The high solubility and lifelong stability of crystallins are crucial to the maintenance of lens transparency and optical properties. Numerous crystallin mutations have been linked to congenital cataract, which is one of the leading causes of newborn blindness. Besides cataract, several crystallin mutations have also been linked to syndromes such as congenital microcornea-cataract syndrome (CMCC). However, the molecular mechanism of CMCC caused by crystallin mutations remains elusive. In the present study, we investigated the mechanism of CMCC caused by the X253R mutation in βB1-crystallin. The exogenously expressed X253R proteins were prone to form p62-negative aggregates in HeLa cells, strongly inhibited cell proliferation and induced cell apoptosis. The intracellular X253R aggregates could be successfully redissolved by lanosterol but not cholesterol. The extra 26 residues at the C-terminus of βB1-crystallin introduced by the X253R mutation had little impact on βB1-crystallin structure and stability, but increased βB1-crystallin hydrophobicity and decreased its solubility. Interestingly, the X253R mutant fully abolished the aggregatory propensity of βB1- and βA3/βB1-crystallins at high temperatures, suggesting that X253R was an aggregation-inhibition mutation of β-crystallin homomers and heteromers in dilute solutions. Our results suggest that an increase in hydrophobicity and a decrease in solubility might be responsible for cataractogenesis induced by the X253R mutation, while the cytotoxic effect of X253R aggregates might contribute to the defects in ocular development. Our results also highlight that, at least in some cases, the aggregatory propensity in dilute solutions could not fully mimic the behaviours of mutated proteins in the crowded cytoplasm of the cells.
Crystallins are the predominant soluble proteins in vertebrate lens. According to their elution positions on size-exclusion chromatography (SEC), crystallins can be categorized into three families (α-, β- and γ-crystallins) with distinct molecular sizes . Among them, α-crystallins, which are large oligomers, belong to the small heat-shock protein family and possess molecular chaperone activity . The main function of α-crystallins in the lens is to protect β/γ-crystallins as well as other proteins against misfolding and aggregation [3,4]. β/γ-Crystallins are the major structural proteins in the lens. Although β- and γ-crystallins share a similar fold in their tertiary structures, they differ significantly in their oligomeric sizes . β-Crystallins exist as homomers or heteromers with sizes between dimers and octa mers, whereas γ-crystallins are exclusively monomers. β-Crystallin heteromers are formed by the acidic (βA) and basic β-crystallin (βB) subunits. Various crystallins generally have an extremely high protein concentration up to 300 mg/ml in the mature lens fibre cells . Crystallins in the lens will last up to several decades after synthesis since the mature lens fibre cells lack organelles and protein-turnover machineries to produce nascent crystallin polypeptides and eliminate damaged proteins . The high solubility, lifelong stability and short-order interactions of various crystallins are thought to be important to the structural and functional maintenance of the lens throughout an individual's lifespan . Consequently, mutations in crystallins account for about half of autosomal dominant cataracts, which is one of the main causes of human blindness in newborns worldwide .
The general feature of congenital cataract caused by crystallin mutations is the formation of large intracellular aggregates in the lens , which will scatter the visible light and thereby retard light transmission. Although inclusion bodies or aggresomes can be observed in the cells with high-level expression of most cataract-causing crystallin mutants , the molecular mechanisms of aggregate formation are diverse at the protein level . As for the structural β- and γ-crystallins, the cataract-causing mutations may influence protein solubility, structure, folding pathway, protein interaction network and sensitivity to denaturants, heat, UV, pH or proteases [7,9–17]. Identification of the defects caused by a disease-linked mutation is important to understand the onset and progression of cataract as well as the development of potential strategies for cataract prevention and treatment.
Besides congenital cataract, mutations in crystallins may also lead to syndromes with both lens and non-lens impairments . The linkage between non-lens disease and α-crystallin mutations is not surprising since α-crystallins have been shown to be multifunctional proteins and are involved in apoptotic events . Interestingly, β/γ-crystallin mutations have been linked to congenital microcornea-cataract syndrome (CMCC) [19–25], implying that these β/γ-crystallins also play an important role in ocular development beyond cataractogenesis. Although it is clear that cataract is a crystallin-aggregation disease, the mechanism of CMCC induced by crystallin mutations remains elusive.
