toward the corneal center. Also, macrophages (Mϕ) are involved in corneal NV. (C) Several externally applied substances can inhibit angiogenesis or drive vessels into regression. Some prominent examples are shown. TKI, tyrosine-kinase inhibitors; NSAID, nonsteroidal anti-inflammatory drugs.
Regarding antilymphangiogenic mediators, evidence is similar: of many effective mediators from experimental studies, only thrombospondin-1 (TSP-1) and soluble VEGFR-2 (sFlk-1) have been shown to be indispensable for maintaining the corneal lymphangiogenic privilege in physiological settings.13,14 Recent advances in investigating corneal avascularity include a novel antiangiogenic system that is orchestrated by the transcription factor FoxC1.15 Also, it has been shown recently that crosstalk between developing nerve fibers and blood vessels inhibits blood vessel sprouting into the cornea (Fig. 1A).16 
Another aspect of corneal avascularity is closely linked to integrity of the corneal epithelium. Unlike the corneal endothelium, the corneal epithelium has an extensive capacity of regeneration upon injury. However, this regeneration relies on functional limbal stem cells (LSCs), which express ATP-binding cassette, subfamily B, member 5 (Abcb5).17,18 Differentiation of these cells into corneal epithelial cells is orchestrated by WNT7A–PAX6 signaling.19 If this differentiation of LSCs into corneal epithelial cells is disturbed, the corneal epithelium becomes metaplastic and displays several skin-like characteristics, for example keratinization, loss of transparency, and subepithelial corneal NV.19 Thus, a functionally intact epithelium is another prerequisite of corneal avascularity. 
Proangiogenic Signaling in the Cornea
Physiologically, antiangiogenic factors outmatch proangiogenic factors in the cornea. In several pathological processes, this equilibrium is disturbed and proangiogenic factors dominate the corneal signaling environment (Figs. 1A, 1B).3,5,20 In experimental models of corneal NV, this process is being used to induce HA or LA. 
Classical proangiogenic factors in the cornea are VEGF and fibroblast growth factor (FGF) isoforms.5 Also, platelet-derived growth factor (PDGF) isoforms,21 (placental growth factor) PLGF,22 angiopoietin-1 (Ang-1),23,24 matrix metalloproteinases (MMPs)8, and CCR2 ligands25 have proangiogenic effects in the cornea in vivo. Ellenberg et al.5 have reviewed further angiogenesis regulating mediators. For VEGF isoforms, TGF-β, IL-1, TNF-α, MMPs, IL-6, and IL-8, we know that they are involved in human corneal neovascular diseases, while for several other transmitters, this has not been shown so far.5,26,27 
Some of these proangiogenic factors induce corneal HA and corneal LA, while others are more specific for either HA or LA. For example, basic FGF (bFGF) induces LA more strongly than HA in corneal pocket assays.28,29 In addition to their different molecular regulation, HA and LA follow different spatial patterns. For example, in the bFGF corneal pocket assay, blood vessels and lymphatics grow toward the bFGF gradient from different sides of the cornea.30 Also, murine corneal HA and LA follow different temporal patterns, and different temporal stages of corneal NV can be regulated by different signaling mechanisms.24,31 Still, it is not entirely clear which transmitters are selective for either corneal HA or LA and which are unselective. The classical concept is that VEGF-A, by binding to VEGFR-2 (Flk-1), primarily acts as a hemangiogenic factor, while VEGF-C and VEGF-D primarily attract lymphatics by binding on VEGFR-3 (Flt-4). However, in the cornea, these processes are interwoven due to macrophage signaling.3 The VEGFR-3 specific ligands do not exclusively stimulate LA in the cornea, but also activate macrophages, which in turn secrete VEGF-A and promote HA.32 Conversely, another study showed that VEGF-A chemotactically attracts macrophages, which secrete VEGF-C and VEGF-D, and induce LA (Fig. 1B).33 Furthermore, it has even been suggested that macrophages themselves can transdifferentiate and integrate into the nascent lymphatic vasculature.34 
Corneal HA and LA is regulated by an intricate interplay between many different signaling pathways. Although many components of these pathways have been identified, their interaction and pathophysiological involvement must be further clarified by in vivo approaches. 
Regression of Blood Vessels
Regression of blood vessels, also called pruning, is an essential process during vessel maturation in embryonic and postnatal development leading to a functional and hierarchically organized vascular system.35 Hyaloid vessels of the eye are a prime example for a vessel type that is temporally formed during embryonic development, but regresses at later stages of development. Although it still is not fully clear whether vessel regression is regulated by an active genetic program or whether it is simply due to withdrawal of prosurvival angiogenic factors that usually maintain the vasculature, pruning is associated with apoptotic endothelial cells.35 In addition, reduced flow in blood vessels36 induces pruning and several molecular mediators, including Rac1,36 FGD5,37 and the Wnt secretion factor Evi38 have been identified in this process. In the cornea, molecular mechanisms of vessel regression are so far not well understood. Corneal blood vessels can regress after photodynamic39 or antiangiogenic therapy,40 whereas corneal lymphatic vessels form and regress spontaneously within the first weeks after birth.41 Likewise, lymphatic vessels in adult mice can regress upon corneal recovery after an inflammatory response, which may partially be mediated by MMP-10 and TGF-β.42 Yet, also in the cornea it remains unclear whether blood and lymphatic vessel regression can be induced by a specific genetic program. 
Therapeutic Inhibition of Corneal NV
The VEGF-A was among the first proangiogenic transmitters to be isolated and is widely viewed as the principal proangiogenic transmitters in pathological NV. In the clinic, this has led to the introduction of VEGF-A–inhibiting drugs used to inhibit tumor angiogenesis or ocular NV.1,20 However, initial euphoria about this new therapeutic approach soon wore off as it turned out that VEGF inhibition alone is not enough to effectively inhibit tumor angiogenesis.1 In the treatment of corneal NV, VEGF-targeting therapy has been shown experimentally to induce a reduction of pathological corneal NV. Approaches for anti-VEGF therapy of corneal NV have been subject of two recent in-depth reviews by Bock et al.3 and Chang et al.20 Some limitations to anti-VEGF therapy have been described. For example, some studies report only a partial efficacy of experimental anti-VEGF therapy and others report neurotoxicity as a limiting factor of these agents.3,20 Another therapeutic compound is aganirsen, an antisense oligonucleotide targeting insulin receptor substrate-1, which has very recently demonstrated safety and efficacy in a phase III study (Fig. 1C).43 
Experimentally-Induced Angiogenesis in the Eye
For several decades, experimentally-induced angiogenesis in the eye has provided a framework for standardized assessment of angiogenesis in vivo. As early as 1974, the rabbit cornea has been used as an experimental model to test proangiogenic substances in a corneal pocket assay.44 In this procedure, a micropocket is cut into the cornea and a pellet containing a proangiogenic factor is implanted into this pocket. This substance then is slowly liberated over the course of several days and induces vessel sprouting from the limbus.45 The classical corneal pocket assay remains a cornerstone of in vivo angiogenesis research (Table 1). It has been supplemented by two different types of ocular angiogenesis models, namely models of retinal angiogenesis and choroidal NV. A strength of retinal models is that physiological angiogenesis during development can be assessed in detail in this context. Much of the knowledge we have of the complicated interplay between tip cells and stalk cells in sprouting angiogenesis stems from retinal angiogenesis models. Also, some pathological states of hypervascularization can be mimicked in the mouse retina.46 Moreover, transgenic models of retinal hypervascularization can be used to investigate spontaneous pathological NV without experimental interference.46 Another ocular angiogenesis model relies on choroidal NV (CNV), which can be externally induced or can arise spontaneously in transgenic mice.47 
Table 1
 
Recent Corneal Pocket Assay Studies
Table 1
 
Recent Corneal Pocket Assay Studies
Mechanism Intervention Reference Model
Inhibition of HA by depletion of natural killer cells; reduced expression of VEGF-A, VEGF-C, and VEGFR3 Anti-NK1.1 antibody 56 bFGF pellet
Inhibition of HA in FMOD−/− mice FMOD−/− mice 109 bFGF pellet
Increased VEGF induced HA in Tie2-CYP2C8-Tr mice Tie2-CYP2C8-Tr mice 110 VEGF pellet
Inhibition of VEGF induced HA H-RN derived peptide from hepatocyte growth factor kringle 1 domain (HGF K1) 58 VEGF pellet
Lcn2 enhances VEGF induced HA Lipocalin 2 (Lcn2) protein administration 111 VEGF pellet
Protein tyrosine kinase 7 (PTK7) forms complex with VEGFR-1, inhibition of HA after PTK7 silencing PTK7 siRNA injection 112 VEGF pellet
Semaphorin 7A, inhibition of HA Sem7A expression vector 113 bFGF pellet
In recent years, this multitude of ocular experimental angiogenesis assays has been complemented by two different assays in the mouse cornea. Over time, the number of studies investigating corneal NV in mouse models has rapidly increased (Fig. 2). The corneal NV models that fueled this increase are corneal injury models (Table 2) and, most recently, spontaneous, transgenic models of corneal NV (Table 3). Transgenic models have long been used in retinal angiogenesis research, but have only recently been introduced in the corneal setting. We propose that with the advent of new models of corneal NV, the mouse cornea can efficiently complement other ocular angiogenesis models to assess angiogenesis in the eye, accelerating in vivo angiogenesis research. 
Figure 2
 
An exponentially increasing number of studies investigates corneal NV In this plot, the cumulative number of new published studies on corneal NV is shown for the last 50 years. The cornea is emerging as a widely-used model to study angiogenesis and the number of new studies has increased strongly over time. Cumulative data, plotted for each year, not smoothed. Source: Pubmed (National Center for Biotechnology Information [NCBI], National Institutes of Health [NIH], Bethesda MD, USA), search query “cornea AND (neovascularization OR angiogenesis).”
Figure 2
 
An exponentially increasing number of studies investigates corneal NV In this plot, the cumulative number of new published studies on corneal NV is shown for the last 50 years. The cornea is emerging as a widely-used model to study angiogenesis and the number of new studies has increased strongly over time. Cumulative data, plotted for each year, not smoothed. Source: Pubmed (National Center for Biotechnology Information [NCBI], National Institutes of Health [NIH], Bethesda MD, USA), search query “cornea AND (neovascularization OR angiogenesis).”
Table 2
 
