- Email: [email protected]
Overexpression, on-column refolding and isotopic labeling of Hahellin from Hahella chejuensis, a putative member of the βγ-crystallin superfamily
Available online at www.sciencedirect.com
Protein Expression and Purification 58 (2008) 269–274 www.elsevier.com/locate/yprep
Overexpression, on-column refolding and isotopic labeling of Hahellin from Hahella chejuensis, a putative member of the bc-crystallin superfamily Atul K. Srivastava a, Yogendra Sharma b,*, Kandala V.R. Chary a,* a
Department of Chemical Sciences, Tata Institute of Fundamental Research, Colaba, Mumbai 400005, India b Center for Cellular and Molecular Biology, Uppal Road, Hyderabad 500007, India Received 11 October 2007, and in revised form 14 December 2007 Available online 3 January 2008
Abstract A gene which encodes a hypothetical protein of 40 kDa has been identified in the genome of a marine bacterium Hahella chejuensis, as a putative member of bc-crystallin superfamily. This hypothetical protein contains a putative bc-crystallin-like domain, along with other domains for carbohydrate binding regions. It is named as Hahellin. A PCR amplified stretch of 92-amino acid residue long protein was cloned into pET21a vector and overexpressed in Escherichia coli strain BL21(DE3)pLysS cells. The recombinant Hahellin, produced as inclusion bodies, was estimated to be around 50% of the total cellular protein content which was solubilized in 8 M urea. The protein was purified and refolded using an anion exchange column. The MALDI-TOF mass spectrometry revealed the purity and monomeric nature of the protein. Further, a method to prepare isotopically (15N/13C) labeled protein with high yield for NMR studies is reported. The uniformly 15N-labeled Hahellin thus produced has been characterized by recording a sensitivity enhanced 2D [15N]–[1H] HSQC spectrum. The well, dispersed peaks in the spectrum confirm that the protein is indeed well folded and suitable for further studies by NMR. Ó 2007 Elsevier Inc. All rights reserved. Keywords: bc-Crystallin; Protein overexpression; Inclusion bodies; On-column refolding; Isotopic labeling; M9 minimal media; HSQC; MALDI-TOF
bc-Crystallin superfamily consists of diverse proteins from many taxa ranging from bacteria to vertebrates [1– 3]. Lens b- and c-crystallins are the archetypes of this superfamily, made of crystallin type Greek key motives [4]. One of the features of the superfamily is that there is no or very little sequence homology among the member proteins, and most of the members have been identified based on structural topology. untiil date, although the structures of a few members have been solved, the structural stabilities and other conformational features of various diverse bc domains have not yet been undertaken. We have been interested in identifying new members of this bc-crystallin superfamily among the genome sequences of * Corresponding authors. Fax: +91 22 2280 4610 (K.V.R. Chary); fax: +91 40 2716 0591 (Y. Sharma). E-mail addresses: [email protected] (Y. Sharma), chary@tifr. res.in (K.V.R. Chary).
1046-5928/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.pep.2007.12.010
various species which have been sequenced recently. In this paper, we describe expression and purification of one such putative bc-crystallin protein from Hahella chejuensis. Hahella chejuensis is a red-pigmented algicidal marine bacterium whose genome has been sequenced recently [5]. This proteobacterium has been isolated from marine sediment samples from Korea [6]. We analyzed the genome of the bacterium to identify proteins which might be responsible for their survival and growth under adverse marine habitat. The focus was on the proteins involved in various forms of stress such as those belonging to lens bc-crystallins since a member homolog Protein S of Myxococcus xanthus has been earlier thought to help in the protection of the bacteria under starved conditions [7]. In this endeavor, we found few polypeptide stretches showing signatures of lens bc-crystallins [8,4,9]. One such putative crystallin homolog is a 40 kDa protein with Accession No. YP_434263. Based on the sequence signature of
270
A.K. Srivastava et al. / Protein Expression and Purification 58 (2008) 269–274
252) protein domain for cloning and overexpression. We have named it as Hahellin (Hahella + crystallin) as a new microbial homolog of the superfamily. In this paper, we report overexpression of Hahellin in Escherichia coli without using any tag or fusing it with other proteins. We describe our attempts for an on-column refolding and purification of the protein from inclusion bodies using ion exchange chromatography [9,10]. We also describe the methods of 13C and 15N isotope labeling of the protein for NMR studies (Figs. 1–5).
Materials and methods Genomic DNA Genomic DNA of H. chejuensis KCTC 2396 was a kind gift from Prof. Jihyun F. Kim (1C Frontier Microbial Genomics and Applications Center, Korea Research Institute of Bioscience and Biotechnology, Republic of Korea) without which this work would not have been possible.
Fig. 1. PCR amplification of Hahellin gene from template DNA of Hahella chejuensis. The product DNA stretch of 273 bp is indicated by an arrow. One hundred base-pair ladder was used as DNA marker.
bc-crystallins, we identified the sequence of the 92-amino acid residue long (Mr = 10.2 kDa, residues number 162–
Cloning and overexpression Forward and reverse primers with Nde1 and BamH1 restriction sites, corresponding to 50 and 30 ends of transcripts, were used to amplify the desired Hahellin gene from genomic DNA of Hahella. The sequences of primers
Fig. 2. Construction of the recombinant plasmid pET21a-Hahellin for expression in E. coli. The insert (Hahellin) was amplified using template DNA from Hahella chejuensis, and inserted into BamH1 and Nde1 cloning sites of pET21a vector. Ap is for ampicillin resistance.
A.K. Srivastava et al. / Protein Expression and Purification 58 (2008) 269–274
Fig. 3. SDS–PAGE analysis of Hahellin shown on a 15% SDS gel: lane 1, molecular weight marker (Kaleidoscope Prestained Standards; Bio-Rad); lane 2, uninduced culture; lane 3, induced culture with 1 mM IPTG; lane 4, purified Hahellin after ion exchange and gel filtration column.
Fig. 4. Size exclusion chromatography was done using Superdex-75 16/60 prep grade column in 10 mM Tris–Cl, 1 mM EDTA, 300 mM NaCl, pH 6.5. The molecular mass standard used was: (1) bovine serum albumin (66 kDa), (2) carbonic anhydrase (29 kDa), (3) cytochrome C from horse heart (12.4 kDa) and (4) aprotinin (6.5 kDa). The flow rate was 0.5 ml/ minute. The single peak corresponds to single monomer.
