Differential Expression of N- and B-Cadherin during Lens Development

Abstract

purpose. To analyze the dynamics of N- and B-cadherin cell adhesion molecule expression and cytoskeletal interaction during embryonic chick lens development.

methods. Localization of N- and B-cadherin, F-actin, and connexin 56 were determined by immunohistochemistry of developing lenses or immunocytochemistry of differentiating primary lens cultures. Biochemical analysis of cytoskeletal linkage of N- or B-cadherin was assessed by differential detergent extraction, electrophoresis, and immunoblotting.

results. The results indicate that although both cadherins are expressed throughout lens development, N-cadherin expression detected was similar in both lens epithelial and fiber cells, whereas B-cadherin was preferentially localized to the lens fiber cells. During differentiation, both cadherins become increasingly associated with the lens cytoskeleton, as indicated biochemically by a transition from largely Triton X-100–soluble to Triton X-100–insoluble pools and immunocytologically by cadherin localization to cell–cell borders and colocalization with the actin cytoskeleton. Although a significant fraction of N-cadherin remains Triton X-100–soluble as the lens cells differentiate, B-cadherin becomes resistant to extraction by both Triton X-100 as well as RIPA buffers. As detected immunocytochemically in lens cell cultures, the temporal localization of N-cadherin to cell–cell interfaces precedes that of B-cadherin. Furthermore, temporal localization of B-cadherin, as opposed to N-cadherin, to cell–cell borders more closely parallels that of connexin 56 in vitro as well as in vivo.

conclusions. These results suggest that while both N- and B-cadherin are expressed during lens cell differentiation, both their patterns of expression as well as their cytoskeletal association differ between epithelial and fiber cells.

During development of the ocular lens, epithelial cells in a predetermined region of the head ectoderm differentiate to form the lens placode, an ectodermal thickening that then invaginates to form the lens pit and then the lens vesicle. ^{1 2} The lens vesicle then becomes partitioned into an anterior portion that retains an epithelial phenotype and a posterior portion that undergoes differentiation to form the primary fibers. During subsequent development and continuing throughout adulthood, the anterior epithelium provides a continuous source of undifferentiated cells that migrate toward the equator of the lens where the cells initiate differentiation into secondary fiber cells that form concentric layers overlying the core of primary fiber cells. This morphologic differentiation of epithelial cells into elongated fiber cells is accompanied by biochemical changes, in both primary and secondary fibers, including the synthesis of lens differentiation markers such as the crystallins, connexins, and filensin. ^{2 3 4 5} Because of its unique pattern of growth, the periphery or cortex of the embryonic lens consists of newly differentiating fiber cells, whereas the core or center contains terminally differentiating or differentiated fiber cells. Although terminally differentiated fiber cells normally may persist for the life of the organism, failure to establish or maintain proper lens fiber organization and biochemical differentiation may be a cause of congenital and age-onset cataracts, a major cause of visual impairment. The dynamic changes in cellular relationships during the course of lens development and the maintenance of proper intercellular relationships in the mature lens are likely to be dependent on cell–cell interactions mediated by cell adhesion molecules (CAMs). 

One major family of CAMs that is likely to play a critical role in lens development is the cadherins, Ca²⁺-dependent, cell–cell adhesion molecules that have been shown to mediate tissue morphogenesis. ^{6 7 8} Cadherins mediate their homophilic cell–cell adhesions via highly conserved extracellular and cytoplasmic domains. ^{6 7 9} The extracellular portion of cadherins is responsible for mediating specific homophilic interactions, whereas the intracellular domain is required for cadherin binding to the cytoplasmic proteins β- or γ-catenin (plakoglobin). This complex is in turn linked to the actin cytoskeleton via α-catenin andα -actinin, an event required for cadherin participation in adherens junction formation. ^{10 11 12 13 14} However, cadherins are also found in cytosolic and membrane pools, ^{10 11} where they may sequester β-catenin and plakoglobin, which can also function independently of cadherins. ^{9 15 16 17} Other than acting as a regulator of catenin levels, it is unknown what functions cadherins might mediate when they are unlinked to the cytoskeleton. ⁹ Linkage to the cytoskeleton is required for stable cadherin–cadherin binding, ¹⁸ and mutational studies with cadherins that lack the cytoplasmic domain and are incapable of binding catenin clearly indicate the importance of cadherin binding to catenins and of linkage to the cytoskeleton for certain cadherin functions. ^{19 20 21}  

