Functional Properties of Hyalocytes under PDGF-Rich Conditions

Abstract

purpose. To investigate the functional properties and intracellular signaling of hyalocytes under platelet-derived growth factor (PDGF)–rich conditions.

methods. The hyalocytes were isolated from bovine eyes and identified by immunocytochemistry and electron microscope. The expression of PDGF receptor α/β and its phosphorylation in response to PDGF-BB was analyzed by Western blot analysis. PDGF-BB–induced proliferation and migration were evaluated by thymidine uptake and Boyden’s chemotaxis assay. The expression of the urokinase-type plasminogen activator (uPA) gene and the fibrinolytic activity were assessed by Northern blotting and fibrin zymography. An in vitro type I collagen gel contraction assay was performed to determine the effect of PDGF-BB on cellular contraction.

results. The hyalocytes were immunocytochemically positive for S-100 and negative for glial fibrillary acidic protein (GFAP) and cytokeratin, as previously described. The electron microscope demonstrated that hyalocytes possess lysosome-like granules, mitochondria, and micropinocytotic vesicles in their cytoplasm. The hyalocytes expressed PDGF receptor α and β, both of which were immediately phosphorylated in response to PDGF-BB. PDGF-BB also activated p85 PI3-kinase, p44/p42 mitogen-activated protein (MAP) kinase and p38 MAP kinase. PDGF-BB induced thymidine uptake and migration in a concentration-dependent (0–10 ng/mL) manner. Inhibitors of the respective kinases prohibited PDGF-BB–dependent thymidine uptake and migration with the exception of the p44/p42 MAP kinase inhibitor, which displayed no inhibitory effects on migration. PDGF-BB increased uPA gene expression and fibrinolytic activity. Collagen gel contraction observed under PDGF-BB–rich conditions was not prohibited by the respective inhibitors investigated.

conclusions. The hyalocytes demonstrated macrophage-like characteristics and may have both physiologic and pathologic roles, such as the maintenance of vitreous transparency through fibrinolytic activity and the pathogenesis of proliferative-vitreoretinal diseases through cellular proliferation and vitreous hyper-contraction.

The human vitreous body makes up the greatest volume of the eye. The body is a clear, jellylike substance composed mainly of water and delicate collagenous skeleton associated with hyaluronic acid, possibly derived from the hyalocytes. The vitreous not only offers support to the structures within the eye but helps maintain the transparency of the media. One of the major causes of the disruption of the transparency is vitreous hemorrhage, which usually originates from the retinal vessels and occurs in many diseases such as diabetic retinopathy and hypertensive retinopathy, or as an initial sign of retinal tear formation. Whereas a small hemorrhage does not usually affect visual acuity, a large hemorrhage creates a massive fibrin clot, resulting in a sudden and severe loss of vision in the affected eye. The fibrin clot in the vitreous body is usually resolved gradually over time; however, the precise mechanisms remain to be elucidated. 

Hannover ¹ was the first to describe the cells in the vitreous cavity, which were later named hyalocytes. ² It is reported that the hyalocytes mainly subsist in the vitreous cortex, vitreous base, and posterior chamber. ^{3 4} Histologically, the hyalocytes have characteristics similar to those of macrophages. They possess lysosomes, mitochondria, ribosomes, and micropinocytotic vesicles, ³ and express leukocyte-associated antigens CD45, CD11a, CD64, and S-100 and major histocompatibility complex class I and II, but do not express CD68, CD11b, CD14, glial fibrillary acidic protein (GFAP), and cytokeratin. ⁵ Although most of the facilities and roles of hyalocytes remain unclear, the hyalocytes are thought to be one of the macrophage lineage and are likely to be broadly associated with physiological and pathologic conditions, including the maintenance of the vitreous as avascular and transparent tissue. ^{6 7 8 9}  