Non-VEGF–Based Strategies to Inhibit Corneal NV in Injury Models
Table 2
 
Non-VEGF–Based Strategies to Inhibit Corneal NV in Injury Models
Mechanism/Molecules HA LA Intervention Reference Model
Inhibition of VEGF, bFGF, TGF-β1, and EGF; upregulation of Tsp-1, Tsp-2, and ADAMTS-1 Largazole (class I HDAC inhibitor) 114 Alkali
CXCR4, CXCR7, and others TC14012 (CXCR4 antagonist and CXCR7 agonist) 115 Alkali
Inhibition of HA and LA in angiopoietin-2−/− mice Ang-2−/− mice 116 Suture
C57BL/6 and FVB mice N/A 63 Alkali and suture
TNF-α Anti–TNF-α antibody 64 Alkali
Negative regulation of HA, CXCL1, and CCL2 MMP12 KO mice 117 Alkali
Reduced TGF-β and IL6 expression Rapamycin 65 Alkali
Angiopoietin-like protein 2; regulation of F4/80 and IL-1β expression Angptl2−/− mice and K14-Angptl2 mice, subconjunctival Angptl2 siRNA injection 118 Suture
Intermediate-conductance Ca(2+)-activated K(+) channel (KCa3.1), inhibition of EGF-induced HA TRAM-34 119 Alkali
SDF1, reduced VEGF and c-Kit expression Neutralizing anti–SDF-1α antibody 120 Alkali
Inhibition of LA, but activation of HA by Ang-2 blockage L1-10, an Ang-2-specific inhibitor 121 Suture
Galectin-3 increases phosphorylation of VEGFR-2, positive regulation of HA Gal3−/− and Mgat5−/− mice 122 Suture
IRS-1, inhibition of HA and LA GS-101 (Aganirsen), an antisense oligonucleotide against IRS-1 123 Suture
Increased expression of Robo4 upon HSV infection; Robo4 ligand Slit2 inhibits HA Injection of Slit protein 74 HSV-infection
Inhibition of inflammation Corticosteroids 124 Suture
Programmed death ligand-1 (PD-L1), negative regulator of HA PD-L1−/− mice 67 Suture
Toll-like receptor 4 (TLR4), activation of HA by TLR4 ligand HMGB1 TLR4−/− and TLR2−/− mice, topical application of HMGB1 125 Alkali
Very late antigen-1 (VLA-1) VLA-1 neutralizing antibody 126 Suture
Neuropilin 2 (NRP2) NRP2 artificial miRNA 127 Suture
Table 3
 
Mouse Models Showing Spontaneous Corneal NV
Table 3
 
Mouse Models Showing Spontaneous Corneal NV
Mouse Model Mechanism HA LA Time Course Intervention Special Features Reference
Destrin−/− mice, corn1 mice Pro-angiogenic VEGFR-3 signaling, deficient in sVEGFR-1 Starts at 20 days Rescue by VEGFR-3 neutralization or application of sVEGFR-1 Spontaneous corneal HA and LA, differentially regulated by genetic background, different time kinetics of HA and LA 10, 77, 78, 94, 96, 97
TSP-1−/− mice Loss of antiangiogenic signaling of CD36+ macrophages Starts at 6 mo Rescued by wild-type bone marrow cells TSP-1 inhibits VEGF-C release and LA by binding to its receptor CD36 on monocytes 13
CD36−/− mice Loss of TSP-1 – CD36 interaction Late onset (approximately 1 y) None Upregulation of VEGF-A is observed when CD36 is knocked out 13, 84
KLEIP/KLHL20−/− mice Proangiogenic miR-204 – Ang-1 pathway Median time 8 wk Anti-VEGF therapy not effective Includes macrophage infiltration and spontaneous corneal NV, accelerated by superficial abrasion 24, 62
LRIG1−/− mice LRIG1 acts as a negative regulator of STAT3 Median time 18 mo STAT3 inhibition by STA21 or wild-type bone marrow cells rescue phenotype Spontaneous, can be induced by injury 85
K5.Stat3C mice Proinflammatory STAT3 signaling Similar to LRIG1−/− None Similar phenotype to LRIG1 deficient mice 85
Jam-A−/− mice VEGF expression and TGF-β activation Starting early but low prevalence Inhibition of HA by DC101 (anti VEGFR-2 antibody) Spontaneous, can be accelerated by injury, approximately 50% prevalence at 6–12 mo 86, 87
ADAMTS9+/− mice ADAMTS9 is an antiangiogenic metalloprotease Median time approximately 6 wk None 80% prevalence in heterozygous mice (only in C57BL/6 strain). Null mice are not viable 128
Le-Cre;Cited2loxP/loxP Presumably via Pax6 and Klf4 8 mo None High penetrance 129
MxCre;RBPf/f Downstream of Notch receptor Approximately 2 mo None Spontaneous NV affects various ocular and extraocular tissues (cornea, retina, liver, lung) 130
Pax6 overexpression using the Aldh3a1 promoter Chi3l4, Flt-1, Wif1 Approximately 6–12 wk None Up to 69% prevalence, associated with corneal inflammation 88
Pax6+/− mice Presumably due to epithelial defects; sVEGFR-1 deficiency Early phenotype starting at 2 weeks Rescued by application of sVEGFR-1 Incomplete prevalence (different prevalence reports), several anterior eye defects including corneal NV 10, 89, 90
Global FoxC1−/− mice and neural-crest specific FoxC1−/− Via VEGF – VEGFR-2 signaling Detectable at 15 days after birth Inhibition of HA by blocking VEGFR-2 signaling Linked to disrupted epithelial architecture; FoxC1−/− mice show spontaneous corneal NV, FoxC1± mice show increased corneal NV upon injury 15
LeCre;VEGFR-2loxP/loxP Lack of sVEGFR-2 At birth Inhibition of LA by blocking VEGFR-3 signaling 100% prevalence in N = 30 mice 14
pCre;VEGFR-1loxP/loxP Lack of sVEGFR-1 Starting 2 d after pCre injection None Fully vascularized at 14 days, high prevalence in n = 10 mice 10
C57BL/6 mice Unknown P6–P14 None During development, the mouse cornea presents lymphatic vessel sprouting and regression 41
Nude mice, nu/nu, and hairless mutant mouse strain, SKH1;hr/hr Proangiogenic factors probably coming from the epithelium 6–12 wk old mice None Spontaneous corneal NV associated with absence of hair in 2 different genetic mouse models 76, 92
Heme oxygenase (HO)-2 MMP-2 Not applicable shRNA injection Local knockdown, not transgenic. Negative regulation of HA 131
Two aspects form the main rationale to add new corneal models to the available ocular angiogenesis models. First, corneal models are the basis to investigate pathological angiogenesis in a previously avascular tissue, as opposed to a multitude of studies investigating physiological angiogenesis in the retina. Studies that investigate pathological hypervascularization in the retina do not start off with a previously avascular scaffold, like the cornea. In the cornea, de novo vessel sprouting can be directly studied, while in the retina, the change in a preexisting vascular network is the focus of investigation of pathological angiogenesis. 
Second, angiogenesis-mediating transmitters have been shown to exert different functions in different ocular tissues. Consequently, different context-dependent actions of proangiogenic transmitters can be assessed in different ocular angiogenesis assays. An example is the widely discussed role of angiopoietin-1 in angiogenesis. Angiopoietin-1 has been described as a stabilizing factor for the quiescent vasculature and as an angiogenesis-inducing factor.48,49 It is solidly established that administration of angiopoietin-1 causes NV.50 However, in the eye, opposite roles have been demonstrated for angiopoietin-1: For the choroid, there is evidence for a proangiogenic and an antiangiogenic role of angiopoietin-1.51,52 For the cornea, there is no such debate. At least four studies describe proangiogenic functions of angiopoietin-1 and/or its receptor Tie2 in the cornea.23,24,53,54 This example shows that well-characterized angiogenesis-related transmitters still may have different roles in different ocular tissues. Consequently, it is worth the effort to broaden the choice of available tools to investigate experimental angiogenesis in the eye. 
We now will describe and compare the three different approaches to study angiogenesis in the cornea. These approaches are summarized in Figure 3
Figure 3
 
Common experimental approaches to induce corneal NV in mice In this Figure, the four most frequently used approaches for NV induction in murine corneas are shown. (A) corneal pocket assay (depicted here is the growth of blood vessels and lymphatics from opposite sides as observed with bFGF pellets). (B) Suture injury model. (C) Chemical injury by alkali burn. (D) Spontaneous transgenic model. Red arrows represent lymphatics, green arrows represent blood vessels.
Figure 3
 