271
designed based on the sequence of the Hahellin gene with the restriction sites Nde1 and BamH1 were as follows: Forward primer: 50 -CATATGGGAGAGAAAACCGT CAAACT-30 Reverse primer: 50 -GGATCCTCAATTCGTCGTCTCC ATTTTGG-30 PCR was performed according to the following thermocycle: initial denaturation at 94 °C for 2 min; 35 cycles of denaturation at 94 °C for 1 min; annealing at 50 °C for 1 min; and extension at 72 °C for 2 min. The amplified PCR product was gel eluted using Qiagen gel extraction kit (Qiagen) to remove unused primers and dNTPs. The purified PCR product was blunt ligated to a TOPO vector. The insert fragment was released by restriction digestion and subcloned into pET21a (Novagen). The construct was then transformed into expression host E. coli BL21 (DE3) cells. E. coli strain DH5a was used for cloning and maintaining the constructs. The recombinant expression plasmid was subsequently verified by DNA sequencing done by an agency MWGAG Biotech (Bangalore, India). Escherichia coli BL21(DE3) cells containing pET21aHahellin recombinant plasmid were inoculated from a single colony and grown overnight in Luria–Bertani (LB)1 media containing 100 lg/ml ampicillin at 37 °C by shaking at 225 rpm. After 12 h, 2.5 ml of culture was transferred into 250 ml of LB media with 100 lg/ml ampicillin. After growing the culture until A600 reached 0.6 in LB medium, the heterologous protein synthesis was induced by the addition of IPTG (isopropyl-b-D-thiogalactopyranoside) to a final concentration of 1 mM. After further incubation for 4 h at 37 °C, cells were pelleted by centrifugation at 6000 rpm for 20 min at 4 °C. The cell pellet (1.2 g) was suspended in 5 ml of a lysis buffer (50 mM Tris–Cl, 2 mM EDTA, 100 mM KCl, 1 mM MgCl2 containing 0.5 mM PMSF, pH 8.0). The cell suspension was extensively sonicated (20 1 min, Branson sonifier 450) on ice. The resulting cell lysate was centrifuged at 16,000 rpm for 20 min. The supernatant and the pellet were checked for the presence of the protein. The protein predominantly overexpressed in the form of inclusion bodies. Therefore, inclusion bodies were used for protein purification.
Purification and refolding of protein from inclusion bodies The procedure of on-column refolding was used to refold and purify Hahellin from the inclusion bodies. The inclusion bodies (frozen pellets) were thawed and suspended in 30 ml of a washing buffer (100 mM Tris–Cl, 5 mM EDTA, 4% Triton X-100, 2 M urea, pH 8.0) and washed 4–5 times with the same buffer until a clear supernatant was obtained. Then the inclusion bodies were washed with wash buffer lacking urea and Triton X-100 to remove the denaturant and detergent. The pI of the proFig. 5. MALDI-TOF mass spectrum of purified Hahellin acquired in linear mode. The prominent peak at 10.24 kDa corresponds to monomer species and a very small peak at 20.6 kDa corresponds to dimer.
1 Abbreviations used: LB, Luria–Bertani; DSS, 2,2-dimethyl-2-silapentane-5-sulfonates.
272
A.K. Srivastava et al. / Protein Expression and Purification 58 (2008) 269–274
tein was calculated from amino acid sequence data as 5.2 (Expasy ProtParam tool). The washed inclusion bodies were dissolved in 20 ml of buffer A (10 mM Tris–Cl, 1 mM EDTA, pH 9.5) containing 8 M urea, stirred continuously to solubilize the protein and incubated at 4 °C for 12 h to ensure complete denaturation. The suspension was centrifuged at 16,000 rpm for 30 min at 4 °C to remove insoluble particles to get a clear solution. A Q-Sepharose (Pharmacia) anion exchanger column (1.5 10 cm.) was equilibrated with 5 bed volumes of buffer A containing 8 M urea. The protein solubilized in urea was loaded onto the equilibrated column with flow rate 25 ml/hour. The column was washed with one bed volume of buffer A containing 8 M urea. After binding the protein the column was treated with a gradient of urea (8 to 0 M) in two bed volumes of buffer A to gradually remove the denaturant. The column was washed with two bed volumes of buffer A to ensure the complete removal of urea. The bound protein was eluted by multi-step elution, each step of one bed volume, with buffer containing 100 mM, 200 mM and then 300 mM NaCl. Flow rate was 25 ml/h. In such purification procedure the last step was size exclusion chromatography. Thus, we could get pure protein in its native state (i.e. in the absence of any denaturant), as established later by recording a 2D [15N-1H] HSQC. Concentration of the protein was estimated using an extinction coefficient value (at 280 nm) of 5960 M1 cm1, which was determined from its amino acid sequence using Expasy ProtParam tool (www.expasy.org). Size exclusion chromatography Gel filtration was performed using a 120 ml Superdex-75 16/60 (Pharmacia) column. Protein in buffer A containing 300 mM NaCl was loaded onto column equilibrated in the same buffer at room temperature. The volume of sample loaded was 2 ml and concentration was about 0.7 mM. Flow rate was 0.5 ml/min. For calculating the molecular mass the column was calibrated by molecular mass standards which were albumin from bovine serum albumin (66 kDa), carbonic anhydrase (29 kDa), cytochrome C from horse heart (12.4 kDa) and aprotinin (6.5 kDa) obtained from Sigma (Gel filtration molecular weight markers). Hahellin was found to be a monomer which is revealed from size exclusion chromatogram. Preparation of
15
N-labeled Hahellin
Uniformly 15N-labeled Hahellin was prepared by growing E. coli BL21 (DE3) cells transformed with the pET21a expression vector containing the Hahellin domain in 1 L of M9 minimal medium containing 15NH4Cl as the sole source of nitrogen. The M9 medium contained the following: M9 salt, 0.05 mM CaCl2, 2.0 mM MgSO4, 0.4% glucose, 0.04% yeast extract and 0.04% CAS amino acid. The other steps of overexpression and purification were the same as described above.