The first CMCC-linked crystallin mutation is X253R in βB1 identified in a U.K. family about 10 years ago . The X253R mutation results in an elongation of the C-terminus of βB1 by 26 extra amino acid residues. β-Crystallins possess extended N- and C-terminus when compared with γ-crystallins . Among the seven β-crystallins, βB1 has the longest N- and C-terminal extensions. The N-terminal extension of βB1 is important for βB1 folding, its protection on βA3 and the molecular chaperone action of αA [27,28]. However, the role of the C-terminal extension remains unclear. It would be interesting to know how the elongated C-terminal extension affects βB1 structure, stability and function. In the present study, we evaluated the effects of the additional 26 residues introduced by the X253R mutation on βB1 cellular distribution, structure, stability and protective effect on βA3. Solubility and cellular experiments were used to check whether the mutant could mimic disease phenotype at the protein and cellular levels. Biophysical experiments at low protein concentrates were used to seek the underlying mechanism. We showed that, at low protein concentrations, X253R is a beneficial mutation to help βB1 as well as βA3/βB1 to fight against aggregation under stressed conditions. However, compared with the wild-type (WT) βB1, the X253R mutant had more hydrophobic exposure, lower solubility and more intracellular aggregates. Our results suggest that the behaviours of crystallins might differ under high and low protein concentration conditions and protein expression level of the mutant in a specific tissue might contribute to the disease phenotype for a given mutation.
MATERIALS AND METHODS
IPTG, ultra-pure guanidine hydrochloride (GdnHCl), 1-anilinonaphthalene-8-sulfonate (ANS), SDS, DTT, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) and trypsin were purchased from Sigma–Aldrich. Lipofectamine™ 2000, Hoechst 33342, Dulbecco's modified Eagle's medium (DMEM) and FBS were purchased from Invitrogen. The antibody against p62 was obtained from Abcam. The Cell Counting Kit-8 (CCK-8) was purchased from Beyotime. All other chemicals were local products of analytical grade.
Protein expression and purification
The WT CRYBB1 and CRYBA3 genes were cloned from the human lens cDNA library as described previously . The X253R-encoding mutant of CRYBB1 was obtained by overlap extension using PCR. The primers used for the extending fragment at the C-terminus of βB1 were: forward, 5′-CACAGAGCCCCCCAAGCGAGTCCACACCTCACTCTGCT-ACCTTGCCCCAACCCTTCTTC-3′ and reverse, 5′-TTATTT-GCCTGGGAAAAATGGGGGAAATAATTGAACATGAAG-AAGGGTTGGGGCAAG-3′. The CRYBB1, CRYBA3 and X253R-encoding mutant CRYBB1 genes were inserted into the prokaryotic vector pET28a containing a His-tag fused to the N-terminus of the protein and a thrombin cleavage site after the His-tag. The His-tagged WT and mutated proteins were overexpressed in Escherichia coli BL21(DE3) as described previously . In brief, the recombinant proteins were purified by Ni2+–nitrilotriacetate column affinity chromatography, following by gel-filtration chromatography. The final products were >98% pure as evaluated by SDS/12.5% PAGE. The proteins were in buffer A containing 20 mM sodium phosphate, 150 mM NaCl, 1 mM EDTA and 1 mM DTT. The protein concentration was determined by the Bradford method . A protein concentration of 5.8 μM was used for most spectroscopic experiments unless otherwise indicated. Most experiments were performed using the His-tagged proteins since the additional His-tag had no effect on the spectral features and thermal aggregation of the WT and mutated proteins. The His-tag was removed by 0.1 mg of thrombin in a 100 μl solution containing 2–3 mg of recombinant proteins at 4°C for 1 h. The purity of the tag-free proteins was analysed by SDS/12.5% PAGE.