Common experimental approaches to induce corneal NV in mice In this Figure, the four most frequently used approaches for NV induction in murine corneas are shown. (A) corneal pocket assay (depicted here is the growth of blood vessels and lymphatics from opposite sides as observed with bFGF pellets). (B) Suture injury model. (C) Chemical injury by alkali burn. (D) Spontaneous transgenic model. Red arrows represent lymphatics, green arrows represent blood vessels.
Corneal Pocket Assays and Injury Models
Corneal Pocket Assay
The classical angiogenesis assay in the mouse cornea is the corneal (micro)pocket assay in which a pellet is implanted into the cornea (Table 1). The pellet contains one or more proangiogenic factors, which are gradually set free over the course of several days. As a result, corneal limbal blood vessels are activated and start to grow toward the pellet.45,55,56 To assess hemangiogenesis and lymphangiogenesis, VEGF-A or bFGF pellets are used, while to assess lymphangiogenesis only, VEGF-C pellets can be used.45,57 The appropriate experimental protocols have been reviewed extensively.45,55,57 The extent of corneal NV typically is assessed 5 to 7 days after pellet implantation.45,56,58 Even though the experimental conditions in corneal pocket assays are highly standardized, there still is considerable variation in angiogenic and lymphangiogenic response depending on the genetic background of the mouse strain.5961 This shows that while corneal pocket assays are a good method for initial screening of proangiogenic substances, care should be taken when quantitatively comparing the results. 
External Injury and Corneal Transplantation
The corneal epithelium has a pronounced capacity for regeneration after injury. Therefore, in mice, superficial corneal abrasion that leaves the corneal limbus intact, completely heals within 7 to 14 days.62 However, more severe external injury or prolonged epithelial defects (e.g., by infection) can disturb corneal integrity in such a way that stromal NV occurs. This effect is used in injury models for NV in murine corneas, which we will explain in detail in the following section. 
Most routinely used injury models rely on suture injury or alkali burn (Table 2).63 Here, we briefly present the most widely used injury approaches. In alkali burn models, a specified amount of NaOH (mostly 0.1 N or 1 N) is applied for a specified time (typically 10–30 seconds).6466 In suture models, one or more nylon sutures are placed eccentrically in the corneal stroma.67,68 These nylon threads remain in the cornea and induce inflammation and NV (HA and LA). Another mechanical injury approach makes use of extensive removal of the corneal epithelium, including the limbus.69 This epithelial debridement must be distinguished from superficial epithelial abrasion, which does not induce corneal NV in wild-type mice and completely heals within a few days.62 Another way to induce corneal NV by injury aims at damaging the LSCs, which are needed for corneal epithelial regeneration and contribute to corneal avascularity. In patients with LSC deficiency (LSCD), corneal NV is observed commonly.18 Three approaches to induce LSCD with corneal NV in mouse models use chemical injury by benzalkonium chloride,70 transplantation of corneas lacking Abcb5-positive LSCs,17 or UV-B irradiation.71 Other similar models use thermal cauterization. 
Another very useful approach is the corneal transplantation model, which can be used to study corneal NV in a clinically highly relevant setting. After allograft transplantation, corneal inflammation and NV occur, both of which can be pharmacologically attenuated.72,73 It is not clear whether corneal NV in these models is primarily due to inflammation and/or due to mechanical injury. Therefore, we subsume this model in the category of external injury models. 
Another very elegant, yet rarely applied approach is based on corneal infection, for example by external application of HSV-1 virus particles to the murine cornea.74,75 
These injury-based models are well standardized and are used widely in different variations. Their major disadvantage is that the damage inflicted upon the cornea is so grave that not only angiogenesis is induced, but also a number of other processes, for example wound healing, epithelial proliferation, and inflammation. Still, some of these approaches are highly relevant models for clinically relevant scenarios. For example, the investigation of angiogenesis in the context of corneal transplantation can yield important insights into mechanisms of allograft failure in patients. Furthermore, the interplay between inflammation and angiogenesis can be studied efficiently in corneal injury and corneal infection models, which, therefore, are highly useful for translation of knowledge to the clinic. 
Transgenic Mouse Models in Corneal NV Research
Two decades ago, spontaneously occurring corneal NV has been demonstrated in two mouse models.76 Yet, this discovery has not stimulated the use of spontaneous models of corneal NV in angiogenesis research. Recently, however, several genetically engineered mice developing spontaneous corneal NV were presented. Here, relevant transgenic models of corneal NV are explained (see also Table 3 for a summary of these models). 
Destrin−/− Mice (Corn1)
Destrin−/−/corn1 mice carry a homozygous mutation in the destrin gene and develop spontaneous corneal HA and LA.77,78 Both processes can be diminished by blocking VEGF–VEGFR-3 signaling: efficient inhibition of NV is achieved by external application of a blocking anti–VEGFR-3 antibody or by application of sVEGFR-1 (for which these corneas are deficient).10,78 Corneal HA reaches 100% prevalence at 4 weeks after birth and does not regress within 12 months, while prevalence of corneal LA is incomplete (reaches 60% at 3 months and decreases to approximately 15% at 12 months).78 Notably, the destrin−/− mouse model shows a low-inflammatory form of angiogenesis. Inflammatory activity as measured by the number of CD45-positive cells in corneal sections of vascularized corn1 corneas is much lower than in vascularized corneas after allograft transplantation.78 Although corneal HA partially regresses after application of anti–VEGFR-3 antibody, some degree of pathological HA still is present.78 Yet, external application of sVEGFR-1 completely inhibits corneal NV in this model.10 These findings show that the destrin−/−/corn1 model can be used to characterize antiangiogenic potential of pharmacological inhibitors. Because the destrin−/−/corn1 phenotype has an early onset and a high penetrance, it is very well suited for angiogenesis research. 
TSP-1−/− Mice and CD36−/− Mice
In 2004, it was reported that knockout mice for TSP-1 and/or TSP-2 do not show spontaneous hemangiogenesis.79 Interestingly, six years later, the same research group found out that TSP-1−/− mice do show spontaneously lymphatically vascularized corneas at the age of six months, while control corneas are devoid of lymphatics at that time point.13 Additionally, TSP-1−/− mice also show decreased intraocular pressure (IOP).80 However, it has not yet been investigated whether decreased IOP is linked to increased ocular lymphangiogenesis this model. 
With TSP-1−/− mice, a model to investigate mechanisms of spontaneous, isolated LA is available. Since the description of the TSP-1−/− model, several interesting facets of this ocular phenotype have been reported. For example, the corneal and lacrimal gland phenotype of the TSP-1−/− mouse resembles human Sjögren's syndrome.81 Therefore, TSP-1−/− mice have been used to investigate pathophysiological mechanisms of Sjögren's syndrome.81,82 
In the TSP-1−/− model, prolymphangiogenic signaling has been shown to critically depend on the presence of CD36-positive monocytes/macrophages.13 This is consistent with the observation that maintenance of lymphatic vessels depends on corneal stromal macrophages in a corneal injury model.83 Additionally, spontaneous corneal LA in TSP-1−/− mice parallels spontaneous corneal LA in CD36−/− mice.84 However, corneal LA in CD36−/− mice takes approximately 1 year to develop.84 
In short, two main features make the TSP-1−/− model well suited for angiogenesis research: First, its proximity to Sjögren's syndrome qualifies it as a disease-oriented model (as opposed to corneal injury models and corneal pocket assays). Second, the complex interaction of macrophages and lymphatics in this model can be used to investigate further basic mechanisms of lymphangiogenesis. 
LRIG1−/− Mice and K5.Stat3C Transgenic Mice
The protein leucine-rich repeats and immunoglobulin-like domains 1 (LRIG1) is a marker of corneal epithelial LSCs and ocular epithelial basal cells.85 Its knockout leads to spontaneous corneal epithelial dysplasia and stromal NV.85 The mechanism behind this phenotype is disinhibition of a proinflammatory Stat3 pathway after LRIG1 knockout. K5.Stat3C mice, which express a constitutively active form of Stat3, show a comparable phenotype. Consequently, LRIG1−/− mice and K5.Stat3C transgenic mice can serve as models to investigate mechanisms of spontaneous inflammatory angiogenesis. 
Inducible VEGFR-1 and VEGFR-2 Knockout Mice
From two recent studies, we know that sVEGFR-1 and sVEGFR-2 are important mediators of corneal angiogenic or lymphangiogenic privilege.10,14 In these studies, inducible pCre,VEGFR-1loxP/loxP and LeCre,VEGFR-2loxP/loxP mice were created that lacked these soluble receptors and consequently showed spontaneous hem- or lymphangiogenesis in the cornea. 
Jam-A−/− Mice
Knockout of junctional adhesion molecule A (Jam-A) results in spontaneous corneal NV, inflammation, and opacification. These changes are linked to activation of various proinflammatory and proangiogenic systems, among them the TGF-β pathway and the VEGF-A–VEGFR-2 pathway.86 Interestingly, in this model of corneal NV, a contribution of bFGF is unlikely because presence of Jam-A has been shown to be essential for bFGF-induced angiogenesis.87 However, a limitation for the practical use of these mice is the incomplete penetrance of this phenotype.86 Still, mechanistically, the Jam-A−/− model is a very attractive model for angiogenesis research. 
FoxC1−/− or NC-FoxC1−/−, and FoxC1+/− or NC-FoxC1+/− Mice
In 2011, Seo et al.15 identified the transcription factor FOXC1 as an essential mediator of corneal clarity and avascularity, and described the following knockout phenotype. Global knockout mice and neural crest (NC) specific knockout mice show spontaneous corneal HA and LA dependent on VEGFR-2 signaling. Interestingly, corneal HA starts during embryonic development and the cornea is completely covered by blood vessels at birth (P0). Also, lymphatic vessels reach deep into the cornea at P0. Moreover, heterozygous mice also show highly increased angiogenic sprout formation in the cornea, but do not become fully vascularized. For angiogenesis research, this model is highly attractive because it facilitates experimental procedures in two ways. First, in these mice, corneal NV is present at birth and second, heterozygous mice also can be used for research, which triples biological sample size per litter. 
C57BL/6 Mice
Recently, Zhang et al.41 systematically investigated the murine corneal development in C57BL/6 mice. They discovered that during postnatal days P6 and P14, LYVE-1–positive lymphatic vessels reach deep into the corneal stroma. After P21, these vessels regress and the cornea remains devoid of blood vessels. This process of lymphangiogenesis and subsequent spontaneous regression is exceptional and offers wide-ranging research perspectives. For example, a full molecular characterization of signaling pathways involved in this process would possibly elucidate new basic mechanisms of lymphangiogenesis. Although this mouse model is not a transgenic model, we subsume it here because of the spontaneous vessel growth and vessel regression seen in these mice. 
Pax6 Overexpression or Heterozygocity
Corneal overexpression of Pax6 using the aldehyde dehydrogenase family 3, subfamily A1 (Aldh3a1) promoter results in epithelial abnormalities and inflammatory stromal NV.88 This combination of epithelial and stromal phenotypes is consistent with the notion that the corneal epithelium helps to maintain corneal avascularity. Remarkably, some of these features (including stromal NV) also occur in Pax6+/− mice (for which one study reports partial prevalence89 and another study reports 100% prevalence90). A recent study shed light on this situation because it was shown that Pax6 is the single most important regulator of corneal epithelial cell identity.19 Apparently, an impairment of tightly regulated Pax6 signaling either way (partial loss of function or gain of function) compromises corneal epithelial quiescence, leading to secondary stromal NV. One mechanism of stromal NV in Pax6+/− mice is deficiency for sVEGFR-1 and external application of sVEGFR-1 rescues Pax6+/− phenotype.10 For practical use in angiogenesis research, Pax6+/− mice are partly suited: although phenotype onset occurs as early as 2 weeks after birth, penetrance is incomplete.89 In Pax6 overexpressing mice, prevalence of neovascular phenotype differs depending on the genetic background and is always below 70%.88 To establish these mice as standard models of corneal NV, breeding could optimize this prevalence. 
KLEIP−/− Mice (KLHL20−/−)
Homozygous mice deficient for the intracellular protein KLEIP/KLHL20 show a progressive corneal dystrophy with spontaneous stromal HA and LA, reaching 100% prevalence at 4 months after birth.62,91 Spontaneous corneal NV in this model is promoted by angiopoietin-1, which in turn is upregulated because its inhibition by miR-204 is lost in this model.24 It can, therefore, serve as a starting point to evaluate non-VEGF targeting therapy approaches for pathological angiogenesis. Corneal NV in KLEIP−/− mice is associated with presence of macrophages in the dystrophic cornea, but it is presently unknown whether these macrophages are required for HA and/or LA in this model.24,62 In practice, KLEIP−/− corneal NV is well suited for angiogenesis research because it has a high penetrance and the time course of phenotype progression is well characterized. 
Nude Mice and Hairless SKH1;hr/hr Mice
Lastly, we want to point out an almost 25-year-old finding by Niederkorn et al.76 They reported that immunodeficient nude (nu/nu) mice show subepithelial corneal NV. This phenotype is not due to immunodeficiency because severe combined immunodeficient (SCID) mice do not show any corneal blood vessels. Rather, the phenotype seems to be related to the lack of hair because another hairless mutant strain (SKH1;hr/hr) displayed a similar phenotype. Later, this phenotype was roughly linked to epithelium-derived proangiogenic factors.92 To our knowledge, the exact molecular mechanisms of corneal NV in hairless mice have not been investigated by today's state of the art methods. This example shows that spontaneous corneal NV in transgenic mice offers a significant and currently unused potential for angiogenesis research. 
Advantages and Limitations of Transgenic Models of Corneal NV
Similar Etiologic Factors in Different Transgenic Models of Corneal NV
In transgenic mouse models, corneal stromal NV often is linked to a disturbed corneal epithelial architecture and the transformation of corneal epithelium toward skin-like epithelium. For example, this is the case in LRIG1−/−, KLEIP−/−, Pax6+/−, destrin−/−/corn1, and Notch1−/− mice (Notch1−/− only after corneal injury).24,78,85,90,93 A comparable change is characteristic of some human corneal diseases, for example xerophthalmia. Two etiologic factors contributing to these changes of corneal epithelial architecture in mouse models are LSC deficiency and disturbed control of epithelial cell proliferation.17,94 
The LSC deficiency, corneal epithelial dysplasia, and stromal NV are observed in mice deficient for LRIG1, a moderately specific LSC marker and regulator of inflammation.85 Interestingly, the corneal phenotype of LRIG1−/− and KLEIP−/− mice is similar morphologically and on an expression level. Our unpublished data show that LRIG1 expression is reduced 2.3 fold in late-stage KLEIP−/− vascularized corneas (P = 0.0018, N = 3 + 3 mice, see the methods of Kather et al.24), suggesting an involvement of LSCs in KLEIP−/− corneal phenotype development. 
Furthermore, control of corneal epithelial cell proliferations requires functional actin cytoskeleton dynamics.94 For example, corneal epithelial dystrophy in destrin−/−/corn1 is due to compromised intracellular actin cytoskeleton remodeling.94 Similarly, the KLEIP protein has been linked to regulation of actin cytoskeleton assembly so that also in KLEIP−/− corneal NV development, corneal epithelial actin polymerization might be an etiologic factor.95 
These findings show that different transgenic mouse models of corneal NV may share mechanisms of phenotype development. This emphasizes the usefulness of spontaneous corneal NV models in angiogenesis research, because these models make it possible to identify and investigate several intertwined signaling pathways that ultimately lead to corneal NV. 
Comparison of Transgenic Models to Corneal Pocket Assays
In corneal pocket assays, one or more proangiogenic transmitters are directly implanted into the cornea, and stimulate the growth of blood vessels and/or lymphatics toward a gradient. However, in basic and translational angiogenesis research, the reductionist approaches that centered on single-transmitter mechanisms of angiogenesis are replaced by a more comprehensive view entailing interconnected regulatory mechanisms of angiogenesis. In transgenic corneal NV models, proangiogenic signaling is partly due to canonical VEGF signaling (like the destrin−/−/corn1 model) or due to VEGF-independent signaling (like in late stage KLEIP−/− corneal NV). In several transgenic corneal NV models, an involvement of several different proangiogenic mechanisms is conceivable. Thereby, transgenic models offer a possibility to investigate these intertwined mechanisms in a well-characterized and experimentally accessible setting. 
It should be noted that, as mentioned above, the extent of the angiogenic and lymphangiogenic response in corneal pocket assays depends on the genetic background of the strain. This also has been reported with transgenic models of corneal NV, for example in destrin−/− (corn1) mice96,97 In this model, the corneal phenotype is more pronounced in A.BY than in C57BL/6 mice.97 
Comparison of Transgenic Models to Corneal Injury Models
In classical corneal injury models, corneal NV is induced by a clearly defined stimulus and follows a clearly defined temporal pattern. However, a potential issue is that injury initiates a plethora of repair processes whose exact relationship to corneal NV is unclear. Transgenic mouse models of corneal NV open up a different approach because they are not dependent on experimental intervention, but show angiogenesis secondary to spontaneous pathological processes. Several fields profit from the use of transgenic corneal NV models, for example the investigation of inflammatory mechanisms. Corneal NV in external injury models is generally linked to a significant inflammatory activity. Correspondingly, it has been shown that anti-inflammatory treatment reduces corneal NV in injury models.26,98,99 In contrast, in transgenic models of corneal NV, high-inflammatory (such as TSP-1−/−) and low-inflammatory (such as destrin−/−/corn1) models are available. Also, in other transgenic models of corneal NV, like the KLEIP−/− model, inflammation is present but the mechanistic link to angiogenesis has not been investigated yet. Therefore, transgenic models of corneal NV offer a convenient framework to investigate new mechanisms that link inflammation to angiogenesis in the cornea. 
Still, transgenic corneal NV models will complement, but not replace, existing corneal NV models. 
Comparison of Transgenic Models to Retinal and Choroidal Models
In retinal and choroidal NV models, transgenic mice have long been incorporated into angiogenesis research. Physiological angiogenesis in these tissues has been used to elucidate basic mechanisms of angiogenesis, such as paracrine signaling in angiogenic sprouts. Also, retinal models of pathological hypervascularization have yielded insights into the pathogenesis of various clinically relevant diseases, such as diabetic retinopathy or retinopathy of prematurity. Thus, these models already have proven to be very useful, but transgenic models of corneal NV still offer different perspectives: Compared to retinal and choroidal models, corneal NV models start off with an avascular tissue but still represent pathological angiogenesis. While the focus of retinal models of pathological angiogenesis is the transition of a healthy vasculature into an aberrant and harmful vascular network, corneal NV model allow to easily image the pathological de novo vascularization of a tissue. 
Furthermore, transgenic mouse models of corneal NV differ from the widely established transgenic models of choroidal NV in one important aspect: Whereas in choroidal NV models, phenotype prevalence is incomplete and limits experimental feasibility, this is not the case with all transgenic corneal NV models. For example, LRIG1−/− and KLEIP−/− mice develop corneal NV in a well-characterized time course and with 100% prevalence.24,85 In FoxC1−/− mice, corneal NV develops even in heterozygous mice, which allows for an even higher yield of samples.15 
Therefore, transgenic models of corneal NV can complement retinal and choroidal NV models and help to clarify mechanisms of de novo angiogenesis. Because the extent of this form of angiogenesis can be easily quantified in the cornea, we will describe the most relevant imaging models in the following section. 
Imaging and Quantifying Corneal Neovessels
As the cornea is a thin, easily accessible, and physiologically transparent tissue, imaging blood vessels and lymphatics is relatively easy in the cornea. Using confocal microscopy, the complete corneal vasculature can be imaged. Blood vessels and lymphatics inside the corneal stroma can be fully captured in detail (Fig. 4). Using two-photon microscopy, it is even possible to capture cellular migration in vivo in real time. For example, a corneal NV model made it possible to capture immune cell immigration into lymphatic vessels for the first time.100 
Figure 4
 
