Increased Association of Deamidated αA-N101D with Lens membrane of transgenic αAN101D vs. wild type αA mice: potential effects on intracellular ionic imbalance and membrane disorganization

Abstract We have generated two mouse models, in one by inserting the human lens αAN101D transgene in CRYαAN101D mice, and in the other by inserting human wild-type αA-transgene in CRYαAWT mice. The CRYαAN101D mice developed cortical cataract at about 7-months of age relative to CRYαAWT mice. The objective of the study was to determine the following relative changes in the lenses of CRYαAN101D- vs. CRYαAWT mice: age-related changes with specific emphasis on protein insolubilization, relative membrane-association of αAN101D vs. WTαA proteins, and changes in intracellular ionic imbalance and membrane organization. Methods Lenses of varying ages from CRYαAWT and CRYαAN101D mice were compared for an age-related protein insolubilization. The relative lens membrane-association of the αAN101D- and WTαA proteins in the two types of mice was determined by immunohistochemical-, immunogold-labeling-, and western blot analyses. The relative levels of membrane-binding of recombinant αAN101D- and WTαA proteins was determined by an in vitro assay, and the levels of intracellular Ca2+ uptake and Na, K-ATPase mRNA were determined in the cultured epithelial cells from lenses of the two types of mice. Results Compared to the lenses of CRYαAWT, the lenses of CRYαAN101D mice exhibited: (A) An increase in age-related protein insolubilization beginning at about 4-months of age. (B) A greater lens membrane-association of αAN101D- relative to WTαA protein during immunogold-labeling- and western blot analyses, including relatively a greater membrane swelling in the CRYαAN101D lenses. (C) During in vitro assay, the greater levels of binding αAN101D- relative to WTαA protein to membranes was observed. (D) The 75% lower level of Na, K-ATPase mRNA but 1.5X greater Ca2+ uptake were observed in cultured lens epithelial cells of CRYαAN101D- than those of CRYαAWT mice. Conclusions The results show that an increased lens membrane association of αAN101D-−relative WTαA protein in CRYαAN101D mice than CRYαAWT mice occurs, which causes intracellular ionic imbalance, and in turn, membrane swelling that potentially leads to cortical opacity.


Background
Although the cornea is the primary refractive tissue performing 70-80% of refraction of the eye, the major function of the lens is in accommodation and to partly help in the refraction. The lens accommodative function gradually diminishes with age, and is almost completely lost at age of > 50 years. The lens transparency plays an important role in focusing light on to the retina, but this role is gradually lost as it develops age-related opacity. Several unique factors maintain lens transparency for up to > 60 year of our life time. These include: cellular homeostasis among only two types of cells (epithelial and fiber cells) [1], an orderly terminal differentiation of epithelial to fiber cells with precise organelles loss [2], the unique interactions among crystallins [3], with almost no protein turnover [4], the specialized lens metabolism [5], specific interactions among α-crystallin and membrane [6], the precise maintenance of intracellular and extracellular ionic concentrations [7], the low levels of cellular water and oxygen in the lens inner cortex and nuclear regions [8], and a unique membrane lipid composition [9]. Alterations among some of these lens unique factors play direct or indirect roles in pathogenesis of cataracts (e.g., pediatric-and age-related cataracts). However, additional cataract-causative factors are also identified, which include mutations in crystallins [10], oxidative insults of crystallins, the loss of redox balance of glutathione [11], extensive truncations of α-, β-, and γ-crystallins [12][13][14][15][16][17][18][19][20], a variety of post-translational modifications with deamidation as being the most abundant [21][22][23][24][25], and the loss of membrane integrity [7,26,27]. These factors individually or in combination also cause lens opacity through altered lens cellular structures and contents, ionic imbalance, increased water and oxygen levels, loss of natural interactions among crystallins, and crystallins' unfolding, degradation and cross-linking.
Our focus in this study is the potential roles of deamidation of Asn 101 of αA crystallin to Asp that introduces negative charges and shown to alter their hydrophobicity, tertiary structures, crystallin-crystallin interactions, and leads to aggregation and cross-linking [21][22][23][24][25][26][27]. In this study, the deamidation of Asn 101 to Asp in in a mouse model was studied to determine phenotypic and molecular changes within the lens due to deamidation of a single nucleotide change in CRYAA crystallin gene. This site was chosen because our past study showed that only deamidation of Asn localized at specific sites in crystallins (e.g., deamidation of N101 but not of N123 residues in αA-crystallin [24], and of N146 but not of N78 of αB-crystallin) exhibited the above-described deamidation-induced effects [25]. To show the potential effects of deamidation in vivo, we have generated mouse models by inserting the human lens αA-N101D transgene in CRYαA N101D mice, and human lens wild-type αA-transgene in CRYαA WT mice (to act as a control). The CRYαA N101D mice developed cortical cataract at about 7-months of age relative to CRYαA WT mice [28,29]. This model showed for the first time that in vivo expression of the deamidated αAN101D caused cortical lens opacity, which was due to the disruption of fiber cell structural integrity and protein insolubilization as aggregation [28]. The comparative RNA sequencing and Ingenuity Pathway Analyses (IPA) of lenses from 2-and 4-months old CRYαA N101D -and CRYαA WT mice showed that the genes belonging to cellular assembly and organization, cell cycle and apoptosis networks were altered in αA N101D lenses [29]. This was accompanied with several cellular defects in αA N101D lenses that included defective terminal differentiation (increased proliferation and decreased differentiation) of epithelial cells to fiber cells, and reduced fiber cells denucleation and expressions of Rho A and Na, K-ATPase (the major lens membrane-bound molecular transporter) [29]. The findings also suggested the potential role of lens intracellular ionic imbalance as the major reason for the development of cataract [29]. The above findings suggested that the altered intracellular ionic imbalance could be due to potential loss of membrane integrity that caused cortical opacity at about 7-months of age in the CRYαA N101D mouse model. Therefore, the focus of the present study was to determine whether an increased membrane-association of αA N101D potentially compromises membrane integrity, and causes an ionic imbalance and leads to cataract development.

