L-Type Calcium Channels Play a Critical Role in Maintaining Lens Transparency by Regulating Phosphorylation of Aquaporin-0 and Myosin Light Chain and Expression of Connexins

Homeostasis of intracellular calcium is crucial for lens cytoarchitecture and transparency, however, the identity of specific channel proteins regulating calcium influx within the lens is not completely understood. Here we examined the expression and distribution profiles of L-type calcium channels (LTCCs) and explored their role in morphological integrity and transparency of the mouse lens, using cDNA microarray, RT-PCR, immunoblot, pharmacological inhibitors and immunofluorescence analyses. The results revealed that Ca (V) 1.2 and 1.3 channels are expressed and distributed in both the epithelium and cortical fiber cells in mouse lens. Inhibition of LTCCs with felodipine or nifedipine induces progressive cortical cataract formation with time, in association with decreased lens weight in ex-vivo mouse lenses. Histological analyses of felodipine treated lenses revealed extensive disorganization and swelling of cortical fiber cells resembling the phenotype reported for altered aquaporin-0 activity without detectable cytotoxic effects. Analysis of both soluble and membrane rich fractions from felodipine treated lenses by SDS-PAGE in conjunction with mass spectrometry and immunoblot analyses revealed decreases in β-B1-crystallin, Hsp-90, spectrin and filensin. Significantly, loss of transparency in the felodipine treated lenses was preceded by an increase in aquaporin-0 serine-235 phosphorylation and levels of connexin-50, together with decreases in myosin light chain phosphorylation and the levels of 14-3-3ε, a phosphoprotein-binding regulatory protein. Felodipine treatment led to a significant increase in gene expression of connexin-50 and 46 in the mouse lens. Additionally, felodipine inhibition of LTCCs in primary cultures of mouse lens epithelial cells resulted in decreased intracellular calcium, and decreased actin stress fibers and myosin light chain phosphorylation, without detectable cytotoxic response. Taken together, these observations reveal a crucial role for LTCCs in regulation of expression, activity and stability of aquaporin-0, connexins, cytoskeletal proteins, and the mechanical properties of lens, all of which have a vital role in maintaining lens function and cytoarchitecture.


Introduction
The human ocular lens needs to maintain its transparency and tensile properties over several decades to support normal vision. Any compromise in these characteristics can lead to impaired accommodation and cataract formation which is a global health burden. To achieve transparency, the ocular lens has evolved remarkable structural adaptations including an avascular phenotype, absence of organelles from mature fiber cells, radial packing of fiber cells, and an internal microcirculation system. [1,2] Lens fiber cells also maintain high tensile strength to support deformability during visual accommodation. The differentiated and unusually long and thin fiber cells, which constitute the bulk of the lens, are derived from epithelial cells that exit the cell cycle at the lens equator. Subsequently, they embark on a differentiation process that induces extensive cell elongation, membrane changes, reorganization of the cortical cytoskeleton and cell adhesive complexes, and expression of various fiber cell abundant and specific proteins including the crystallins, cytoskeletal proteins and channel proteins. [3,4,5,6] While the hexagonal geometry, ordered packing, deformability, membrane cytosketal network integrity and channel protein organization of fiber cells are considered critical for optical clarity and focusing ability of the lens, [1,4] the molecular and biochemical mechanisms governing these unique structural specializations and interactions are not well understood.
Our recent work as well as work from other laboratories has documented the expression and distribution of several adaptor proteins in the lens with known roles in linking cytoskeletal proteins to membrane proteins, including the ERM (ezrin, moesin, radixin) proteins, ankyrin-B, NrCAM, periaxin, desmoyokin (AHNAK), beta-IV spectrin, dystroglycan, contactins and Caspr. [7,8,9,10] Importantly, these scaffolding proteins have been shown to regulate membrane clustering and organization of ion channel proteins in neuronal and cardiac muscle cells. [11,12,13,14] Further, the PDZ domain-containing protein AHNAK/desmoyokin has been demonstrated to interact directly and regulate LTCC activity, indicating a potential role for this protein in the organization of LTCCs in lens fibers as well. [15,16,17,18] Given our long term research objective of determining the functional and regulatory significance of cytoskeletal scaffolding proteins involved in fiber cell membrane organization, and our limited understanding of the role of LTCCs in lens fibers, here we have undertaken an analysis of the expression and distribution profile of LTCCs in the lens to determine the role of these channel proteins in lens architecture and transparency.
Calcium homeostasis has been recognized to be critical for lens structural integrity and transparency, [19,20] with elevated levels of lens calcium and the subsequent stimulation of calpain activity reported to cause cataract or lens opacification. [21,22,23,24,25,26] Furthermore, intracellular calcium acting together with calmodulin has been reported to regulate the activities of the aquaporin-0 water channel and connexin gap junctions, as well as the stability of crystallin proteins which are required for lens transparency and architecture. [27,28,29,30,31,32,33,34,35,36,37] Therefore, a tightly regulated calcium influx is critical for homeostasis of lens physiology and function. Although there is indirect evidence for the presence of voltage-gated LTCCs in lens epithelial and fiber cells, limited knowledge is currently available regarding a direct role for these proteins in normal lens transparency, as well as their expression and distribution profiles within lens. [38,39,40,41] Moreover, lens fibers have been reported to contain high levels of phosphorylated myosin light chain (MLC) relative to lens epithelium, with the phosphorylation of MLC being regulated by calcium/calmodulindependant MLC kinase. [42] Significantly, inhibition of MLC kinase activity has been demonstrated to induce lens opacification, indicating that biochemical pathways/mechanisms regulating MLC kinase activity, such as regulation of intracellular calcium flux and activity of calcium/calmodulin dependent MLC kinase, play an important role in lens function. [42] While the LTCCs are known to regulate intracellular calcium flux, their potential role in regulation of MLC phosphorylation, and the activities and expression of aquaporin-0 and connexins in the lens is not well understood.
Based on the significant role that intracellular calcium appears to play in the lens, in this study we determined how impairment of LTCC activity might influence lens transparency and fiber cell structural integrity. This study reveals that LTCC inhibition in exvivo mouse lenses impairs water permeability, cytoskeletal organization and transparency in association with dysregulated aquaporin-0 serine phosphorylation, myosin light chain phosphorylation and expression of connexins, and uncovers an essential role for LTCC activity in maintenance of lens transparency and function.

