TY - JOUR
AU - Mandal, Dipak K.
AB - Abstract Guanidine hydrochloride (GdnHCl)-induced unfolding of bovine spleen galectin-1 (Gal-1) exhibits three-state mechanism involving exclusive, structured tertiary monomer in 0.5 M GdnHCl. Gal-1 has one tryptophan residue (Trp 68) per subunit. Phosphorescence spectra of both Gal-1 dimer and intermediate monomer at 77 K show single (0,0) band at 405.6 nm, characteristics of free tryptophan environment as of N-acetyl-l-tryptophanamide. Unfolded Gal-1 in 4 M GdnHCl gives (0,0) band at longer wavelength (408.6 nm) signifying that Trp 68 is less solvent exposed, being localized in an environment of residual structure. Trifluoroethanol (TFE)- and hexafluoroisopropanol (HFIP)-induced structural changes of Gal-1 dimer and monomer have been investigated by far-UV CD and FTIR. CD results show reversible nature of β-sheet to α-helix transition, with 30% helix in 80% TFE or 40% HFIP for Gal-1 dimer. Temperature-dependent studies show that induced helix entails reduced thermal stability. FTIR results reveal partial β-sheet to α-helix conversion but with quantitative yield. At intermediate TFE concentration, both Gal-1 dimer and monomer aggregate as evidenced by FTIR band at ∼1617/cm, however, the onset of aggregation and stability of aggregates for monomer differ from those of dimer. The results may provide important insights into perturbed folding problem of Gal-1. β-sheet lectin, galectin-1, protein aggregation, protein denaturation, solvent perturbation Folding of proteins is believed to occur through a limited number of partially folded intermediate species on a funnel-shaped energy landscape (1) and structural properties of those non-native states should provide valuable insights into interactions responsible for their formation as well as their role in protein folding (2). Apart from protein folding, these non-native states may be involved as off pathway structures such as misfolded structures or protein aggregation leading to amyloid formation (3). Folding intermediates can be obtained by perturbation of protein structures by chaotropic agent such as guanidine hydrochloride (GdnHCl) or by alcohol. In general, GdnHCl produces intermediates with native-like secondary conformations (4) whereas alcohols are known to produce non-native states with non-native secondary structures (5–8). 2,2,2-Trifluoroethanol (TFE) and 1,1,1,3,3,3-hexafluoroisopropanol (HFIP) are widely used to denature and characterize non-native secondary structures in protein folding studies (5–8). These fluoroalcohols cause disruption of native structures and tend to stabilize α-helical structures of proteins and peptides. Alcohol destabilizes hydrophobic cores of protein because of its non-polar character while it enhances helix formation of protein by minimizing exposure of peptide backbone (9). Individual protein has specific helix forming propensity depending on amino acid sequence (6), and it has been shown that extensive helix formation in TFE does not occur when a protein has a very low intrinsic propensity for helix formation (10). Galectins are one of the most important protein families among the animal lectins, defined by their β-galactoside-binding specificity with highly conserved sequence (11). They participate in a variety of biological processes such as cell adhesion, growth regulation, cancer metastasis, apoptosis, RNA transcription and also in the organism development or the progression of pathological conditions, such as cancerogenesis, immune response or even pathogen entry (12). Galectin-1 was the first member of this family to be identified. It is expressed in many tissues in various species including human, chicken, cow, mouse and rat. It exists as a non-covalently linked homodimer of 14 kDa subunits having two carbohydrate recognition domains per dimer. Crystal structure of galectin-1 from bovine spleen has been determined, which displayed tertiary structure comprising the typical jelly-roll topology similar to that of legume lectin, although there is no detectable sequence similarity (13). Each subunit consists of a five- and a six-stranded anti-parallel β-sheets (Fig. 1A). Both β-sheets extend across the dimer interface, in a 2-fold symmetric fashion, to create the continuous 10- and 12-stranded anti-parallel β-sheets (Fig. 1B). Fig. 1 View largeDownload slide Structure of (A) Gal-1 monomer with Trp 68 highlighted in yellow ball-and-stick and (B) Gal-1 dimer (PDB id: 1slt). Fig. 1 View largeDownload slide Structure of (A) Gal-1 monomer with Trp 68 highlighted in yellow ball-and-stick and (B) Gal-1 dimer (PDB id: 1slt). The structural perturbation induced by chemical denaturant and fluoroalcohol of galectin-1, and its implications in protein folding and aggregation would constitute an interesting theme of study. Previously, a denaturation study in GdnHCl of porcine spleen galectin-1 involving biphasic transition was reported (14). In this article, we have investigated differing environment of single tryptophan residue of bovine spleen galectin-1 (Gal-1) in dimeric, monomeric and unfolded states obtained in GdnHCl-induced unfolding reaction using triplet state emission spectroscopy. Furthermore, we have explored the TFE- and HFIP-induced denaturation of Gal-1 and their role on secondary structure perturbation and onset of aggregation at the level of Gal-1 dimer and monomer by far-UV CD and FTIR. The results provide important insights into the perturbed folding problem of this β-sheet animal lectin. Materials and Methods Materials Divinyl sulfone, diethylaminoethyl sepharose (DEAE Sepharose), dithiothreitol (DTT), deuterium oxide (D2O) and GdnHCl, TFE, HFIP were purchased from Sigma. 