Conformational studies of a hyperthermostable enzyme Sotirios Koutsopoulos 1 , John van der Oost 2 and Willem Norde 1,3 1 Laboratory of Physical Chemistry and Colloid Science, Wageningen University, the Netherlands 2 Laboratory of Microbiology, Wageningen University, the Netherlands 3 Department of Biomedical Engineering, University Medical Center Groningen, the Netherlands Hyperthermophilic microorganisms predominantly belong to the Archaea, the third phylogenetic domain of life [1]. They flourish in environments of extreme temperatures even higher than 100 °C, which until recently were considered as incompatible with life. No multicellular organisms have been found to tolerate temperatures above 60 °C and no unicellular eukarya have been discovered to withstand long-term exposure to temperatures higher than % 70 °C. Pyrococcus furio- sus is an anaerobic hyperthermophile which was dis- covered in geothermally heated marine sediments at 100 °C [2]. It is a very efficient consumer of the organic material found on the sea floor such as pro- teins, peptides and sugar mixtures (e.g. maltose, cello- biose, oligosaccharides and starch), which are fermented and used as carbon source. P. furiosus has a large collection of hyperthermostable enzymes which may be used in important applications in biotechnology. One of them, the extracellular endo- b-1,3-glucanase (LamA), has been isolated and charac- terized [3]. LamA hydrolyzes 1,3-b-glycosyl bonds of polysaccharides such as laminarin. The temperature of maximum activity is 104 °C and the optimal pH % 6.5. LamA is practically inactive at room temperature and shows detectable activity only above 30 °C [4]. The intrinsic fluorescence from LamA’s tryptophans can be used to study its structural characteristics and identify conformational states upon heat and chemical treatment [5]. The fluorescence emission spectrum of proteins depends on the microenvironment of the fluorescent amino acids. Fluorescence spectroscopy is a useful technique for studying partially folded or unfol- ded proteins; NMR and X-ray crystallography are much less practical due to the structural heterogeneity and mobility of the polypeptide chain. In the steady- state fluorescence measurements the sample is Keywords circular dichroism; hyperthermostable protein; steady-state fluorescence; time- resolved fluorescence and anisotropy Correspondence S. Koutsopoulos, Center for Biomedical Engineering, Massachusetts Institute of Technology, NE47-Room 307, 500 Technology Square, Main Street, Cambridge, MA 02139-4307, USA Fax: +1 617 258 5239 Tel: +1 617 324 7612 E-mail: sotiris@mit.edu (Received 15 July 2005, accepted 30 August 2005) doi:10.1111/j.1742-4658.2005.04941.x The structural features of the hyperthermophilic endo-b-1,3-glucanase from Pyrococcus furiosus were studied using circular dichroism, steady-state and time-resolved fluorescence spectroscopy and anisotropy. Upon heat and chemical treatment the folded and denatured states of the protein were characterized by distinguishable spectral profiles that identified a number of conformational states. The fluorescence methods showed that the spec- tral differences arose from changes in the local environment around specific tryptophan residues in the native, partially folded, partially unfolded and completely unfolded state. A structural resemblance was observed between the native protein and the structurally perturbed state which resulted after heat treatment at 110 °C. The enzyme underwent disruption of the native secondary and tertiary structure only after incubation at biologically extre- mely high temperatures (i.e. 150 °C), whilst in the presence of 8 m of guani- dine hydrochloride the protein was partially unfolded. Abbreviations ANS, 8-anilino naphthalene-1-sulfonic acid; CD, circular dichroism; GdnHCl, guanidine hydrochloride; LamA, endo-b-1,3-glucanase. 