Results
Expression and Distribution of Ca (V) 1. 2

and 1.3 Channels in Mouse Lens
To determine the expression of LTCCs in lens tissue, we utilized several approaches including cDNA microarray, RT-PCR and immunoblot analyses. cDNA microarray analysis of RNA derived from a 7-day old C57BL/6J mouse lenses revealed expression of different LTCC genes including Ca(v) 1.1, 1.2, 1.3 and 1.4 at varying levels, together with their auxiliary subunit genes a2d and b (data not shown). Based on the cDNA microarray data, we examined for the expression profile of the pore forming subunit genes of all four LTCC subtypes including Ca (V) 1.1, 1.2, 1.3 and 1.4 along with the expression of their auxiliary subunits (a2d and b) by RT-PCR analysis using RNA derived from P2 and P21 mouse lenses. As shown in Fig. 1A., the expression profiles of Ca (V) 1.2 and 1.3 were readily detectable compared to the expression of Ca (V) 1.1 and 1.4 which is consistent with the known wide range of tissue expression for Ca (V) 1.2 and 1.3 channels relative to Ca (V) 1.1 and 1.4, which exhibit restricted tissue distribution. [43,44,45,46] Both P2 and P21 mouse lenses exhibited expression of the LTCC auxiliary subunit genes including a2d and b2 as shown in Fig. 1B. Expression of Ca (V) 1.2 and 1.3 proteins was then assessed by immunoblotting analysis of membrane rich and soluble fractions derived from three weekold mouse lenses, using the appropriate monoclonal antibodies. As shown in Fig. 1C, Ca (V) 1.2 and 1.3 antibodies yielded specific immunopositive signals only in the membrane enriched lens fraction, identifying proteins of the expected molecular mass (, 250 kDa), together with proteins in the lower range of , 160 kDa which represent truncated forms of LTCCs (as per antibody provider) confirming the presence of Ca (V) 1.2 and 1.3 channel proteins in the mouse lens. Intriguingly, the immunoreactivity with lower or fragmented proteins especially with Ca (V) 1.3 specific antibody was rather intense and the reasons for this are not clear. We also examined the distribution pattern of Ca (V) 1.2 and 1.3 channel proteins in the P21 mouse lens equatorial sections by immunofluorescence-based confocal imaging, which revealed that Ca(V) 1.2 and 1.3 channel proteins are distributed to the lens epithelium and fiber cells of the outer and inner cortical regions ( Fig. 1D and E). In mouse lens equatorial sections, both Ca (V) 1.2 and 1.3 are localized predominantly to the short arms of hexagonally shaped fiber cells ( Fig. 1D and E). Although both channel proteins are localized intensely to the short arms relative to the long arms of fiber cells, there appear to be some distinct differences in the subcellular distribution pattern at the short arms of the fiber cells. While Ca (V) 1.3 distribution along the short arm is localized uniformly to the plasma membrane, Ca (V) 1.2 channel protein exhibits a rather patchy and clustered distribution patter at the short arm of fiber cells ( Fig. 1D and E, right panels). Although the use of both antibodies (Ca (V) 1.2 and 1.3) appears to be associated with some immunostaining in the cytosol of the lens epithelium, the soluble fraction of lens homogenates did not exhibit any immunopositive reactivity to these antibodies by blot analysis indicating some non-specific staining in the cytosolic regions of lens ( Fig. 1D and E). These distribution patterns of LTCCs were confirmed in two independent lenses which were fixed at different times to rule out fixation variability. Since the lens nucleus is very hard, this region did not exhibit much immunopositive staining, as is the case for many other proteins examined.

Inhibition of L-type Calcium Channel Activity Impairs Transparency in Ex-vivo Mouse Lenses
To determine a potential role for LTCCs in maintenance of lens transparency, we examined the effects of dihydropyridine inhibitors felodipine and nifedipine, which are specific antagonists of LTCCs, on organ cultured mouse lenses. [44,46,47] Lenses derived from P21 to P26 day-old mice (male and female animals of the C57BL/6J strain) treated with 10 mM felodipine (dissolved in 0.025% DMSO) exhibited time dependent changes in transparency ( Fig. 2A). While lenses treated with felodipine (10 mM) for 6 hrs remained mostly transparent with a few specimens exhibiting only a spotty subcapsular/cortical haziness or opacity, all lenses developed an intense peripheral ring-like subcapsular opacity after 24 hrs of exposure to drug. By 48 hrs of treatment, the opacity extended throughout the cortical region of lens ( Fig. 2A). All cultured lenses exhibited protein leakage into the medium over the course of drug treatment, with the response being moderately (,10 to 20%) but significantly higher in the drug treated lenses as opposed to control lenses (Fig. 2B). Protein leakage from drug treated lenses did not increase progressively with increasing length of drug exposure (Fig. 2B). Lenses treated with drug for either 6 or 24 hrs exhibited a comparable degree of (,35 mg/ml of culture medium/lens) protein leakage. Intriguingly, however, the lens weight (wet) of drug treated specimens showed a significant decrease as compared to the control lenses after 24 and 48 hrs of incubation (Fig. 2C). Lenses treated with nifedipine (25 mM) which is known to be a less potent inhibitor of LTCC activity relative to felodipine, [47] exhibited no obvious opacity after 24 hrs of drug treatment (data not shown) but turned opaque with the opacity extending into the nucleus as well as the cortical region following a 48 hrs exposure to drug (Fig. 2D). Similar to the findings in felodipine treated lenses, treatment with nifedipine caused significant reductions in wet weight compared to control lenses after 48 hrs of exposure to drug (Fig. 2C). Control lenses were treated with appropriate volumes of vehicle (0.025% DMSO) used to dissolve drugs.

