Cloning and characterization of a thermostable catfish aB-crystallin
with chaperone-like activity at high temperatures
Chung-Ming Yu
a,c, Gu-Gang Chang
b, Hui-Chuan Chang
b, Shyh-Horng Chiou
c,d,*
a
Graduate Institute of Life Sciences, National Defense Medical Center, Taipei, Taiwan
b
Faculty of Life Sciences, Institute of Biochemistry, National Yang-Ming University, Taipei, Taiwan
cLaboratory of Crystallin Research, Institute of Biological Chemistry, Academia, Taipei, Taiwan d
Institute of Biochemical Sciences, National Taiwan University, P.O. Box 23-106, Taipei 10617, Taiwan Received 1 March 2004; accepted in revised form 7 April 2004
Abstract
We have cloned, expressed and characterized catfish aB-crystallin (FaB). Genomic sequence comparison has revealed conservation of intron splicing sites and coding regions, however, the two intron sequences, 50- and 30-untranslated regions of FaB gene are shorter than those reported for other vertebrates. In contrast to mammalian homologues with a subunit association ratio (aA-crystallin/aB-crystallin) of 3:1, a-crystallin from catfish lens showed a ratio of 19:1. The biophysical properties and chaperone-like activity of recombinant FaB and porcine aB-crystallin (PaB) were studied and compared by heat denaturation, circular dichroism, intrinsic and dye-binding fluorescence, gel-filtration, and analytical ultracentrifugation. FaB shows 50% precipitation occurring at 728C that is higher than PaB at 668C. Even though FaB also possesses more surface hydrophilic regions than PaB, FaB still possesses higher chaperone activity to prevent aggregation of alcohol dehydrogenase at 608C. The molecular mass of FaB showed a smaller size (450 kDa) than PaB (550 kDa), which is also confirmed by analytical ultracentrifugation. In addition, FaB possesses better refolding potential after preheating treatment than PaB. FaB also exhibits higher chaperone-like activity than PaB to prevent insulin aggregation induced by dithiothreitol. In contrast to the prevalent notion that fish crystallins generally denature easily, FaB with chaperone-like activity appears to be more stable than mammalian homologues towards thermal and chemical denaturation.
q2004 Elsevier Ltd. All rights reserved.
Keywords: catfish eye lenses; aB-crystallin; thermostability; chaperone-like activity; small heat-shock protein; crystallin aggregation; surface hydrophilicity
1. Introduction
a-Crystallin constitutes a major class of lens proteins present in all vertebrate eye lenses (de Jong and Hendriks, 1986; Wistow and Piatigorsky, 1988). Native a-crystallin from mammalian lenses is commonly isolated as a large water-soluble aggregate with a molecular mass of about 600 – 800 kDa. It consists of two homologous subunits aA
and aB of about 55 – 60% sequence identity (van der Ouderaa et al., 1973, 1974), each with a molecular mass of , 20 kDa and in a binding ratio (aA/aB) of 3:1 for most mammalian lenses (Groenen et al., 1994).
a-Crystallin was shown to possess structural and functional similarities to small heat-shock proteins (sHSP) (Ingolia and Craig, 1982; Hendrick and Hartl, 1993). These proteins were first sequenced for sHSP isolated from Drosophila (Ingolia and Craig, 1982) and found to contain a highly conserved region strikingly similar to the C-terminal two-thirds of a-crystallin sequence, now denoted as a-crystallin domain (Ingolia and Craig, 1982; Wistow, 1985). The role of this conserved region was first suggested to be a highly stable structural unit or a useful building block for proteins under stressful environments (Wistow, 1985). However, in vitro studies of a-crystallin indicated a chaperone-like activity associated with this lens
0014-4835/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. DOI:10.1016/j.exer.2004.04.006
www.elsevier.com/locate/yexer
* Corresponding author. Address: Dr Shyh-Horng Chiou, Institute of Biochemical Sciences, National Taiwan University, P.O. Box 23-106, Taipei 10617, Taiwan, ROC.
E-mail address: [email protected] (S.-H. Chiou).
Abbreviations: ADH, alcohol dehydrogenase; ANS, 1-anilinonaphthalene-8-sulfonic acid; CD, circular dichroism; FaB, catfish aB-crystallin; HPLC, high performance liquid chromatography; PBS, phosphate buffered saline; PCR, polymerase chain reaction; PaB, porcine aB-crystallin; RT-PCR, reverse transcription coupled polymerase chain reaction; sHSP, small heat-shock protein; UTR, untranslated regions.
protein (Horwitz, 1992), which raises the supposition that a-crystallin is essential not only for the structural integrity required for lens transparency, but also for its chaperone-like role in maintaining the solubility of other lens proteins (Horwitz, 1993).
aB-Crystallin has been shown to be expressed in other tissues besides the lens suggesting that it may possess a general cellular function (Ingolia and Craig, 1982; Hendrick and Hartl, 1993). However, zebrafish aB-crystallin is expressed at extremely low levels outside of the lens (Posner et al., 1999). Expression of aB-crystallin was shown to confer protection to cells against thermal (Horwitz, 1992), osmotic (Golenhofen et al., 2002), and oxidative injuries (Klemenz et al., 1991). Most significantly relating to physiological functions is the finding that outside the lens tissue the R120G mutation of aB-crystallin can directly account for inheritable desmin-related myopathies (Vicart et al., 1998). Interestingly, aB-crystallin is also found to be overexpressed in many neurological disorders such as Creutzfeldt – Jacob disease (Renkawek et al., 1992), Alzheimer’s disease (Renkawek et al., 1994), diffuse Lewy body disease (Jackson et al., 1995), and Alexander’s disease (Iwaki et al., 1989). All the available data attest to the general functional significance of aB-crystallin.
Fish represents the oldest and most diverse group of vertebrates (Powers, 1989). The modern fishes comprise two major classes of piscine species, i.e. Osteichthyes or teleostean (bony) fishes, and Chondrichthyes or cartilagi-nous fishes (sharks and skates). The study of lens crystallins from the piscine class is of special interest from the evolutionary point of view because they constitute the early protein forms of vertebrates and are thought to be ancestral to those of land vertebrates. It is especially noteworthy that the abundant presence of various common and specific classes of structurally conserved crystallins in eye lenses of different species of vertebrates constitutes a good model system to unravel the complex process of evolution in structurally homologous proteins (Chiou, 1986; de Jong and Hendriks, 1986). Catfishes reside generally in the dark or underground environment, belonging to species of nocturnal scavengers. They thus constitute a rare species in verte-brates to study the evolutionary effects on the structure and function of aB-crystallin. We have characterized here the structural difference of aB-crystallins between catfish and other vertebrates on the genomic DNA and protein levels. Because the porcine aB-crystallin (PaB) possesses similar biophysical properties and chaperone-like activity to the well-studied aB-crystallins from human and bovine species in our previous studies (Liao et al., 2002b), we have tried to characterize the differences between biophysical properties and chaperone-like activity of catfish aB-crystal-lin (FaB) and PaB. Several important differences were found between aB-crystallins from catfish and porcine lenses, which may point out some important aspects of this sHSP in regard to its interactions with other crystallins and its function in the lens.