Ethics statement
All animal experiments were performed per protocols approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Alabama at Birmingham (Protocol no. 130208393). Mice were housed in a pathogen-free environment at the facility of the University of Alabama at Birmingham.

Materials
Unless stated otherwise, the molecular biology-grade chemicals were purchased from Millipore-Sigma (St. Louis, MO, USA) or Fisher (Atlanta, GA, USA) companies. The Rabbit polyclonal anti-human aquaporin-0 (AQP0) antibody was purchased from Alpha Diagnostics (San Antonio, TX, USA). Additional commercial sources of various chemicals and antibodies used in the study are described throughout the text.

Generation of transgenic mice
The mouse model that expresses a human αA-crystallin gene in which Asn-101 was replaced with Asp is referred to as αA N101D -transgenic mouse model. This model has been considered to be "deamidated" in this study, and the mice expressing αAN101D-transgene is referred here as CRYαA N101D mice. Both mouse models (human lens αA N101D transgenic-and human wild-type αA-transgenic mouse models were generated in Dr. Om Srivastava's laboratory [28]. Independent transgenic (Tg) mouse lines were established from transgenic founders using C57BL/6 mice (Harlan Laboratories, Indianapolis, IN). αAN101D protein expression constituted about 14 and 14.2% of the total αA in the lens WS-and WI-proteins of the αA N101D transgenic mice, respectively [28]. The mouse lenses were extracted after the mice were euthanized using the CO 2 procedure as per approved method by the Institutional Animal Care and Use Committee (IACUC) of the University of Alabama at Birmingham (Protocol no.130208393). Adult (2-3 months) wild type mice (C57BL6) were obtained from the university breeding colony. Animals were kept under a 12/12 h light-dark cycle and had ad libitum access to food and water. We have used three mice from each group of CRYαA WT mice control and αA N101D mice in all the experiments described below.

Isolation of water soluble (WS)-and water insoluble (WI)proteins from mouse lenses
The WS-and WI-protein fractions from lenses of desired ages of CRYαA WT -and CRYαA N101D mice were prepared as previously described by us [28]. All procedures were performed at 5°C unless specified otherwise. The lenses were removed under a dissecting microscope and placed in 5°Ccold buffer A (5 mM Tris-HCl, 1 mM EDTA, 1 mM DTT, pH 7.8, and protease inhibitor cocktail [Roche Life Science, Indianapolis, IN]), and centrifuged at 14,000 x g for 15 min at 5°C to separate the WS-and WI-protein fractions. The supernatant (WS-protein fraction) was collected, and next the pellet (WI-protein fraction) was resuspended in buffer A, centrifuged as above. The recovery of WS-and WIprotein fractions was repeated twice after centrifugation, and the WS supernatants after each centrifugation steps were pooled. The final WI-protein pellet was solubilized in 5 mM Tris-HCl, pH 7.5, containing 8 M urea, 5 mM EDTA, and 5 mM EGTA. The 8 M urea concentration was diluted to 4 M urea with buffer A prior to centrifugation as above. The protein concentrations in these fractions were determined by using a kit (Pierce Biotechnology-Thermo Fisher) using bovine serum albumin as a standard.

Membrane isolation from mouse lenses
The membranes from lenses of 1-and 6-month-old CRYαA WT and CRYαA N101D mice were prepared as described previously [30,31]. Lenses of identical ages from both types of mice were homogenized in buffer B (0.05 M Tris-HCl, pH 8.0 containing 5 mM EDTA, 1 mM DTT, 150 mM NaCl, and protease inhibitor cocktail [Roche, Indianapolis, IN]), and the preparations were centrifuged at 100,000 x g for 30 min using Beckman TL 100 centrifuge with a TLA 100.3 rotor. The supernatant was collected, and the pellets were washed twice with the above buffer B and centrifuged as above. This was followed by three additional washes with buffer B containing 8 M urea and centrifugation as above after each wash. Next, the pellet was washed twice with water and centrifuged as above. The pellet was then washed with 0.1 M cold (5°C) NaOH [30,31]. A final wash of pellet was with water and centrifugation as above to recover the purified lens membrane preparations as pellets.
Purification of recombinant WTαA-and αA-N101Dcrystallins, their conjugation with Alexa Fluor 350 and membrane binding The WTαA-and αA-N101D mutant proteins were expressed in E.coli and purified by a Ni-affinity column chromatographic method as previously described by us [28]. Each protein was labeled with Alexa-350 using a protein labeling kit as suggested by the manufacturer (Molecular Probes, Carlsbad, CA). The binding of Alexa 350-conjugated WT αAand αA-N101D mutant proteins to mouse lens membrane (isolated from C57BL non-transgenic mice) was determined as previously described [32,33]. During the binding assay, the purified lens membrane (containing 2.5 mg protein; isolated from 1 to 3-month old non-transgenic C57 mice) was incubated with increasing but identical concentrations of either Alexa-labelled WT αAor αA-N101D proteins at 37 ο C for 6 h. Next, the incubated preparations were centrifuged at 14,000 X g and the supernatant and pellet (membrane fraction) recovered. After washing the membrane fraction with water and centrifugation as above, the relative levels of fluorescence of membranes incubated with WT αAand αA-N101D mutant proteins were determined using Perkin Elmer Multiplate Reader (Model Victor1420-04).
Determination of intracellular Ca 2+ in epithelial cells in culture from lenses of CRYαA WT -and CRYαA N101D mice To culture epithelial cells, six 5-months old lenses from CRYαA WT -and CRYαA N101D mice were excised and incubated with 0.25% trypsin at 37 ο C for 2.5 h in an incubator with 5% CO 2 -humidified air. Next, the lens cells in trypsin solution were centrifuged at 1200 rpm for 3 min, and trypsin (in the supernatant) was discarded. The lens epithelial cells (recovered as pellet) were suspended in medium 199 (Thermo Fisher Scientific, Grand Island, NY) containing 10% fetal calf serum (Hyclone, Logan, Utah) and 1% antibiotic-antimycotic solution (Thermo Fisher Scientific, Grand Island, NY) in 12-well plates (Corning, Franklin Lakes, NJ). After 24 h, the unattached cells were discarded by washing with the above medium. The old medium was replaced with fresh medium after every 48 h, and the cells were allowed to grow for 7 to 10 days until confluent. Next, the confluent cells were trypsinized and seeded in 12-well plates for intercellular Ca 2+ determination and were allowed to grow for 24 h.
The cells were washed with medium 199 without phenol red, incubated in calcium orange dye (Thermo Fisher Scientific, Grand Island, NY) at a final concentration of 4 μM for 30 min at room temperature as instructed in the manufacturer's protocol. After 30 min, the cells were washed with the above medium, and Ca 2+ indicator was examined under a microscope (Leica DMI 4000B) using a Texas Red filter.

