Molecular Vision 2005; 11:744-748 <>
Received 17 March 2004 | Accepted 4 September 2005 | Published 13 Septmember 2005

Experimental study on relationship between retinal vein occlusion and loss of vitreous gel mass

Jin Ma,1 Yao Ke,1 Dezheng Wu,2 Xiongfei Gu,2 Rulong Gao,2

1Eye Center of Affiliated Second Hospital, School of Medicine, Zhejiang University, Hangzhou, Peoples Republic of China; 2Zhongshan Ophthalmic Center, Zhongshan University, Guangzhou, Peoples Republic of China

Correspondence to: Dr. Jin Ma, Eye Center of Affiliated Second Hospital, School of Medicine, Zhejiang University, 88 Jiefang Road, Hangzhou, 310009, Peoples Republic of China; Phone: 86-571-87783897; FAX: 86-571-87783897; email:


Purpose: To explore the relationship between retinal vein occlusion (RVO) and loss of vitreous gel mass.

Methods: An experimental RVO was induced by a photodynamic method in one eye (experimental eye) of 17 chinchilla rabbits. The contralateral eye was used as an untreated control. The changes in vitreous gel mass and liquefaction were investigated one month later. Changes in molecular properties of type II collagen in vitreous were analyzed by western immunoblot analysis.

Results: An RVO was successfully induced in 13 of 17 chinchilla rabbits (76.5%), which caused loss of vitreous gel mass and loss of elasticity, accompanied by release of a water-like liquid from the gel. The α chains of type II collagen were crosslinked together to form high molecular weight components of β and γ, which weakened the stability of collagen net structure.

Conclusions: Loss of vitreous gel mass occurred in RVO eyes. The crosslinks of vitreous collagen may damage the stability of collagen structure, and promote loss of vitreous gel mass.


Retinal vein occlusion (RVO) is one of the most common vascular diseases in the human eye. Macular edema is a common complication in RVO which leads to loss of visual acuity. The pathogenesis of macular edema associated with RVO is considered to be multifactorial, and is not fully understood. Recently, it has been proposed [1,2] that the vitreoretinal adhesion may be implicated in the pathogenesis of the macular edema [3-6]. Premacular vitreous cortex seems to be the major reason for loss of gel mass, which promotes incomplete vitreous separation from the retina, and results in partial vitreous detachment with traction at the sites of residual vitreomacular adhesion. The traction may cause breakdown of the blood-retinal barrier leading to macular edema. Several studies have even reported favorable anatomic and functional results after vitrectomy and removal of the posterior hyaloid in macular edema associated with RVO [7-9].

The vitreous gel structure is maintained by a three dimensional network of randomly organized, nonbranching, mainly type II collagen fibrils stabilized by interfibrillar hyaluronic acid molecules [10]. Destruction of this three dimensional framework is believed to be the cause of vitreous liquefaction and loss of gel mass.

In order to determine the relationship between RVO and loss of vitreous gel mass, and to explore the mechanisms of vitreous changes during RVO, we investigated vitreous changes with special emphasis on collagen, in an experimental model of RVO in chinchilla rabbits.



Seventeen chinchilla rabbits (34 eyes) were used in this study, each rabbit weighting 2-2.5 kg. The animals were fed standard laboratory chow and treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research.

RVO formation by photodynamic method

The rabbits were anesthetized with a mixture of Ketalar and Rompun. Their pupils were dilated with 1% cyclopentolate eyedrops. Rose Bengal-mediated argon laser retinal vein photothrombosis was used to produce focal RVO in one eye of each rabbit. Rose Bengal solution 50 mg/kg (50 mg/ml in sterile water), was injected in an ear vein immediately before the laser treatment. Argon green laser treatments (300 mW, 0.5 s, 50 μm spot) were used to produce focal RVO. Each vein was treated at a distance of a half to one disc diameter from the optic disc, until the blood flow was completely stopped in the vein. The contralateral eye served as the control, received the same amount of argon laser treatment far from retinal vessels. Fundus photography and fundus fluorescein angiography (FFA) were performed 24 h after the laser treatment to confirm that the treated veins were completely occluded. Sodium fluorescein at a concentration of 10% (0.1 ml/kg) was injected in the ear vein. Any rabbits in which the treated retinal vessel was only partially occluded were excluded from the study.