L-type Calcium Channel Inhibition Induces Disorganization and Swelling of Lens Cortical Fiber Cells in Association with Alterations in Protein Profile
Histological analysis of mouse lenses treated with felodipine (10 mM) for 6 hrs revealed an intact epithelium and outer cortical fibers, but extensive disruption of organization of inner cortical fiber cells (indicated with arrows; Fig. 3B) especially peripheral fibers of the inner cortical region, which displayed extensive swelling compared to control lenses (Fig. 3A, B and C). Extending the length of drug treatment to 24 hrs resulted in extensive disorganization of outer cortical fiber cells along with inner cortical fiber cells of the peripheral region with intact epithelium, and dramatically enhanced fiber cell swelling (Fig. 3 B, arrows) leading to larger intracellular spaces and a disruption in fiber cell organization ( Fig. 3C shows a magnified area in panel B indicated with a square box). Interestingly, treatment of lenses with felodipine for either 6 or 24 hrs did not affect the deeper inner cortical fibers and the nuclear region fiber cells, which remained intact and exhibited compact organization similar to control lenses consistent with the lack of opacity observed in the nuclear regions as assessed by H&E staining (Fig. 3A) and WGA-based fluorescence staining ( Fig. 3B and C). Drug-induced adverse effects or cytotoxicity was evaluated in lenses treated with felodipine for either 6hrs or 24 hrs by detecting apoptosis (green staining, arrows) and nuclei integrity (blue staining) and distribution using TUNEL and Hoechst staining, respectively (Fig. 3D). Both types of analyses confirmed the absence of drug-induced cytotoxic effects in felodipine treated lenses relative to control (DMSO treated) lenses (Fig. 3D).
The above described histological alterations and protein leakage changes noted in felodipine treated mouse lenses (Figs. 2 & 3) led us to evaluate possible alterations in the protein profiles of lens soluble and insoluble fractions. These analyses were performed using SDS-PAGE in conjunction with GelCode blue staining of resolved proteins. As shown in Fig. 4, there were no obvious difference in the protein profile of soluble (A) and membraneenriched insoluble fractions (B) in the 6 hrs felodipine treated lenses as compared to control lenses. However, marked differences were noted for selected proteins in both soluble (Fig. 4C) and insoluble fractions (Fig. 4D) of lenses treated with felodipine for 24 hrs. To identify proteins which exhibited obvious decrease in their band intensity relative to control samples based on SDS-PAGE analysis and gelcode blue staining of felodipine treated lenses, protein bands of interest (indicated with arrow heads) were extracted from SDS-PAGE gels, trypsin digested and subjected to mass spectrometry-based (Synapt G2) identification. The mass spectrometry analysis identified b-B1 crystallin, Hsp-90, filensin, b-catenin, phakinin, spectrin, 14-3-3e and vimentin as proteins which exhibited decrease in levels in drug treated lenses (Table 1).
Deregulated Aquaporin-0 Phosphorylation, Expression of Connexins and Decreased Levels of 14-3-3e Protein Precedes Lens Opacity Induced by LTCC Inhibition As described earlier, felodipine treated lenses exhibit decreases of lens weight (Fig. 2) and extensive fiber cell swelling (Fig. 3) indicating possible changes in water permeability and transport. Since water channel activity of aquaporin-0, which is the most abundant water channel protein in the lens fiber cell membrane, is known to be regulated by calcium and calmodulin in a phosphorylation (Ser235) dependent manner, [29,33,34,35] we examined changes in aquaporin-0 serine235 phosphorylation in drug-treated lenses, using a ser235 phospho-specific polyclonal aquaporin-0 antibody. Aquaporin-0 ser235 phosphorylation which is regulated by protein kinase A and AKAP2 (A-kinase anchor protein-2) was recently demonstrated to be crucial for the water permeability activity of aquaporin-0. [29] The total aquaporin-0 protein levels in the membrane enriched insoluble lens fraction (P21) were found to be similar between felodipine treated and control specimens following 6 or 24 hrs of drug treatment (Fig. 5A, B and E). Additionally, despite extensive disorganization and swelling of fiber cells in drug-treated lenses (Fig. 5F, inserts), aquaporin-0 was found to exhibit a normal pattern of distribution to the fiber cell membrane, as determined by immunofluorescence staining (Fig. 5F, arrows in insert). However, significant increases were evident in the level of ser235 phosphorylated aquaporin-0 in lenses treated with felodipine for 24 hrs ( Fig. 5B and C). We also measured changes in serine-phosphorylated proteins in membrane-enriched fractions from felodipine-treated lenses using a monoclonal antibody specific for the serine-phosphorylated residue in proteins. The same ,28 kDa molecular mass protein which exhibited intense reactivity with ser235 phospho-specific aquaporin-0 antibody also cross reacted with the phosphoserine antibody, exhibiting elevated levels in felodipine treated lenses compared to control lenses (Fig. 5A, B and D). These two observations collectively reveal that inhibition of LTCCs in mouse lenses increases aquaporin-0 ser235 phosphorylation.
In addition to aquaporin-0, we examined possible changes in the distribution and levels of connexins in felodipine-treated lenses, since the gap junction activity of these proteins, which are abundantly expressed in fiber cells, is essential for lens transparency. [5,48] As shown in Fig. 6, connexin-50 is distributed in a precisely organized manner, exhibiting a clustered or patchy localization (Fig. 6D, arrows and insert) at the center of the long arm of hexagonal fiber cells based on immunofluorescence confocal imaging of sections from the lens equatorial plane. On the other hand, felodipine treated lenses exhibit a complete disorganization of connexin-50 distribution in the cortical fibers, with no disruptions noted in nuclear fibers (Fig. 6D). Intriguingly, the levels of connexin-50 were found to exhibit a robust and significant increase in membrane-enriched lens fractions from felodipine treated lenses relative to control lenses ( Fig. 6 A, B and C), starting from 6 hrs of treatment with felodipine. Total aquaporin-0 was immunoblotted as a loading control in experiments designed to assess changes in levels of connexin-50. Connexin-46 protein levels and distribution were not examined in this study. A. Treatment of organ cultured (P21 to P26) mouse lenses with felodipine (10 mM) induced progressive lens opacity (subcapsular to cortical) with increasing time. Control lenses treated with DMSO (0.025%) remained transparent throughout the experiment. B. The felodipine treated lenses showed significant (* P,0.01) increase in protein secretion/leakage (mg/ml media/lens) into the media as compared to control lenses, based on the mean value from an n = 4 to 6 samples. D. Mouse lenses (P21 to P26) treated with nifedipine (25 mM) exhibited cortical and nuclear opacity after 48 hr of drug treatment, while control lenses treated with DMSO alone (0.025%) remained transparent. C. Both felodipine and nifedipine treated mouse lenses showed a significant (* P,0.01; n = 4-6) decrease in lens wet weight (mg/lens) compared to the corresponding controls. doi:10.1371/journal.pone.0064676.g002 Additionally, to determine whether the noted increase in the levels of connexin-50 was due to increased gene expression, felodipine treated lenses were examined for changes in the levels of connexin-50 and 46 transcripts by q-RT-PCR analysis (Fig. 7). This analysis revealed a significant increase in the levels of both connexin-50 and 46 transcripts in felodipine treated lenses relative to control lenses ( Fig. 7A-D). Figures 7A and C depict the q-RT-PCR output traces for both connexin-50 and 46, respectively, along with the GAPDH control as indicated with arrows. Figures 7B and D show quantitative changes in the expression level of connexin-50 and 46 genes, respectively, in the felodipine treated lenses as compared to control lenses, based on results from three independent analyses. Expression levels of the GAPDH (Glyceraldehyde 3-phosphate dehydrogenase) transcript were used to normalize the cDNA concentration of control and drug treated lenses in q-RT-PCR analyses ( Fig. 7A and C).
As described earlier, felodipine treated lenses revealed changes in the profile of different proteins including 14-3-3e and b-B1crystallin, based on SDS-PAGE analysis followed by mass spectrometry -based identification ( Fig. 4C; Table 1)). Since 14-3-3e and other isoforms of this protein are known to interact with and regulate the membrane organization and activity of various channel proteins, [49,50,51,52,53,54] we evaluated the effects of felodipine on the levels of this protein in drug-treated mouse lenses. The levels of 14-3-3e protein decreased progressively and significantly in felodipine-treated lenses starting from 6 hrs onwards, based on immunoblot analysis of soluble lens fractions (Fig. 8A, B & C). On the other hand, the levels of b-B1-crystalin, which is a calcium binding crystallin, [37] exhibited modest increases in the 6 hrs felodipine-treated lenses (soluble fraction) ( Fig. 8A & D), but at 24 hrs following drug treatment, the levels of this protein decreased significantly compared to the levels in control lenses ( Fig. 8B and D). GAPDH was used as a loading control for these analyses.

