Integrin αVβ5-mediated Removal of Apoptotic Cell Debris by the Eye Lens and Its Inhibition by UV Light Exposure*

Accumulation of apoptotic material is toxic and associated with cataract and other disease states. Identification of mechanisms that prevent accumulation of apoptotic debris is important for establishing the etiology of these diseases. The ocular lens is routinely assaulted by UV light that causes lens cell apoptosis and is associated with cataract formation. To date, no molecular mechanism for removal of toxic apoptotic debris has been identified in the lens. Vesicular debris within lens cells exposed to UV light has been observed raising speculation that lens cells themselves could act as phagocytes to remove toxic apoptotic debris. However, phagocytosis has not been confirmed as a function of the intact eye lens, and no mechanism for lens phagocytosis has been established. Here, we demonstrate that the eye lens is capable of phagocytizing extracellular lens cell debris. Using high throughput RNA sequencing and bioinformatics analysis, we establish that lens epithelial cells express members of the integrin αVβ5-mediated phagocytosis pathway and that internalized cell debris co-localizes with αVβ5 and with RAB7 and Rab-interacting lysosomal protein that are required for phagosome maturation and fusion with lysosomes. We demonstrate that the αVβ5 receptor is required for lens epithelial cell phagocytosis and that UV light treatment of lens epithelial cells results in damage to the αVβ5 receptor with concomitant loss of phagocytosis. These data suggest that loss of αVβ5-mediated phagocytosis by the eye lens could result in accumulation of toxic cell debris that could contribute to UV light-induced cataract formation.

Accumulation of apoptotic material is toxic to epithelial cell populations (1)(2)(3)(4) and is associated with development of multiple autoimmune, inflammatory, aging, and degenerative diseases (1)(2)(3)(4)(5)(6)(7)(8)(9)(10). Many of these diseases are associated with exposure of affected tissues to specific environmental insults that result in apoptosis (8). Identification of cell systems that protect against the effects of apoptosis-inducing insults is an important step toward understanding and developing therapies to treat these diseases. An excellent model system to identify these mechanisms is the transparent eye lens that functions to focus light onto the retina where visual information is transmitted to the brain (11). The lens is one of the most environmentally challenged tissues of the body because it lacks protective pigmentation and resides just behind the transparent and surface-exposed cornea (11). Damage to the lens and its components causes age-related cataract formation (12)(13)(14) that remains a leading cause of visual disability despite advances in surgical treatment options (15)(16)(17). Although many lens exposures have been proposed to contribute to age-related cataract formation, long term exposure to UV light is a significant risk factor for the development of human cataracts (18 -20).
UV light has multiple damaging effects on the eye lens (18 -24). A major consequence of UV light exposure to the lens is damage to the light-facing monolayer of anterior epithelial cells (21)(22)(23)(24) that overlie and are required for the homeostasis of the organelle-free fiber cells that make up the lens core (11,25,26). UV light exposure has been shown to result in apoptosis of lens epithelial cells (21,22,(27)(28)(29), and damage to lens epithelial cells and their subcellular components results in cataract formation (22, 24, 30 -32). These data suggest the need for a lens mechanism to remove apoptotic debris generated upon UV light exposure to prevent epithelial cell death and thereby maintain lens transparency.
A common mechanism to remove and detoxify apoptotic cell debris is phagocytosis that utilizes specialized cell surface receptors to recognize and engulf extracellular apoptotic debris associated with tissue homeostasis and remodeling (33). During this process, phagocytosed material is enclosed in a membranebound structure called a phagosome (34). Phagosomes containing internalized material undergo progressive acidification and ultimately fuse with lysosomes where internalized debris is degraded and detoxified (33,35,36). Lysosomes accumulate in the region around the microtubule-organizing center, located in the perinuclear region of the cell (37).
In most tissues, specialized phagocytes such as macrophages and dendritic cells operate to phagocytose apoptotic cells and cell debris (4,38,39). However, the mature lens is an encapsulated tissue with no blood supply or obvious source of specialized phagocytes suggesting the possibility that lens epithelial cells themselves could act as phagocytes that remove apoptotic cells debris resulting from exposure to UV light and other apoptotic agents.
Early studies identified the presence of vesicles containing what was described as apoptosis-like material in UV-treated cultured lens epithelial cells (40) and in the intact lenses of UV-treated rodents (21,31), suggesting the possibility that lens cells could act as phagocytes to remove and detoxify the cell debris in the apoptotic lens caused by UV light exposure. Similar vesicles have also been found in the lens epithelium of UV light-treated rabbit lenses (24) and human age-related cataract lenses (32). However, these pioneering studies did not confirm that the material contained within the observed vesicles was truly of lens epithelial cell or even of apoptotic origin, and they did not establish whether the vesicles originated from actual phagocytosis by lens cells or by some other intracellular vesicleproducing process, including lens cell autophagy, that has recently been established (41)(42)(43). Importantly, these early studies provided no molecular mechanism to account for the formation of the identified vesicles.
Here, we sought to establish that the intact eye lens is indeed capable of phagocytizing apoptotic cell debris, and we sought to identify a potential molecular mechanism for phagocytosis by lens cells. Using dual-color fluorescent labeling of lens epithelial cells and apoptotic lens cell debris, we demonstrate that isolated lens epithelial cells and epithelial cells of the intact eye lens actively phagocytize extracellular apoptotic lens epithelial cell debris. Using a combination of transcriptome profiling, colocalization studies, and functional analysis, we demonstrate that the lens expresses the entire complement of components employed by the integrin ␣V␤5 pathway to mediate phagocytosis and that the function of ␣V␤5 is required for phagocytosis of lens apoptotic cell debris by lens epithelial cells. Finally, we demonstrate that exposure of lens epithelial cells to UV light results in damage to ␣V␤5 in association with loss of lens epithelial cell phagocytosis. These data establish that the intact eye lens removes apoptotic cell debris by lens cell phagocytosis, and they establish that this process requires the actions of the ␣V␤5 phagocytosis receptor. The results suggest that UV light-induced damage to integrin ␣V␤5 and the consequential loss of lens phagocytosis could result in accumulation of toxic apoptotic cell debris, death of lens epithelial cells, and ultimately cataract formation.
