Biochimie 94 (2012) 975e984
Contents lists available at SciVerse ScienceDirect
Biochimie journal homepage: www.elsevier.com/locate/biochi
Research paper
Structural and functional specificity of small heat shock protein HspB1 and HspB4, two cellular partners of HspB5: Role of the in vitro hetero-complex formation in chaperone activity Fériel Skouri-Panet a, Magalie Michiel b,1, Céline Férard a, Elodie Duprat a, Stéphanie Finet a, * a b
IMPMC UMR7590, CNRS, UPMC Paris 6, 4 place Jussieu, Paris, France Laboratoire Stress Oxydants et Cancer, CEA, DSV, iBiTec-S, Gif-sur-Yvette, France
a r t i c l e i n f o
a b s t r a c t
Article history: Received 2 August 2011 Accepted 20 December 2011 Available online 26 December 2011
The ubiquitous small heat shock proteins are essential elements in cellular protection, through a molecular chaperone activity. Among them, human small heat shock protein HspB1, HspB4 and HspB5 are involved in oncogenesis, anti-apoptotic activity and lens transparency. Therefore, these proteins are potential therapeutic targets in many diseases. Their general chaperone activity is related to their dynamic and multiple oligomeric structures, which are still poorly understood. The tissue selective distribution of HspB1 and HspB4, two cellular partners of HspB5, suggests that these two proteins might have evolved to play distinct physiological functions. Moreover, hetero-complex formation seems to be favoured in vivo, yet the functional specificity of the HspB1eHspB5 and HspB4eHspB5 hetero-complexes compared to the homo-oligomers remains unclear in the stress response pathway. A powerful approach combining biochemistry, biophysics and bioinformatics, allowed us to compare the different assemblies, with a special emphasis on the structural data, subunit exchange properties, activity and sequence evolution. We showed that they all exhibit different properties, from structural organization in physiological versus stress conditions, to chaperone-like activity, whatever the level of sequence conservation. Subunit exchange kinetics leading to HspB1eHspB5 or HspB4eHspB5 hetero-complex formation is also different between these two complexes: HspB5 exchanges more rapidly subunits with HspB1 than with HspB4. The relative sequence conservation in the sHSP superfamily does hide important structural heterogeneity and flexibility, which confer an enlarged range of different surface necessary to efficiently form complexes with various stress-induced cellular targets. Our data suggest that the formation of hetero-complexes could be an original evolutionary strategy to gain new cellular functions. Ó 2012 Elsevier Masson SAS. All rights reserved.
Keywords: Mammal small heat shock proteins HspB1eHspB4eHspB5 oligomers Chaperone activity Sequence analysis X-ray and light scattering
1. Introduction The small heat shock proteins (sHSP) belong to the ubiquitous heat shock protein class involved in cellular response to various stresses. They are widely found across species and are expressed in many cell types and tissues, at specific stages of the development and differentiation and/or induced by environmental stresses. Gene duplications early in vertebrate evolution contributed to the generation of 15 or more paralogous sHSP genes in vertebrates [1]. Ten mammalian sHSP are known [2] including HspB1 (Hsp27), HspB4 (aA-crystallin) and HspB5 (aB-crystallin). The human sHSP
* Corresponding author. Tel.: þ33 144279817; fax: þ33 144273785. E-mail address: stephanie.fi
[email protected] (S. Finet). 1 Present address: Université de Cergy, Laboratoire SATIE, Neuville sur Oise, Cergy Pontoise, France. 0300-9084/$ e see front matter Ó 2012 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.biochi.2011.12.018
members share a cytoprotective molecular chaperone function, as they are able to prevent thermal and chemical-induced aggregation of a variety of misfolded proteins by their incorporation into large soluble complexes. The sHSP are involved in important functional pathways in cells (cytoskeleton organization, apoptosis and protein turnover) and the modulation of the sHSP expression and/or activity is found associated to several pathologies: cancers, neurological diseases (neuropathies, Alzheimer or Parkinson diseases) and cataracts. They are thus becoming important pathological biomarkers [3e5] and potential therapeutic targets [6]. In particular, HspB1 is upregulated in various types of cancer [7e10] and promotes tumour survival through anti-apoptotic and pro-angiogenesis activities [11e15]. Several protein sequence modifications, such as mutations, ageing-associated or inappropriate post-translational modifications, have been identified to be responsible for the molecular
976
F. Skouri-Panet et al. / Biochimie 94 (2012) 975e984
chaperone dysfunction and thus leading to pathologies [16e21]. Among them a missense mutation in the HspB5 gene, R120G, responsible for a desmin-related myopathy and a cataract [21], causes desmin aggregation due to altered filament interactions [22]. The R120 mutation was shown to be associated with a significant ability to aggregate in cultured cells, correlated to an increased oligomeric size and a partial loss of the in vitro chaperone-like activity [23,24]. Functional sHSP form large, homo or hetero-oligomeric complexes, comprising 2 to 40 subunits (SU) of 12e43 kDa, monodisperse (constant SU number) or polydisperse (variable SU number). The sHSP superfamily, comprising more than 400 protein members (at least one in any living species) [25,26], is characterized by a conserved “alpha-crystallin domain” (ACD, 90 amino acids) with N-ter and C-ter extensions, both highly variable in length and amino acid composition. The structural information for full-length sequence of vertebrate sHSP is still lacking. The available 3D structures mainly correspond to partial assemblies, with truncated or unresolved N-ter and C-ter extensions, although these extensions might be involved in regulation of assembly, activity and substrate specificity [27e31]. The structures, from crystallography or NMR, indicated that the ACD formed a conserved b-sandwich topology, with up to 9 b-strands and a dimeric basic building block [28e37]. The different dimeric arrangements between metazoan or nonmetazoan species are formed through the switch of one side of the ACD sandwich from one sheet to the other, resulting in the formation of a deep groove, opened or closed, at the interface. The ACD side pockets, which are candidate sites for binding the nonnative protein targets [27,28,38,39] have been so far only observed in the open state, occupied or empty. However, these partial structures did not allow clarifying the involvement of the N-ter and C-ter extensions in the mammalian quaternary structure, the dynamic oligomerization and the chaperone activity. Neither heterocomplex nor sHSP/target 3D structure have been resolved yet. Furthermore the sHSP possess the original property to continuously exchange their SU while either keeping the same average SU number or changing their oligomeric state under stress [38e41]. In the eye lens, the functional entity is a hetero-complex HspB4eHspB5 (lens alpha-crystallins) with a 3:1 ratio in most vertebrate lens. This complex undergoes an irreversible structural transition at 60 C that involves an important SU reorganization accompanied by a doubling of size [42]. The hetero-complex has previously shown a higher stability in temperature and pressure compared to the human HspB5 homo-oligomers [43]. HspB4 and HspB5 are only co-expressed in the vertebrate eye lens whereas HspB1 and HspB5 are constitutively expressed and found colocalized in many tissues including skeletal muscles [44e46]. Both are also over-expressed under stress conditions. However it remains unclear whether in vivo hetero-complex formation is systematic or even enhanced [6] and if it confers any functional advantage. Mammalian two-hybrid system assays have shown interactions between HspB1, HspB4 and HspB5 [47]. HspB4 can form a hetero-complex in vitro with HspB5 at 37 C by extensive SU exchange [42] and the interactions between HspB1 and HspB5 enhance resistance and stability at high temperatures [48]. However several points have to be clarified: the role of multiple sHSP proteins, specifically expressed and localized (tissue and subcellular level) associated with the ability to form different heterocomplexes; the role of the hetero-complex formation and more generally conformational plasticity (compared to those of homocomplexes), which could be a way to resist to a wide range of deleterious environments; and the evolutionary conservation of these properties among mammals. The aim of this work was to compare two partners of the HspB5, i.e. HspB1 and HspB4, present in different cells (HspB4 especially in