Impaired Myosin Light Chain Phosphorylation Precedes Lens Opacity Induced by LTCC Inhibition
Myosin light chain (MLC) which is a regulatory subunit of myosin II and whose phosphorylation is regulated by calcium/ calmodulin-dependent MLC kinase, is critical for actomyosinbased contractile and mechanical activity. [55] We evaluated MLC phosphorylation to determine whether impairment of LTCC activity by felodipine indeed influences MLC phosphorylation in lens tissue. This was done by immunoblotting analysis using a phospho-specific MLC antibody (Fig. 9). In lenses treated with 10 mM felodipine for either 6 or 24 hrs, there was a significant (.80%) decrease in MLC phosphorylation compared to control lenses ( Fig. 9 A, B and C). In contrast, the levels of total MLC protein were significantly elevated in the same fractions from the felodipine treated lenses, based on immunoblotting analysis and as compared to control lenses ( Fig. 8 A, B and D). This response was found to be much robust in lenses treated for 6 hrs with felodipine as compared to lenses treated with the drug for 24 hrs (Fig. 9D). For both phospho-MLC and total MLC analyses, a-tubulin was immunoblotted as a loading control.

LTTC Inhibition in Mouse Lens Epithelial Cells Disrupts Actin Cytoskeletal Organization and Decreases Intracellular Calcium and MLC Phosphorylation
To obtain further insight into the influence of LTCC activity on the regulation of MLC phosphorylation, actin cytoskeletal organization and intracellular calcium, lens primary epithelial cells, derived from P26 C57BL/6J mice were treated with felodipine (10 mM) for either 6 or 24 hrs in serum free medium. The FURA-2/AM-loaded lens epithelial cells were evaluated by digital imaging microscopy for baseline fluorescence at 340/ 380 nm excitation in the presence and absence of thapsigargin. As shown in Fig. 10, while control cells exhibited a robust and sharp spike/elevation in the calcium-dependent FURA-2/AM fluorescence with addition of thapsigargin, the felodipine-treated lens epithelial cells from both 6 and 24 hrs treatments failed to show any significant calcium signaling in response to thapsigargin, indicating a significant reduction in total intracellular calcium in the LTCC-inhibited cells. These conclusions were based on recordings with several individual cells. Felodipine treated lens epithelial cells were examined for changes in cell viability and cytotoxicity using in vivo fluorescein diacetate hydrolysis and propidium iodide labeling. As shown in the supplemental figure ( Figure S1), the felodipine treated lens epithelial cells were found to be viable based on in vivo imaging of fluorescein diacetate hydrolysis -dependent green fluorescence and lack of propidium iodide staining, similar to what is noted for control cells. These data confirm lack of drug-induced cytotoxic effects in lens epithelial cells ( Figure S1) and are consistent with the noted lack of apoptosis in the felodipine treated lenses described earlier (Fig. 3D).
Importantly, the felodipine treated lens epithelial cells revealed a dramatic decrease in actin stress fibers and MLC phosphorylation as assessed by rhodamine phalloidin staining (red) and phospho-MLC antibody immunofluorescence staining (green), respectively (Fig. 11). While control cells exhibit strong actin stress fibers (red staining) extending from one end to another end and distributed throughout the cell body, barely any actin stress fibers were evident in the felodipine treated cells, indicating druginduced actin depolymerization. Consistent with the changes observed in actin stress fibers, drug treated lens epithelial cells exhibited very little staining for phospho-MLC as compared with the intense and filamentous staining distributing along the actin stress fibers in control cells (Fig. 11). The merging of actin stress fiber staining with phospho-MLC staining confirms perfect colocalization of these two proteins (data not shown). The blue staining depicts cell nuclei labeled with Hoechst. Representative images of individual cells derived from triplicate analyses are shown. . Felodipine-induced changes in protein profile of ex-vivo organ cultured mouse lenses. Felodipine (10 mM for 6 hrs or 24 hrs) and the corresponding DMSO treated control mouse lenses (p21 to P26) were homogenized and separated into soluble (100,0006g supernatant) and insoluble (membrane-enriched) fractions. Equal amounts of protein from soluble (A, C) and insoluble (B, D) fractions were then separated by gradient SDS-PAGE (4-20%) and stained with GelCode blue. Protein bands that exhibited obvious differences in staining intensity between the felodipine treated and control lenses (annotated using numbers and arrowheads in the gel images) were extracted, trypsinized and identified using a Synapt G2 mass spectrometer. See Table 1 for a listing of the proteins identified by mass spectrometry. C1 to C3 and F1 to F3 are individual specimens from control and felodipine treated lenses, respectively. doi:10.1371/journal.pone.0064676.g004