Preparation of Chicken Primary Lens Epithelial Cells-Primary chicken lens epithelial cell cultures were prepared from the lenses of embryonic day 10 (E10) 3 White Leghorn embryonated chicken eggs (Charles River Laboratories, Storrs, CT) using the method of Menko et al. (45). Briefly, primary lens cells were isolated from chicken lenses by trypsinization and agitation. Cells were plated onto glass bottom dishes coated with mouse laminin (catalog no. 23017015, Invitrogen) and cultured in Medium 199 (catalog no. 11150067, Invitrogen) supplemented with 10% FBS (catalog no. 10437028, Invitrogen) and penicillin/streptomycin antibiotic mix (50 units/ml; catalog no. 154140, Invitrogen).
Chicken Lens Explants-E13 chicken lens tissue explants were prepared by surgical removal of the lens epithelium from the bulk of the lens fibers by use of fine forceps. Explants were cultured in serum-free M199 media containing penicillin/ streptomycin. Explants were immediately incubated with beads for 16 h in serum-free media in 35-mm 2 glass bottom tissue culture dishes. After 16 h, the explants were washed three times with PBS, fixed in 4% formaldehyde, and counterstained with ␣-tubulin and DAPI nuclear stain as described below.
Ex Vivo Chick Lens Culture-E13 chicken lenses were prepared by surgical removal of the lens from the vitreous by anterior approach. Lenses were cultured in 96-well tissue culture plates in serum-free M199 media and immediately incubated with GFP-labeled primary chicken lens epithelial cell debris for 4 h. Following the 4-h incubation, lenses were washed three times with PBS and prepared for cryosectioning and immunolabeling as described below.
Assays for Phagocytosis of Fluorescent Labeled Substrates-2.0-m yellow-green ( ex / em ϭ 505/515 nm) carboxylated FluoSpheres (catalog no. F8827, Invitrogen) (hereafter referred to as beads) were vortexed and in all cases were added to cells at a concentration of 5.05 million beads/ml of culture media. Fluorescein-labeled attenuated bacterial particles (Vybrant, catalog no. V-6694, Invitrogen) were prepared according to the manufacturer's instructions and vortexed, and in all cases 100 l of the fluorescein-labeled bacterial particle suspension was added per ml of culture media.
SRA 01/04 cells were plated at a density of 150,000 cells/well on glass bottom 35-mm 2 tissue culture dishes (catalog no. D35-20-0-N, In Vitro Scientific, Sunnyvale, CA), and beads or fluorescein-labeled bacterial particles were added as described above, and at indicated times the cells were washed three times with PBS and fixed in 3.7% formaldehyde. Cells were counterstained with ␣-tubulin (catalog no. ab18251, Abcam, Cambridge, UK) and stained with 300 nM DAPI nuclear stain (catalog no. D1306, Invitrogen) as described in detail below. Primary chicken lens epithelial cells were plated onto 12-well glass bottom multiwell plates (catalog no. P12G-1.5-14-F, MatTek Corp., Ashland, MA) coated with mouse laminin or onto laminin-coated 12-mm round coverslips (catalog no. 354087, BD BioCoat, BD Biosciences). Beads or fluorescein-labeled bacterial particles (Vybrant) were added as described above, and at the indicated times the cells were washed three times with PBS and fixed in 3.7% formaldehyde. Cells were then counterstained with ␣-tubulin and DAPI nuclear stain.
Post-fixation, cells/explants were washed three times with ice-cold PBS, permeabilized using 0.25% Triton X-100 for 15 min at room temperature, and blocked in blocking buffer (1% BSA (catalog no. BP1600, Fisher), 0.2% Tween (catalog no. P1379, Sigma), and PBS) for 30 min. Post-blocking, cells were incubated with polyclonal rabbit primary antibody against ␣-tubulin (1:1000) and/or monoclonal ␣V␤5 antibody (catalog no. PIF6, Developmental Studies Hybridoma Bank, University of Iowa, Iowa City) overnight at 4°C in blocking buffer. Cells were washed three times in PBS and incubated with fluorescently conjugated anti-rabbit or anti-mouse secondary anti-body (1:2000, Alexa-Fluor 488 or 555, catalog nos. A11001, A11008, A21422, and A21428, Invitrogen) in blocking buffer for 1 h at room temperature. Cells were washed in PBS three times, and the nuclei were counterstained using 300 nm of DAPI for 5 min followed by three washes with PBS.
Transmission Electron Microscopy Analysis of Phagocytosis by Lens Epithelial Cells-For transmission electron microscopy examination of phagocytosis of beads, beads were added to chicken primary cells grown on coverslips for 16 h followed by overnight fixation in 2% glutaraldehyde in PBS at 4°C. Postfixation, cells were rinsed in three washes of 0.1 M phosphate buffer and then post-fixed in 2% osmium tetroxide in 0.1 M phosphate buffer for 1 h. After buffer washes, the cells were dehydrated through a series of graded ethanol and embedded using EMbed (Electron Microscopy Sciences, Hatfield, PA) overnight in a 64°C oven. Silver/gold sections were cut on a Leica Ultracut R (Leica, Buffalo Grove, IL) and stained in uranyl acetate and lead citrate, and images were captured by a Gatan Orius SC 200S CCD camera (Gatan, Pleasanton, CA) in a JEOL 1400 electron microscope (FEI, Hillsboro, OR).
Primary chicken lens epithelial cells were reverse-transfected with pAcGFP-1 or pDsRed Monomer-C1 vectors (Clontech) using Lipofectamine 2000 (Invitrogen) transfection reagent according to the manufacturer's instructions. Transfection efficiency was ϳ80% in primary chicken lens epithelial cells transfected with pDsRed.
