This protocol describes both in vivo and ex vivo methods to fully visualize and characterize hyaloid vessels, a model of vascular regression in mouse eyes, using optical coherence tomography and fundus fluorescein angiography for the live imaging and ex vivo isolation and subsequent flat mount of hyaloid for quantitative analysis.
In the eye, the embryonic hyaloid vessels nourish the developing lens and retina and regress when the retinal vessels develop. Persistent or failed regression of hyaloid vessels can be seen in diseases such as persistent hyperplastic primary vitreous (PHPV), leading to an obstructed light path and impaired visual function. Understanding the mechanisms underlying the hyaloid vessel regression may lead to new molecular insights into the vascular regression process and potential new ways to manage diseases with persistent hyaloid vessels. Here we describe the procedures for imaging hyaloid in live mice with optical coherence tomography (OCT) and fundus fluorescein angiography (FFA) and a detailed technical protocol of isolating and flat-mounting hyaloid ex vivo for quantitative analysis. Low-density lipoprotein receptor-related protein 5 (LRP5) knockout mice were used as an experimental model of persistent hyaloid vessels, to illustrate the techniques. Together, these techniques may facilitate a thorough assessment of hyaloid vessels as an experimental model of vascular regression and studies on the mechanism of persistent hyaloid vessels.
The blood supply in the eye is essential to ensure the normal development of the retina and surrounding ocular tissues and to equip a proper visual function. There are three vascular beds in the eye: the retinal vasculature, the choroid, and a transient embryonic circulatory network of hyaloid vessels. The development of the ocular vasculature requires spatial and temporal coordination throughout embryogenesis and tissue maturation. Among the three vascular beds, the hyaloid vasculature is the first functional blood supply system to provide nutrition and oxygen to the newly formed embryonic lens and the developing retina. Hyaloid vessels regress at the same time that the retina vasculatures develop and mature1. The regression of hyaloid vasculature is pivotal to allow a clear visual pathway for the development of visual function; hence, this vascular regression process is as important as the growth of retinal vasculature. Impaired hyaloid regression may lead to eye diseases. Moreover, the regression of hyaloid vessels provides a model system to investigate the cellular and molecular mechanisms involved in the regulation of vascular regression, which may have implications for the angiogenic regulation in other organs as well.
The hyaloid vasculature, derived from the hyaloid artery (HA), is composed of vasa hyaloidea propria (VHP), tunica vasculosa lentis (TVL), and pupillary membrane (PM). It provides nourishment to the developing retina, the primary vitreous, and the lens during embryonic development2. Arising from the HA, VHP branches anteriorly through the vitreous to the lens. The TVL cups the posterior surface of the lens capsule, and anastomoses to the PM, which connects to the anterior ciliary arteries, covering the anterior surface of the lens2,3, resulting in the formation of a network of vessels in the PM3,4,5. Interestingly, there are no veins in the hyaloid vasculature, and the system makes use of choroidal veins to accomplish venous drainage.
In the human embryo, the hyaloid vasculature is nearly complete at approximately the ninth week of gestation and starts to regress when the first retinal vessels appear, during the fourth month of gestation2. Beginning with atrophy of the VHP, regression of the capillary networks of the TVL, the PM, and lastly, the HA occurs subsequently2,3. Meanwhile, the primary vitreous retracts and the secondary vitreous starts to form, composed of the extracellular matrix components, including collagen fibers. By the sixth month of gestation, the primary vitreous is reduced to a small transparent canal extending from the optic nerve disc to the lens, called the Cloquet’s canal or hyaloid canal, and the secondary vitreous becomes the main component of the posterior segment2,3. The hyaloid circulation vanishes mostly at 35 to 36 weeks of gestation, just before birth3.
Unlike humans, in whom hyaloid vasculature is completely regressed at birth, the mouse hyaloid vascular system starts to regress after birth. As the mouse retina is born avascular and retinal vessels develop postnatally, hyaloid vessels regress concurrently from postnatal day (P) 4 and are mostly completely regressed by P216 (Figure 1). The PM disappears first between P10 and P12, and the VHP disappears between P12 and P16, while a small number of TVL and HA cells remain even at P16, and by P21 the hyaloid vascular system regression is almost complete6. In the meantime, retinal vasculature begins developing after birth. The superficial layer of vascular plexus fully extends to the peripheral retina at P7–P8, the deep layer (located in the outer plexiform layer) develops from P7–P12, and finally, the intermediate plexus in the inner plexiform layer develops between P12 and P157. As the retinal vasculature develops, it gradually replaces the function of concomitantly regressing hyaloid vessels, providing nutrition and oxygen to the developing eye. The postnatal occurrence of hyaloid vessel regression in mice provides an easily accessible experimental model to observe and study the hyaloid vasculature, as well as the molecular basis governing vascular regression processes under both physiological and pathological conditions8.
