Spectral transmission of the optical media of the human eye with respect to keratitis and cataract formation

The spectral transmissions of cornea, aqueous humour, lens and vitreous humour of human eyes were measured in the range from 250 nm to 800 nm in 1 nm steps using a high resolution double monochromator. Cornea and lens are good cut off filters; 1% transmission was measured in various cornea samples between 291 and 298 nm, and in different lenses between 389 and 414 nm. Aqueous humour and vitreous humour showed high transparency (80%) at lambda > 320 nm. The wavelength range of keratitis effective irradiance was found to be completely within the wavelength range of absorption of the cornea, that of cataract effective irradiance was found to be at the short wavelength end partly outside the wavelength range of absorption of the lens. It may therefore be supposed that the action spectrum of cataract obtained by animal experiments is applicable to the human eye only with certain reservations.

Effect of vitreous fluidity on the measurement of blood-retinal barrier permeability using contrast-enhanced MRI

Breakdown of the blood-retinal barrier (BRB), frequently an early clinical sign in retinopathy, can be accurately determined using contrast-enhanced MRI. However, increased vitreous fluidity with age and disease may affect the accuracy of the MRI method. We compared the permeability surface area product per area of leaky retina in eyes with normal vitreous (5.42 +/- 0.48 x 10(-4) cm/min, mean +/- SEM, n = 5) to the contralateral gas-compressed vitrectomized eyes (5.41 +/- 0.54 x 10(-4) cm/min, n = 5). The effect of vitrectomy was not significant (P = 0.325) using a Wilcoxon matched pairs signed rank test on the signed differences of the PS' values.

Role of the vitreous in the prognosis of peripheral uveitis

We studied 116 eyes with peripheral uveitis (mean follow-up, 5.0 years) to ascertain the effect of the vitreous on prognosis. At the initial examination, the prevalence of partial posterior vitreous detachment was significantly higher in eyes with macular edema (27%) than in those without (7%, P = .01). During follow-up, the prevalence of no posterior vitreous detachment initially was significantly higher in eyes that developed macular edema (78%) than in those that did not (22%, P = .01). The prevalence of no or partial posterior vitreous detachment was also significantly higher in eyes with decreased visual acuity of three or more Snellen lines (57% and 38%, respectively) compared with those with visual acuity decreases of two Snellen lines or less (21% and 11%; P = .02 and P = .01, respectively). Eyes with peripheral uveitis with vitreous adhesion may have a less favorable prognosis compared with those with complete posterior vitreous detachment.

Spectral sensitivity of the blood-retinal barrier at the pigment epithelium for blue light in the 400-500 nm range

To specify the spectral sensitivity of the retinal pigment epithelium (RPE) for blue light damage, pigmented rabbits were exposed to light of 408, 418, 439, 455, 485, and 500 nm (half-peak bandwidth approximately 12 nm). The range of radiant exposure was 15-275 J cm-2 (1.7-19 mW cm-2 for 0.5-5 h). Vitreous fluorophotometry was used to functionally evaluate the blood-retinal barrier at the RPE in vivo, and electron microscopy to visualize RPE ultrastructure in vitro. A significant increase in permeability of the blood-retinal barrier was seen only after exposure to light of 418 nm. Radiant exposure at threshold for permeability increase was 18 J cm-2. Electron microscopy of the RPE demonstrated dispersion and clumping of melanin granules. The results suggest that the RPE is most sensitive to light in the range 412-425 nm, possibly due to damage-mediating chromophores such as cytochrome c oxidase and lipofuscin.

Extracellular dopamine concentration in the retina of the clawed frog, Xenopus laevis

Dopamine reaches targets in the outer retina of the clawed frog (Xenopus laevis) by diffusion from a network of dopaminergic cells and processes located predominantly at the junction of inner nuclear and inner plexiform layers. We obtained values for the steady-state release, uptake, and extracellular concentration of dopamine in the retina by a combination of HPLC (with electrochemical detection), scintillation spectroscopy, and fast-scan cyclic voltammetry. Vitreal concentrations of dopamine varied from 564 +/- 109 nM in light-adapted eyes near the time of subjective dawn to 156 +/- 12 nM in dark-adapted eyes. The data are consistent with a simple model for steady-state dopamine diffusion from an appropriately sited thin-sheet source. This model was used to generate a profile of extracellular dopamine concentration as a function of retinal depth. The model predicted an increase in the dopamine concentration from the vitreous to the layer of dopaminergic cells, remaining constant from that layer to the distal tips of the photoreceptors. This prediction was borne out by comparing fast-scan voltammetric measures of dopamine at the distal tips of the receptors with the vitreal concentrations determined by HPLC using electrochemical detection.