2. Materials and methods 2.1. Materials
Equine liver alcohol dehydrogenase (ADH) was obtained from Sigma. g-Crystallin was obtained from porcine eye lenses. All restriction enzymes were purchased from New England Biolabs. 1-Anilinonaphthalene-8-sulfonic acid (ANS) was purchased from Molecular Probes.
2.2. Preparation of mRNA and intron sequence analysis of FaB
Total RNA of catfish lenses was purified using the TRIZOL reagent (Invitrogen) following the protocol from the manufacturer. The 50- and 30-untranslated region (UTR) sequences were determined by the protocols from SMART Race cDNA amplification system (Clontech, BD Bios-ciences). The chromosomal DNA of catfish used in intron sequence determination was prepared according to the protocols of TRIZOL reagent after total RNA preparation. The first and second introns of FaB gene were amplified from catfish chromosome DNA by using two primer pairs based on the exon sequences covering the first and second introns of FaB gene. These polymerase chain reaction (PCR) products were cloned with pGEM-Teasy vector (Promega) for sequencing.
2.3. Construction and expression of cDNA clones for FaB and PaB
The coding region for FaB gene was amplified from catfish lens cDNA mixture and subcloned into pET21b expression vector (Novagen) with Nde I and Hind III cutting sites. Expression clone of PaB was constructed as described previously (Liao et al., 1998).
FaB and PaB were expressed in Escherichia coli BL21 (DE3) cells and purified by the methods described previously (Liao et al., 1998) with some modifications for FaB expression and purification. Because the recombinant FaB after gel-filtration analysis was easily precipitated in 50% acetonitrile used in the reverse-phase high-perform-ance liquid chromatography (HPLC) system, the precipi-tates containing FaB were centrifuged at 300 £ g for 10 min and washed once with 100% acetonitrile. The FaB precipitates could be dissolved in acidic ddH2O (0·1% TFA). The lyophilized recombinant FaB or PaB was dissolved in phosphate-buffered saline (PBS) with 8Murea and introduced to TSK HW-55 gel-filtration column for protein refolding and renaturation. The fractions containing purified FaB or PaB were pooled and concentrated by VIVASPIN Concentrator (100 kDa molecular-mass cut-off, Vivascience). The protein concentrations were determined by absorbance measurements using extinction coefficients calculated from amino acid sequence data of each protein (Gill and von Hippel, 1989). These purified proteins were
stored at 4 or 2 208C for short or long term storage, respectively.
2.4. Thermal stability analysis
Temperature-dependent aggregation of FaB and PaB were monitored in Ultraspec 4000 UV/VIS spectropho-tometer equipped with a Peltier type temperature controller (Amersham Pharmacia). The heating rate was 18C min21. The temperature range for protein heating treatment was from 25 to 958C. The protein aggregation was monitored by light scattering at 360 nm. Protein concentration used was 0·7 mg ml21in PBS. First derivative calculation was used to estimate the temperature at which 50% precipitation had occurred.
2.5. Circular dichroism spectra analysis
Circular dichroism (CD) spectra were performed on a JASCO J-715 spectropolarimeter. The temperatures for CD measurement were controlled and maintained by a thermo-static, circulating water bath. The far-UV CD spectra were recorded as the mean of 5 accumulations with a 0·02 cm path length water-jacket cell from 250 to 190 nm and the near-UV CD spectra as the mean of 10 accumulations with a 1 cm path length water-jacket cell from 360 to 250 nm under constant N2 flush and constant-temperature water flow. All CD spectra were corrected for their respective buffer blanks. Spectra were also noise-reduced by poly-nomial curve fitting program supplied by the manufacturer, with the precaution of not over-smoothing the data points. The CD data were expressed as molar ellipticity in ½u (degrees cm2dmol21). The wavelength scan data was recorded in 0·2 nm steps with 1·0 nm bandwidth, 50 nm min21 scanning speed and 2 sec response time. Protein concentrations used in both CD spectra were 0·7 mg ml21in 10 mMphosphate buffer, pH 7·4.
2.6. Fluorescence measurements
Intrinsic and ANS-binding fluorescence were recorded by an F4500 fluorescence spectrophotometer (Hitachi) with circulating water bath to maintain desired temperatures. The cuvette with a 0·3 cm £ 0·3 cm cross section and 60 ml of protein sample each were used for measurements. The intrinsic fluorescence of Trp residues was measured with an excitation wavelength at 295 nm. ANS was used to determine the surface hydrophobicity of protein molecule (Johnson et al., 1979) and measured with an excitation wavelength at 395 nm. To determine the temperature-dependent changes of Trp intrinsic and ANS-binding fluorescence, the equilibrium time of protein samples for each specified temperature was 30 min in a thermostatic controller before measurements. The protein concentration used was 100 mg ml21 in PBS. ANS solution (10 ml of 10 mM stock solution) was thoroughly mixed with 1 ml
protein samples, incubated for 5 min at specified tempera-tures to establish equilibrium. The emission spectra were recorded from 300 to 400 nm and from 400 to 600 nm for the Trp intrinsic and ANS-binding fluorescence measure-ments, respectively. The excitation and emission bandwidth were set at 5 nm. All quantum yields of fluorescence spectra were corrected for their respective buffer blanks or ANS only at different temperatures.
2.7. Gel-permeation FPLC analysis
Multimeric sizes of recombinant FaB and PaB were evaluated by an analytical Superose-6 HR 10/30 column with FPLC system (Amersham Pharmarcia). Molecular-weight standards for gel-filtration (Sigma) were used for calibration. The protein concentration used was 1 mg ml21 in PBS and 0·4 ml sample was auto-injected. The column was eluted at 0·4 ml min21 with PBS. All samples were passed through the syringe-driven filter (pore size 0·45 mm) before analysis. The width-of-half-height of each sample was determined by the program equipped in FPLC system and used to estimate the polydisperse property of FaB and PaB.