Failure of hyaloid regression can be seen in diseases such as PHPV, which is a rare congenital developmental anomaly of the eye resulting from a failed or incomplete regression of the embryological, primary vitreous and hyaloid vasculature9. The mechanisms regulating the regression process of hyaloid vasculature are complicated and broadly studied. One major molecular pathway essential for the normal regression of hyaloid vessels is the Wnt signaling pathway10, as genetic mutations in this pathway affecting both Wnt ligand and receptors have been linked with PHPV in humans9. Experimental studies identified a Wnt ligand, Wnt7b, which is produced by macrophages around hyaloid vessels in the developing eye to mediate this regression process. Wnt7b activates Wnt signaling by binding with the receptors frizzled4 (FZD4)/LRP5 in adjacent endothelial cells to initiate cell apoptosis, leading to the regression of hyaloid vessels10. As a result, Wnt7b-deficient mice show a persistence of hyaloid vessels10. Similarly, a nonconventional Wnt ligand, Norrin (encoded by the Ndp gene), also binds to FZD4/LRP5 to induce the hyaloid vessel regression during development. Ndpy/-, Lrp5-/-, and Fzd4-/- mice all display postponed hyaloid vessel regression, supporting a critical regulatory role of Wnt signaling11,12,13,14,15,16. Moreover, another Wnt coreceptor LRP6 overlaps with LRP5 in their function on modulating the Wnt signaling pathway in hyaloid vascular endothelial cells17. Other factors that may also contribute to hyaloid regression include the hypoxia-inducible factor18,19, vascular endothelial growth factor20,21, collagen-1822,23, Arf24, angiopoietin-225, and bone morphogenetic protein-426. In this paper, we use Lrp5-/- mice as a model of persistent hyaloid vessels to demonstrate the techniques of assessing and characterizing hyaloid vasculature through both in vivo and ex vivo methods.
The visualization of hyaloid vasculature in vivo and ex vivo is essential for studying the mechanisms of hyaloid vessel regression. Current methods to observe hyaloid vasculature mainly focus on visualizing and analyzing the VHP and HA, through OCT and FFA images, eye cross sections, and the hyaloid flat mount. OCT and FFA are powerful in vivo imaging tools, allowing longitudinal observation in live animals after they have opened their eyes. Moreover, isolated hyaloid flat mount provides visualization of the whole hyaloid vasculature and a means to achieve an accurate quantification of the vessel numbers. Yet the delicate and fragile nature of hyaloid vessels and the resulting technical difficulties of its isolation may have limited its use in eye research somewhat10,17,27. In this paper, we provide a detailed protocol of the visualization of hyaloid vessels, combining both in vivo live retinal imaging and ex vivo isolated hyaloid flat mount to enhance the feasibility of these techniques. This protocol has been adapted with modification and expansion from previous publications on the in vivo method of live fundus and OCT imaging28 and the ex vivo method of isolated hyaloid flat mount11.
All animals were treated in accordance with the Association for Research in Vision and Ophthalmology (ARVO) Statement for the Use of Animals in Ophthalmic and Vision Research for animal experiments, following the Guidelines of the National Institutes of Health (NIH) regarding the care and use of animals for experimental procedures and the regulations set forth by the Institutional Animal Care and Use Committee (IACUC) at Boston Children’s Hospital. Lrp5-/- mice (stock no. 005823; Jackson Laboratory) and its wild-type (WT) control C57BL/6J mice (stock no. 000664; Jackson Laboratory) were used for this study.
