Molecular Vision 2002; 8:226-234 <http://www.molvis.org/molvis/v8/a29/> Received 30 May 2002 | Accepted 9 July 2002 | Published 10 July 2002

Download Reprint

Translocation of macromolecules into whole rat lenses in culture

Daniel L. Boyle, Paul Carman, Larry Takemoto
 
 

Division of Biology, Kansas State University, Manhattan, KS

Correspondence to: Daniel L. Boyle, Division of Biology, Ackert Hall, Kansas State University, Manhattan, KS, 66506; Phone: (785) 532-0134; FAX: (785) 532-66653; email: dboyle@ksu.edu

Abstract

Purpose: Little is known about the endocytosis and transcytosis of macromolecules into lens epithelium and fiber cells. The objective of this study was to determine if proteins (α-crystallins, β-crystallins, and γ-crystallins), carbohydrate (dextran), and plasmid DNA translocate from culture medium into these parts of the lens, with and without prior encapsulation into liposomes.

Methods: α-Crystallins, β-crystallins, γ-crystallins, and dextran were coupled with the fluorochrome Texas red, and plasmid DNA was labeled with propidium iodide. Adult rat lenses were incubated in medium containing one of these components with and without prior encapsulation of the macromolecule in commercially available liposomes (BioPORTER for α-crystallins, β-crystallins, γ-crystallins, and dextran; GenePORTER for plasmid DNA). Translocation of fluorescent macromolecule from the medium into the lens capsule, epithelium and fiber cells was monitored by confocal microscopy.

Results: α-Crystallins, β-crystallins, γ-crystallins, and dextran were present in the capsule, epithelium, and fiber cells after 5 h of incubation. Translocation of fluorescent protein macromolecules into the epithelium was greatly facilitated by encapsulation in BioPORTER liposomes. These macromolecules were localized within the cytoplasm of epithelium and fiber cells. Plasmid DNA was localized to the epithelium, but not the fiber cells. Prior encapsulation of plasmid DNA into GenePORTER liposomes did not increase the intensity of fluorescence localized in epithelium. Without encapsulation, plasmid DNA preferentially localized to the nuclei of epithelial cells, while after encapsulation, plasmid DNA preferentially localized to the cytoplasm.

Conclusions: After incubation with cultured lenses, large macromolecules comprised of proteins and carbohydrates were localized within the cytoplasm of epithelial cells and fiber cells. Prior encapsulation of protein macromolecules into BioPORTER liposomes facilitated the translocation of macromolecules into the cytoplasm of epithelium. Incubation of lenses with plasmid DNA resulted in localization to the epithelium, but not fiber cells. Localization of plasmid DNA was not facilitated by prior encapsulation in GenePORTER. Encapsulated DNA preferentially localized to the cytoplasm of epithelial cells, while without encapsulation, plasmid DNA localizes to the nuclei of epithelial cells. Together, these studies demonstrate that macromolecules of potential biological importance can readily pass through the lens capsule into epithelial cells and in some cases transcytose through the epithelium into fiber cells of the cortex. Furthermore, these studies suggest that prior encapsulation of protein macromolecules may be a possible therapeutic delivery system of physiologically important macromolecules into the epithelium and/or fiber cells of the intact lens.

Introduction

Little is known about the translocation of macromolecules through the lens capsule and into lens epithelium and fiber cells. Past studies have shown that small molecular weight components such as glucose [1], an amino acid [2], and some large molecular weight macromolecules such as some lens crystallins [3] and egg albumin [4] can pass through the isolated lens capsule. Furthermore, antibodies to lens epithelium-derived growth factor (LEDGF) can lyse epithelial cells when incubated with the intact lens [5], indirectly suggesting that the immunoglobulin molecules can pass through the anterior portion of the lens capsule. These studies suggest that the lens capsule is permeable to some macromolecules.