NMR sample preparation The sample for NMR spectroscopy was prepared by concentrating the protein solution in 10 mM Tris–Cl, 300 mM NaCl, and 20 mM CaCl2 in volume of 540 ll. The sample was concentrated using a 10 ml Amicon stirred cell with ultrafiltration membrane disk of 3 kDa cut-off. The final sample volume of 600 ll was made by adding 60 ll of 2H2O. The concentration of the protein was estimated to be 0.8 mM or 8.3 mg/ml. NMR experiment NMR experiment was carried out on a Varian INOVA 600 MHz NMR spectrometer equipped with a pulsed field gradient unit and triple resonance probe with an actively shielded z-gradient, operating at a 1H frequency of 600 MHz. Sensitivity enhanced 2D [15N–1H] HSQC at pH 6.7, temperature 25 °C, was recorded with 1H carrier placed on H2O resonance (4.72 ppm) and 15N carrier at 115 ppm. The sample was found to be very stable and did not change or degrade with time for more than a month. This was periodically checked by recording 2D [15N–1H] HSQC. 1H chemical shifts were calibrated relative to 2,2-dimethyl-2-silapentane-5-sulfonates (DSS) at 298 K (0.0 ppm), 15N chemical shifts were calibrated with respect to an external standard of 15NH4Cl (2.9 N in 1 M HCl). Spectra were processed using VNMR 6.1B software. Mass spectrometry by MALDI-TOF The protein solution (0.5 ll of 20 lM) was mixed with an equal volume of matrix (saturated solution of cyano-4-hydroxycinnamic acid in 35% acetonitrile and 65% TFA (0.1% (v/v) in water)), followed by application of the protein matrix mixture onto the MALDI target plate and drying at room temperature to form a co-crystal of protein and matrix. The dried sample was then analyzed by MALDI-TOF using a Micromass (UK) TofSpec 2E, fitted with a 337 nm laser in its linear mode. Results and discussion Identification of Hahellin Using the sequence signature of bc-crystallin domains, we identified a protein in the genome of Hahella chejuensis with Accession No. YP_434263. This protein is 450 residues long and has a domain (FA58C, Coagulation factor 5/8 C-terminal domain or discoidin domain), a RGD-like domain and a putative bc-crystallin-like domain. We selected a 92 residue long stretch forming a putative crystallin-like domain for cloning and overexpression.
A.K. Srivastava et al. / Protein Expression and Purification 58 (2008) 269–274
Identification of recombinant Hahellin plasmid The DNA stretch containing the Hahellin sequence was amplified by PCR using the genomic DNA of H. chejuensis. After gel elution of the DNA, it was cloned with the Nde1 and BamH1 restriction sites of a pET21a vector. The sequence of the recombinant plasmid was confirmed by DNA sequencing. Overexpression of protein in the form of inclusion bodies The transformed E. coli BL21(DE3) cells containing pET21a expression vector with the Hahellin gene on LB plates were first transferred to an M9 minimal medium agar plate and then grown in M9 minimal media culture. This enhanced the growth as compared to the inoculum directly from LB agar plate. The protein expression was induced by IPTG (isopropyl-b-D-thiogalactopyranoside). After analysis of the soluble and insoluble fractions of the E. coli cell extract, protein was found to be mostly expressed as insoluble inclusion bodies. Inclusion bodies, as it is well known, have both advantages and disadvantages [10]. The inclusion bodies are usually resistant to proteolytic degradation [11], often contain almost exclusively the overexpressed protein and therefore are good starting point for protein purification. Purification and refolding of protein The cell lysate was centrifuged after extensive sonication to get inclusion bodies. As the inclusion bodies are expected to contain some membrane proteins and enzymes, plasmid DNA and some non-protein molecules and some other molecules [12], which might induce aggregation in the desired protein [13], the pellet was washed first with a washing buffer (100 mM Tris–Cl, 5 mM EDTA, 4% Triton X-100, 2 M urea, pH 8.0). After several washings with the above buffer the protein was free of above impurities. Since the protein molecules in inclusion bodies are in an aggregated and physiologically inactive form, it was necessary to have correct folding of the protein for biophysical and structural characterization. For this purpose, the inclusion bodies were solubilized in buffer A (10 mM Tris–Cl, 1 mM EDTA, pH 9.5) containing 8 M urea, stirred continuously to solubilize protein and incubated at 4 °C for 12 h to completely denature the protein. The fact that we were not successful in refolding the denatured protein by rapid dilution and the dialysis methods in presence of stabilizing agents like arginine and glycerol (each in concentration of 0.1, 0.25, 0.5 and 0.75 M), etc. [14], we opted for the on-column refolding approach [15–17]. In this method, we used Q-Sepharose anion exchanger resin and loaded the solubilized inclusion bodies. The denatured protein molecules (based on its pI value 5.2) bind to the resin and fold properly when the denaturant concentration is gradually reduced to zero. Immobilization caused by the binding of the protein onto
273
the column prevents possible aggregation that would have resulted otherwise [18]. After the denaturant concentration was reduced to zero, the column was eluted with varying NaCl concentrations (100, 200 and 300 mM). This avoids the co-elution of other undesired proteins. Most of the target protein was eluted in the first bed volume in 300 mM NaCl. With this procedure, we could get the protein with some minor impurities of higher molecular weights. To remove most of the minor impurities, the fractions were collected, pooled and concentrated in an Amicon ultrafiltration cell for size exclusion chromatography using Superdex-75 (Pharmacia) column pre-equilibrated with buffer. The purity of the protein was checked on SDS–PAGE and found to be pure for biophysical studies (Fig. 3). The yield of the protein was estimated to be 16 mg per 1 L of LB media and 10 mg per 1 L of M9 minimal media containing 15NH4Cl as the sole source of nitrogen [19]. The wet cell weights were about 5 and 4 g per liter of LB and M9 minimal media, respectively. MALDI-TOF mass spectrometry MALDI-TOF mass spectrometry was used to determine the oligomerization state of the protein, if any. The mass spectrometric analysis showed two peaks in the spectrum; one at 10,251 m/z that corresponds to the monomeric state, and other at 20,556 m/z which corresponds to the dimer peak. The dimer peak is relatively very weak (Fig. 5). It confirmed that protein exists largely in monomeric state in solution under the experimental conditions. NMR characterization Hahellin consists of single polypeptide chain containing 92 amino acid residues. The amino acid sequence of the protein is shown below (Scheme 1). NMR is one of the techniques to determine whether a protein is folded or not. The unfolded protein molecule gives rise to a collapsed spectrum where several peaks come in a narrow chemical shift range [20]. We used HSQC to see if the refolded protein is properly folded after purification and is suitable for structural characterization. As seen in Fig. 6, the folded nature of the protein molecule is revealed by the well-dispersed [15N–1H] HSQC spectrum, both along the proton (1H) and nitrogen (15N) chemical shift axes. The protein is rich in serine residues (12/92; 13%) with only two proline residues. There are no cysteine or tryptophan residues. The N-terminal amino acid residues with exchangeable backbone amide proton (HN) and proline which do not have amide proton do not show any MGEKTVKLYE DTHFKGYSVE LPVGDYNLSS LISRGALNDD LSSARVPSGL RLEVFQHNNF KGVRDFYTSD AAELSRDNDA SSVRVSKMET TN
Scheme 1.