Cell culture, immunofluorescence and cell viability assay
The CRYBB1 and X253R-encoding mutant CRYBB1 genes were inserted into the eukaryotic vector pEGFP-C3 or pcDNA3.1-N-Flag containing HindIII and BamHI sites. In the recombinant plasmids, the EGFP or FLAG tag was fused to the N-terminus of the proteins to avoid interference with the C-terminal extension of βB1. The plasmids were obtained using the Plasmid Maxiprep kit (Vigorous) and verified by DNA sequencing. The WT and mutated βB1 proteins were exogenously expressed in the HeLa cells by transient transfection. The intracellular distribution of the overexpressed WT and mutated proteins and cell viability were determined using the same procedures as those reported recently . In brief, the HeLa cells were cultured in DMEM with the addition of 10% FBS at 37°C. The cells were seeded on glass slides treated with 2.5 M NaOH and 60% ethanol and grown for 24 h. The recombinant plasmids containing the WT and mutated CRYBB1 genes were transiently transfected into the cells with Lipofectamine™ 2000 according to the manufacturer's instructions. After transfection for 4 h, the cells were transferred to fresh high-glucose DMEM to remove the residual recombinant plasmids and transfection reagent and cultivated for 24 h. Then, transfected cells were fixed using 4% paraformaldehyde for 40 min and washed three times in PBS. The nuclei were dyed with Hoechst 33342 and aggresomes were stained with mouse antibody against p62. The fixed cells in the glass slides were visualized using a Carl Zeiss LSM 710 confocal microscope. Fluorescence images were recorded using excitation wavelengths of 405, 488 and 649 nm for Hoechst 33342, EGFP and Alexa Fluor 649-conjugated goat antibody against p62. The confocal imaging experiments were performed using a pinhole size of 50 μm and a ×63/1.40 numerical aperture (NA) oil DIC M27 Plan Apochromat objective. Co-localization and fluorescence intensity data were analysed by ImageJ (NIH) . Intensity correlation analysis was obtained by calculating the Pearson's correlation coefficient (Rr), where perfect correlation is 1. At least ten cells with X253R aggregates were used for the calculation of Rr. The cell viability was determined by CCK-8. The HeLa cells with or without the overexpression of βB1 were cultivated for a given time and reacted with the CCK-8 reagent for 1 h at 37°C. Then, the absorbance at 450 nm was measured on a Bio-Rad Laboratories Model-680 microplate reader. The cells transfected with pEGFP-C3 was used as a control.
The far-UV and near-UV circular dichroism (CD) spectra were measured on a Jasco-715 spectrophotometer using a 1 mm path-length cell and a 10 mm path-length cell respectively. The fluorescence spectra were recorded on a Hitachi F-2500 spectrofluorimeter with a 200 μl cuvette. The intrinsic tryptophan fluorescence was measured using an excitation wavelength of 295 nm and an emission wavelength range of 300–400 nm. Parameter A of tryptophan fluorescence spectra was calculated by dividing the fluorescence intensity at 320 nm by that at 365 nm (I320/I365) . Phase diagram analysis of the tryptophan fluorescence spectra was performed by plotting I320 against I365 as described previously . The ANS extrinsic fluorescence was determined using an excitation wavelength of 380 nm and an emission wavelength range of 400–600 nm. The ANS stock solution was added to the protein solutions to reach a final protein/ANS molar ratio of 1:75 and the solutions were equilibrated for 30 min in the dark at room temperature before measurements. Resonance Rayleigh light scattering was measured using an excitation wavelength of 295 nm . The turbidity of protein solutions was determined by the absorbance at 400 nm on an Ultraspec 4300 pro UV/visible spectrophotometer. The SEC analysis was achieved on a Superdex 75HR 10/30 column equipped on an ÄKTA FPLC (GE Healthcare) with an elution rate of 0.4 ml/min and 100 μl of protein solutions. Most spectroscopic experiments were conducted at 25°C with a protein concentration of 5.8 μM (approximately 0.2 mg/ml) in buffer A except that 29.0 μM (approximately 1 mg/ml) was used for near-UV CD measurements.