5484 FEBS Journal 272 (2005) 5484–5496 ª 2005 FEBS constantly illuminated and the emission is recorded. Time-resolved measurements are performed with expo- sure of the sample to a picosecond light pulse and recording of the intensity decay in the nanosecond timescale [6]. The fluorescence and anisotropy decays contain information on the shape, rigidity, compact- ness, fluorophore dynamics and rotational motion of the protein [6,7]. Even in the absence of structural data, valuable information about the local and global dynamics of LamA can be inferred from inspection of the fluorescence decays alone. In this study, the structural characteristics of the hyperthermostable LamA are investigated at extreme temperatures and high concentrations of guanidine hydrochloride (GdnHCl). The spectroscopic analysis will enable us to characterize the thermally and chem- ically denatured states of LamA. Using a combination of circular dichroism, steady-state and time-resolved spectroscopy and anisotropy we will show that it is possible to observe conformations of partially struc- tured, partially unfolded and completely unfolded states, depending on the treatment. Results The hyperthermophilic LamA is a single-domain protein with a molar mass of 30 085 Da. Experimental data from mass spectroscopy (MALDI TOF) and size exclu- sion chromatography showed that LamA in solution is a monomer. LamA contains 11 tryptophans homogen- ously distributed over the amino acid sequence (Fig. 1). For the graphical representation a molecular simulation, software was utilized [8] assuming structural similarity of LamA with a homologous 1,3-1,4-b-glucanase from Bacillus licheniformis and with a j-carrageenase frag- ment from Pseudoalteromonas carrageenovora whose crystal structures are known (PDB entries 1GBG and 1DYP, respectively) [9,10]. According to the model, the shape of LamA is globular-ellipsoid with calculated dimensions of 4.6 nm · 3.2 nm · 3.4 nm. For the selec- tion of the best model preliminary analysis of the NMR solution structure of LamA as well as spectroscopic data from this work were taken into consideration. Investiga- tion of proteins with multiple tryptophans results in emission spectra that represent the contribution from all emitting groups. Nevertheless, valuable information can be obtained from analyses of the conformational states of LamA upon heat treatment and in the presence of GdnHCl. At the experimental conditions employed in this work LamA shows a calorimetric transition at 109 °C which represents denaturation [11] and main- tains its structural integrity at high concentrations of GdnHCl up to 5.5 m. Circular dichroism (CD) The secondary and tertiary structural features of LamA were studied by far- and near-UV CD, respect- ively. As may be seen in Fig. 2 (top, curve a), the far- UV CD spectrum of native LamA exhibits a broad negative peak at 217 nm and a positive absorption dif- ference band from % 207 nm. This spectral profile is characteristic of proteins predominantly consisting of b-structures. The spectral analysis revealed that native LamA consists of b-sheets and turns up to % 96% (Table 1). Heating the protein solution up to 98 °C followed by cooling resulted in restoring the spectral ellipticity (spectrum coincided with curve a in the top panel of Fig. 2). However, heating at and above the Fig. 1. Graphic display of the structure of LamA using molecular modeling. The enzymatic cleft is located on the top of the structural representation. Secondary structural elements (A) and the position (B) of the tryptophans in the three-dimensional structure (graphs were generated with Swiss PDB Viewer). S. Koutsopoulos et al. CD, fluorescence and anisotropy of LamA FEBS Journal 272 (2005) 5484–5496 ª 2005 FEBS 5485 denaturation temperature (e.g. 110 °C) did not result in recovering the spectral features of the native protein (Fig. 2; top, curve b). Notably, heating at such high temperatures did not unfold the hyperthermostable protein. The features of the native state could still be observed in the denatured sample, illustrating the per- sistence