Western blot and Immunohistochemical analyses
The WS-and WI-proteins and membrane fractions isolated from lenses were analyzed for their immunoreactivity with anti-aquaporin-0 antibody (to visualize the membrane intrinsic protein), and Mouse anti-His monoclonal antibody ([Novagen, Madison, WI], to visualize WTαA and αA N101D ) during Western blot analysis. The SDS-PAGE analysis was carried out as described by Laemmli [34].
The confocal immunohistochemical analysis of lens axial sections of WTαA and αA N101D was carried out as previously described by us [28]. The analysis was performed at the High-Resolution Imaging core facility of the University of Alabama at Birmingham.

Localization by Immunohistochemical-transmission Electron microscopic method
The analysis was performed at the High-Resolution Imaging core facility of the University of Alabama at Birmingham. His-tagged αA WT -and αA N101D -crystallins were localized in lens cells by an Aurion immunogold method and the reagents used were Aurion Conventional Immunogold reagents (Electron Microscopy Science [PA]). Lenses of desired ages were fixed in phosphatebuffered saline, pH 7.4 containing 4% paraformaldehyde and 0.05% glutaraldehyde (Electron Microscopy Sciences, Hatfield, PA) for 2 h at room temperature, and then overnight at 4°C. The fixed lenses were washed with water (Millipore, Billerica, MA). Samples were dehydrated by ascending ethanol gradient series followed by infiltration overnight at 4°C with absolute ethanol: London Resin (LR) white (1:1). Next, the samples were incubated overnight with pure LR white resin on a rotating platform. The lenses were removed and transferred to gelatin capsules containing fresh LR white and allowed to polymerize for 24 h at 45-50°C. Ultra-thin (silver gold to light gold) LR white lens sections were collected on nickel mesh grids. The color of sections was silver-gold to light gold, and based on their color, the thickness was estimated to between 70 and 80 nm. For immunogold-labeling, the protocol as described in Electron Microscopy Sciences (Hatfield, PA) was precisely followed. To inactivate aldehyde groups present after aldehyde fixation, the samples on grids were incubated on 0.05 M glycine in PBS buffer for 10-20 min. Next, the grids were transferred onto drops of the matching Aurion blocking solution for 15 min, and then were washed for 15 min in incubation solution (PBS containing 0.1% bovine serum albumin and 15 mM NaN3, pH 7.3). This was followed by a 2X wash in incubation buffer, each time for 5 min. The grids were incubated with two primary antibodies (Mouse anti-His monoclonal antibody and Rabbit anti-aquaporin-0 polyclonal antibody for 1 h. In controls, the primary antibodies were omitted. The grids were then washed 6X (5 min each time) with the incubation solution and transferred to following secondary antibody conjugates {(goat anti-Rabbit EM grade conjugate 25 nm diameter) and (goat antimouse EM grade conjugate 10 nm diameter)} and were incubated for 30 min to 2 h. The grids were washed on drops of incubation solution for 6X (5 min each time). The grids were washed twice with PBS for 5 min, postfixed in 2% glutaraldehyde in PBS for 5 min, and finally washed with distilled water and contrasted according to standard procedures. Lens sections were imaged using an FEI 120kv Spirit TEM (FEI-Thermo Fisher), and images were collected using an AMT (AMT-Woburn, MA) digital camera.

RNA extraction and reverse transcription-quantitative polymerase chain reaction (RT-PCR [qPCR])
RNA was extracted with Trizol reagent (Invitrogen) from cultured lens epithelial cells from CRYαA N101D and CRYαA WT mice, and all the samples were analyzed in triplicates. Real-time PCR quantifications were performed using the BIO-RAD iCycler iQ system (Bio-Rad, Hercules, CA), using a 96-well reaction plate for a total volume of 25 μL. RNA was extracted as described above. Primers were designed using Primer3 for the following genes: Atp1a2 Forward-5'CGGGAGCCATAAGGGTTTGT 3′, and Atp1a2 Reverse-5'GCACTGACTTGGCTGTTGTG 3′.
The ACTB gene was used for normalization. The reaction mixture included 12.5 μL of Real-Time SYBR Green PCR master mix, 2.5 μL of reverse transcription product, 1 μL of forward and reverse primer and 8 μL of DNase/RNase free water. The reaction mixtures were initially heated to 95°C for 10 min to activate the polymerase, followed by 40 cycles, which consisted of a denaturation step at 95°C for 15 s, annealing at 55°C for 60 s and an elongation step at 72°C. The qRT-PCR data were analyzed by the comparative ΔCt method.