Protein denaturation by heat or GdnHCl
Protein denaturation experiments were conducted using a protein concentration of 5.8 μM in buffer A. Thermal denaturation was performed by heating the protein solutions continuously from 26°C to 86°C using a water bath connected to the spectrophotometers. The thermal transition curves were obtained by monitoring the changes in intrinsic tryptophan fluorescence, light scattering and turbidity every 2°C after 2 min equilibration at the given temperature. Denaturation of the proteins by GdnHCl was performed by incubating the proteins at 4°C in buffer A containing various concentrations of GdnHCl ranging from 0.0 M to 6.0 M overnight. The transition curves were obtained by detecting the changes in far-UV CD, intrinsic tryptophan fluorescence, extrinsic ANS fluorescence, light scattering and turbidity.
Protein thermal aggregation and refolding kinetics
The time course of thermal aggregation kinetics was monitored by heating the proteins at 75°C continuously and the turbidity (D400) data were recorded every 2 s for 30 min. The βB1–βA3 heterodimer was prepared by incubating equimolar of βB1 and βA3 at 37°C for 3–4 h. During refolding, the samples were incubated in buffer A containing 4 M GdnHCl at 25°C for 3–4 h to fully denature the proteins. The time course of refolding kinetics was obtained by a fast 1:40 manual dilution of the GdnHCl-denatured proteins into buffer A and the D400 data were recorded immediately after dilution. The dead time of manual dilution was approximately 2 s. The final GdnHCl and protein concentrations were 0.1 M and 5.8 μM respectively.
Limited proteolysis by trypsin was performed as described previously . In brief, the protein solutions with a protein concentration of 5.8 μM were treated by trypsin with a protein to protease ratio of 1:20 or 1:50 (w/w). The samples were incubated at 37°C for 2 h or 4 h and the tryptic fragments were analysed by SDS/12.5% PAGE.
Solubility of the WT and mutated proteins was determined using the method described elsewhere . In brief, the protein solutions in microminiature were centrifuged by a 10 kDa molecular-mass cut-off Millipore concentrator at 11000 g and 4°C. The protein concentration was determined every 15 min until the maximum concentration was reached. The final value was calculated from three repetitions and represented as the mean ± S.E.M.
Monolayer surface pressure measurements
Details of monolayer surface pressure experiments were the same as those reported recently . In brief, the membrane-binding ability of proteins was examined by the change in the surface pressure (π) of the Langmuir monolayer on an NIMA 9000 Microbalance at 22°C. A circular Teflon trough was filled with 6 ml of buffer A and the subphase was stirred with a magnetic rotor continuously. The DPPC monolayer was prepared by spreading DPPC dissolved in chloroform/methanol (3:1, v/v) at the air/water interface to reach a given initial surface pressure (πi). The protein solution with a final protein concentration of 200 nM was injected into the subphase from a sample hole after πi reached a stable value. Then, the time course of the change in surface pressure (Δπ) was recorded continuously for 2 h.
The X253R mutant forms intracellular aggregates and induces cell death
The X253R mutation of βB1 has been identified to be the cause of CMCC in a U.K. family, implying that the mutation might not only induce protein aggregation in the lens but also affect ocular development. Since crystallin mutants exhibit similar patterns in various cell lines including lens cells , HeLa cells were used to avoid misleading by endogenous WT proteins. As shown in Figure 1(A) and Supplementary Figure S1(A), the overexpressed WT βB1 showed a similar distribution pattern to that of the control cells overexpressing EGFP. The X253R mutant had a high propensity to form granule-like aggregates (Figure 1C) distributed in both the cytoplasm and nucleus of the cells (Supplementary Figure S1A). The difference between the WT and mutated proteins was more likely to be caused by their dissimilar propensities to aggregate, not by their expression levels in the HeLa cells (Supplementary Figure S1B). By measuring the GFP fluorescence intensity, protein expression level could be analysed at the single-cell level. The results showed that cells containing X253R aggregates had a relative higher protein expression level than cells without X253R aggregates (Supplementary Figure S1C), implying that the behaviour of X253R was strongly dependent on protein concentration. Intracellular crystallin aggregates usually co-localized with the p62-positive aggresomes [8,30,36]. However, most of the X253R aggregates were in distinct cellular foci from the p62-positive aggresomes (Figure 1B). The average Pearson's correlation coefficient (Rr) was 0.15±0.03 for cells containing X253R aggregates. Thus, the X253R proteins might form aggregates using a dissimilar mechanism to those well-defined instable crystallin mutants. It is noteworthy that recombinant X253R could only be obtained by culturing the E. coli cells at 10°C. Most X253R molecules were in the inclusion bodies when overexpressed in E. coli cells at 37°C, whereas the WT protein did not (results not shown). Thus, it seems that X253R mutation induced βB1 aggregation in both eukaryotic and prokaryotic cells.