The anterior portion of the lens capsule is in direct contact with the aqueous humor. This portion of the capsule is lined by a simple cuboidal epithelium. The basal epithelial surface is in contact with the lens capsule, while the apical surface is in contact with cortical fiber cells. The apical-lateral membranes of epithelium form tight junctions that exclude the passage of molecules from the aqueous humor into the lens cortex. Previous studies [6,7] have shown that small molecules such as lanthanum nitrate (molecular weight 430) are blocked from passage into the intercellular space of anterior cortical fiber cells by tight junctions (zonulae occludentes) that join epithelial cells. Alternatively, it has been suggested by Lo et al. [8], that a large molecule such as horseradish peroxidase can be internalized into epithelial cells via a mechanism involving endocytosis at the apical end of lateral membranes of epithelial cells. These scientists also reported that the endocytic activity of different species varied, with rats having much less activity than rabbits or guinea pigs. It is possible that some of the internalized macromolecule could be transcytosed to the apical membrane of epithelial cells that border superficial cortical cells, resulting in transcellular passage of the macromolecule through the lens epithelium. Based upon this possibility, transcytosis of large molecular weight molecules from the basal-lateral to apical side of the epithelium might be facilitated by prior encapsulation of the molecules within liposomes [9], with a defined charge that takes advantage of the net negative charge on the surface of many mammalian cells.

Finally, it has been thought that fiber cells in both the cortex and nucleus can only communicate with other fiber cells through gap junctions. The restricted pore size of the gap junctions excludes intercellular passage of large molecular weight macromolecules greater than approximately 1.5 kDa [10-12]. However, Shestopalov et al. [13] have recently shown that in embryonic chicken lenses, autofluorescent proteins initially localized within one fiber cell of the nucleus, can pass into the interior of neighboring fiber cells, in a time dependent manner. This study indicates that a novel cell-cell fusion pathway may exist in the lens core for the movement of large macromolecules from one cell to another. This mechanism was not observed in the superficial cortex. However, cell-cell fusions have been observed in the adult frog cortex [14]. These studies [13-16] raise the intriguing possibility that fiber cell fusion may be utilized by all lenses for intercellular passage of large molecular weight components from one cell to another.

To explore the possible translocation of large molecular weight molecules in culture medium into cells of various parts of the lens, we have used confocal microscopy to monitor real time movement of fluorescently labeled proteins, carbohydrate and DNA through the capsule and into the epithelium and fiber cells of cultured rat lenses. The results show that these macromolecules can translocate through the capsule into cells of the epithelium. Quite surprisingly, proteins and carbohydrate transcytosed the epithelium and were internalized by cortical fiber cells.

Methods

Native α-, β-, and γ-crystallins were prepared from fetal bovine lenses using HPLC gel permeation chromatography as previously described [17]. After dialysis against distilled water and lyophilization, protein was determined according to Bradford [18], using bovine serum albumin as standard.

A stock solution of Texas red sulfonyl chloride dye (Molecular Probes, Eugene, OR) was prepared by adding 1.0 ml of dimethylformamide to 1 mg of Texas red dye. Approximately 50 μl of Texas red stock was added to 2.5 mg of protein dissolved in 0.25 ml of 0.1 M sodium bicarbonate buffer, pH 9.0. This solution was light protected and incubated at 4 °C for 1 h with constant agitation followed by exhaustive dialysis against distilled water. The dialyzed material was concentrated using a Micron microconcentrator (Amicon, Beverly, MA), then stored light protected at -20 °C until use. Dextran-Texas red (approximate molecular weight 70,000) was obtained from Molecular Probes.

A DNA plasmid containing the gene for enhanced yellow-green variant of fluorescent protein from Aequorea victoria, fused to the CMV (PCMV1E) promoter was obtained from Clontech Laboratories, Palo Alto, CA. To follow the translocation of plasmid DNA into the lens, 250 μg of plasmid was added to 0.5 ml of phosphate buffered saline, pH 7.4, containing 1.5 μM propidium iodide (Molecular Probes). This solution was then dialyzed exhaustively against distilled water in the dark, followed by concentration using a Micron microcentrator and stored until use as previously described.

BioPORTER lipid preparations were obtained from Gene Therapy Systems, San Diego, CA. BioPORTER lipid preparations were used to encapsulate protein and carbohydrate into liposomes. A stock preparation of BioPORTER was prepared by adding 250 μl of methanol to one vial of dried BioPORTER supplied by the manufacturer and vortexing the vial for 10-20 s at high speed. Aliquots (25 μl) of the stock were added to 1.5 ml Eppendorf tubes, dried under a hood at RT and stored at -20 °C until use. To encapsulate macromolecules, 250 μg of macromolecule described previously was added to one tube of dried BioPORTER and then vortexed at high speed for 10 s. This was then added to one culture dish containing 2 ml of serum free medium and one whole lens.