the lens and HspB1 mainly in muscles) in order to better understand their different molecular interactions, subunit exchange properties and biological function. This study consisted in measuring the size and the chaperone-like activity of the different homo and hetero-complexes either close to their physiological conditions or under stress conditions (temperature and pressure). We used complementary experimental techniques such as small angle X-ray scattering (SAXS), light scattering (DLS and SEC-MALS) and chromatography, combined to an in silico analysis of the evolutionary conservation (or variability) of mammalian sequence properties. Altogether, our results allowed us to explore the original sHSP functioning and to provide new insights on hetero-assembly properties. 2. Material and methods 2.1. Protein purification
aN (bovine native a-crystallin hetero-complex) and bbLb (bovine bLowb-crystallin, the lowest chromatographic fraction of the bLowcrystallin (Low refers to low molecular weight)) are purified by size exclusion chromatography (Superdex S200PG column) from the cortical fractions of calf eye lens (given by a slaughter-house at Coutances, France) [42]. Recombinant human HspB5 and human or bovine HspB4 purification protocols are similar to the ones detailed previously [49,50]. In order to obtain wild type full-length sequence (without any tag) we deleted the 159 first N-ter nucleotides (corresponding to His Tag/S tag/enterokinase site configuration) from the human HspB1 sequence cloned in pet30b vector, kindly given by P. Arrigo (Université Claude Bernard Lyon1, FR). Shortly, all the plasmids were transformed into E. coli BL21 (DE3) strain and the same protocol has been used for expression and protein purification [23,24,43]. After growth and IPTG induction for 4 h at 37 C, cells were lysed by sonication in a 150 mM phosphate buffer (22 mM Na2HPO4, 28 mM KH2PO4, 70 mM KCl, 1.3 mM EDTA, 3 mM NaN3, 3 mM DTT) supplemented with antiproteases cocktail VII and benzonase nuclease (0.05 U/ml culture) (Calbiochem, Merck chemical ltd, Nottingham, UK). The soluble protein fraction was obtained by centrifugation. An additional step for nucleic acid precipitation by 0.08% Poly Ethylene Imine (PEI, Qiagen) was carried out. The last soluble fraction was dialysed against phosphate buffer and finally filtered, before loading on a preparative S300 sephacryl column (GE Healthcare, Vélizy, FR) for a size exclusion chromatography purification step. This single step was sufficient, as checked by SDS-PAGE and mass spectrometry, to obtain pure proteins in their native oligomeric structure. The concentrations of the individual stock solutions are measured with a Nanodrop spectrophotometer (wavelength at 280 nm), and an extinction coefficient equal to 1.51, 2.30, 1.77, 0.68, 0.69, 0.85 for CS, bLb, HspB1, HspB4, HspB5 and aN, respectively. 2.2. Hetero-complex formation HspB1eHspB4, HspB1eHspB5 and HspB4eHspB5 heterocomplexes were formed by mixing each protein, ratio 1:1 (w:w), and incubating at 37 C from 1 h to 20 h. The final concentrations for the assays were 0.5 mg/ml for DLS analysis, 1.2 mg/ml for chromatography and 2 mg/ml for SAXS. 2.3. Dynamic light scattering (DLS) For complex characterization and activity assays, DLS measurements were performed on a DynaPro instrument (Wyatt Technology Corp., Santa Barbara, CA) as described previously
F. Skouri-Panet et al. / Biochimie 94 (2012) 975e984
[24,51]. The laser light wavelength was 830 nm and the scattered light was collected at 90 by an optic fibre until a detector. All samples were filtered through 0.22 mm filters. Protein concentrations were adjusted in the 0.1e2 mg/ml range. Sample volumes varied from 20 to 80 ml, the correlation time s was 0.5 ms and the acquisition time was 10 s for a total measurement up to 1 h. The sample temperature was controlled with a Peltier cell from 4 C up to 80 C with a precision of 0.1 C. The scattered intensity (Is) was recorded for all the samples over a 1 h period for each temperature. The hydrodynamic radius (Rh) and the polydispersity (% P) were obtained using regularization methods (Dynamics software, version 6) [24,51].
2.4. Size exclusion chromatography and Multi angle light scattering (SEC-MALS) Molecular mass (MM) determination was performed by SECMALS experiments at room temperature. Size exclusion chromatography was carried out on an AKTA Purifier chromatography system (GE Healthcare, www.gelifesciences.com) with a silica KW 804 column (Shodex, Munich, Germany), connected to a three angle laser light scattering device and a refractive index detector (Treos MiniDawn Wyatt Technology, Santa Barbara, CA). Samples of 50 mL at around 1 mg/ml were pre-incubated for 1 h at the desired temperature (4 Ce60 C), cooled on ice for a few minutes (to block the subunit exchange process) and filtered at 4 C for 10 min on Nanosep MF GFP 0.2 mm centrifugal devices (Pall Life Science, Port Washington, NY) before loading at room temperature. The structural transition reversibility was previously checked by DLS, by heating the proteins 1 h at each temperature, before coming back to 20 C for 1 h (data not shown). 2.5. Anion exchange chromatography Subunit exchange between HspB5 and HspB1 or HspB5 and HspB4 (1:1 ratio, 1.2 mg/ml for each sHSP) was analyzed using HiTrap ANX FF columns 1 ml (GE Healthcare, http://www. gelifesciences.com) at 20 C. The equilibration buffer was 20 mM Tris pH 6.8 and the elution buffer was the same including 1 M NaCl. The elution rate was 1 ml/min, the injected volume was 100 ml and the NaCl gradient was fixed at 1.5% per ml. After mixing at 4 C where exchange does not take place, the samples were directly injected (20 C) or incubated at 37 C for one to 20 h, before loading on the anion exchange column.
2.6. Small angle X-ray scattering (SAXS) For complex characterization, data collection was carried out at the ID2 beamline at the European Radiation Synchrotron Facility (ESRF, Grenoble, France) [52]. The 2D detector was a fibre optically coupled FReLoN CCD based on Kodak image sensor, 4 4 binning mode, as a function of the scattering parameter q, where q ¼ 2ps ¼ (4p/l)sinq, s is the amplitude of the scattering vector, l is the wavelength of the X-rays (12.4 keV for the capillary experiments and 16.5 keV for the pressure experiments) and 2q the scattering angle. The sample to detector distance was set to 1.5 m or 2 m depending upon the experiment. The radius of gyration Rg was calculated using the Guinier approximation: I(q) ¼ I(0) exp(q2 R2g/ 3). The concentrations were from 1.1 to 10 mg/ml. The sample volume was 40 ml for the temperature experiments (in a 1.6 to 1.8 mm-diameter quartz capillary) and 80e100 ml in the pressure cell [43]. The temperature and pressure cells are thermostated with 0.1e0.5 C precision respectively.