Postvitrectomy hypotonia: the role of the vitreous and retinochoroidal lesions

The pathomechanisms of hypotonia after vitrectomy remain obscure. To examine the possible escape of intraocular fluid through the ocular wall, hydraulic fluid conductivity was measured across preparations of retina, pigmented epithelium, and choroid isolated from rabbit eyes 1, 3, 5, or 7 weeks after the production of laser or cryopexy lesions. The hydraulic conductivity measured in a modified Fischbarg-Bourguet chamber, was 0 in controls. At 1 and 3 weeks after transpupillary diode laser coagulation of the equatorial retina, hydraulic conductivity was measurable. However, the conductivity of these preparations returned to 0 by 5 weeks. After transscleral cryopexy, conductivity remained elevated for 7 weeks. Break-up of residual cortical vitreous with hyaluronidase increased the conductivity of "mature" cryopexy lesions to the values obtained 1 week after injury. These results suggest that there is escape of fluid across retinochoroidal lesions; fluid conductance typically decreases over time; an intact cortical gel can hinder conductance and seal the retinochoroidal leak. Thus, the risk factors for hypotonia appear to include extensive retinochoroidal lesions and the absence of a normal cortical vitreous in aphakia and/or after aggressive vitrectomy.

Vitreous body-derived mitogenic activity for retinal pigment epithelial cells. A further characterization

To obtain a better insight into the possible role of vitreous body in proliferative vitreoretinopathy (PVR), we examined the influence of bovine vitreous body fractions separated by fast protein liquid chromatography (FPLC) on the growth of porcine retinal pigment epithelial (RPE) cells in vitro. Tetrazolium colorimetric assay (MTT assay) and cell counting were used for quantification. After fractionating proteins of the physiological vitreous body by FPLC gel chromatography (Superose 12/Superdex 75), we determined the mitogenic effect of the resulting fractions by either cell counting or MTT assay. With both columns we were capable of separating two fractions (20-40 kDa and 1-2 kDa) that induced RPE cell proliferation. A comparison of the mitogenesis and the protein content of the fractions indicated that we were not dealing with a nonspecific protein effect. The low-molecular fraction weighing less than 2 kDa was of particular interest to us. This was further separated by FPLC reversed-phase chromatography.

[Experimental study on local cooling of the eyeball in ocular surgery (4). The local cooling effect on uveal blood circulation, intraocular pressure, and intravitreous pressure]

The effects of local cooling on the uveal blood circulation, intraocular pressure, and intravitreous pressure were evaluated. Saline solutions maintained at 10 degrees C and 35 degrees C were dripped on the rabbit ocular surface. After 5-minute dripping of 10 degrees C saline solution, the temperature decreased by 10.6 degrees C at the ciliary body, 7.7 degrees C at the choroid and 2.9 degrees C at the central region of the vitreous. Hydrogen clearance with electrolytically generated hydrogen gas was employed to measure the tissue blood flow, which fell to 82.9% of normo-thermal value at the ciliary body and 86.3% at the choroid. The intravitreous pressure dropped by 2.8 mmHg. After 30-minute dripping of 10 degrees C saline solution, the intraocular pressure dropped by 3.0 mmHg. We consider that the reduction of intraocular and intravitreous pressure during local cooling of the ocular surface is related to decrease in the uveal blood flow.

Light transmittance of ocular media in living rabbit eyes

PURPOSE

The purpose of this study was to measure the transmittance of electromagnetic radiation, particularly visible light, through the ocular media of living and whole rabbit eyes. Previous determinations have been carried out on excised cadaver eyes.

METHODS

A specially designed fiberoptic probe (outer diameter, 0.9 mm) was placed in the vitreous in front of the retina using a microsurgical technique. In eight living albino rabbits (under general anesthesia), ocular transmittance was determined in the wavelength range 350 to 1100 nm using a reversed beam path (from vitreous to cornea).