The X253R mutant of βB1 forms intracellular aggregates and induces cell death
Cell viability and apoptosis assays were performed to investigate how the X253R mutation affects cell survival. As shown in Figure 1(D), the proliferation of HeLa cells was unaffected by exogenously expressed WT βB1, whereas it was significantly retarded by X253R. Cell viability began to fall after 60 h of cultivation for the X253R group, implying that long-term cultivation might induce cell death. To prove this hypothesis, the cells were double-stained with propidium iodide (PI) and annexin V and cell death was analysed by bivariate flow cytometry (Figures 1E and 1F). The exogenously expressed X253R mutant significantly increased the percentages of apoptotic and necrotic cells. These observations suggested that the exogenously expressed X253R mutant exhibited the toxic effect on cell survival and induced cell death, probably via the formation of abnormal intracellular aggregates.
The X253R mutation increases βB1 hydrophobicity and decreases βB1 solubility
Among various β/γ-crystallins, βB1 has the longest N- and C-terminal extensions (Figure 2A). The C-terminal extension of βB1 is probably flexible since its electron density is poor in the crystal structure of truncated βB1 . The extra 26 residues at the C-terminus of βB1 introduced by the X253R mutation are mainly composed of hydrophobic residues (Figures 2A and 2B). To study whether the elongated C-terminal extension affects βB1's biophysical properties, recombinant proteins were produced from E. coli. Consistent with recent observations , a comparison between the His-tagged and tag-free proteins indicated that the existence of a His-tag at the N-terminus did not affect the structural features and stabilities of both the WT βB1 and X253R mutant (Supplementary Figure S2). Thus, His-tagged proteins were used for most of biophysical analysis in the present study. The solubility of βB1 in water was significantly reduced by the X253R mutation (Figure 2C). Monolayer surface pressure experiments were conducted to further examine whether the elongated hydrophobic C-terminal extension of X253R could endow βB1 membrane-insertion ability. As shown in Figure 2(D), the WT βB1 had a very low propensity to insert into the DPPC monolayer as reflected by the slow change in surface pressure (Δπ), consistent with the fact that βB1 is a highly soluble cytoplasmic protein. At an initial surface pressure (πi) of approximately 20 mN/m, the X253R mutant exhibited a rapid increase in Δπ within 20 min, which indicated that X253R could insert into the DPPC monolayer in vitro. The critical surface pressures (πc) of the two proteins were obtained from the linear relationship between Δπ and πi (Figure 2E). The relative higher πc of X253R also suggested that X253R had stronger membrane-insertion ability than the WT βB1. However, the πc values of both proteins were lower than that of lipid bilayers in the cells (approximately 31 mN/m) . The πc values of the WT and mutated βB1 were also smaller than that of α-crystallin, which has been shown to have membrane-bound fractions in the cells . Thus, the increased membrane-insertion ability induced by X253R was not likely to occur in the cells, which was evident by no obvious membrane distribution of both proteins in the cells (Figure 1A). Because no increase in π can be observed for the non-specific interactions between soluble proteins and the charged head group of phospholipids , the monolayer surface pressure experiments suggested that the elongated C-terminal extension of X253R introduced extra hydrophobicity to βB1.