1. Part I: In vivo imaging of hyaloid vessels using a rodent retinal imaging system
2. Part II: Ex vivo visualization of hyaloid vessels
In vivo imaging of hyaloid vessels in live mice
Figure 3A reveals cross-sectional views of OCT images for the retina and hyaloid tissues in 3-month-old WT and Lrp5-/- mice, an animal model with persistent hyaloid. The WT eye shows the absence of hyaloid tissue, whereas the Lrp5-/- eye shows two persistent hyaloid vessels derived from the optic nerve head. Figure 3B displays FFA images of persistent hyaloid vessels (green) in the fluorescent field in 6-week-old Lrp5-/- mice. The WT mouse shows no remnant of hyaloid vessels, and the Lrp5-/- mouse shows eight branches of hyaloid vessels in the vitreous body.
Ex vivo visualization of hyaloid vessels
Figure 4A,B demonstrates isolated hyaloid vessels as visualized in flat mounts, where vascular cells and associated macrophages were stained by their nuclei with DAPI staining (blue). The HA is at the center of each image, and hyaloid vessels are revealed by the DAPI-derived discontinuous lines. Each line represents one vessel of VHP. In Lrp5-/- mice, higher numbers of remaining hyaloid vessels were observed at P8 in flat mounts (Figure 4A). The WT mice had an average of 12 branches of hyaloid vessels at P8, whereas the age-matched Lrp5-/- pups showed around 25 branches of hyaloid vessels, demonstrating a significantly impaired regression of hyaloid vasculature (Figure 4B). In addition, delayed and incomplete retina vascular development, another characteristic often associated with persistent hyaloid vessels, was also observed in the Lrp5-/- pups (Figure 4C,D). Figure 5 shows the remaining hyaloid vessels in cross sections of Lrp5-/- eyes at P8, whereas the WT eyes do not display hyaloid vessels.
Figure 1: Schematic diagram depicting the developmental regression of hyaloid vasculature in mouse eyes. Hyaloid vessels and branches, including VHP, TVL, and PM, are derived from the HA, and occupy much of the space between the lens and the immature retina at birth (P0). Hyaloid involution in mice starts with the regression of PM capillaries as early as P4. At P8, PM, VHP, and TVL layers are continuously regressing, coinciding with the complete formation of the superficial layer of the retinal vessels. By P12, the involution of the PM layer is complete, whereas the atrophy of VHP and TVL is still in progress. In the meantime, a deep layer of retinal vessels starts to form during P7–P12. By P16, the regression of the hyaloid system is partially completed (with the remaining TVL vessels and the HA left), and the intermediate layer of retinal vascular plexus continues to develop. The retinal vasculature is fully mature by P21 and takes over the role of nourishing retinal tissues from the hyaloid vessels, which are now mostly regressed. At P21, the vitreous, in the absence of hyaloid vessels, shows a clear visual pathway. Dashed red lines represent the regressing vessels. VHP = vasa hyaloidea propria; TVL = tunica vasculosa lentis; PM = pupillary membrane; HA = hyaloid artery. Please click here to view a larger version of this figure.
Figure 2: Schematic diagram depicting the sequential procedure of hyaloid vessel isolation for ex vivo flat-mounting visualization. (A) Injection of gelatin into the enucleated eye through four injecting points (red dots) at the limbus, to solidify the vitreous body. (B) Removal of the cornea and optic nerve, careful dissection of the iris, and peeling off the sclera-choroid-RPE complex. (C) Flipping the remaining retinal cup with the lens to make the retina side face up; then, injecting PBS between the retina and the vitreous body to separate them, followed by peeling off the retinal layer. (D) Flipping upside down again the rest of the tissue, containing the lens with the surrounding hyaloid. Lifting the lens slightly to loosen the connection of TVL and VHP. (E) Cutting at the HA between the TVL and the VHP, and the removal of the lens and TVL. Flat-mounting VHP layer. VHP = vasa hyaloidea propria; TVL = tunica vasculosa lentis; PM = pupillary membrane; HA = hyaloid artery. Please click here to view a larger version of this figure.
Figure 3: In vivo imaging of hyaloid vessels with optical coherence tomography (OCT) and fundus fluorescence angiography (FFA). A: Representative OCT images of WT and Lrp5-/- mice at 3-month-old showing persistent hyaloid vessels in the vitreous space in Lrp5-/- eyes. B: Representative FFA images of WT and Lrp5-/- mice at 6-weeks-old. Mice were injected with 1mg fluorescein sodium in 0.1ml saline per mouse after anesthesia, and images were taken 5 min after injection, focused on hyaloid vessels in Lrp5-/- mice or vitreous body in WT (in the absence of hyaloid vessels). Persistent hyaloid vessels were visualized in Lrp5-/- but not WT eyes. Red arrows indicate hyaloid vessels. Both scale bars: 100 μm (A). Please click here to view a larger version of this figure.