Platelet-derived growth factor (PDGF) is the variously functionalized cytokine primarily isolated from the platelets and consists of combinations of two peptides: PDGF-A chain and -B chain (AA, AB, and BB dimmers). Needless to say, PDGF increases in the injured microenvironment, as in vitreous hemorrhage, for example. ^{10 11} Locally released PDGF affect the functional property of many types of cells including monocytes-macrophages. PDGF-BB is the predominant PDGF isoform under investigation with respect to its potential for promoting the healing of wounds ¹² and is already known to be increased in the eyes with diseases such as proliferative vitreoretinal diseases including diabetic retinopathy. ^{10 13 14 15} In contrast, the PDGF specific cell surface receptor consists of dimmers from two subunits α (PDGFRα) and β (PDGFRβ). Because the PDGF-B chain is capable of binding to both PDGFRα and β, the signaling pathway from PDGF-BB includes all signaling induced by PDGF. On the basis of these findings, it is theoretically possible that locally released PDGF-BB may affect the functional properties of hyalocytes thought to possess macrophage-like characteristics. 

In the present study, we addressed the contribution of PDGF-BB to the functional alteration of hyalocytes and elucidated a part of their physiological and pathologic roles. 

Materials and Methods

Antibodies against S-100, GFAP and cytokeratin were purchased from Sigma-Aldrich (Tokyo, Japan), Dako Cytomation (Glostrup, Denmark) and Santa Cruz Biotechnology (Santa Cruz, CA), respectively. Recombinant human PDGF-BB was also obtained from Sigma-Aldrich. Protein kinase inhibitors against mitogen activated kinase (MAPK) 1 (PD98059), phosphatidylinositol-3-kinase (PI3K; wortmannin), and p38 MAPK (SB203580) were purchased from CalBiochem (La Jolla, CA). Anti-phospho-specific antibodies against ERK1 (p44)/ERK2 (p42) and p38 were obtained from New England Biolabs (Beverly, MA). Anti-phospho-specific antibody against p85 and non-phosphorylation–specific antibodies against ERK1, p38, and p85 were purchased from Santa Cruz Biotechnology, and horseradish peroxidase-conjugated secondary antibodies were obtained from Bio-Rad (Hercules, CA). 

Cell Culture

Bovine eyes were obtained from a local abattoir. The posterior part of the vitreous body was extracted and washed once in Dulbecco’s modified Eagle’s medium (DMEM; Nakalai Tesque, Kyoto, Japan). The vitreous was chopped into several pieces and incubated in type I collagen-coated dishes poured DMEM with 20% fetal bovine serum (FBS; Invitrogen-Gibco, San Diego, CA) for 1 week. The cells that had proliferated on the dishes were trypsinized and then subcultured in type I collagen–coated dishes containing DMEM with 20% FBS. This condition was used for the growth condition of the hyalocytes. Cultured hyalocytes obtained between passages 4 and 8, which had shown no obvious morphologic change were used for the following experiments. Retinal pigment epithelial (RPE) cells and retinal glial cells were also isolated from bovine eyes, as previously described. ^{16 17 18}  

Immunocytochemistry

Hyalocytes were seeded and cultured in fibronectin-coated chamber slides (Nalge Nunc, Rochester, NY) and used for immunocytochemistry. The cells were washed twice in phosphate buffered saline (PBS), dried out completely, and fixed in −20°C acetone for 10 minutes. The cells were then redried and incubated for 20 minutes in PBS containing 1% skim milk (Nakalai Tesque). S-100, GFAP, and keratin were detected with respective antibodies (anti-S-100 antibody, Sigma-Aldrich; anti-GFAP antibody, Dako Cytomation; and anti-pan-cytokeratin antibody, Santa Cruz Biotechnology) using the avidin-biotin-peroxidase complex method, as previously described. ¹⁹  

Electron Microscopy

The 1-week-cultured vitreous in which the cells proliferated homogeneously was fixed in 1% glutaraldehyde and 1% paraformaldehyde in phosphate buffer. The vitreous was then postfixed in veronal acetate buffer osmium tetroxide (2%), dehydrated in ethanol and water, and embedded in Epon. Ultrathin sections were cut from the blocks and mounted on copper grids. The specimens were observed with a electron microscope (JEM 100CX; JEOL, Tokyo, Japan), as previously described. ²⁰  