Corneal blood vessels and lymphatics imaged with confocal microscopy. Compared to other mouse tissues, the cornea is an ideal organ to study experimentally induced angiogenesis, because it can easily be imaged in conventional and confocal fluorescence microscopy. Using confocal microscopy, the corneal tissue can be imaged as a whole, so that the complete corneal vasculature can be analyzed. In this Figure, CD31-positive blood and lymphatic vessels (A) and LYVE-1–positive lymphatics (B) in a KLEIP−/− cornea are shown. (C) Shows the overlay of CD31 and LYVE-1 staining.
Figure 4
 
Corneal blood vessels and lymphatics imaged with confocal microscopy. Compared to other mouse tissues, the cornea is an ideal organ to study experimentally induced angiogenesis, because it can easily be imaged in conventional and confocal fluorescence microscopy. Using confocal microscopy, the corneal tissue can be imaged as a whole, so that the complete corneal vasculature can be analyzed. In this Figure, CD31-positive blood and lymphatic vessels (A) and LYVE-1–positive lymphatics (B) in a KLEIP−/− cornea are shown. (C) Shows the overlay of CD31 and LYVE-1 staining.
Ex vivo, corneal specimens usually are assessed quantitatively in a planar dimension (in whole mount samples, Fig. 5) or in an apical-basal dimension (in tissue sections, Fig. 6). 
Figure 5
 
Manual quantification of NV in corneal whole mounts. In mouse models of corneal NV, such as the transgenic KLEIP−/− model, vessels sprout from the vascular plexus at the limbus toward the corneal center. The extent of corneal NV can be quantified by measuring the remaining avascular area. This is shown in (A) for blood vessels and in (B) for lymphatics (area was delineated by hand in sample images). (C) A DAPI staining of a control cornea is shown. Measuring DAPI-positive area permits normalization of area measurements to total corneal area.
Figure 5
 
Manual quantification of NV in corneal whole mounts. In mouse models of corneal NV, such as the transgenic KLEIP−/− model, vessels sprout from the vascular plexus at the limbus toward the corneal center. The extent of corneal NV can be quantified by measuring the remaining avascular area. This is shown in (A) for blood vessels and in (B) for lymphatics (area was delineated by hand in sample images). (C) A DAPI staining of a control cornea is shown. Measuring DAPI-positive area permits normalization of area measurements to total corneal area.
Figure 6
 
Different vascularization patterns can be distinguished quantitatively. Corneal NV can occur in different parts of the corneal stroma (e.g., subepithelial or deep stromal corneal NV). These vascularization patterns can be distinguished by quantitative measurements in corneal sections. This image shows subepithelial corneal NV in a KLEIP−/− mouse. Bowman's layer and Descemet's membrane have been delineated with a dashed line and the distance of each vessel from Bowman's layer is indicated by dotted lines. Quantifying these measurements permits exact representations of the distribution of corneal neovessels.
Figure 6
 