Assay for Phagocytosis of Apoptotic Lens Cell Debris-48 h following transfection, the pAcGFP-C1 overexpressing SRA 01/04 or primary chicken lens epithelial cells were treated with 2 M staurosporine (catalog no. S6942, Sigma) for 4 h to induce apoptosis. We acknowledge that staurosporine may also cause programmed necrosis (necroptosis); therefore, we optimized our staurosporine treatment to ensure that the cells were undergoing apoptosis. However, we cannot rule out that some necrotic cells were also added. After 4 h, the GFP-labeled apoptotic cells were washed with PBS and harvested into PBS, and equal amounts of apoptotic cell suspension (100 l) were added to pDsRed-overexpressing SRA 01/04 or pDsRed-overexpressing primary chicken lens epithelial cells plated onto glass-bottom 35-mm 2 tissue culture dishes. At the indicated times, pDsRed-overexpressing SRA 01/04 or pDsRed overexpressing primary chicken lens epithelial cells were fixed with 3.7% paraformaldehyde and counterstained with the nuclear dye DAPI (300 nm). Cells were visualized using a Zeiss LSM700 confocal microscope (Carl Zeiss Microscopy, Jena, Germany).
Immunostaining of E13 Chicken Lens Sections-Freshly isolated or ex vivo cultured E13 chicken lenses were washed in PBS containing calcium and magnesium (PBS-CM) and fixed for 18 h in 3.7% paraformaldehyde/PBS-CM at 4°C and then dehydrated in 30% sucrose/PBS-CM for at least 24 h or longer. Midsagittal lens sections were prepared by cryosectioning (20-m thick sections). Sections were permeabilized using 0.25% Triton X-100 (catalog no. T1001, Anatrace, Santa Clara, CA) in PBS for 10 min at room temperature, followed by three washes in PBS. Nonspecific antibody binding was blocked using 2.5% horse serum (catalog no. S-2012, Vector Laboratories, Burlingame, CA) for 1 h at room temperature. Primary ␣V␤5, RAB7 (catalog no. SC-10767, Santa Cruz Biotechnology), or RILP (catalog no. SC98331, Santa Cruz Biotechnology) antibodies were diluted 1:50 in 1% bovine serum albumin, 0.25% Triton X-100 in PBS, and the sections were incubated overnight at 4°C. Sections were then washed three times in PBS followed by fluorescently conjugated anti-mouse and/or anti-rabbit secondary antibody (1:2000, Alexa-Fluor 488 or 555, catalog nos. A11001, A11008, A21422, and A21428, Invitrogen) diluted in 1% BSA, 0.25% Triton X-100, PBS, and incubated for 1 h at room temperature and where indicated co-stained with 1:80 Texas Red-X phalloidin (catalog no. T7471, Invitrogen). Sections were finally visualized using a Zeiss LSM700 confocal microscope.
Antibody Inhibition of Lens Epithelial Cell Phagocytosis-For antibody inhibition studies, dual fluorescently labeled SRA 01/04 or primary chicken lens epithelial cells were prepared as described above. The pDsRed-overexpressing SRA 01/04 or pDsRed-overexpressing primary chicken lens epithelial cells were pre-incubated with either ␣V␤5, ␤1-integrin (catalog no. CSAT, Developmental Studies Hybridoma Bank), or ␣V␤3 antibody (catalog no. ab78289, Abcam) diluted in media for 30 min; after 30 min, the cells were washed in PBS, and fresh antibody was added along with 100 l of GFP-labeled apoptotic cell debris. All antibodies were added at a concentration of 3 g/ml.
For quantitation of phagocytosis in the presence of inhibiting antibodies or following UV light exposure (described below), non-transfected SRA 01/04 cells were plated at 100,000 cells in glass bottom dishes as described previously. Following challenge with GFP-labeled apoptotic debris, cells were washed three times in PBS containing calcium and magnesium (PBS-CM), and the extracellular fluorescence was quenched using 0.2% trypan blue solution (catalog no. T8154, Sigma) diluted in PBS-CM for 10 min at room temperature as described previously (46). Cells were then fixed and stained with ␣-tubulin antibody as described above. Cells were visualized using a Zeiss LSM700 confocal microscope; at least 100 cells per treatment were imaged, and the number of cells containing GFP-labeled apoptotic cell debris was counted and expressed as % phagocytic cells. Differences among treatments were determined using Student's t test assuming equal variance where p Ͻ 0.01 was considered statistically significant.
Exposure of Lens Epithelial Cells to UVA-SRA 01/04 cells were seeded onto 60-mm 2 tissue culture dishes at 500,000 cells or 150,000 cells on 35-mm glass bottom dishes in complete media and exposed to 6.3 or 9.5 J UVA using a UVA LED array ( 365 nm , High-Power UV LED Tool, Seraph Robotics). The array was centered 10 cm above the plate, and cells were exposed in phenol-free, serum-free DMEM (catalog no. 11054-020, Invitrogen). Following UV treatment of 60-mm dishes, total protein was prepared in Nonidet P-40 lysis buffer supplemented with a protease inhibitor mixture (catalog no. P8340, Sigma). The effect of UV light exposure on the ␣V␤5 receptor was analyzed using SDS-PAGE and Western blotting, as described previously (47). The antibodies used are as follows: ITGAV (catalog no. SC-6617-R, Santa Cruz Biotechnology); ITGB5 (catalog no. SC-14010, Santa Cruz Biotechnology); and GAPDH (catalog no. SC-25778, Santa Cruz Biotechnology). The secondary antibody used was fluorescently conjugated anti-rabbit secondary antibody (1:5000, catalog no. A21428, Dylight-800 IgG, Thermo Scientific). Western blots were visualized using a LI-COR Odyssey (LI-COR Biosciences, Lincoln, NE).