977
2.7. In vitro chaperone-like assays Chaperone-like activity was assayed by DLS, by monitoring the thermal aggregation of the target protein (bbLb-crystallin at 60 C or Citrate synthase (Porcine heart CS, Sigma) at 48 C) alone and in the presence of the different sHSP [24,51]. The scattered intensity, Is(t,T), was recorded as a function of the time t (over a 1-h period) and the temperature T. The normalised scattered intensity difference, DIs(t,T), was used: DIs(t,T) ¼ [Is(target þ sHSP)(t,T) Is (sHSP)(t,T)]/Is(sHSP)(0,T). The final concentrations were 2.2 mg/ml for the CS, 0.5 mg/ml for the bbLb and 0.5 mg/ml for the sHSP, either homo or hetero-oligomers, in 150 mM phosphate buffer pH 6.8. 2.8. Sequence analysis The HspB1, HspB4 and HspB5 homologous protein sequences from mammalian species were retrieved from Uniprot (http:// www.uniprot.org) and NCBI protein sequence database (http:// www.ncbi.nlm.nih.gov/protein). We used an iterative similarity search approach, based on the HMMER suite [53]. It starts from the multiple alignment of the ACD sequences of the 10 human sHSP (HspB1-10) [2], from which a Hidden Markov Model (HMM) profile is built, calibrated and further used for the similarity search. At each step of this iterative approach, the new similar mammalian sequences are automatically split into regions (N-ter, ACD e which shares sequence similarity with the profile e, C-ter), and the ACD sequences are added to the initial multiple sequence alignment; a new HMM-profile is then build and calibrated. The procedure ends when no new significantly similar sequences are detected; the similarity significance is indicated by the E-value (expectation value), with a threshold set to 10e29 (a lower E-value being correlated to a higher significance). We build two sequence datasets. The first one comprises all of the complete sequences retrieved by this iterative procedure: this is the complete mammalian dataset. The second one is a subpart of the first one, considering the mammalian species for which HspB1, HspB4 and HspB5 complete sequences are together available. In order to analyze the sequence conservation within each orthologous group (HspB1, HspB4, HspB5) and between them, we have computed a pair-wise sequence identity percent (%ID) in two ways: (i) for a given group, each sequence is aligned and compared to the others. This intra-group approach gives the distribution of the %ID within each of the 3 orthologous groups; (ii) for a given pair of groups (for example HspB1/HspB4), each sequence of the first group is aligned and compared to each sequence of the second one. This inter-group approach gives the distribution of the %ID between the orthologous groups (HspB1/HspB4, HspB1/HspB5, HspB4/HspB5). These sequences share several group-related and large indels (Fig. S2), in particular for the N-ter and C-ter regions; therefore, we considered only amino acid sites with no gap to compute the %ID between two sequences. The statistical analysis (tests of normality and mean comparisons) of pI (isoelectric point) and length of the sequences was achieved with the R package (http://www.rproject.org). 3. Results 3.1. Preparation of recombinant sHSP The sHSP generally form large complexes of different sizes, shapes, polydispersity and exhibit a great sensitivity to their environment (i.e. temperature, pH, salt.). This point is particularly important in the context of the preparation of the samples: depending on the expression/purification protocol, the results could reveal large discrepancies corresponding to structural-
978
F. Skouri-Panet et al. / Biochimie 94 (2012) 975e984
induced modifications during the different steps. The emphasis was on finding a purification protocol able to produce the proteins without any tag, with no denaturation at any step and no extreme buffer composition since all these procedures were known to modify the oligomeric states of sHSP [48,54,55]. Our experimental procedure comprised a single purification step by gel filtration chromatography (Fig. 1A), and the expressed proteins were always kept in phosphate buffer pH 6.8. SDS-PAGE showed one single band with the expected apparent molecular mass. Protein identity was verified by mass spectrometry analysis (data not shown).
Table 1 HspB1, HspB4 and HspB5 assembly size as a function of temperature. The hydrodynamic radii (Rh in nm) and polydispersity percentage (polyD), determined by DLS, are indicated only for soluble non aggregated complexes (first line of each box). The molecular mass (MM in kDa) of the chromatography peak top and the deduced subunit (SU) number are indicated after 1 h at the specified temperature (second line). In hooks, the minimum/maximum SU number measured at half peak (given the population distribution or polydispersity). ND: not determined (HspB1 conformation being reversible); NC: not considered (sHsp assemblies not separated from soluble aggregates by SEC).
10 100 120 140 elution volume (ml)
1 0,8
1
20°C 37°C 48°C 55°C
0,6 0,4
0,5
0,2 0 6
D 10
7 8 9 elution volume (ml)
HspB4-B5 20°C 37°C, 1h 37°C, 20h
5 B5
B1+B5
1,5
HspB4
B4+B5
0 10
400
B4
10
200
B1
0 0
10 20 30 elution volume (ml)
40
0 0
10 20 30 elution volume (ml)
NaCl gradient (mM)
OD at 280nm (mAU)
160
HspB1-B5 20°C 37°C, 1h 37°C, 20h B5
within peaks (Fig. 1B). An asymmetrical peak (enlargement toward small elution volumes, like for HspB4 at 20 C) indicates a broad range of mass distribution and a heterogeneous sample. The evolution of HspB1, HspB4 and HspB5 size was followed from 20 up to 60 C (Table 1, Fig. 1B for HspB4 and see Ref. [24] for HspB5). The oligomer Rh and polydispersity were checked by DLS at all temperatures, like their reversibility upon return to ambient temperature. In previous works, HspB1, HspB4 and HSpB5 were reported to form oligomers of around 24 SU [54], 26 and 28 SU respectively [56] but the temperature at which the measurements
refractive index (AU)
20
20
60 C
6
OD at 280nm (AU)
30
C
HspB5 7.6 (30%) 618/31 [27e39] 7.3 (22%) 608/30 [28e36] 8.3 (22%) 725/36 [32e43] 9.8 (37%) 867/43 [37e53] 11.2 (23%) 1790/NC
molar mass (10 g/mol)
40
0 80
HspB4 8.0 (33%) 552/28 [20e43] 7.4 (18%) 468/24 [20e29] 8.0 (15%) 619/31 [27e36] 9.3 (15%) 914/46 [38e59] 11.6 (23%) 1920/NC
55 C
B
HspB1 HspB4 HspB5
50
HspB1 7.5 (31%) 445/20 [17e23] 8.4 (28%) 539/24 [19e29] 9.5 (26%) ND 10.4 (25%) ND 11.0 (27%) ND
48 C
The HspB1, HspB4 and HspB5 homo-oligomers were investigated as a function of temperature by DLS, SEC-MALS and SAXS. ID2 SAXS beamline allowed also the study of pressure effects on these assemblies. These techniques give complementary parameters related to the mean size (the DLS hydrodynamic radius Rh, the SECMALS molecular mass MM and the deduced subunit number SU, and the SAXS radius of gyration Rg) and the polydispersity of the sample (the standard deviation, a measure of the dispersion from the mean value). The DLS and the MALS devices allow the analysis of samples by separating soluble populations of different Rh and mass. With these techniques, the sHSP, forming soluble high molecular weight assemblies, are likely separated from soluble aggregates formed at high temperature. In such a case, for sake of clarity, only the major population Rh and MM are indicated in the Table 1. SEC-MALS elution profiles reveal the mass distribution
60
Sample Rh (polyD) MM/SU Rh (polyD) MM/SU Rh (polyD) MM/SU Rh (polyD) MM/SU Rh (polyD) MM/SU
37 C
3.2. Oligomerization states of the individual sHPS and their structural transitions under stress
A 70
T 20 C
0 40
Fig. 1. (A) sHSP purification by SEC experiments using a S300 sephacryl column at 20 C. This single step allowed separating the different sHSP from the large soluble aggregates (corresponding to the elution volume around 100 ml) and from the bacterial proteins (elution volume after 160 ml) (B) SEC-MALS experiments of HspB4 as a function of temperature: differential refractive index elution profiles (continuous lines) and corresponding molar masses (open symbols). Each sample was incubated for 1 h at the chosen temperature (black and B: 20 C; dark grey and ,: 37 C; grey and >: 48 C; light grey and Δ: 55 C) before injection on the column. (C, D) subunit exchange capacity of sHSP using anion exchange chromatography. Elution profiles (OD at 280 nm, continuous lines) with NaCl gradient (dotted line) of mixtures of HspB5 and HspB1 (B) or HspB4 (C), in ratio 1:1 (w:w). In the absence of incubation (20 C, black), HspB5 was not retained on the column, whereas HspB1 or HspB4 were eluted around 33 and 27 ml (330 and 280 mM NaCl) respectively. 1 h and 20 h incubation at 37 C are indicated in dark grey and light grey respectively.