In lens development, cadherins have been best characterized in the chicken, where the cadherins that have been identified are N-cadherin, B-cadherin, and L-CAM. ^{22 23 24 25 26} The cadherin names refer to N for neural cadherin, B for brain cadherin, and L for liver CAM, where these proteins were originally identified. Although L-CAM is expressed in the embryonic ectoderm and lens placode, it is downregulated once the lens vesicle separates from the overlying ectoderm. ^{25 26} B-cadherin, as indicated by an initial study of its expression during chick development, is expressed in the lens placode and lens vesicle and after primary fiber formation becomes lost from lens epithelial cells but is retained in lens fiber cells. ²⁵ However, the detailed developmental analysis of B-cadherin in the lens or its association with the lens cytoskeleton has not been previously examined. In contrast, N-cadherin has been more extensively studied in the lens, where it is found in association with both lens epithelial cells and fiber cells, ^{22 23 24} suggesting that it might play an important role throughout lens development. Furthermore, N-cadherin has been implicated in the formation of lens epithelial cell monolayers in vitro and in the gap junction–mediated dye transfer between lentoid cells. ^{14 27 28} Taken together, these data would support the hypothesis that cadherins play an important role in lens development, but the distinct functions of the different cadherins expressed in the lens remains to be determined. To begin to address this issue, we undertook an analysis of N- and B-cadherin expression and cytoskeletal interaction during lens differentiation in vivo and in vitro. The results indicate that although both N- and B-cadherin increase their association with the lens cytoskeleton during the course of differentiation, this effect is more pronounced for B-cadherin, whose temporal expression also appears to be more specifically correlated with that of a specific marker of lens cell differentiation, connexin 56. The results indicate that individual cadherins are likely to play distinct roles in lens development. 

Methods

Preparation of Lens Tissues and Primary Lens Cell Cultures

Experiments conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Whole lenses were isolated from 50 fertilized white Leghorn chicken eggs at embryonic day 10 (E10) or E11, trypsinized, and plated on 35-mm dishes or 8-well slides (Falcon Labware, Lincoln Park, NJ) as described, except that 6.67μ g/mL mouse laminin (Gibco-BRL, Grand Island, NY) was used as substrate. ²⁹ After 3 days of culture, the cells were refed and maintained for an additional 5 days on medium 199 (Gibco-BRL) containing 10% fetal calf serum-1% penicillin-streptomycin-actinomycin mixture-1% glutamine (Sigma, St. Louis, MO). 

Microdissection of Chick Embryo Lenses

Lenses from E10 or E11 were used interchangeably for microdissection studies because the cadherin fractionation patterns were similar in these lenses. Whole lenses, isolated from E10 or E11 chick embryos as described above, were microdissected as described in detail previously into anterior central epithelial cells (EC), equatorial epithelium (EQ), peripheral fibers (FP), and central fibers (FC). ³⁰  