of a stable network of b-structures up to % 87%. Monitoring the ellipticity at 220 nm showed the beginning of the thermal transition which indicated that at 110 °C (i.e., just above the denaturation point of the protein) residual secondary structure was still present (Fig. 2, inset). The CD spectrum of LamA at 110 °C closely resembled the one recorded for the same sample after cooling to room temperature. Heat incubation at 150 °C for 30 min resulted in collapsed secondary structure and the polypeptide chain appeared to be unordered (Fig. 2; top, curve c). Severe changes in the secondary structure were also observed in the presence of 8 m GdnHCl but the effect could not be quantified (Fig. 2; top, curve d). This finding is in contrast to a previous study where it was reported that the presence of 7.9 m GdnHCl did not alter LamA’s secondary structural characteristics [12]. In the near-UV region the differences between the native and the thermally denatured states were more noticable. The CD spectrum of native LamA shows two minima in ellipticity at % 295 nm and 265 nm. The bands arose from the aromatic residues fixed in an asymmetric environment. The CD spectrum of de- natured LamA after heating at 110 °C resembled the one of native LamA but the intensities of the bands were lower. After heat incubation at 150 °C the spec- trum of LamA had very little and no ellipticity at 295 nm and 262 nm, respectively (Fig. 2; bottom, curve c), suggesting disruption of the tertiary structure. In the presence of 8 m GdnHCl the near-UV CD spec- tral profile of LamA showed decreased ellipticity of the bands around 295 and 262 nm and increased ellip- ticity of the positive band around 285 nm (Fig. 2; bottom, curve d). These changes, although significant, strongly suggest that even at 8 m GdnHCl the protein did not completely unfold. These results are in agree- ment with data reported by Chiaraluce et al. [12]. Steady-state fluorescence spectroscopy The fluorescence emission spectra of LamA recorded after excitation at 300 nm are typical for a multitryp- tophan protein [6]. The native protein shows a maxi- mum at 335 nm (Fig. 3; curve a). After heating of the protein solution to 110 °C the maximum intensity shif- ted to 344 nm. This indicates partial exposure of tryp- tophan(s) to water, possibly due to a structural distortion. Incubation at 150 °C shifted the emission maximum to 356 nm suggesting significant exposure of tryptophans and possibly collapsed tertiary structure. Interesting features were also revealed from the -15 -10 -5 0 5 10 15 20 25 190 200 210 220 230 240 250 260 Wavelength (nm) [ θ 01 x ] 4- mc ged( 2 lom 1- ) (a) (b) (c) (d) 260 280 300 320 -0.4 -0.2 0.0 0.2 0.4 (b) (d) (c) (a) (c) (b) (d) (a) [ θ 01 x ] 4- mc ged( 2 lom 1- ) Wavelen g th (nm) -15 -10 -5 0 5 10 30 50 70 90 110 Temperature ( o C) [ θ 01 x ] 4- mc ged( 2 lom 1- ) Fig. 2. Circular dichroism of 0.25 mgÆmL )1 LamA in 10 mM sodium phosphate buffer at pH 7.0 in the far-UV (top) and near-UV (bottom) region of the spectrum. Lines represent: (a) LamA in the native form, (b) heat denatured at 110 °C, (c) after heat incubation at 150 °C, and (d) in the presence of 8 M GdnHCl. Spectra were recor- ded at 20 °C. Inset: the thermal transition of LamA monitored by the molar ellipticity at 220 nm. Table 1. Secondary structure content of LamA in 0.01 M sodium phosphate at pH 7.0 in the native state, after heat treatment and in the presence of 8 M GdnHCl as calculated from CD spectra using the program CONTIN. Sample a-helix (%) b-sheet (%) b-turn (%) Unordered (%) LamA in solution 0.5 ± 0.3 74.5 ± 2.5 21.0 ± 2.0 4.0 ± 2.0 LamA (110 °C) 4.0 ± 1.2 65.0 ± 3.0 22.0 ± 1.5 10.0 ± 4.2 LamA (150 °C) 1.0 ± 0.4 43.0 ± 1.7 17.0 ± 1.0 39.0 ± 2.0 CD, fluorescence and anisotropy of LamA S. Koutsopoulos et al . 