2.8. Analytical ultracentrifugation analysis
The molar-mass distribution of FaB and PaB under various preheating treatments was estimated by a Beckman-Coulter XL-A analytical ultracentrifuge with an An60Ti rotor. Sedimentation velocity was performed at 208C and 25 000 rpm with standard double sectors aluminium cen-trepieces. The UV absorption of the cells was scanned every 5 min for 2 h. The data were analyzed with the SedFit program (Schuck, 2000). The solvent density and viscosity were corrected with the UltraScan version 5·0 (Demeler and Saber, 1998). A partial specific volume of 0·705 and 0·718 were used for FaB and PaB, respectively (Harpaz et al., 1994). All samples were visually checked for clarity after ultracentrifugation. The protein concentration used was 1 mg ml21in PBS.
2.9. Chaperone-like activity assay
Chaperone-like activities of FaB and PaB were analyzed by measuring the capability to prevent the aggregation of ADH denatured by heating treatment or the reduction of disulfide bonds in insulin. The aggregation assays with different substrate proteins were performed essentially as described previously (Horwitz et al., 1998a). The extent of aggregation was measured as a function of time by monitoring the light scattering at 360 nm in a Pharmacia F-4000 fluorescence spectrophotometer with Peltier type controller and recorded at 1-s intervals. The aggregations of ADH (0·4 mg ml21) in PBS with FaB or PaB in a molar ratio of 2:1 (ADH/aB-crystallin) were monitored at 50 and 608C. The dithiothreitol (DTT)-induced aggregation of
insulin (0·4 mg ml21) in PBS was studied at 25 and 378C with various molar ratios of insulin to FaB or PaB. The suppression of aggregation was calculated by the percent (%) change of absorbance at 360 nm in the absence and presence of aB-crystallins.
3. Results and discussion
3.1. Genomic structure of FaB gene
Our recent interest has been focused on one teleostean species, i.e. the catfish, a nocturnal scavenger which is commonly raised in the local aquacultures of Taiwan. We have amplified cDNAs constructed from the lenses of catfishes by PCR methodology in order to aid in the genomic analysis of the major eye lens protein with chaperone-like function, i.e. a-crystallin consisting of aA-and aB-crystallin subunits. Important insights into the evolutionary history of different protein families are being discovered through the analysis of molecular sequences.
The FaB gene encodes a protein of 172 amino acids in length. The genomic structure and all intron/exon bound-aries of aB-crystallins reported for vertebrate species including catfish are conserved. The three exons of FaB gene code for residues 1 – 64, 65 – 105, and 106 – 172 of FaB protein sequence, respectively. The first and second introns are located between position 70 and 71, and between position 111 and 112, respectively, of the alignment in Fig. 1. However, the lengths of the two intron sequences of FaB gene (375 and 111 bps in length, respectively) were much shorter than the homologues from other vertebrates (Table 1). The 50and 30-UTR sequences of FaB gene were also found slightly shorter than homologous sequences from other vertebrates (Table 1). The general observation in genomic analysis is that more complex organisms generally have more introns with longer lengths (Logsdon, 1998). The numbers of protein-coding genes do not increase exponen-tially in complex organisms according to the results of genome projects currently established (Chervitz et al., 1998; Roest Crollius et al., 2000; Rubin et al., 2000), but the intron density, size and sequence complexity correlate well with their developmental complexity (Mattick and Gagen, 2001). The gene and protein sequence comparison shown for aB-crystallins below seem to indicate that catfish belongs to fishes of ancient evolutionary status.
3.2. Protein sequence alignment and comparison
Based on conserved domain homology search by blastp module in BLAST server (http://www.ncbi.nlm.nih.gov/ blast), FaB, which is similar to its homologues, could be separated into three distinct structural domains, the N-terminal domain (residues 1 – 59, which was three amino acids shorter than PaB), C-terminal a-crystallin
domain (residues 60 – 159) and C-terminal extension (residues 160 – 172).
FaB shows high sequence similarity with other verte-brate species based on protein sequence alignment. Protein sequence alignment of FaB with other vertebrate homol-ogues is analyzed by CLUSTAL V program (Higgins et al., 1991) in the MegAlign module of Lasergene package (DNASTAR, Inc., Madison, WI, USA) (Fig. 1). A distinct gap was introduced between residues #44 (valine) and #45 (tyrosine) of FaB, which corresponded to a conserved SPF (Ser-Pro-Phe) tripeptide in the sequences of other ver-tebrates except zebrafish. This tripeptide contains a Ser residue that could be phosphorylated in vivo (Chiesa et al., 1987). Another conserved Ser residue at position 19 in mammalian aB-crystallins was absent in FaB, which was also found to be lost in zebrafish aB-crystallin (Posner et al., 1999). The Ser residue at position 62 of the alignment in Fig. 1 is conserved in all vertebrate aB-crystallins, which was also shown to be phosphorylated in vivo (Chiesa et al., 1987). However, the physiological significance of phos-phorylation at these conserved serine residues remains unknown. It does not appear to play some essential role in chaperone-like activity as judged by the substitution or lack of serine residues in FaB.
Table 2 summarizes the basic parameters obtained by comparison of protein sequences of aB-crystallins from catfish and other vertebrates. It is noteworthy that W9 was conserved in all vertebrates except birds. The most conserved Trp residue was located at position 63 of the alignment in Fig. 1. In addition, another conserved Tyr residue was located at position 125 of the alignment in Fig. 1. The estimated molar extinction coefficient ð1280Þ of
FaB based on amino acid contents (Gill and von Hippel, 1989) was 27 880M21cm21which is two times higher than those of mammalian vertebrates (13 940M21cm21) because of its high Trp content (4 versus 2 residues per molecule). In addition, zebrafish possesses the highest 1280
(33 000M21cm21) due to the highest content of Tyr
residues. Note that all Trp residues found in aB-crystallins are located at the N-terminal domain (Fig. 1). In general, the estimated pI values of mammals were higher than birds and fishes. The contribution of these conserved Trp and Tyr residues to the structure and chaperone-like activity of aB-crystallin is unknown; however, these residues could be good targets for site-specific mutagenesis in biophysical studies (Liang et al., 1999).
The systematic pair-wise protein sequence comparison of FaB and 12 homologous aB-crystallin sequences from four major classes of vertebrates as templates for the construction of a phylogenetic tree were performed by using a combination of distance matrix and approximate parsimony methods (Hein, 1990) (Fig. 2). It is noteworthy that the phylogenetic tree based on the sequence divergence among these protein sequences indeed exemplifies the dissimilarity between piscine aB-crystallins (e.g. catfish, dogfish and zebrafish) and other terrestrial vertebrate species (Fig. 2).