PDGF Receptor Phosphorylation

Subconfluent hyalocytes were starved in DMEM containing 1% FBS for 8 hours. Cells were then stimulated with 10 ng/mL PDGF-BB for the indicated time. The hyalocytes were washed once in cold 1× PBS and lysed in 1× Laemmli buffer (50 mM Tris [pH 6.8], 2% SDS, and 10% glycerol) containing protease inhibitors (1 mmol/L phenylmethfluorideylsulfonyl [PMSF], 2 μg/mL aprotinin, 10 μg/mL leupeptin, 1 mM NaF, and 0.5 mM Na₃VO₄). ²¹ Cell lysates were incubated with anti-PDGF receptor α or β polyclonal agarose-conjugated antibody in a 1× immunoprecipitation buffer (1× 1% Triton X-100, 150 mM NaCl, 10 mM Tris [pH 7.4], 1 mM EDTA, 1 mM EGTA, 0.2 mM Na₃VO₄, 0.2 mM PMSF, 0.2 mM NaF, and 0.5% NP-40) for 24 hours. Immunoprecipitates were washed with a 1× immunoprecipitation buffer and centrifuged. Proteins were extracted with a 2× electrophoresis sample buffer (1× 125 mM Tris-HCl [pH 6.8], 2% SDS, 5% glycerol, 0.003% bromophenol blue, and 1% β-mercaptoethanol). The extraction was subjected to 6% SDS-PAGE and transferred to nitrocellulose filters (New England Biolabs). The blots were incubated with anti-phosphotyrosine antibody (4G10) followed by incubation with HRP-conjugated secondary antibody. Visualization was performed using enhanced chemiluminescence (ECL; Amersham, Arlington Heights, IL) detection system according to the manufacturer’s instructions and the density of the signals was measured and analyzed using analysis software (NIH image, version 1.55; available by ftp at zippy.nimh.nih.gov/ or at http://rsb.info.nih.gov/nih-image; developed by Wayne Rasband, National Institutes of Health, Bethesda, MD). 

p44/p42, p38 MAPK, and p85 PI3K Phosphorylation

PDGF-BB-induced MAPK and PI3K phosphorylation was evaluated by Western blot analysis. After 8 hours of starvation in DMEM containing 1% FBS, the cells were stimulated with PDGF-BB at the indicated dose and time. Cells were lysed as described earlier. Phospho-MAPK/PI3K/p38 and total MAPK/PI3K/p38 were detected with the respective antibodies. Visualization was performed as described earlier. 

[³H]-Thymidine Uptake

The hyalocytes were seeded into 24-well plates at a density of 1 × 10⁴ cells/well. The media were replaced by DMEM with 3% fatal fetal bovine serum (FBS) the next day. After 24 hours, the cells were stimulated with PDGF-BB for 18 hours. [³H]-thymidine was then added (0.25 μCi/well) for an additional 5 hours, after which the cells were washed, fixed, and lysed. Incorporated [³H]-thymidine was determined by scintillation counting, as previously described. ²¹  

Chemotaxis Assay

The assay was performed using a 24-well Boyden chemotaxis chamber (8 μm-sized pores; Chemotaxicell, Kurabo, Japan), with a slight modification to the previously described method. ²² The cells grown in DMEM containing 20% FBS until subconfluence were harvested briefly, and the suspension (1.5 × 10⁵ cells/300 μL DMEM containing 1% FBS) was placed in the upper chamber. The membrane separating each pair of wells was precoated with fibronectin and the chemoattractants (PDGF-BB at 0, 1, 10 ng/mL in DMEM containing 1% FBS) placed into the wells of the lower chamber. After incubation for 4 hours at 30°C, the upper surface of the filter was scraped with a cotton-tipped stick to remove nonmigrated cells, and the membrane with the migrated cells were fixed and stained (Giemsa staining Diff-Quick stain kit; Dade Behring AG, Dudingen, Switzerland) and mounted onto glass slides. For quantitative analysis, the cells were observed using a 100× objective on a light microscope. Three random objective fields of stained cells were counted for each well, and the mean number of migrating cells per one field was calculated. The number of migrated cells, obtained from six independent wells for each group, was compared using ANOVA and the Dunnett test. 