2-Iodoacetamide was purchased from E. Merck, Germany. Sepharose 4B of Pharmacia was used for preparation of lactosyl sepharose 4B. TFE-OD was obtained by distilling TFE with D2O. Fresh bovine spleen was purchased from local slaughterhouse. All other reagents used were of analytical grade. Double-distilled water was used throughout. Protein purification and chemical modification Galectin-1 was purified from bovine spleen by anion exchange chromatography on DEAE sepharose column followed by affinity chromatography on lactosyl sepharose 4B column (15). Lactosyl sepharose 4B was prepared from sepharose 4B by activation with divinylsulfone and coupling with lactose (16). Concentration of galectin-1 from bovine spleen (Gal-1) was determined spectrophotometrically using A1%, 1 cm = 6.5 at 280 nm and expressed in terms of monomer (Mr = 14,500) (15). Free –SH group of cysteine residues of Galectin-1 was chemically modified to S-acetamido group (-S-CH2CONH2) via treatment with 2-iodoacetamide (17). Excess free iodoacetamide was removed by extensive dialysis against PBS (20 mM sodium phosphate containing 0.15 M NaCl, pH 7.2). Chemical denaturation experiments Gal-1 samples (50 µg/ml) were prepared in requisite concentration of GdnHCl in DPBS buffer (20 mM sodium phosphate containing 0.15 M NaCl and 2 mM DTT, pH 7.2). The samples were kept at room temperature for 18 h to ensure equilibrium and the unfolding reactions were monitored by steady state intrinsic (tryptophan) as well as extrinsic (ANS) fluorescence. Size-exclusion chromatography GdnHCl-induced denaturation was characterized by size-exclusion chromatography, performed on Superose-12 10/300 GL column (bed volume: 24 ml) attached to a Waters HPLC system. An aliquot of 200 µl of a protein sample (40 µM) prepared by incubation with required concentration of GdnHCl in DPBS buffer, pH 7.2 as described above was injected into the column. The column was preequilibrated with the same buffer in which the protein sample was prepared. The flow rate was 0.5 ml/min, and eluent was checked on-line for absorption at 280 nm by Waters 2489 UV-Visible detector. Steady-state absorption and fluorescence measurements Ultraviolet absorption was measured in a Hitachi U 4100 double-beam spectrophotometer and steady-state fluorescence measurements were done with a Hitachi F-7000 spectrofluorimeter (equipped with a 150 W xenon lamp), both using Sigma cuvette (volume: 2 ml; path length: 1 cm). The excitation and emission band pass was 5 nm each, and scan speed was fixed at 60 nm/min. Gal-1 samples were prepared in DPBS buffer, pH 7.2 with requisite amount of GdnHCl. For tryptophan fluorescence measurement, samples were excited at 280 nm and scanned between 300 and 400 nm. For ANS fluorescence, these previously prepared protein samples were mixed with 100 µM ANS solution, kept for 5 min before excitation at 370 nm and emissions were scanned in the region 400–600 nm. All spectra were corrected by subtraction of appropriate blanks. Phosphorescence measurement at 77 K Phosphorescence spectra of galectin samples in presence of various concentrations of GdnHCl were recorded in freezing condition (at 77 K) in a Hitachi F-7000 spectrofluorimeter equipped with phosphorescence accessories using a Dewar system having a 5 mm o.d. quartz tube. All the samples of Gal-1 dimer and monomer were made in 40% ethylene glycol in appropriate buffer. Protein concentration was 40 μM in each case. Samples were frozen in liquid nitrogen immediately after mixing with ethylene glycol. Excitation wavelength was set at 280 nm with 10-nm band pass. Emission band pass was 2.5 nm. All spectra were found to be reproducible and free from any polarization artifacts. CD measurement CD measurements were done on a J-815 spectropolarimeter (Jasco, Japan) equipped with a Peltier-type temperature controller in far-UV region (190–260 nm). Gal-1 samples were prepared in 20 mM DPBS buffer, pH 7.2. Samples of Gal-1 derivative were made in PBS, pH 7.2. Modified and unmodified Gal-1 monomer was prepared after incubation for 18 h in presence of 0.5 M GdnHCl in PBS and DPBS buffer, respectively. Requisite amount of TFE or HFIP was added directly for experiments with the Gal-1 dimer. For Gal-1 monomer, a 0.5 M GdnHCl solution in 80% TFE/HFIP was prepared and it was used as stock TFE/HFIP solution for measurement of CD spectra of monomer. All the spectra were recorded in a 1-mm path-length cell using a scan speed of 50 nm/min with response time of 2 s and averaged over five scans to eliminate signal noise. For thermal reversibility studies in the range of 20–60°C, Gal-1 samples in TFE (or HFIP) were gradually heated to higher temperatures followed by a gradual decrease to initial temperatures. To test the reversibility of alcohol perturbation process, derivative of Gal-1 dimer in PBS, pH 7.2 was mixed with TFE (or HFIP) to a final concentration of 80% TFE (or 60% HFIP) and the mixture was dialysed extensively against the same buffer to remove TFE/HFIP from sample. The samples obtained after completion of dialysis were examined by far-UV CD. FTIR measurement FTIR absorption spectra were obtained on a FT/IR-680 plus spectrometer (Jasco, Japan) at room temperature at 4/cm resolution as an average of 512 scans. Before measuring FTIR spectra, native dimeric Gal-1 in DPBS buffer was diluted in D2O containing 20 mM DTT and concentrated in Vivaspin 6 concentrator (GE Healthcare). This process was repeated several times so that all water molecules were replaced by D2O and the final protein concentration was ∼6 mg/ml. The native sample, prepared in D2O containing DTT, was kept for 15 min and then placed in a sealed cell composed of two ZnSe windows and a Teflon spacer of 100-µm path length. Gal-1 monomer was generated by incubation of native Gal-1 in 0.5 M GdnHCl in D2O containing DTT at room temperature for 18 h and thereafter concentrated against 0.5 M GdnHCl in D2O containing 20 mM DTT to replace the water molecules. For measurements with different TFE-perturbed