Different vascularization patterns can be distinguished quantitatively. Corneal NV can occur in different parts of the corneal stroma (e.g., subepithelial or deep stromal corneal NV). These vascularization patterns can be distinguished by quantitative measurements in corneal sections. This image shows subepithelial corneal NV in a KLEIP−/− mouse. Bowman's layer and Descemet's membrane have been delineated with a dashed line and the distance of each vessel from Bowman's layer is indicated by dotted lines. Quantifying these measurements permits exact representations of the distribution of corneal neovessels.
In a planar dimension, the most common parameter for NV is area covered by vessels, which can be measured either manually or semiautomatically. Bock et al.101 have introduced a semiautomatic, threshold-based method for morphometry of angiogenesis and lymphangiogenesis in corneal whole mounts. This method excels manual counting techniques in accuracy, objectivity, and reproducibility.101 Blacher et al. have extended this toolkit to measurement of morphological features, such as density of vessel extremities or density of branching points in the corneal neovascular plexus.102 A similar method has been used by Dastjerdi et al.103 to quantify corneal NV in slit-lamp images. Several other image analysis approaches for blood vessel segmentation have been implemented for images of the retinal vasculature, but not yet for the cornea.104 
Analysis of vessel coverage of the cornea has yielded surprising results. For example, it has been shown that vascularization in the nasal part of the cornea develops in a different way than vascularization in the temporal part.105 Also, lymphatic vessels in the nasal part of the cornea show more lymphatic valves than vessels in the temporal cornea.106 Additionally, the morphology of the growing vasculature differs depending on the etiologic factor.107 These findings highlight the benefit of accurately analyzing planar vascular growth patterns in mouse corneas. 
In an apical-basal direction, different patterns of corneal NV can be distinguished, for example subepithelial and deep stromal NV. This parallels blood vessel growth in the mouse retina, where different apical-basal blood vessel distribution patterns have been described.46 
Clinical Perspectives
In recent years, antiangiogenic therapy has been a widely-discussed issue in the context of corneal NV, retinal hypervascularization, and also in the context of extraocular diseases.20,108 In the cornea, NV is associated with loss of visual acuity and allograft failure.4,6 In extraocular pathologies, antiangiogenic therapy offers strong therapeutic perspectives. However, until now, most antiangiogenic therapy approaches have been of limited efficacy. A possible reason for this discrepancy between the high expectations and the limited success of antiangiogenic therapy could be the focus on therapeutic options targeting only one transmitter. In vivo models, such as transgenic spontaneous corneal NV models, make it possible to investigate blood vessel growth that develops under real pathophysiological conditions, with minimal to no experimental intervention. Transgenic mouse models of corneal NV have been used to identify several NV-related signaling cascades, such as TSP-1 - CD36 - VEGF-C pathway, STAT3 signaling, miR-204–Ang-1 axis and others (Table 3). Yet, it still must be investigated how these transmitters affect corneal NV in humans. 
In summary, transgenic models of corneal NV can be used to characterize the complex interaction between various proangiogenic pathways, thereby helping to overcome the limitations of current antiangiogenic therapy in the clinic. 
Acknowledgments
Sandra Cramer-Stoll and Nicole Woik provided technical support with corneal samples and confocal microscopy. 
Supported by grants from Deutsche Forschungsgemeinschaft (SFB/TR23 “Vascular Differentiation and Remodeling” [JNK and JK] and IRTG 1874 “DIAMICOM” [JK]) and Studienstiftung des deutschen Volkes (JNK), and by the Core Facility Live Cell Imaging Mannheim at the CBTM (DFG INST 91027/10-1 FUGG). 
Disclosure: J.N. Kather, None; J. Kroll, None 
References
Carmeliet P. Angiogenesis in life, disease and medicine. Nature. 2005; 438: 932–936. [CrossRef] [PubMed]
Kather JN Kroll J. Rho guanine exchange factors in blood vessels: fine-tuners of angiogenesis and vascular function. Exp Cell Res. 2013; 319: 1289–1297. [CrossRef] [PubMed]
Bock F Maruyama K Regenfuss B Novel anti(lymph)angiogenic treatment strategies for corneal and ocular surface diseases. Prog Retin Eye Res. 2013; 34: 89–124. [CrossRef] [PubMed]
Bachmann B Taylor RS Cursiefen C. The association between corneal NV and visual acuity: a systematic review. Acta Ophthalmol. 2013; 91: 12–19. [CrossRef] [PubMed]
Ellenberg D Azar DT Hallak JA Novel aspects of corneal angiogenic and lymphangiogenic privilege. Prog Retin Eye Res. 2010; 29: 208–248. [CrossRef] [PubMed]
Bachmann B Taylor RS Cursiefen C. Corneal neovascularization as a risk factor for graft failure and rejection after keratoplasty: an evidence-based meta-analysis. Ophthalmology. 2010; 117: 1300–1305. [CrossRef] [PubMed]
Cursiefen C Schlotzer-Schrehardt U Kuchle M Lymphatic vessels in vascularized human corneas: immunohistochemical investigation using LYVE-1 and podoplanin. Invest Ophthalmol Vis Sci. 2002; 43: 2127–2135. [PubMed]
Azar DT. Corneal angiogenic privilege: angiogenic and antiangiogenic factors in corneal avascularity, vasculogenesis, and wound healing (an American Ophthalmological Society thesis). Trans Am Ophthalmol Soc. 2006; 104: 264–302. [PubMed]
Chang JH Gabison EE Kato T Azar DT. Corneal neovascularization. Curr Opin Ophthalmol. 2001; 12: 242–249. [CrossRef] [PubMed]
Ambati BK Nozaki M Singh N Corneal avascularity is due to soluble VEGF receptor-1. Nature. 2006; 443: 993–997. [CrossRef] [PubMed]
Cursiefen C Chen L Saint-Geniez M Nonvascular VEGF receptor 3 expression by corneal epithelium maintains avascularity and vision. Proc Natl Acad Sci U S A. 2006; 103: 11405–11410. [CrossRef] [PubMed]
Makino Y Cao R Svensson K Inhibitory PAS domain protein is a negative regulator of hypoxia-inducible gene expression. Nature. 2001; 414: 550–554. [CrossRef] [PubMed]
Cursiefen C Maruyama K Bock F Thrombospondin 1 inhibits inflammatory lymphangiogenesis by CD36 ligation on monocytes. J Exp Med. 2011; 208: 1083–1092. [CrossRef] [PubMed]
Albuquerque RJ Hayashi T Cho WG Alternatively spliced vascular endothelial growth factor receptor-2 is an essential endogenous inhibitor of lymphatic vessel growth. Nat Med. 2009; 15: 1023–1030. [CrossRef] [PubMed]
Seo S Singh HP Lacal PM Forkhead box transcription factor FoxC1 preserves corneal transparency by regulating vascular growth. Proc Natl Acad Sci U S A. 2012; 109: 2015–2020. [CrossRef] [PubMed]
Ferrari G Hajrasouliha AR Sadrai Z Ueno H Chauhan SK Nerves Dana R. and neovessels inhibit each other in the cornea. Invest Ophthalmol Vis Sci. 2013; 54: 813–820. [CrossRef] [PubMed]
Ksander BR Kolovou PE Wilson BJ ABCB5 is a limbal stem cell gene required for corneal development and repair. Nature. 2014; 511: 353–357. [CrossRef] [PubMed]
Ahmad S. Concise review: limbal stem cell deficiency, dysfunction, and distress. Stem Cells Transl Med. 2012; 1: 110–115. [CrossRef] [PubMed]
Ouyang H Xue Y Lin Y WNT7A and PAX6 define corneal epithelium homeostasis and pathogenesis. Nature. 2014; 511: 358–361. [CrossRef] [PubMed]
Chang JH Garg NK Lunde E Han KY Jain S Azar DT. Corneal neovascularization: an anti-VEGF therapy review. Surv Ophthalmol. 2012; 57: 415–429. [CrossRef] [PubMed]
Cao R Brakenhielm E Pawliuk R Angiogenic synergism, vascular stability and improvement of hind-limb ischemia by a combination of PDGF-BB and FGF-2. Nat Med. 2003; 9: 604–613. [CrossRef] [PubMed]
Tarallo V Bogdanovich S Hirano Y Inhibition of choroidal and corneal pathologic neovascularization by Plgf1-de gene transfer. Invest Ophthalmol Vis Sci. 2012; 53: 7989–7996. [CrossRef] [PubMed]
Morisada T Oike Y Yamada Y Angiopoietin-1 promotes LYVE-1-positive lymphatic vessel formation. Blood. 2005; 105: 4649–4656. [CrossRef] [PubMed]
Kather JN Friedrich J Woik N Angiopoietin-1 is regulated by miR-204 and contributes to corneal neovascularization in KLEIP-deficient mice. Invest Ophthalmol Vis Sci. 2014; 55: 4295–4303. [CrossRef] [PubMed]
Ambati BK Joussen AM Kuziel WA Adamis AP Ambati J. Inhibition of corneal neovascularization by genetic ablation of CCR2. Cornea. 2003; 22: 465–467. [CrossRef] [PubMed]
Dana R. Comparison of topical interleukin-1 vs tumor necrosis factor-alpha blockade with corticosteroid therapy on murine corneal inflammation, neovascularization, and transplant survival (an American Ophthalmological Society thesis). Trans Am Ophthalmol Soc. 2007; 105: 330–343. [PubMed]
Wallace GR John Curnow S, Wloka K, Salmon M, Murray PI. The role of chemokines and their receptors in ocular disease. Prog Retin Eye Res. 2004; 23: 435–448. [CrossRef] [PubMed]
Chung ES Saban DR Chauhan SK Dana R. Regulation of blood vessel versus lymphatic vessel growth in the cornea. Invest Ophthalmol Vis Sci. 2009; 50: 1613–1618. [CrossRef] [PubMed]
Chang LK Garcia-Cardena G Farnebo F Dose-dependent response of FGF-2 for lymphangiogenesis. Proc Natl Acad Sci U S A. 2004; 101: 11658–11663. [CrossRef] [PubMed]
Hajrasouliha AR Sadrai Z Chauhan SK Dana R. b-FGF induces corneal blood and lymphatic vessel growth in a spatially distinct pattern. Cornea. 2012; 31: 804–809. [CrossRef] [PubMed]
Yuen D Pytowski B Chen L. Combined blockade of VEGFR-2 and VEGFR-3 inhibits inflammatory lymphangiogenesis in early and middle stages. Invest Ophthalmol Vis Sci. 2011; 52: 2593–2597. [CrossRef] [PubMed]
Chung ES Chauhan SK Jin Y Contribution of macrophages to angiogenesis induced by vascular endothelial growth factor receptor-3-specific ligands. Am J Pathol. 2009; 175: 1984–1992. [CrossRef] [PubMed]
Cursiefen C Chen L Borges LP VEGF-A stimulates lymphangiogenesis and hemangiogenesis in inflammatory neovascularization via macrophage recruitment. J Clin Invest. 2004; 113: 1040–1050. [CrossRef] [PubMed]
Maruyama K Ii M Cursiefen C Inflammation-induced lymphangiogenesis in the cornea arises from CD11b-positive macrophages. J Clin Invest. 2005; 115: 2363–2372. [CrossRef] [PubMed]
Wacker A Gerhardt H. Endothelial development taking shape. Curr Opin Cell Biol. 2011; 23: 676–685. [PubMed]
Chen Q Jiang L Li C Haemodynamics-driven developmental pruning of brain vasculature in zebrafish. PLoS Biol. 2012; 10: e1001374. [CrossRef] [PubMed]
Cheng C Haasdijk R Tempel D Endothelial cell-specific FGD5 involvement in vascular pruning defines neovessel fate in mice. Circulation. 2012; 125: 3142–3158. [CrossRef] [PubMed]
Korn C Scholz B Hu J Endothelial cell-derived non-canonical Wnt ligands control vascular pruning in angiogenesis. Development. 2014; 141: 1757–1766. [CrossRef] [PubMed]
Bucher F Bi Y Gehlsen U Hos D Cursiefen C Bock F. Regression of mature lymphatic vessels in the cornea by photodynamic therapy. Br J Ophthalmol. 2014; 98: 391–395. [CrossRef] [PubMed]
Koenig Y Bock F Kruse FE Stock K Cursiefen C. Angioregressive pretreatment of mature corneal blood vessels before keratoplasty: fine-needle vessel coagulation combined with anti-VEGFs. Cornea. 2012; 31: 887–892. [CrossRef] [PubMed]
Zhang H Hu X Tse J Tilahun F Qiu M Chen L. Spontaneous lymphatic vessel formation and regression in the murine cornea. Invest Ophthalmol Vis Sci. 2011; 52: 334–338. [CrossRef] [PubMed]
Kelley PM Steele MM Tempero RM. Regressed lymphatic vessels develop during corneal repair. Lab Invest. 2011; 91: 1643–1651. [CrossRef] [PubMed]
Cursiefen C Viaud E Bock F Aganirsen antisense oligonucleotide eye drops inhibit keratitis-induced corneal neovascularization and reduce need for transplantation: the I-CAN Study. Ophthalmology. 2014; 121: 1683–1692. [CrossRef] [PubMed]
Gimbrone MA Jr Cotran RS Leapman SB Folkman J. Tumor growth and neovascularization: an experimental model using the rabbit cornea. J Natl Cancer Inst. 1974; 52: 413–427. [PubMed]
Rogers MS Birsner AE D'Amato RJ. The mouse cornea micropocket angiogenesis assay. Nat Protoc. 2007; 2: 2545–2550. [CrossRef] [PubMed]
Stahl A Connor KM Sapieha P The mouse retina as an angiogenesis model. Invest Ophthalmol Vis Sci. 2010; 51: 2813–2826. [CrossRef] [PubMed]
Grossniklaus HE Kang SJ Berglin L. Animal models of choroidal and retinal neovascularization. Prog Retin Eye Res. 2010; 29: 500–519. [CrossRef] [PubMed]
Suri C Jones PF Patan S Requisite role of angiopoietin-1, a ligand for the TIE2 receptor, during embryonic angiogenesis. Cell. 1996; 87: 1171–1180. [CrossRef] [PubMed]
Augustin HG Koh GY Thurston G Alitalo K. Control of vascular morphogenesis and homeostasis through the angiopoietin-Tie system. Nat Rev Mol Cell Biol. 2009; 10: 165–177. [CrossRef] [PubMed]
Koh GY. Orchestral actions of angiopoietin-1 in vascular regeneration. Trends Mol Med. 2013; 19: 31–39. [CrossRef] [PubMed]
Oliner JD Bready J Nguyen L AMG 386, a selective angiopoietin 1/2-neutralizing peptibody, inhibits angiogenesis in models of ocular neovascular diseases. Invest Ophthalmol Vis Sci. 2012; 53: 2170–2180. [CrossRef] [PubMed]
Lee J Park DY Park do Y Angiopoietin-1 suppresses choroidal neovascularization and vascular leakage. Invest Ophthalmol Vis Sci. 2014; 55: 2191–2199. [CrossRef] [PubMed]
Kanayama S Nishida K Yamato M Analysis of angiogenesis induced by cultured corneal and oral mucosal epithelial cell sheets in vitro. Exp Eye Res. 2007; 85: 772–781. [CrossRef] [PubMed]
Singh N Macnamara E Rashid S Systemic soluble Tie2 expression inhibits and regresses corneal neovascularization. Biochem Biophys Res Commun. 2005; 332: 194–199. [CrossRef] [PubMed]
Tang Z Zhang F Li Y A mouse model of the cornea pocket assay for angiogenesis study. J Vis Exp. 2011; 54: e3077.
Lee H Schlereth SL Park EY Emami-Naeini P Chauhan SK Dana R. A novel pro-angiogenic function for interferon-gamma-secreting natural killer cells. Invest Ophthalmol Vis Sci. 2014; 55: 2885–2892. [CrossRef] [PubMed]
Cao R Lim S Ji H Mouse corneal lymphangiogenesis model. Nat Protoc. 2011; 6: 817–826. [CrossRef] [PubMed]