GenePORTER lipid preparations were obtained from Gene Therapy Systems, San Diego, CA, and used to encapsulate plasmid DNA. A stock preparation of GenePORTER was prepared according to the manufacture instructions. Aliquots (5 μl) of the stock were added to 1.5 ml Eppendorf tubes and stored at -20 °C until use. To encapsulate DNA, 10 μg of plasmid DNA was diluted into 125 μl of serum free culture medium and added to one tube of GenePORTER, then vortexed at high speed for 10 s. This was then added to one culture dish containing 2 ml of serum free medium and one whole lens.

Adult Sprague Dawley male rats 3 to 4 months old and approximately 300 to 400 g were euthanized by carbon dioxide inhalation. Guidelines by the Institute for Laboratory Animal Research were followed. The lenses were removed from enucleated eyes, and then incubated at 37 °C in a Delta T Dish (Bioptechs, Butler, PA) containing 2.0 ml of sterile Eagle's Minimum Essential Media (MEM) containing 52 mM sodium bicarbonate, 2 mM L-glutamine, 100 units/ml penicillin, 0.1 mg/ml streptomycin, and 0.1 mg/ml of phenol red. Prior to adding fluorescently labeled macromolecules, all lenses were assessed by light and fluorescence microscopy for capsule integrity and lens epithelial cell viability. Any lenses that had enlarged Y sutures, areas of discontinuity of lens epithelium, SYTOX Green (Molecular Probes) labeling of nuclei, red autofluorescence or tears in the capsule with visible protrusion of fiber cells were removed from these experiments. Fluorescently labeled macromolecules included crystallins or dextran, or plasmid DNA, with or without encapsulation in BioPORTER or GenePORTER liposomes. Localization of fluorescently labeled macromolecule over time was assessed on at least 6 different lenses using a Zeiss laser scanning confocal microscope (model 410, equipped with an Axiovert 100 inverted microscope, an argon-krypton 488/568/647 laser, Plan Neofluar 20x/0.5, Ft488/568 dichroic beam splitter, KP 600 line selection filter, LP 590 emission filter, and the software package LSM, ver. 3.993, Zeiss, Thornwood, NY). Individual optical sections and z-scans were taken with the pinhole set at 14 (full width at half-maximum, 5 μm). Digital image files in tagged image file format (TIFF) were imported into image editing software (Photoshop version 4.0; Adobe Systems, Mountain View, CA) for labeling, image sizing, brightness, and contrast adjustments. A 290 min time series was taken of the central anterior epithelium with an optical section (5 μm thick), scanned every 10 min. To determine depth of macromolecule penetration into the lens, a z-scan at 90° to the corresponding optic section was taken. The pixel intensity was quantified using Scion image Beta 4.0.2 software (Scion Corporation, Frederick, MD) by taking the average of a 20 μm square encompassing the epithelium. Temperature of the incubation medium was kept constant at 37 °C using a Delta T heater system (Bioptechs, Butler, PA). After 290 min of incubation and monitoring by confocal microscopy, the lenses were fixed by overnight incubation in 2% (v/v) paraformaldehyde in PBS pH 7.4, followed by vibratome cutting into 100 μm sections and analysis with confocal microscopy using a 63x/1.4 (full width at half-maximum, 0.65 μm) or a 40x/1.3 (full width at half-maximum, 0.78 μm) objective.

Results

Figure 1 shows the results of confocal, real time analysis of the lens epithelial layer from lenses incubated with fluorescently labeled α-crystallins, with or without encapsulation with BioPORTER liposomes. Without encapsulation (Figure 1A), a faint fluorescence was first seen at approximately 270 min after the addition of α-crystallins, followed by a gradual increase in fluorescence up to 290 min. Encapsulation with BioPORTER liposomes (Figure 1B) resulted in the appearance of fluorescence at an earlier time (approximately 180 min). In addition, the fluorescence intensity was much greater with encapsulation than without at any given time. Quantization of pixel intensity at different time periods (Figure 2) confirmed that fluorescence intensity increased with increasing time, indicating that increasing amounts of α-crystallins were endocytosed and/or accumulating into epithelial cells. This also showed that BioPORTER facilitated the uptake of fluorescent α-crystallins into epithelial cells.

