Multiple Aggregates and Aggresomes of C-Terminal Truncated Human αA-Crystallins in Mammalian Cells and Protection by αB-Crystallin

Background Cleavage of 11 (αA162), 5 (αA168) and 1 (αA172) residues from the C-terminus of αA-crystallin creates structurally and functionally different proteins. The formation of these post-translationally modified αA-crystallins is enhanced in diabetes. In the present study, the fate of the truncated αA-crystallins expressed in living mammalian cells in the presence and absence of native αA- or αB-crystallin has been studied by laser scanning confocal microscopy (LSM). Methodology/Principal Findings YFP tagged αAwt, αA162, αA168 and αA172, were individually transfected or co-transfected with CFP tagged αAwt or αBwt, expressed in HeLa cells and studied by LSM. Difference in protein aggregation was not caused by different level of α-crystallin expression because Western blotting results showed nearly same level of expression of the various α-crystallins. The FRET-acceptor photo-bleaching protocol was followed to study in situ protein-protein interaction. αA172 interacted with αAwt and αBwt better than αA168 and αA162, interaction of αBwt being two-fold stronger than that of αAwt. Furthermore, aggresomes were detected in cells individually expressing αA162 and αA168 constructs and co-expression with αBwt significantly sequestered the aggresomes. There was no sequestration of aggresomes with αAwt co-expression with the truncated constructs, αA162 and αA168. Double immunocytochemistry technique was used for co-localization of γ-tubulin with αA-crystallin to demonstrate the perinuclear aggregates were aggresomes. Conclusions/Significance αA172 showed the strongest interaction with both αAwt and αBwt. Native αB-crystallin provided protection to partially unfolded truncated αA-crystallins whereas native αA-crystallin did not. Aggresomes were detected in cells expressing αA162 and αA168 and αBwt co-expression with these constructs diminished the aggresome formation. Co-localization of γ-tubulin in perinuclear aggregates validates for aggresomes.