Results
Lens Epithelial Cells Phagocytize Multiple Substrates-To evaluate whether lens cells were capable of phagocytosis, we first examined the ability of human SRA 01/04 lens epithelial cells to internalize fluorescently labeled polystyrene beads over a period of 16 h (Fig. 1A). Because the polystyrene beads are a non-degradable substrate, they allowed the tracking and visu-alization of their subcellular accumulation over time. Confocal imaging of the cells using ␣-tubulin (red) as a counterstain revealed significant time-dependent uptake of the beads (greenish/white) by the cells (Fig. 1A). By 16 h, a large numbers of beads accumulate at the perinuclear region of the cells. The perinuclear clustering of the internalized beads suggests maturation of the phagosome and transport to lysosomes, because the beads are a non-degradable material; here, they accumulate at the point of fusion with lysosomes (Fig. 1A). Orthogonal images obtained by confocal Z-stacking were used to confirm that substrates were aligned in the same plane as the cells as confirmation that they are internalized and not merely externally interacting with the cells. In addition to the beads, the cells also were capable of phagocytosis of fluorescein-labeled (green) bacterial particles demonstrating that phagocytosis by the cells is not limited to the beads (Fig. 1B). To rule out the possibility that the ability to phagocytize was limited to the transformed SRA 01/04 human lens cells, the ability of primary embryonic chicken lens epithelial cells to phagocytize beads (Fig. 1C) and the fluorescein-labeled (green) bacterial particles (Fig. 1D) was also examined. As shown in Fig. 1, C and D, the primary lens cells were capable of phagocytizing both substrates. To ensure that phagocytosis was not restricted to isolated lens epithelial cells grown in the presence of serum, we also examined the ability of explanted portions of central embryonic chicken lens epithelial tissue attached to the lens capsule and cultured in the absence of serum to phagocytize the beads (Fig. 1E). As detected for the human SRA 01/04 cells and the primary chicken lens cells, phagocytosis of beads by lens epithelial cells comprising the excised portions of the chicken lens epithelium was detected under serumfree conditions. Finally, to further confirm that the beads were internalized in the cells, primary chicken lens epithelial cells were incubated with beads for 16 h and subsequently examined by transmission electron microscopy ( Fig. 1, F-H). Smooth rounded beads are visible within the cells (Fig. 1, F and G), whereas extracellular beads have a crenulated appearance, compare Fig. 1, F and G (internal spheres), with Fig. 1H (external spheres).
We acknowledge that the fluorescently labeled beads and the fluorescently labeled attenuated bacteria examined here are not necessarily normal substrates of the lens in vivo. The beads were chosen as a non-degradable substrate easily tracked and detected by confocal and transmission electron microscopy. The bacterial aggregates were employed because they are a degradable substrate in contrast to the beads and mimic apoptotic cellular material by having surface-exposed phosphatidylserine that is a classic "eat-me" signal also required for phagocytosis of apoptotic debris (48). The data obtained using these substrates are likely to be reflective of the actual phagocytic function of lens epithelial cells because these are well characterized substrates for phagocytosis in other studies (49,50).
Lens Epithelial Cells Phagocytize Exogenously Added Apoptotic Lens Epithelial Cell Debris-Because labeled polystyrene beads or labeled bacterial particles are not likely normal substrates for lens epithelial cell phagocytosis, phagocytosis of apoptotic lens cell debris was examined in human SRA 01/04 ( Fig.  2A), primary embryonic chicken lens epithelial cells (Fig. 2B), and ex vivo cultured E13 chicken lenses (Fig. 2C). SRA 01/04 or primary chicken lens cells were transiently transfected with a CMV promoter-driven green fluorescent protein (GFP) or red fluorescent protein (pDsRed) expression plasmids. Green (GFP-overexpressing) SRA 01/04 or green (GFP-overexpressing) primary chicken lens cells were subjected to staurosporine treatment to induce apoptosis, and the resulting cells and debris were harvested in PBS buffer and hereafter referred to as GFP-labeled apoptotic cell debris. GFP-labeled apoptotic cell debris was added to red (pDsRed overexpressing) SRA 01/04 cells, red (pDsRed overexpressing) primary chicken lens cells, or intact ex vivo E13 chicken cultured lenses for the indicated times. Phagocytosis by SRA 01/04 cells and primary chicken lens cells was determined by localization of GFP-labeled apoptotic cell debris in the red (pDsRed-overexpressing) cells. Phagocytosis by ex vivo cultured lenses was determined by fixing and sectioning ex vivo lenses and localizing GFP-labeled apoptotic cell debris within the boundaries of lens cells counterstained with the nuclear stain DAPI (blue) and F-actin (red) fluorescent staining and using orthogonal confocal fluorescent imaging. Human SRA 01/04 ( Fig. 2A), primary chicken lens epithelial cells (Fig. 2B), and ex vivo cultured lenses (Fig. 2C) were all capable of phagocytizing apoptotic cell debris. Orthog-onal images obtained by Z-stacking confirmed that the GFPlabeled apoptotic cell debris aligned in the same plane as the phagocytizing cells (Fig. 2, A and B, red) consistent with internalization. As a control, SRA 01/04 lens epithelial cells were also challenged with GFP-labeled apoptotic debris and kept on ice to inhibit the active process of phagocytosis. Results showed that although low levels of binding of GFP-labeled debris are apparent, internalization is almost completely blocked (data not shown) suggesting that the phagocytosis examined in this study is an active process that is energy-dependent.