Sun Y Su L Wang Z Xu Y Xu X. H-RN a peptide derived from hepatocyte growth factor, inhibits corneal neovascularization by inducing endothelial apoptosis and arresting the cell cycle. BMC Cell Biol. 2013; 14: 8. [CrossRef] [PubMed]
Rohan RM Fernandez A Udagawa T Yuan J D'Amato RJ. Genetic heterogeneity of angiogenesis in mice. FASEB J. 2000; 14: 871–876. [PubMed]
Regenfuss B Onderka J Bock F Hos D Maruyama K Cursiefen C. Genetic heterogeneity of lymphangiogenesis in different mouse strains. Am J Pathol. 2010; 177: 501–510. [CrossRef] [PubMed]
Nakao S Maruyama K Zandi S Lymphangiogenesis and angiogenesis: concurrence and/or dependence? Studies in inbred mouse strains. FASEB J. 2010; 24: 504–513. [CrossRef] [PubMed]
Hahn N Dietz CT Kuhl S Vossmerbaeumer U Kroll J. KLEIP deficiency in mice causes progressive corneal neovascular dystrophy. Invest Ophthalmol Vis Sci. 2012; 53: 3260–3268. [CrossRef] [PubMed]
Giacomini C Ferrari G Bignami F Rama P. Alkali burn versus suture-induced corneal neovascularization in C57BL/6 mice: an overview of two common animal models of corneal neovascularization. Exp Eye Res. 2014; 121: 1–4. [CrossRef] [PubMed]
Cade F Paschalis EI Regatieri CV Vavvas DG Dana R Dohlman CH. Alkali burn to the eye: protection using TNF-alpha inhibition. Cornea. 2014; 33: 382–389. [CrossRef] [PubMed]
Shin YJ Hyon JY Choi WS Chemical injury-induced corneal opacity and neovascularization reduced by rapamycin via TGF-beta1/ERK pathways regulation. Invest Ophthalmol Vis Sci. 2013; 54: 4452–4458. [CrossRef] [PubMed]
Ferrari G Bignami F Giacomini C Franchini S Rama P. Safety and efficacy of topical infliximab in a mouse model of ocular surface scarring. Invest Ophthalmol Vis Sci. 2013; 54: 1680–1688. [CrossRef] [PubMed]
Jin Y Chauhan SK El Annan J Sage PT Sharpe AH Dana R. A novel function for programmed death ligand-1 regulation of angiogenesis. Am J Pathol. 2011; 178: 1922–1929. [CrossRef] [PubMed]
Dana MR Zhu SN Yamada J. Topical modulation of interleukin-1 activity in corneal neovascularization. Cornea. 1998; 17: 403–409. [CrossRef] [PubMed]
Sivak JM Ostriker AC Woolfenden A Pharmacologic uncoupling of angiogenesis and inflammation during initiation of pathological corneal neovascularization. J Biol Chem. 2011; 286: 44965–44975. [CrossRef] [PubMed]
Lin Z He H Zhou T A mouse model of limbal stem cell deficiency induced by topical medication with the preservative benzalkonium chloride. Invest Ophthalmol Vis Sci. 2013; 54: 6314–6325. [CrossRef] [PubMed]
Das P Pereira JA Chaklader M Phenotypic alteration of limbal niche-associated limbal epithelial stem cell deficiency by ultraviolet-B exposure-induced phototoxicity in mice. Biochem Cell Biol. 2013; 91: 165–175. [CrossRef] [PubMed]
Mills RA Coster DJ Williams KA. Effect of immunosuppression on outcome measures in a model of rat limbal transplantation. Invest Ophthalmol Vis Sci. 2002; 43: 647–655. [PubMed]
Cursiefen C Cao J Chen L Inhibition of hemangiogenesis and lymphangiogenesis after normal-risk corneal transplantation by neutralizing VEGF promotes graft survival. Invest Ophthalmol Vis Sci. 2004; 45: 2666–2673. [CrossRef] [PubMed]
Mulik S Sharma S Suryawanshi A Activation of endothelial roundabout receptor 4 reduces the severity of virus-induced keratitis. J Immunol. 2011; 186: 7195–7204. [CrossRef] [PubMed]
Xia L Zhang S Cao Z Hu Y Yang H Wang D. Interleukin-17 enhanced immunoinflammatory lesions in a mouse model of recurrent herpetic keratitis. Microbes Infect. 2013; 15: 126–139. [CrossRef] [PubMed]
Niederkorn JY Ubelaker JE Martin JM. Vascularization of corneas of hairless mutant mice. Invest Ophthalmol Vis Sci. 1990; 31: 948–953. [PubMed]
Smith RS Hawes NL Kuhlmann SD Corn1: a mouse model for corneal surface disease and neovascularization. Invest Ophthalmol Vis Sci. 1996; 37: 397–404. [PubMed]
Cursiefen C Ikeda S Nishina PM Spontaneous corneal hem- and lymphangiogenesis in mice with destrin-mutation depend on VEGFR3 signaling. Am J Pathol. 2005; 166: 1367–1377. [CrossRef] [PubMed]
Cursiefen C Masli S Ng TF Roles of thrombospondin-1 and −2 in regulating corneal and iris angiogenesis. Invest Ophthalmol Vis Sci. 2004; 45: 1117–1124. [CrossRef] [PubMed]
Haddadin RI Oh DJ Kang MH Thrombospondin-1 (TSP1)-null and TSP2-null mice exhibit lower intraocular pressures. Invest Ophthalmol Vis Sci. 2012; 53: 6708–6717. [CrossRef] [PubMed]
Turpie B Yoshimura T Gulati A Rios JD Dartt DA Masli S. Sjogren's syndrome-like ocular surface disease in thrombospondin-1 deficient mice. Am J Pathol. 2009; 175: 1136–1147. [CrossRef] [PubMed]
Contreras-Ruiz L Regenfuss B Mir FA Kearns J Masli S. Conjunctival inflammation in thrombospondin-1 deficient mouse model of Sjogren's syndrome. PLoS One. 2013; 8: e75937. [CrossRef] [PubMed]
Maruyama K Nakazawa T Cursiefen C The maintenance of lymphatic vessels in the cornea is dependent on the presence of macrophages. Invest Ophthalmol Vis Sci. 2012; 53: 3145–3153. [CrossRef] [PubMed]
Mwaikambo BR Sennlaub F Ong H Chemtob S Hardy P. Genetic ablation of CD36 induces age-related corneal neovascularization. Cornea. 2008; 27: 1037–1041. [CrossRef] [PubMed]
Nakamura T Hamuro J Takaishi M LRIG1 inhibits STAT3-dependent inflammation to maintain corneal homeostasis. J Clin Invest. 2014; 124: 385–397. [CrossRef] [PubMed]
Chatterjee S Wang Y Duncan MK Naik UP. Junctional adhesion molecule-A regulates vascular endothelial growth factor receptor-2 signaling-dependent mouse corneal wound healing. PLoS One. 2013; 8: e63674. [CrossRef] [PubMed]
Cooke VG Naik MU Naik UP. Fibroblast growth factor-2 failed to induce angiogenesis in junctional adhesion molecule-A-deficient mice. Arterioscler Thromb Vasc Biol. 2006; 26: 2005–2011. [CrossRef] [PubMed]
Davis J Piatigorsky J. Overexpression of Pax6 in mouse cornea directly alters corneal epithelial cells: changes in immune function, vascularization, and differentiation. Invest Ophthalmol Vis Sci. 2011; 52: 4158–4168. [CrossRef] [PubMed]
Ramaesh T Collinson JM Ramaesh K Kaufman MH West JD Dhillon B. Corneal abnormalities in Pax6+/- small eye mice mimic human aniridia-related keratopathy. Invest Ophthalmol Vis Sci. 2003; 44: 1871–1878. [CrossRef] [PubMed]
Baulmann DC Ohlmann A Flugel-Koch C Goswami S Cvekl A Tamm ER. Pax6 heterozygous eyes show defects in chamber angle differentiation that are associated with a wide spectrum of other anterior eye segment abnormalities. Mech Dev. 2002; 118: 3–17. [CrossRef] [PubMed]
Nacak TG Alajati A Leptien K The BTB-Kelch protein KLEIP controls endothelial migration and sprouting angiogenesis. Circ Res. 2007; 100: 1155–1163. [CrossRef] [PubMed]
Kaminska GM Niederkorn JY. Spontaneous corneal neovascularization in nude mice. Local imbalance between angiogenic and anti-angiogenic factors. Invest Ophthalmol Vis Sci. 1993; 34: 222–230. [PubMed]
Vauclair S Majo F Durham AD Ghyselinck NB Barrandon Y Radtke F. Corneal epithelial cell fate is maintained during repair by Notch1 signaling via the regulation of vitamin A metabolism. Dev Cell. 2007; 13: 242–253. [CrossRef] [PubMed]
Ikeda S Cunningham LA Boggess D Aberrant actin cytoskeleton leads to accelerated proliferation of corneal epithelial cells in mice deficient for destrin (actin depolymerizing factor). Hum Mol Genet. 2003; 12: 1029–1037. [CrossRef] [PubMed]
Hara T Ishida H Raziuddin R Dorkhom S Kamijo K Miki T. Novel kelch-like protein, KLEIP, is involved in actin assembly at cell-cell contact sites of Madin-Darby canine kidney cells. Mol Biol Cell. 2004; 15: 1172–1184. [CrossRef] [PubMed]
Kawakami-Schulz SV Sattler SG Doebley AL Ikeda A Ikeda S. Genetic modification of corneal neovascularization in Dstn (corn1) mice. Mamm Genome. 2013; 24: 349–357. [CrossRef] [PubMed]
Kawakami-Schulz SV Verdoni AM Sattler SG Ikeda A Ikeda S. Differences in corneal phenotypes between destrin mutants are due to allelic difference and modified by genetic background. Mol Vis. 2012; 18: 606–616. [PubMed]
Jin Y Arita M Zhang Q Anti-angiogenesis effect of the novel anti-inflammatory and pro-resolving lipid mediators. Invest Ophthalmol Vis Sci. 2009; 50: 4743–4752. [CrossRef] [PubMed]
Cho YK Uehara H Young JR Archer B Ambati BK. Effect of glucocorticoid (triamcinolone acetonide) pretreatment in a murine penetrating keratoplasty and suture model. Cornea. 2012; 31: 1468–1475. [CrossRef] [PubMed]
Steven P Bock F Huttmann G Cursiefen C. Intravital two-photon microscopy of immune cell dynamics in corneal lymphatic vessels. PLoS One. 2011; 6: e26253. [CrossRef] [PubMed]
Bock F Onderka J Hos D Horn F Martus P Cursiefen C. Improved semiautomatic method for morphometry of angiogenesis and lymphangiogenesis in corneal flatmounts. Exp Eye Res. 2008; 87: 462–470. [CrossRef] [PubMed]
Blacher S Detry B Bruyere F Foidart JM Noel A. Additional parameters for the morphometry of angiogenesis and lymphangiogenesis in corneal flat mounts. Exp Eye Res. 2009; 89: 274–276. [CrossRef] [PubMed]
Dastjerdi MH Al-Arfaj KM Nallasamy N Topical bevacizumab in the treatment of corneal neovascularization: results of a prospective, open-label, noncomparative study. Arch Ophthalmol. 2009; 127: 381–389. [CrossRef] [PubMed]
Fraz MM Remagnino P Hoppe A Blood vessel segmentation methodologies in retinal images--a survey. Comput Methods Programs Biomed. 2012; 108: 407–433. [CrossRef] [PubMed]
Ecoiffier T Yuen D Chen L. Differential distribution of blood and lymphatic vessels in the murine cornea. Invest Ophthalmol Vis Sci. 2010; 51: 2436–2440. [CrossRef] [PubMed]
Truong T Huang E Yuen D Chen L. Corneal lymphatic valve formation in relation to lymphangiogenesis. Invest Ophthalmol Vis Sci. 2014; 55: 1876–1883. [CrossRef] [PubMed]
Shi W Ming C Liu J Wang T Gao H. Features of corneal neovascularization and lymphangiogenesis induced by different etiological factors in mice. Graefes Arch Clin Exp Ophthalmol. 2011; 249: 55–67. [CrossRef] [PubMed]
Cursiefen C Colin J Dana R Consensus statement on indications for antiangiogenic therapy in the management of corneal diseases associated with neovascularisation: outcome of an expert roundtable. Br J Ophthalmol. 2012; 96: 3–9. [CrossRef] [PubMed]
Adini I Ghosh K Adini A Melanocyte-secreted fibromodulin promotes an angiogenic microenvironment. J Clin Invest. 2014; 124: 425–436. [CrossRef] [PubMed]
Panigrahy D Kalish BT Huang S Epoxyeicosanoids promote organ and tissue regeneration. Proc Natl Acad Sci U S A. 2013; 110: 13528–13533. [CrossRef] [PubMed]
Yang J McNeish B Butterfield C Moses MA. Lipocalin 2 is a novel regulator of angiogenesis in human breast cancer. FASEB J. 2013; 27: 45–50. [CrossRef] [PubMed]
Lee HK Chauhan SK Kay E Dana R. Flt-1 regulates vascular endothelial cell migration via a protein tyrosine kinase-7-dependent pathway. Blood. 2011; 117: 5762–5771. [CrossRef] [PubMed]
Ghanem RC Han KY Rojas J Semaphorin 7A promotes angiogenesis in an experimental corneal neovascularization model. Curr Eye Res. 2011; 36: 989–996. [CrossRef] [PubMed]
Zhou H Jiang S Chen J Ren X Jin J Su SB. Largazole, an inhibitor of class I histone deacetylases, attenuates inflammatory corneal neovascularization. Eur J Pharmacol. 2014;.
Shen M Yuan F Jin J Yuan Y. The effect of TC14012 on alkali burn-induced corneal neovascularization in mice. Ophthalmic Res. 2014; 52: 17–24. [CrossRef] [PubMed]
Yuen D Grimaldo S Sessa R Role of angiopoietin-2 in corneal lymphangiogenesis. Invest Ophthalmol Vis Sci. 2014; 55: 3320–3327. [CrossRef] [PubMed]
Chan MF Li J Bertrand A Protective effects of matrix metalloproteinase-12 following corneal injury. J Cell Sci. 2013; 126: 3948–3960. [CrossRef] [PubMed]
Toyono T Usui T Yokoo S Angiopoietin-like protein 2 is a potent hemangiogenic and lymphangiogenic factor in corneal inflammation. Invest Ophthalmol Vis Sci. 2013; 54: 4278–4285. [CrossRef] [PubMed]
Yang H Li X Ma J Blockade of the intermediate-conductance Ca(2+)-activated K+ channel inhibits the angiogenesis induced by epidermal growth factor in the treatment of corneal alkali burn. Exp Eye Res. 2013; 110: 76–87. [CrossRef] [PubMed]
Liu GQ Lu PR Li LB Zhang XG. Inhibited experimental corneal neovascularization by neutralizing anti-SDF-1alpha antibody. Int J Ophthalmol. 2012; 5: 7–12. [PubMed]
Yan ZX Jiang ZH Liu NF. Angiopoietin-2 promotes inflammatory lymphangiogenesis and its effect can be blocked by the specific inhibitor L1-10. Am J Physiol Heart Circ Physiol. 2012; 302: H215–223. [CrossRef] [PubMed]
Markowska AI Jefferies KC Panjwani N. Galectin-3 protein modulates cell surface expression and activation of vascular endothelial growth factor receptor 2 in human endothelial cells. J Biol Chem. 2011; 286: 29913–29921. [CrossRef] [PubMed]
Hos D Regenfuss B Bock F Onderka J Cursiefen C. Blockade of insulin receptor substrate-1 inhibits corneal lymphangiogenesis. Invest Ophthalmol Vis Sci. 2011; 52: 5778–5785. [CrossRef] [PubMed]
Hos D Saban DR Bock F Suppression of inflammatory corneal lymphangiogenesis by application of topical corticosteroids. Arch Ophthalmol. 2011; 129: 445–452. [CrossRef] [PubMed]
Lin Q Yang XP Fang D High-mobility group box-1 mediates toll-like receptor 4-dependent angiogenesis. Arterioscler Thromb Vasc Biol. 2011; 31: 1024–1032. [CrossRef] [PubMed]
Grimaldo S Yuen D Ecoiffier T Chen L. Very late antigen-1 mediates corneal lymphangiogenesis. Invest Ophthalmol Vis Sci. 2011; 52: 4808–4812. [CrossRef] [PubMed]
Tang XL Sun JF Wang XY Du LL Liu P. Blocking neuropilin-2 enhances corneal allograft survival by selectively inhibiting lymphangiogenesis on vascularized beds. Mol Vis. 2010; 16: 2354–2361. [PubMed]
Koo BH Coe DM Dixon LJ ADAMTS9 is a cell-autonomously acting, antiangiogenic metalloprotease expressed by microvascular endothelial cells. Am J Pathol. 2010; 176: 1494–1504. [CrossRef] [PubMed]
Chen Y Carlson EC Chen ZY Conditional deletion of Cited2 results in defective corneal epithelial morphogenesis and maintenance. Dev Biol. 2009; 334: 243–252. [CrossRef] [PubMed]
Dou GR Wang YC Hu XB RBP-J, the transcription factor downstream of Notch receptors, is essential for the maintenance of vascular homeostasis in adult mice. FASEB J. 2008; 22: 1606–1617. [CrossRef] [PubMed]
Bellner L Patil KA Castellano K Halilovic A Dunn MW Schwartzman ML. Targeted suppression of HO-2 gene expression impairs the innate anti-inflammatory and repair responses of the cornea to injury. Mol Vis. 2011; 17: 1144–1152. [PubMed]
Figure 1
 