systems, required amount of stock TFE-OD (100% in case of dimer and 90% TFE-OD containing 0.5 M GdnHCl for monomer) was added in the respective samples, kept for 15 min and thereafter FTIR spectra were recorded. In all cases appropriate blanks were subtracted and spectra in the range of 1,700–1,600/cm were examined. Results and Discussion Galectin-1 (Gal-1) from bovine spleen exists as a dimer in its native state. Structural perturbations of Gal-1 with concomitant unfolding, as induced by chemical denaturant such as GdnHCl, and by fluoroalcohols (TFE and HFIP) have been investigated using size-exclusion HPLC, fluorescence, low temperature phosphorescence, far-UV CD and FTIR. Gal-1 contains six cysteine (free-SH) residues per monomer. To prevent Gal-1 from oxidation during unfolding reaction, the experiments were carried out either with Gal-1 in presence of DTT or with iodoacetamide-alkylated Gal-1 (S-acetamido Gal-1). GdnHCl- induced unfolding of bovine spleen Gal-1 Figure 2A shows a biphasic denaturation curve of Gal-1 as monitored by intrinsic fluorescence. Gal-1 dimer shows a characteristic emission maximum at 340 nm, which undergoes a small shift to 343 nm for the intermediate species in 0.5 M GdnHCl, and finally to 347 nm for the completely unfolded protein in 4 M GdnHCl. Figure 2B depicts a plot of ANS fluorescence intensity at 470 nm as a function of GdnHCl concentration. It is seen that the fluorescence intensity at 470 nm for the ANS–native dimer complex is negligible which, however, increases dramatically in GdnHCl concentration range of 0.25–0.75 M (∼20-fold in 0.5 M GdnHCl). ANS fluorescence then falls and becomes very small in ≥3 M GdnHCl. The result indicates the existence of compact molten globule-like intermediate around 0.5 M GdnHCl. The size of the intermediate in 0.5 M GdnHCl was characterized by size exclusion HPLC (Fig. 3). The elution profile shows that the intermediate species elutes as a single peak at larger elution volume compared to that for native dimer, with an estimated mass of 14.5 kDa which corresponds to a monomer. It may be mentioned that low concentration of GdnHCl can stabilize the folded protein toward unfolding (18, 19); however, with further increase in concentration, the function of GdnHCl as a denaturant dominates. GdnHCl is ionic. It has been suggested by Mayr and Schmid (18) that the stabilization of ribonuclease T1 at low concentration (0.3 M) of GdnHCl arises from the electrostatic binding of guanidinium ions to specific cation-binding sites of the protein. Under sub-denaturing concentrations of GdnHCl, the entropic effect brought about by non-covalent interactions between GdnHCl and protein molecules as well as the electrostatic effect contribute to the stability of a protein (19). An exclusive intermediate of Gal-1 ∼0.5 M GdnHCl is formed by subunit dissociation without monomer unfolding. The formation of compact monomeric intermediate may be attributed, at least in part, to the stabilizing effect of GdnHCl under low concentration. The monomeric intermediate is completely unfolded in ≥3 M GdnHCl. The size-exclusion profile of Gal-1 in 4 M GdnHCl (Fig. 3) shows a single peak at lower elution volume than that of the dimer, which suggests that Gal-1 monomer unfolds and forms a species with an increased apparent size, as has been observed for GroES unfolding (20). The secondary β-sheet structure of native Gal-1 remains unperturbed in the intermediate monomer as monitored by far-UV CD and FTIR (described later). It has been observed that the pattern of GdnHCl-induced unfolding reaction of S-acetamido Gal-1 remains similar (data not shown). This type of three-state unfolding mechanism in GdnHCl is similar to that reported previously for galectin-1 from porcine spleen (14) however, the GdnHCl concentration (0.5 M) for the intermediate in the unfolding reaction of bovine spleen Gal-1 is considerably lower than that required (1.75 M) for porcine spleen galectin-1. Fig. 2 View largeDownload slide (A) GdnHCl-induced denaturation curve of bovine spleen galectin-1 (Gal-1). Excitation wavelength, 280 nm. (B) A plot of ANS fluorescence intensity at 470 nm as a function of GdnHCl concentration, Excitation wavelength, 370 nm. Protein concentration was 50 μg/ml and ANS was 100 μM in each case. Excitation and emission band pass, 5 nm each; scan rate, 60 nm/min. The spectra were corrected for appropriate blanks. Fig. 2 View largeDownload slide (A) GdnHCl-induced denaturation curve of bovine spleen galectin-1 (Gal-1). Excitation wavelength, 280 nm. (B) A plot of ANS fluorescence intensity at 470 nm as a function of GdnHCl concentration, Excitation wavelength, 370 nm. Protein concentration was 50 μg/ml and ANS was 100 μM in each case. Excitation and emission band pass, 5 nm each; scan rate, 60 nm/min. The spectra were corrected for appropriate blanks. Fig. 3 View largeDownload slide Size exclusion elution profile on Superose-12 10/300 GL column of Gal-1 in 0 M (a), 0.5 M (b) and 4 M (c) GdnHCl in DPBS buffer (20 mM sodium phosphate containing 0.15 M NaCl and 2 mM DTT, pH 7.2). Inset: molecular weight calibration curve. The column was calibrated with the following proteins (from left to right): native galectin-1 (28 kDa), soybean trypsin inhibitor (20.1 kDa) and cytochrome c (12 kDa). Elution position of galectin-1 in 0.5 M GdnHCl is marked by an arrow. Fig. 3 View largeDownload slide Size exclusion elution profile on Superose-12 10/300 GL column of Gal-1 in 0 M (a), 0.5 M (b) and 4 M (c) GdnHCl in DPBS buffer (20 mM sodium phosphate containing 0.15 M NaCl and 2 mM DTT, pH 7.2). Inset: molecular weight calibration curve. The column was calibrated with the following proteins (from left to right): native galectin-1 (28 kDa), soybean trypsin inhibitor (20.1 kDa) and cytochrome c (12 kDa). Elution position of galectin-1 in 0.5 M GdnHCl is marked