Introduction
A major protein of the vertebrate eye lens, namely a-crystallin, consists of two homologous 20 kDa subunits, namely aAand aBcrystallins [1][2][3]. These two proteins are members of the small heat-shock protein (sHsp) family and have the ability to operate as molecular chaperones by binding to partially unfolded target proteins and preventing them from aggregation [4][5][6][7]. The Cterminal extension and the predominantly hydrophilic flexible Cterminal tail of aA-crystallin play a vital role in the oligomerization [8,9] as well as for ensuring solubility of the protein assemblies formed with target proteins.
Post-translational modifications of lens crystallins are believed to play a major role in the development of human senile cataract. Cleavage of amino acid residues at specific sites in the C-terminal end of aA-crystallin constitutes the major form of modification that leads to structural and functional changes in this sHsp/molecular chaperone [10][11][12][13][14][15][16]. In human aA-crystallin, 13 cleavage sites have been identified and the residues 162, 168 and 172 being the major ones [16]. Cleavage of serine from the C-terminus, which forms truncated aA172, is the most prevalent form of modification that occurs in human eye lens crystallins [10,15,16]. Our earlier studies have shown increased formation of aA172 in diabetic human lenses; the total level of aA172 increased from about 30% in non-diabetic lenses to about 50% in diabetic lenses [16]. Cleavage of 1, 5, and 11 residues showed diverse effects on oligomerization and chaperone function [17]. Chaperone activity of aA172 was 28-46% higher than that of aAwt and the oligomeric size was increased by 12% [17]. On the other hand, aA168 and aAwt had similar chaperone activity and molecular mass whereas aA162 behaved quite differently by showing 80-100% decrease in chaperone activity and 42% decrease in molecular mass. However, it should be emphasized that these results were obtained by studying homoaggregates, but, in human lenses they may exist as homoaggregates as well as heteroaggregates in association with native aA-crystallin and/or aB-crystallin. As heteroaggregates, the truncated aAcrystallins are expected to behave differently. The ability to associate with native aAor aB-crystallin is dictated by the strength of the interactions between them. In a previous in vitro study with recombinant aBwt, aAwt and the C-terminal truncated aAcrystallins and by utilizing fluorescent chemical probes in fluorescence resonance energy transfer (FRET) analysis, we have observed C-terminal truncation affecting interaction with aAwt and aBwt [18]. However, mapping the interactions in living mammalian cells has not been done before. In addition, the present study was aimed to show, whether truncated aA-crystallins tend to aggregate in living cells and, if so, will co-expression with either aAwt or aBwt suppress aggregation? The present study also showed whether cleavage of the C-terminal residues of aA-crystallin affects its interaction with native aAand aB-crystallins in mammalian cells.  Fig. 1A). Randomly selected fields of 50 cells were counted and % of cells containing aggregation was calculated for each group. The values of all the truncated constructs are statistically significant at p,0.05 (t-test) compared to the aAwt group. C: Immunofluorescence analysis of aA-crystallin over expression in HeLa cells. aA-crystallin is abundantly localized in the cytoplasm, however, nuclear localization as in the form of foci were seen in aAwt, aA162 and aA168 individually expressing cells. In cells transfected with empty vector, there was no staining indicates that there was no endogenous expression of aA-crystallin. Cells containing aggregates were stained intensely in aA162 and aA168 expressing cells. An enlarged view from aA168 expressing cells shows nuclear foci (bottom right panel) and stained cytoplasmic aggregates. The images are representative of four similar images obtained in three independent experiments. doi:10.1371/journal.pone.0019876.g001 Aggresomes are spherical or ribbon like structures localized in the perinuclear region. Protein quality control systems, such as molecular chaperones and ubiquitin-proteasome system (UPS) degrade or refold the abnormal proteins and prevents the toxic accumulation of small protein aggregates. However, when the protein quality control system is overwhelmed or evaded, the resulting small aggregates are dispersed throughout the cell and they are actively cleared via transport to intracellular inclusion bodies (IBs). These IBs are termed aggresomes or aggresome-like inclusions. These structures are conserved from yeast to mammalian cells and act as storage bins for protein aggregates [19][20][21]. The formation of aggresomes is believed to serve as cytoprotective function by refolding or degradation of unfolded or misfolded proteins [22] and they are produced around the microtubule organizing center (MTOC) for degradation [19]. In this paper, we have demonstrated aggresomes are evident in aA162 and aA168 individually expressing constructs and in co-expression of aAwt with these truncated constructs. Co-expression of aBwt with these truncated constructs significantly diminished the aggresome formation.

Materials
The Cyan (pAmCyan1-Cl or CFP) and Yellow (pZsYellow1-Cl or YFP) expression vectors were obtained from Clonetech ( Palo Alto, CA ), HeLa cells were obtained from the American Type Culture Collection (

Construction of CFP and YFP-tagged a-crystallins vectors
The full-length human aBcrystallin wild-type, aAcrystallin wild-type and the C-terminal truncated aA-crystallins (aA162, aA168 and aA172) genes were PCR amplified using the appropriate primers containing restriction sites, Xho I and Hind III and cloned into the C-terminal end of the mammalian expression vectors, pAmCyan1-C1 (CFP) or pZsYellow1-C1 (YFP) driven by CMV promoter. In the present study, both human aBwt and aAwt were sub-cloned into the CFP vector for the expression of crystallin genes in cyan color and aAwt, aA172, aA168 and aA162 were subcloned into the YFP vector for their expression in yellow color. All the constructs were confirmed by restriction digestion analysis and sequenced at the UAMS DNA sequencing core facility.