F. Skouri-Panet et al. / Biochimie 94 (2012) 975e984
were done was not clearly mentioned. From our analysis, using soft purification protocol and precise temperature-dependent mass determination, we found 20, 28 and 31-mers at 20 C and 24, 24 and 30-mers at 37 C for HspB1, HspB4 and HspB5 respectively (Table 1). We showed that HspB1 exhibits a continuous increase of the oligomeric size and an important, although relatively stable, polydispersity with increasing temperature. At 48 C, HspB1 remained soluble at least for 15 min, then HspB1 was prone to aggregate (data not shown). So the Rh and polydispersity stability is then indicative only for the molecular species remaining in solution. During the initial phase, the structural transitions of the HspB1 major soluble assemblies were almost totally reversible when returning to room temperature (data not shown). Thus SECMALS experiments (performed at ambient temperature after incubation at the chosen temperature) allowed us to determine the HspB1 MM at 20 and 37 C only. HspB5, and even more HspB4, remain soluble at higher temperature (up to 60 C) than HspB1. From ambient to physiological temperature, they form smaller and less heterogeneous assemblies, corresponding to smaller Rh, polydispersity, MM and SU number. For HspB4, this structural modification is important (as evidenced by SEC-MALS on Fig. 1B since the corresponding peak becomes symmetrical and elutes latter). For these experiments, no reversibility was observed (checked by DLS) for HspB4 and HspB5 within the time needed for the measurement. They both form soluble high molecular weight assemblies with increasing temperature, and at 60 C, a highest proportion of soluble aggregates were measured by DLS for HspB1 (13%) compared to HspB5 (6%) and HspB4 (1%). In parallel, the HspB5 partners were studied by SAXS as a function of temperature and hydrostatic pressure (Fig. 2 and supplementary data Fig. S1). In Fig. 2A the sHSP radius of gyration is plotted as a function of temperature, from 15 to 60 C for HspB4 and HspB5, and 15e48 C for HspB1 (HspB1 becoming insoluble and unstable above). On SAXS curves (Fig. S1), an increase of size corresponds to an increase of Rg and of the intensity at the origin, and a displacement of the minima towards smaller q; an increase of polydispersity induces a loss of the minima. For HspB1, the Rg continuously increases from 20 to 48 C. The HspB4 exhibits a structural reorganization around 37 C corresponding to a smaller and less polydisperse oligomer (Fig. 2A and Fig. S1B). HspB5 is stable from 20 C to 37 C with a continuous increase from 43 C to 60 C (Fig. 2A and [43]). HspB4 is the most stable in temperature as HspB1 and HspB5 are prone to produce high molecular weight aggregates at moderately elevated temperatures (Fig. S1a). At the end of the cycle (48 C for HspB1 or 60 C for HspB4 and HspB5), the samples were returned to 15 C, to evaluate the reversibility of temperature-induced structural modifications (the reversibility was checked by DLS at all the temperatures). At 60 C, the SAXS experiments showed that the transitions were not reversible for
A
3.3. Subunit exchange and formation of hetero-complexes at physiological temperature Subunit exchange capacity of homo-oligomers and heterooligomers sHsps was studied using an anion exchange chromatography, DLS and SAXS. The anion exchange chromatography allowed characterizing the hetero-oligomer formation from individuals having a different charge [24]. The isoelectric point (pI) of HspB1, HspB4 and HspB5 are respectively 5.98, 5.77 and 6.8. The HspB1eHspB5 and HspB4eHspB5 oligomers were here analysed (Fig. 1C and D) but not the formation of HspB1eHspB4 hetero-complex, the pI of individuals being too close. The elution profiles showed that, in the absence of incubation at 37 C (20 C, black curves in Fig. 1C and D) in Tris buffer pH 6.8, HspB5 is not retained on the column whereas HspB1 and HspB4 are eluted around 33 and 27 ml (330 and 270 mM NaCl) respectively. After 1 h incubation at 37 C (dark grey curves), there is neither HspB5 nor HspB1 homo-complexes in solution, meaning that the SU exchange for the HspB1eHspB5 mixture (Fig. 1C) has occurred and is complete. In addition no evolution is observed after a longer time (see the 20 h-incubation light grey curve). On the contrary, in the case of the mixture HspB4eHspB5 (Fig. 1D), no SU exchange is evidenced after 1 h incubation at 37 C: the profile is similar to the initial one (20 C) and the completed exchange is reached after 20 h incubation at 37 C (light grey curve). The hetero-complexes are formed through this SU exchange mechanism with total disappearance of the individual
C 20°C 37°C
HspB1
10
Rg (nm)
HspB4 and HspB5, while HspB1 conformation at 48 C allowed almost a total reversibility when decreasing the temperature to 15 C (Rg close to the initial value). In Fig. 2C, the variation of HspB4 Rg with increasing hydrostatic pressure at 20 C shows a plateau until 200 MPa before increasing, similarly to HspB5 and aN [43]. HspB1, more sensitive to pressure, exhibits an oligomeric modification from 150 MPa (Fig. 2B). This means that lens proteins are stable over a wider range of pressure [43]. At 37 C, the HspB1 pressure evolution is small until 200 MPa, and it is more important between 200 and 300 MPa. This behaviour is rather different than for lens sHSP [43], as shown also for HspB4 (in Fig. 2C) where HspB4 undergo an important structural reorganisation between 100 and 200 MPa. For all the oligomers, the pressure elevation results in increased size and polydispersity. The pressure cycle is almost totally reversible i.e. as the scattering curves are recovered upon decompression essentially at physiological temperature. Our results demonstrate that irreversible temperature-induced and reversible pressure-induced association is a common mechanism for HspB1, HspB4 and HspB5, resulting in the formation of larger, more polydisperse and more dynamical assemblies under stress.
B
11
979
20°C 37°C
HspB4
9 8 7 HspB1 HspB4 HspB5
6 5 10
20
30
40
50
Temperature (°C)
60
0
50
100
150
200
Pressure (MPa)
250
300 0
50
100
150
200
250
300
Pressure (MPa)
Fig. 2. SAXS data: Evolution of the radius of gyration (Rg) as a function of temperature for HspB1, HspB4 and HspB5 (A) and as a function of pressure for HspB1 (B) and HspB4 (C).
980
F. Skouri-Panet et al. / Biochimie 94 (2012) 975e984
A
parents after only 1 h for HspB1. This is an unexpected result since the SU exchange rate between HspB1 and HspB5 is much faster (at least 20 times) than the rate to form HspB4eHspB5 complex in the same conditions. The formation of HspB1eHspB4 and HspB1eHspB5 heterocomplexes was characterized by DLS and SAXS after 1 h incubation at 37 C. The reconstituted HspB4eHspB5 mixture in different ratio and the native a-crystallin properties have been extensively described before [43,57,58], and the slow exchange SU rate of the HspB4eHspB5 mixture was not appropriate to study these oligomers by SAXS. The SAXS intensities of HspB1eHspB4 and HspB1eHspB5 are compared to the individual oligomers (Fig. 3). In the case of HspB1eHspB4, the DLS parameters, with a Rh of 8.5 nm, and the whole SAXS curves, are closer to HspB1 (Rh ¼ 8.4 nm) and could not fit to the average of the two isolated parent properties. The HspB1eHspB4 hetero-complex, ratio 1:1, keeps an overall conformation which is close to HspB1 while composed of 2 types of SU. In the case of HspB1eHspB5, the SAXS curve and DLS analysis indicate an intermediate size and shape (Rh of 7.7 nm) between the two individual parents, with Rh 7.3 and 8.4 nm, respectively. Altogether, our data reveal that exchange is total and induces the formation of a new, unique and stable hetero-assembly intermediate between the two homo-oligomers.