274
A.K. Srivastava et al. / Protein Expression and Purification 58 (2008) 269–274
chejuensis. This work was supported by a grant (Y.S) from the Department of Science and Technology (DST), Govt. of India. References
Fig. 6. Sensitivity enhanced 2D [15N–1H] HSQC of uniformly 15N-labeled Hahellin. The spectrum was acquired with 2048 data points in t2 and 64 in t1 dimension with spectral widths of 11 and 28 ppm along 1H and 15N dimensions, respectively. The number of transients collected for each t1 increment was 2 and the total experiment time was 5 min. The data were multiplied with shifted sine-square bell window function both along t1 and t2 axes, and zero-filled to 4096 along t2 dimension and to 256 along t1 dimension, prior to 2D Fourier transform.
15
N–1H correlation in the [15N–1H] HSQC spectrum. Therefore, one should expect to see at least 89 peaks or 15 N–1H correlations in the [15N–1H] HSQC spectrum recorded in a mixed solvent of 90% H2O and 10% 2H2O. Fig. 6 shows [15N–1H] HSQC spectrum recorded on a uniformly 15N-labeled protein. The dispersion of the peaks suggests the well folded structure of the protein, which accounts for around 87 peaks. We describe the refolding of a bc-crystallin domain of a novel member protein using an on-column refolding procedure. The protein obtained is refolded as seen with a welldispersed 2D [15N–1H] HSQC spectrum. Further structural characterization of Hahellin is underway. Acknowledgments The facilities provided by the National Facility for High Field NMR, supported by DST, DBT, CSIR, and Tata Institute of Fundamental Research, Mumbai, India, are greatly acknowledged. We thank Prof. Jihyun F. Kim, 21C Frontier Microbial Genomics and Applications, Center, Korea Research Institute of Bioscience and Biotechnology, Republic of Korea, for providing the DNA of Hahella
[1] N.H. Lubsen, H.J. Aarts, J.G. Schoenmakers, The evolution of lenticular proteins: the beta- and gamma-crystallin super gene family, Prog. Biophys. Mol. Biol. 51 (1988) 47–76. [2] R. Jaenicke, C. Slingsby, Lens crystallins and their microbial homologs: structure, stability, and function, Crit. Rev. Biochem. Mol. Biol. 36 (2001) 435–499. [3] H. Bloemendal, J.W. de, R. Jaenicke, N.H. Lubsen, C. Slingsby, A. Tardieu, Aging and vision: structure, stability and function of lens crystallins, Prog. Biophys. Mol. Biol. 86 (2004) 407–485. [4] G. Wistow, Evolution of a protein superfamily: relationships between vertebrate lens crystallins and microorganism dormancy proteins, J. Mol. Evol. 30 (1990) 140–145. [5] H. Jeong, J.H. Yim, C. Lee, S.H. Choi, Y.K. Park, S.H. Yoon, C.G. Hur, H.Y. Kang, D. Kim, H.H. Lee, K.H. Park, S.H. Park, H.S. Park, H.K. Lee, T.K. Oh, J.F. Kim, Genomic blueprint of Hahella chejuensis, a marine microbe producing an algicidal agent, Nucleic Acids Res. 33 (2005) 7066–7073. [6] H.K. Lee, J. Chun, E.Y. Moon, S.H. Ko, D.S. Lee, H.S. Lee, K.S. Bae, Hahella chejuensis gen. nov., sp. nov., an extracellular-polysaccharide-producing marine bacterium, Int. J. Syst. Evol. Microbiol. 51 (2001) 661–666. [7] G. Wistow, L. Summers, T. Blundell, Myxococcus xanthus spore coat protein S may have a similar structure to vertebrate lens beta gammacrystallins, Nature 315 (1985) 771–773. [8] G. Wistow, B. Turnell, L. Summers, C. Slingsby, D. Moss, L. Miller, P. Lindley, T. Blundell, X-ray analysis of the eye lens protein gammaII crystallin at 1.9 A resolution, J. Mol. Biol. 170 (1983) 175–202. [9] J. Graw, The crystallins: genes, proteins and diseases, Biol. Chem. 378 (1997) 1331–1348. [10] S.C. Makrides, Strategies for achieving high-level expression of genes in Escherichia coli, Microbiol. Rev. 60 (1996) 512–538. [11] T. Grune, T. Jung, K. Merker, K.J. Davies, Decreased proteolysis caused by protein aggregates, inclusion bodies, plaques, lipofuscin, ceroid, and ‘aggresomes’ during oxidative stress, aging, and disease, Int. J. Biochem. Cell Biol. 36 (2004) 2519–2530. [12] P. Valax, G. Georgiou, Molecular characterization of beta-lactamase inclusion bodies produced in Escherichia coli. 1. Composition, Biotechnol. Prog. 9 (1993) 539–547. [13] S.M. Singh, A.K. Panda, Solubilization and refolding of bacterial inclusion body proteins, J. Biosci. Bioeng. 99 (2005) 303–310. [14] A.P. Middelberg, Preparative protein refolding, Trends Biotechnol. 20 (2002) 437–443. [15] A. Jungbauer, W. Kaar, R. Schlegl, Folding and refolding of proteins in chromatographic beds, Curr. Opin. Biotechnol. 15 (2004) 487–494. [16] M. Li, Z.G. Su, J.C. Janson, In vitro protein refolding by chromatographic procedures, Protein Expr. Purif. 33 (2004) 1–10. [17] M.K. Jobby, Y. Sharma, Purification of a crystallin domain of Yersinia crystallin from inclusion bodies and its comparison to native protein from the soluble fraction, Biomed. Chromatogr. 20 (2006) 956–963. [18] J. Suttnar, J.E. Dyr, E. Hamsikova, J. Novak, V. Vonka, Procedure for refolding and purification of recombinant proteins from Escherichia coli inclusion bodies using a strong anion exchanger, J. Chromatogr. B Biomed. Appl. 656 (1994) 123–126. [19] J. Sambrook et al., Molecular Cloning: A Laboratory Manual A.1, second ed., Plainview, NY, 1989. [20] Y. Bai, J. Chung, H.J. Dyson, P.E. Wright, Structural and dynamic characterization of an unfolded state of poplar apo-plastocyanin formed under nondenaturing conditions, Protein Sci. 10 (2001) 1056– 1066.