The surface hydrophilicity analysis for the distribution of amino acids along primary protein sequences of FaB and PaB was analyzed (Fig. 3). The mean Kyte – Doolittle hydropathy of the polypeptide (Kyte and Doolittle, 1982) of FaB possessed higher hydrophilic content (0·7) than that of PaB (0·5), especially in the N-terminal segment (0·4 and 0·1 for FaB and PaB, respectively) and C-terminal extension region (1·2 and 0·7 for FaB and PaB, respectively). In addition, FaB also appears to possess a more hydrophilic a-crystallin domain than PaB (0·8 versus 0·7).
Fig. 1. Multiple sequence alignment of 13 aB-crystallin protein sequences from species of different vertebrate classes. The identical amino acid residues among various sequences based on the first one (catfish) are shaded in black blocks. The gaps denoted in dashes were introduced for optimal alignment and maximum homology for the aligned sequences. The splicing sites of aB-crystallins are conserved and the first and second introns are located between position 70 and 71, and between position 111 and 112, respectively. These aB-crystallin sequences shown are from catfish (Clarias batrachus), zebrafish (Danio rerio), spiny dogfish (Squalus acanthias), bullfrog (Rana catesbeiana), human (Homo sapiens), bovine (Bos taurus), porcine (Sus scrofa), mouse (Mus musculus), rat (Rattus norvegicus), blind rat (Spalax judaei), rabbit (Oryctolagus cuniculus), golden hamster (Mesocricetus auratus), chicken (Gallus gallus), and duck (Anas platyrhynchos) in GenBank. Amino acid residues are denoted by one-letter symbols.
Table 1
The length of non-coding regions of aB-crystallin in different genomic sequences from various vertebrates (from GenBank)
Accession number Species Intron I (bp) Intron II (bp) 50-UTR (bp) 30-UTR (bp) M73741 Mouse 1048 1656 475 138 U04320 Rat 1020 1981 44 653 U16124 Duck 1492 1111 39 702 AY184812 Catfish 375 111 29 31
The comparison points to the fact that FaB is more hydrophilic than PaB, resulting in a higher solubility of FaB compared to PaB especially at high temperatures (see below).
3.3. Expression and purification of catfishaB-crystallin The gel-filtration of lens crystallins from catfish showed that there were three major peaks (data not shown), which is
Table 2
Comparison of basic parameters based on protein sequences of aB-crystallins from catfish and other vertebrates Trp (W) content and position Estimated extinction coefficient ð1280Þ (M21cm21) Protein length Estimated pI Molecular weight
Tyr (Y) content and position
Zebrafish 4 (W9, W52, W57, W58) 33 000 168 5·77 19977·5 8 (Y10, Y21, Y27, Y42, Y46, Y47, Y50, Y119) Catfish 4 (W9, W15, W54, W57) 27 880 172 6·44 19831·1 4 (Y41, Y45, Y94, Y119)
Dogfish 3 (W9, W50, W62) 20 910 177 6·38 20254·1 3 (Y49, Y115, Y124)
Bullfrog 2 (W9, W58) 17 900 173 6·31 20078·8 5 (Y15, Y47, Y50, Y111, Y120) Chicken 2 (W17, W59) 12 660 174 6·31 20019·9 1 (Y121)
Porcine 2 (W9, W60) 13 940 175 6·76 20128·8 2 (Y48, Y122) Human 2 (W9, W60) 13 940 175 6·76 20158·9 2 (Y48, Y122)
The basic parameters were obtained from the analysis by using the ExPASy server with ProtParam tool.
Fig. 2. Phylogenetic tree of 13 aB-crystallins from species of different vertebrates classes based on protein sequences. Analysis of sequence data was carried out in MegAlign program module in Lasergene package (DNASTAR, Inc., Madison, WI, USA). The percent divergence is calculated by comparing sequence pairs in relation to the relative positions in the phylogenetic tree, in contrast to the percent identity, which is estimated by comparing percent sequence identity directly without accounting for phylogenetic relationships. A phylogenetic tree (A) was then constructed based on the percent divergence (B) between protein sequences using a combination of distance matrix and approximate parsimony methods in the phylogeny generation program ofHein (1990). The tree was built using the CLUSTAL V program and weighted residue-weight table. The length in each branch represents the sequence distance between aligned pairs. The scale beneath the tree measures the distance between sequences (in millions of years). The dotted lines shown in (A) point to the fact that the sequence distance based on protein sequence comparison is not proportional to the scale.
similar to our previous studies in carp (Chiou et al., 1986). a-Crystallin from catfish lenses was shown to co-elute with b-crystallins (Fig. 4, Lane 3), similar to our previous study for carp a-crystallin (Chiou et al., 1986). However, the ratio of aA and aB subunits of catfish a-crystallin was found to be about 19:1 by densitometric analysis (Scion Image beta 4·02 for Win, Scion Co., Frederick, Maryland, USA) on the SDS-PAGE of the fractions containing catfish a-crystallin, whereas a ratio of 3:1 was found in all mammalian homologues (Groenen et al., 1994). Because of the low abundance of FaB in catfish lens, we could not obtain enough proteins for detailed structural and functional analyses. Therefore, we construct the recombinant FaB for biophysical and chaperone-like activity assays. It is noted that catfish aA-crystallin could not be refolded after denaturation, which hampered its isolation and purification by C4 reverse phase HPLC system used for mammalian aA-crystallin purification and characterization (Perry and Abraham, 1986). The purification of recombinant catfish aA-crystallin will be reported later using another protocol described byMerck et al. (1992).
FaB was overexpressed in E. coli with an apparent mole-cular mass of 20 kDa (Fig. 4, Lane 1). The purity of recombi-nant FaB from reversed-phase HPLC was shown to be higher than 99% (Fig. 4, Lane 2). Analysis by mass spectrometry (19 831·6 Da) also confirmed its correct molecular mass (19 831·1 Da from sequence). It is noted that the lyophilized FaB from reversed-phase HPLC can be dissolved in 8Murea
and refolded by gel-filtration chromatography to get a multimeric homoaggregate of about 450 kDa.
3.4. Comparison of thermal stability
In order to compare the difference of thermostability, the aggregation of FaB and PaB was monitored by light scattering at 360 nm (Fig. 5) from 25 to 858C. The PaB did
Fig. 3. Hydrophilicity analysis of FaB (red column) and PaB (black column). The hydrophilicity plots of FaB and PaB were performed by using Protean module in Lasergene package (DNASTAR, Inc., Madison, WI, USA). The hydrophilicity scores shown in the plot (window size ¼ 9 amino acids) were based on the hydrophilic properties of amino acid residues in the protein primary sequence, higher positive scores indicating more hydrophilic. Three blank spaces were inserted to FaB protein sequence between V44 and Y45 in order to fill up the gap between the sequence alignments of the two aB-crystallins.