B
3.4. In vitro chaperone-like activity studied by DLS The chaperone-like activity of the sHSP complexes, i.e. their ability in suppressing the heat-induced aggregation of a target protein, was investigated by DLS (Fig. 4). It consists of following the scattered intensity of the different samples, heated as a function of time (up to 3600 s). An increase of intensity corresponds to the formation of larger complexes or aggregates. When the aggregates are too large, they could saturate the detector. This was observed for the target protein alone, quite immediately for the bbLb and after 2000 s for the CS. Chaperone assays were performed by adding the different sHSP as molecular chaperones to the targets, either bbLb or CS, before heating. First, as we have previously characterized aN [42], HspB4 and HspB5 (Table 1) at 60 C, we wanted to compare the ability of the homo-oligomers and their native hetero-oligomer aN to suppress the heat-induced aggregation of the eye lens target protein, the bbLb-crystallin (Fig. 4A). A differential chaperone-like activity is observed at 60 C: the native aN (ie HspB4eHspB5 hetero-complex ratio 3:1) is more efficient than HspB4, whereas HspB5 alone was a poor chaperone for bbLb-crystallin. This result is perfectly consistent with previous results performed for different ratios of HspB4 and HspB5 and different targets [57,58]. Second, the
A
Fig. 4. Chaperone-like activity of the sHsp samples (homo and hetero-oligomers) against the aggregation of (A) bbLb at 60 C and (B) CS at 48 C, followed by DLS. Differential efficiency of sHsp protection: HspB5 < HspB4 < aN for bbLb, and HspB1 < HspB1eB4 < HspB1eB5 < HspB4 < HspB5 for CS.
chaperone-like activity of HspB1, HspB4, HspB5, HspB1-HspB4 or HspB1-HspB5 was also tested against the CS aggregation at 48 C (Fig. 4B). Previously to the activity test, the hetero-oligomers size and stability of the HspB1eHspB4 and HspB1eHspB5 have been checked at 48 C. It was found that the HspB1eHspB4 and HspB1eHspB5 Rh at 48 C were 9.0 nm and 8.4 nm respectively. In addition, the DLS intensities are constant for both hetero-oligomers for more than 1 h (data not shown). As HspB1 alone is prone to aggregate at this temperature after 15e20 min, this means that both HspB4 and HspB5 are able to avoid the temperature aggregation of HspB1, which may result from a stabilization effect of
-1
B
10
HspB4 HspB1 HspB4+HspB1
-2
HspB5 HspB1 HspB5+HspB1 -2
10
I(q) (mm )
10 -1
-1
I(q) (mm )
-1
10
-3
10
-4
-4
10
10
-5
10
0.0
-3
10
-5
0.5
1.0
1.5 -1
q (nm )
2.0
10
0.0
0.5
1.0
1.5
2.0
-1
q (nm )
Fig. 3. Scattered intensity (I(q) as a function of scattering parameter q) of hetero-complexes: HspB1 þ HspB5 (A) and HspB1 þ HspB4 (B), ratio 1:1 (w:w), compared to the individual homo-oligomers HspB1, HspB4 and HspB5. All the proteins were incubated 1 h at 37 C before the measurement at 37 C.
F. Skouri-Panet et al. / Biochimie 94 (2012) 975e984
981
Fig. 5. Distribution of the isoelectric point (pI) of mammalian HspB1, HspB4 and HspB5 amino acid sequences. The pI values were computed for each protein region (N-ter and C-ter extensions, ACD domain) and for the full-length sequences. For a given distribution: the box indicates the second and third quartiles, and the median value (bold line); the lines show the largest/smallest observation that falls within a distance of 1.5 times the box size from the nearest box hinge. If any observations fall farther away, the additional points are considered “extreme” values and are shown separately. This dataset comprises the amino acid sequences of HspB1, HspB4 and HspB5 from 8 species; for details on this dataset, see the legend of the Figure S3.
HspB4 and HspB5 on HspB1. The individual HspB4 and HspB5 were good chaperones as both intensities were very stable during all the experiment. On the contrary, although HspB1 alone is able to remain soluble without aggregates (i.e. low and constant DLS scattered intensity) at least for 15 min at 48 C, the CSeHspB1 mixture intensity increases rapidly at 48 C, corresponding to the inability of HspB1 to avoid the CS aggregation. HspB1 is not a good chaperone for CS. In the case of hetero-oligomers, CS was partially rescued by HspB1eHspB4, corresponding to an averaged activity between the two parent oligomers. On the other hand, HspB1eHspB5 has good chaperone efficiency as the intensity was stable until 2000e2500 s, close to the individual HspB5. These results clearly show that, within the hetero-complexes formed with HspB5, HspB1 has gained in chaperone-like activity. 3.5. Sequence properties of mammalian HspB1, HspB4 and HspB5 Understanding structural and functional properties of sHSP relies also on the analysis of the sequences and their evolution. The HspB1, HspB4 and HspB5 complete protein sequences from mammalian species were retrieved by an iterative similarity search approach (described in the Material and Method section). These 70 sequences constitute the complete mammalian HspB1/B4/B5 dataset composed as follow: 10 HspB1 (i.e. the orthologous protein sequence from 10 mammalian species), 45 HspB4 and 15 HspB5. These three paralogous sequences are together available for 8 mammalian species. These results reflect the current state of sequence databases; our data thus do not prove that these three paralogous sequences are missing in some mammalian species. In this paper, we mainly describe and discuss the amino acid properties of these 24 sequences; we refer to those of the complete mammalian sequence dataset in order to assess that this small dataset is a representative sample of the mammalian sequences. We analyzed the evolutionary conservation or variability of our 24 sequence set (within and between the three orthologous sequence groups), with multiple sequence alignment of each protein region (N-ter, ACD, C-ter); we also identified the statistical specificity of their amino acid properties (pI, length). The distribution of these properties is presented in Fig. 5 (pI) and S3 (sequence length and identity). The intra-group analyses highlight the high conservation of sequence and properties of both HspB4 and HspB5 proteins. Indeed
the sequence identity percent within these two protein groups are 92.9% on average (without any sequence below 85%) and 95.6% on average (no sequence below 89%), respectively. Their pI and length are also conserved, considering the full-length proteins but also each of their regions (N-ter, ACD, C-ter). Moreover, these two groups of sequences tend to be widely conserved in mammalian species, as revealed by the analysis of the complete mammalian sequence dataset: their sequence identity is higher than 81% (HspB4) and 88% (HspB5), respectively. On the other hand, the HspB1 group shares more sequence variability. The identity between the 8 mammalian HspB1 sequences is higher than 75% and are 85.1% on average (higher than 70% for the complete mammalian sequence dataset); their pI and length are also more variable than those of HspB4 and HspB5 proteins, in particular for the extensions (N-ter length and C-ter pI). The inter-group analyses indicate the high conservation of sequence and properties between HspB4 and HspB5 mammalian proteins (57% identity on average for the 8 common species; 53.3% identity on average for the complete dataset); the HspB1 sequences are less similar (HspB1/HspB4: 41.4% and HspB1/HspB5: 46.2% identity on average), and several properties of HspB1 significantly differ from those of HspB4 and HspB5. In particular, the pI of the Nter of the HspB1 proteins is high (9.8 as median value for the 8 HspB1 N-ter sequences); the N-ter of HspB4 and HspB5 are about 5.5 and 6.4, respectively. The length of HspB1 proteins also significantly differs from that of HspB4 and HspB5 (the full-length HspB1 proteins are longer than 30 amino acids than HspB4 and HspB5); this is directly correlated with the respective length of their N-ter extension. The HspB1 sequences are more similar to HspB5 than to HspB4. However, HspB1 and HspB4 share no significant difference for the pI of their C-ter (from 5 to 6.3 and about 6.8, respectively) and full-length sequences; the pI of the C-ter of the HspB5 proteins is high (about 10.5 observed for the 8 sequences). No significant difference was reported between ACD sequence properties (length and pI) of mammalian HspB1, HspB4 and HspB5. These results are highly similar with those obtained with the complete mammalian sequence dataset (data not shown). 4. Discussion sHSP are a widespread and essential protein superfamily since they are found to be involved in many cellular pathways. As