Cell culture and transfection
HeLa cells were cultured in MEM medium (Invitrogen, Carlsbad, CA) supplemented with 10% FBS and penicillin/ streptomycin (100 mg/ml), at 37uC in 5% CO 2 humidified chamber. About 1.0610 5 cells / ml were seeded into each 35 mm, sterile glass bottomed single well poly-d-lysine treated plates (MetTek Corporation, Ashland, MA, USA) and cultured in 2 ml of growth medium for transient transfection. The overnight adherent cells were transfected with Lipofectamine 2000 (Invitrogen, Rockville, MD) according to the manufacture's protocol. Briefly, each well was transfected alone or co-transfected with total 2 mg/well of pAmCyan1-C1 (CFP), and/or pZsYellow1-C1 (YFP) plasmids encoding the respective crystallin gene along with 5 ml of Lipofectamine 2000. After 6 h, transfected medium was removed and replaced with fresh medium was containing 10% FBS. After 48 h transfection, cells were examined for laser scanning confocal microscopic study. Transfected cells showing aggregates were typically counted at 640 magnification. Fields were randomly chosen and about 300 cells were counted per experiment and repeated at least three times and counts were blindly performed.

Western blotting for aAand aB-crystallins expressed in HeLa cells
After 48 hours transfection, cells were lysed with lysis buffer containing 50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 0.02% sodium azide, 0.1% SDS, 1% NP-40, 0.5% sodium deoxycholate and 0.1 mM EDTA supplemented with cock-tail protease inhibitors and 3 M urea. Further, cells were sonicated and the protein concentration was measured by BCA assay method. For each sample, 5 mg of protein was loaded into 12% SDS-PAGE and electroblotted to nitrocellulose membrane. The blots were blocked with 5% non-fat dry milk prepared in TBST (Trisbuffered saline supplemented with 0.1% Tween 20) and subsequently incubated with primary antibody for aA-crystallin (monoclonal, Abcam, 1: 2000), aB-crystallin (rabbit polyclonal, Abcam, 1:2000) for one hour at room temperature. Blots were washed with TBST for three times and incubated with appropriate HRP-conjugated secondary antibodies (1 in 5000, Santa Cruz Biotechnology Inc, CA) for one hour at room temperature. Enhanced chemiluminescence substrate was used and the signal was detected by exposing the blots on films. For loading control, blots were stripped with Restore Western Blot stripping buffer (Thermo Scientific Inc, IL) and re-probed with a rabbit polyclonal antibody against b-actin (Abcam, 1: 10000) for 1 hour at room temperature.

Immunofluorescence microscopy
Cells were grown on 35-mm cover glass bottom dishes. After 48 hours transfection, cells were washed with PBS, fixed with 4% paraformaldehyde for 20 minutes at room temperature (RT) and permeabilized with 0.5% Triton X-100 for 10 minutes at RT. Cells were blocked with 3% Normal Goat Serum (NGS) for one hour at RT and labeled with primary antibody for aA-crystallin in 3% NGS (1 in 500) for overnight at 4uC and subsequently incubated with Alexa Fluor 594 goat-anti-mouse IgG secondary antibody (Molecular Probes) diluted in 3% NGS (1 in 500) for one hour at RT and washed with PBS. For double immunofluorescence, cells were fixed with 4% paraformaldehyde and permeabilized in 0.5% Triton X-100 and blocked with 10% normal goat serum (NGS) and simultaneously incubated with the two primary antibodies, aA-crystallin (Mouse monoclonal, 1 : 500) and ctubulin (Rabbit polyclonal, Abcam 1: 500) diluted in 5% NGS for overnight at 4uC and washed with PBS for five times. The cells were then stained with Alexa Fluor 594 Goat anti-mouse (1:500) and Alexa Fluor 488 Goat anti-rabbit (1:500) diluted in 5% NGS for one hour at RT and washed with PBS for three times. Hoechst 33342 (Molecular Probes) was used to stain the nuclei. The images were acquired with an LSM 510 Meta Carl Zeiss Confocal microscope at 663 objective and analyzed using AIM Imaging Software.