Fig. 4. SDS-PAGE analysis of recombinant FaB and native catfish lens a-crystallin. Lane 1, overexpressed crude lysate of recombinant FaB; Lane 2, purified recombinant FaB; Lane 3, a-crystallin fraction isolated from lenses. The arrows indicate aA and aB subunits of FaB. The included region with character ‘}’ indicates the b-crystallins associated with catfish lens a-crystallin. The ratio of aA/aB is about 19:1, which is dramatically different from 3:1 for most mammalian a-crystallin. Standard protein markers (in kDa) are shown on the left lane: phosphorylase b (94), bovine serum albumin (67), ovalbumin (43), carbonic anhydrase (30), soybean trypsin inhibitor (20), and lysozyme (14).
not aggregate up to 658C with a 50% precipitation occurring at 668C, whereas the FaB remained soluble up to 708C and showed a 50% precipitation occurring at 728C (Fig. 5). These results suggested that the thermostability of FaB was higher than those of PaB. The temperature for 50% aggregate formation of PaB is similar to that of human aB-crystallin (64·58C), which has been found to correlate with the structural transition temperature in far-UV CD analysis (Perng et al., 1999).
3.5. Comparison of secondary and tertiary structures Far-UV CD can be used to estimate secondary structure contents, whereas near-UV CD can be used to monitor the microenvironments of aromatic amino acids present in native proteins (Dinner et al., 2000). The far- and near-UV CD spectra measured with temperature and wavelength scan of FaB and PaB were used to compare the structural difference between these two proteins (Fig. 6). The wavelength scan CD spectra in the far- and near-UV ranges of FaB and PaB showed that the molar ellipticity of each spectrum decreased at elevated temperatures (Fig. 6 and data not shown for near-UV CD). Secondary-structure estimation based on far-UV CD spectra for FaB and PaB under various temperatures were analyzed by SELCON 3 (Sreerama et al., 1999) in the CDPro package. FaB contained higher molar ellipticity (more negative) and higher b-sheet content (32%) than PaB (28%) at 258C (Fig. 6). These results are also consistent with the prevailing evidence that a-crystallin possesses mostly b-sheet second-ary structures (Bloemendal et al., 1999). PaB contains twice as much random coil conformations (44%) than FaB (22%) at 608C. However, FaB possesses more b-sheet structure at
608C (42%) than at 258C (32%). These data also suggested that FaB is more thermostable than PaB.
3.6. Intrinsic tryptophan fluorescence spectra and surface hydrophobicity
The intrinsic fluorescence of Trp residue can provide useful information of the conformational changes within a protein. The emission wavelength maxima (Trp lmax) and
fluorescence intensity at emission wavelength maxima ðITrpÞ were recorded to compare the conformational difference between FaB and PaB. Because FaB and PaB possess 4 and 2 Trp residues in the N-terminal segment, it is not surprising to find that the ITrpof FaB is higher than PaB at
258C; however, the ITrpof FaB was lower than PaB at 608C
(Fig. 7(A)). The Trp lmax of FaB showed no significant
changes from 25 (340·8 nm) to 608C (341·4 nm), however, the Trp lmax of PaB are significantly red-shifted from 25
(338·8 nm) to 608C (344·5 nm) (Fig. 7(A)). The change (in intensity) observed in FaB upon heating to 608C is probably due to a local displacement of a quenching group, since no corresponding conformational change is observed in CD (Fig. 6), while the redshift in the Trplmaxof PaB shows that
its unfolding transition already commences at 608C, which is also corroborated by CD (Fig. 6).
We have also used ANS as an environmentally sensitive fluorophore to probe the surface hydrophobicity differences in FaB and PaB under different temperatures. The emission wavelength maxima (ANSlmax) and fluorescence intensity
at emission wavelength maxima ðIANSÞ were recorded to
compare the surface hydrophobicity difference between FaB and PaB. PaB showed not only higher IANS but also
smaller ANS lmax than FaB at different temperatures
(Fig. 7(B)). It suggests that PaB possesses more hydro-phobic surface than FaB at temperatures below 608C.
Fig. 5. Temperature stability of FaB and PaB. (A) Thermal stability of recombinant FaB (K) and PaB (W). Temperatures at 50% precipitation during heating process were calculated to be 66 and 728C for PaB and FaB, respectively.
Fig. 6. The far-UV CD spectra of FaB and PaB at different temperatures. (B), FaB at 258C; (A), PaB at 258C; (O), FaB at 608C; (K), PaB at 608C. Note that PaB at 608C shows partially unfolded structure with CD minimum shifting to 207 nm, indicating a loss of b-sheet structure.
The ANS lmax of FaB and PaB were both slightly
red-shifted from 25 to 608C and the IANS also showed
temperature dependent decrease in both aB-crystallins (Fig. 7(B)). These data appeared to suggest that FaB exposed much less hydrophobic regions, which is consistent with the predicted results of hydrophilicity analysis (Fig. 3).
3.7. Analytical gel-filtration chromatography
The quaternary conformation of FaB and PaB was analyzed by analytical gel-filtration chromatography. The estimated molecular mass of FaB was 450 kDa, which was
smaller than PaB (550 kDa) but the width-of-half-height (volume) of FaB (1·5 ml) was larger than PaB (1·3 ml) (Fig. 8). Therefore, FaB is more dispersed than PaB regarding its aggregation state. It indicated that FaB possesses a larger size distribution than PaB. After preheating for 30 min, we found that FaB could maintained its oligomerization as non-preheated FaB after preheating treatment at 608C; however, PaB preheated at 508C possessed a larger molecular mass (580 kDa) than non-preheated PaB but with a similar width-of-half-height (Fig. 8). Thus, preheating treatment increases the molecular size of PaB without affecting its size distribution but no effects on the molecular size and size distribution of FaB. Similar preheating effect on molecular mass of PaB was also found in human aB-crystallin (Sun and Liang, 1998). These results indicated that the refolding of FaB after preheating treatment is better than PaB.
3.8. Analytical ultracentrifugation
In order to obtain hydrodynamic information about the size and shape of aB-crystallin, FaB and PaB were analyzed by the sedimentation velocity in an analytical ultracentrifuge. Continuous sedimentation coefficient anal-ysis of the sedimentation data was analyzed by the SEDFIT program (Schuck, 2000) (Fig. 9). The A – C plots shown in Fig. 9 indicated that the fitting results were consistent and reliable. The curves of aB-crystallins revealed a polydis-perse system with a broad distribution of different size species similar to our previous report (Liao et al., 2002a). The sedimentation coefficient ðSÞ of the FaB was about 12S;
Fig. 7. Intrinsic tryptophen (Trp) and ANS-binding fluorescence spectra of FaB and PaB. (A) The intrinsic Trp fluorescence spectra of FaB at 25 (B) and 608C (O), and PaB at 25 (A) and 608C (K). The greater decrease of Trp intrinsic fluorescence in FaB as compared to PaB at 608C is probably due to higher exposure of Trp residues to solvent by its higher surface hydrophilicity. (B) The ANS-binding fluorescence spectra of FaB at 25 (B) and 608C (O), and PaB at 25 (A) and 608C (K). Note that ANS-binding fluorescence is lower for FaB than PaB at 608C, even though FaB possesses higher chaperone-like activity at 608C than PaB.