5486 FEBS Journal 272 (2005) 5484–5496 ª 2005 FEBS respective fluorescence intensities. Heating LamA to 110 °C resulted in decreased emission. The effect of thermal treatment was more pronounced after incuba- tion at 150 °C and subsequent recording of the fluores- cence emission at 20 °C (i.e., the intensity decreased threefold as compared to that of native LamA). The emission maximum of LamA in 8 m GdnHCl was observed at 350 nm with two-fold increased intensity (Fig. 3). 8-Anilino naphthalene-1-sulfonic acid (ANS) fluorescence spectroscopy measurements Coherence and integrity of the external surface of LamA upon thermal and chemical treatment were tes- ted by measuring the exposure of the hydrophobic groups to the solvent. The fluorescence intensity of ANS is quenched in aqueous solution, but in contact with nonpolar groups a striking emission enhancement is observed [13,14]. Depending on the treatment, the interaction of LamA with ANS resulted in notable dif- ferences in the fluorescence emission of the probe (Fig. 4). Heating LamA at 110 °C resulted in 12-fold increased intensity relative to that of the native state. After incubation at 150 °C the intensity was similar to that of native LamA but the ANS emission maximum was clearly blue-shifted to 460 nm (Fig. 4; curve c) which suggests increased exposure of hydrophobic groups. In the presence of 8 m GdnHCl the ANS fluor- escence could not be measured, probably due to the interaction of ANS with the denaturant. Time-resolved fluorescence decay In an attempt to understand the origin of the differ- ences observed in the steady-state fluorescence spectra, we inspected the time-resolved profiles. The decays were best fitted by five components according to Eqn (4) (Experimental procedures), except in the case of LamA in 8 m GdnHCl where four exponents were sufficient. The lifetimes (s) and their fractional contri- butions (a) associated with the decays are summarized in Table 2. Heat and chemical treatment of LamA (Fig. 5; curves b–d) resulted in fluorescence decays that relaxed at longer lifetimes as compared to that of the native state (Fig. 5; curve a). This can also be seen in Table 2, from the increased contribution (a i ) of the longest lifetimes (s i ) on the average fluorescence life- times, <s>. LamA thermally treated at 110 °C has a dynamic fluorescence profile that clearly differs from that of the native protein. The differences are striking as compared to the information obtained from the steady-state spectra (Fig. 3; curves a and b). Compar- ison of the decays justifies the dynamic diversity of the tryptophans’ local microenvironment owing to conformational changes. Heat treatment at 110 °C and incubation at 150 °C resulted in similar decay profiles. However, inspection of the resolved parameters shows that after heat treatment at 110 °C, the short fluores- cence lifetimes, s 1 –s 3 , are shorter relative to those found for LamA after incubation at 150 °C. The pic- ture is reversed at longer lifetimes (Table 2). In the 0 20 40 60 80 100 120 410 460 510 560 Wavelen g th (nm) ).u.a( ytisnetni ecnecseroulf SNA (b) (c) (d) (a) Fig. 4. Binding of ANS to LamA before and after heat and chemical treatment. Spectra of 0.1 mgÆmL )1 LamA in sodium phosphate buf- fer at pH 7.0 in the presence of 50 l M ANS were recorded at 20 °C after excitation at 380 nm. The lines represent the ANS fluor- escence of LamA (a) in the native state, (b) thermally denatured at 110 °C, (c) incubated at 150 °C, and (d) in the presence of 8 M GdnHCl. 0 50 100 150 200 250 300 350 305 325 345 365 385 Wavenumber (nm) ).u.a( ytisnetnI noissimE ecneseroulF (a) (b) (c) (d) Fig. 3. Normalized steady-state fluorescence emission spectra of LamA in sodium phosphate buffer at pH 7.0. Curve (a) refers to the native state, (b) and (c) to LamA after heat treatment at 110 °Cand 150 °C, respectively, and (d) in the presence of 8 M GdnHCl. Spec- tra of 0.025 mgÆmL )1 LamA were recorded at 20 °C; the excitation wavelength was 300 nm. S. Koutsopoulos et al. CD, fluorescence and anisotropy of LamA FEBS Journal 272 (2005) 5484–5496 ª 2005 FEBS 5487 presence of 8 m GdnHCl the decay differs from that of