Discussion
This study reveals the expression and distribution profile of Ca (V) 1.2 and 1.3 voltage-gated LTCCs in mouse lens epithelium and fiber cells, and demonstrates that pharmacological inhibition of these LTCCs impairs lens transparency, together with increased aquaporin-0 serine-235 phosphorylation which is known to directly modulate the water permeability activity of aquaporin-0. Treatment of mouse lenses with LTCC inhibitor did not cause any detectable cytotoxicity, but was found to upregulate gene expression of connexin-50 and -46 which are key fiber cell gap junction proteins involved in regulation of lens microcirculation. Additionally, impairment of LTCC activity in the mouse lens leads to decreased phosphorylation of MLC, which is a regulatory subunit of myosin II required for maintenance of lens cytoarchitecture and mechanical properties. Cultured mouse lens epithelial cells treated with the LTCC inhibitor felodipine also exhibit decreased actin stress fibers, MLC phosphorylation, and decreased intracellular calcium, in the absence of any cytotoxic responses.
These observations collectively uncover a critical role for LTCCs in homeostasis of lens transparency and architecture.
Intracellular calcium is an essential and versatile intracellular messenger and regulates variety of cellular functions. [56,57] Therefore, a tightly regulated calcium influx is required for homeostasis of cell and organ function. Some of our earlier intriguing observations in the ocular lens including identification of various scaffolding proteins involved in ion channel protein membrane clustering, [7] force-induced changes in MLC phosphorylation (unpublished data) and high levels of phosphorylated MLC in lens fibers [42] along with the known rapid changes in lens shape during accommodation, motivated us to examine whether LTCCs which regulate intracellular calcium, might play a significant role in lens transparency and structural integrity. The role of LTCCs in neuronal and cardiac muscle excitability, contraction coupling, various biochemical responses and gene expression has been well understood and demonstrated to have a vital role in physiology and function of various organs. [45,58,59] [45,60,61] The activity of these channel proteins is regulated by various extracellular cues and intracellular signaling mechanisms. Our data based on using multiple approaches revealed the expression and distribution of Ca (V) 1.2 and 1.3 in both lens epithelium and cortical fiber cells. Interestingly, the distribution of these channel proteins in fiber cells was discretely and intensely localized to the short arms of the hexagonal lens fibers. Additionally, relative to Ca (V) 1.3, which is distributed uniformly to the short arm, Ca (V) 1.2 was found to be clustered at the short arm. The reason for these differences in the distribution pattern of these two LTCCs within lens fiber cells is not clear at present. However, Ca (V) 1.2 and 1.3 are known to influence different cellular responses despite their similar activity as it relates to regulation of intracellular calcium. [45,60,61].
The role of LTCCs in maintenance of normal lens transparency and structural integrity is not understood. The data from this study demonstrate that LTCC activity is critical for normal lens transparency based on felodipine and nifedipine induced opacity in ex-vivo lenses. The subcapsular and cortical opacity induced by these agents is correlated with the intense distribution of LTCCs in the cortical fibers. Additionally, this body of work regarding the importance of LTCC activity for normal lens transparency further supports the earlier reported clinical observation of nifedipine use posing an increased risk for cataract formation in human subjects. [62,63] Intriguingly, the felodipine and nifedipine treated mouse lenses which exhibited cortical opacity, also revealed decreases in lens weight without any noticeable cytotoxic response. Although there was an increase in protein leakage from the LTCC inhibited lenses, the amount of protein leakage from these lenses was not very large as compared to the protein losses from control lenses. Additionally, the control lenses did not exhibit any opacification despite the noted protein leakage during lens incubation. Therefore, it is possible that in addition to protein leakage, altered microcirculation and water permeability could be partly responsible for the observed decreases in lens weight. The basis for this latter speculation is the fact that histological changes including disorganization and swelling of cortical fibers noted in the LTCC inhibited lenses appeared very similar to the changes reported in the mouse lenses with altered aquaporin-0 water channel activity in ex vivo. [29].
In contrast to our data on LTCC inhibitor induced opacification in ex-vivo normal lenses, other LTCC inhibitors including verapamil and nifedipine have been reported to delay formation of diabetic and radiation-induced cataracts in rodents. [38,39,40,41] Diabetic and radiation-induced cataract lenses have been reported to contain elevated levels of intracellular calcium which is expected to activate calpain and caspases and cause protein degradation and cell death, respectively. [38,41,64] Additionally, diabetic lenses are reported to have elevated levels of sugar alcohols such as sorbitol and xylitol which can lead to osmotic imbalance and lens opacification. [65] Since inhibition of LTCCs is expected to decrease intracellular calcium in lens epithelial cells as documented in our studies (Fig. 10) and those from other laboratories [66], LTCC inhibitors are presumed to have a protective response in delaying lens opacification in diabetic and radiation-exposed rodents as noted in several published studies. [  Additionally, elevated levels of polyols are thought to cause osmotic imbalance leading to water accumulation in extracellular spaces in the diabetic lenses. [67] It is plausible that the LTCC inhibitor mediated increase in Ser235 phosphorylation of aquaporin-0 noted in this study with expected subsequent enhancement in aquaporin-0 water permeability [29] might mediate a protective effect towards delaying cataract formation in diabetic lenses by increased influx of water from the extracellular environment into cell interior and subsequently coupling to the lens microcirculation system. However, in contrast to sugar and radiation-induced cataracts, inhibition of LTCCs in normal lenses, per se appears to cause a variety of biochemical and metabolic insults due to decreased intracellular calcium including transcriptional changes, changes in cytoskeletal organization and integrity, deregulated water channel, gap junction and myosin II activities. Collectively these effects appear to impair lens transparency under normal conditions.
Our studies revealed increased levels of connexin-50 in the membrane-enriched fraction of lenses treated with felodipine relative to control lenses, with the increases preceding any noticeable lens opacification. Interestingly, the increased levels of connexin-50 in the felodipine treated lenses appear to be the result of transcriptional upregulation based on the q-RT-PCR results (Fig. 7). Additionally, as shown in this study, the expression of connexin-46 gene was also found to be upregulated under LTCCs inhibition indicating a close regulatory interaction between the homeostasis of intracellular calcium and transcriptional control of connexin gene expression in the lens tissue. On the other hand, while the total aquaporin-0 levels were found to be similar between the felodipine treated versus control lenses, the levels of ser235 phosphorylated aquaporin-0 were elevated significantly in felodipine treated lenses. It has been reported that Aquaporin-0 ser235 phosphorylation is partly regulated by calcium/calmodulin and PKA. [29,34] Calmodulin has been shown to bind to the Cterminal domain of aquaporin-0 near ser235 in a calcium dependent manner [28,33,34,68] and mask this site for ser235 phosphorylation by PKA. [29] This blockage of PKA-regulated ser235 phosphorylation of aquaporin-0 by calcium/calmodulin is thought to impair aquaporin-0 water permeability activity by closing the channel pore. [29] On the other hand, the absence or decreased levels of intracellular calcium impairs the complex formation of calcium/calmodulin with aquaporin-0, facilitating the PKA/AKAP2 mediated ser235 phosphorylation and activation of the pore activity of aquaporin-0 for water, thereby increasing influx of water. [29] Therefore the increase in aquaporin-0 ser235 phosphorylation observed in our studies upon LTCC inhibition, when taken together with increased fiber cell swelling in cortical fibers, strongly suggests that the LTCC regulated calcium influx directly or indirectly influences aquaporin-0 water channel activity and fiber cell water permeability. Interestingly, aquaporin-0 has been shown to interact directly connexin-50 and this interaction has been shown to enhance connexin-50 gap junction coupling. [69,70] Therefore, it is conceivable that there may be also some possible relationship between aquaporin-0 phosphorylation and connexin-50 expression in the LTCC inhibited lenses.
It is also noteworthy that 14-3-3e, a serine/threonine phosphoprotein binding protein which is known to regulate activity and biogenesis of LTCC and other channel proteins, [49,50,51,52,53,54] was found to be decreased in the felodipine treated lenses before any significant lens opacification had occurred. Interestingly, the protein levels of this broad family of regulatory proteins are reported to be decreased upon inhibition of Na-K ATPase activity in the lens, indicating their close functional association with transport or channel protein activity in lens. [71,72] Additionally, PKCc interaction with 14-3-3e has been shown to influence connexin-43 gap junction activity in lens epithelium. [73].
Our observation of a marked decrease in MLC phosphorylation in LTCC inhibited lenses together with significantly increased levels of total MLC reveals the importance of LTCC regulated intracellular calcium in regulation of myosin II activity in lens. This conclusion was also partly supported by the data obtained from lens epithelial cells in which inhibition of LTCCs was associated with decreased intracellular calcium, and decreased actin stress fibers and MLC phosphorylation ( Fig. 10 and 11). It is possible that the elevated levels of total MLC in the felodipine treated lenses could be an adaptive response to the severely compromised activity of MLC. Additionally, in support of the importance of myosin II activity for lens transparency and architecture, in our previous study we found that inhibition of MLC kinase which is a calcium/calmodulin-dependent kinase, leads to lens opacification. [42] Although the precise role of the robust increases in levels of phosphorylated MLC in differentiating lens fibers reported earlier is yet to be understood, [42] changes in lens shape induced by physical force caused decreases in MLC phosphorylation, suggesting an involvement of force sensitive ion channels in regulation of intracellular calcium (our unpublished data). One of the unique characteristics of lens is its rapid deformability during accommodation. Whether lens deformability and the associated biomechanical and shape changes in fiber cell is partly influenced by intracellular calcium and contractile properties are not known. The role of LTCC activity in cardiac contractile coupling is recognized to be important for heart function. [58,61,74] Therefore, additional studies are necessary to explore and delineate the role of LTCC activity in lens contractile and tensile properties during lens accommodation using primate lenses. In addition, it is well recognized that myosin II activity is essential for lens function based on cataract formation in both human and mouse models with mutation in nonmuscle myosin II-A (MYH9). [75,76] In future studies, we plan to determine whether inhibition of LTTC activity influences lens stiffness to unravel the significance of LTCCs in maintaining the lens mechanical properties and function. Finally, in this initial study we demonstrate the importance of LTCC activity for maintenance of lens transparency. The precise role, significance and mechanistic understanding of the noted changes in aquaporin-0 phosphorylation, expression of connexins and levels of 14-3-3e which occur upon inhibition of LTCC, however, require additional in vivo and gene targeted studies.