Integrin ␣V␤5 Receptor and Other Members of the ␣V␤5 Phagocytosis Pathway Are Expressed by the Lens-Because lens epithelial cells and intact lenses were shown to phagocytize apoptotic debris, we next sought to investigate whether lens epithelial cells express transcripts that encode proteins that could orchestrate phagocytosis within lens epithelial cells. Data were obtained from mining of high throughput RNA sequencing data obtained using microdissected E13 chicken lens (Fig. 3A) (51). Here, we investigated transcript levels of established phagocytosis receptors and associated members of the ␣V␤5 pathway. Functional clustering and transcript abundance analysis revealed robust transcript levels of the ␣V␤5 subunits integrin ␣V (ITGAV) and integrin ␤5 (ITGB5). Consistent with a poten-  DECEMBER 18, 2015 • VOLUME 290 • NUMBER 51 tial lens function for ␣V␤5, multiple transcripts encoding proteins established to function in the ␣V␤5-mediated phagocytosis pathway were also expressed by the lens, including MFGE8 (milk fat globule-EGF factor) (52), MERTK (MER proto-oncogene tyrosine kinase) (53), and GAS6 (growth arrest-specific 6) (54) (Fig. 3B). We also included in our analysis ITGB3 (integrin ␤3) (55) that was used as a control in our functional analysis of ␣V␤5-mediated phagocytosis. Fluorescent confocal microscopic analysis of E13 chicken lens midsagittal cryosections stained with a previously characterized ␣V␤5-specific antibody (46) revealed expression of ␣V␤5 heterodimeric protein throughout the central and equatorial regions of the lens epithelium (Fig. 3C). ␣V␤5 was also detected, albeit at lower levels, in cortical lens fiber cells and central lens fiber cells.

Lens Phagocytosis of Apoptotic Lens Cell Debris
Integrin ␣V␤5 Co-localizes with Internalized GFP-labeled Apoptotic Lens Cell Debris-Based on the established function for phagocytosis of the ␣V␤5 receptor (46, 55-58) and its detection by RNA sequencing and immunostaining in the lens, we next sought to establish a potential role for the ␣V␤5 receptor in lens cell phagocytosis of apoptotic lens cell debris. GFP-labeled apoptotic cell debris was prepared as described above and added to SRA 01/04 human lens epithelial cells, primary chicken lens epithelial cells, or E13 ex vivo cultured chicken lenses for 4 h. Internalized GFP-labeled apoptotic cell debris and ␣V␤5 were co-localized by immunostaining with ␣V␤5specific antibody. Orthogonal images obtained by Z-stacking (0.96-m slices) demonstrate that internalized ␣V␤5 co-localized with GFP-labeled apoptotic cell debris in human SRA 01/04 lens epithelial cells (Fig. 4A), primary chicken lens epithelial cells (Fig. 4B), and the ex vivo cultured chicken lenses (Fig. 4C) as evidenced by overlapping punctate staining of green (GFP-labeled apoptotic cell debris) and red fluorescence (␣V␤5-specific antibody) appearing as yellow surrounded by ␣V␤5-specific staining. Not all GFP-labeled particles shown in Fig. 4C are associated with ␣V␤5. These particles could have been internalized by other phagocytosis receptors such as ␣V␤3 (1-3), or they could have been internalized by ␣V␤5 that was subsequently recycled or degraded. Consistent with recycling of ␣V␤5, there are also numerous ␣V␤5 (Fig. 2C, red) puncta in the cell that are not associated with the GFP-labeled debris. In SRA 01/04 cells (Fig. 4A) and the ex vivo cultured lens (Fig. 4C), a distinct red band of ␣V␤5 surrounding GFP-labeled apoptotic debris is evident in the zoomed images.
Integrin ␣V␤5 and Internalized GFP-labeled Apoptotic Lens Cell Debris Also Co-localize with the Phagosome Markers RAB7 and RILP in E13 Chicken Lens Epithelial Cells-Because phagocytized and internalized material is transported to lysosomes by a process called phagosome maturation (33-34), we next sought to establish whether ␣V␤5, which is internalized during phagocytosis, co-localizes with the RAB7 GTPase and its effector RILP. RAB7 GTPase is a molecular switch that facilitates phagosome maturation through regulation of fusion events, although RILP functions as its downstream effector required for final step fusion with lysosomes (59). Both molecules are employed in this study as intracellular markers of late phagosome maturation (33,34). Fluorescent confocal microscopic analysis of untreated E13 chicken lens sections stained with RAB7-or RILP-specific antibodies revealed the expression of both RAB7 (Fig. 5A) and RILP (Fig. 5B) throughout the central and equatorial regions of the lens epithelium. RAB7 and RILP were also detected in cortical lens fiber cells. Importantly, RAB7 (Fig. 5C) and RILP (Fig. 5D) co-localize with integrin ␣V␤5 suggesting that members of the phagosome maturation path- way participate in post-internalization processing of material phagocytized by ␣V␤5.
Because RAB7 and RILP have overlapping functions for phagosome maturation and for endolysosomal trafficking, we next determined whether both molecules co-localized with the internalized GFP-labeled debris in the ex vivo cultured lenses to determine whether phagosome maturation is involved following ␣V␤5-mediated phagocytosis of apoptotic material demonstrated in Fig. 2C. We stained mid-sagittal lens sections from lenses previously shown to phagocytize GFP-labeled apoptotic cell debris with either RAB7 or RILP. The internalized GFP-labeled apoptotic cell debris co-localized with RAB7 (red) (Fig.  5E) and also with RILP (red) (Fig. 5F) as indicated by yellow puncta in the merged image (Fig. 5, E and F) suggesting the involvement of phagosome maturation for degradation following phagocytosis of the apoptotic debris.