Results
Age-related protein Insolubilization in lenses of CRYαA N101D and CRYαA WT mice To determine at what age there is change in the protein profiles in lenses of CRYαA N101D and CRYαA WT mice occurred, a comparative analysis of WS-proteins and WI-proteins from the lenses of the two types of mice of different ages was carried out (Fig. 1). The WS-and WIproteins from lenses of different ages (1-, 3-, 4-, 5-and 7-months) were analyzed by SDS-PAGE. The WS-protein profiles from the lenses of the CRYαA N101D and CRYαA WT mice were almost identical until 3-months of age, except lens preparations from ages of 4-, 5-and 7-months of CRYαA N101D mice exhibited relatively greater levels of aggregated protein of M r > 30 kDa and higher relative to same-aged lenses from CRYαA WT mice (Lanes 4 and 5 in Fig. 1b). Additionally, on quantification, the relatively increasing levels of WS-proteins showed age-related water insolubilization beginning at 4-months of age in the lenses of αA N101D mice (Table 1). Between 4-to 7-months of age, relatively about 5 to 10% higher proteins became insoluble in lenses of CRYαA N101D. To determine changes in individual crystallins due to their insolubilization, the WS-protein fraction from 7-month-old lenses was fractionated by a size-exclusion HPLC using a G-4000PWXL column (Tosoh Biosciences, fractionation range of protein with M r 's between 1X10 4 to 1X10 7 Da). The comparative protein elution profiles at 280 nm of 7-month old lenses of αA N101D -mice showed an increased protein in the void volume peak (representing WS-HMW proteins), and reduced βand γ-crystallin peaks relative to lenses of CRYαA WTmice (the differences shown in green in Fig. 2a). The void volume peak in WS-protein fraction was also higher in the 7-month old lenses relative to 1-month old lenses of αA N101D -mice (Results not shown), suggesting a relatively increased HMW protein aggregate formation with aging. On western blot analysis of the individual column fractions nos. 6 to 9 (constituting the void volume-HMW-protein peak) with an anti-His antibody, the levels of Hisimmunoreactive protein were higher in 7-month old CRYαA N101D lenses compared to the identical aged CRYαA WT lenses (Fig. 2b). Additionally, because the immunoreactive peak in the WT lenses was in the fractions no. 8 and 9 whereas it was in the fractions no. 7 and 8 in the αA N101D lenses that suggested that the HMW proteins ). Note that a greater insolubilization and aggregated proteins (M r > 30 kDa) were seen in WI-proteins of lenses of 4month and older CRYαA N101D mice compared to age-matched lenses from CRYαA WT mice. The Table 1. shows quantification of protein levels in the WS-and WI-protein fractions of lenses of different ages from CRYαA N101D and CRYαA WT mice. Gel images are not cropped of αA N101D lenses showed a higher molecular weights relative to the HMW proteins from WT lenses. On quantification of Western blot images with Image J (Fig. 2c), the intensity of the immunoreactive HMW proteins of αA N101D was about 20% greater relative to WT lenses. Together, the results suggested a greater aggregation with higher M r of the HMW-proteins in CRYαA N101D lenses compared to the identical aged CRYαA WT lenses.

Identification proteins present in water insoluble-urea soluble (WI-US) -and water insoluble urea insoluble (WI-UI) protein fractions of lenses of CRYαA WT and CRYαA N101D mice
To Identify the insolubilized proteins in WTαA vs. αA N101D lenses, the WI-proteins from 5-month-old mice were further fractionated into WI-US-and WI-UIprotein fractions, and examined by SDS-PAGE (Fig. 3) followed by their protein compositional analysis by mass spectrometry. SDS-PAGE analysis showed that both WI-US-and WI-UI-protein fractions from CRYαA N101D lenses contained greater levels of protein species including aggregated proteins (M r > 30 kDa) [Identified as a, and c in Fig. 3] relative to the same fractions from lenses of CRYαA WT mice (Identified as b and d in Fig. 3). The mass spectrometric analysis was carried out at the following two levels: (i) In the first level analysis, determination of the total protein compositions in the WI-USand WI-UI protein fractions of the two types of lenses (Supplemental Tables A [Comparative protein Table D]. The rationale of the two levels of analysis was to determine the relative proteins compositions due to the greater insolublization of proteins in CRYαA N101D lenses relative to CRYA AWT lenses (Fig. 1, Table 1). Our expectation was that the level 1 comparative examination would identify the total proteins that underwent insolubilization, and existed in the US-and UI-protein fractions, whereas the level 2 analysis would selectively identify those proteins that formed aggregates (M r > 30 kDa) in the US-and UI-fractions. The rationale was that the information would implicate potential roles of specific crystallins in the aggregation and therefore, in the cataractogenic mechanism. The proteins detected in the WI-US-protein fractions of CRYαA N101D lenses but were absent in the WT lenses are described in Supplemental Table A. Together, the results show that the WI-US fraction of CRYαA N101D lenses was enriched in several histones, which could be due to the lack of denucleation relative to WT lenses. Absence of Retinal dehydrogenase in transgenic lens fraction. The proteins present in the WI-UI-protein fractions of CRYαA N101D lenses but were absent in WT lenses are described in Supplemental Table B. In summary, the results again show that the majority of histones that existed in CRYαA N101D lenses were absent in the WT lenses, which could be due to the lack of denucleation in the lenses of former mice. Also, specifically αBand βB2-crystallin became insoluble as their levels were higher even in the WI-UI-fraction of lenses of CRYαA N101D relative to WT lenses. As noted above, the purpose of the second level of mass spectrometric analysis was to elucidate the comparative compositions of aggregated proteins (M r > 30 kDa) in WI-US-and WI-UI-protein fractions of CRYαA N101D and CRYαA WT lenses [Supplemental Tables C and D]. On comparison, the major proteins present as aggregates (M r > 30 kDa) in WI-US fraction of CRYαA N101D but absent in CRYαA WT were (Supplemental Table C): βB3and γCcrystallins, collagen alpha-1(IV) chain and -alpha-2(IV) chain and nestin. In contrast, the exclusively present major proteins in WI-US fraction of CRYαA WT were: γC-, γD-, γE-γF-crystallins. The above list describes the selective proteins that were water insoluble-urea soluble and became the part of the complexes with M r > 30 kDa in CRYαA N101D lenses. The greater abundance of αA-, and βB1-crystallins in the aggregated form suggested their potential involvement in the aggregation process along with βB3and γCcrystallins.
On comparison of major proteins that existed in WI-UI protein fraction as > 30 kDa aggregates in CRYαA N101D not in the CRYαA WT included [Supplemental Table D]: γB-, γDand γE-crystallins, and nestin. In the WI-UI fraction, the greater abundance of proteins in CRYαA N101D compared to CRYαA WT were: αA-crystallin and lens fiber Fig. 2 Size-exclusion HPLC (using a G-4000PWXL column) and SDS-PAGE analysis of WS-HMW proteins eluted in the void volume. a HPLCprotein elution profiles at 280 nm of WS-proteins from lenses of 5-month-old CRYαA N101D -and CRYαA WT mice. The green region shows the difference in the A280 profiles of WS-proteins from the two type of lenses. b Western blot analysis of the void volume peaks (constituted by the fraction no. 6 to 9 in [A]) following HPLC separation of WS-proteins from lenses of CRYαA N101D -and CRYαA WT mice. Note that in the CRYαA WT lenses, the αA-immunoreactive bands were in fractions no. 8 and 9 whereas it were in fraction no. 7 and 8 in CRYαA N101D lenses, suggesting a higher M r of HMW proteins in the latter. c Quantification of the Western Blot using Image J. Gel images are not cropped intrinsic proteins. Together, the results showed that the proteins that remained urea insoluble and were possibly associated with the membrane of CRYαA N101D lenses included: γB-, γDand γE-crystallins, and nestin (Nestin is an intermediate filament protein).
Increased association of αAN101D with Lens membrane in the outer cortical Fiber cells relative WTαA in CRYYAA WT lenses Our previous report [28] showed an increased levels and abnormal deposition of αA N101D within the outer cortical region in CRYαA N101D lenses compared CRYαA WT lenses. This suggested a relatively greater membrane binding of αA N101D, which was further investigated in experiments as described below.