Fundus examination

The examinations of fundus photographs and FFA were taken before laser photothrombosis, and after laser photothrombosis at 1 day, 4 days, 7 days, 10 days, and 15 days thereafter. The process of RVO were evaluated by all those continual examinations. Eyes were subsequently enucleated 1 month after it was established that RVO had formed.

Gel/liquid vitreous separation

Animals were killed with an overdose of pentobarbital sodium. The enucleated eyes were incised circumferentially about 3 mm posterior to the corneal limbus. The posterior portion, including the sclera, the choroid, and retina, was separated cleanly from the vitreous body. The total vitreous was dissected carefully from the ciliary body, the zonules, and the lens with blunt forceps. The gel/liquid vitreous separation was performed according to the method described by Ueon [11]. Each vitreous sample was poured onto a plastic, resin-coated fiberglass net (mesh opening approximately 1.5 mm; net size 10x16 cm) that was positioned on top of a rectangular filter paper (Fisher No. 1, 10x16 cm; Fisher, Pittsburgh, PA). Liquified vitreous ran through the mesh onto the filter paper, which absorbed the liquid immediately. The gel component of the vitreous was separated from the liquefied vitreous by the net. The gel portion held by the net was transferred to a preweighed plastic dish (8 cm diameter). Separation was done at room temperature for 30 s. The gel vitreous was weighed and the percentage of gel vitreous was calculated as the wet weight of the separated gel vitreous divided by the initial wet weight of the vitreous.

Gel properties

In order to evaluate loss of gel mass, the eyes were suspended by forceps after the posterior ocular portion was separated from the vitreous body and the gel length was measured (Figure 1).

Vitreous collagen extraction

Collagen was extracted from the rabbit vitreous, using a modification of the Seery and Davison method [12]. A total vitreous was mixed with 1 ml protease inhibitor cocktail (0.5 mg/l leupeptin, 1 mM Na2EDTA, 0.7 mg/l pepstatin, and 0.2 mM phenylmethylsulfonyl fluoride; Boehringer Mannheim, Indianapolis, IN), and homogenized at 0 °C in a Ten Broeck tissue grinder (Corning Glass Works, Corning, NY). The homogenate was centrifuged (100,000x g, 45 min) for 45 min at 4 °C in an ultracentrifuge (L3-50, Beckman, Palo Alto, CA). The insoluble residue, which included the collagen, was resuspended in 0.5 M acetic acid, treated with pepsin (10 mg of enzyme per 100 mg sample; pepsin, twice crystallized, Sigma, St. Louis, MO), and agitated overnight at 4 °C. After the enzyme treatment and centrifugation (11,000x g for 30 min), the pH of the supernatant was adjusted to 7.2 by the addition of 1 M NaOH, which then was dialyzed against 0.01 M Na2HPO4 at 4 °C overnight, dialyzed against distilled water at 4 °C overnight, and freeze-dried to yield solubilized vitreous collagen.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)

The vitreous collagen was dissolved in 0.5 M acetic acid, and electrophoresis sample buffer (2% sodium dodecyl sulfate (SDS), 0.0625 M Tris-HCL, pH 6.8, 10% glycerol, 5% 2-mercaptoethanol, and 0.025% bromophenol blue) was added to the collagen solutions [13]. The mixture was placed in boiling water for 5 min and centrifuged at 2,000x g for 5 min. The sample solution (20 μl) was loaded onto a polyacrylamide gradient gel (4-20%, Mini-Protean II Ready Gels, Bio-Rad, Richmond, CA). Constant voltage (200 volts) electrophoresis with running buffer (0.124 M Tris, 0.959 M glycine, and 0.5% SDS) was conducted for 45 min. The gel was stained with Coomassie blue (Sigma). After being stained, the gel was dried using a gel drier and scanned by the spectrophotometer with a gel-holder attachment. The molecular size of the collagen peptide was calculated from the migration distance of the peptide band on SDS-PAGE. Bovine dermal type I collagen (Vitreogen 100, 3.0 mg/ml; Celtrix Laboratories, Palo Alto, CA) was used as the molecular size standard for collagen [14].