To determine if fluorescently labeled α-crystallins were transcytosed from epithelium into deeper layers of the lens, confocal microscopy was used to visualize fluorescence in a z-scan perpendicular to the plane of focus in Figure 1. This z-scan consisted of the media, capsule, epithelium, and the cortical fiber cells. Figure 3A,B both show that after 290 min of incubation with fluorescently labeled α-crystallins, the fluorescence was localized into the anterior superficial cortex. This penetration was not due simply to cell death, since epithelial cells were viable after 290 min of incubation, as determined by their ability to exclude SYTOX green dye (Molecular Probes).

The images seen in Figure 1, Figure 2, and Figure 3 showed that the fluorescently labeled α-crystallins were able to enter the epithelium and cortex of the intact lens, but at this resolution (5 μm thick optical section), it was difficult to determine if the protein was intracellular or intercellular. To obtain higher resolution confocal images, the intact lens was fixed with paraformaldehyde, then vibratome sectioned, and the individual sections analyzed by confocal microscopy using a 63x/1.4 N.A. objective. Figure 4A shows a focal plane that contains the epithelium. The fluorescence preferentially appears in the cytoplasm (arrows), as compared with the nuclei (arrowheads). The same fluorescence pattern was seen in the epithelial cells of lenses treated with α-crystallins encapsulated with BioPORTER liposomes (results not shown). Figure 4B,C show representative confocal images taken from a vibratome section containing cortical fiber cells from the same lens. Fluorescence areas (arrows) corresponding to cytoplasm of fiber cells are seen in the focal plane, showing the intracellular location of fluorescently labeled α-crystallins. A DIC image of the same area (Figure 4C) confirms that the fluorescence correspond to the cytoplasm of individual fiber cells cut in cross-section. The results of Figure 1, Figure 2, Figure 3, and Figure 4 show that fluorescently labeled α-crystallins were able to transcytose the epithelium, to be internalized by superficial cortical fiber cells of the lens.

To determine if other macromolecules of differing size and charge could be internalized, Texas red labeled β-crystallins, γ-crystallins, and dextran were incubated with intact lenses, without and with prior encapsulation with BioPORTER liposomes. Figure 5A,C,E show that all three classes of macromolecules are internalized into epithelial cells, where preferential localization is in the cytoplasm (arrows) and not nucleus (arrowheads). In Figure 5A, fluorescent γ-crystallins can be seen in individual superficial fiber cells (large arrows). Furthermore, the results of Figure 5B,D,F show that these same molecules transcytose the epithelium into the cortex. The results of Figure 5 involve β-crystallins, γ-crystallins, and dextran that had been encapsulated with BioPORTER liposomes prior to incubation with lenses. Similar results, but of lower fluorescence intensity, were also seen with the same protein macromolecules that had not been encapsulated (results not shown).

Since the lens crystallins and dextran were internalized by cells of the epithelium and cortical fiber cells of the intact lens, it might also be possible for other types of biologically important macromolecules to exhibit the same behavior. This would be especially relevant to nucleic acids, which could be used to express functionally important proteins in epithelium and/or cortical fiber cells of lenses. To test this possibility, a 4.7 kb plasmid was labeled with propidium iodide, and then incubated with intact lenses, with and without encapsulation with GenePORTER liposomes, which have been specifically designed to facilitate entry of DNA into mammalian cells. Figure 6 shows confocal, real time analysis of the fluorescently labeled plasmid in the focal plane containing central epithelial cells. Without encapsulation (Figure 6A), fluorescence is first seen after approximately 100 min of incubation, with a gradual increase in the intensity during the remaining time of incubation. The encapsulated plasmid (Figure 6B) shows very low if any entry into the central epithelial layer.

In contrast to the results of lenses incubated with α-crystallins, z-scan analysis shows that lenses incubated with plasmid DNA without (Figure 7A) or with (Figure 7C) prior encapsulation with GenePORTER liposomes showed localization only in the epithelial, with no detectable fluorescence in the cortex. Without prior encapsulation, fluorescently labeled DNA was internalized by the majority of epithelial cells (Figure 7A,B) and localized to the nucleus (arrowheads), even in more equatorial areas of epithelium not shown in this image. In comparison, encapsulated DNA (Figure 6C) was not present in the central epithelium in this lens. However, regions more equatorial (Figure 7C) showed regions where patches of epithelium were fluorescent while adjacent epithelium did not internalize DNA (larger arrow). Furthermore, fluorescently labeled DNA was preferentially localized to the cytoplasm (Figure 7D, arrow), and not the nucleus (arrowhead).