RESULTS

A maximum optical transmittance of 94% to 96% (standard deviation, 2%-3%) was found between 630 and 730 nm (reflection losses in the cornea-air interface excluded). In the blue portion of the spectrum, transmittance decreased rapidly for shorter wavelengths, and was 50% at 400 nm and less than 1% at 380 nm. In the infrared part of the spectrum, transmittance was close to 90% up to 900 nm but declined at longer wavelengths, coinciding with the absorption in pure water. Calibration recordings showed a 1% to 2% accuracy of the method.

CONCLUSIONS

This experimental technique using an intraocular fiberoptic probe yields a high accuracy and indicates that light transmittance is very high in vivo and superior to that reported from cadaver eyes.

Regulation of lens cell growth and polarity by an embryo-specific growth factor and by inhibitors of lens cell proliferation and differentiation

We used a double-label method, which monitors the rate at which cells enter S-phase of the cell cycle, to identify factors that control the growth of chicken embryo lens epithelial cells in vivo. With this assay, we identified a mitogen for lens epithelial cells in the anterior segment of the embryonic eye. When the anterior chamber was opened briefly, by tearing the cornea or displacing the lens, the growth-promoting activity was lost. None of the purified growth factors tested replaced this growth activity, including EGF, bFGF, PDGF, IGF-1, IGF-2, TGF beta and mixtures of these factors. However, chicken embryo serum or plasma did cause chicken embryo lens epithelial cells to progress through the cell cycle. The activity in serum was destroyed by heat and protease treatment. It was most active in serum from 10-day embryos, decreased with subsequent development and was undetectable from 2 days after hatching through adulthood. When embryo serum or plasma was mixed with vitreous humor or IGF-1, agents that induce lens fiber cell formation, cell elongation was prevented. In contrast to the mitogenic activity in serum, this inhibitor of differentiation was insensitive to trypsin treatment. We also identified an activity in vitreous humor that inhibited the growth-promoting agent in embryo serum. Plasma proteins readily enter the anterior chamber of the eye of chicken embryos.(ABSTRACT TRUNCATED AT 250 WORDS)

Evaluation of cytologic specimens obtained during experimental vitreous biopsy

Vitreous specimens can be useful for diagnosis of intraocular infection, inflammation, and neoplasms. Concern has been raised that obtaining vitreous specimens through a guillotine cutter might result in suboptimal cytologic changes. To determine if aspiration yields better cytologic information than vitrectomy, the authors performed experimental vitreous biopsies on rabbit eyes with vitritis to compare specimens taken by aspiration or vitrectomy with cutting rates of 100, 300, 600 per minute. The specimens were processed by cytospin preparations and stained with Papanicolaou and May-Grünwald-Giemsa stain. There was no difference in the adequacy of the specimens. Cell loss or damage to cell morphologic features when obtaining specimens through aspiration or vitrectomy at different cutting rates could not be differentiated by a blinded cytologic evaluation. A theoretical model of shear stress on cells passing through a guillotine cutter was also developed. The experimental and theoretical data show that vitrectomy with a cutting rate as fast as 600 per minute yields an adequate specimens with a sufficient number of well preserved cells to make definite cytologic interpretations, and that vitreous aspiration is not necessary.

Rheology of the vitreous body. Part I: Viscoelasticity of human vitreous

The rheological properties of the vitreous body of the eye are believed to be a function of composition and to differ among species, as well as to vary regionally within the vitreous. These properties are essential to the mechanical functioning of the eye. Although there are gross, qualitative data on vitreous rheology available in the literature, quantitative rheological measurements on human vitreous and on eyes of other species are sparse and incomplete. The aim of the research reported in this series of papers is to study the rheological behavior of human, bovine, and porcine vitreous, to measure the macromolecular and electrolyte content of these samples, and to correlate and compare these values for the different species as a function of location in the vitreous. In this paper, the rheological model used to correlate viscoelasticity of the vitreous is presented, and a detailed description of the rheological instrumentation and methods of analysis used is given. Data on the rheological properties of human vitreous, as a function of location within the eye, are presented. The results show that in the human eye there are significant differences in a number of the viscoelastic parameters as a function of location within the vitreous body.

Does a cytoplasmic bridge connect the CNS axon with the inner loop of myelin?