by an arrow. Tryptophan environment in different structural states of bovine spleen Gal-1 as monitored by phosphorescence Gal-1 possesses a single tryptophan residue (Trp 68) in a subunit, which appears to be critical for distinguishing the galactose ring from the glucose ring, because of its preference for the axial C4-OH responsible for intimate C−H−π-cloud interactions (21), and has been found to be instrumental for stacking of the lactose moiety (22) Trp 68 thus plays a critical role in galectin–glycan interaction, and thereby governs ligand binding and galectin function. Compared with the fluorescence spectra of tryptophan residues which are generally broad, phosphorescence studies at 77 K provide more structured spectra with characteristic (0,0) band, and hence more useful to probe the tryptophan environment in different structural states involved in Gal-1 unfolding in GdnHCl. The position of the (0,0) band could be rationalized in terms of the solvent exposure of the tryptophan residue, though the local charges and the rigidity of the environment may also play their role. The position of (0,0) band in protein samples ranges from 403 to 420 nm (23). The shift of phosphorescence emission to shorter wavelength can be attributed to the lower polarizability of the environment and the less stabilization of the triplet state by rigid solvation geometry for the exposed tryptophans whereas the red-shifted phosphorescence occurs due to the tryptophan residues located in a buried polarizable environment that stabilizes the triplet state more than the ground state (24). The phosphorescence spectra for Gal-1 dimer, monomer and unfolded Gal-1 in 40% ethylene glycol at 77 K are shown in Fig. 4. It is observed that both native dimer and intermediate monomer give single (0,0) band at 405.6 nm (spectra b and c; Fig. 4) in the blue-shifted region which indicates that Trp 68 is highly solvent exposed in solution. To compare the degree of solvent exposure of Trp 68, the phosphorescence emission was measured with N-acetyl-l-tryptophanamide (NATA)—a free tryptophan analogue, which shows (0,0) band at nearly same position at 406.2 nm (spectrum a; Fig. 4). These results demonstrate that Trp 68 environment in native dimer/intermediate monomer resembles very closely to that of free tryptophan in solution. This observation is also consistent with the location of Trp 68 in crystal structure (Fig. 1A). Interestingly, when Gal-1 was fully denatured in ≥4 M GdnHCl, the (0,0) band of the unfolded Gal-1 with complete loss of secondary β-structure (data not shown) exhibits a significant (3 nm) red shift to 408.6 nm (spectrum d, Fig. 4). To determine whether this occurs due to wide difference in solvent conditions, the phosphorescence of NATA was measured in 4 M GdnHCl, that shows a negligible shift of (0,0) band (spectra a and e, Fig. 4). It therefore follows that Trp 68 of Gal-1 in GdnHCl-induced unfolded state does not resemble free tryptophan environment of native protein, but is less solvent-exposed. The denaturant-induced unfolded state, once well-modelled by an unstructured random coil; is now considered to be more complex, and comprises a broad structural ensemble called the denatured state ensemble (DSE) which may hold important clues into the ‘folding code’ (25). Recent work on DSE shows that the pattern of deviations from random coil under high GdnHCl concentration indicates that non-polar interactions persist at high denaturant concentration (26). The phosphorescence red shift in the GdnHCl-induced unfolded state of Gal-1 might occur due to the Trp 68 being localized in an environment of flickering element of residual structure constituting a non-native hydrophobic cluster around the tryptophan. The present results thus conform to and provide further insight into the DSE model. Fig. 4 View largeDownload slide Phosphorescence spectra of NATA in 0 M (a) GdnHCl; Gal-1 in 0 M (b), 0.5 M (c), 4 M (d) GdnHCl; NATA in 4 M (e) GdnHCl. Excitation wavelength, 280 nm; excitation and emission band passes were 10 and 2.5 nm, respectively. Fig. 4 View largeDownload slide Phosphorescence spectra of NATA in 0 M (a) GdnHCl; Gal-1 in 0 M (b), 0.5 M (c), 4 M (d) GdnHCl; NATA in 4 M (e) GdnHCl. Excitation wavelength, 280 nm; excitation and emission band passes were 10 and 2.5 nm, respectively. TFE- and HFIP-induced conformational change of Gal-1 dimer as monitored by far-UV CD Crystal structure of native Gal-1 dimer shows that its secondary structure is predominantly β-sheet (61%) with no α-helix (Fig. 1B) (13). Figure 5A shows the far-UV CD spectra of Gal-1 in increasing concentrations of TFE at 20°C. In absence of TFE, the protein exhibits a typical β-sheet band shape with characteristic negative peak at 217 nm. With addition of TFE up to 20%, this band shape remains unaltered. Thereafter the spectrum begins to change significantly and in presence of 30% TFE, it assumes α helical band shape characterized by two negative peaks at 219 and 207 nm. With further increase in TFE concentration up to 80%, the α-helical band shape persists with progressive rise of intensity, indicating more helical transformation. Figure 5B shows far-UV CD spectra of Gal-1 with increasing HFIP concentration. In contrast to TFE, limited addition of 10% HFIP induces change in spectral band shape which broadens with increase in negative intensity. The CD spectra acquire a prominent α helical band shape at HFIP concentration of 30% onward. Fig. 5 View largeDownload slide Far-UV CD spectra of Gal-1 dimer in presence of (A) 0% (a), 10% (b), 20% (c), 30% (d), 40% (e), 60% (f), 80% (g) TFE and (B) 0% (h), 10% (i), 30% (j), 40% (k), 60% (l) HFIP at 20°C. Protein concentration was 0.28 mg/ml. Scan speed was fixed at 50 nm/min. Corresponding buffer spectra were subtracted in each case and at least five scans were performed. (C) Plot