Protein Expression and Purification 58 (2008) 269–274 www.elsevier.com/locate/yprep
Overexpression, on-column refolding and isotopic labeling of Hahellin from Hahella chejuensis, a putative member of the bc-crystallin superfamily Atul K. Srivastava a, Yogendra Sharma b,*, Kandala V.R. Chary a,* a
Department of Chemical Sciences, Tata Institute of Fundamental Research, Colaba, Mumbai 400005, India b Center for Cellular and Molecular Biology, Uppal Road, Hyderabad 500007, India Received 11 October 2007, and in revised form 14 December 2007 Available online 3 January 2008
Abstract A gene which encodes a hypothetical protein of 40 kDa has been identified in the genome of a marine bacterium Hahella chejuensis, as a putative member of bc-crystallin superfamily. This hypothetical protein contains a putative bc-crystallin-like domain, along with other domains for carbohydrate binding regions. It is named as Hahellin. A PCR amplified stretch of 92-amino acid residue long protein was cloned into pET21a vector and overexpressed in Escherichia coli strain BL21(DE3)pLysS cells. The recombinant Hahellin, produced as inclusion bodies, was estimated to be around 50% of the total cellular protein content which was solubilized in 8 M urea. The protein was purified and refolded using an anion exchange column. The MALDI-TOF mass spectrometry revealed the purity and monomeric nature of the protein. Further, a method to prepare isotopically (15N/13C) labeled protein with high yield for NMR studies is reported. The uniformly 15N-labeled Hahellin thus produced has been characterized by recording a sensitivity enhanced 2D [15N]–[1H] HSQC spectrum. The well, dispersed peaks in the spectrum confirm that the protein is indeed well folded and suitable for further studies by NMR. Ó 2007 Elsevier Inc. All rights reserved. Keywords: bc-Crystallin; Protein overexpression; Inclusion bodies; On-column refolding; Isotopic labeling; M9 minimal media; HSQC; MALDI-TOF
bc-Crystallin superfamily consists of diverse proteins from many taxa ranging from bacteria to vertebrates [1– 3]. Lens b- and c-crystallins are the archetypes of this superfamily, made of crystallin type Greek key motives [4]. One of the features of the superfamily is that there is no or very little sequence homology among the member proteins, and most of the members have been identified based on structural topology. untiil date, although the structures of a few members have been solved, the structural stabilities and other conformational features of various diverse bc domains have not yet been undertaken. We have been interested in identifying new members of this bc-crystallin superfamily among the genome sequences of * Corresponding authors. Fax: +91 22 2280 4610 (K.V.R. Chary); fax: +91 40 2716 0591 (Y. Sharma). E-mail addresses: [email protected] (Y. Sharma), chary@tifr. res.in (K.V.R. Chary).
1046-5928/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.pep.2007.12.010
various species which have been sequenced recently. In this paper, we describe expression and purification of one such putative bc-crystallin protein from Hahella chejuensis. Hahella chejuensis is a red-pigmented algicidal marine bacterium whose genome has been sequenced recently [5]. This proteobacterium has been isolated from marine sediment samples from Korea [6]. We analyzed the genome of the bacterium to identify proteins which might be responsible for their survival and growth under adverse marine habitat. The focus was on the proteins involved in various forms of stress such as those belonging to lens bc-crystallins since a member homolog Protein S of Myxococcus xanthus has been earlier thought to help in the protection of the bacteria under starved conditions [7]. In this endeavor, we found few polypeptide stretches showing signatures of lens bc-crystallins [8,4,9]. One such putative crystallin homolog is a 40 kDa protein with Accession No. YP_434263. Based on the sequence signature of
270
A.K. Srivastava et al. / Protein Expression and Purification 58 (2008) 269–274
252) protein domain for cloning and overexpression. We have named it as Hahellin (Hahella + crystallin) as a new microbial homolog of the superfamily. In this paper, we report overexpression of Hahellin in Escherichia coli without using any tag or fusing it with other proteins. We describe our attempts for an on-column refolding and purification of the protein from inclusion bodies using ion exchange chromatography [9,10]. We also describe the methods of 13C and 15N isotope labeling of the protein for NMR studies (Figs. 1–5).
Materials and methods Genomic DNA Genomic DNA of H. chejuensis KCTC 2396 was a kind gift from Prof. Jihyun F. Kim (1C Frontier Microbial Genomics and Applications Center, Korea Research Institute of Bioscience and Biotechnology, Republic of Korea) without which this work would not have been possible.
Fig. 1. PCR amplification of Hahellin gene from template DNA of Hahella chejuensis. The product DNA stretch of 273 bp is indicated by an arrow. One hundred base-pair ladder was used as DNA marker.
bc-crystallins, we identified the sequence of the 92-amino acid residue long (Mr = 10.2 kDa, residues number 162–
Cloning and overexpression Forward and reverse primers with Nde1 and BamH1 restriction sites, corresponding to 50 and 30 ends of transcripts, were used to amplify the desired Hahellin gene from genomic DNA of Hahella. The sequences of primers
Fig. 2. Construction of the recombinant plasmid pET21a-Hahellin for expression in E. coli. The insert (Hahellin) was amplified using template DNA from Hahella chejuensis, and inserted into BamH1 and Nde1 cloning sites of pET21a vector. Ap is for ampicillin resistance.
A.K. Srivastava et al. / Protein Expression and Purification 58 (2008) 269–274
Fig. 3. SDS–PAGE analysis of Hahellin shown on a 15% SDS gel: lane 1, molecular weight marker (Kaleidoscope Prestained Standards; Bio-Rad); lane 2, uninduced culture; lane 3, induced culture with 1 mM IPTG; lane 4, purified Hahellin after ion exchange and gel filtration column.
Fig. 4. Size exclusion chromatography was done using Superdex-75 16/60 prep grade column in 10 mM Tris–Cl, 1 mM EDTA, 300 mM NaCl, pH 6.5. The molecular mass standard used was: (1) bovine serum albumin (66 kDa), (2) carbonic anhydrase (29 kDa), (3) cytochrome C from horse heart (12.4 kDa) and (4) aprotinin (6.5 kDa). The flow rate was 0.5 ml/ minute. The single peak corresponds to single monomer.