Northern Blot Analysis

Gene expression of uPA was evaluated by Northern blot analysis, as previously described. ²³ Total RNA was purified from cultured hyalocytes using the acid guanidine thiocyanate-phenol-chloroform extraction (AGPC) method. Radioactive probes were generated using labeling kits (Multiprime; Amersham) and [³²P]dCTP (NEN, Boston, MA). Quantitation of Northern blot analysis was performed using a computing phosphorescence imager (PhosphorImager with ImageQuant software analysis; Molecular Dynamics, Sunnyvale, CA). Lane loading differences were normalized by rehybridization with radiolabeled 36B4 cDNA probe as an internal control gene. ²³  

Fibrin Zymography

The fibrinolytic activity was assayed by direct zymography. ²⁴ Cells were lysed in Laemmli buffer as described earlier. Each sample and recombinant urokinase-type plasminogen activator (u-PA; Chemicon, CA) as a positive control were applied for gel electrophoresis. The gel was washed for 2 hours with 2.5% Triton X to remove SDS and placed on the fibrin plate containing plasminogen-rich fibrinogen (15 mg/mL; Sigma-Aldrich), thrombin (0.125 U/mL; Mitsubishi Pharma, Tokyo, Japan) with or without leupeptin (30 μM; Cal Biochem) which is a selective inhibitor of plasmin activity, ²⁵ and incubated for 4 hours at 37°C to allow fibrinolysis to occur. The fibrin plate was then stained with Coomassie blue and destained with 40% methanol and 10% acetic acid. 

Collagen Gel Contraction

The contraction assay was performed as previously described, with some modifications. ²⁶ Bovine hyalocytes were collected by treatment of cultures with trypsin-EDTA for 3 minutes, washed with unsupplemented MEM, and resuspended in MEM at a density of 2.2 × 10⁶ cells mL. Type I collagen (Koken Co., Ltd., Tokyo, Japan), reconstitution buffer, hyalocytes suspension, and water were mixed on ice at a ratio of 7:1:1:1:1 (final concentration of type I collagen gel, 1.9 mg/mL; final cell density, 2 × 10⁵ cells/mL. The resultant mixture (0.5 mL) was added to a 24-multiwell plate (Nunc, Roskilde, Denmark), and the formation of collagen gels induced by incubation at 37° centigrade under 5% CO₂ for 60 minutes). After gelatinization, 0.5 mL of DMEM was added to each well. The gels were then freed from the walls of the culture wells with a microspatula. The diameter of the collagen gel was measured with a ruler 48 hours after stimulation. For quantitative purposes, contraction data are presented as the change in diameter. 

Results

Isolation and Identification of the Hyalocytes

When the vitreous was extracted, the hyalocytes were present mainly on the vitreous gel surface and accumulated like branching blood vessels (Fig. 1A) , as previously described. ²⁷ Some of the cells that proliferated in the vitreous gel attached and formed colonies on the dishes within 1 week (Fig. 1B) . Passaged cells were immunocytochemically identified as hyalocytes expressing S-100 protein, but neither expressed GFAP or cytokeratin (Fig. 1C) , in agreement with previous reports. ¹⁹ Under the electron microscope, hyalocytes displayed an infinite form and possessed lysosome-like granules, mitochondria, and micropinocytotic vesicles, presenting the characteristics of macrophage lineage (Fig. 1D) . ^{3 5 28 29}  

PDGF Receptor Expression and Phosphorylation

Although hyalocytes are thought to be of the macrophage lineage, it is not known whether they express PDGF receptors. To address PDGF receptor expression by cultured hyalocytes and its activation in response to PDGF-BB, quiescent hyalocytes were treated with recombinant PDGF-BB at the concentration of 10 ng/mL. Total cell lysates were immunoprecipitated with antibodies against PDGF receptor α and β, respectively. As shown in Fig. 2 , autophosphorylation of PDGF receptor α was detected within 2 minutes and the maximum level achieved 15 minutes after PDGF-stimulation. In contrast, autophosphorylation of PDGF receptor β reached its maximum level within 2 minutes and declined thereafter. To confirm the protein levels of PDGF receptors, the membrane was reprobed with antibodies against PDGF receptors. Prominent expression of both isoform was detected. These protein levels, however, gradually decreased after treatment with PDGF, indicating the existence of a degradation mechanism after phosphorylation. 