native and heat-treated LamA. In this case, the data analysis showed that the shortest and the longest life- times observed in the other samples could not be resolved. Instead, the major contribution to the decay arises from tryptophans relaxing at medium and relat- ively long lifetimes. Time-resolved anisotropy decay Two exponential terms were required to describe the anisotropy decays of LamA according to Eqn (6). The fitting parameters are summarized in Table 3. The fluorescence is mainly depolarized by the rapid local motion of the tryptophans and by the overall rotation of the entire protein. The diversity of the anisotropy decays observed for each sample (Fig. 6) suggests a different depopulation mechanism of the excited state depending on the protein conformation. The aniso- tropy of native LamA decays slower relative to that after heat and chemical treatment. Data analysis revealed two rotational correlation times at / 1 ¼ 260 ps and / 2 ¼ 18.9 ns with amplitudes b 1 ¼ 0.038 and b 2 ¼ 0.122, respectively. The shortest correlation time is associated with the rapid internal flexibility of a population of indole rings, which depends on the microenvironment that the tryptophans reside in, in the protein. The longer component is relevant to the rotational diffusion of the protein from which the hydrodynamic size may be calculated using the Einstein–Stokes equation (u ¼ 4pR 3 h g=3kT; where g is the viscosity of the medium, k is the Boltzmann con- stant and T is the absolute temperature). The hydro- dynamic radius, R h , of native LamA was found to be 2.63 nm. This value is in good agreement with the pro- tein size of the model, especially if the size of the hydration layer surrounding the protein in solution is taken into account. After thermal denaturation at 110 °C, the anisotropy decay was found to be consid- erably different from that observed for the native state (Fig. 6; curves a and b). This is also shown in the short correlation time resolved at / 1 ¼ 434 ps which is longer than that observed in native LamA but which has a larger amplitude. The longest rotational correla- tion time is slightly longer than that of the native state but the difference in the calculated hydrodynamic radii does not document size expansion (Table 3). After incubation at 150 °C and in the presence of 8 m GdnHCl, the anisotropy decayed very fast and calcula- tions on the size of the protein could not be implemen- ted. In the case of incubation at 150 °C the longest correlation time, which was observed in the native state and in LamA after heating at 110 °C, could not Table 2. Calculated fluorescence decay parameters of LamA in sodium phosphate buffer solution 0.01 M at pH 7.0 upon excitation at 295 nm. Standard errors for the calculation of the decay components are given in parentheses. Sample a 1 (%) s 1 (ns) a 2 (%) s 2 (ns) a 3 (%) s 3 (ns) a 4 (%) s 4 (ns) a 5 (%) s 5 (ns) <s> (ns) LamA in solution 32.0 (± 2.6) 0.028 (± 0.004) 27.6 (± 1.3) 0.233 (± 0.034) 30.7 (± 2.8) 0.612 (± 0.046) 9.0 (± 0.8) 1.463 (± 0.053) 0.6 (± 0.3) 5.521 (± 0.091) 0.43 LamA (110 °C) 31.8 (± 2.4) 0.037 (± 0.004) 26.4 (± 2.2) 0.355 (± 0.044) 22.2 (± 2.5) 1.278 (± 0.072) 15.6 (± 1.7) 4.051 (± 0.066) 3.9 (± 0.4) 8.235 (± 0.064) 1.34 LamA (150 °C) 27.2 (± 3.0) 0.051 (± 0.006) 19.6 (± 1.4) 0.348 (± 0.037) 25.2 (± 1.9) 1.455 (± 0.040) 24.1 (± 2.6) 3.831 (± 0.035) 3.8 (± 2.5) 7.761 (± 0.083) 1.67 LamA in 8 M GdnHCl 15.4 (± 3.1) 0.282 (± 0.018) 13.4 (± 1.5) 0.712 (± 0.059) 32.2 (± 3.6) 1.737 (± 0.075) 39.0 (± 2.7) 3.981 (± 0.062) – – 2.25 CD, fluorescence and anisotropy of LamA S. Koutsopoulos et al . 