(i). Immunohistochemical Analyses of Lenses from
CRYαA N101D and CRYαA WT Mice The purpose of the experiments was to determine relative levels of αAN101D and WTαA in the outer cortical regions of CRYαA N101D -vs. CRYαA WT lenses. This was examined by immunohistochemical analysis of 5months old lenses of the two types of mice using anti-His monoclonal (for detection of WTαA and αA N101D [green fluorescence]) -and polyclonal anti-aquaporin 0 (for membrane detection [red fluorescence])-antibodies (Fig. 4). The axial sections (at 10X magnification) showed an irregular and greater deposition of Histagged αA (Green) in the lens outer cortex of CRYαA N101D mice (Shown by an arrow in Fig. 4a) relative to CRYαA WT mice (Shown by an arrow in Fig. 4b).
Similarly, the equatorial sections (at 40X magnification) also exhibited a greater immunoreactive green fluorescence in the outer cortex of the CRYαA N101D lens relative to the CRYαA WT lens (shown by arrows in Fig. 4c  and d). Together, the results suggested the abnormally greater levels of association of αA N101D in the outer cortical regions, and potentially with the fiber cell membranes in the CRYαA N101D lenses relative to those of CRYαA WT lenses. The rationale for the next experiment was that if the greater membrane-association of αA-N101D occurs in vivo in CRYαA N101D lenses compared to CRYαA WT lenses, the difference in their levels could also be determined in the purified membrane fractions by western blot analysis. The expectation was that following the step-wise membrane purification by using 8 M urea (to dissociate noncovalently-bound membrane proteins), and by the final wash with 0.1 N NaOH (to remove non-membranous extrinsic proteins) [30,31], the purified membrane would show relative levels of membrane-association of αA N101D vs. WTαA in the two types of mice. To normalize the levels of the relative association during the membrane preparations, two lenses of 1-month-old and two lenses from 6-month old from CRYαA N101D and CRYαA WT mice were identically processed, using identical volumes of buffers at each steps during membrane purification (See Methods). Next, Western blot analysis using anti-His-and anti-aquaporin 0-antibodies were used to determine the relative levels of membrane-association of WTαA and αAN101D at different purification steps (Results not shown). To simplify the western blot results of fractions recovered during the sequential steps of membrane purification, only the results of immunoblots with anti-His antibody but not with anti-aquaporin-0 are shown in Fig. 5. However, the western blot profiles with anti-aquaporin-0 were almost identical to anti-His antibody results. In Fig.   5, a, b, e and f show SDS-PAGE analysis followed by Coomassie blue-stained gels exhibiting relative levels of protein bands in preparations at different membrane purification steps in lenses at two different age groups (1 and 6 months). In  membrane preparation. Lane 9 of 6-month old lenses represents the crude lens WS-homogenate. The results show that the green fluorescence representing WTαA in CRYαA WT mice was entirely disappeared on urea solubilization in 1-and 6-month old lenses (lanes 1 to 5 in both left and right panels), whereas it was still present in these lenses until 0.1 N NaOH wash (lane 6 in left and right panels). In contrast, the green fluorescence still existed in lane 6 of membranes from 1-and 6-month-old CRYαA N101D lenses. Together, the results suggest that αA N101D was tightly bound and at the higher levels to lens membrane of CRYαA N101D lenses relative to CRYαA WT lenses.
On Image J-quantification of the Western blots (Fig. 5 i  and j), the lanes 4 and 5 (urea soluble fractions) of 1-month old lenses showed higher levels (2.5X) of immunoreactivity with anti-His antibody in the CRYαA N101D lenses (shown  Fig. 5. i Relative levels of immunoreactive WTαA lenses (blue) and αA-N101D αA-(red) during membrane purification from 1-month old lenses. j Relative levels of immunoreactive WTαA lenses (blue) and αA N101D αA-(red) during membrane purification from 6-months old lenses. Note that relatively higher levels of αA N101D than WTαA was associated with purified membranes in lanes 4 (in 1-month old) and lane 5 (in both 1-and 6-monthold) of the two types of lenses. Gel images are not cropped in red) compared to those from CRYαA WT lenses (blue). Similarly in Fig. 5j, among the lanes 4 and 5 containing same fractions from 6-month old lenses (as described in 1month old lenses), the lane 5 showed a greater immunoreactive level of CRYαA N101D lenses (red) compared to CRYαA WT lenses (blue). Additionally, the lane 6 (representing membrane remaining after two urea washes, right panel) of 6-month CRYαA N101D lenses exhibited about 2X greater immunoreactivity than CRYαA WT lenses (Quantification results not shown). Together, the results show that relative to CRYαA WT , higher levels of CRYαA N101D were tightly associated with the lens membranes of 1-and 6month old CRYαA N101D mice. To examine whether αA N101D show a greater binding affinity to the lens membrane relative to WTαA-crystallin, the binding of the two recombinant proteins to purified lens membrane was determined. The recombinant WTαA-and αAN101D proteins were labeled with Alexa 350 using a protein labeling kit by the procedure described by the manufacturer (Molecular Probes, Thermo fisher Scientific). The two labeled-proteins were purified by a size-exclusion HPLC column and were analyzed by SDS-PAGE. Figure 6a shows the Coomassie blue-stained WT αA (lane 1), αA N101D protein (lane 2), and the purified lens membrane from nontransgenic C57 mice (lane 3). The Fig. 6b shows the images of the two