Detergent Extraction of Chick Embryo Lenses

Triton-soluble proteins were obtained by immediately extracting the lens fractions for 1 to 1.5 hours with 100 μl of ice cold 10 mM imidazole-100 mM NaCl-1 mM MgCl₂-5 mM Na₂EDTA-1% Triton X-100, pH 7.4, containing 50μ g/mL aprotinin (Sigma), 25 μg/mL soybean trypsin inhibitor (Sigma), 100 μM benzamidine (Fisher Scientific, Fairlawn, NJ), 5μ g/mL leupeptin (Calbiochem, La Jolla, CA), and 0.5 mM phenylmethylsulfonyl fluoride (Calbiochem). The Triton-soluble and -insoluble fractions were separated by centrifugation at 12,000g for 10 minutes. The resulting Triton-insoluble pellet was briefly rinsed with 100 μl Triton extraction buffer, repelleted by centrifugation at 12,000g for 10 minutes, and was subsequently reextracted for an additional 15 minutes with 100 μl ice cold 50 mM Tris-150 mM NaCl-5 mM Na₂EDTA-1% Triton X-100– 0.1% sodium deoxycholate-0.1% SDS, pH 8 (RIPA buffer), containing the above protease inhibitor cocktail. The RIPA-insoluble and -soluble fractions were separated by centrifugation at 12,000g for 10 minutes. The resulting RIPA-insoluble pellet was rinsed with 100 μl RIPA buffer as described above and solublilized in modified Laemmli sample buffer without bromphenol blue or β-mercaptoethanol. 

Analysis of Lens Proteins by SDS-PAGE and Western Immunoblotting

Triton-soluble, RIPA-soluble, and RIPA-insoluble proteins were brought to 1× modified Laemmli sample buffer, and protein contents of these lysates were determined by BCA (Pierce, Rockford, IL). Lysates were then prepared for reduced or nonreduced SDS-PAGE for N- or B-cadherin analysis, respectively, followed by electrotransfer and Western blotting. To compare the relative proportions of detergent-soluble and -insoluble cadherins within each region of the microdissected lens, detergent-soluble and -insoluble fractions were loaded at equivalent proportions of the original tissue by volume. The Triton-soluble lysates were loaded at equal protein (30 μg). 

For N-cadherin analysis, membranes were blocked for 1 hour with 5% nonfat milk (NFM) in 10 mM Tris-150 mM NaCl (TBS), pH 7.4, and probed with 6B3, an N-cadherin—specific monoclonal antibody (1:2000 in 3% NFM in TBS; Moore, Knudsen, and Grunwald, unpublished results, 1996) for 1 hour. After three washes in TBS containing 0.1% Tween 20 (TBS-Tween), the membranes were incubated with 1:5000 sheep anti-mouse antibodies conjugated to horseradish peroxidase (Amersham, Arlington Heights, IL) for 1 hour, followed by three additional washes in TBS-Tween. N-cadherin was then visualized by chemiluminescence using ECL + substrate (Amersham) and luminography followed by densitometric analysis for cadherins using a one-dimensional gel analysis program (Eastman-Kodak, Rochester, NY). For B-cadherin analysis, membranes were processed as described for N-cadherin, except that the B-cadherin–reactive monoclonal antibody 6D5 ^{31 32} (1:1000 in 3% NFM) was used instead of 6B3, and the blocking and primary antibody solutions contained 0.05% Tween 20 and 1 mM CaCl₂ to maximize immunoreactivity of this conformation-sensitive antibody. 

Immunohistochemical and Immunocytochemical Procedures

Chick embryos, heads, or eyes were isolated and processed for frozen sections according to Bronner-Fraser et al. ³³ Briefly, tissues were processed by overnight fixation in methanol at 4°C, rehydration in 10 mM HEPES-150 mM NaCl-3 mM KCl-5.6 mM glucose-1 mM CaCl₂ (HBSG-Ca²⁺) for 1 hour, and sequential equilibration in 0.5% sucrose and 15% sucrose in HBSG-Ca²⁺ for 2 to 4 hours and overnight, respectively. Tissues were embedded and frozen in O.C.T. compound, and 20-μm sections, either parallel or perpendicular to the visual plane, were obtained and processed for immunofluorescence. 