Figure 4: Visualization of delayed hyaloid vessel regression and delayed retinal vasculature development in Lrp5-/- mice. (A) Representative images of flat-mounted hyaloid vessels stained with DAPI (blue) in WT and Lrp5-/- eyes at P8. Arrows (white) indicate the hyaloid artery. (B) Quantification of the numbers of hyaloid vessels branching from the hyaloid artery in WT and Lrp5-/- eyes. (C) Representative flat-mounted retinas stained with isolectin-IB4 (red) for the vasculature in WT and Lrp5-/- eyes at P8. White dashed lines indicate the retina edge. Yellow lines indicate the vascularized area edge. (D) Quantification of vascular coverage of the superficial retinal vascular plexus in WT and Lrp5-/- eyes. n = 8–12/group. The scale bars in panels A and C = 1 mm. Data are shown as mean ± SEM. Two-tailed Student’s t-test was used for statistical analysis. **P < 0.01. This figure is modified with permission from Wang et al.27. Please click here to view a larger version of this figure.
Figure 5: Visualization of hyaloid vessels in cross sections of Lrp5-/- eyes. Representative images of cross sections of eyes isolated from WT and Lrp5-/- mice at P8. The eyes were enucleated and embedded in optimal cutting temperature, and sections were cut using cryostat. The cross sections were stained with DAPI (blue) to show the nuclei and isolectin-IB4 (red) to show the blood vessels. White arrows indicate hyaloid vessels. The scale bar = 500 μm. This figure is adapted with permission from Chen et al.16. Please click here to view a larger version of this figure.
Techniques to assess and characterize hyaloid vessels are intuitive and necessary procedures to observe the hyaloid vessel regression in animal models, to allow studies on the mechanisms underlying the vascular regression during development. While the in vivo retinal imaging allows the longitudinal observation of hyaloid regression in the same animal, access to a rodent fundus imaging system for OCT and FFA may be a limiting factor. In addition, in vivo imaging in live mice is not feasible before they open their eyes. Therefore, this methodology is not applicable during eye development in the neonatal stage. On the other hand, while imaging cross sections of isolated eyes may be used for any age of the mouse and has a low technical barrier, it is not quantitative when there are only a few visible hyaloid vessels, and the chance of obtaining an ideal image depends largely on the angle of sectioning (Figure 5). In comparison, the isolation of hyaloid vessels for flat mount visualization allows the complete imaging of the whole hyaloid vessel and accurate quantification, yet technical challenges may exist due to the delicate and fragile nature of hyaloid vessels. We hope the protocol detailed in this paper helps overcome these challenges.
Several critical steps in the hyaloid isolation protocol include the injection of the gelatin solution, the removal of the retina, and the final flat mounting. The technical difficulty of handling hyaloid vasculature lies in its delicate nature, the almost liquid-like state of vitreous where hyaloid vessels reside, and its close connection with the adjacent lens and the retina. Injecting gelatin solution intravitreally is key to solidifying the vitreous body containing hyaloid vessels and, thereby, making their texture firmer for easier dissection and handling. Gelatin is a gelling agent forming transparent elastic thermoreversible gels. By injecting the liquid form of gelatin (at room temperature or at 37 °C) into the vitreous space and cooling it to a lower temperature (4 °C), the vitreous body transforms into a firmer gel cup. Performing multiple gelatin injections into the eye allows the full dispersion of a sufficient amount of the gelatin into the vitreous space, to form a uniform and round gelatinized hyaloid tissue cup. Another technical difficulty lies in removing the retina without damaging hyaloid tissue. Unlike the choroid, which is relatively easy to dissect apart, the gelatinized vitreous body containing hyaloid vessels is challenging to separate from the nearby retina without damaging the hyaloid. We found that injecting PBS into the space between the vitreous body and the retina generated a liquid buffer layer, making it much easier to separate these two tissues. This is somewhat similar to the hydrodissection technique used in cataract surgery to separate the capsule and the cataract cortex. The final flat mounting is the last technically difficult step of the protocol. Warming the gelatinized hyaloid tissue in a PBS droplet on a slide warmer converts it back to a more liquid-like form to allow easier flat mounting. Proper arrangement and orientation of the hyaloid gel cup is still essential to ensure its even flattening after melting and drying.