PDGF-BB Activation of p44/p42 MAPK, p38 MAPK, and p85 PI3K

To examine whether p44/p42 MAPK, p38 MAPK, and PI3K pathways lie downstream of PDGF receptor activation in hyalocytes, we investigated PDGF-induced p44/p42 MAPK, p38 MAPK, and p85 PI3K phosphorylation. Quiescent hyalocytes showed a slight phosphorylation of the proteins; however, PDGF-BB stimulated phosphorylation within 2 minutes. Maximum activation was achieved after 5 minutes and maintained up to at least 60 minutes in p44/p42 MAPK and p38 MAPK. In contrast, p85 PI3K was gradually phosphorylated in a time-dependent manner and maintained up to at least 60 minutes (Fig. 3) . 

PDGF-BB–Induced Thymidine Uptake

To examine whether PDGF-BB induces the proliferation of hyalocytes, we investigated thymidine uptake in hyalocytes. PDGF-BB–induced thymidine uptake in the hyalocytes occurred in a concentration-dependent (0–10 ng/mL) manner. Induction ratios compared with the control were 131% at the concentration of 0.1 ng/mL of PDGF, 208% at 1.0 ng/mL, and 247% at 10 ng/mL (Fig. 4A) . Inhibitors against p44/p42 MAPK, p38 MAPK, and PI3K significantly restrained the induction as shown in Figure 4B . In particular, PD98059 restrained the uptake by 72.0% and 73.6% for nonstimulated and stimulated control, respectively. Wortmannin and SB203580 also significantly restrained PDGF-BB–induced thymidine uptake, indicating all three signaling pathways investigated were involved in PDGF-BB–dependent hyalocytes proliferation. 

PDGF-BB–Induced Migration

PDGF-BB promoted the significant migration of hyalocytes, as shown in Figure 5A . The number of migrated cells to the backside of the membrane occurred in a concentration-dependent (0–10 ng/mL) manner (Fig. 5B) . Although Wortmannin and SB203580 significantly inhibited the migration (30.0%, P < 0.01; and 16.0%, P < 0.05, respectively), PD98059 had no apparent effect on PDGF-induced hyalocytes’ migration. 

Fibrin Zymography

Fibrin zymography, using total cell lysate, was performed to examine the fibrinolytic activity of hyalocytes in the PDGF-rich condition. The fibrinolytic molecule was detected at approximately 50 kDa, which appeared to be urokinase-type plasminogen activator (u-PA) dependent. The fibrinolytic activity was prominently intensified by PDGF-BB (Fig. 6A) . The fibrin plate containing leupeptin, a selective inhibitor of plasmin activity, ²⁵ apparently diminished the fibrinolytic activity by both PDGF-treated sample and recombinant uPA, indicating that PDGF-BB–induced fibrinolytic activity is mediated through the activation of plasmin. However, another fibrinolytic band (upper positions) remained, even in the presence of leupeptin. These signals may be independent of plasmin activity, but we could not identified its origin in the present study. 

PDGF-Induced uPA Gene Expression

The effect of PDGF-BB on u-PA gene expression was confirmed by Northern blot analysis. As shown in Figure 6B , PDGF-BB (10 ng/mL) upregulated u-PA gene expression within 1 hour, and the maximum effect was achieved 4 hours after stimulation (Fig. 6B) . 