of β-sheet content (open square) and change of α-helix content (filled square) for Gal-1 dimer as a function of TFE concentration. (D) Plot of β-sheet content (open triangle) and change of α-helix content (filled triangle) for Gal-1 dimer as a function of HFIP concentration. Secondary structure components were estimated from analysis of far-UV CD spectra using CDNN software, version 2.1. See the text for details. Fig. 5 View largeDownload slide Far-UV CD spectra of Gal-1 dimer in presence of (A) 0% (a), 10% (b), 20% (c), 30% (d), 40% (e), 60% (f), 80% (g) TFE and (B) 0% (h), 10% (i), 30% (j), 40% (k), 60% (l) HFIP at 20°C. Protein concentration was 0.28 mg/ml. Scan speed was fixed at 50 nm/min. Corresponding buffer spectra were subtracted in each case and at least five scans were performed. (C) Plot of β-sheet content (open square) and change of α-helix content (filled square) for Gal-1 dimer as a function of TFE concentration. (D) Plot of β-sheet content (open triangle) and change of α-helix content (filled triangle) for Gal-1 dimer as a function of HFIP concentration. Secondary structure components were estimated from analysis of far-UV CD spectra using CDNN software, version 2.1. See the text for details. The secondary structure components were estimated by deconvolution of far-UV CD spectra using CDNN software (27). The result shows predominant β-sheet secondary structure component in native Gal-1 (40%), while overestimating the helical component (11%) when compared with crystal structure and FTIR analysis (shown below), which shows no α-helix. To depict β-sheet to α-helix transformation in presence of TFE/HFIP using far-UV CD, the overestimated helical component is taken as a background data, based on which the progressive change in α-helix content is estimated. Figure 5C shows a plot of change of secondary structure elements (β-sheet, α-helix) as a function of TFE concentration. With progressive addition of TFE, helical component quantitatively increases at the expense of sheet structure and finally reaches to 30% α-helix at 80% TFE. Plot of percentage of secondary structure components against HFIP concentration is shown in Fig. 5D. For HFIP, quantitative helix induction (∼30%) occurs in 40% HFIP—a much lower amount compared to TFE perturbation. This implies that HFIP exerts more perturbing influence than TFE for helical conversion. More HFIP, however, does not lead to further helix transformation. The results suggest that Gal-1 sequence must have some helical propensity, though it appears to be much less when compared with β-sheet legume lectins such as soybean agglutinin (28) and concanavalin A (8) which possess similar ‘jelly roll’ toppology as Gal-1 (13). The β-sheet to α-helix transformation was monitored at different temperatures within 20–60°C. Figure 6A represents far-UV CD spectra of Gal-1 in presence of 50% TFE at different temperatures. As shown, the spectra exhibit α-helical band shape at all temperatures (20–60°C). However, the intensity of the α-helical signal decreases with rise in temperature. These results imply that induced helix entails reduced thermal stability. The reversibility of the heat induced α-helix conversion has also been examined. It is observed that the decreased intensity of induced α-helix at higher temperature is completely regained upon lowering the temperature (Fig. 6A). The results clearly indicate that the heat induced α-helical conversion is completely reversible. Fig. 6 View largeDownload slide (A) Far-UV CD spectra of Gal-1 dimer in presence of 50% TFE on heating progressively from 20°C (a) to 40°C (b) to 60°C (c), and thereafter cooling back to 40°C (d) and 20°C (e). (B) Normalized far-UV CD spectra at pH 7.2 of dimeric Gal-1 derivative without TFE (f) and sample of Gal-1 derivative after TFE removal (g) from mixture of protein and 80% TFE. Fig. 6 View largeDownload slide (A) Far-UV CD spectra of Gal-1 dimer in presence of 50% TFE on heating progressively from 20°C (a) to 40°C (b) to 60°C (c), and thereafter cooling back to 40°C (d) and 20°C (e). (B) Normalized far-UV CD spectra at pH 7.2 of dimeric Gal-1 derivative without TFE (f) and sample of Gal-1 derivative after TFE removal (g) from mixture of protein and 80% TFE. The reversibility of TFE- and HFIP-induced transition of native β-sheet to non-native α-helix conformation for Gal-1 or S-acetamido Gal-1 was examined. After generation of maximum amount of helical form, fluoroalcohol (TFE/HFIP) was removed by extensive dialysis, and the resulting protein samples were characterized by CD. Figure 6B shows far-UV CD spectra of S-acetamido Gal-1 before TFE addition and after TFE removal. It is seen that with removal of TFE, α-helical band disappears, and instead native-like typical β-sheet band shape (with a single negative peak at 217 nm) reappears. Similar results are obtained with HFIP-induced transition (data not shown). The results reveal the reversible nature of conformational transition of Gal-1 under the influence of fluoroalcohols. Theoretical computational studies, using a 2D lattice model, suggest that fluoroalcohol mainly weakens non-local hydrophobic interactions and slightly favours local helical interactions (9). It seems that non-local intramolecular interaction responsible for native β-conformation of Gal-1 is restored on removal of fluoroalcohol. Structural perturbation of Gal-1 monomer in TFE and HFIP As described before, Gal-1 monomer was generated exclusively in 0.5 M GdnHCl. Under this condition, far-UV CD spectra could not be recorded <205 nm. Figure 7A shows the far-UV CD spectra of Gal-1 monomer in presence of various concentrations of TFE. Without TFE, the CD spectrum exhibits typical β-sheet band shape (negative peak at 217 nm) as for Gal-1 dimer. In presence of 10–20% TFE, the band shape changes to a negative extremum at 226 nm for an atypical