Integrin ␣V␤5 Is Required for Phagocytosis of Apoptotic Cell Debris by Lens Epithelial Cells-Having demonstrated that the ␣V␤5 receptor co-localizes with internalized GFP-labeled apoptotic cell debris in both cultured lens epithelial cells and ex vivo cultured chick lenses, we next sought to determine the requirement for the ␣V␤5 receptor in the phagocytosis of GFP-labeled FIGURE 4. Co-localization of GFP-labeled apoptotic lens cell debris with integrin ␣V␤5 in phagocytosing SRA 01/04, primary chicken lens epithelial cells, and the ex vivo cultured intact chicken lens. A, confocal fluorescent images of cultured human lens epithelial cells (SRA01/04) 4 h following incubation with GFP-labeled apoptotic cell debris (green). Cells are immunolabeled with integrin ␣V␤5 (red) and counterstained with ␣-tubulin (white) and the nuclear stain DAPI (blue). Ortho refers to orthogonal images obtained by Z-stacking (0.96-m intervals) to determine co-planarity of the cells and internalized GFP-labeled debris. Examples of internalized integrin ␣V␤5 co-localized with GFP-labeled apoptotic cell debris are evident as yellow puncta and outlined in the orthogonal image by a dashed circle. The dashed box 2D zoom below shows individual channels in grayscale with the merged image in color. Arrows point to a green GFP-labeled apoptotic cell debris surrounded by a ring of red ␣V␤5. B, confocal fluorescent images of primary chicken lens epithelial cells (PLEC) 4 h following incubation with GFP-labeled apoptotic cell debris (green). Cells are immunolabeled with integrin ␣V␤5 (red) and counterstained with ␣-tubulin (white) and the nuclear stain DAPI (blue). In the 2D zoom image, individual channels are shown in grayscale and the merged image in color; arrows point to integrin ␣V␤5 (red) puncta that co-localize with internalized (green) GFP-labeled apoptotic debris (yellow puncta in merged image). C, confocal fluorescent images of midsagittal lens sections prepared from ex vivo cultured chicken lenses challenged with GFP-labeled apoptotic cell debris for 4 h. Sections are immunolabeled with integrin ␣V␤5 (red) and counterstained with the nuclear stain DAPI (blue), individual channels are shown in grayscale with the merged image in color. The dashed boxes and circles all represent areas of co-localization between integrin ␣V␤5 and GFP-labeled apoptotic cell debris (yellow puncta). Co-planarity of ␣V␤5 and GFP-labeled debris is indicated by a dashed white circle in the orthogonal projection. 2D indicates that the confocal image shown is from a single plane of focus. Scale bars are 20 m where indicated. DECEMBER 18, 2015 • VOLUME 290 • NUMBER 51 apoptotic cell debris. Human SRA 01/04 epithelial cells and primary chicken lens epithelial cells were pre-incubated with a function-blocking antibody specific for ␣V␤5, and the ability of the antibody-treated cells to internalize GFP-labeled apoptotic cell debris was monitored. Pre-treatment with the ␣V␤5 antibody markedly reduced the internalization of apoptotic cell debris in both human SRA 01/04 epithelial cells and primary chicken lens epithelial cells (Fig. 6, A and B) consistent with a requirement for ␣V␤5 for phagocytosis of apoptotic cell debris by lens epithelial cells. To provide quantitation of the observed inhibition of phagocytosis by the lens epithelial cells, nontransfected human SRA 01/04 epithelial cells were pretreated FIGURE 5. In vivo expression of late endosome trafficking markers RAB7 and RILP and their co-localization with integrin ␣V␤5 in vivo and GFP-labeled apoptotic cell debris ex vivo. A and B, confocal fluorescent images demonstrating RAB7 (green) (A) and RILP (green) (B) protein expression throughout E13 chicken lenses by immunolabeling midsagittal E13 lens sections (20 m) with RAB7-or RILP-specific antibodies and counterstaining with the nuclear stain DAPI (blue) and F-actin (white). C and D, confocal fluorescent images immunostained with RAB7 (red) (C) or RILP (red) and ␣V␤5 (green) (D) and counterstained with the nuclear stain DAPI (blue). Co-localization of RAB7 or RILP with integrin ␣V␤5 (yellow puncta on merged image) is highlighted using dashed boxes and circles containing zoomed areas of the merged image. Ortho refers to orthogonal images obtained by Z-stacking (0.96-m intervals) to determine co-planarity of the cells and internalized RAB7/RILP and ␣V␤5. E and F, confocal fluorescent images of midsagittal lens sections (20 m) prepared from ex vivo cultured chicken lenses that were challenged with GFP-labeled apoptotic cell debris (green) for 4 h. Sections were immunolabeled with RAB7 (red) (E) or RILP (red) (F) and counterstained with the nuclear stain DAPI (blue). Individual channels for C-F are shown in grayscale with the merged image in color. Areas of co-localization between RAB7 or RILP with internalized GFP-labeled apoptotic cell debris are highlighted using dashed boxes and circles. The dashed box on the 2D image and the dashed circle on the orthogonal image represent the same area of the original en face image. Co-planarity of co-localized structures between RAB7 or RILP and internalized GFP-labeled apoptotic cell debris is indicated by a dashed white circle in the orthogonal projection below. 2D indicates that the confocal image shown is from a single plane of focus. Scale bars are 20 m where indicated.

Lens Phagocytosis of Apoptotic Lens Cell Debris
with equal amounts of the function-blocking ␣V␤5 antibody, the function-blocking ␤1-integrin antibody, or ␣V␤3 antibodies as controls. Pretreated cells were then challenged with GFP-labeled apoptotic cell debris. Extracellular fluorescence was quenched using trypan blue, and the cells were stained with ␣-tubulin. At least 100 cells per treatment were imaged, and the FIGURE 6. Immuno-inhibition of integrin ␣V␤5 blocks internalization of apoptotic cell debris by lens epithelial cells. A and B, confocal fluorescent images of ␣-tubulin (red) immunolabeled SRA 01/04 human lens epithelial cells (A) and pDsRed-overexpressing (red) primary chicken lens epithelial cells (PLEC) (B) counterstained with the nuclear stain DAPI (blue) following a 4-h incubation with GFP-labeled apoptotic cell debris (green) in the presence or absence of a function-blocking integrin ␣V␤5 antibody. Ortho refers to orthogonal images obtained by Z-stacking (0.96-m intervals) to determine co-planarity of the cells (red) and internalized GFP-labeled apoptotic cell debris (green). Scale bars are 20 m where indicated. C and D, representative graphs showing quantitation of inhibition of phagocytosis using the integrin ␣V␤5 antibody. SRA 01/04 were pre-incubated for 30 min with function-blocking integrin ␣V␤5 (C and D), ␤1-integrin (C), or ␣V␤3 antibodies (D) before a 4-h challenge with GFP-labeled apoptotic cell debris. At least 100 cells were imaged per treatment, and the number of cells with internalized GFP-labeled apoptotic cell debris were counted and expressed as % phagocytic cells. Differences among treatments were determined using Student's t test assuming equal variance where p Ͻ 0.01 was considered statistically significant. *, p Ͻ 0.001 compared with control untreated cells. DECEMBER 18, 2015 • VOLUME 290 • NUMBER 51 number of cells with internalized GFP-labeled apoptotic cell debris was counted and expressed as % cells containing apoptotic debris. Pretreatment of the cells with the function-blocking ␣V␤5 antibody resulted in a 40% (Fig. 6C) or 50% decrease in phagocytosis (p Ͻ 0.001) (Fig. 6D) relative to control cells. By contrast, pretreatment with either a ␤1-integrin (Fig. 6C) or an ␣V␤3 antibody (Fig. 6D) had no effect on phagocytosis.