271
designed based on the sequence of the Hahellin gene with the restriction sites Nde1 and BamH1 were as follows: Forward primer: 50 -CATATGGGAGAGAAAACCGT CAAACT-30 Reverse primer: 50 -GGATCCTCAATTCGTCGTCTCC ATTTTGG-30 PCR was performed according to the following thermocycle: initial denaturation at 94 °C for 2 min; 35 cycles of denaturation at 94 °C for 1 min; annealing at 50 °C for 1 min; and extension at 72 °C for 2 min. The amplified PCR product was gel eluted using Qiagen gel extraction kit (Qiagen) to remove unused primers and dNTPs. The purified PCR product was blunt ligated to a TOPO vector. The insert fragment was released by restriction digestion and subcloned into pET21a (Novagen). The construct was then transformed into expression host E. coli BL21 (DE3) cells. E. coli strain DH5a was used for cloning and maintaining the constructs. The recombinant expression plasmid was subsequently verified by DNA sequencing done by an agency MWGAG Biotech (Bangalore, India). Escherichia coli BL21(DE3) cells containing pET21aHahellin recombinant plasmid were inoculated from a single colony and grown overnight in Luria–Bertani (LB)1 media containing 100 lg/ml ampicillin at 37 °C by shaking at 225 rpm. After 12 h, 2.5 ml of culture was transferred into 250 ml of LB media with 100 lg/ml ampicillin. After growing the culture until A600 reached 0.6 in LB medium, the heterologous protein synthesis was induced by the addition of IPTG (isopropyl-b-D-thiogalactopyranoside) to a final concentration of 1 mM. After further incubation for 4 h at 37 °C, cells were pelleted by centrifugation at 6000 rpm for 20 min at 4 °C. The cell pellet (1.2 g) was suspended in 5 ml of a lysis buffer (50 mM Tris–Cl, 2 mM EDTA, 100 mM KCl, 1 mM MgCl2 containing 0.5 mM PMSF, pH 8.0). The cell suspension was extensively sonicated (20 1 min, Branson sonifier 450) on ice. The resulting cell lysate was centrifuged at 16,000 rpm for 20 min. The supernatant and the pellet were checked for the presence of the protein. The protein predominantly overexpressed in the form of inclusion bodies. Therefore, inclusion bodies were used for protein purification.
Purification and refolding of protein from inclusion bodies The procedure of on-column refolding was used to refold and purify Hahellin from the inclusion bodies. The inclusion bodies (frozen pellets) were thawed and suspended in 30 ml of a washing buffer (100 mM Tris–Cl, 5 mM EDTA, 4% Triton X-100, 2 M urea, pH 8.0) and washed 4–5 times with the same buffer until a clear supernatant was obtained. Then the inclusion bodies were washed with wash buffer lacking urea and Triton X-100 to remove the denaturant and detergent. The pI of the proFig. 5. MALDI-TOF mass spectrum of purified Hahellin acquired in linear mode. The prominent peak at 10.24 kDa corresponds to monomer species and a very small peak at 20.6 kDa corresponds to dimer.
1 Abbreviations used: LB, Luria–Bertani; DSS, 2,2-dimethyl-2-silapentane-5-sulfonates.
272
A.K. Srivastava et al. / Protein Expression and Purification 58 (2008) 269–274
tein was calculated from amino acid sequence data as 5.2 (Expasy ProtParam tool). The washed inclusion bodies were dissolved in 20 ml of buffer A (10 mM Tris–Cl, 1 mM EDTA, pH 9.5) containing 8 M urea, stirred continuously to solubilize the protein and incubated at 4 °C for 12 h to ensure complete denaturation. The suspension was centrifuged at 16,000 rpm for 30 min at 4 °C to remove insoluble particles to get a clear solution. A Q-Sepharose (Pharmacia) anion exchanger column (1.5 10 cm.) was equilibrated with 5 bed volumes of buffer A containing 8 M urea. The protein solubilized in urea was loaded onto the equilibrated column with flow rate 25 ml/hour. The column was washed with one bed volume of buffer A containing 8 M urea. After binding the protein the column was treated with a gradient of urea (8 to 0 M) in two bed volumes of buffer A to gradually remove the denaturant. The column was washed with two bed volumes of buffer A to ensure the complete removal of urea. The bound protein was eluted by multi-step elution, each step of one bed volume, with buffer containing 100 mM, 200 mM and then 300 mM NaCl. Flow rate was 25 ml/h. In such purification procedure the last step was size exclusion chromatography. Thus, we could get pure protein in its native state (i.e. in the absence of any denaturant), as established later by recording a 2D [15N-1H] HSQC. Concentration of the protein was estimated using an extinction coefficient value (at 280 nm) of 5960 M1 cm1, which was determined from its amino acid sequence using Expasy ProtParam tool (www.expasy.org). Size exclusion chromatography Gel filtration was performed using a 120 ml Superdex-75 16/60 (Pharmacia) column. Protein in buffer A containing 300 mM NaCl was loaded onto column equilibrated in the same buffer at room temperature. The volume of sample loaded was 2 ml and concentration was about 0.7 mM. Flow rate was 0.5 ml/min. For calculating the molecular mass the column was calibrated by molecular mass standards which were albumin from bovine serum albumin (66 kDa), carbonic anhydrase (29 kDa), cytochrome C from horse heart (12.4 kDa) and aprotinin (6.5 kDa) obtained from Sigma (Gel filtration molecular weight markers). Hahellin was found to be a monomer which is revealed from size exclusion chromatogram. Preparation of
15
N-labeled Hahellin
Uniformly 15N-labeled Hahellin was prepared by growing E. coli BL21 (DE3) cells transformed with the pET21a expression vector containing the Hahellin domain in 1 L of M9 minimal medium containing 15NH4Cl as the sole source of nitrogen. The M9 medium contained the following: M9 salt, 0.05 mM CaCl2, 2.0 mM MgSO4, 0.4% glucose, 0.04% yeast extract and 0.04% CAS amino acid. The other steps of overexpression and purification were the same as described above.