β-sheet with significant loss of intensity. The decreased intensity may probably arise due to aggregation under this condition. On further addition of TFE (≥30%), the far-UV CD spectra tend to assume α-helical band shape with progressive increase of signal intensity up to 50% TFE. Deconvolution of the CD spectral data (205–260nm) shows that 17% α-helix is induced solely from β-sheet component. An estimate of smaller yield in this case may be due to insufficient range of data. The pattern of the far-UV CD spectra of Gal-1 monomer in presence of HFIP is similar as for TFE perturbation (Fig. 7B). A somewhat larger yield of helical conformation (∼27%) results mainly from the β-sheet structure. Fig. 7 View largeDownload slide Far-UV CD spectra of Gal-1 monomer in presence of (A) 0% (a), 10% (b), 20% (c), 30% (d), 40% (e), 50% (f), 60% (g) TFE and (B) 0% (h), 10% (i), 20% (j), 30% (k), 40% (l), 50% (m), 60% (n) HFIP at 20°C. Protein concentration was 0.28 mg/ml. Scan speed was fixed at 50 nm/min. Corresponding buffer spectra were subtracted in each case and at least five scans were performed. Fig. 7 View largeDownload slide Far-UV CD spectra of Gal-1 monomer in presence of (A) 0% (a), 10% (b), 20% (c), 30% (d), 40% (e), 50% (f), 60% (g) TFE and (B) 0% (h), 10% (i), 20% (j), 30% (k), 40% (l), 50% (m), 60% (n) HFIP at 20°C. Protein concentration was 0.28 mg/ml. Scan speed was fixed at 50 nm/min. Corresponding buffer spectra were subtracted in each case and at least five scans were performed. TFE-induced secondary structure change of Gal-1 as monitored by FTIR FTIR spectroscopy serves as a useful technique to probe secondary structure components of proteins through investigation of mainly amide I absorption band which is contributed primarily by stretching vibrations of the peptide C = O linkage. This is because different secondary structure elements (β-sheet/α-helix) have definite pattern of hydrogen bonding involving C = O groups (29). Figure 8A shows the FTIR spectra of native Gal-1 dimer in absence and in presence of different concentrations of TFE-OD. Native Gal-1 exhibits the amide I′ band (N-deuterated) at 1630/cm characteristic of β-sheet structure. Analysis by protein secondary structure estimation software (JASCO, version 1.01.03) shows 56% β-sheet with no α-helix (Table 1), which is in close agreement with crystallographic data. In presence of 80% TFE-OD, a new amide I′ peak is obtained at 1651/cm indicating at least partial transformation to α-helical conformation. Analysis of the spectra reveals 35% helix which results solely from the change of β-sheet as the yield of α-helix is almost same as the decrease in β-sheet content (Table 1). The FTIR spectra for Gal-1 monomer (in 0.5 M GdnHCl) show similar pattern of amide I′ band without and with 80% TFE-OD (Fig. 8B). As for Gal-1 dimer, there occurs partial but quantitative conversion of β-sheet to α-helical conformation (Table 1). These results are also in good agreement with those obtained from far-UV CD studies. Fig. 8 View largeDownload slide FTIR amide I′ spectra of (A) Gal-1 dimer in presence of 0% (a), 20% (b), 40% (c) and 80% (d) TFE-OD; and (B) Gal-1 monomer with 0% (e), 40% (f) and 80% TFE-OD (g). Spectra were recorded at 4/cm resolution as an average of 512 scans. Corresponding buffer spectra were subtracted in each case and absorption data were normalized. Fig. 8 View largeDownload slide FTIR amide I′ spectra of (A) Gal-1 dimer in presence of 0% (a), 20% (b), 40% (c) and 80% (d) TFE-OD; and (B) Gal-1 monomer with 0% (e), 40% (f) and 80% TFE-OD (g). Spectra were recorded at 4/cm resolution as an average of 512 scans. Corresponding buffer spectra were subtracted in each case and absorption data were normalized. Table I. Estimation of β-sheet and α-helix from analysis of FTIR spectra of Gal-1 dimer and monomer in absence and presence of TFE. System  β-Sheet  α-Helix  Dimer  56  0  Dimer + 80% TFE-OD  21  35  Monomer  57  0  Monomer + 80% TFE-OD  25  33  System  β-Sheet  α-Helix  Dimer  56  0  Dimer + 80% TFE-OD  21  35  Monomer  57  0  Monomer + 80% TFE-OD  25  33  View Large Protein aggregation of Gal-1 in presence of fluoroalcohols Protein aggregation is associated with numerous pathogenic conditions such as Alzheimer’s, Parkinson’s and Creutzfeldt-Jakob diseases. A well-studied class of protein aggregates is the highly structured amyloid fibrils (30). Protein aggregation occurs in response to several stresses including changes in solvent conditions, and chemical co-solvents such as TFE are frequently employed in these studies (31). The molecular mechanism for protein aggregation is, however, poorly understood. It has been suggested that amyloid fibril formation from the native state takes place through conformational changes leading to the formation of sticky amyloid-prone, partially structured intermediate(s) (32). We have observed that both Gal-1 monomer and dimer give visible aggregation at intermediate concentrations of TFE. FTIR spectra (Fig. 8, spectra b and f) show a prominent amide I′ band at ∼1617/cm characteristic of extensive intermolecular β-sheet of amyloid fibrils (33). Far-UV CD studies indicate that, under relatively low TFE concentrations (10–20%), Gal-1 monomer destabilization leads to a non-native conformation which is prone to aggregation (Fig. 7). Analysis using prediction software–TANGO (http://tango.crg.es/) (34) and AGGRESCAN (http://bioinf.uab.es/aggrescan/) (35) delineates a few regions of Gal-1 sequence as probable ‘aggregation hotspots’ which include β-strand residues, 30–34 (FLLNL), 55–60 (VNTIVC), 84–91 (VVEVCISF) constituting mainly the hydrophobic side chains. However, under high TFE concentrations, the aggregation peak is observed to disappear with formation of helical structure. This might be possible as the large loss of tertiary structure disfavours presumed