Laser scanning confocal microscope studies
A Zeiss Meta LSM 510 Laser Scanning Microscope (Carl Zeiss Inc., Thornwood, NY) with 663 oil-immersion objective (plan Apochromat, NA 1.4) (University of Arkansas for Medical Sciences core facility) was utilized. To visualize CFP and YFP fluorescence, cells expressing fluorescent proteins were excited with appropriate laser beam and filtered with both dichromatic band pass filters, captured at 12 bit 5126512 multi-track channel images with CCD cameras with the following configurations: for CFP channel, the cells were excited with 458 nm filter by argonion laser and the emission intensity was collected using band pass (BP) 475-525 nm filters and for YFP channel, the cells were excited with 514 nm filter by argon-ion laser and the emission intensity was collected using BP 530-600 nm filters. Both the CFP and YFP was excited using argon-ion laser at 25 mW, 2.0 and 0.5% exposure respectively. All images were taken at room temperature.

FRET analysis by live acceptor photobleaching method
The acceptor photobleaching method is one of the accurate methods available to determine the interaction between two proteins based on the increased intensity of donor fluorescence at the time of acceptor bleaching. In this method, the acceptor fluorescence was bleached with the help of high intensity argon laser light (100% exposure at 514 beam). A series of prebleaching and post-bleaching donor and acceptor signal collecting protocols were automated for the acquisition of pre-bleach and post-bleach images and noted the increased level of donor intensities due to de-quenching and decreased level of acceptor signal due to photo-bleaching. The increased donor (CFP) fluorescence intensity and decreased acceptor (YFP) fluorescence intensity is the sign for the occurrence of protein-protein interaction. The FRET efficiency was calculated based on ten images taken from each construct examined and each experimental condition was performed 3 times and values were averaged. The FRET efficiency (E) was calculated by: E = 12(Ipre/Ipost)6100%, where Ipre is pre-bleach fluorescence intensity and Ipost is post-bleach fluorescence intensity.

Aggresome staining
HeLa cells were grown on glass bottom 35 mm dishes and transfected with YFP-tagged aAwt, aA162, aA168 and aA172 constructs individually and or co-transfection with CFP-tagged aA-wt or aB-wt. After 48 h transfection, cells were fixed in 4% paraformaldehyde for 30 minutes at room temperature (RT) and permeabilized with 0.5% Triton X-100 in 16 assay buffer for 30 minutes on ice. Cells were washed with 16assay buffer for two times and stained with ProteoStat Aggresome dye (Enzo Life Sciences, PA) for 30 minutes at RT and washed with 16 assay buffer. The stained cells were examined with an LSM 510 Meta Confocal microscope and images were captured at 663 objective with red filter.

Statistical analysis
A two-tailed Student's t-test was used to calculate the significance between the wild-type and the truncated aA-crystallin groups. The p value,0.05 was considered as significant.

Individually expressed aAwt and its C-terminal residues cleaved proteins in HeLa cells
HeLa cells were transfected with YFP-DNA constructs for aAwt and the three truncated aA-crystallins individually and laser scanning confocal microscopic (LSM) images were taken after 48 h (Fig. 1A). LSM images of the CFP vector alone and YFP vector alone showed full expression of each vector in both the nucleus and the cytoplasm with no evidence of aggregation (data not shown). Expression of YFP-aAwt and YFP-truncated aA-crystallins was mostly confined to the cytoplasm. However, the immunostained cells show that nuclear localization in the form of foci is evident in aAwt and the truncated constructs overexpressed cells (Fig. 1C). In addition, there was an evidence for the presence of protein aggregates predominantly in the cells expressing aA168 and aA162 and to a lesser extent in cells expressing aA172. In the immunofluorescence microscopic images, these aggregates were stained intensely than the diffuse staining pattern seen in aAwt and aA172 expressing cells (Fig. 1C). Moreover, the shape of the cells expressing these truncated aA-crystallins appeared abnormal and distorted. When aAwt, aA172, aA168, and aA162 were expressed individually, nearly 3, 27, 55 and 74% cells, respectively, had significant level of aggregates (Fig. 1B). Western blotting with anti-aA antibody showed nearly equal level of expression of aA-wt and each of the truncated aA-crystallin (Fig. 2). Thus, the difference in the levels of protein aggregation was apparently not due to different level of expression of the various forms of aA-crystallin. Furthermore, the immunofluorescence results suggest that there was no endogenous expression of aA-crystallin in empty vector transfected HeLa cells (Fig. 1C).