Sarah Luse (1959) reported over 30 years ago on the presence of a bridge connecting the axon to the myelin sheath in the central nervous system (CNS). This notion has not been accepted in the literature. Wolman (1992) found that the progress of demyelination in some viral diseases affecting the CNS fits the concept of Luse, as the process occurred primarily along the major dense line of myelin, which is in continuity with the cytoplasm of the oligodendroglial cell. Injection of Lucifer yellow (LY) and horseradish peroxidase (HRP) into the vitreous of guinea pigs, with and without iontophoresis, resulted in labeling of the nerve axons and myelin. Labeling of myelin by HRP occurred along the major dense line which indicated that a transient or permanent cytoplasmic bridge connects axons and myelin in the optic nerve.

[Photodynamically induced vitreous liquefaction in vivo]

Photodynamically induced vitreous liquefaction in rabbit eye was investigated. Photosensitizer, riboflavine phosphate, was injected into the vitreous cavity of the rabbit before white-light irradiation. After the irradiation (0, 1, 3, 6 hr) the rabbit vitreous body was separated into gel and liquid portions. The liquid vitreous body was weighed, and the vitreous liquefaction percentage was calculated. One hour irradiation caused 38% of liquefaction of the eye; 3 hr, 50% liquefaction; 6 hr, 59% liquefaction. Although irradiated control eye (without photosensitizer) and the dark adapted control eye (non-irradiated) showed 10-15% liquefaction throughout the experimental period, the liquefaction percentage of the experimental vitreous bodies was significantly larger than that of the control vitreous. Inhibition experiment showed that the radical scavengers (Cu, Zn-superoxide dismutase, catalase, and mannitol) could suppress the photodynamically induced vitreous liquefaction. Results indicated that free radicals, including hydroxyl radical and superoxide anion, which are generated by photosensitizer and visible light irradiation, may contribute to the age-related vitreous liquefaction of humans.

Growth regulation of the mammalian ocular lens by vitreous humor

Experiments were performed in our laboratory to study the effects of a mammalian 8 kD vitreous humor (VH) factor on the DNA synthesis and mitosis of the epithelial cells of organ cultured rabbit lens. The 8 kD polypeptide factor was purified from mature rabbit vitreous humor by liquid chromatography. Proliferative activities of the epithelial cells of organ cultured lenses were stimulated by 3% rabbit serum. The data from our experiments depicted that the 8 kD VH factor effectively inhibits DNA synthesis and mitosis by the epithelial cells of the organ cultured lens. Our experiments also showed that this 8 kD VH factor can maintain its growth inhibitory activity even when heated for 3 min at 95 degrees C. The growth inhibitory effect of the 8 kD VH factor was dose dependent. Using iodinated vitreal proteins it was demonstrated that the VH proteins are able to enter or bind to lens epithelial cells. The growth inhibitory effect of the 8 kD VH factor was also tested on tissue cultured lens epithelial cells. These experiments showed that the 8 kD VH factor has no growth inhibitory effect on the tissue cultured lens epithelial cells. This experiment has been repeated many times using different concentrations of the factor. These observations suggest that the 8 kD VH factor may have receptors in the lens capsular material (extracellular matrix) and the factor-receptor binding is essential for the growth inhibitory effect.

[Conductivity of sodium hyaluronate solution]

In order to elucidate the property of water contained by the stereoic network in the vitreous gel, the conductivity of sodium hyaluronate solutions was measured in various concentrations at 25 degrees C and the relationship between concentrations of sodium hyaluronate and conductivity was evaluated. Each conductivity curve was divided into three stages. Equivalent conductivity was highly dependent upon the concentration of sodium hyaluronate at high and low concentrations, but was rather independent of it in intermediate concentrations. These results suggest that there are three types of water dependent upon the concentration of sodium hyaluronate, indicating physical variety, and that sodium hyaluronate solution in the intermediate concentration may possibly have a buffer homeostatic action on the ionic environment in the vitreous body.

Biological activities of vitreous gel, retrohyaloid fluid and subretinal fluid from diabetic and non-diabetic eyes

This study compared the effects of vitreous gel, retrohyaloid fluid and subretinal fluid from diabetic and non-diabetic eyes on the proliferation and migration of retinal microvascular cells in vitro. Intraocular fluids were obtained from eyes undergoing repair of retinal detachment, due either to proliferative diabetic retinopathy or rhegmatogenous retinal detachment associated with a degree of proliferative vitreoretinopathy. The results demonstrated that the intraocular stimulatory activity for the proliferation of retinal microvascular endothelial cells varied between the different ocular compartments. The mitogenic and migrational activity in vitreous gel was greater than that of either the subretinal or retrohyaloid fluids of the same eye, and the activity of subretinal fluid was intermediate between that of the vitreous gel and the retrohyaloid fluid. There was no significant difference between the activities of the samples from diabetic and non-diabetic eyes.