Fig. 8. Analytical gel-filtration analysis of the recombinant FaB and PaB. The molecular masses for FaB (A), 608C preheated FaB (K), PaB (B) and 508C preheated PaB (O) were 450, 450, 550 and 580 kDa, respectively. The width-of-half-height in volume of FaB, 608C preheated FaB, PaB and 508C preheated PaB were 1·5, 1·5, 1·3 and 1·3 ml, respectively. The symbols (L) indicate the elution time of molecular-mass standards, 1, blue dextran (, 2000 kDa); 2, bovine thyroglobulin (669 kDa); 3, horse spleen apoferritin (443 kDa); 4, sweet potato b-amylase (200 kDa); 5, yeast alcohol dehydrogenase (150 kDa).
which was smaller than PaB ð14·5SÞ (Fig. 9(D)). The data are compatible with the gel-filtration results (Fig. 8) that FaB is smaller in size than PaB. The sedimentation coefficient ðSÞ of the FaB preheated at 608C is the same
as non-preheated FaB and PaB preheated at 508C possesses a larger S value ð15·5SÞ than non-preheated PaB, which are compatible to the results of gel-filtration after preheating treatment.
Fig. 9. Analytical ultracentrifugation analysis of FaB and PaB. (A) Trace of absorbance at 280 nm during sedimentation (line) and from fitted data (circle) of non-preheated FaB. (B) Fitting residuals of the data shown in (A). (C) Continuous cðSÞ plot of the analyzed result from (A). (D) Comparison of cðSÞ plots for the FaB (B), FaB preheated at 608C (A), PaB (O), and PaB preheated at 508C (K). The sedimentation coefficient ðSÞ of the native and preheated FaB and PaB were 12S; 12S; 14·5S and 15·5S; respectively.
Both results from gel-filtration and analytical ultracen-trifugation indicate that FaB possesses better refolding potential than PaB.
3.9. Comparison of chaperone-like activity
The chaperone-like activities of FaB and PaB against thermal aggregation were performed by using ADH as substrate at 50 (Fig. 10(A)) and 608C (Fig. 10(B)). At a molar ratio of 2:1 (ADH/aB-crystallins), the suppression of ADH aggregation at 508C by PaB (about 83%) was better than FaB (about 66%). The data are consistent with the ANS-binding experiments (Fig. 7(B)), showing that a-crystallin with more hydrophobic surface possesses better chaperone-like activity (Raman and Rao, 1994). However, even though FaB was less hydrophobic as revealed by its lower IANS than PaB at 608C (Fig. 7(B)),
FaB possessed better chaperone-like activity than the PaB (about 43 versus 16%) (Fig. 10(B)). The contradictory data obtained for assays at two different temperatures might be
accounted for by the instability of PaB at 608C (Liao et al., 2002a,b) (Fig. 5). These data also suggest that FaB is more thermostable than PaB.
Reduction of disulfide linkage between insulin A and B chains leads to the aggregation of the B chains (Sanger, 1949). The aggregation could be suppressed by a-crystal-lin (Farahbakhsh et al., 1995). The chaperone-like activities of FaB and PaB were also analyzed by DTT-induced insulin aggregation (Fig. 10(C)). It was surprising to find that PaB could not prevent DTT-induced insulin aggregation at 258C with a molar ratio of 28:1 (insulin/ aB-crystallin) in contrast to 32% protective ability observed for FaB (data not shown). At 378C and a molar ratio of 28:1 (insulin/aB-crystallin), FaB could totally prevent the insulin aggregation induced by DTT, however, PaB had just about 6% chaperone-like activity (Fig. 10(C)). At a molar ratio of 56:1, FaB still possessed 54% chaperone-like activity while PaB possessed 81% chaperone-like activity only at a high molar ratio of 7:1 (Fig. 10(C)). These data indeed suggest that FaB possesses
Fig. 10. Comparison of chaperone-like activity between FaB and PaB. (A) ADH aggregation at 508C with a molar ratio of 2:1 (ADH/aB-crystallin), (B), ADH alone; (O), ADH plus FaB; (V), ADH plus PaB. (B) ADH aggregation at 608C with a molar ratio of 2:1 (ADH/aB-crystallin), (B), ADH alone; (O), ADH plus FaB; (V), ADH plus PaB. (C) Insulin aggregation at 378C, (B), insulin only; (O), insulin plus FaB with a molar ratio of 28:1 (insulin/aB-crystallin); (K), insulin plus FaB with a molar ratio of 56:1; (V), insulin plus PaB with a molar ratio of 28:1; (S), insulin plus PaB with a molar ratio of 7:1. It is noted that FaB at a much lower chaperone/substrate ratio of 1:28 can completely inhibit DTT-induced insulin aggregation than PaB (1:7).
abnormally higher chaperone-like activity than PaB in the DTT-induced insulin aggregation, even though PaB possesses larger surface hydrophobic regions than FaB at 25 and 378C (Fig. 7(B)).
There is no reason why the two methods should give identical results, even though most reports in the literature imply that the two methods (heat-induced ADH aggregation and DTT-induced insulin aggregation) should provide the same results (Plater et al., 1996; Horwitz et al., 1998b; Lindner et al., 1998) In contrast to the chaperone-like activity to prevent heat-induced protein aggregation (Rao et al., 1993; Raman and Rao, 1994), there is no mechanistic basis to correlate surface hydrophobicity with the chaper-one-like activity against chemical denaturation, such as DTT-induced insulin aggregation.
4. Conclusion
In this report, we have cloned, expressed and character-ized FaB. The comparison between FaB and PaB was carried out regarding their thermostability, protein confor-mation and chaperone-like activity. cDNAs of FaB gene was constructed from catfish lenses in order to aid in the genomic analysis of aB-crystallin. The structure – function characterization of FaB bears special biological significance because it belongs to an authentic member of the sHSPs family with chaperone-like activity. FaB possesses higher structural stability than PaB as judged by thermostability measurements (Fig. 5), secondary structure predictions based on CD measurements (Fig. 6), and chaperone-like activity by suppression of ADH aggregation at 608C (Fig. 10(B)). The distinct properties found in FaB in contrast to most mammalian homologous aB-crystallins coupled with detailed sequence information will provide some deeper insights in the elucidation of structure – function relationship of a-crystallin, and the mechanism underlying crystallin thermostability and the accompanying chaperone-like activity in the near future. Site-specific mutagenesis of FaB on several putative essential sites for aB-crystallin thermostability and chaperone-like activity is currently in progress.