The protocol of isolating hyaloid vessels may be further modified in several ways. The concentration of gelatin solution we used is 5%, yet higher or lower concentrations may also work, depending on the preference of the researcher to achieve a firmer or softer texture for handling. The DAPI staining of cellular nuclei may also be modified with isolectin-IB4 or other endothelial cell or macrophage markers, to better distinguish vascular endothelial cells from macrophages. A word of caution: the flat-mounted hyaloid vessels are very delicate and only loosely adhered to the slide; hence, the hyaloid slides need to be handled very gently and carefully if rinsing is needed during staining. This isolation protocol also has its limitations, including the complexity of performing the fine dissection and the difficulty of the long-term preservation of the samples due to the nature of the fragile tissue. Nevertheless, isolating and flat-mounting hyaloid vessels is still the most comprehensive and intuitive way to study hyaloid vasculature, to promote ocular and angiogenesis studies10,17,27,29,30,31.
The hyaloid vasculature provides an excellent experimental model of studying developmental vessel growth and regression, relevant for research fields in ophthalmology, angiogenesis, and programmed cell death, with which this paper aims to assist. Future modifications of the isolation protocol may be directed at improving the feasibility of a technical dissection procedure. Overall, the combined in vivo imaging and ex vivo isolation of hyaloid vessels is an advantageous method to allow the full assessment and characterization of hyaloid vessels in mice as a useful model of vascular regression.
The authors have nothing to disclose.
This work was supported by National Institutes of Health (NIH) grants (R01 EY024963 and EY028100) to J.C. Z.W. was supported by a Knights Templar Eye Foundation Career Starter Grant. The hyaloid isolation procedure described in this study was adapted with modification from protocols generously shared by Drs. Richard Lang, Toshihide Kurihara, and Lois Smith, to whom the authors are thankful.
AK-Fluor (fluorescein injection, USP) | Akorn | 17478-253-10 | |
Anti-CD31 antibody | Abcam | ab28364 | |
Antifade mounting medium | Thermo Fisher | S2828 | |
Antifade Mounting Medium with DAPI | Vector Laboratories | H-1200 | |
Artificial tear eyedrop | Systane | N/A | |
Bovine serum albumin (BSA) | Sigma-Aldrich | A2058 | |
C57BL/6J mice | The Jackson Laboratory | Stock NO: 000664 | |
Calcium chloride (CaCl2) | Sigma-Aldrich | C1016 | |
Cryostat | Leica | CM3050S | |
Cryostat | Leica | CM3050 S | |
Cyclopentolate hydrochloride and phenylephrine hydrochloride eyedrop | Cyclomydril | N/A | |
Gelatin | Sigma-Aldrich | G9382 | |
Goat anti-Rabbit IgG (H+L) Cross-Adsorbed Secondary Antibody, Alexa Fluor 488 | ThermoFisher Scientific | A-11008 | |
Heating board | Lab-Line Instruments Inc. | N/A | |
Isolectin GS-IB4, 594 conjugate | ThermoFisher Scientific | I21413 | |
Ketamine hydrochloride injection | KetaVed | NDC 50989-996-06 | |
Lrp5-/- mice | The Jackson Laboratory | Stock NO. 005823 | Developed by Deltagen Inc., San Mateo, CA |
Micron IV and OCT | Phoenix Research Labs | N/A | Imaging software: InSight |
Microscope | Zeiss | discovery v8 | |
Microsurgery forceps | Scanlan International | 4004-05 | |
Microsurgery scissors | Scanlan International | 6006-44 | |
Optimal cutting temperature compound | Tissue-Tek | 4583 | |
Optimal cutting temperature compound | Agar Scientific | AGR1180 | |
Paraformaldehyde (16%) | Electron Microscopy Sciences | 15710 | |
Peel-A-Way disposable embedding molds (tissue molds) | Fisher Scientific | 12-20 | |
Phosphate-buffered saline (PBS) buffer (10X) | Teknova | P0496 | |
Slide cover glass | Premiere | 94-2222-10 | |
Superfrost microscope slides | Fisherbrand | 12-550-15 | |
Triton X-100 | Sigma-Aldrich | X100 | |
Xylazine sterile solution | Akorn: AnaSed | NDC: 59399-110-20 |