Regulation of physiological avascularity, induction of corneal NV and therapeutic intervention in the cornea. Whenever the balance of pro- and antiangiogenic transmitters in the cornea is disturbed, corneal NV occurs. In this Figure, the progression from physiological avascularity to complete vascularization is shown. Blue represents DAPI-positive epithelial cell nuclei, red lines represent lymphatics, green lines represent blood vessels. I, epithelium; II, Bowman's layer; III, stroma; IV, Descemet's membrane; V, endothelium. (A) A number of mediators (orange) actively maintain corneal avascularity. Green margin indicates the inhibition of blood vessel growth, red margin indicates the inhibition of lymphatic vessel growth. (B) The most relevant proangiogenic factors are shown in yellow. Various studies have shown that if the concentration of any one of these transmitters is experimentally elevated to a certain degree, blood vessels and/or lymphatics sprout from the corneal limbal plexus toward the corneal center. Also, macrophages (Mϕ) are involved in corneal NV. (C) Several externally applied substances can inhibit angiogenesis or drive vessels into regression. Some prominent examples are shown. TKI, tyrosine-kinase inhibitors; NSAID, nonsteroidal anti-inflammatory drugs.
Figure 1
 
Regulation of physiological avascularity, induction of corneal NV and therapeutic intervention in the cornea. Whenever the balance of pro- and antiangiogenic transmitters in the cornea is disturbed, corneal NV occurs. In this Figure, the progression from physiological avascularity to complete vascularization is shown. Blue represents DAPI-positive epithelial cell nuclei, red lines represent lymphatics, green lines represent blood vessels. I, epithelium; II, Bowman's layer; III, stroma; IV, Descemet's membrane; V, endothelium. (A) A number of mediators (orange) actively maintain corneal avascularity. Green margin indicates the inhibition of blood vessel growth, red margin indicates the inhibition of lymphatic vessel growth. (B) The most relevant proangiogenic factors are shown in yellow. Various studies have shown that if the concentration of any one of these transmitters is experimentally elevated to a certain degree, blood vessels and/or lymphatics sprout from the corneal limbal plexus toward the corneal center. Also, macrophages (Mϕ) are involved in corneal NV. (C) Several externally applied substances can inhibit angiogenesis or drive vessels into regression. Some prominent examples are shown. TKI, tyrosine-kinase inhibitors; NSAID, nonsteroidal anti-inflammatory drugs.
Figure 2
 
An exponentially increasing number of studies investigates corneal NV In this plot, the cumulative number of new published studies on corneal NV is shown for the last 50 years. The cornea is emerging as a widely-used model to study angiogenesis and the number of new studies has increased strongly over time. Cumulative data, plotted for each year, not smoothed. Source: Pubmed (National Center for Biotechnology Information [NCBI], National Institutes of Health [NIH], Bethesda MD, USA), search query “cornea AND (neovascularization OR angiogenesis).”
Figure 2
 
An exponentially increasing number of studies investigates corneal NV In this plot, the cumulative number of new published studies on corneal NV is shown for the last 50 years. The cornea is emerging as a widely-used model to study angiogenesis and the number of new studies has increased strongly over time. Cumulative data, plotted for each year, not smoothed. Source: Pubmed (National Center for Biotechnology Information [NCBI], National Institutes of Health [NIH], Bethesda MD, USA), search query “cornea AND (neovascularization OR angiogenesis).”
Figure 3
 
Common experimental approaches to induce corneal NV in mice In this Figure, the four most frequently used approaches for NV induction in murine corneas are shown. (A) corneal pocket assay (depicted here is the growth of blood vessels and lymphatics from opposite sides as observed with bFGF pellets). (B) Suture injury model. (C) Chemical injury by alkali burn. (D) Spontaneous transgenic model. Red arrows represent lymphatics, green arrows represent blood vessels.
Figure 3
 