982
F. Skouri-Panet et al. / Biochimie 94 (2012) 975e984
a consequence sHSP are very good candidates for therapeutic strategies but numerous publications concerning human HspB1, HspB4 or HspB5 and their role in cellular stress management have shown contradictory structural and functional properties. In this present study, full-length proteins were produced as close as possible to physiological conditions in order to purify functional sHSP oligomers. Their structure and function have been characterized in vitro under physiological controlled parameters or stressful conditions (temperature and pressure). All the sHSP studied form large soluble oligomers at physiological conditions, generally larger and more polydisperse (appearance of minor populations of soluble aggregates) with increasing stress, related to a greater SU exchange and a differential chaperone activity. This activity relies on the target and the type of assembly either homo or hetero-complexes. The chaperone activity of the different homooligomers is found correlated to their stability in temperature. HspB4 exhibits an important structural reorganization at physiological temperature and is less sensitive to aggregation at elevated temperature. HspB4 is a very good chaperone for the physiological lens targets (which is consistent with its lens-specific localization), as well as for the model CS target. HspB1 aggregates at lower temperature than the two lens sHSP and it is a poor chaperone for CS. HspB1 has the fastest temperature reversibility, due to a greater exchange capacity. The hetero-oligomers formed with HspB1 are different in size, shape and chaperone activity depending on the partner. Although a conformation close to HspB1, HspB1eHspB4 has an intermediate activity between the two parents. However, this association is not physiological as the two proteins were not found co-localized up to now. For HspB1eHspB5, its structural organization is intermediate between the two parents, but the activity is close to HspB5, which is the best chaperone for CS. For the aN, the exchange between the 2 SU also induces a hetero-oligomer formation with an original structural organisation and an optimised chaperone activity. In the lens, homo-oligomers are never found, only the hetero-complex is present. The formation of sHSP hetero-oligomers could be a general strategy for increasing the activity with a better reactivity at the cell level because it could allow the formation of new types of assembly with original surfaces leading to new potential functions. The resulting hetero-oligomer could have kept the advantages of each parent within a unique complex and results in fast and efficient molecular chaperone upon over-expression. These results illustrate all the possibilities in acquiring new properties through the structural organization and the SU exchange. When considering the sHSP amino acid sequences (length i.e. number of amino acids, charge), HspB5 is closer to HspB4 than to HspB1, whatever the mammalian species; this observation, totally compatible with our experimental results, also applies to the sequence extensions (N-ter and C-ter). Both the HspB4 and HspB5 sequence properties are conserved among mammals, whereas those of the HspB1 sequences (especially for their extensions) are variable. This evolutionary conservation of HspB4 and HspB5 is probably related to the evolutionary stability of the protein content of the lens (on the contrary to other tissues e.g. muscle). In spite of the fact that HspB4 and HspB5 sequences are closer, all the three studied sHSP are able to exchange their SU. The efficiency of the chaperone-like activity is supposed to be linked to this SU exchange capacity and the accessibility of the binding sites. This would imply some structural and functional plasticity, conferred by the N-ter and C-ter extensions: whereas the complete mammalian HspB1, HspB4 and HspB5 proteins have isoelectric point between 5.8 and 6.8, HspB1 N-ter and HspB5 C-ter are very basic, corresponding to probable disordered extensions, as high charge and low hydropathy are characteristic of disordered proteins [59]. In the same way, N-ter of mammalian HspB1 is much more important in
length that of HspB4 and HspB5 (whose number of amino acids for both full-length and extensions are similar). The recent dimeric ACD structures suggested that the ACD dimer is able to fluctuate between open- and closed- conformations, with a closed state being favoured in the oligomer. In addition, it is known that the N-ter and C-ter extensions are also essential to the stability of oligomers and the chaperone activity [27e31]. But the entire wild type structure is still a matter for discussion in the large oligomer formation and functioning. The chaperone activity, related to the SU exchange capacity, implies a structural plasticity. This plasticity is supposed to depend upon the accessibility of the extensions, which exhibit low structural features and high evolutionary variability. Disordered regions do not require strict sequence conservation and for the sHSP family this structural flexibility can provide a wide range of conformational states. The higher accessibility of binding sites could allow the binding of many differently shaped proteins that have been misfolded in distinct ways [60e62]. On the other hand, this structural variability, corresponding to a larger accessibility of the hydrophobic regions, could favour the sHSP insoluble aggregation itself. A balance has to be reached between the two competing processes, assembly formation with non-native protein targets to protect the cell and sHSP aggregation. This is what was probably observed for HspB1 with a fast SU exchange but a larger tendency to aggregate. This could reflect a specialization of the function of HspB1 within the various species of mammals. Our results provide new insights on specific properties of each of these sHSP that have specific cellular implications. Plasticity is observed at all scales: polydispersity, combined hetero-oligomer formation, high number of copies and under-functionalization of the extensions. As known for the immune system, a functional variability allows an optimum recognition of a maximum of nonnative protein targets. sHSP have a large spectra of targets. Targets to be rescued are stress related and dependent upon development stage, environment or species. Stress induced structural modifications are probably an adaptation process and heterocomplex formation a part of it. sHSP are known to exchange SU with other sHSP in order to form hetero-oligomers, which could improve their dynamical structure, their stability and their chaperone efficiency. The capacity to form the hetero-complexes could be a selective evolutionary advantage. Acknowledgements Authors wish to acknowledge T. Narayanan and M. Fernandez for the high support at the ID2 beamline at the ESRF, and Romy Chen-Min-Tao and Christophe Habib for their preliminary work on the sHSP sequences. This work was supported by the French Ministry of Research (M.M.), the ’University Pierre et Marie Curie’, UPMC (E.D., F.S.-P., S.F.), and the ’Centre National de la Recherche Scientifique’, CNRS (F.S.-P., S.F.). Appendix. Supplementary data Supplementary data related to this article can be found online at doi:10.1016/j.biochi.2011.12.018. References [1] E. Franck, O. Madsen, T. van Rheede, G. Ricard, M.A. Huynen, W.W. de Jong, Evolutionary diversity of vertebrate small heat shock proteins, J. Mol. Evol. 59 (2004) 792e805. [2] G. Kappé, E. Franck, P. Verschuure, W.C. Boelens, J.A.M. Leunissen, W.W. de Jong, The human genome encodes 10 alpha-crystallin-related small heat shock proteins: HspB1eB10, Cell Stress Chaperon. 8 (1) (2003) 53e61.