5488 FEBS Journal 272 (2005) 5484–5496 ª 2005 FEBS be resolved. Instead, a medium-lived component was found at 3.8 ns. It should be noted that the long rota- tional correlation time observed for LamA in the presence of 8 m GdnHCl should be corrected by a factor 2.3 when compared to the respective lifetimes of LamA in guanidine-free solutions [15]. This is due to the difference between the viscosity of the solution in the presence and in the absence of 8 m GdnHCl, i.e., Table 3. Anisotropy decay parameters of LamA in sodium phosphate buffer solution 0.01 M at pH 7.0. Values in parentheses are the range at 67% confidence intervals. ND, not determined. Sample b 1 / 1 (ns) b 2 / 2 (ns) R h (nm) r o h LamA in solution 0.038 (0.027–0.048) 0.26 (0.24–0.28) 0.122 (0.111–0.132) 18.90 (17.97–19.84) 2.63 (2.59–2.68) 0.16 23.5° (21.2–24.9) LamA (110 °C) 0.054 (0.043–0.067) 0.43 (0.41–0.46) 0.071 (0.061–0.084) 19.39 (18.64–20.14) 2.66 (2.62–2.69) 0.13 32.4° (31.7–32.9) LamA (150 °C) 0.098 (0.092–0.115) 0.34 (0.32–0.36) 0.062 (0.054–0.075) 3.83 (3.67–3.98) – 0.16 39.7° (39.4–40.4) LamA in 8 M GdnHCl 0.095 (0.063–0.120) 2.14 (1.93–2.34) 0.016 (0.003–ND) 17.96 (12.32–23.59) – 0.11 49.0° (ND)52.9) 1 10 100 1000 10000 100000 1000000 0 5 10 15 20 25 30 Time (ns) ytisnetnI ecnecseroulF (a) (b) (c) (d) -0,04 0 0,04 0102030 -0,04 0 0,04 0102030 -0,04 0 0,04 0102030 -0,04 0 0,04 Residuals 0102030 -0,3 0 0,3 0102030 -0,3 0 0,3 0102030 -0,3 0 0,3 0102030 -0,3 0 0,3 0102030 (a) (c) (b) (d) (a) (c) (b) (d) Autocorrelation Fig. 5. Time-resolved fluorescence decay of 0.25 mgÆmL )1 LamA in sodium phosphate buffer pH 7.0 at 20 °C on excitation at 295 nm. The y axis is in a logarithmic scale. The lines represent the fluorescence decay of (a) LamA in the native state, (b) heated at 110 °C, (c) incubated at 150 °C, and (d) in the presence of 8 M GdnHCl. S. Koutsopoulos et al. CD, fluorescence and anisotropy of LamA FEBS Journal 272 (2005) 5484–5496 ª 2005 FEBS 5489 2.39 cP and 1.02 cP at 20 °C, respectively [16]. Short life- times are not affected by the viscosity of the medium. The fundamental anisotropy, r o , representing the total anisotropy in the absence of rotation (at t ¼ 0), is equal to the sum of the amplitudes, b i , of the fluoro- phores. For excitation at 295 nm the theoretical time- zero anisotropy is about 0.3 [6]. This is higher than the values obtained for the tryptophans in LamA. Depend- ing on the protein conformation and the freedom of the tryptophans to rotate in the protein matrix, the r o may be reduced as a result of subpicosecond motions that are too fast to be detected [17,18], noncollinearity of the absorption and emission dipoles [6,18,19], and intertryptophan energy migration [20,21]. The rotation angle, h, of the tryptophans attached in the protein backbone may be calculated from the amplitude b 1 of the fast motion [7]: 1 À b fast r o ¼ 3 cos 2 h À 1 2 ð1Þ The average cones of rotation of the tryptophans in LamA (Table 3) increase from 23.5° in the native state to 32.4° in the thermally denatured state, to 39.7° in the unfolded state after incubation at 150 °C, to 49.0° in the perturbed conformation in the presence of 8 m GdnHCl. The increase of the rotational freedom in the heat and chemically treated LamA illustrates the fast anisotropy decays observed in Fig. 6. 0,00 0,05 0,10 0,15 0,20 0 5 10 15 20 25 30 Time (ns) yportosinA (a) (b) (c) (d) -10 0 10 0102030 -10 0 10 0102030 -10 0 10 0102030 -10 0 10 0102030 -0,2 0 0,2 0102030 -0,2 0 0,2 0102030 -0,2 0 0,2 0102030 -0,2 0 0,2 0102030 (a) (c) (b) (d) (a) (c) (b) (d) Residuals Autocorrelation Fig. 6. Time-resolved anisotropy of 0.25 mgÆmL )1 LamA in sodium phosphate buffer at pH 7.0 at 20 °C. The decays represent (a) native LamA, (b) LamA heat treated at 110 °C, (c) after incubation at 150 °C, and (d) in the presence of 8 M GdnHCl. The lines represent fitting of the anisotropy exponential decay with two components as shown in Table 3. CD, fluorescence and anisotropy of LamA S. Koutsopoulos et al . 