Alexa 350-labeled proteins under a UV trans-illuminator [Lane 1: Images of Alexa 350-labeled WTαA, and lane 2: Alexa 350-labeled αAN101D). During the binding assay, the purified lens membrane (containing 2.5 mg protein; isolated from 1 to 3-month old non-transgenic C57BL mice) was incubated with increasing but identical concentrations of either Alexa-labelled WT αAor αA N101D proteins at 37 ο C for 6 h (See details in Methods). A relatively higher levels (> 1.5X) of binding of αAN101D protein relative to WTαA protein with membrane preparation was observed (Fig. 6c). The values reported are the average of triplicate assays. (iv).Immunogold-Labeling for Relative Localization of αA-WT and αAN101D in Lens Membranes of CRYαA N101D and CRYαA WT Mice To ascertain the relative levels association αAN101D vs. WTαA to the lens membrane in vivo, the immunogoldlabeling experiment was carried out (See details in Methods). (A) and (B) in Fig. 7 show lens membranes from CRYαA N101D and CRYαA WT mice at 500 nm magnification, and (C) and (D) from these lenses at 100 nm magnification, respectively. The bigger gold particles (25 nm, red arrows) the smaller gold particles (10 nm, yellow arrows) represented the aquaporin-0 and the His-tagged αAN101D and WTαA, respectively. As shown in the representative images in (A) to (D), the 25 nm gold particles (representing aquaporin-0, identified by red arrows) were bound to membranes. On counting the membrane-associated 10 nm particles (representing His-tagged αAN101D and WTαA), almost the same c Binding of a WT αA, and αA N101D with purified lens membrane (2.5 mg protein; isolated from 1-to 3-month old non-transgenic C57 mice). During the binding assay, the protein mixtures were incubated with increasing but identical concentrations of either Alexa-labelled WTαA-or αA-N101D at 37 ο C for 6 h, centrifuged at 14,000Xg and the supernatant and pellet (membrane fraction) recovered. After washing the membrane fraction with water and centrifugation as above, the relative fluorescence of membranes incubated with WT αAand αA-N101D mutant proteins was determined. The values reported are the average of triplicate assays numbers of the particle were found to be associated with membranes of both CRYαA N101D and CRYαA WT lenses, suggesting that the His-tagged αAN101D and WTαA were bound to the membranes of the two types of lenses. Our previous study [28] showed that αAN101D protein constituted about 14 and 14.2% of the total αAcrystallin in the WSand WI-proteins, respectively in the lenses of CRYαA N101D mice. Therefore, an argument can be made that although an almost equal number of 10 nm and 25 nm particles were associated with membranes of the two type of lenses, a higher number of gold particle representing αAN101D relative to WTαA were associated with the membrane.
Another interesting observation was that the membranes of CRYαA N101D lenses were about 2X more swollen relative to those of CRYαA WT lenses [Fig. 7, compare (A) to (B) and (C) to (D)]. The width of the membrane was quantified using Image J as shown in Fig.  7e. The swelling could represent water intake within the lens cells due to the potential ionic imbalance in the CRYαA N101D lenses compared to CRYαA WT lenses. Such a possibility of ionic imbalance was further determined as described below.
Na, K-ATPase and Ca 2+ levels in cultured epithelial cells from lenses of CRYαA N101D and CRYαA WT mice Sodium-potassium-adenosine triphosphatase (Na, K-ATPase) has been recognized for its role in regulating electrolyte concentrations in the lens, and the electrolyte balance is vital to lens transparency [35,36]. In addition, calcium has been reported to control both sodium and potassium permeability through lens membranes [37]. In our previous study [29], we showed that the expression of Na,K-ATPase at the protein level was drastically reduced in CRYαA N101D lenses relative to CRYαA WT lenses. Next, the levels of Na, K-ATPase mRNA, and Ca 2+ levels were determined in cultured epithelial cells from lenses of CRYαA N101D and CRYαA WT mice. Both (A) and (B) in Fig. 8 show intracellular Ca 2+ levels in the presence of calcium orange in cultured epithelial cells from CRYαA N101D and CRYαA WT , respectively. Only a few CRYαA N101D epithelial Fig. 7 Immunogold-labeling to determine relative localization of αA-WT and αA N101D in lens membranes of CRYαA N101D and CRYαA WT mice. a and c show membranes of lenses from CRYαA N101D (at 500 nm and 100 nm magnification respectively) and (b) and (d) from CRYαA WT (at 500 nm and 100 nm magnification respectively). The bigger particles (25 nm, red arrows) represented the aquaporin 0 whereas the smaller gold particles (10 nm, yellow arrows) represented the His-tagged αA N101D and WTαA-. As shown in the representative images in (a) to (d), both 10 nm and the 25 nm gold particles were bound to membranes.e: Quantification of width of membranes from lenses of CRYαA N101D and CRYαA WT mice. Note that the lens membranes of CRYαA N101D mice were about 2X wider that those from CRYαA N101D mice, suggesting membrane swelling of the former lenses cells showed the Ca 2+ uptake, which was possibly due to our previous finding that the lens cells contained only about 14% of αAN101D mutant protein [28]. In this experiment, 100 cells from the cultures of two types of lenses were counted. On quantification by Image J, the number of cells exhibiting calcium orange uptake were 1.5X greater in cells of CRYαA N101D lenses relative to cells from CRYαA WT lenses (Fig. 8b). On the determination of levels of mRNA of Na, K-ATPase in these cells, its level was 75% lower in the CRYαA N101D lens cells than CRYαA WT lens cells (Fig. 8c).