Vitreous collagen on the gel was transblotted onto a nitrocellulose membrane (Bio-Rad) using the semi-dry method [15]. The membrane was incubated at 37 °C for 1 h with 3% bovine serum albumin in phosphate-buffered saline (PBS, pH 7.2) for blocking, and reacted with mouse anti-human type II collagen monoclonal antibody (Chemicon, Temecula, CA) at 37 °C for 1 h. This antibody reacts with pepsin-solubilized rabbit type II collagen, but has no cross-reactivity with type V or IX collagen [16]. The membrane was washed with PBS containing 0.05% Tween-20 and then reacted with peroxidase-conjugated sheep anti-mouse IgG (Sigma) at 37 °C for 1 h. The bands on the membrane were visualized with diaminobenzidine dihydrochloride (DAB; Cappel, Durham, NC) containing 0.1% hydrogen peroxide.

Data analysis

Statistical analysis was conducted with a nonparametric procedure because the population of the data could not be assumed to have a normal distribution. The Mann-Whitney U test was used for comparing the medians of the two groups.


Fundus examination

Flow of blood through the retinal vein was interrupted 10 to 20 min after photocoagulation. The retinal vein distal to occlusion point became more tortuous, dark, and dilated with very small retinal hemorrhages nearby. The retinal vein was not fully occluded in four experimental rabbits (8 eyes) the first day after photocoagulation, and they were excluded from this study. Four days after photocoagulation, no remarkable changes in fundus examination were detected in the remaining 13 of 17 treated animals. After 7 days, retinal bleeding was gradually absorbed. The RVO and laser treatment did not cause any vitreous hemorrhages in this study. On day 10, blood flow occlusion reversed partly. By day 15, the RVO reversed completely in all experimental eyes (Figure 2).

Vitreous liquid and gel evaluation

In the 13 rabbits with successful RVO formation, the mass of the vitreous gel was 58% (median; range: 53%-72%) of the total vitreous mass at 1 month after photocoagulation. This proportion was significantly lower than that of the contralateral eyes (median: 78%; range: 69%-86%; Z=-2.191, p<0.05 by Mann-Whitney U test).

Figure 1 shows the eye hung by forceps after the posterior ocular portion was separated from the vitreous body 1 month after RVO. Any liquified vitreous is poured out of the posterior segment prior to the measurement of the gel length. The gel length was significantly shorter in experimental eyes (median: 1.7 cm; range: 1.50-2.1 cm) than in controls (median: 2.2 cm; range: 1.90-2.4 cm, Z=-1.989; p<0.05 by Mann-Whitney U-test). These findings indicating a loss in the length of the vitreous gel supported the previous finding of a loss of mass of the vitreous gel in the experimental eye.

Vitreous collagen evaluation

Figure 3 shows SDS-PAGE and immunoblotting of rabbit vitreous collagen. The control sample (lane 2) showed an intense band of collagen peptide, typical of vitreous collagen [12,17,18]. Immunoblotting revealed the band to be the α-chain of type II collagen (lane 4). A less intensive band in the middle, the β-component of the type II collagen [17,18], was also visible. However, the γ-component of the type II collagen, a less intense band on the top, remained indistinct in the control samples (lanes 2 and 4), which was detected clearly in experimental ones (lanes 3 and 5). SDS-PAGE further revealed estimated molecular weights of the α-, β-, and γ-components of type II collagen of 101 kDa, 205 kDa, and 303 kDa, respectively. These values agreed with related reports [13]. The density of the α-chain of type II in the controls (lane 4) was similar to that of experimental eyes (lane 5; Z=-1.478; p>0.05), as measured by the optical density of the immunoblotted bands. However, bands of both the β- and γ-components of the experimental eyes appeared darker than those of the controls. The immunoblotted density ratios of the β- and γ-components to all components of type II collagen in experimental eyes were significantly higher than in the controls (Table 1), indicating the formation of a more crosslinked component with higher molecular weight forms of vitreous collagen.