hydrophobically driven interactions for aggregate formation and shifts the balance from intermolecular β-sheet formation to intramolecular helical segment formation. It is interesting to note that at 40% TFE, aggregate of Gal-1 monomer remains intact (Fig. 8B, spectrum f) while it begins to disappear in case of Gal-1 dimer (Fig. 8A, spectrum c). The results show that the onset of aggregation and the stability of aggregates for Gal-1 monomer differ from those of dimer. These reflect the relative ease of tertiary structure disruption for Gal-1 monomer leading to extensive intermolecular β-sheet of amyloid fibrils (33). Conclusion The GdnHCl-induced denaturation of Gal-1 reveals a three-state mechanism that offers important and interesting insights as regards the environment of its single tryptophan residue. While the native Gal-1 dimer and the intermediate monomer resemble free tryptophan environment in solution, the GdnHCl-induced unfolded state depicts a relatively buried tryptophan environment. Unlike the native state of a protein, which has a unique structure, the unfolded state may comprise a complex state of a broad structural ensemble. As the starting point for the refolding of a protein, this DSE may hold important clues into the folding code (25). The present results signify the residual structure in the DSE which may be important in setting up the topology of a fold. The fluoroalcohols such as TFE and HFIP also denature globular proteins, typically leading to the formation of non-native α-helical structure. The present results show reversible nature of β-sheet to α-helix transition of Gal-1. It has been suggested that TFE-induced helicity is indicative of α-helical propensity based on the amino acid sequence (6), which might suggest a possibility of non-hierarchical model of protein folding. When present at intermediate concentrations, TFE induces the aggregation of Gal-1. It has been proposed that TFE mimics the environment occurring in proximity to biological membranes. Thus, such in vitro studies might reveal conformations involved in protein folding, transport and degradation pathways in living cells or adopted under different stress and disease conditions (36, 37). Funding This work was supported by research grant (No. 02/(0028)/11/EMR-II) from the Council of Scientific and Industrial Research (CSIR), Government of India. Conflict of interest None declared. Abbreviations Abbreviations CD circular dichroism DTT dithiothreitol FTIR Fourier transform infrared Gal-1 galectin-1 from bovine spleen GdnHCl guanidine hydrochloride HFIP 1,1,1,3,3,3-hexafluoroisopropanol HPLC high performance liquid chromatography NATA N-acetyl-l-tryptophanamide TFE 2,2,2-trifluoroethanol References 1 Dill KA,  Chan HS.  From Levinthal to pathways to funnels,  Nat. Struct. Biol. ,  1997, vol.  4 (pg.  10- 19) Google Scholar CrossRef Search ADS PubMed  2 Ptitsyn OB.  Kinetic and equilibrium intermediates in protein folding,  Protein Eng. ,  1994, vol.  7 (pg.  593- 596) Google Scholar CrossRef Search ADS PubMed  3 Dobson CM.  Protein folding and misfolding,  Nature ,  2003, vol.  426 (pg.  884- 890) Google Scholar CrossRef Search ADS PubMed  4 Chatterjee A,  Mandal DK.  Denaturant-induced equilibrium unfolding of concanavalin A is expressed by a three-state mechanism and provides an estimate of its protein stability,  Biochim. Biophys. Acta ,  2003, vol.  1648 (pg.  174- 183) Google Scholar CrossRef Search ADS PubMed  5 Konno T,  Iwashita J,  Nagayama K.  Fluorinated alcohol, the third group of cosolvents that stabilize the molten-globule state relative to a highly denatured state of cytochrome c,  Protein Sci. ,  2000, vol.  9 (pg.  564- 569) Google Scholar CrossRef Search ADS PubMed  6 Shiraki K,  Nishikawa K,  Goto Y.  Trifluoroethanol-induced stabilization of the alpha- helical structure of beta- lactoglobulin: implication for non-hierarchical protein folding,  J. Mol. Biol. ,  1995, vol.  245 (pg.  180- 194) Google Scholar CrossRef Search ADS PubMed  7 Naseem F,  Khan RH.  Characterization of a common intermediate of pea lectin in the folding pathway induced by TFE and HFIP,  Biochim. Biophys. Acta ,  2005, vol.  1723 (pg.  192- 200) Google Scholar CrossRef Search ADS PubMed  8 Xu Q,  Keiderling TA.  Trifluoroethanol-induced unfolding of concanavalin A: equilibrium and time-resolved optical spectroscopic studies,  Biochemistry ,  2005, vol.  44 (pg.  7976- 7987) Google Scholar CrossRef Search ADS PubMed  9 Thomas PD,  Dill KA.  Local and nonlocal interactions in globular proteins and mechanisms of alcohol denaturation,  Protein Sci. ,  1993, vol.  2 (pg.  2050- 2065) Google Scholar CrossRef Search ADS PubMed  10 Schönbrunner N,  Wey J,  Engels J,  Georg H,  Kiefhaber T.  Native-like β-structure in a trifluoroethanol-induced partially folded state of the all-β-sheet protein tendamistat,  J. Mol. Biol. ,  1996, vol.  260 (pg.  432- 445) Google Scholar CrossRef Search ADS PubMed  11 Lis H,  Sharon N.  Lectins: carbohydrate-specific proteins that mediate cellular recognition,  Chem. Rev. ,  1998, vol.  98 (pg.  637- 674) Google Scholar CrossRef Search ADS PubMed  12 Liu FT,  Patterson RJ,  Wang JL.  Intracellular functions of galectins,  Biochim. Biophys. Acta ,  2002, vol.  1572 (pg.  263- 273) Google Scholar CrossRef Search ADS PubMed  13 Liao D,  Kapadia G,  Ahmed H,  Vasta GR,  Herzberg O.  Structure of S-lectin, a developmentally regulated vertebrate β-galactoside-binding protein,  Proc. Natl Acad. Sci. USA ,  1994, vol.  91 (pg.  1428- 1432) Google Scholar CrossRef Search ADS   14 Iglesias MM,  Elola MT,  Martinez V,  Fink N,  Todel CW.  Identification of an equilibrium intermediate in the unfolding process of galectin-1, which retains its carbohydrate-binding specificity,  Biochim. Biophys. Acta ,  2003, vol.  1648 (pg.  164- 173) Google Scholar CrossRef Search ADS PubMed  15 Ahmed H,  Fink NE,  Pohl J,  Vasta GR.  