Collagen Gel Contraction

We hypothesized that the hyalocytes could be involved in proliferative vitreoretinal diseases by means of the contraction of the vitreous and the proliferative membrane. We thus confirmed the functional property of hyalocytes under PDGF-rich condition in collagen gel. The control gel showed no apparent contraction, whereas PDGF-BB significantly lessened (10% compared with control, P < 0.05) the collagen gel diameter (Fig. 7) . In the absence of PDGF-BB, the collagen gel diameter was not influenced by the inhibitors investigated (PD98059, wortmannin, and SB203580). In contrast, PD98059 demonstrated further shrinking of collagen gel contracted by PDGF-BB (17.8% contraction compared with control in the presence of PDGF-BB, P < 0.01). 

Discussion

In the present study, we confirmed the histologic characteristics of hyalocytes showing macrophage-like property in agreement with previous reports. ^{3 5 7 30} Although the hyalocytes are supposed to be one of the macrophage lineages, their functional properties largely remain to be elucidated. We thus examined the functional properties of hyalocytes under a PDGF-rich condition similar to that which occurs in various intraocular diseases, including vitreous hemorrhage and proliferative vitreoretinal diseases. 

Accumulating studies using both light and electron microscopes have indicated that the hyalocytes are morphologically very similar to macrophages, and the most popular concept about their origin proposes that the hyalocytes originate from the blood monocytes. ³¹ Accordingly, the hyalocytes are assumed to be resident macrophages of the vitreous cavity under physiological conditions. However, it has been reported that the hyalocytes may be composed of at least two different morphologic populations. ³⁰ One population of hyalocytes are distributed on the epithelial surface of the ciliary body where the cells are stellate with some short processes, and the other is distributed randomly on the vitreous surface in the posterior part of the eyeball, where they are elongated in shape with a spherical perikaryon and a few stout processes. In the present study, we used the cells in the posterior part of the vitreous, because we supposed these cells at least to be involved in the pathogenesis of various vitreoretinal diseases. Although the functional differences of these two types are not completely understood, the former seems to be concerned with the regression of the tunica vasculosa lentis on the embryonic lens. ³² In addition, the hyalocytes distributed on the epithelial surface of the ciliary body are associated with the loss of capillary integrity, leakage of erythrocytes into the vitreal compartment and phagocytosis of the apoptotic endothelium. ^{8 9 33} It also has been reported that the hyalocytes are directly responsible for the endothelial cell death in this system. ⁹ Furthermore, Zhu et al. ⁶ reported that hyalocytes secrete some antiangiogenic factors. Considering this evidence, the hyalocytes are thought to be active in maintaining the vitreous cavity as an avascular and transparent tissue. In the present study, we demonstrated the fibrinolytic activity of hyalocytes. In the vitreous cavity, the humoral element circulates poorly because of the gelatinous structure of the vitreous. Therefore, a vitreous hemorrhage tends to remain in the cavity for a long time. The fibrinous material not only affects the visual acuity by its opacification, but the fibrinous membranes on the retinal surface are likely to function as a scaffolding for infiltrating cells, resulting in fibrous membrane formation and retinal traction. It is thus reasonable that hyalocytes digest the fibrinous material in the vitreous cavity as a physiological function, indicating one of the physiological properties of the cells as the maintainer of vitreous transparency. 

In contrast, it is thought that the hyalocytes themselves may be involved in vitreoretinal interface diseases such as epiretinal membrane formation, macular hole, and diabetic macular edema. ^{34 35} Faulborn et al. ³⁶ reported that human hyalocytes in diabetic eyes have a different shape compared with those in normal eyes, and their number seemed to be increased. Certainly it has been reported that PDGF increases in such hemorrhagic and/or proliferative pathosis. ^{10 11 13 14 15} In addition to PDGF, many factors, including transforming growth factor-(TGF)-β, have been thought to contribute to the pathogenesis of proliferative vitreoretinal diseases. ^{37 38 39} It is possible that the interaction between these factors may result in a vicious cycle. In the present study, PDGF enhanced the plasmin activity of hyalocytes. Plasmin is known to promote the release of latency-associated peptides from latent TGF-β1, leading to the activation of this cytokine. ^{40 41} TGF-β is known to be a multifunctional cytokine that modulates the proliferation and differentiation of many cell types and is considered to bring about the increase of extracellular matrix production, resulting in the formation and contraction of proliferative membranes. ^{26 39 42 43} In the presence of a high concentration of TGF-β in diseases such as proliferative diabetic retinopathy and proliferative vitreoretinopathy, TGF-β is known to encourage the proliferation of retinal endothelial cells and fibroblasts and to act on the contraction of proliferative membranes as a consequence. ^{37 38 39} Therefore, in proliferative diseases, plasmin activated by the hyalocytes’ plasminogen activator would activate TGF-β, which might exacerbate the pathosis by the formation of a fibrous membrane and its cicatricial contraction. 