It has been postulated that intractable macular edema associated with RVO unresponsive to laser treatment may be due to posterior hyaloid traction [3,4]. The abnormality of retinal microcirculation in RVO may cause the breakdown of the blood-retinal barriers. Serum components could leak into the vitreous gel because of increased vascular permeability, which induce vitreous liquefaction and condensation that often are associated with loss of vitreous gel mass. Vitreous contraction may result in partial vitreous detachment with traction at the sites of residual vitreoretinal adhesion, which in turn aggravates the breakdown of blood-retinal barrier, resulting in increased accumulation of extracellular fluid in the macula. Several clinical studies also support the fact that the presence of a posterior vitreous detachment is associated with a significantly reduced risk of developing macular edema in RVO patients [3,4,19]. Thus, an understanding of the mechanisms of vitreous changes during RVO may enable us to control these harmful changes and should be immensely helpful to patients with RVO. Only limited information is available about vitreous changes during RVO, and the mechanisms are poorly understood.

Our results demonstrated that RVO response causes loss of gel mass and possible loss of elasticity. The changes of vitreous gel structure were that a water-like liquid separated from the vitreous humor due to changes in the rheological properties of the gel component. The loss of gel mass may result from many potential causes including (but not limited to) syneresis, clotting, shrinkage of the three dimensional meshwork, cracking or tearing the meshwork, reduced gel strength, reduced elasticity, viscosity changes, and protein modifications that break down the meshwork [10,20-22]. These findings coincide with clinical observations [20], suggesting that this experimental model for RVO is probably useful in investigating the mechanisms of vitreous changes in human RVO eyes. As to the pathogenesis of the loss of vitreous gel mass, the damage to the stability of vitreous structure should be given more attention. The gel structure of the vitreous body is believed to be maintained by a meshwork of randomly organized collagen fibrils stabilized by hyaluronic acid molecules between the fibrils [10]. The crosslinks of collagen may damage the gel structure integrity, promote loss of gel mass, and promote the release a water-like liquid from the gel. Our results demonstrated that RVO can cause increased amounts of high molecular weight compoments in vitreous collagen (β- and γ-components), presumably due to extensive molecular crosslinks, which were evident by SDS-PAGE analyses (Figure 3). Thus, we speculate that the crosslinks of collagen may damage the gel structure integrity, promoting the loss of gel mass, and releasing a water-like liquid from the gel. However, further studies are needed to understand the mechanisms of crosslinks of collagen molecules in eyes with RVO. Several investigations reported [21] that the serum-induced crosslinks of collagens may play a role in the vitreous contraction seen in different vitreoretinal disorders. Serum components such as fibronectin (FN) and transglutaminase (TG) in the vitreous should increase in concentration because of increased vascular permeability. FN and TG were reported [20,23-25] to contribute to the formation of crosslinks of vitreous collagen (and other factors possibly contributing to crosslink formation during ocular diseases) can induce collagen gel contraction. The formation of collagen-FN-collagen crosslinks catalyzed by TG may play a major role in the vitreous contraction observed in several vitreoretinal disorders [23,26].

In conclusion, the present study demonstrated vitreous changes during experimental RVO and suggests possible mechanisms. Further studies are essential to clarify the nature of the RVO-induced changes in the vitreous. This knowledge may enable us to prevent vitreous changes and subsequent deterioration of visual function in eyes with RVO.


We thank Jiekai Jiang (Zhejiang University) for helpful discussions. Also, we thank Feng Wen (Zhongshan University) for his technical support.


1. Kado M, Jalkh AE, Yoshida A, Takahashi M, Wazen N, Trempe CL, Schepens CL. Vitreous changes and macular edema in central retinal vein occlusion. Ophthalmic Surg 1990; 21:544-9.

2. Roldan M, Serrano JM. Macular edema and vitreous detachment. Ann Ophthalmol 1989; 21:141-8.

3. Takahashi MK, Hikichi T, Akiba J, Yoshida A, Trempe CL. Role of the vitreous and macular edema in branch retinal vein occlusion. Ophthalmic Surg Lasers 1997; 28:294-9.

4. Hikichi T, Konno S, Trempe CL. Role of the vitreous in central retinal vein occlusion. Retina 1995; 15:29-33.

5. Glacet-Bernard A, Coscas G, Chabanel A, Zourdani A, Lelong F, Samama MM. Prognostic factors for retinal vein occlusion: prospective study of 175 cases. Ophthalmology 1996; 103:551-60.