Common experimental approaches to induce corneal NV in mice In this Figure, the four most frequently used approaches for NV induction in murine corneas are shown. (A) corneal pocket assay (depicted here is the growth of blood vessels and lymphatics from opposite sides as observed with bFGF pellets). (B) Suture injury model. (C) Chemical injury by alkali burn. (D) Spontaneous transgenic model. Red arrows represent lymphatics, green arrows represent blood vessels.
Figure 4
 
Corneal blood vessels and lymphatics imaged with confocal microscopy. Compared to other mouse tissues, the cornea is an ideal organ to study experimentally induced angiogenesis, because it can easily be imaged in conventional and confocal fluorescence microscopy. Using confocal microscopy, the corneal tissue can be imaged as a whole, so that the complete corneal vasculature can be analyzed. In this Figure, CD31-positive blood and lymphatic vessels (A) and LYVE-1–positive lymphatics (B) in a KLEIP−/− cornea are shown. (C) Shows the overlay of CD31 and LYVE-1 staining.
Figure 4
 
Corneal blood vessels and lymphatics imaged with confocal microscopy. Compared to other mouse tissues, the cornea is an ideal organ to study experimentally induced angiogenesis, because it can easily be imaged in conventional and confocal fluorescence microscopy. Using confocal microscopy, the corneal tissue can be imaged as a whole, so that the complete corneal vasculature can be analyzed. In this Figure, CD31-positive blood and lymphatic vessels (A) and LYVE-1–positive lymphatics (B) in a KLEIP−/− cornea are shown. (C) Shows the overlay of CD31 and LYVE-1 staining.
Figure 5
 
Manual quantification of NV in corneal whole mounts. In mouse models of corneal NV, such as the transgenic KLEIP−/− model, vessels sprout from the vascular plexus at the limbus toward the corneal center. The extent of corneal NV can be quantified by measuring the remaining avascular area. This is shown in (A) for blood vessels and in (B) for lymphatics (area was delineated by hand in sample images). (C) A DAPI staining of a control cornea is shown. Measuring DAPI-positive area permits normalization of area measurements to total corneal area.
Figure 5
 
Manual quantification of NV in corneal whole mounts. In mouse models of corneal NV, such as the transgenic KLEIP−/− model, vessels sprout from the vascular plexus at the limbus toward the corneal center. The extent of corneal NV can be quantified by measuring the remaining avascular area. This is shown in (A) for blood vessels and in (B) for lymphatics (area was delineated by hand in sample images). (C) A DAPI staining of a control cornea is shown. Measuring DAPI-positive area permits normalization of area measurements to total corneal area.
Figure 6
 
Different vascularization patterns can be distinguished quantitatively. Corneal NV can occur in different parts of the corneal stroma (e.g., subepithelial or deep stromal corneal NV). These vascularization patterns can be distinguished by quantitative measurements in corneal sections. This image shows subepithelial corneal NV in a KLEIP−/− mouse. Bowman's layer and Descemet's membrane have been delineated with a dashed line and the distance of each vessel from Bowman's layer is indicated by dotted lines. Quantifying these measurements permits exact representations of the distribution of corneal neovessels.
Figure 6
 
Different vascularization patterns can be distinguished quantitatively. Corneal NV can occur in different parts of the corneal stroma (e.g., subepithelial or deep stromal corneal NV). These vascularization patterns can be distinguished by quantitative measurements in corneal sections. This image shows subepithelial corneal NV in a KLEIP−/− mouse. Bowman's layer and Descemet's membrane have been delineated with a dashed line and the distance of each vessel from Bowman's layer is indicated by dotted lines. Quantifying these measurements permits exact representations of the distribution of corneal neovessels.
Table 1
 
Recent Corneal Pocket Assay Studies
Table 1
 
Recent Corneal Pocket Assay Studies
Mechanism Intervention Reference Model
Inhibition of HA by depletion of natural killer cells; reduced expression of VEGF-A, VEGF-C, and VEGFR3 Anti-NK1.1 antibody 56 bFGF pellet
Inhibition of HA in FMOD−/− mice FMOD−/− mice 109 bFGF pellet
Increased VEGF induced HA in Tie2-CYP2C8-Tr mice Tie2-CYP2C8-Tr mice 110 VEGF pellet
Inhibition of VEGF induced HA H-RN derived peptide from hepatocyte growth factor kringle 1 domain (HGF K1) 58 VEGF pellet
Lcn2 enhances VEGF induced HA Lipocalin 2 (Lcn2) protein administration 111 VEGF pellet
Protein tyrosine kinase 7 (PTK7) forms complex with VEGFR-1, inhibition of HA after PTK7 silencing PTK7 siRNA injection 112 VEGF pellet
Semaphorin 7A, inhibition of HA Sem7A expression vector 113 bFGF pellet
Table 2
 
Non-VEGF–Based Strategies to Inhibit Corneal NV in Injury Models
Table 2
 
Non-VEGF–Based Strategies to Inhibit Corneal NV in Injury Models
Mechanism/Molecules HA LA Intervention Reference Model
Inhibition of VEGF, bFGF, TGF-β1, and EGF; upregulation of Tsp-1, Tsp-2, and ADAMTS-1 Largazole (class I HDAC inhibitor) 114 Alkali
CXCR4, CXCR7, and others TC14012 (CXCR4 antagonist and CXCR7 agonist) 115 Alkali
Inhibition of HA and LA in angiopoietin-2−/− mice Ang-2−/− mice 116 Suture
C57BL/6 and FVB mice N/A 63 Alkali and suture
TNF-α Anti–TNF-α antibody 64 Alkali
Negative regulation of HA, CXCL1, and CCL2 MMP12 KO mice 117 Alkali
Reduced TGF-β and IL6 expression Rapamycin 65 Alkali
Angiopoietin-like protein 2; regulation of F4/80 and IL-1β expression Angptl2−/− mice and K14-Angptl2 mice, subconjunctival Angptl2 siRNA injection 118 Suture
Intermediate-conductance Ca(2+)-activated K(+) channel (KCa3.1), inhibition of EGF-induced HA TRAM-34 119 Alkali
SDF1, reduced VEGF and c-Kit expression Neutralizing anti–SDF-1α antibody 120 Alkali
Inhibition of LA, but activation of HA by Ang-2 blockage L1-10, an Ang-2-specific inhibitor 121 Suture
Galectin-3 increases phosphorylation of VEGFR-2, positive regulation of HA Gal3−/− and Mgat5−/− mice 122 Suture
IRS-1, inhibition of HA and LA GS-101 (Aganirsen), an antisense oligonucleotide against IRS-1 123 Suture
Increased expression of Robo4 upon HSV infection; Robo4 ligand Slit2 inhibits HA Injection of Slit protein 74 HSV-infection
Inhibition of inflammation Corticosteroids 124 Suture
Programmed death ligand-1 (PD-L1), negative regulator of HA PD-L1−/− mice 67 Suture
Toll-like receptor 4 (TLR4), activation of HA by TLR4 ligand HMGB1 TLR4−/− and TLR2−/− mice, topical application of HMGB1 125 Alkali
Very late antigen-1 (VLA-1) VLA-1 neutralizing antibody 126 Suture
Neuropilin 2 (NRP2) NRP2 artificial miRNA 127 Suture
Table 3
 
Mouse Models Showing Spontaneous Corneal NV
Table 3
 
Mouse Models Showing Spontaneous Corneal NV
Mouse Model Mechanism HA LA Time Course Intervention Special Features Reference
Destrin−/− mice, corn1 mice Pro-angiogenic VEGFR-3 signaling, deficient in sVEGFR-1 Starts at 20 days Rescue by VEGFR-3 neutralization or application of sVEGFR-1 Spontaneous corneal HA and LA, differentially regulated by genetic background, different time kinetics of HA and LA 10, 77, 78, 94, 96, 97
TSP-1−/− mice Loss of antiangiogenic signaling of CD36+ macrophages Starts at 6 mo Rescued by wild-type bone marrow cells TSP-1 inhibits VEGF-C release and LA by binding to its receptor CD36 on monocytes 13
CD36−/− mice Loss of TSP-1 – CD36 interaction Late onset (approximately 1 y) None Upregulation of VEGF-A is observed when CD36 is knocked out 13, 84
KLEIP/KLHL20−/− mice Proangiogenic miR-204 – Ang-1 pathway Median time 8 wk Anti-VEGF therapy not effective Includes macrophage infiltration and spontaneous corneal NV, accelerated by superficial abrasion 24, 62
LRIG1−/− mice LRIG1 acts as a negative regulator of STAT3 Median time 18 mo STAT3 inhibition by STA21 or wild-type bone marrow cells rescue phenotype Spontaneous, can be induced by injury 85
K5.Stat3C mice Proinflammatory STAT3 signaling Similar to LRIG1−/− None Similar phenotype to LRIG1 deficient mice 85
Jam-A−/− mice VEGF expression and TGF-β activation Starting early but low prevalence Inhibition of HA by DC101 (anti VEGFR-2 antibody) Spontaneous, can be accelerated by injury, approximately 50% prevalence at 6–12 mo 86, 87
ADAMTS9+/− mice ADAMTS9 is an antiangiogenic metalloprotease Median time approximately 6 wk None 80% prevalence in heterozygous mice (only in C57BL/6 strain). Null mice are not viable 128
Le-Cre;Cited2loxP/loxP Presumably via Pax6 and Klf4 8 mo None High penetrance 129
MxCre;RBPf/f Downstream of Notch receptor Approximately 2 mo None Spontaneous NV affects various ocular and extraocular tissues (cornea, retina, liver, lung) 130
Pax6 overexpression using the Aldh3a1 promoter Chi3l4, Flt-1, Wif1 Approximately 6–12 wk None Up to 69% prevalence, associated with corneal inflammation 88
Pax6+/− mice Presumably due to epithelial defects; sVEGFR-1 deficiency Early phenotype starting at 2 weeks Rescued by application of sVEGFR-1 Incomplete prevalence (different prevalence reports), several anterior eye defects including corneal NV 10, 89, 90
Global FoxC1−/− mice and neural-crest specific FoxC1−/− Via VEGF – VEGFR-2 signaling Detectable at 15 days after birth Inhibition of HA by blocking VEGFR-2 signaling Linked to disrupted epithelial architecture; FoxC1−/− mice show spontaneous corneal NV, FoxC1± mice show increased corneal NV upon injury 15
LeCre;VEGFR-2loxP/loxP Lack of sVEGFR-2 At birth Inhibition of LA by blocking VEGFR-3 signaling 100% prevalence in N = 30 mice 14
pCre;VEGFR-1loxP/loxP Lack of sVEGFR-1 Starting 2 d after pCre injection None Fully vascularized at 14 days, high prevalence in n = 10 mice 10
C57BL/6 mice Unknown P6–P14 None During development, the mouse cornea presents lymphatic vessel sprouting and regression 41
Nude mice, nu/nu, and hairless mutant mouse strain, SKH1;hr/hr Proangiogenic factors probably coming from the epithelium 6–12 wk old mice None Spontaneous corneal NV associated with absence of hair in 2 different genetic mouse models 76, 92
Heme oxygenase (HO)-2 MMP-2 Not applicable shRNA injection Local knockdown, not transgenic. Negative regulation of HA 131
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×