Tissue sections were permeabilized with 0.5% Triton X-100 in HBSG-Ca²⁺ for 15 minutes, washed three times with HBSG-Ca²⁺, and blocked 1 hour with 10% normal goat serum (NGS; Sigma) in HBSG-Ca²⁺. Hybridoma-conditioned medium containing the N-cadherin–specific antibody NCD-2 ³⁴ was used neat with the tissue sections, whereas ascites containing the B-cadherin–reactive 6D5 was used at a 1:1000 dilution. Also, in preliminary experiments, B-cadherin–specific antibody 5A6 ²⁵ produced staining patterns identical to those of the B-cadherin–reactive 6D5. In some cases, tissues were double-stained for connexin 56, using an anti–connexin 56 antiserum, ³⁵ and for B-cadherin. Tissues were incubated 1 hour with cadherin or connexin 56 antibody, washed three times, and exposed to appropriate secondary antibodies. These were 1:150 donkey anti-rat or 1:100 goat anti-mouse antibodies conjugated to Lissamine rhodamine sulfonyl chloride (Jackson ImmunoResearch Laboratories, West Grove, PA), 1:500 goat anti-mouse antibodies conjugated to Oregon Green (Molecular Probes, Eugene, OR), and 1:500 goat anti-rabbit antibodies conjugated to Alexa 568 (Molecular Probes). Negative controls consisted of tissue sections from which the primary antibody was omitted. After five additional washes, three in 10% NGS and two in HBSG-Ca²⁺, slides were coverslipped and examined using a Nikon Optiphot microscope (Garden City, NJ) fitted with fluorescence filters. 

After specific days in culture, primary lens cells were washed three times with PBS freshly supplemented with 1 mM CaCl₂ and 1 mM MgCl₂, fixed 10 minutes with 3.7% paraformaldehyde in PBS-1 mM CaCl₂-1 mM MgCl₂, and washed three times more with PBS-1 mM CaCl₂-1 mM MgCl₂ before storage at 4°C. Immunocytochemical staining of primary lens cultures were performed as described above, except that 1:500 NCD-2 and 1:250 6D5 in 10% NGS in PBS-1 mM CaCl₂-1 mM MgCl₂ were used to probe for N- and B-cadherin, respectively. Lens cultures were also stained for chick connexin 56, using 1:500 connexin 56 antiserum, followed by 1:100 dilution of the appropriate rhodamine-conjugated secondary antibody (Jackson ImmunoResearch Laboratories), respectively. In some instances, the cultures were double-stained with either N- or B-cadherin and fluorescein isothiocyanate–conjugated phalloidin (Molecular Probes) to visualize actin filaments or with connexin 56 antiserum, followed by 1:100 fluorescein dichlorotiazine conjugated to goat anti-rabbit antibodies (Molecular Probes). Slides were coverslipped and examined by conventional or confocal microscopy. 