The X253R mutation increases βB1 hydrophobicity and decreases βB1 solubility
The potency of the elongated C-terminus to form extra ordered structures was analysed by PONDR scores, which predict the disordered region in proteins . As the control, the N- and C-terminal extensions of βB1 were predicted to be intrinsically disordered by most algorithms in the PONDR website (Supplementary Figure S3). As for X253R, the elongated C-terminal extension became partially ordered as predicted by most PONDR scores. Biophysical techniques were applied to the recombinant proteins to evaluate the effect of the mutation on βB1 structure. SEC analysis indicated that the two proteins had similar elution positions, suggesting that the mutation did not significantly affect the oligomeric status of βB1 (Figure 2F). The X253R mutant had a larger ellipticity in the far-UV CD spectrum (Figure 2G). However, the difference in their CD spectra suggested that the extra 26 residues were more likely to form random coils or disordered structures. The almost superimposed near-UV CD (Figure 2H) and tryptophan fluorescence spectra (Figure 2I) of the two proteins indicated that the mutation had no influence on the microenvironments around the aromatic residues. The exposed hydrophobic exposure was evaluated by ANS fluorescence (Figure 2J). Compared with the WT βB1, the mutant had more binding sites of the hydrophobicity probe ANS as revealed by approximately 2-fold increase in ANS fluorescence. Thus, these spectroscopic studies suggested that the extra 26 residues in X253R were more likely to be flexible or disordered and thereby they introduced extra hydrophobicity to βB1 by the additional hydrophobic residues (Figure 2B).
The X253R mutation has minor impact on βB1 stability but inhibits βB1 thermal aggregation
Cataract-causing mutations usually lead to instability of proteins in fighting against denaturation induced by various physical and chemical stresses. In the present study, we evaluated the effect of X253R mutation on βB1 stability against heat treatment, chemical denaturants and proteolysis. As shown in Figures 3(A)–3(D), there were considerable deviations of the thermal transition curves between X253R and the WT βB1 when a protein concentration of 5.8 μM (approximately 0.2 mg/ml) was used. βB1 began to partially unfolded and aggregate when heated at high temperatures above 70°C (Figure 3E), consistent with those previous and recent observations [11,13,19]. Unlike the WT protein, no aggregates could be identified even though X253R was heated at extreme high temperatures as revealed by the small change in turbidity. The thermal aggregation of both βB1 and X253R was protein-concentration-dependent (Supplementary Figure S4). When heated at 70°C, the critical concentration of βB1 thermal aggregation was below 0.05 mg/ml, whereas that of X253R was approximately 0.5 mg/ml. At a protein concentration of 5.8 μM, X253R began to unfold at a much lower temperature than the WT protein. At temperatures above 70°C where the WT βB1 began to aggregate, X253R continued to unfold. At around 80°C, the maximum emission wavelength of tryptophan fluorescence (Emax) of X253R reached a value of approximately 350 nm (Figure 3A), which is close to the Emax of free tryptophan fluorophores in water . Thus, the results in Figure 3 suggested that at low protein concentrations, the X253R mutation modified the irreversible thermal denaturation of βB1 from aggregation of partially unfolded molecules to complete loss of ordered structures at extreme high temperatures. Phase diagram analysis of the tryptophan fluorescence spectra also indicated that the thermal denaturation of βB1 involved the accumulation of an aggregation-prone intermediate appeared at approximately 68°C, whereas X253R exhibited a two-state process without the appearance of any intermediates (Supplementary Figure S5). Quantitative analysis by fitting the Parameter A transition curves indicated that the midpoint temperatures of thermal transition (Tm) were 71.3±0.2°C and 72.7±0.3°C for the WT and mutated protein respectively. The changes in enthalpy of thermal denaturation (ΔHm) were 870±96 kJ/mol and 281±51 kJ/mol for the WT and mutated proteins respectively. The increase in Tm and decrease in ΔHm were caused by the change in the reversibility of thermal denaturation induced by the X253R mutation. βB1 has been shown to protect βA3 in the heteromer to fight against aggregation [27,28]. Thermal aggregation kinetic studies (Figure 3F) indicated that the protective effect of X253R on βA3 was much stronger than the WT βB1 and could completely inhibit βA3 aggregation at high temperatures.