Discussion

In spite of its obvious importance, very little, if anything, is known about the transcytosis of biologically important macromolecules across the epithelium and into fiber cells. It has been generally accepted that most macromolecules such as proteins can pass through the anterior lens capsule, where they are blocked from further passage into the lens by tight junctions surrounding epithelial cells [6-8,19]. Based upon "wash-out" experiments, Lo et al. [6,8,19] suggested that some of these macromolecules such as horseradish peroxidase (HRP) may pass through the capsule at the equatorial regions that are devoid of tight junctions, then pass through the intercellular space between epithelium and superficial fiber cells of the cortex. However, these studies [8,19] did not report that HRP was internalized by anterior fiber cells. Alternatively, it has also been shown [8] that macromolecules were endocytosed by epithelial cells and differentiating fiber cells. Such a process may be particularly important for the homeostasis of epithelial cells, which are metabolically active. In addition to low molecular weight metabolites and ions, these cells may also require the internalization of large molecular weight components such as proteins that are known to exist in the aqueous humor [20].

The latter mechanism is consistent with our observation that a very large protein complex such as α-crystallins (approximate molecular weight 600-900 kDa [21]) and β-crystallins (approximate molecular weight 50-150 kDa [21]) can readily enter the cytoplasm of epithelial cells present in cultured lenses. While it is possible that some crystallins are entering the lens at the equatorial region, then passing through the intercellular space between epithelium and cortex as suggested by Lo et al. [19], the small size of this intercellular space (approximately 20 nm [22]) may exclude the passage of a very large molecular weight complexes such as α- and β-crystallins. In addition, previous reports [8] indicated that HRP, after a similar time period to these experiments, was not present intracellularly in anterior cortical fiber cells. Furthermore, an extensive literature search did not produce any references reporting that macromolecules may actually transcytose the epithelium and be internalized by cortical fiber cells. HRP endocytosis by lens epithelium was reported to localize to secondary lysosomes [8], indicating that this macromolecule was targeted for digestion by epithelial cells once internalized and not transcytosed to the underlying fiber cells. The appearance of Texas red labeled macromolecules in fiber cells in these experiments indicate that these molecules possibly transcytosed the lens epithelium and were internalized by fiber cells. This was not believed to be due to the proteolysis of macromolecules in the medium or epithelial cells and the subsequent passage of peptide fragments or free dye into deeper layers of the lens through gap junction. In support of this theory, SDS-PAGE demonstrated no detectable breakdown of fluorescently labeled α-crystallin in cultured whole lenses (results not shown). These results raise the intriguing possibility that selective macromolecules may actually transcytose the epithelium and subsequently be internalized by the underlying fiber cells.

This possibility is consistent with our results showing movement of fluorescently labeled macromolecules into epithelial cells and the underlying fibers. This is also consistent with the finding that fluorescent protein internalization into epithelium is enhanced by prior encapsulation in BioPORTER liposomes. It is most probable that liposomes fuse with the first lipid membrane that they come in contact with. In these studies, encapsulated macromolecules most likely fused with the plasma membrane of epithelial cells and did not penetrate to deeper layers or pass for any significant distance between lens fiber cells. In addition, visualization of the liposomes by confocal microscopy revealed vesicles with an average diameter of approximately 1 μm (results not shown). Based upon the value of 20 nm for the intercellular space between lens epithelial and fiber cells (22), it is probable that the encapsulated α-crystallin cannot enter the intercellular space between epithelial cells and fiber cells, or between the intercellular spaces between adjoining fiber cells. More probable is a mechanism involving initial binding of the cationic liposomes to the probable anionic surface of epithelial cells, followed by membrane fusion, endocytosis and possible transcytosis of macromolecules in a manner analogous to the processes known to occur within epithelial and endothelial cells in other tissues of the body [23,24]. However, the possible transcytosis of macromolecules across the epithelium is not absolutely dependent upon the net charge of the macromolecule, since we have also found that macromolecules with differing net charges such as the α, β, and γ-crystallins, as well as dextran molecules, can all be internalized without prior encapsulation. Encapsulation with BioPORTER liposomes accelerates the internalization process of proteins, consistent with the use of liposomes that optimize the membrane binding and/or fusion processes to deliver macromolecules into cells.