NMR sample preparation The sample for NMR spectroscopy was prepared by concentrating the protein solution in 10 mM Tris–Cl, 300 mM NaCl, and 20 mM CaCl2 in volume of 540 ll. The sample was concentrated using a 10 ml Amicon stirred cell with ultrafiltration membrane disk of 3 kDa cut-off. The final sample volume of 600 ll was made by adding 60 ll of 2H2O. The concentration of the protein was estimated to be 0.8 mM or 8.3 mg/ml. NMR experiment NMR experiment was carried out on a Varian INOVA 600 MHz NMR spectrometer equipped with a pulsed field gradient unit and triple resonance probe with an actively shielded z-gradient, operating at a 1H frequency of 600 MHz. Sensitivity enhanced 2D [15N–1H] HSQC at pH 6.7, temperature 25 °C, was recorded with 1H carrier placed on H2O resonance (4.72 ppm) and 15N carrier at 115 ppm. The sample was found to be very stable and did not change or degrade with time for more than a month. This was periodically checked by recording 2D [15N–1H] HSQC. 1H chemical shifts were calibrated relative to 2,2-dimethyl-2-silapentane-5-sulfonates (DSS) at 298 K (0.0 ppm), 15N chemical shifts were calibrated with respect to an external standard of 15NH4Cl (2.9 N in 1 M HCl). Spectra were processed using VNMR 6.1B software. Mass spectrometry by MALDI-TOF The protein solution (0.5 ll of 20 lM) was mixed with an equal volume of matrix (saturated solution of cyano-4-hydroxycinnamic acid in 35% acetonitrile and 65% TFA (0.1% (v/v) in water)), followed by application of the protein matrix mixture onto the MALDI target plate and drying at room temperature to form a co-crystal of protein and matrix. The dried sample was then analyzed by MALDI-TOF using a Micromass (UK) TofSpec 2E, fitted with a 337 nm laser in its linear mode. Results and discussion Identification of Hahellin Using the sequence signature of bc-crystallin domains, we identified a protein in the genome of Hahella chejuensis with Accession No. YP_434263. This protein is 450 residues long and has a domain (FA58C, Coagulation factor 5/8 C-terminal domain or discoidin domain), a RGD-like domain and a putative bc-crystallin-like domain. We selected a 92 residue long stretch forming a putative crystallin-like domain for cloning and overexpression.
A.K. Srivastava et al. / Protein Expression and Purification 58 (2008) 269–274
Identification of recombinant Hahellin plasmid The DNA stretch containing the Hahellin sequence was amplified by PCR using the genomic DNA of H. chejuensis. After gel elution of the DNA, it was cloned with the Nde1 and BamH1 restriction sites of a pET21a vector. The sequence of the recombinant plasmid was confirmed by DNA sequencing. Overexpression of protein in the form of inclusion bodies The transformed E. coli BL21(DE3) cells containing pET21a expression vector with the Hahellin gene on LB plates were first transferred to an M9 minimal medium agar plate and then grown in M9 minimal media culture. This enhanced the growth as compared to the inoculum directly from LB agar plate. The protein expression was induced by IPTG (isopropyl-b-D-thiogalactopyranoside). After analysis of the soluble and insoluble fractions of the E. coli cell extract, protein was found to be mostly expressed as insoluble inclusion bodies. Inclusion bodies, as it is well known, have both advantages and disadvantages [10]. The inclusion bodies are usually resistant to proteolytic degradation [11], often contain almost exclusively the overexpressed protein and therefore are good starting point for protein purification. Purification and refolding of protein The cell lysate was centrifuged after extensive sonication to get inclusion bodies. As the inclusion bodies are expected to contain some membrane proteins and enzymes, plasmid DNA and some non-protein molecules and some other molecules [12], which might induce aggregation in the desired protein [13], the pellet was washed first with a washing buffer (100 mM Tris–Cl, 5 mM EDTA, 4% Triton X-100, 2 M urea, pH 8.0). After several washings with the above buffer the protein was free of above impurities. Since the protein molecules in inclusion bodies are in an aggregated and physiologically inactive form, it was necessary to have correct folding of the protein for biophysical and structural characterization. For this purpose, the inclusion bodies were solubilized in buffer A (10 mM Tris–Cl, 1 mM EDTA, pH 9.5) containing 8 M urea, stirred continuously to solubilize protein and incubated at 4 °C for 12 h to completely denature the protein. The fact that we were not successful in refolding the denatured protein by rapid dilution and the dialysis methods in presence of stabilizing agents like arginine and glycerol (each in concentration of 0.1, 0.25, 0.5 and 0.75 M), etc. [14], we opted for the on-column refolding approach [15–17]. In this method, we used Q-Sepharose anion exchanger resin and loaded the solubilized inclusion bodies. The denatured protein molecules (based on its pI value 5.2) bind to the resin and fold properly when the denaturant concentration is gradually reduced to zero. Immobilization caused by the binding of the protein onto
273
the column prevents possible aggregation that would have resulted otherwise [18]. After the denaturant concentration was reduced to zero, the column was eluted with varying NaCl concentrations (100, 200 and 300 mM). This avoids the co-elution of other undesired proteins. Most of the target protein was eluted in the first bed volume in 300 mM NaCl. With this procedure, we could get the protein with some minor impurities of higher molecular weights. To remove most of the minor impurities, the fractions were collected, pooled and concentrated in an Amicon ultrafiltration cell for size exclusion chromatography using Superdex-75 (Pharmacia) column pre-equilibrated with buffer. The purity of the protein was checked on SDS–PAGE and found to be pure for biophysical studies (Fig. 3). The yield of the protein was estimated to be 16 mg per 1 L of LB media and 10 mg per 1 L of M9 minimal media containing 15NH4Cl as the sole source of nitrogen [19]. The wet cell weights were about 5 and 4 g per liter of LB and M9 minimal media, respectively. MALDI-TOF mass spectrometry MALDI-TOF mass spectrometry was used to determine the oligomerization state of the protein, if any. The mass spectrometric analysis showed two peaks in the spectrum; one at 10,251 m/z that corresponds to the monomeric state, and other at 20,556 m/z which corresponds to the dimer peak. The dimer peak is relatively very weak (Fig. 5). It confirmed that protein exists largely in monomeric state in solution under the experimental conditions. NMR characterization Hahellin consists of single polypeptide chain containing 92 amino acid residues. The amino acid sequence of the protein is shown below (Scheme 1). NMR is one of the techniques to determine whether a protein is folded or not. The unfolded protein molecule gives rise to a collapsed spectrum where several peaks come in a narrow chemical shift range [20]. We used HSQC to see if the refolded protein is properly folded after purification and is suitable for structural characterization. As seen in Fig. 6, the folded nature of the protein molecule is revealed by the well-dispersed [15N–1H] HSQC spectrum, both along the proton (1H) and nitrogen (15N) chemical shift axes. The protein is rich in serine residues (12/92; 13%) with only two proline residues. There are no cysteine or tryptophan residues. The N-terminal amino acid residues with exchangeable backbone amide proton (HN) and proline which do not have amide proton do not show any MGEKTVKLYE DTHFKGYSVE LPVGDYNLSS LISRGALNDD LSSARVPSGL RLEVFQHNNF KGVRDFYTSD AAELSRDNDA SSVRVSKMET TN
Scheme 1.