6. Sebag J, Balazs EA. Pathogenesis of cystoid macular edema: an anatomic consideration of vitreoretinal adhesions. Surv Ophthalmol 1984; 28 Suppl:493-8.

7. Kumagai K, Ogino N, Furukawa M, Demizu S, Atsumi K, Kurihara H, Ishigooka H. [Vitreous surgery for macular edema in branch retinal vein occlusion]. Nippon Ganka Gakkai Zasshi 2002; 106:701-7.

8. Tachi N, Hashimoto Y, Ogino N. Vitrectomy for macular edema combined with retinal vein occlusion. Doc Ophthalmol 1999; 97:465-9.

9. Sekiryu T, Yamauchi T, Enaida H, Hara Y, Furuta M. Retina tomography after vitrectomy for macular edema of central retinal vein occlusion. Ophthalmic Surg Lasers 2000 May-Jun; 31:198-202.

10. Bishop PN. Structural macromolecules and supramolecular organisation of the vitreous gel. Prog Retin Eye Res 2000; 19:323-44.

11. Ueno N, Sebag J, Hirokawa H, Chakrabarti B. Effects of visible-light irradiation on vitreous structure in the presence of a photosensitizer. Exp Eye Res 1987; 44:863-70.

12. Seery CM, Davison PF. Collagens of the bovine vitreous. Invest Ophthalmol Vis Sci 1991; 32:1540-50.

13. Hikichi T, Akiba J, Ueno N, Yoshida A, Chakrabarti B. Cross-linking of vitreous collagen and degradation of hyaluronic acid induced by bilirubin-sensitized photochemical reaction. Jpn J Ophthalmol 1997; 41:154-9.

14. Kuhn K. The classical collagens: type I, II, and III. In: Mayne R, Burgenson RE, editors. Structure and function of collagen types. Orlando: Academic Press; 1987. p 1-42.

15. Kyhse-Andersen J. Electroblotting of multiple gels: a simple apparatus without buffer tank for rapid transfer of proteins from polyacrylamide to nitrocellulose. J Biochem Biophys Methods 1984; 10:203-9.

16. Glant TT, Hadhazy C, Mikecz K, Sipos A. Appearance and persistence of fibronectin in cartilage. Specific interaction of fibronectin with collagen type II. Histochemistry 1985; 82:149-58.

17. Liang JN, Chakrabarti B. Spectroscopic studies on pepsin-solubilized vitreous and cartilage collagens. Curr Eye Res 1981; 1:175-81.

18. Miller EJ. Structural studies on cartilage collagen employing limited cleavage and solubilization with pepsin. Biochemistry 1972; 11:4903-9.

19. Kishi S. [Vitreous and macular diseases]. Nippon Ganka Gakkai Zasshi 2003; 107:813-34.

20. Sebag J. The vitreous: structure, function, and pathobiology. New York: Springer-Verlag; 1989.

21. Akiba J, Ueno N, Chakrabarti B. Molecular mechanisms of posterior vitreous detachment. Graefes Arch Clin Exp Ophthalmol 1993; 231:408-12.

22. Akiba J, Ueno N, Chakrabarti B. Age-related changes in the molecular properties of vitreous collagen. Curr Eye Res 1993; 12:951-4.

23. Akiba J, Kakehashi A, Ueno N, Tano Y, Chakrabarti B. Serum-induced collagen gel contraction. Graefes Arch Clin Exp Ophthalmol 1995; 233:430-4.

24. Mosher DF. Cross-linking of fibronectin to collagenous proteins. Mol Cell Biochem 1984; 58:63-8.

25. Mosher DF, Schad PE, Vann JM. Cross-linking of collagen and fibronectin by factor XIIIa. Localization of participating glutaminyl residues to a tryptic fragment of fibronectin. J Biol Chem 1980; 255:1181-8.

26. Li X, Shi X, Fan J. Posterior vitreous detachment with plasmin in the isolated human eye. Graefes Arch Clin Exp Ophthalmol 2002; 240:56-62.

Ma, Mol Vis 2005; 11:744-748 <>
©2005 Molecular Vision <>
ISSN 1090-0535