Acknowledgements
This work was supported in part by Academia and the National Science Council (NSC Grants 89-2311-B-002-098, 90-2311-B-002-022 and 91-2311-B-002-015 to S.-H. Chiou), Taipei, Taiwan. We thank Dr P.-Y. Chan for reading and detailed discussion on the manuscript before submission. This report will be submitted as part of a dissertation by C.-M.Y. to Graduate Institute of Life Sciences, National Defense Medical Center in partial fulfilment of the PhD degree.
References
Bloemendal, M., Toumadje, A., Johnson Jr, W.C., 1999. Bovine lens crystallins do contain helical structure: a circular dichroism study. Biochim. Biophys. Acta 1432, 234 – 238.
Chervitz, S.A., Aravind, L., Sherlock, G., Ball, C.A., Koonin, E.V., Dwight, S.S., Harris, M.A., Dolinski, K., Mohr, S., Smith, T., Weng, S., Cherry, J.M., Botstein, D., 1998. Comparison of the complete protein sets of worm and yeast: orthology and divergence. Science 282, 2022 – 2028. Chiesa, R., Gawinowicz-Kolks, M.A., Kleiman, N.J., Spector, A., 1987.
The phosphorylation sites of the B2 chain of bovine a-crystallin. Biochem. Biophys. Res. Commun. 144, 1340 – 1347.
Chiou, S.H., 1986. Phylogenetic comparison of lens crystallins from the vertebrate and invertebrate-convergent or divergent evolution? FEBS Lett. 201, 69 – 73.
Chiou, S.H., Chang, T., Chang, W.C., Kuo, J., Lo, T.B., 1986. Characterization of lens crystallins and their mRNA from the carp lenses. Biochim. Biophys. Acta 871, 324 – 328.
de Jong, W.W., Hendriks, W., 1986. The eye lens crystallins: ambiguity as evolutionary strategy. J. Mol. Evol. 24, 121 – 129.
Demeler, B., Saber, H., 1998. Determination of molecular parameters by fitting sedimentation data to finite-element solutions of the Lamm equation. Biophys. J. 74, 444 – 454.
Dinner, A.R., Sali, A., Smith, L.J., Dobson, C.M., Karplus, M., 2000. Understanding protein folding via free-energy surfaces from theory and experiment. Trends Biochem. Sci. 25, 331 – 339.
Farahbakhsh, Z.T., Huang, Q.L., Ding, L.L., Altenbach, C., Steinhoff, H.J., Horwitz, J., Hubbell, W.L., 1995. Interaction of a-crystallin with spin-labeled peptides. Biochemistry 34, 509 – 516.
Gill, S.C., von Hippel, P.H., 1989. Calculation of protein extinction coefficients from amino acid sequence data. Anal. Biochem. 182, 319 – 326.
Golenhofen, N., Ness, W., Wawrousek, E.F., Drenckhahn, D., 2002. Expression and induction of the stress protein aB-crystallin in vascular endothelial cells. Histochem. Cell Biol. 117, 203 – 209.
Groenen, P.J., Merck, K.B., de Jong, W.W., Bloemendal, H., 1994. Structure and modifications of the junior chaperone a-crystallin. From lens transparency to molecular pathology. Eur. J. Biochem. 225, 1 – 19. Harpaz, Y., Gerstein, M., Chothia, C., 1994. Volume changes on protein
folding. Structure 2, 641 – 649.
Hein, J., 1990. Unified approach to alignment and phylogenies. Methods Enzymol. 183, 626 – 645.
Hendrick, J.P., Hartl, F.U., 1993. Molecular chaperone functions of heat-shock proteins. Ann. Rev. Biochem. 62, 349 – 384.
Higgins, D.G., Bleasby, A.J., Fuchs, R., 1991. CLUSTAL V: improved software for multiple sequence alignment. CABIOS 5, 151 – 153. Horwitz, J., 1992. a-Cystallin can function as a molecular chaperone. Proc.
Nat. Acad. Sci. USA 89, 10449 – 10453.
Horwitz, J., 1993. Proctor lecture. The function of a-crystallin. Invest. Ophthalmol. Vis. Sci. 34, 10 – 22.
Horwitz, J., Huang, Q.L., Ding, L., Bova, M.P., 1998a. Lens a-crystallin: chaperone-like properties. Methods Enzymol. 290, 365 – 383. Horwitz, J., Bova, M., Huang, Q.L., Ding, L., Yaron, O., Lowman, S.,
1998b. Mutation of aB-crystallin: effects on chaperone-like activity. Int. J. Biol. Macromol. 22, 263 – 269.
Ingolia, T.D., Craig, E.A., 1982. Four small Drosophila heat-shock proteins are related to each other and to mammalian a-crystallin. Proc. Nat. Acad. Sci. USA 79, 2360 – 2364.
Iwaki, T., Kume-Iwaki, A., Liem, R.K., Goldman, J.E., 1989. aB-Cystallin is expressed in non-lenticular tissues and accumulates in Alexander’s disease brain. Cell 57, 71 – 78.
Jackson, M., Gentleman, S., Lennox, G., Ward, L., Gray, T., Randall, K., Morrell, K., Lowe, J., 1995. The cortical neuritic pathology of Huntington’s disease. Neuropathol. Appl. Neurobiol. 21, 18 – 26.
Johnson, J.D., El-Bayoumi, M.A., Weber, L.D., Tulinsky, A., 1979. Interaction of a-chymotrypsin with the fluorescent probe 1-anilino-naphthalene-8-sulfonate in solution. Biochemistry 18, 1292 – 1296. Klemenz, R., Frohli, E., Steiger, R.H., Schafer, R., Aoyama, A., 1991.
aB-Cystallin is a small heat-shock protein. Proc. Nat. Acad. Sci. USA 88, 3652 – 3656.
Kyte, J., Doolittle, R.F., 1982. A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 157, 105 – 132. Liang, J.J., Sun, T.X., Akhtar, N.J., 1999. Spectral contribution of the
individual tryptophan of aB-crystallin: a study by site-directed mutagenesis. Protein Sci. 8, 2761 – 2764.
Liao, J.H., Hung, C.C., Lee, J.S., Wu, S.H., Chiou, S.H., 1998. Characterization, cloning, and expression of porcine aB-crystallin. Biochem. Biophys. Res. Commun. 244, 131 – 137.