Co-expression of CFP-aAwt with YFP-aAwt and YFPtruncated aA-crystallins
HeLa cells were co-transfected with CFP-aAwt and each of the YFP-aA-truncated and the images were collected after 48 h (Fig. 3A). Co-expression of CFP and YFP vectors alone in HeLa cells resulted in the presence of both the vectors in the cytoplasm as well as the nucleus with no apparent aggregation (data not shown). As expected, protein expression was confined to the cytoplasm only when CFP-and YFP-tagged proteins were coexpressed (Fig. 3A). Co-expression with aAwt has not improved the appearance of the cells and has not decreased protein aggregation within the cells. Significant protein aggregation was seen in the cells expressing aA162, aA168, and aA172, as shown by 81, 68 and 54% of the cells, respectively, having significant protein aggregates (Fig. 3B). This shows actual increase in protein aggregation, probably due to co-aggregation of truncated aAcrystallins with aAwt. Western blotting with anti-aA antibody showed nearly equal level of total aA expression (anti-aA antibody does not distinguish between aAwt and the truncated forms) (Fig. 4). Co-expression of CFPaBwt with YFPaAwt and YFPtruncated aA-crystallins HeLa cells were transfected with pairs of CFPaBwt and YFPaAwt and all the truncated aA-crystallins and the images were collected after 48 h (Fig. 5A). By co-expression with aB, there was clear indication of improvement in cell morphology even in the presence of truncated aAcrystallins like aA168 and aA162. Cells carrying protein aggregates also decreased substantially as indicated by a decrease in the cells containing aggregates to 1, 4, 34, and 55%, respectively for CFPaB/YFPaAwt, CFPaBwt/ YFPaA172, CFPaBwt/YFPaA168 and CFPaBwt/YFPaA162 (Fig. 5B). Thus, the presence of aB-crystallin has shown significant inhibition of protein aggregation, although cells expressing aA162 still remained vulnerable to aggregation. Western blotting with both anti-aA and anti-aB antibodies showed almost the same level of expression of the various aA-crystallins and aBwt (Fig. 6).

Results of in situ FRET studies by LSM image analysis for homologous and heterologous interactions
The acceptor photo-bleaching method was used to determine the intensities of interactions (FRET efficiency) of the C-terminal truncated aA-crystallins with aAwt and aBwt. It is expected that when the acceptor fluorescence is completely bleached the donor fluorescence intensity increases proportionately and this increase is considered a measure of the interaction between the two proteins. Co-expression of CFP and YFP vectors only followed by photobleaching of the acceptor YFP showed no increase in the donor CFP fluorescence intensity which is indicative of the lack of interactions between the vectors alone. Pre-bleach and post-bleach LSM images of CFPaAwt/YFPaA-truncated and CFPaBwt/YFPaA-truncated showed complete or nearly complete photo-bleaching; Fig. 7 illustrates, as an example, photo-bleaching of YFPaA172 (acceptor) and increase in fluorescence intensity of CFPaBwt (donor). FRET efficiency values were generated from LSM images, calculated as discussed in 'Methods'. These values were generated for homologous interactions where aAwt interacts with C-terminal truncated aAcrystallins and also for heterologous interactions where aBwt interacts with C-terminal truncated aA-crystallins ( Fig. 8) As expected, negative control (vectors alone) showed very little interaction while positive controls (aAwt/aAwt and aBwt/aAwt) showed significant interaction. However, in aAwt/aA162 and aAwt/aA168 FRET efficiencies were nearly 30% lower than that of aAwt/aAwt whereas in aAwt/aA172 FRET efficiency was 50% higher. CFPaBwt/ YFPaAwt and CFPaBwt/YFP-truncated aA-crystallins also showed complete photo-bleaching. FRET efficiency in aBwt/aA168 was slightly higher and in aBwt/aA172 two-fold higher than in aBwt/ aAwt. Moreover, the overall interaction of the C-terminal truncated aA-crystallins with aBwt was two-fold higher than with aAwt.