Lens Phagocytosis of Apoptotic Lens Cell Debris
UVA Treatment of Lens Epithelial Cells Results in Loss of Both ␣V␤5 Subunits ITGAV and ITGB5 and Reduced Ability of Lens Epithelial Cells to Phagocytize Apoptotic Cell Debris-Because UV light has been shown to cause apoptosis in lens cells and in whole lenses (21,22,(27)(28)(29) and because UV light exposure is associated with cataract formation (18 -20), we sought to determine the effect of UVA light on the subunits of the integrin ␣V␤5 receptor and on the process of lens cell phagocytosis of apoptotic debris. To test this, human SRA 01/04 lens epithelial cells were exposed to UVA light using a UVA LED array. Cells were exposed to a total of 6.3 or 9.5 J of UVA light as these doses were previously reported to have significant effects on lens epithelial cell morphology, apoptosis, and viability (60). Western blot analysis of total protein extracts from UVA-exposed SRA 01/04 cells using specific antibodies revealed dramatically reduced levels of full-length native ITGAV as well as formation of higher molecular weight aggregates; ITGB5 levels also dropped markedly with no aggregate formation. GAPDH is shown as an internal standard (Fig. 7A). Phagocytosis of GFPlabeled apoptotic cell debris following identical UVA treatment was quantified as described previously. The number of cells capable of phagocytosis of apoptotic debris following UVA exposure was significantly reduced relative to control untreated cells with a 32% reduction following a 6.3 J exposure (p Ͻ 001) (Fig. 7B) and a 62% reduction following a 9.5 J exposure (p Ͻ 001) (Fig. 7B). UV light treatment also resulted in significant changes in cell morphology accompanied by the appearance of cellular projections (6.3 J) and cellular blebbing (9.5 J) (Fig. 7C).

Discussion
The eye lens is under constant stress in the form of environmental insults ranging from UV light (18 -20) to smoke exposure (61,62). Consistently, the eye lens has evolved multiple protective and repair systems to preserve its transparent function in the face of environmental insults (63). These include abundant anti-oxidant systems (64 -66), protein protective systems (67)(68)(69), protein salvaging systems (70,71), protein repair systems (72)(73)(74), as well autophagy systems for clearance of damaged organelles (41)(42)(43). Loss of the functions of these systems results in a range of lens effects including, but not limited to, accumulation of damaged and modified proteins (75)(76)(77), apoptosis of lens cells (78,79), and cataract formation (73,80). Because these systems are not confined to the eye lens, their study in the lens provides important information for their functions in more complex tissues and disease states.
The majority of protective and repair systems in the eye lens are contained within the light facing single monolayer of cells called the lens epithelium. The lens epithelium covers the organelle-free fiber cells that are essential for lens transparency. The lens epithelium is the first part of the lens exposed to environmental insult. It is essential for the homeostasis of the entire eye lens (11,25,26), and damage to lens epithelial cells is associated with cataract formation (22, 24, 30 -32). The function of the entire lens is dependent on the activities of multiple epithelial cell components, including ion transporters (25,26), protein synthesis (67), mitochondrial functions (14,73,80,81), and lipid biogenesis (82); therefore, maintaining the homeostasis of the epithelium is critical for the entire lens.
A major insult to the lens epithelium is UV light exposure (21)(22)(23)(24). The human cornea absorbs all incident light below 300 nm wavelength and ϳ92% of UVB between 300 and 315 nm (24,83,84). Most of the UV light transmitted to the lens is therefore UVA light; this is primarily absorbed by the anterior lens epithelium, and long term UVA exposure is believed to contribute to lens damage (24,83,84). UV light exposure of lens epithelial cells is associated with apoptosis of lens epithelial cells (21,22,(27)(28)(29) and human cataract formation (18 -20); however, the mechanisms used by lens epithelial cells to deal with the downstream effects of UV light and other insults, i.e. apoptosis and accumulation of apoptotic debris to ensure epithelial cell homeostasis and lens transparency, remain largely unknown. Because UV light and other insults result in apoptosis of lens epithelial cells, we sought a mechanism for the lens epithelium to remove apoptotic cells and debris to ensure lens function and thereby preserve lens transparency.
In many non-lens systems, phagocytosis is a major mechanism to remove and thereby detoxify apoptotic cells and debris (1-4). However, the lens is an encapsulated tissue with no easily discernable external source of macrophages, dendrite-like cells, A, Western analysis of integrin ␣V, integrin ␤5, and GAPDH protein levels following UVA exposure (6.3 J compared with 9.5 J). B and C, SRA 01/04 cells were treated with UVA and challenged with GFPlabeled apoptotic cell debris for 4 h, and phagocytosis was quantified (B). At least 100 cells were imaged per treatment; the number of cells with internalized GFPlabeled apoptotic cell debris were counted and expressed as % phagocytic cells. Differences among treatments were determined using Student's t test assuming equal variance where p Ͻ 0.01 was considered statistically significant. *, p Ͻ 0.001 compared with control untreated cells. Representative confocal images of SRA 01/04 cells following UVA treatment and challenge with GFP-labeled apoptotic cell debris (green) are shown below. C, cells were counterstained with the nuclear stain DAPI (blue) and ␣-tubulin (red).
or other specialized phagocytic cells that could remove apoptotic lens epithelial cell debris and thereby prevent secondary necrosis that arises in epithelial cell populations containing latent apoptotic debris (1)(2)(3)(4). We therefore hypothesized that lens epithelial cells themselves could function as endogenous phagocytes of the lens that could remove apoptotic debris generated by lens exposure to UV light and other environmental insults. Consistently, other investigators have detected apoptosis-like material contained within lens epithelial cell vesicles in isolated lens cells treated with UV light (40) and in the lenses of rodents treated with UV light (21,30).