Results

Distinct Localization of N- and B-Cadherin during Lens Differentiation In Vivo

The localization patterns of N- and B-cadherin were determined by immunohistochemical analysis at various stages of lens development. Our observations indicated that the earliest distinction between N- and B-cadherin localization occurs during primary fiber cell differentiation after lens vesicle formation. Compared with control tissue sections, both N- and B-cadherin staining was observed in the E4 chick embryo lens vesicle (Fig. 1) . However, although N-cadherin was clearly detectable in both the epithelial cells as well as the fiber cells (Fig. 1A) , by this stage of development B-cadherin localization had become prominent in the differentiating primary fiber cells and was barely detectable in the anterior epithelium (Fig. 1B) . Distinct localization patterns of N- and B-cadherin in the lens are especially evident when compared with the neighboring ectoderm of the presumptive cornea, where B-cadherin but no N-cadherin was detected (Fig. 1) . The distinct patterns of N- and B-cadherin localization also were evident at subsequent stages of development. Day 10 (E10) of lens development is a dynamic stage that already has represented all the four major stages of lens cell differentiation: undifferentiated anterior epithelial progenitor cells; equatorial epithelial cells, which proliferate and then initiate differentiation; the differentiating peripheral fiber cells; and the terminally differentiated central fiber cells. At this and later stages of development, the difference between N- and B-cadherin localization in the anterior epithelium and fiber cells is especially pronounced (Fig. 2) . The anterior epithelium of the lens is positive for N-cadherin, but no B-cadherin was detected (Fig. 2A versus Fig. 2D ). The strong N- and B-cadherin reactivity in the endothelium and epithelium, respectively, of the cornea, served as internal controls (Figs. 2A 2D) . In contrast to the lens epithelium, there was prominent staining of both peripheral (Figs. 2B 2E) and central (Figs. 2C 2F) fiber cells for both N- (Figs. 2B 2C) and B-cadherin (Figs. 2E 2F) . This distinct localization of N- and B-cadherin expression between undifferentiated anterior epithelial cells and differentiated fiber cells persisted through at least E20 of chick development (data not shown). 

Increased Association of N- and B-Cadherin with the Cytoskeleton during Lens Cell Differentiation In Vivo

As discussed above, association of cadherins with the actin cytoskeleton is of functional significance because this interaction is required for the full function of cadherins and their assembly of adherens junctions. To determine whether changes in cadherin–cytoskeletal associations occurred during the course of lens cell differentiation, E10 chick embryo lenses were microdissected into the four distinct developmental regions described above. Each fraction was then independently analyzed by sequential detergent extraction, SDS-PAGE, and immunoblotting to determine the extent of cadherin–cytoskeleton association as indicated by Triton X-100 detergent insolubility. ^{10 11 36} A two-stage detergent extraction was used, first with Triton X-100 to separate cell extracts into Triton-soluble (unlinked to the actin cytoskeleton) and Triton-insoluble (linked to the actin cytoskeleton) fractions, after which the Triton-insoluble cytoskeletal pellet was reextracted with the more stringent RIPA buffer to further fractionate cytoskeleton-linked proteins into RIPA-soluble and -insoluble fractions. In the central undifferentiated anterior epithelium, N-cadherin was found to be largely Triton-soluble, with very little Triton-insoluble and no RIPA-insoluble N-cadherin detected (Fig. 3) . A small increase in Triton insolubility for N-cadherin was observed in the equatorial fraction. However, when fiber cell fractions were examined, it was found that most of the N-cadherin was now found in the Triton-insoluble but RIPA-soluble fraction, indicating a shift from a cadherin pool that is unlinked to the cytoskeleton to one that is linked to the cytoskeleton. Examination of the partitioning of B-cadherin after sequential extractions of the various lens fractions showed a similar trend toward increasing cytoskeletal association, with two important distinctions (Fig. 3) . First, although a significant minority of the N-cadherin pool remains Triton-soluble even in differentiated fiber cells, only traces of B-cadherin could be detected in these fractions, with essentially all the B-cadherin being Triton-insoluble (Figs. 3A 3B) . Second, the majority of the B-cadherin in fiber cells was not only Triton-insoluble but was also RIPA-insoluble and therefore resistant to even stringent detergent extraction (Fig. 3A) . These results indicate distinct interactions of N- and B-cadherin with the lens cytoskeleton and may reflect the distinct functions each plays in lens cell differentiation. 