5490 FEBS Journal 272 (2005) 5484–5496 ª 2005 FEBS Discussion The hyperthermostable LamA shows a heat denatura- tion transition at 109 °C and only a partly unfolded structure at 7.9 m GdnHCl [11,12]. In this study, the structural characteristics of LamA were thoroughly investigated upon thermal and chemical treatment. The spectroscopic data suggested different conformations depending on the temperature of the treatment. It was shown that after cooling to room temperature, the thermally denatured LamA did not refold to the native conformation but to a compact form with defined structure that is different from that of the native and of the chemically denatured states. Such a conformation resembles the features of a molten globule exhibiting native-like secondary structure but different tertiary structure. LamA’s irreversible denaturation is con- firmed by calorimetric experiments (S. Koutsopoulos, J. van der Oost & W. Norde, unpublished data). Both the secondary and tertiary structures irreversibly col- lapsed only after prolonged heating at 150 °C. The interaction of LamA with GdnHCl solutions did not show significant changes in the spectroscopic character- istics of the protein up to % 5.5 m GdnHCl. Severe changes in LamA’s secondary and tertiary structure were observed in the presence of 8 m GdnHCl. Inspection of the far-UV CD spectra showed minor changes in the secondary structure of LamA upon heat denaturation at 110 °C while significant changes were observed upon incubation at 150 °C and in the pres- ence of 8 m GdnHCl. Moreover, the near-UV bands at 262 nm and 295 nm of LamA were notably decreased upon heat treatment at 110 °C, and significantly sup- pressed after heat incubation at 150 °C or in the pres- ence of 8 m GdnHCl. The intensities of the bands decrease when aromatic residues become more distant from each other due to loose structure. The conclusions drawn from fluorescence spectro- scopy are in line with CD. The fluorescence emission of native LamA showed maximum intensity at 335 nm indicating moderate interaction of the tryptophans with the solvent. The emission profile of thermally denatured LamA at 110 °C suggests that cooling to room temperature did not result in refolding to the native conformation but rather to a native-like form. The red-shift in the emission maximum indicates increased tryptophan exposure. It should be noted that when the emission maxima are correlated with tryptophan exposure to water, the interaction often originates from penetration of water molecules into the interior of the protein. This is true especially for struc- tural distortions induced by heat treatment. In such cases, the red-shift of the emission originates from larger accessibility of tryptophans to both internal and external water. The red-shift to 356 nm, which was observed upon heat incubation at 150 °C, suggests protein unfolding. In the presence of 8 m GdnHCl the emission maximum was observed at 350 nm which is red-shifted as compared to the respective maximum of the native state, but not as much as that of the unfolded protein. Analysis of the fluorescence properties of multitryp- tophan proteins is a difficult task even when the struc- ture is known. The emission spectrum represents the average of local quenching and complicated resonance energy transfer phenomena. Apart from the influence of the polar solvent, which decreases the fluorescence emission of exposed tryptophans, in the protein matrix tryptophans can be quenched by neighboring carboxyl groups, histidine, methionine, phenylalanine, lysine, etc. [22]. Energy transfer from one tryptophan to another tryptophan or to a tyrosine decreases the fluorescence quantum yield of the donor [23]. LamA has a single cysteine that is likely to play a critical role as sulfhydryl groups are notorious quenchers of the proximal tryptophans [24]. After thermal denaturation at 110 °C, the fluorescence intensity moderately decreased while incubation at 150 °C resulted in sub- stantial three-fold decreased emission. This observation and the red-shift of the emission maximum at 356 nm suggest that in this conformation, the tryptophans are quenched, possibly due to contact with water. The residual intensity may imply that even in the case of an extensively hydrated unstructured backbone it is possible that tryptophan(s) belong to