To test the hypothesis that lens epithelial cells are capable of phagocytosis, and to provide a mechanism for possible phagocytosis by lens epithelial cells, we verified that cultured human lens epithelial cells, primary chick lens epithelial cells, and ex vivo cultured chicken lenses are capable of phagocytizing fluorescently labeled and exogenously added apoptotic debris prepared from lens epithelial cells (Figs. 1 and 2). We went on to data mine our RNA sequencing analysis of the chicken lens (51) to identify that the members of the integrin ␣V␤5 phagocytosis pathway are robustly expressed by the lens (Fig. 3). We also found that ␣V␤5 co-localizes with internalized labeled apoptotic debris and the phagosome maturation markers RAB7 and RILP within lens epithelial cells (Figs. 4 and 5). We also found that pre-treatment of cells with an ␣V␤5 function-blocking antibody significantly reduces phagocytosis by lens epithelial cells (Fig. 7) consistent with the established function of ␣V␤5 for phagocytosis in other tissues, including the retinal pigmented epithelium (46,(55)(56)(57) and trabecular meshwork (85). Finally, we demonstrate that exposure of lens epithelial cells to UVA light results in degradation of ␣V␤5 and reduced lens epithelial cell phagocytosis (Fig. 7). We believe that the data shown in Fig. 7 demonstrating degradation of ␣V␤5 concomitant with decreased phagocytosis suggest that ␣V␤5 in lens epithelial cells is effective at removing apoptotic bodies in the intact lens up to a certain level of UV exposure, and only after an exposure threshold is reached is phagocytosis blocked by the loss of ␣V␤5. Further studies will be required to establish the relationship between UV light exposure, ␣V␤5 degradation, and lens phagocytosis.
Our results suggest a novel phagocytosis function for the lens that serves to remove toxic apoptotic debris from the lens epithelium. They also provide evidence that lens phagocytosis is mediated by members of the ␣V␤5 pathway. Because UV light causes apoptosis of lens epithelial cells and cataract formation, our results suggest that damage to ␣V␤5 and loss of lens phagocytosis of apoptotic cell debris could contribute to UV lightinduced cataract formation. Because we demonstrated that the intact lens is capable of phagocytizing exogenously added debris, our results suggest that the lens could also phagocytize apoptotic cell and/or other debris in the aqueous humor that could be generated by other eye tissues, including the cornea.
Although our results demonstrate that cells of the intact lens are capable of phagocytosis using an ␣V␤5-dependent phagocytosis pathway, further studies will be required to establish a direct cause and effect relationship between apoptosis of lens epithelial cells, phagocytosis by the lens, UV light exposure, and cataract formation. Although we provide evidence that ␣V␤5 is required for lens cell phagocytosis, we cannot rule out that other phagocytic pathways and receptors also function to coordinate phagocytosis by lens cells. Indeed our RNA sequencing analysis also indicated that the lens expresses the scavenger receptors CD36 (86), SCARB1 (87), and others (data not shown).
Our identification of ␣V␤5 as a phagocytosis receptor for the lens is a new lens function for this important integrin. We are confident that ␣V␤5 plays a similar role in the lenses of multiple species because both human and chicken lens epithelial cells were examined in our study, and previous studies have identified ␣V␤5 to be expressed in the human (46,56,88), rat (46), and mouse (55). Indeed, the ␣V and ␤5 subunits that make up heterodimeric ␣V␤5 are conserved among vertebrate phyla, and their respective protein sequences for either subunit have greater than 75% protein sequence homology between mouse, chicken, and human species sequences (NCBI, Clustal Omega, data not shown).
␣V and ␤5 integrin subunit levels in lens epithelial cells have been reported to be regulated by TGF␤ (89). Because TGF␤ is also associated with the differentiation of lens epithelial cells, it is likely that ␣V␤5 has multiple roles in the lens, including cellular adhesion through binding to vitronectin (90). Other investigators have examined the phenotype of mouse lenses deleted for the ␣V subunit of ␣V␤5, and they found no apparent developmental defects (91). This result is consistent with ␣V␤5 having a protective and maintenance role in the adult lens through its phagocytosis function. Many other studies have identified that lens protective and repair systems do not provide a developmental phenotype but do result in cataract formation upon lens exposure to environmental stress (73,92).
In summary, we demonstrate that lens epithelial cells are capable of phagocytizing autologous apoptotic cell debris through an integrin ␣V␤5-mediated pathway. These results suggest phagocytosis of apoptotic material is an important function for the lens that is likely required for the preservation of lens homeostasis and transparency. The results also show that exposure of the lens to UV light results in damage to ␣V␤5 and reduced lens phagocytosis. This could result in accumulation of toxic apoptotic cell debris that could contribute to UV light-induced cataract formation. These eye lens data provide insight into those mechanisms that could be important for the maintenance of other tissues upon environmental insult.
Author Contributions-M. K. conceived of and coordinated the study and wrote the paper. D. C. and L. A. B. designed, performed, and analyzed the experiments in Figs. 1-7. O. B. designed, performed, and analyzed the experiment shown in Fig. 4. D. C. and L. A. B. assisted in writing the manuscript, and all authors reviewed the results and final manuscript and approved it for submission.