Cultures of human retinal pigment epithelium. Modulation of extracellular matrix

The extracellular matrix of retinal pigment epithelial (RPE) cells is an important component of the fibroelastic membranes formed in proliferative vitreoretinopathy (PVR). These membranes consist of cells intermingled with fibrillar proteins, such as elastin and collagen. Because RPE cells may be exposed to vitreous and serum-derived factors in conditions that lead to PVR, the effects of vitreous and serum on the deposition of fibrillar proteins by RPE cells were studied using radioactive labeling and indirect immunofluorescence techniques. The RPE cells, modulated by combined vitreous and serum into a fibrocytic-like appearance, deposited less fibroelastic proteins per cell. However, as a result of simultaneous cell proliferation, the total deposition of fibroelastic proteins per culture was increased. Indirect immunofluorescence studies suggest that it is type I collagen that is altered in the matrix of these modulated cell cultures. This information is important in pathologic conditions characterized by intraocular cell proliferation and fibroelastic membrane formation; these phenomena are seen in many vitreoproliferative disorders.

Visualization of water movement in the living rabbit eye

Water enriched with the stable isotope 17O (H2(17)O) shortens the transverse relaxation time (T2) of protons in water and can therefore be used as the contrast agent for proton magnetic resonance (MR) imaging. This agent can be given topically or intravenously to demonstrate water movement in the eye. Topical H2(17)O (0.05-0.1 ml/eye, 10% enrichment) entered the anterior chamber within 5 min and dissipated from the chamber in a single-exponential fashion (flow-rate constant k = 0.1 min-1), principally due to an exchange with the iridic circulation. No H2(17)O was detected in the vitreous. Intravenous administration of H2(17)O (1 ml/kg, 10% enrichment) resulted in rapid entry (less than 20 min) of the agent into the aqueous chamber. Again, no H2(17)O was detected in the vitreous. The lens region, on the other hand, showed an increase in image intensity with time that reached a plateau after 40 min. Although these findings are preliminary, acetazolamide (20 mg/kg injected intravenously) appeared to affect iridic circulation, possibly through vasoconstriction. Potential application of this H2(17)O-enhanced MR imaging technique is discussed.

New ocular micromanipulator for measurements of retinal and vitreous physiologic parameters in the mammalian eye

A new ocular micromanipulator for introducing a microelectrode into the mammalian eye is described. Included is a technical description of the device, incorporating arched tracks and DC motors to provide for angular motion around a scleral entrance point. Applications for in vivo measurements of local PO2 in the retina and vitreous of miniature pigs and cats are shown. The new design features of this micromanipulator allow for smooth microelectrode movements over large retinal areas, facilitating the rapid collection of data in many different areas of the retina and vitreous.

Migration of rabbit corneal endothelial cells on vitreous strands. An organ culture experiment

Anterior vitreous synechia, sometimes seen as a complication after anterior segment surgery, has been mimicked in organ culture experiments. A specimen consisting of a cornea with a vitreous strand adhering to the posterior surface has been kept in culture medium for 6 and 12 days. Migration of corneal endothelial cells onto the surface of the vitreous strand started after 6 days, and the surface of the strand was partly or almost totally covered by corneal endothelium after 12 days. At this stage the cells were elongated and directed along the strand. In clinical situations a vitreous strand adhering to the cornea may exert traction on the retina. This traction may be caused by contraction of actin filaments in endothelial cells covering the strand. This experimental model may be used in attempts to modify the behaviour of the cells by pharmacological manipulation.

[Intraocular light transmission characteristics using a microfiber system]

There have been some reports about light transmission into the vitreous, but few that have been measured directly in vivo. The author attempted to measure visible light and ultraviolet light using a microfiber system in the vitreous. Microfibers 0.6 mm in diameter were made. The microfiber system can be inserted into the vitreous at the pars plana and can measure the characteristics and intensity of the transmitted light. We used rabbit, monkey and aphakic rabbit eyes in this study. The light transmission into the vitreous of normal rabbits and monkey eyes gradually decreased from 400 nm to 360 nm, but the light transmission of the aphakic rabbit eyes did not distinctly decreased. This tendency was found in vivo and in vitro. It was considered that this system is useful to determine the light transmission characteristics of various eye conditions.