In this study, we demonstrated the promoting effect of PDGF-BB on the proliferation of cultured hyalocytes. Moreover, PDGF-BB induced the contraction of collagen gel containing the hyalocytes, indicating the involvement of hyalocytes in the pathogenesis of various vitreoretinal diseases. Of note, whereas kinase inhibitors such as PD98059 (MEK1 inhibitor), wortmannin (PI3-kinase inhibitor), and SB203580 (p38 MAPK inhibitor) prominently inhibited PDGF-dependent proliferation of hyalocytes respectively, only PD98059 rather promoted further contraction of the collagen gel induced by PDGF-BB. From these results, the signaling pathway leading to the proliferation and contraction thus seems to come from different directions, and the p44/p42 MAPK pathway, not the PI3-kinase or p38 MAP-kinase pathways, may be a diverging point. The exact mechanisms are still unclear from the present study. More precise examination regarding the intracellular signaling pathway is needed, to elucidate the control of the vitreoretinal diseases by various agents. In addition, we used type I collagen in both two- and three-dimensional experiments in this study. However, the main component of collagen in the vitreous is type II collagen. It is thus impossible to interpret directly the results obtained in this study, to explain pathogenesis in vivo. Further research is required for this area to be fully understood. 

In the present study, we confirmed the histologic characteristics of the hyalocytes and demonstrated both the physiological and pathologic roles of the hyalocytes in part. The hyalocytes appear to be one of the maintainers of vitreous transparency. However, if the pathologic conditions continue over a long period or exceed a certain limit, the hyalocytes themselves may exacerbate the disease conditions. 

 Supported in part by Grant-in-Aid for Scientific Research 15591861 from the Ministry of Education, Science, Sports and Culture, Japan.

 Submitted for publication October 2, 2003; revised February 3, 2004; accepted March 9, 2004.

 Disclosure: Y. Noda, None; Y. Hata, None; T. Hisatomi, None; Y. Nakamura, None; K. Hirayama, None; M. Miura, None; S. Nakao, None; K. Fujisawa, None; T. Sakamoto, None; T. Ishibashi, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Histologic characteristics of hyalocytes. (A) In the extracted vitreous, the hyalocytes were present mainly on the vitreous gel surface and accumulated like branching blood vessels. (B) Some of the cells that proliferated in the vitreous gel attached to the dishes and formed colonies within 1 week. (C) Immunocytochemistry showed that hyalocytes had a different stainability from RPE and glial cells. Hyalocytes expressed S-100 protein, but expressed neither GFAP nor cytokeratin. In contrast, RPE cells expressed S-100 protein and cytokeratin, and retinal glial cells expressed GFAP. (D) Under an electron microscope, hyalocytes assumed many forms and possessed lysosome-like granules, mitochondria, and micropinocytotic vesicles, presenting the characteristics of a macrophage lineage.

Figure 2.

 

PDGF-BB–induced phosphorylation of PDGF receptor α and β. Quiescent hyalocytes were treated with recombinant PDGF-BB at a concentration of 10 ng/mL. Total cell lysates were immunoprecipitated with antibodies against PDGF receptor α and β. Autophosphorylation of PDGF receptor α was detected within 2 minutes, and the maximum level achieved 15 minutes after stimulation. In contrast, autophosphorylation of PDGF receptor β reached its maximum level within 2 minutes and declined thereafter. To confirm the protein levels of PDGF receptors, the membrane was reprobed with antibodies against PDGF receptors. Prominent expression of both isoforms was detected. These protein levels, however, gradually decreased after treatment with PDGF. Reproducible results were obtained from three independent experiments; representative data are presented.