Detection of Aggresomes in truncated aA-crystallin expressed cells
Aggresomes are known to serve as storage bins of misfolded or aggregated proteins. Since the truncated aA-crystallin forms intracellular aggregate, we sought to investigate whether truncated aA-crystallin expression forms aggresomes, we stained the cells with Proteostat Aggresome dye after 48 hour transfection. This   FRET efficiency demonstrates the interaction between the aA and aB subunits of a-crystallin. The interaction was strong between the wild types of aA and aB subunits. The interaction between the truncated constructs, aA162 and aA168 with aAwt and aBwt were lower compared to aA172 expression. The results are expressed as mean 6 Standard Deviation (SD). doi:10.1371/journal.pone.0019876.g008 dye has been used to detect misfolded and aggregated proteins within aggresomes and inclusion bodies in cells [23]. Interestingly, bright punctate staining for aggresomes localized specifically in the perinuclear or juxta-nuclear sites of the cells individually expressing aA162 and aA168 but not in aAwt and aA172 and these results is very similar with the positive control where cells treated with a potent cell-permeable proteasome inhibitor, MG-132 (5 mM) for 15 hours. (Fig. 9). There was no decrease in the number of cells containing aggresomes in co-expression of aA-wt with aA162 and aA168 constructs (Fig. 10). In contrast, coexpression of aB-wt significantly diminished the aggresome formation in cells expressing with aA162 and aA168 constructs (Fig. 11).
To further validate the perinuclear inclusions as aA-crystallinpositive aggresomes as judged by staining with ProteoStat Aggresome dye, transfected cells were subjected to double immunostaining with aA-crystallin and c-tubulin antibodies. The c-tubulin has been previously shown to co-localize with aggresomes in the microtubule organizing center (MTOC) [24][25][26][27]. We compared the aA-crystallin and c-tubulin staining in the transfected cells and our results suggest a strong degree of overlap in staining the perinuclear region. The data is consistent with the interpretation of only the two truncated versions of aA-crystallins, aA162 and aA168 form aggresomes but not in cells expressing with aAwt and aA172 ( Fig. 12 and Fig. 13).