Ethics Statement
Animals. This study was carried out in strict accordance with the recommendations of the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and the Association for Research in Vision and Ophthalmology. The protocol was approved by the Committee for Ethics of Animal Experiments at the Duke University School of Medicine (Protocol Number: A007-11-01).  Mice C57BL/6J strain mice obtained from Taconic Laboratories (Hudson, NY) were bred in-house at the Duke University Medical Center Vivarium facility and used in this study. Lenses obtained from weaned mice (male and female animals aged between P21 to P26 days) were used in the organ culture studies. Animals were euthanized using intraperitoneal injections of Euthasol.

Lens Organ Cultures
Freshly dissected lenses from 3-week-old mice were incubated at 37uC under 5% CO 2 in Dulbecco's Modified Eagle Medium (Low Glucose DMEM, 298 mOsm) containing penicillin (100 U/ml) and streptomycin (100 mg/ml). Lenses were stabilized in the culture media for 16 hrs and pre-screened to rule out any potential dissection-associated damages by assessing for protein leakage into the media as described earlier. [42] Pre-screened clear lenses were cultured in control media or media containing felodipine or nifedipine dissolved in dimethyl sulfoxide (DMSO, at a final concentration of 0.025%). Lenses treated with DMSO (0.025%) alone served as controls for drug treatments. Media were changed every day for both the control (DMSO treated) and drug-treated lenses, lens transparency was monitored and graded daily using a dissecting light microscope and images were recorded under dark illumination using Canon digital camera (Zeiss AXio Cam ERc 5 s). Felodipine (10 mM) treated lenses were harvested at 6, 24 and 48 hrs intervals, while nifedipine (25 mM) treated lenses were harvested after 48 hrs. DMSO treated control lenses were harvested at the respective time intervals mentioned above. Following drug treatment, lenses were fixed and stored at the appropriate conditions for analysis of histological changes by hematoxilin and Eosin (H&E) staining, immunohistochemistry, total protein profiles and immunoblotting.