Discussion
Several past studies have shown in vitro effects of deamidation of crystallins on their structural properties including those in αA-, and αB-crystallins [21][22][23][24][25]. It has been reported that deamidation of Asn and Gln was the major modification identified in several human cataractous and aged lenses and these totaled 66% of the modification in the water-soluble and water-insoluble protein fractions that was analyzed by 2D LC/MS [23]. The mass spectrometric analysis found that there is negligible (less than 1%) deamidation at αA N101 site in both aged and cataractous human lenses [38]. These studies suggested that because of low levels of deamidation of αAat N101 to D in normal and cataractous lenses, the αAN101D might not play a significant role in cataract development. However, additional studies suggest otherwise. For example, our in vitro studies showed significant altered structural and functional properties of αA-crystallin on deamidation of N101 residue but not of N123 residue [24]. We also showed that the WS-protein fraction from 50 to 70 year-old human donors contained αAfragments with deamidation of N101 to D [39]. This finding is significant because recent studies have also shown an increasing role of crystallin fragments in cataract development [40,41]. In the present study, the cortical cataract development in mice on the introduction of αA-N101D transgene further show significance of deamidation of this site and altered changes in the lens. However, the exact in vivo molecular mechanism of αAN101D-induced crystallin's aggregation is yet to be fully understood.
Previously we showed that the three recombinant deamidated αA-mutants (N101D, N123D, and N101D/N123D) exhibited reduced levels of chaperone activity, alterations in secondary and tertiary structures, and larger aggregates relative to WT-αA-crystallin [24,25]. Among the above three mutants, the maximally affected and altered properties were observed in the recombinant αAN101D mutant [25]. Additionally, our recent results show that in vitro, the deamidated αA-, and αB-crystallins facilitated greater interaction with βA3-crystallin, leading to the formation of larger aggregates, which might contribute to the lens cataractogenic mechanism [42]. As a further extension of our previous studies [28,29], in some studies the 7-month old lenses were chosen because of the development of cortical cataract at about 7-month of age in the αA-N101D mice relative to αA wt mice. In other experiments, lenses from 5month old of both types of mice were used to determine the progression of phenotypic changes in lenses to determine their significance in cataractogenic mechanism.
The present study show that the introduction of αAN101D trans-gene in a mouse model resulted in the following major in vivo effects in lenses of CRYαA N101Drelative to CRYαA WT mice: (A) An age-related difference in protein profiles with an increasing association αA N101D with WI-protein fraction suggesting its insolubilization after 4-months of age. (B) The WS-HMW protein fraction showed a higher level of proteins with a greater M r . (C) Mass spectrometric analysis showed preferential insolubilization of αA-, αB-, γDand γEcrystallins, and nestin, which remained insoluble even in 8 M urea. (D) The tight association of αAN101D with membranes relative to WTαA, which could not be fully dissociated with 8 M urea treatment. (E) In vitro, αA N101 a showed greater affinity and binding to lens membranes relative WTαA. (F) The greater number of immunogoldlabeled αA N101 relative WTαA binding to membrane along with relatively greater swelling of lens membranes, suggesting the potential water uptake due to intracellular ionic imbalance, and (G) The ionic imbalance was suggested by the greater Ca 2+ uptake and 75% reduction in mRNA levels of Na, K-ATPase in the epithelial cells cultured from CRYαA N101D lenses relative to those from CRYαA WT lenses. Our mass spectrometry analysis showed that retinal dehydrogenase was absent in the N101D mice. It has been shown earlier that Aldh1a1(−/ −) knock-out mice developed lens opacification later in life [43]. Retinal dehydrogenase 1 may protect the lens against cataract formation by detoxifying aldehyde products on lipid peroxidation in both cornea and lens. It has been shown that antimalarial drug chloroquine which binds and inhibit retinal dehydrogenase 1 [44] induce cataract in rats [45]. Together, these findings suggest altered membrane integrity (possibly due to greater levels of αA N1010D binding to membrane than WTαA) resulting in intracellular ionic imbalance in CRYαA N101D lenses, which could play a major role in the cortical cataract development.
Among the lens crystallins, only α-crystallin show an association with the membrane in both normal and cataractous lenses [6,[46][47][48][49][50]. Lens membranes contain both a high-affinity saturable and low-affinity non-saturable αcrystallin-binding sites [46,[50][51][52]. Alpha-crystallin binding to native membranes was enhanced on stripping of extrinsic proteins from the lens membrane surface to expose lipid moieties [32,33], which contradicted a previous report that the crystallin mostly interacts with lens membrane MP26 protein [53]. Even after stripping extrinsic membrane proteins by alkali-urea treatment, the fulllength αA-, and αB-crystallins remained associated with membranes of both bovine and human lenses [6]. Additionally, αB-crystallin showed three-fold higher binding to lens membrane relative to αA-crystallin, and their binding was affected by the residual membrane-associated proteins, suggesting that their binding behaviors were affected by an intrinsic lens peptides [6]. A large-scale association of proteins with cell membranes in the lens nucleus (mostly in the barrier region) occurs after middle age in human lenses [48], and such association was enhanced by mild thermal stress [49]. The in vitro studies further supported this because the binding capacity of αcrystallin from older