An unexpected result our study is that a large molecular weight complex such as α-crystallins, after apparently passing through the epithelial cell layer, can then pass into the interior of superficial fiber cells of the cortex, where the macromolecule can further pass into the interior of deeper fiber cells. Although the mechanism of this process in the lens is currently unknown, it is possible that it might also involve an endocytic mechanism similar to what might be occurring at the basal-lateral membrane surface of lens epithelial cells [8,17]. Alternatively, it is possible that once macromolecules transcytose the epithelium and are internalized by superficial fiber cells, intercellular passage of macromolecules into deeper fiber cells may, at least in part, be mediated by regions of fiber cell/fiber cell membrane fusion. Areas of membrane fusion between cortical and nuclear fiber cells have been reported in the adult frog lens [14] and in nuclear fiber cells in the developing chick lens [13]. Whatever the mechanism of passage, it is clear that the theory involving fiber cell-to-fiber cell passage of only low molecular components via gap junctions must be revised, to include the observation that large molecular weight proteins such as lens crystallins and dextran in the current study and green fluorescent protein in a previous study [13] can indeed pass between fiber cells.

Perhaps the most exciting result of the current study involves the observation that in addition to proteins and carbohydrate, plasmid DNA of sufficient size (4.7 kb) to code for proteins can also pass through the capsule and be internalized into epithelial cells. Based upon the results of Figure 7, without encapsulation almost all DNA is localized to the nucleus. After prior encapsulation, the nuclear translocation of DNA is inhibited and localization was to the cytoplasm of epithelial cells. These results, in conjunction with those of the other macromolecules and previous reports [8], indicate that different macromolecules may have different fates once internalized by lens epithelial cells. Some may be targeted for lysosomal breakdown, others may translocate to the nucleus, while still others may be targeted for delivery to the underlying fiber cells.

Whatever the mechanisms involved in protein and DNA internalization, trafficking and localization, the results of the current study clearly show that these biologically important macromolecules can be directly internalized into the intact lens in culture. Based upon this observation, it may be possible in future studies, using intraocular injection, to selectively internalize functionally important proteins such as enzymes, growth factors, chaperones, etc. into the lens epithelium, where they can be used to revitalize metabolic processes that have been attenuated by aging and/or cataractogenesis. Numerous studies have strongly suggested that cataractogenesis may not only involve posttranslational modification within the fiber cells, but may also involve changes in fundamental metabolic processes occurring in the epithelium [25-27]. Even more effective may be the internalization of plasmids coding for key metabolic enzymes and/or molecular chaperones. The feasibility of this approach was suggested in a study [28] which reported that the possible expression of GFP in lens epithelium after intravitreal injection of rBV-CMV GFP in mice. In the absence of encapsulation, the present study shows that DNA is preferentially localized to the nucleus of epithelial cells, where it can be readily transcribed. After translation in the cytosol, the newly synthesized protein may not only be able to augment the metabolism of epithelial cells, but based upon our results, it may also be able to pass into the interior of fiber cells, where it could exert the same therapeutic effects as in the epithelium. In this manner, it may be possible in future studies to develop strategies directed towards the rehabilitation of specific metabolic processes that are known to be diminished during the aging and/or cataractogenic processes of the intact lens.

Acknowledgements

Supported by a grant from the National Eye Institute.

References

1. Fels IG. Permeability of the anterior bovine lens capsule. Exp Eye Res 1970; 10:8-14.

2. Fels IG. Permeability of the lens capsule to tryptophan. Exp Eye Res 1971; 12:51-9.

3. Hockwin O, Poonawalla N, Noll E, Licht W. [Permeability of the isolated lens capsule to amino acids and water-soluble proteins]. Albrecht Von Graefes Arch Klin Exp Ophthalmol 1973; 188:175-81.

4. Friedenwald JS. Permeability of the lens capsule. Arch Ophthalmol 1930; 3:182-93.

5. Ayaki M, Sueno T, Singh DP, Chylack LT Jr, Shinohara T. Antibodies to lens epithelium-derived growth factor (LEDGF) kill epithelial cells of whole lenses in organ culture. Exp Eye Res 1999; 69:139-42.