Galectin-1 from bovine spleen: iochemical characterization, carbohydrate specifity and tissue-specific isoform profiles,  J. Biochem. ,  1996, vol.  120 (pg.  1007- 1019) Google Scholar CrossRef Search ADS PubMed  16 Teichberg VI,  Aberdam D,  Erez U,  Pinelli E.  Affinity-repulsion chromatography. Principle and application to lectins,  J. Biol. Chem. ,  1988, vol.  263 (pg.  14086- 14092) Google Scholar PubMed  17 Stowell SR,  Cho M,  Feasley CL,  Arthur CM,  Song X,  Colucci JK,  Karmakar S,  Mehta P,  Dias-Baruffi M,  McEver RP,  Cummings RD.  Ligand reduces galectin-1 sensitivity to oxidative inactivation by enhancing dimer formation,  J. Biol. Chem. ,  2009, vol.  284 (pg.  4989- 4999) Google Scholar CrossRef Search ADS PubMed  18 Mayr LM,  Schmid FX.  Stabilization of a protein by guanidinium chloride,  Biochemistry ,  1993, vol.  32 (pg.  7994- 7998) Google Scholar CrossRef Search ADS PubMed  19 Bhuyan AK.  Protein stabilization by urea and guanidine hydrochloride,  Biochemistry ,  2002, vol.  41 (pg.  13386- 13394) Google Scholar CrossRef Search ADS PubMed  20 Higurashi T,  Nosaka K,  Mizobata T,  Nagai J,  Kawata Y.  Unfolding and refolding of Escherichia coli chaperonin GroES is expressed by a three-state model,  J. Mol. Biol. ,  1999, vol.  291 (pg.  703- 713) Google Scholar CrossRef Search ADS PubMed  21 Meynier M,  Guerlesquin F,  Roche P.  Computational studies of human galectin-1: role of conserved tryptophan residue in stacking interaction with carbohydrate ligands,  J. Biomol. Struct. Dyn. ,  2009, vol.  27 (pg.  49- 57) Google Scholar CrossRef Search ADS PubMed  22 Lopez-Lucendo MF,  Solis D,  Andre S,  Hirabayashi J,  Kasai K,  Kaltner H,  Gabius HJ,  Romero A.  Growth-regulatory human galectin-1: crystallographic characterisation of the structural changes induced by single-site mutations and their impact on the thermodynamics of ligand binding,  J. Mol. Biol. ,  2004, vol.  343 (pg.  957- 970) Google Scholar CrossRef Search ADS PubMed  23 Strambini EG,  Strambini GB.  Tryptophan phosphorescence as a monitor of protein conformation in molecular films,  Biosens. Bioelectron. ,  2000, vol.  15 (pg.  483- 490) Google Scholar CrossRef Search ADS PubMed  24 Ozarowski A,  Barry JK,  Matthews KS,  Maki AH.  Ligand-induced conformational changes in lactose repressor: a phosphorescence and ODMR study of single-tryptophan mutants,  Biochemistry ,  1999, vol.  38 (pg.  6715- 6722) Google Scholar CrossRef Search ADS PubMed  25 Baldwin RL.  A new perspective on unfolded proteins,  Adv. Protein Chem. ,  2002, vol.  62 (pg.  361- 367) Google Scholar PubMed  26 Bowler BE.  Residual structure in unfolded proteins,  Curr. Opin. Struct. Biol. ,  2012, vol.  22 (pg.  4- 13) Google Scholar CrossRef Search ADS PubMed  27 Böhm G. ,  CDNN: CD Spectra Deconvolution Software Version 2.1 ,  1997 Halle University of Halle-Wittenberg 28 Molla AR,  Mandal DK.  Trifluoroethanol-induced conformational change of tetrameric and monomeric soybean agglutinin: role of structural organization and implication for protein folding and stability,  Biochimie ,  2013, vol.  95 (pg.  204- 214) Google Scholar CrossRef Search ADS PubMed  29 Barth A.  Infrared spectroscopy of proteins,  Biochim. Biophys. Acta ,  2007, vol.  1767 (pg.  1073- 1101) Google Scholar CrossRef Search ADS PubMed  30 Chiti F,  Dobson CM.  Protein misfolding, functional amyloid, and human disease,  Annu. Rev. Biochem. ,  2006, vol.  75 (pg.  333- 366) Google Scholar CrossRef Search ADS PubMed  31 Otzen DE.  Amyloid formation in surfactants and alcohols: membrane mimetics or structural switchers?,  Curr. Protein Pept. Sci. ,  2010, vol.  11 (pg.  355- 371) Google Scholar CrossRef Search ADS PubMed  32 Srisailam S,  Kumar TKS,  Rajalingam D,  Kathir KM,  Sheu HS,  Jan FJ,  Chao PC,  Yu C.  Amyloid-like fibril formation in an all β-barrel protein. Partially structured intermediate state(s) is a precursor for fibril formation,  J. Biol. Chem. ,  2003, vol.  278 (pg.  17701- 17709) Google Scholar CrossRef Search ADS PubMed  33 Zandomeneghi G,  Krebs MRH,  Mccammon MG,  Fändrich M.  FTIR reveals structural differences between native β-sheet proteins and amyloid fibrils,  Protein Sci. ,  2004, vol.  13 (pg.  3314- 3321) Google Scholar CrossRef Search ADS PubMed  34 Fernandez-Escamilla AM,  Rousseau F,  Schymkowitz J,  Serrano L.  Prediction of sequence-dependent and mutational effects on the aggregation of peptides and proteins,  Nat. Biotech. ,  2004, vol.  22 (pg.  1302- 1306) Google Scholar CrossRef Search ADS   35 Conchilli-Sole’ O,  de Groot NS,  Avile’s FX,  Vendrell J,  Daura X,  Ventura S.  AGGRESCAN: a server for the prediction and evaluation of “hot-spots” of aggregation in polypeptides,  BMC Bioinf. ,  2007, vol.  8 pg.  65  Google Scholar CrossRef Search ADS   36 Munishkina LA,  Phelan C,  Uversky VN,  Fink AL.  Conformational behaviour and aggregation of α-synuclein in organic solvents: modeling the effects of membranes,  Biochemistry ,  2003, vol.  42 (pg.  2720- 2730) Google Scholar CrossRef Search ADS PubMed  37 Perham M,  Liao J,  Wittung-Stafshede P.  Different effects of alcohols on conformational switchovers in α-helical and β-sheet protein models,  Biochemistry ,  2006, vol.  45 (pg.  7740- 7749) Google Scholar CrossRef Search ADS PubMed  © The Authors 2013. Published by Oxford University Press on behalf of the Japanese Biochemical Society. All rights reserved
TI - Denaturation of bovine spleen galectin-1 in guanidine hydrochloride and fluoroalcohols: structural characterization and implications for protein folding
JO - The Journal of Biochemistry
DO - 10.1093/jb/mvt084
DA - 2013-09-13
UR - https://www.deepdyve.com/lp/oxford-university-press/denaturation-of-bovine-spleen-galectin-1-in-guanidine-hydrochloride-pDzfZX08ta
SP - 531
EP - 540
VL - 154
IS - 6
DP - DeepDyve
ER -