F. Skouri-Panet et al. / Biochimie 94 (2012) 975e984 [3] M.A. Fanelli, F.D.C. Carrion, J. Dekker, J. Schoemaker, D.R. Ciocca, Serological detection of heat shock protein hsp27 in normal and breast cancer patients, Cancer Epidemiol. Biomarkers Prev. 7 (1998) 791e795. [4] Z. Rui, J. Jian-Guo, T. Yuan-Peng, P. Hui, R. Bing-Gen, Use of serological proteomic methods to find biomarkers associated with breast cancer, Proteomics 3 (2003) 433e439. [5] Q. Huang, J. Ye, Q. Huang, W. Chen, L. Wang, W. Lin, J. Lin, X. Lin, Heat shock protein 27 is over-expressed in tumor tissues and increased in sera of patients with gastric adenocarcinoma, Clin. Chem. Lab. Med. 48 (2) (2010) 263e269. [6] A.-P. Arrigo, S. Simon, B. Gibert, C. Kretz-Remy, M. Nivon, A. Czekalla, D. Guillet, M. Moulin, C. Diaz-Latoud, P. Vicart, Hsp27 (HspB1) and alphaBcrystallin (HspB5) as therapeutic targets, FEBS Lett. 581 (19) (2007) 3665e3674. [7] L.M. Vargas-Roig, F.E. Gago, O. Tello, J.C. Anzar, D.R. Ciocca, Heat shock protein expression and drug resistance in breast cancer patients treated with induction chemotherapy, Int. J. Cancer 79 (1998) 468e475. [8] P.A. Cornford, A.R. Dodson, K.F. Parsons, A.D. Desmond, A. Woolfenden, M. Fordham, J.P. Neoptolemos, Y.Q. Ke, C.S. Foster, Heat shock protein expression independently predicts clinical outcome in prostate cancer, Cancer Res. 60 (2000) 7099e7105. [9] J.H. Lee, D. Sun, K.J. Cho, M.S. Kim, M.H. Hong, I.K. Kim, J.S. Lee, J.H. Lee, Overexpression of human 27 kDa heat shock protein in laryngeal cancer cells confers chemoresistance associated with cell growth delay, J. Cancer Res. Clin. Oncol. 133 (2007) 37e46. [10] A.A. Romani, P. Crafa, S. Desenzani, G. Graiani, C. Lagrasta, M. Sianesi, P. Soliani, A.F. Borghetti, The expression of hsp27 is associated with poor clinical outcome in intrahepatic cholangiocarcinoma, BMC Cancer 7 (2007) 232. [11] C. Garrido, A. Fromentin, B. Bonnotte, N. Favre, M. Moutet, A.P. Arrigo, P. Mehlen, E. Solary, Heat shock protein 27 enhances the tumorigenicity of immunogenic rat colon carcinoma cell clones, Cancer Res. 58 (1998) 5495e5499. [12] S.M. Keezer, S.E. Ivie, H.C. Krutzsch, A. Tandle, S.K. Libutti, D.D. Roberts, Angiogenesis inhibitors target the endothelial cell cytoskeleton through altered regulation of heat shock protein 27 and cofilin, Cancer Res. 63 (2003) 6405e6412. [13] H. Schepers, M. Geugien, M. van der Toorn, A.L. Bryantsev, H.H. Kampinga, B.J.L. Eggen, E. Vellenga, Hsp27 protects AML cells against VP-16-induced apoptosis through modulation of p38 and c-Jun, Exp. Hematol. 33 (2005) 660e670. [14] R. Wu, H. Kausar, P. Johnson, D.E. Montoya-Durango, M. Merchant, M.J. Rane, Hsp27 regulates Akt activation and polymorphonuclear leukocyte apoptosis by scaffolding MK2 to Akt signal complex, J. Biol. Chem. 282 (2007) 21598e21608. [15] M.T. Aloy, E. Hadchity, C. Bionda, C. Diaz-Latoud, L. Claude, R. Rousson, A.P. Arrigo, C. Rodriguez-Lafrasse, Protective role of Hsp27 protein against gamma radiation-induced apoptosis and radiosensitization effects of Hsp27 gene silencing in different human tumor cells, Int. J. Radiat. Oncol. Biol. Phys. 70 (2008) 543e553. [16] A. Tardieu, M. Michiel, F. Skouri-Panet, E. Duprat, S. Simon, S. Finet, From Sequence to Pathology: The Required Integrity of the Structure-DynamicsFunction Relationships at the Molecular Level, in: S. Simon, A.P. Arrigo (Eds.), Small Stress Proteins in Human Diseases, Nova Sciences Publishers Inc., New York, 2010, pp. 215e244. [17] O.V. Evgrafov, I. Mersiyanova, J. Irobi, L. Van Den Bosch, I. Dierick, C.L. Leung, O. Schagina, N. Verpoorten, K. Van Impe, V. Fedotov, E. Dadali, M. AuerGrumbach, C. Windpassinger, K. Wagner, Z. Mitrovic, D. Hilton-Jones, K. Talbot, J.-J. Martin, N. Vasserman, S. Tverskaya, A. Polyakov, R.K.H. Liem, J. Gettemans, W. Robberecht, P. De Jonghe, V. Timmerman, Mutant small heatshock protein 27 causes axonal Charcot-Marie-Tooth disease and distal hereditary motor neuropathy, Nat. Gen. 36 (6) (2004) 602e606. [18] R. Benndorf, M.J. Welsh, Shocking degeneration, Nat. Genet. 36 (2004) 547e548. [19] I. Dierick, J. Irobi, P. De Jonghe, V. Timmerman, Small heat shock proteins in inherited peripheral neuropathies, Ann. Med. 37 (2005) 413e422. [20] M. Litt, P. Kramer, D.M. LaMorticella, W. Murphey, E.W. Lovrien, R.G. Weleber, Autosomal dominant congenital cataract associated with a missense mutation in the human alpha crystallin gene CRYAA, Hum. Mol. Genet. 7 (3) (1998) 471e474. [21] P. Vicart, A. Caron, P. Guicheney, Z. Li, M.C. Prevost, A. Faure, D. Chateau, F. Chapon, F. Tome, J.M. Dupret, D. Paulin, M. Fardeau, A missense mutation in the aB-crystallin chaperone gene causes a desmin-related myopathy, Nat. Gent. 20 (1) (1998) 92e95. [22] M.D. Perng, S.F. Wen, I.P. van den, A.R. Prescott, R.A. Quinlan, Desmin aggregate formation by R120G alphaB-crystallin is caused by altered filament interactions and is dependent upon network status in cells, Mol. Biol. Cell. 15 (5) (2004) 2335e2346. [23] S. Simon, M. Michiel, F. Skouri-Panet, J.P. Lechaire, P. Vicart, A. Tardieu, Residue R120 is essential for the quaternary structure and functional integrity of human alphaB-crystallin, Biochemistry 46 (33) (2007) 9605. [24] M. Michiel, F. Skouri-Panet, E. Duprat, S. Simon, C. Férard, A. Tardieu, S. Finet, Abnormal assemblies and subunit exchange of alphaB-crystallin R120 mutants could be associated to destabilization of the dimeric substructure, Biochemistry 48 (2) (2009) 442e453. [25] T. Kriehuber, T. Rattei, T. Weinmaier, A. Bepperling, M. Haslbeck, J. Buchner, Independent evolution of the core domain and its flanking sequences in small heat shock proteins, FASEB J. 24 (10) (2010) 3633e3642.