Distinct Regulation of N- and B-Cadherin during Lens Cell Differentiation In Vitro

In vitro, primary chick embryo lens cultures differentiate and express δ-crystallin and the fiber cell differentiation–specific protein aquaporin-0 (MIP28), as they form multilayer lentoid structures, providing an excellent model system for analysis of different stages of lens cell differentiation. ^{29 37} Differentiation in culture can be described as a multistage process, first forming an epithelial monolayer of cuboidal packed cells and then initiating formation of lentoid bodies. To determine whether expression of N- and B-cadherins were differentially regulated during this process, we examined cadherin distribution by immunocytochemistry in the early stages of culture before organization of the monolayer, after formation of the hexagonally packed monolayer but preceding lentoid formation and in the final stage when lentoids predominate in the cultures. Furthermore, because differentiating lens cells express the gap junction protein connexin 56, ^{3 4 35} we assessed the relative differentiation state of our cultures by monitoring connexin 56 staining. During early stages of lens epithelial cell monolayer formation, N-cadherin was localized to cell–cell borders (Fig. 4A ), as indicated by colocalization with the cortical actin cytoskeleton (Fig. 4B) . In contrast, at this stage of differentiation in vitro, B-cadherin was not detectable at cell–cell borders (Fig. 4C). 

In later stage cultures, when monolayers of lens epithelial cells had organized, N-cadherin was found to be prominently localized to the cell–cell borders of the hexagonally packed, cuboidal monolayer (Fig. 4E) , and B-cadherin could also now be detected, albeit more weakly than N-cadherin, at regions of cell–cell contact (Fig. 4G) . When the distribution of connexin 56 was examined at this stage in culture, only a small percentage of the N-cadherin–positive cells were found to contain connexin 56 at their cell–cell borders (Figs. 5A 5B 5C) . In contrast, in those regions of the cultures where B-cadherin could be detected at cell–cell borders, we often detected a similar distribution of connexin 56 (Figs. 5D 5E 5F) . In late stage cultures where lentoids predominate, both N- and B-cadherins (Figs. 6A 6D ), as well as connexin 56 (Fig. 6G) , were all detected at cell–cell borders with the cortical actin cytoskeleton (Figs. 6B 6E 6H) . Taken together, these results indicate that localization of B-cadherin to cell–cell borders of differentiating lens epithelial cells in vitro occurs after that of N-cadherin and that the distribution of B-cadherin is more temporally correlated with that of the lens differentiation marker connexin 56. 

To confirm the temporal relationship between B-cadherin and connexin 56, sections of E4 chick embryo lens, in which primary fiber formation has been initiated, were double-stained for these proteins. Both B-cadherin and connexin 56 were clearly detected in elongating primary fiber cells, whereas little, if any, staining was detected for either of these proteins in the undifferentiated epithelial cells (Fig. 7) . Thus, the localization of B-cadherin and connexin 56 are temporally linked during lens differentiation in vivo as well as in vitro. 

Discussion

Although different cadherins are highly homologous in both structure and function to each other, the distinct patterns of cadherin expression observed during embryonic development suggest that subtle differences must exist in the function of each type of cadherin. The results reported here indicate that of the cadherins expressed in the chick embryo lens, B-cadherin is preferentially expressed in differentiated fiber cells and is also more resistant to stringent detergent extraction. In contrast, N-cadherin is more ubiquitously distributed within the lens and is localized to both undifferentiated epithelial cells as well as differentiated fiber cells. These results confirm and extend those previously reported for these two cadherins in the lens. ^{22 24 25} Among the specific functions that have been attributed to cadherins is a structural role in the assembly of adherens junctions. ^{13 14 38} Lens fiber cells are joined together by very complex junctional assemblies that occupy much of the plasma membrane, and our results suggest that both N- and B-cadherin become increasingly incorporated into these junctional complexes, as indicated by the shift from Triton-soluble to -insoluble fractions. The linkage of proteins with the cytoskeleton has been associated with relative resistance to nonionic detergent extraction, such as with the Triton X-100 extraction used here. ³⁶ These cytoskeleton-linked proteins, although Triton-insoluble, may often be solubilized with a more stringent detergent treatment such as RIPA buffer. ^{39 40} In fact, one study examining the detergent extraction of integrin subunits has shown that these receptors can be fractionated into RIPA-soluble and -insoluble pools and that this partitioning can be modulated by procedures that will affect integrin phosphorylation and, potentially, function. ⁴¹ Interestingly, B-cadherin appears to be particularly resistant to RIPA extraction (in comparison with N-cadherin) in differentiating lens fibers. Although this RIPA insolubility could reflect a tighter association of B- than N-cadherin with a similar cytoskeletal component, an alternative explanation may be that B- and N-cadherin are in part associated with distinct cytoskeletal elements. A third possibility is the existence of a linkage of B-cadherin to an as yet unidentified component of lens cells. Additional studies systematically characterizing this RIPA-insoluble fraction would be required to further understand the functional implications of this detergent resistance for B-cadherin. 