Liao, J.H., Lee, J.S., Chiou, S.H., 2002a. C-terminal lysine truncation increases thermostability and enhances chaperone-like function of porcine aB-crystallin. Biochem. Biophys. Res. Commun. 297, 309 – 316.
Liao, J.H., Lee, J.S., Chiou, S.H., 2002b. Distinct roles of aA- and aB-crystallins under thermal and UV stresses. Biochem. Biophys. Res. Commun. 295, 854 – 861.
Lindner, R.A., Kapur, A., Mariani, M., Titmuss, S.J., Carver, J.A., 1998. Structural alterations of a-crystallin during its chaperone action. Eur. J. Biochem. 258, 170 – 183.
Logsdon Jr, J.M., 1998. The recent origins of spliceosomal introns revisited. Curr. Opin. Genet. Dev. 8, 637 – 648.
Mattick, J.S., Gagen, M.J., 2001. The evolution of controlled multitasked gene networks: the role of introns and other non-coding RNAs in the development of complex organisms. Mol. Biol. Evol. 18, 1611 – 1630.
Merck, K.B., De Haard-Hoekman, W.A., Oude Essink, B.B., Bloemendal, H., De Jong, W.W., 1992. Expression and aggregation of recombinant aA-crystallin and its two domains. Biochim. Biophys. Acta 1130, 267 – 276.
Perng, M.D., Muchowski, P.J., van Den, I.P., Wu, G.J., Hutcheson, A.M., Clark, J.I., Quinlan, R.A., 1999. The cardiomyopathy and lens cataract mutation in aB-crystallin alters its protein structure, chaperone activity, and interaction with intermediate filaments in vitro. J. Biol. Chem. 274, 33235– 33243.
Perry, R.E., Abraham, E.C., 1986. High-performance liquid chromato-graphic separation of lens crystallins and their subunits. J. Chromatogr. 351, 103 – 110.
Plater, M.L., Goode, D., Crabbe, M.J., 1996. Effects of site-directed mutations on the chaperone-like activity of aB-crystallin. J. Biol. Chem. 271, 28558 – 28566.
Posner, M., Kantorow, M., Horwitz, J., 1999. Cloning, sequencing and differential expression of aB-crystallin in the zebrafish, Danio rerio. Biochim. Biophys. Acta 1447, 271 – 277.
Powers, D.A., 1989. Fish as model systems. Science 246, 352 – 358. Raman, B., Rao, C.M., 1994. Chaperone-like activity and quaternary
structure of a-crystallin. J. Biol. Chem. 269, 27264– 27268.
Rao, P.V., Horwitz, J., Zigler Jr, J.S., 1993. a-Cystallin, a molecular chaperone, forms a stable complex with carbonic anhydrase upon heat denaturation. Biochem. Biophys. Res. Commun. 190, 786 – 793. Renkawek, K., de Jong, W.W., Merck, K.B., Frenken, C.W., van Workum,
F.P., Bosman, G.J., 1992. aB-Cystallin is present in reactive glia in Creutzfeldt-Jakob disease. Acta Neuropathol. (Berl.) 83, 324 – 327. Renkawek, K., Voorter, C.E., Bosman, G.J., van Workum, F.P., de Jong,
W.W., 1994. Expression of aB-crystallin in Alzheimer’s disease. Acta Neuropathol. (Berl.) 87, 155 – 160.
Roest Crollius, H., Jaillon, O., Bernot, A., Dasilva, C., Bouneau, L., Fischer, C., Fizames, C., Wincker, P., Brottier, P., Quetier, F., Saurin, W., Weissenbach, J., 2000. Estimate of human gene number provided by genome-wide analysis using Tetraodon nigroviridis DNA sequence. Nat. Genet. 25, 235 – 238.
Rubin, G.M., Yandell, M.D., Wortman, J.R., Gabor Miklos, G.L., Nelson, C.R., Hariharan, I.K., Fortini, M.E., Li, P.W., Apweiler, R., Fleischmann, W., Cherry, J.M., Henikoff, S., Skupski, M.P., Misra, S., Ashburner, M., Birney, E., Boguski, M.S., Brody, T., Brokstein, P., Celniker, S.E., Chervitz, S.A., Coates, D., Cravchik, A., Gabrielian, A., Galle, R.F., Gelbart, W.M., George, R.A., Goldstein, L.S., Gong, F., Guan, P., Harris, N.L., Hay, B.A., Hoskins, R.A., Li, J., Li, Z., Hynes, R.O., Jones, S.J., Kuehl, P.M., Lemaitre, B., Littleton, J.T., Morrison, D.K., Mungall, C., O’Farrell, P.H., Pickeral, O.K., Shue, C., Vosshall, L.B., Zhang, J., Zhao, Q., Zheng, X.H., Lewis, S., 2000. Comparative genomics of the eukaryotes. Science 287, 2204 – 2215.
Sanger, F., 1949. Fractionation of oxidized insulin. Biochem. J. 44, 126 – 128.
Schuck, P., 2000. Size-distribution analysis of macromolecules by sedimentation velocity ultracentrifugation and lamm equation model-ing. Biophys. J. 78, 1606 – 1619.
Sreerama, N., Venyaminov, S.Y., Woody, R.W., 1999. Estimation of the number of a-helical and b-trand segments in proteins using circular dichroism spectroscopy. Protein Sci. 8, 370 – 380.
Sun, T.X., Liang, J.J., 1998. Intermolecular exchange and stabilization of recombinant human aA- and aB-crystallin. J. Biol. Chem. 273, 286 – 290.
van der Ouderaa, F.J., de Jong, W.W., Bloemendal, H., 1973. The amino-acid sequence of the aA2 chain of bovine a-crystallin. Eur. J. Biochem. 39, 207 – 222.
van der Ouderaa, F.J., De Jong, W.W., Hilderink, A., Bloemendal, H., 1974. The amino-acids sequence of the aB2 chain of bovine a-crystallin. Eur. J. Biochem. 49, 157 – 168.
Vicart, P., Caron, A., Guicheney, P., Li, Z., Prevost, M.C., Faure, A., Chateau, D., Chapon, F., Tome, F., Dupret, J.M., Paulin, D., Fardeau, M., 1998. A missense mutation in the aB-crystallin chaperone gene causes a desmin-related myopathy. Nat. Genet. 20, 92 – 95.
Wistow, G., 1985. Domain structure and evolution in a-crystallins and small heat-shock proteins. FEBS Lett. 181, 1 – 6.
Wistow, G.J., Piatigorsky, J., 1988. Lens crystallins: the evolution and expression of proteins for a highly specialized tissue. Ann. Rev. Biochem. 57, 479 – 504.