Discussion
In the present study, we have studied various CFP and YFP tagged a-crystallins expression in HeLa cells. Human lens epithelial cells may have been a preferable choice, however, the level of expression of aAcrystallin and aB-crystallin is expected to be too low in the available human epithelial cell lines to obtain significant LSM signal. Moreover, the epithelial cells are known to have endogenous aB-crystallin and it would have complicated the study. All the three truncated aA-crystallins investigated in this study showed various extent of protein aggregation when each construct was transfected in HeLa cells, either individually or with aAwt or aBwt, for 48 hours (Fig. 1, 3, & 5). It was not possible to ascertain the actual level of aggregated protein in individual cells,   instead, depended solely on visual assessment. There is no accurate method is available to quantify the intracellular protein aggregations in the cells. We used Dynamic Light Scattering (DLS) assay to measure the aggregations from the lysed samples, but we do not see any light scattering in these samples. The reason is DLS technology will not able to track the aggregation of protein of interest when more than one protein present in the lysed sample. This makes sense, because the aggregates form a high molecular weight complex and also the size of the particles in the complex overlaps with the other particles present in the sample. DLS technique is often applied to aggregation studies on purified proteins which is 95% purity or higher. In fact, the LSM images provided in figures 1A, 3A, and 5A give the true appearance of the living HeLa cells with different levels of aggregates.
The question arises as to what makes the truncated aAcrystallins to aggregate in mammalian cells. Lack of protein stability and conformational changes could be two major factors that could influence aggregation. In an earlier study, we have tested the stability of the truncated aA-crystallins at 25 and 37uC by measuring light scattering for 30 minutes [17]. All the three truncated aA-crystallins were stable at 25uC and only aA162 was slightly unstable at 37uC. However, all the three truncated aA-crystallins, were under different degree of unfolding stress when they were expressed individually for 48 h in HeLa cells. As homoaggregates, while they were not associated with aAwt or aBwt, they exhibited conformational changes, aA162 showing the largest change [17]. Moreover, aA162 showed three-fold increase in the a-helical content accompanied by a loss in b-sheet conformation and an increase in random coil conformation [17].
Thus, it appears that the conformational changes make these truncated aA-crystallins aggregation prone. Association with aAwt did not prevent their susceptibility to aggregation (Fig. 3). However, interaction with aBwt significantly diminished the aggregation of each of the truncated aA-crystallin, aA168 and aA172 showing the most effect and aA162 showing the least effect (Fig. 5). aB-crystallin is known to be a better molecular chaperone than aA-crystallin and it readily recognizes partially unfolded structures and prevent them from aggregation. Structural studies suggest that aA172 and aA168 are partially unfolded [17] and, so, aB-crystallin binds effectively to these polypeptides and the aggregation process is nearly completely prevented in aA172 and significantly decreased in aA168. In the case of aA162, there is strong evidence for the presence of fully unfolded structural  entities [17] which explains why aB-crystallin failed to recognize them and, thus, unable to completely prevent protein aggregation.
Earlier in vitro FRET studies performed in our laboratory have confirmed C-terminal truncated aA-crystallins having weak interactions with both aAwt and aBwt [18]. This was exceptionally true for aA162 interacting with aBwt because the subunit exchange rate (k) was 0.65610 24 S 21 as compared to 4.1610 24 S 21 for aAwt interacting with aBwt [18]. The decreased interaction of the truncated aA-crystallins with aAwt or aBwt is expected to increase the probability of the truncated aA-crystallins existing as homoaggregates rather than heteroaggregates. This may put these truncated aA-crystallins in aggregation mode in the cell. Small heat-shock proteins (sHsps) like aAand aB-crystallins may also operate by a different mechanism by which preformed protein aggregates are dissociated first by sHsps for subsequent re-folding by the Hsp70 chaperone machine. However, it is uncertain whether such a pathway exists in these cells.
Dysregulation of the degradation of misfolded and aggregated proteins or the protein quality control pathway has been implicated in Cystic Fibrosis, many neurodegenerative diseases and cancer [28][29][30][31]. In the present study, we have used a dye, ProteoStat Aggresome dye for the detection of aggresomes. This dye has been used to detect the aggregated proteins and peptides within aggresomes and related inclusion bodies in cells and tissues [23]. Furthermore, this dye is non-fluorescent in solution but becomes brightly fluorescent upon binding to the tertiary structure of aggregated proteins [23]. Formation of aggresomes in the truncated constructs, aA162 and aA168, suggests that it may be a cellular response and the inhibition of aggresomes in C-terminal truncated constructs co-expressed with aBwt strongly suggests that chaperones are able to refold the aggregated proteins. Moreover, there was no sequestration of aggresomes in aA-wt co-expression with these truncated constructs suggest that it co-aggregates with the unfolded protein products of the aA162 and aA168 and localized into the perinuclear sites of the cells. Our results on formation of aggresomes were similar with another report on a myopathy-causing aB-crystallin mutant, R120G forms aggresomes in cell culture models [32].
It has been reported that multiple aggregates or pre-aggresome particles may be an intermediate step in aggresome formation which can proceed further upon inhibition of proteasome [33]. The present study documented both multiple aggregates and typical perinuclear localized aggresomes in HeLa cells overexpressing the C-terminal truncated aA-crystallin genes. Aggresomes are special protective structures that fundamentally differ from other multiple aggregates, that some of which cause cellular toxicity. They are formed around centrosome/microtubule organizing center (MTOC), a sub-cellular region is robustly enriched with chaperones and components of UPS. [19]. Indeed it has been reported that there is a close correlation between aggresome formation and cell survival. [34]. More studies are needed in this direction to elucidate the role of aggresomes and the signaling pathway in diseases associated with aA-crystallin mutants.