274
A.K. Srivastava et al. / Protein Expression and Purification 58 (2008) 269–274
chejuensis. This work was supported by a grant (Y.S) from the Department of Science and Technology (DST), Govt. of India. References
Fig. 6. Sensitivity enhanced 2D [15N–1H] HSQC of uniformly 15N-labeled Hahellin. The spectrum was acquired with 2048 data points in t2 and 64 in t1 dimension with spectral widths of 11 and 28 ppm along 1H and 15N dimensions, respectively. The number of transients collected for each t1 increment was 2 and the total experiment time was 5 min. The data were multiplied with shifted sine-square bell window function both along t1 and t2 axes, and zero-filled to 4096 along t2 dimension and to 256 along t1 dimension, prior to 2D Fourier transform.
15
N–1H correlation in the [15N–1H] HSQC spectrum. Therefore, one should expect to see at least 89 peaks or 15 N–1H correlations in the [15N–1H] HSQC spectrum recorded in a mixed solvent of 90% H2O and 10% 2H2O. Fig. 6 shows [15N–1H] HSQC spectrum recorded on a uniformly 15N-labeled protein. The dispersion of the peaks suggests the well folded structure of the protein, which accounts for around 87 peaks. We describe the refolding of a bc-crystallin domain of a novel member protein using an on-column refolding procedure. The protein obtained is refolded as seen with a welldispersed 2D [15N–1H] HSQC spectrum. Further structural characterization of Hahellin is underway. Acknowledgments The facilities provided by the National Facility for High Field NMR, supported by DST, DBT, CSIR, and Tata Institute of Fundamental Research, Mumbai, India, are greatly acknowledged. We thank Prof. Jihyun F. Kim, 21C Frontier Microbial Genomics and Applications, Center, Korea Research Institute of Bioscience and Biotechnology, Republic of Korea, for providing the DNA of Hahella
[1] N.H. Lubsen, H.J. Aarts, J.G. Schoenmakers, The evolution of lenticular proteins: the beta- and gamma-crystallin super gene family, Prog. Biophys. Mol. Biol. 51 (1988) 47–76. [2] R. Jaenicke, C. Slingsby, Lens crystallins and their microbial homologs: structure, stability, and function, Crit. Rev. Biochem. Mol. Biol. 36 (2001) 435–499. [3] H. Bloemendal, J.W. de, R. Jaenicke, N.H. Lubsen, C. Slingsby, A. Tardieu, Aging and vision: structure, stability and function of lens crystallins, Prog. Biophys. Mol. Biol. 86 (2004) 407–485. [4] G. Wistow, Evolution of a protein superfamily: relationships between vertebrate lens crystallins and microorganism dormancy proteins, J. Mol. Evol. 30 (1990) 140–145. [5] H. Jeong, J.H. Yim, C. Lee, S.H. Choi, Y.K. Park, S.H. Yoon, C.G. Hur, H.Y. Kang, D. Kim, H.H. Lee, K.H. Park, S.H. Park, H.S. Park, H.K. Lee, T.K. Oh, J.F. Kim, Genomic blueprint of Hahella chejuensis, a marine microbe producing an algicidal agent, Nucleic Acids Res. 33 (2005) 7066–7073. [6] H.K. Lee, J. Chun, E.Y. Moon, S.H. Ko, D.S. Lee, H.S. Lee, K.S. Bae, Hahella chejuensis gen. nov., sp. nov., an extracellular-polysaccharide-producing marine bacterium, Int. J. Syst. Evol. Microbiol. 51 (2001) 661–666. [7] G. Wistow, L. Summers, T. Blundell, Myxococcus xanthus spore coat protein S may have a similar structure to vertebrate lens beta gammacrystallins, Nature 315 (1985) 771–773. [8] G. Wistow, B. Turnell, L. Summers, C. Slingsby, D. Moss, L. Miller, P. Lindley, T. Blundell, X-ray analysis of the eye lens protein gammaII crystallin at 1.9 A resolution, J. Mol. Biol. 170 (1983) 175–202. [9] J. Graw, The crystallins: genes, proteins and diseases, Biol. Chem. 378 (1997) 1331–1348. [10] S.C. Makrides, Strategies for achieving high-level expression of genes in Escherichia coli, Microbiol. Rev. 60 (1996) 512–538. [11] T. Grune, T. Jung, K. Merker, K.J. Davies, Decreased proteolysis caused by protein aggregates, inclusion bodies, plaques, lipofuscin, ceroid, and ‘aggresomes’ during oxidative stress, aging, and disease, Int. J. Biochem. Cell Biol. 36 (2004) 2519–2530. [12] P. Valax, G. Georgiou, Molecular characterization of beta-lactamase inclusion bodies produced in Escherichia coli. 1. Composition, Biotechnol. Prog. 9 (1993) 539–547. [13] S.M. Singh, A.K. Panda, Solubilization and refolding of bacterial inclusion body proteins, J. Biosci. Bioeng. 99 (2005) 303–310. [14] A.P. Middelberg, Preparative protein refolding, Trends Biotechnol. 20 (2002) 437–443. [15] A. Jungbauer, W. Kaar, R. Schlegl, Folding and refolding of proteins in chromatographic beds, Curr. Opin. Biotechnol. 15 (2004) 487–494. [16] M. Li, Z.G. Su, J.C. Janson, In vitro protein refolding by chromatographic procedures, Protein Expr. Purif. 33 (2004) 1–10. [17] M.K. Jobby, Y. Sharma, Purification of a crystallin domain of Yersinia crystallin from inclusion bodies and its comparison to native protein from the soluble fraction, Biomed. Chromatogr. 20 (2006) 956–963. [18] J. Suttnar, J.E. Dyr, E. Hamsikova, J. Novak, V. Vonka, Procedure for refolding and purification of recombinant proteins from Escherichia coli inclusion bodies using a strong anion exchanger, J. Chromatogr. B Biomed. Appl. 656 (1994) 123–126. [19] J. Sambrook et al., Molecular Cloning: A Laboratory Manual A.1, second ed., Plainview, NY, 1989. [20] Y. Bai, J. Chung, H.J. Dyson, P.E. Wright, Structural and dynamic characterization of an unfolded state of poplar apo-plastocyanin formed under nondenaturing conditions, Protein Sci. 10 (2001) 1056– 1066.