The extreme detergent insolubility that we observed for B-cadherin, relative to N-cadherin, suggests a distinct function for B-cadherin in the lens. This finding, coupled with the retention of N-cadherin expression in the undifferentiated epithelium, suggests a role in earlier phases of lens development such as those undergone by epithelial cells, which include posterior migration and the initiation of differentiation. Indeed, consistent with this hypothesis, our previous studies have demonstrated that blockade of N-cadherin function with the inhibitory antibody NCD-2 prevents the formation of packed cuboidal monolayers in vitro. ²⁷ Thus, lens cells may require the continuous expression of N-cadherin to maintain signaling and/or structural activities in both epithelial and fiber cells, whereas B-cadherin may be more specifically required for signaling and/or structural adhesive interactions necessary for continued progression to a more highly differentiated state. 

The mature lens contains an extensive system of gap junctions. In light of the previously demonstrated role of E-cadherin in gap junction assembly in a variety of epithelial cell types, ^{42 43 44} one additional function of cadherins during lens differentiation may be in the targeting of gap junction proteins and the assembly of specific gap junctions. Indeed in one in vitro study, treatment of preformed chick lentoids with function blocking anti–N-cadherin antibodies prevented gap junction–mediated dye transfer between cells. ²⁸ Likewise, our findings that the timing of B-cadherin, rather than N-cadherin, most closely parallels the similar timing of connexin 56 localization, both in vitro as well as in vivo, suggests a relationship might exist between these proteins. However, further experiments will be required to delineate the precise relationship between B-cadherin and connexin 56. 

Our results are consistent with the recent observation that connexin 56 does not appear to colocalize with N-cadherin in the lens, as determined by immunoelectron microscopy. ⁴⁵ Although we did not examine cadherin distribution at this high a level of resolution, our results do indicate that in lens fiber cells in vivo as well as in lentoids in vitro, both N- and B-cadherin ultimately become concentrated at cell–cell borders in proximity to the cortical actin cytoskeleton. Clearly, our results demonstrate that the associations of B- and N-cadherin with the lens cytoskeleton are distinct; however, further studies will be required to determine the mechanism by which this occurs and the functional consequences of this difference. 

 Supported by National Institutes of Health Grant R01EY10695 (GBG, ASM) and NIH National Research Service Award Fellowship F3206787 (LL).

 Submitted for publication December 7, 1999; revised April 14, 2000; accepted May 10, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

Immunohistochemical analysis of N- and B-cadherin localization in embryonic day 4 (E4) lenses. E4 chick embryo heads were fixed, processed, and immunostained for N-cadherin (A), B-cadherin (B), or with secondary antibody alone (C) as described in Methods. Note localization of N-cadherin in both epithelial and primary fiber cells, although B-cadherin was only detected in the differentiating cells within the lens. E, lens epithelial cells; F, primary fiber cells; C, cornea. Bar, 50 μm.

Figure 2.

 

Immunohistochemical analysis of N- and B- cadherin in E10 lenses. Lenses were isolated from E10 chick embryos and were fixed, processed, and immunostained for N- (A) through (C) or B-cadherin (D) through (F) as described in Methods. The panels represent staining for cadherins in the cornea and the lens anterior epithelium (A, E) and in cross sections of peripheral lens fibers (B) and (E) and of more centrally located lens fibers (C, F). Note the localization of N-cadherin in both the corneal endothelium and all regions of the lens, whereas B-cadherin is detected in the corneal epithelium and in lens fiber cells. In (A, D), C, cornea; E, lens epithelium. Bar, 50 μm.