6. Lo WK, Harding CV. Tight junctions in the lens epithelia of human and frog: freeze-fracture and protein tracer studies. Invest Ophthalmol Vis Sci 1983; 24:396-402.

7. Johnson MJ, Unakar NJ. Alterations in lens permeability during galactose cataract development in rat. Lens Eye Toxic Res 1992; 9:93-113.

8. Lo WK, Zhang W. Endocytosis of macromolecules in the lenses of guinea pig and rabbit. Lens Eye Toxic Res 1989; 6:603-12.

9. Zelphati O, Wang Y, Kitada S, Reed JC, Felgner PL, Corbeil J. Intracellular delivery of proteins with a new lipid-mediated delivery system. J Biol Chem 2001; 276:35103-10.

10. Kistler J, Evans C, Donaldson P, Bullivant S, Bond J, Eastwood S, Roos M, Dong Y, Gruijters T, Engel A. Ocular lens gap junctions: protein expression, assembly, and structure-function analysis. Microsc Res Tech 1995; 31:347-56.

11. Peracchia C, Girsch SJ, Bernardini G, Peracchia LL. Lens junctions are communicating junctions. Curr Eye Res 1985; 4:1155-69.

12. Harris AL, Walter A, Paul D, Goodenough DA, Zimmerberg J. Ion channels in single bilayers induced by rat connexin32. Brain Res Mol Brain Res 1992; 15:269-80.

13. Shestopalov VI, Bassnett S. Expression of autofluorescent proteins reveals a novel protein permeable pathway between cells in the lens core. J Cell Sci 2000; 113:1913-21.

14. Kuszak JR, Macsai MS, Bloom KJ, Rae JL, Weinstein RS. Cell-to-cell fusion of lens fiber cells in situ: correlative light, scanning electron microscopic, and freeze-fracture studies. J Ultrastruct Res 1985; 93:144-60.

15. Kuszak JR, Novak LA, Brown HG. An ultrastructural analysis of the epithelial-fiber interface (EFI) in primate lenses. Exp Eye Res 1995; 61:579-97.

16. Kuszak JR, Ennesser CA, Bertram BA, Imherr-McMannis S, Jones-Rufer LS, Weinstein RS. The contribution of cell-to-cell fusion to the ordered structure of the crystalline lens. Lens Eye Toxic Res 1989; 6:639-73.

17. Hansen JS, Takemoto DJ, Takemoto LJ. Monoclonal antiserum to gamma crystallin: Characterization and use as a probe of changes in the aging human lens. Lens Res 1985; 2:207-19.

18. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976; 72:248-54.

19. Lo WK. In vivo and in vitro observations on permeability and diffusion pathways of tracers in rat and frog lenses. Exp Eye Res 1987; 45:393-406.

20. Davson H. Physiology of the eye. 4th ed. Edinburgh: Churchill Livingstone; 1980. p. 18.

21. Harding JJ, Crabbe MJ. The lens: Development, proteins, metabolism and cataract. In: Davson H, editor. The Eye. Vol IB. Orlando: Academic Press; 1984. p. 207-492.

22. Philipson BT, Hanninen L, Balazs EA. Cell contacts in human and bovine lenses. Exp Eye Res 1975; 21:205-19.

23. Okamoto CT. Endocytosis and transcytosis. Adv Drug Deliv Rev 1998; 29:215-28.

24. Matveev S, Li X, Everson W, Smart EJ. The role of caveolae and caveolin in vesicle-dependent and vesicle-independent trafficking. Adv Drug Deliv Rev 2001; 49:237-50.

25. Karim AK, Jacob TJ, Thompson GM. The human anterior lens capsule: cell density, morphology and mitotic index in normal and cataractous lenses. Exp Eye Res 1987; 45:865-74.

26. Straatsma BR, Lightfoot DO, Barke RM, Horwitz J. Lens capsule and epithelium in age-related cataract. Am J Ophthalmol 1991; 112:283-96.

27. Kantorow M, Horwitz J, Carper D. Up-regulation of osteonectin/SPARC in age-related cataractous human lens epithelia. Mol Vis 1998; 4:17 <http://www.molvis.org/molvis/v4/a17/>.

28. Haeseleer F, Imanishi Y, Saperstein DA, Palczewski K. Gene transfer mediated by recombinant baculovirus into mouse eye. Invest Ophthalmol Vis Sci 2001; 42:3294-300.