Figure 3.

 

PDGF-BB activated p85 PI3K, p44/p42 MAPK, and p38 MAPK. (A) Quiescent hyalocytes showed no significant phosphorylation of p85 PI3K, whereas PDGF-BB stimulated phosphorylation of p85 PI3K within 2 minutes, which gradually intensified in a time-dependent manner up to at least 60 minutes. (B) PDGF-BB stimulated phosphorylation of p44/p42 MAPK within 2 minutes, and maximum activation was achieved after 5 minutes and maintained up to at least 60 minutes. (C) PDGF-BB stimulated phosphorylation of p38 MAPK within 2 minutes. Maximum activation was achieved after 5 minutes and declined thereafter. Reproducible results were obtained from three independent experiments; representative data are presented.

Figure 4.

 

PDGF-BB induced thymidine uptake and the inhibitory effect of the agents. (A) PDGF-BB induced thymidine uptake in hyalocytes in a concentration-dependent (0–10 ng/mL) manner (n = 4, **P < 0.01). The induction rate was 247% at a concentration of 10 ng/mL compared with the control. (B) MAPK and PI3K inhibitors restrained the uptake. PD98059 (PD) in particular, restrained the uptake by 72.0% and 73.6% for nonstimulated and stimulated control, respectively. Wortmannin (WM) and SB203580 (SB) also restrained uptake by 42.6% and 43.6% compared with with the nonstimulated control, and 57.9% and 67.3% compared with with stimulated control (n = 4, **P < 0.01). Representative data from three independent experiments are presented.

Figure 5.

 

PDGF-BB induced migration of hyalocytes. (A) Cells migrating to the backside of the membrane (arrows) were visualized by Giemsa staining. PDGF-BB significantly activated the migration of hyalocytes (pore [arrowheads] size on the membrane: 8 μm). (B) The number of migrating cells was counted. Wortmannin (WM) and SB203580 (SB) significantly inhibited the migration, whereas PD98059 (PD) had no apparent inhibitory effect (n = 6, *P < 0.05, **P < 0.01). Reproducible results were obtained from three independent experiments; representative data are presented.

Figure 6.

 

PDGF-BB stimulated uPA-plasmin activity and the expression of the u-PA gene. (A) Fibrin zymography was performed with total cell lysate. Total cell lysate from quiescent hyalocytes slightly lysed the fibrin plate (lane 1). On the contrary, lysate from cells stimulated with PDGF-BB (10 ng/mL, 24 hours) significantly lysed the fibrin plate and the fibrinolytic molecule detected was approximately 50 kDa (lane 2), in accordance with the u-PA activator (uPA, lane 5). The fibrin plate containing leupeptin (30 μM, lanes 3, 4, 6), which is a selective inhibitor of plasmin activity, apparently diminished fibrinolytic activity. UI: unidentified. (B) The effect of PDGF-BB on u-PA gene expression was confirmed by Northern blot analysis. The membrane was rehybridized with a 36B4 probe. Reproducible results were obtained from three independent experiments; representative data are presented.

Figure 7.

 

Collagen gel contraction by hyalocytes in the presence of PDGF-BB. An in vitro collagen gel contraction assay was performed. Hyalocytes were embedded in type I collagen gel (n = 4). Forty-eight hours after stimulation in the presence or absence of inhibitors, the gel was photographed (top), and the diameter of the collagen gel was measured (bottom). A control gel showed no apparent contraction, whereas PDGF-BB significantly shortened the collagen gel diameter. Wortmannin and SB203580 did not influence PDGF-BB–induced contraction, whereas PD98059 induced further contraction compared with the control gel in the presence of PDGF-BB (*P < 0.05, **P < 0.01). Reproducible results were obtained from three independent experiments; representative data are presented.

The authors thank Tadahisa Kagimoto and Eve E. Sockett for excellent help.