Figure 3.

 

Biochemical analysis of N- and B-cadherin cytoskeletal association in fractions of microdissected E10 or E11 lenses. E10 or E11 lenses were microdissected into anterior central epithelial (EC), equatorial epithelial (EQ), peripheral fiber (FP), and central fiber (FC) fractions, detergent extracted, and prepared for SDS-PAGE and Western immunoblotting for N- or B-cadherin as described in Methods. (A) Typical blot pattern of the partitioning of N- and B-cadherin into Triton-soluble, RIPA-soluble, and RIPA-insoluble pools in the various microdissected lens fraction. (B) Average results of densitometric scans from four sets of experiments (±SEM). Note the shift of both N- and B-cadherin from a Triton-soluble to a Triton-insoluble pool during lens cell differentiation as well as the greater resistance to RIPA solubilization of B-cadherin compared with N-cadherin.

Figure 4.

 

Immunocytochemical analysis of N- and B-cadherin localization during early stages of lens differentiation in vitro. Lens cell cultures were established from E10 or E11 lenses and double-stained for N- (A, E) or B-cadherin (C, G) and filamentous actin (B, D, F, H) as described in Methods. (A through D) Early undifferentiated monolayers; (E) through (H) the differentiating hexagonal cuboidal packed monolayers. The same field using the same focal plane was photographed in pairs of panels (A, B), (C, D), (E, F), and (G, H). Note the earlier appearance of N-cadherin at cell–cell borders compared with the later localization of B-cadherin. Bar, 50μ m.

Figure 5.

 

Immunocytochemical analysis of N-cadherin, B-cadherin, and connexin 56 during early stages of lens differentiation in vitro. Differentiating hexagonal cuboidal packed monolayers were immunostained for N-cadherin (red, A), B-cadherin (red, D), or connexin 56 (green, B, E) and examined by confocal microscopy as described in Methods. Note that in the merged confocal images of N-cadherin staining and connexin 56 staining (C), there was little overlap of N-cadherin and connexin 56 staining, seen as yellow or yellow-orange staining at cell–cell borders. In contrast, those regions containing intense B-cadherin cell–cell border staining also showed a similar localization of connexin 56, as indicated by the more extensive yellow or yellow-orange staining at cell–cell borders in (F). Bar, 50 μm.

Figure 6.

 

Immunocytochemical analysis of N-cadherin, B-cadherin, and connexin 56 localization in differentiated lentoids in vitro. Lens cell cultures containing lentoids were established and examined by confocal microscopy after immunostaining for N-cadherin (red, A), B-cadherin (red, D), connexin 56 (red, G), or filamentous actin (green, B, E, H) as described in Methods. Double-stained, merged confocal images for actin and for N-cadherin (C), B-cadherin (F), or connexin 56 (I) is also shown. Note the similar staining patterns of all four proteins in these differentiated lentoids. Bar, 50 μm.

Figure 7.

 

Localization of B-cadherin and connexin 56 during primary lens fiber formation in vivo. Lenses from E4 chick embryos were processed for immunohistochemistry and stained for detection of B-cadherin (A) and connexin 56 (B) as described in Methods. Note the similar distribution pattern of B-cadherin and connexin 56 in differentiating primary lens fiber cells, whereas B-cadherin also is detected in the cornea. In (C), C, cornea; E, lens epithelium; F, primary fiber cells. Bar, 50 μm.

The authors thank Lucy Reed for expert technical assistance and Masatoshi Takeichi, Louis Reichardt, Christoph Redies, Karen Knudsen, and Jean Jiang for gifts of antibodies.