Cell Viability and Cytotoxicity
To evaluate the effects of felodipine on the viability of mouse lens epithelial cells, cells were grown on gelatin-coated 12 well plastic plates in the presence of DMEM and FBS as described above. After treatment with 10 mM felodipine for 6 or 24 hrs in the absence of FBS, cells were rinsed twice with DMEM and then treated with fluorescein diacetate (2 mg/mL ) and propidium iodide (0.6 mg/mL) for 3 minutes. Following this, viable and dead/ damaged cells, which stain green and red, respectively, were viewed and imaged using a Zeiss Observer.D1 fluorescence microscope.

Intracellular Calcium Assay
Changes in the levels of intracellular calcium were determined using FURA-2/AM in mouse lens epithelial cells treated with felodipine, as previously described. [78] Briefly, mouse lens epithelial cells were seeded in glass-bottomed 35-mm Petri dishes and were allowed to grow for three days in DMEM with FBS (10%). The cells were then rinsed with DMEM (no FBS) and treated with felodipine (10 mM) for 6 to 24 hrs in the absence of FBS. Following this, the cells were incubated with FURA-2/AM at a final concentration of 4 mM for 30 min at 37uC in the dark. The cells were then washed twice with Hanks' balanced salt solution followed by recording base-line measurements of relative intracellular concentration in control and felodipine-treated cells using digital imaging microscopy at 340/380 nm excitation. After obtaining base-line measurements, thapsigargin (1 mM) was added to the culture media, and measurements were continued for over 3 minutes at 5 seconds interval. Intracellular calcium levels in control and drug treated cells in the presence and absence of thapsigargin was expressed as the ratio of emitted fluorescence of cytoplasmic FURA-2/AM at 340/380-nm light excitations. We collected data from 5-10 cells per high power microscope field. The data were collected and analyzed with SimplePCI 6 (Hamamatsu Corp., Serwickley, PA).

RT-PCR
To determine the expression profile of genes encoding L-type Calcium channels and their auxiliary subunits in mouse lens tissue, we performed RT-PCR analyses using total RNA extracted from the P2 and P21-day-old mouse (C57BL/6 strain) lenses using RNeasy Micro kit as we described earlier. [7] The Advantage RTfor-PCR Kit was used to synthesize first-strand cDNA from total RNA isolates, followed by amplification of cDNA pools for the genes of interest, using an AdvantageH 2 PCR Kit. Minus RT (reverse transcriptase) controls were included in the RT-PCR reactions. The mouse-specific forward and reverse oligonucleotide primer sets were used in PCR reactions to amplify cDNA species corresponding to various LTCC subunits ( Table 2): The PCR gene products were extracted and sequenced to confirm identity.

q-RT-PCR
Real-time quantification of connexin-50 and -46 gene transcripts in felodipine treated and control mouse lenses was performed using a Prism 7700 Sequence Detection System (Applied Biosystem, Inc), as previously described by us. [77] Briefly, first strand cDNA pools from the drug treated and control lenses were normalized relative to the housekeeping gene GAPDH. PCR reactions were done in triplicate using PCR Master Mix as described earlier. [77] The sequences of oligonucleotide primer sets of connexin-50 and -46 are listed in Table 2. The fold difference in connexin-50 and connexin-46 gene expression between control and felodipine treated lenses was normalized to GAPDH and calculated by the comparative threshold (C T ) method, as described by the manufacturer (Prism 7700 Sequence Detection System; Applied Biosystem, Inc.).

Histological Analysis
The organ cultured lenses were fixed and stored in 10% buffered formalin at room temperature until further processing. The specimens were subsequently dehydrated and embedded in paraffin, and 5-micron thick sections were cut in equatorial plane and mounted on the charged glass slides. These sections were subjected to H & E staining. Images were captured using Zeiss Axioplan2 microscopy (Zeiss Inc. Thornwood, NY. USA).

Analysis of Lens Protein Profile in Cytosolic and Membrane Enriched Fractions
Control and drug treated cultured mouse lenses were homogenized using a glass homogenizer and cold (4uC) hypotonic buffer containing 10 mM Tris buffer pH 7.4, 0.2 mM MgCl2, 5 mM Nethylmaleimide, 2.0 mM Na 3 VO 4 , 10 mM NaF, 60 mM phenyl methyl sulfonyl fluoride (PMSF), 0.4 mM iodoacetamide, Protease inhibitor cocktail (complete, Mini, EDTA-free) and PhosSTOP Phosphatase Inhibitor Cocktail (one each/10 ml buffer), Roche (Basel, Switzerland)). Homogenates were centrifuged at 8006g for 10 min at 4uC, and supernatant fractions were centrifuged further at 100,0006g for one hr at 4uC. The 100,0006g supernatant was designated as the cytosolic fraction, while the insoluble pellets derived were re-suspended in hypotonic buffer, washed twice with same buffer and the final membrane-enriched protein pellets were suspended in urea sample buffer containing 8 M urea, 20 mM Tris, 23 mM glycine, 10 mM dithiothreitol (DTT), and saturated sucrose along with protease and phosphatase inhibitors. Protein concentrations were estimated in both the cytosolic and membrane-enriched fractions by the Bradford method using protein assay reagent. Protein samples were separated by gradient SDS-PAGE (4-20% Criterion XT precast gels) using 16 MOPs buffer (Invitrogen) and stained overnight with GelCodeH Blue stain reagent. After destaining with Milli Q pure water, gels were imaged using a Fotodyne image system, and protein bands of interest were excised from the gels and trypsin digested for further characterization by mass spectrometry. Briefly, gel slices were subjected to in-gel tryptic digestion using the In-Gel Tryptic digestion kit (Pierce), as per manufacturer's instructions. This digestion process included both reduction and alkylation of protein samples.