Effect of systemic acetazolamide on the fluid movement across the aqueous-vitreous interface

In an attempt to study the effect of systemic acetazolamide on the fluid flow between the aqueous and vitreous in normal eyes, 50 mg kg(-1) acetazolamide was given intravenously every hour for 3 hr and the time change of the aqueous flow rate was calculated in two groups of rabbits, applying one of the following two different methods to each: the fluorescein method II of Jones and Maurice, a classical fluorometric method, or the more recently developed Johnson-Maurice method which entails intravitreal injection of FITC-dextran and measurements of its concentration in the anterior chamber many days after the injection. The flow rate after acetazolamide calculated by the fluorescein method II of Jones and Maurice in one group of rabbits was reduced to 55 +/- 5% (mean +/- S.E. n = 9) of the control on the average. When calculated by the Johnson-Maurice method in another group of rabbits, the reduction was to 80 +/- 4% (n = 11) of the control rate. The difference between the above figures was significant (P less than 0.005). Furthermore, the effect of acetazolamide calculated by the first procedure was significantly greater than that calculated by the second at 1 and 2.33 hr and at later times after the acetazolamide injection (P less than 0.05-0.01). On the other hand, the outflow pressure was reduced by 53-60% in both groups. The difference between the flow rates after acetazolamide determined by the above two methods was best explained by assuming that the FITC-dextran movement from the vitreous into the aqueous was reduced by about 25% after acetazolamide administration.(ABSTRACT TRUNCATED AT 250 WORDS)

Studies on the possible role of vitreous humor on protein synthesis and morphology of organ cultured adult rabbit lens. II. Epithelial cells

The protein synthetic activities of epithelial cells of lenses organ-cultured without adhered vitreous humor manifest significant increase compared to the epithelial cells of lenses incubated with attached vitreous humor. This effect is not due to trauma of vitreous removal, as the addition of freeze-dried vitreous humor to the culture medium of lenses without attached vitreous humor could inhibit protein synthesis. However, the protein synthesis inhibitor in the vitreous humor has no visible effect on the lens morphology. It was also found that the factor from vitreous humor has no effect on mRNA production or cell-free protein synthesis. Thus, it seems that the effect on protein synthesis must be mediated via some other pathway.

Ultrastructural and immunocytochemical changes in retinal pigment epithelium, retinal glia, and fibroblasts in vitreous culture

Retinal pigment epithelial (RPE) cells, retinal glial, and fibroblasts, three cell types believed to play a role in the pathogenesis of epiretinal membrane formation, were maintained in vitreous culture to determine the influence of vitreous on their ultrastructure and expression of cytokeratin, glial fibrillary acidic protein (GFAP), vimentin, and glutamine synthetase (GS). Using a highly sensitive, preembedding technique for the immunolocalization of these antigens at the ultrastructural level, most RPE cells were found to lose cytokeratin and vimentin within 1 day after seeding on irradiated vitreous. The percentage of keratin-positive cells then increased with time in culture. If the vitreous was placed on RPE cells cultured in monolayer instead of placing the cells on the vitreous, keratin and vimentin were expressed these intermediate filament proteins diminished with time. Glutamine synthetase was found in RPE cells grown in monolayer with or without a vitreous overlay, but not in RPE cells grown on the surface of vitreous. Retinal glial grown on vitreous showed a time-dependent decrease in the number of cells expressing GFAP and a corresponding increase in cells expressing vimentin or GS. Some fibroblasts in vitreous culture expressed vimentin but not the other antigens evaluated. A substantial number of cells in each culture did not stain positively for cytokeratin, GFAP, vimentin, or GS. All three cell types showed phenotypic diversity at the ultrastructural level with each cell type being capable of assuming the same morphologic appearance under certain conditions. These results demonstrate the phenotypic plasticity of RPE cells, retinal glia, and fibroblasts when grown in contact with vitreous and provide further evidence that neither ultrastructure, intermediate filament protein expression, nor the presence of GS is sufficient to determine the cell type of origin of cells in epiretinal membranes.