983
[26] P. Poulain, J.C. Gelly, D. Flatters, Detection and architecture of small heat shock protein monomers, PloS One 5 (4) (2010) e9990. [27] J.G. Ghosh, M.R. Estrada, J.I. Clark, Interactive domains for chaperone activity in the small heat shock protein, human alphaB crystalline, Biochemistry 44 (45) (2005) 14854e14869. [28] S. Jehle, P. Rajagopal, B. Bardiaux, S. Markovic, R. Kuhne, J.R. Stout, V.A. Higman, R.E. Klevit, B.-J. van Rossum, H. Oschkinat, Solid-state NMR and SAXS studies provide a structural basis for the activation of alphaB-crystallin oligomers, Nat. Struct. Mol. Biol. 17 (99) (2010) 1037e1042. [29] K. Takeda, T. Hayashi, T. Abe, Y. Hirano, Y. Hanazono, A. Yohda, K. Miki, Dimer structure and conformational variability in the N-terminal region of an archaeal small heat shock protein, StHsp14.0, J. Struct. Biol. 174 (2011) 92e99. [30] A.R. Clark, C.E. Naylor, C. Bagnéris, N.H. Keep, C. Slingsby, Crystal structure of R120G disease mutant of human alphaB-crystallin domain dimer shows closure of a groove, J. Mol. Biol. 408 (1) (2011) 118e134. [31] E. Basha, K.L. Friedrich, E. Vierling, The N-terminal arm of small heat shock proteins is important for both chaperone activity and substrate specificity, J. Biol. Chem. 281 (52) (2006) 39943e39952. [32] K.K. Kim, R. Kim, S.H. Kim, Crystal structure of a small heat-shock protein, Nature 394 (6693) (1998) 595e599. [33] R.L. van Montfort, E. Basha, K.L. Friedrich, C. Slingsby, E. Vierling, Crystal structure and assembly of a eukaryotic small heat shock protein, Nat. Struct. Biol. 8 (12) (2001) 1025e1030. [34] R. Stamler, G. Kappe, W. Boelens, C. Slingsby, Wrapping the alpha-crystallin domain fold in a chaperone assembly, J. Mol. Biol. 353 (1) (2005) 68e79. [35] E. Hilario, F.J.M. Martin, M.C. Bertolini, L. Fan, Crystal structures of Xanthomonas small heat shock protein provide a structural basis for an active molecular chaperone oligomer, J. Mol. Biol. 408 (1) (2011) 74e86. [36] C. Bagnéris, O.A. Bateman, C.E. Naylor, N. Cronin, W.C. Boelens, N.H. Keep, Crystal structures of alpha-crystallin domain dimers of alphaB-crystallin and Hsp20, J. Mol. Biol. 392 (2010) 1242e1252. [37] A. Laganowsky, J.L.P. Benesch, M. Landau, L. Ding, M.R. Sawaya, D. Cascio, Q. Huang, C.V. Robinson, J. Horwitz, D. Eisenberg, Crystal structures of truncated alphaA and alphaB crystallins reveal structural mechanisms of polydispersity important for eye lens function, Prot. Sci. 5 (2010) 1031e1043. [38] S. Jehle, B. van Rossum, J.R. Stout, S.M. Noguchi, K. Falber, K. Rehbein, H. Oschkinat, R.E. Klevit, P. Rajagopal, AlphaB-crystallin: a hybrid solid-state/ solution-state NMR investigation reveals structural aspects of the heterogeneous oligomer, J. Mol. Biol. 385 (2009) 1481e1497. [39] K.K. Sharma, G.S. Kumar, A.S. Murphy, K. Kester, Identification of 1,1’-bi(4anilino)naphthalene-5,5’-disulfonic acid binding sequences in alpha-crystallin, J. Biol. Chem. 273 (25) (1998) 15474e15478. [40] M.P. Bova, L.L. Ding, J. Horwitz, B.K. Fung, Subunit exchange of alphaA-crystallin, J. Biol. Chem. 272 (47) (1997) 29511e29517. [41] M.P. Bova, H.S. McHaourab, Y. Han, B.K. Fung, Subunit exchange of small heat shock proteins. Analysis of oligomer formation of alphaA-crystallin and Hsp27 by fluorescence resonance energy transfer and site-directed truncations, J. Biol. Chem. 275 (2) (2000) 1035e1042. [42] T. Putilina, F. Skouri-Panet, K. Prat, N.H. Lubsen, A. Tardieu, Subunit exchange demonstrates a differential chaperone activity of calf alpha-crystallin toward beta LOW- and individual gamma-crystallins, J. Biol. Chem. 278 (16) (2003) 13747e13756. [43] F. Skouri-Panet, S. Quevillon-Cheruel, M. Michiel, A. Tardieu, S. Finet, sHSPs under temperature and pressure: the opposite behaviour of lens alphacrystallins and HSP26, Biochim. Biophys. Acta 1764 (2006) 372e383. [44] M. Gernold, U. Knauf, M. Gaestel, J. Stahl, P.-M. Kloetzel, Development and tissue-specific distribution of mouse small heat shock protein hsp25, Dev. Genet. 14 (1993) 103e111. [45] Y. Inaguma, S. Goto, H. Shinohara, K. Hasegawa, K. Ohshima, K. Kato, Physiological and pathological changes In levels of the two small stress proteins, HSP27 and aB crystallin, in rat hindlimb muscles, J. Biochem. 114 (1993) 378e384. [46] F.A.J.M. van de Klundert, M.L.J. Gijsen, P.R.L.A. van den IJssel, L.H.E.H. Snoeckx, W.W. de Jong, AlphaB-crystallin and hsp25 in neonatal cardiac cellsedifferences in cellular localization under stress conditions, Eur. J. Cell Biol. 75 (1998) 38e45. [47] L. Fu, J.J.-N. Liang, Detection of protein-protein interactions among lens crystallins in a mammalian two-hybrid system assay, J. Biol. Chem. 277 (6) (2002) 4256e4260. [48] L. Fu, J.J.-N. Liang, Enhanced stability of alphaB-crystallin in the presence of small heat shock protein Hsp27, Biochem. Biophys. Res. Commun. 302 (4) (2003) 710e714. [49] J. Horwitz, Alpha-crystallin, Review. Exp. Eye Res. 76 (2003) 145e153. [50] K.B. Merck, W.A. De Haard-Hoekman, B.B. Oude Essink, H. Bloemendal, W.W. De Jong, Expression and aggregation of recombinant aA-crystallin and its two domains, Biochim. Biophys. Acta 1130 (3) (1992) 267e276. [51] M. Michiel, F. Skouri-Panet, E. Duprat, A. Tardieu, K. Lampi, S. Finet, Aggregation of deamidated bB2-crystallin and incomplete rescue by a-chaperone, Exp. Eye Res. 90 (6) (2010) 688e698. [52] T. Narayanan, O. Diat, P. Bösecke, SAXS and USAXS on the high brilliance beamline at the ESRF, Nucl. Instrum. Methods in Phys. Res. Sec. A 467e468 (2) (2002) 1005e1009. [53] S.R. Eddy, Profile hidden Markov models, Bioinformatics 14 (1998) 755e763. [54] T. Rogalla, M. Ehrnsperger, X. Preville, A. Kotlyarovi, G. Lutsch, C. Ducasse, C. Paul, M. Wieske, P.-A. Arrigo, J. Buchner, M. Gaestel, Regulation of Hsp27
984
[55] [56] [57] [58]
F. Skouri-Panet et al. / Biochimie 94 (2012) 975e984 oligomerization, chaperone function, and protective activity against oxidative stress/tumor necrosis factor a by phosphorylation, J. Biol. Chem. 274 (27) (1999) 18947e18956. B. Lejl-Garolla, G. Mauk, Self-association of a small heat shock protein, J. Mol. Biol. 345 (2005) 631e642. J.L.P. Benesch, M. Ayoub, C.V. Robinson, J.A. Aquilina, J. Biol. Chem. 238 (42) (2008) 28513e28517. T.-X. Sun, J.J.-N. Liang, Intermolecular exchange and stabilization of recombinant human aA and aB-crystallin, J. Biol. Chem. 273 (1) (1998) 286e290. P.N.B.S. Srinivas, P.Y. Reddy, G.B. Reddy, Significance of a-crystallin heteropolymer with a 3:1 aA/aB ratio: chaperone-like activity, structure and hydrophobicity, Biochem. J. 414 (2008) 453e460.
[59] V.N. Uversky, J.R. Gillespie, A.L. Fink, Why are “natively unfolded” proteins unstructured under physiologic conditions? Proteins 41 (2000) 415e427. [60] M. Haslbeck, A. Ignatiou, H. Saibil, S. Helmich, E. Frenzl, T. Stromer, J. Buchner, A domain in the N-terminal part of Hsp26 is essential for chaperone function and oligomerization, J. Mol. Biol. 343 (2004) 445e455. [61] A.K. Dunker, C.J. Oldfiel, J. Meng, P. Romero, J.Y. Yang, J.W. Chen, V. Vacic, Z. Obradovic, V.N. Uversky, The unfoldomics decade: an update on intrinsically disordered proteins, BMC Genomics 9 (Suppl 2) (2008) S1. [62] N. Jaya, V. Garcia, E. Vierling, Substrate binding site flexibility of the small heat shock proteins molecular chaperones, Proc. Nat. Acad. Sci. USA 106 (2009) 15604e15609.