Mass Spectrometry
Peptides obtained from in-gel-trypsin-digests where analyzed using a nanoAcquity UPLC system coupled to a Synapt G2 HDMS mass spectrometer (Waters Corp, Milford, MA). Peptides were separated on a 100 mm6100 mm column containing 1.7 mm C18 BEH particles (Waters), and using a 30-min gradient of 5 to 32% acetonitrile with 0.1% formic acid at a flow rate of 0.3 ml/ min and a column temperature of 35uC. For each sample, we conducted a data-dependent analysis (DDA) using a 0.8 s MS scan followed by MS/MS acquisition of the top three ions with charge greater than one. MS/MS scans for each ion used an isolation window of ,3 Da, a maximum of 2 s per precursor, and dynamic exclusion for 120 s within 1.2 Da. DDA data were converted to searchable files using ProteinLynx Global Server 2.4 (Waters Corporation) and searched against NCBI mouse database (February 2011) using Mascot server 2.2 with the following parameters: maximum one missed cleavage site, carbamidomethylation at Cys residues as fixed modification and Met oxidation, Nterminal acetylation, Asn, Gln deamidation as variable modifications. Precursor ion mass tolerance was set to 20 ppm, while fragment mass tolerance to 0.2 Da. Acceptance criteria for protein identification required identification of at least two peptides for each protein with a confidence interval percentage, (CI %) of over 99.9%, corresponding to a false discovery rate of 0.1%.

Immunoblotting
To determine the presence of Ca (V) 1.2 and Ca (V) 1.3 proteins in mouse lens, lens membrane-enriched protein fractions were isolated as mentioned above from P21 mice. Similarly cytosolic and membrane-enriched protein fractions were prepared from control (DMSO incubated) and drug treated lenses. Equal amounts of proteins were resolved by SDS-PAGE using 5, 8 or 15% acrylamide containing gels, followed by electrophoretic transfer to nitrocellulose membranes. Membranes were blocked with 5% milk protein, in Tris buffer containing 0.1% Tween-20 (washing buffer) and then incubated overnight with appropriate monoclonal or polyclonal primary antibodies, at a 1:1000 dilution, followed by the respective secondary antibodies as described earlier. [7] Immunoblots were then developed using enhanced chemiluminescence ECL prime reagents (Amersham Biosciences, Buckinghamshire, UK; C. No. RPN2232), and scanned using a FOTODYNE Gel Doc scanner equipped with TL100 software, densitometry analysis was carried out using ImageJ software.

MLC Phosphorylation
Following harvest, control (DMSO treated) and drug treated lenses were incubated with 10% ice cold trichloroacetic acid (TCA) for 5 minutes, washed with several changes of ice cold water, followed by diethyl ether, air dried and stored at -80C until further processing. For extraction of total protein, lens specimens were homogenized in 8 M urea buffer containing 20 mM Tris, 23 mM glycine, 10 mM dithiothreitol and saturated sucrose along with protease and phosphatase inhibitors. Lysate was centrifuged at 16,000 xg for 15 min, and protein content analyzed in the supernatants. Equal amounts of protein (100 mg) from the supernatants derived from control and drug treated samples were resolved by urea/glycerol-polyacrylamide gel electrophoresis and immunoblotted using anti-phospho-specific MLC rabbit polyclonal antibody, as we described earlier. [42] Immunoblots were developed using enhanced chemiluminescence ECL prime reagents. Total MLC levels were also determined in the same samples by immunoblot analysis using anti-MLC antibody.

Immunofluorescence
Tissue sections derived from either paraffin embedding or cryofixation were used in immunofluorescence analyses. The paraffin embedded lens sections were de-paraffinized and immunostained as we described earlier. [7] Slides were viewed and photographed using a Nikon Eclipse 90i confocal laser-scanning microscope. For double labeling, appropriate polyclonal and monoclonal antibodies were mixed and incubated with the tissue specimens as described above in conjunction with Alexa Fluor 488 and 568 conjugated secondary antibodies and Hoechst 33258.
For Cryosection immunostaining, freshly dissected mouse lenses (p21-old) were fixed in 4% paraformaldehyde in phosphatebuffered saline (PBS) at 4C, overnight. Fixed lens tissues were then transferred sequentially into 5% and 30% sucrose in PBS. Tissues embedded in optimal cutting temperature medium (OCT) were sectioned using a cryotome (Microme HM 550, Walldorf, Germany), equatorial or sagittal plane (5-10 micron thick) tissue sections were mounted on charged glass slides and stored at -80C until further processing. Cryosections were immunostained as we described earlier using appropriate primary and secondary antibodies. [7].
Mouse lens epithelial cells grown on gelatin-coated glass coverslips were stained for F-actin and phospho-MLC using rhodamine phalloidin and phospho-MLC antibody, respectively as previously described. [77].

TUNEL Assay
In situ terminal transferase dUTP nick end labeling (TUNEL) staining was performed using an ApopTag Plus Fluorescein kit to evaluate and compare apoptotic cell death in mouse lens sections derived from felodipine treated and control lenses as described previously. [79] Apoptotic cells were detected using a fluorescence microscope (Zeiss Axioplan-II). In the same specimens, cell nuclei were labeled with Hoechst stain. The DNase-1 treated lens sections were used as a positive control for this assay.

Statistical Analysis
Where required the Student's t-test was used to determine significance of differences noted between control (DMSO treated lenses) and drug treated lenses. Values are provided as Mean 6 S.E.M, and a P,0.05 was considered statistically significant for all comparisons. Figure S1 Cell viability and cytotoxicity in mouse lens epithelial cells treated with felodipine. To evaluate felodipine-induced effects on cell viability and cytotoxicity, mouse lens primary epithelial cells were grown on gelatin-coated plastic dishes, serum starved overnight and treated with felodipine (10 mM) for 6 or 24 hrs. Following this, cells were rinsed with DMEM and treated with fluorescein diacetate (2 mg/well) and propedium iodide (0.6 mg/well) for 3 min prior to imaging the green fluorescence derived from in vivo fluorescein diacetate hydrolysis and propidium iodide red nuclei staining in live cells. As shown in the figure, both control and felodipine treated cells were found to exhibit comparable viability and fluorescein diacetatebased green fluorescence. Additionally, there was no propidium iodide incorporation into cell nuclei in either the drug treated or the control cells, confirming absence of drug-induced cytotoxic effects. Scale bar indicates image magnification. The images are representative of triplicate analyses. (TIF)