lenses to lipids increased with age and decreased in diabetic donors who were treated with insulin [50]. This implied that under diabetic conditions, abnormal binding of α-crystallin to lens membrane occurred. Such information in the literature about membrane binding of native vs. post-translationally modified crystallins including the deamidated αAN101D species is Fig. 8 Determination of levels of intracellular Ca 2+ and Na, K-ATPase mRNA in cultured epithelial cells from lenses of CRYαA N101D and CRYαA WT mice. a Left and right panels show intracellular Ca 2+ staining following uptake from calcium orange in cells from CRYαA WT -and CRYαA N101D mice, respectively. b Quantification by Image J of the number of cells that showed positive intracellular Ca 2+ -staining following uptake from calcium orange in CRYαA WT -and CRYαA N101D cells. c Relative levels of Na, K-ATPase mRNA in epithelial cells from CRYαA WT -and CRYαA N101D mice as determined by the QRT-PCR method presently lacking. Therefore, the results of the present study showing relatively increased binding of αAN101D relative to WTαA are highly significant.
The RNA sequence and IPA data of our previous study [29] further support the findings of the present study. This study [29] showed that the genes belonging to gene expression, cellular assembly, and organization, and cell cycle and apoptosis networks were altered, and specifically, the tight junction-signaling and Rho A signaling were among the top three canonical pathways that were affected in the CRYαA N101D lenses relative to CRYαA WT lenses. The present study showed an increased association of αAN101D to membrane, and this could lead to potential ionic imbalance affecting tight junction assembly and RhoA GTPase expression. This in turn causes increased proliferation and decreased of differentiation and denucleation of epithelial cells, and an accumulation of nuclei and nuclear debris in the lens anterior inner cortex and fiber cell degeneration. Some of these phenotypic changes could be cause or effects, but together could be responsible for the age-related cortical cataract development in CRYαA N101D lenses.
To maintain ionic balance within lens cells, a permeability barrier close to the surface of the lens is responsible for the continuous sodium extrusion via Na, K-ATPase-mediated active transport [35][36][37]. Without an active sodium extrusion, lens sodium and calcium contents are shown to increase resulting in lens swelling that leads to loss of lens transparency [35]. Similarly, an excessive intracellular Ca 2+ levels can be detrimental to lens cells, and its increased levels play an important role in development of cortical cataract [37]. Therefore, homeostasis of Na + , K +, Ca 2+ and other ions within the lens has been recognized as of fundamental importance in lens pathophysiology. These have been altered as shown in our present and our previous studies [29]. It is also possible that the increased Ca 2+ levels could in turn lead to calpain activation and proteolysis of crystallins, which will be investigated in future.
Similar to our study, other studies have shown that an increased membrane binding of αcrystallin in the pathogenesis of different forms of cataracts. The high molecular weight complexes (HMWCs), comprised of α-crystallin and other crystallins, accumulate with aging and show a greater membrane binding capacity than native α-crystallin [50]. Other mutants of αA-crystallin, like the αA N101D mutant, also exhibit a greater membrane binding than corresponding wild-type species [54]. For example, in the αAR116Cassociated congenital cataracts, an increased membrane binding capacity along with changes in complex polydispersity, and the reduction of subunit exchange were considered potential factors in the cataract pathogenesis [54]. Similarly, αA-crystallin R49C neo mutation influenced the architecture of lens fiber cell membranes and caused posterior and nuclear cataracts in mice [55].
Interactions between proteins and the cell membrane are an integral aspect of many biological processes, which are influenced by compositions of both membrane lipids and protein structure [56]. Reports have shown the age-related lipid compositional changes in the lens membrane, which might affect α-crystallin binding, i.e., in the nucleus of the human lenses, the levels of glycerophospholipids declined steadily by age 40 as opposed to the levels of ceramides and dihydroceramides increased approximately 100 fold during middle-age [57,58]. Further, it has been shown that because of the elevation of sphingolipid levels with species, age, and cataract, lipid hydrocarbon chain order, or stiffness increased. Therefore, the increased membrane stiffness caused an increase in light-scattering, reduced calcium pump activity, altered protein-lipid interactions, and perhaps slow fiber cell elongation [60]. Presently, whether similar changes occur in αA N101D lenses are not known.
Alpha A-and αB-crystallins differently associate with the cellular membrane, i.e. αA-crystallin may interact exclusively with membrane phospholipids, and thereby unaffected by the presence of extrinsic proteins on the membrane, whereas these proteins may act as conduits for αB-crystallin to bind to the membrane [58]. Presently, the specific binding mechanism of αAN101D to the membrane and age-related changes in lipid composition in lenses CRYαA N101D vs. CRYαA WT are unknown, and these are presently the focus of our investigations.

Conclusions
The results presented in this study suggest that an increased association of αAN101D relative WTαA with the lens membrane causes a possible loss of membrane integrity, leading to an ionic imbalance, and in turn, to membrane swelling, cellular disorganization and finally cortical opacity. Our future study will determine the specific binding site in the αAN101D relative WTαA, and changes in the membrane compositions that might facilitate the increased binding of the deamidated crystallin with the membrane.