Effect of the X253R mutation on βB1 thermal stability
We further evaluated the effect of X253R mutation on βB1 stability against proteolysis and chemical denaturation. No significant difference was observed in the susceptibility to the limited protease trypsin (Supplementary Figure S6). The ionic denaturant GdnHCl unfolds proteins via disrupting both hydrophobic and electrostatic interactions. As shown in Figure 4, βB1 unfolding induced by GdnHCl was a multi-state process with the accumulation of an intermediate at approximately 2.0 M GdnHCl, consistent with the previous result . There were some deviations in the transition curves from far-UV CD and ANS fluorescence, which were caused by the difference in the spectral features of the native proteins (Figure 2). No aggregation-prone states could be observed during the denaturation of both proteins. Phase diagram analysis indicated that the mutation had no impact on βB1 unfolding pathway (Supplementary Figure S7). When monitored by tryptophan fluorescence, the transition curves of the two proteins were almost identical except that the curves of the WT protein shifted slightly to higher denaturant concentrations at GdnHCl concentration above 2.0 M. This suggested that the X253R mutation slightly destabilized the intermediate accumulated at approximately 2.0 M. However, no significant difference was observed in the thermodynamic parameters when fitting the data in Figure 3(C) using the method described previously  (results not shown).
Effect of the X253R mutation on βB1 folding
The intracellular X253R aggregates can be redissolved by lanosterol
Recently, we have identified that lanosterol can redissolve protein aggregates formed by various cataract-causing crystallin mutants . Unlike those well-studied crystallin mutants, X253R formed intracellular aggregates distinct from p62-positive aggresomes (Figure 1). To investigate whether lanosterol could reverse p62-negative aggregates, we treated the HeLa cells containing exogenously expressed X253R by various concentrations of lanosterol (Figure 5). As the control, cholesterol had no impact on the intracellular aggregation of X253R. Lanosterol could effectively reduce the percentage of cells with X253R aggregates in a concentration-dependent manner. Thus, it seems that lanosterol had a general effect on dissolving both p62-positive  and p62-negative (the present study) intracellular aggregates.
Intracellular X253R aggregates redissolved by lanosterol
Acting as the dominant structural proteins in the lens, β/γ-crystallins are proposed to be essential to the maintenance of lens transparency via their high solubility in water, lifelong stability and short-order interactions . Consequently, mutations in β/γ-crystallins usually lead to isolated autosomal dominant congenital cataracts. However, non-lens defects such as CMCC have also been identified to be caused by some β/γ-crystallin mutations. X253R in βB1 is the first crystallin mutation that was associated with CMCC in 2005 . Besides X253R in βB1, several mutations in αA, βA4, γC, γD have also been linked to the onset of CMCC [19,21,22,24,42–47]. Despite the increasing list of CMCC-causing mutations, the underlying molecular mechanism remains elusive. We hypothesize that CMCC-causing mutations may have dual effects: promoting crystallin aggregation and impairing cell survival. The former one is the main cause of cataract, whereas the later one may contribute to both lens cell and non-lens cell impairments. This hypothesis was verified by the results of the present study. The X253R mutation not only formed intracellular aggregates, but also inhibited cell proliferation and induced cell death. The observation of X253R aggregates redissolved by lanosterol suggested that lanosterol might have a general dissolving effect on aggregates formed by various crystallin mutants.
α-Crystallins have been identified as regulators of cell apoptosis , and thereby it is not surprising for cell death modulated by α-crystallin mutations. For example the R116C mutation in αA was shown to affect the protective effect of αA on inhibition of cell death in 2002 . Later in 2006, the R116C mutation in αA was linked to CMCC in an Indian family . Recently, it was shown that the G129C mutation in γC promotes cell apoptosis and the aggregated G129C proteins are toxic to the cells . Although mutations in γC have been associated with CMCC, the G129C mutation has been reported to associate with isolated cataract . Unlike R116C in αA, our results show that the X253R mutation significantly induced cell death under normal but not stressed conditions. Moreover, the percentage of apoptotic cells induced by the X253R mutant of βB1 (approximately 30%) was much larger than by the G129C mutant of γC (approximately 10%). Thus, our results suggest that the strong promotion effect on cell death under normal cell conditions might contribute to the abnormalities of both lens and non-lens cells in CMCC patients with X253R mutation.
Our biophysical studies suggested that the most possible reason for the intracellular aggregation of X253R was the hydrophobicity introduced by the extra 26 residues at the C-terminus of βB1. Cataract-causing mutations in β/γ-crystallins may lead to protein aggregation via diverse mechanisms such as a decrease in protein solubility, modifications of native structure and protein interaction networks, alteration of the folding pathway and an increase in susceptibility to denaturants, heat, UV, pH or proteases [7,9–17]. Despite these diverse effects, it is certain that a cataract-causing mutation will increase the protein-aggregatory propensity under a certain physiological or pathological condition. In some cases, a mutation could even increase protein stability, which can be regarded as a stability-beneficial mutation [11,13]. The stability-beneficial mutations are likely to occur at functional amino acid residues, which may be important for protein interactions. The correlation between aggregatory propensity and instability is also controversial since the mutant with the highest aggregatory propensity may not be the one with the lowest stability . Our results show that X253R could also be regarded as an aggregation-inhibition mutation since the mutated protein did not aggregate under all extreme conditions tested where the WT protein would aggregate. The major defects caused by the X253R mutation were the increase in hydrophobicity and decrease in solubility in water. Although the X253R mutation could effectively inhibit βB1 aggregation in diluted solutions, the extra hydrophobicity might facilitate the mutated proteins to aggregate in the crowded cytoplasm of human cells. The crowding environment has been proposed to promote protein association/aggregation . The human lens fibre cells are one of the tissues with the highest protein concentration and βB1 occupies approximately 9% of the total soluble crystallins . In this case, the significantly decrease in βB1 solubility might contribute to the intracellular aggregation induced by the X253R mutation.
Among various β/γ-crystallins, βB1 possesses the longest N- and C-terminal extensions. Previous studies have shown that the N-terminus of βB1 will degrade along with the increase in age . The N-terminal extension contributes little to βB1 structure and stability, but is important to βB1 folding, its protective effect on βA3 and the action of α-crystallins . It is interesting to find that, similar to the long N-terminal extension, the artificially extended C-terminus could also help βB1 as well as βB1/βA3 to fight against aggregation in diluted solutions. The dissimilar effects of the mutation on heat- and GdnHCl-induced denaturation suggested that the change in hydrophobicity accounted for the aggregation-inhibition effect of X253R. Due to the lack of high-resolution structure of the C-terminal extension and heteromeric β-crystallins, the structural basis of how the elongated C-terminal extension protected β-crystallins against aggregation is unclear. A possible explanation is that the hydrophobic elongated C-terminal extension might interact with the exposed hydrophobic areas during denaturation and thereby inhibit the adhesion of partially unfolded molecules via hydrophobic interactions. Further structural study is needed to elucidate the structural basis of the actions of the N- and C-terminal extensions of βB1.
Yong-Bin Yan conceived and designed the experiments. Xiao-Yao Leng, Liang-Bo Qi and Yi-Bo Xi performed the protein experiments. Hai-Yun Li and Jing Wang performed the cellular experiments. Xiao-Yao Leng, Hai-Yun Li and Yong-Bin Yan analysed the data. Yong-Bin Yan wrote the paper.
This work was supported by the Ministry of Science and Technology of China [grant number 2012CB917304]; the State Key Laboratory of Membrane Biology (to Y.-B.Y.); and the Tsinghua-Peking Joint Center for Life Sciences (to H.-Y.L. and Y.-B.X.).
Cell Counting Kit-8
congenital microcornea-cataract syndrome
Dulbecco's modified Eagle's medium
maximum emission wavelength of tryptophan fluorescence
protein data bank
These authors contributed equally to this work.