a locally struc- tured domain. The twofold increase of the fluorescence intensity in the presence of 8 m GdnHCl probably ori- ginates from relocation of tryptophans in the three dimensional structure of the protein. In the new posi- tions the interactions of the tryptophans with quench- ing groups are weaker and ⁄ or the intertryptophan distances are longer than that required for energy transfer [25,26]. Both mechanisms increase the fluores- cence quantum yield, which overwhelms the quenching effect of the solvent-exposed tryptophans. Hence, from both the emission maximum and the fluorescence intensity it is concluded that even at 8 m GdnHCl there is a residual structure in LamA that involves buried tryptophan residue(s). Notable differences in LamA before and after ther- mal and chemical treatment were also observed upon interaction with ANS (Fig. 4). The heat-denatured state is probably characterized by a structural distor- tion from which dissolved ANS accessed hydrophobic groups that were previously located in the interior of the protein [27]. This interaction led to a significantly S. Koutsopoulos et al. CD, fluorescence and anisotropy of LamA FEBS Journal 272 (2005) 5484–5496 ª 2005 FEBS 5491 increased intensity. The unfolded state upon incuba- tion of LamA at 150 °C was justified by the blue-shift of the ANS emission maximum. Time-resolved fluorescence gave insight into the tryptophans’ relaxation dynamics. Conformational changes were justified by simple inspection of the fluor- escence decays. The fluorescence of the heat- and chem- ically-treated LamA decayed at longer lifetimes. This is typical for proteins with solvent exposed tryptophans [28]. Each of the five (four in the presence of 8 m GdnHCl), lifetimes, s i , resolved represents a class of tryptophans in a specific microenvironment [29–31], and the respective pre-exponentials, a i , are related to the fraction of tryptophans in each class [28,32,33]. In native LamA the extremely short lifetime (28 ps), which accounts for one third of the total fluorescence intensity, can be assigned to very efficient energy trans- fer or to strong static quenching from amino acid(s) (e.g. cysteine) in the vicinity of the emitting trypto- phans. The amplitude of the longest lifetime, s 5 ,in native LamA at 5.5 ns probably represents water- exposed tryptophans and contributes very little to the total fluorescence. The picture is reversed after heat and chemical treatment, where the contribution from the longest lifetimes is significantly increased. After heating at 110 °C the tryptophans, character- ized by extremely short-lived relaxation in the native state, were now decayed at a slightly longer lifetime (37 ps). Notably, the amplitude, a 1, of the tryptophans emitting at the shortest lifetime is similar to that resolved for native LamA. The amplitude for trypto- phans that decay at longer lifetimes was markedly increased, which suggests that the slightly exposed tryp- tophans of the native protein become more exposed in the molten globule and therefore more solvent- quenched. Studies in helical peptides and in small b-structured proteins show that the amplitudes for each decay com- ponent vary with the secondary structure [34,35]. The fluorescence from tryptophans belonging to an exten- ded b-conformation decays with significant contribu- tion from intermediate lifetimes. This is the case for native LamA. Interestingly, the apparent increase of a-helices and the decrease of sheets and strands upon heat treatment at 110 °C and in the presence of 8 m GdnHCl, as evidenced by far-UV CD, were confirmed by the time-resolved fluorescence measurements: the increased pre-exponential term a 5 of the longest decay time and the decreased contribution of the intermediate components (i.e., a 2 –a 4 for the native and a 2 –a 3 for the heat and chemically treated LamA) sug- gest decreased b-structures and increased helical con- tent [34]. The contribution, a i , of the longest lifetime compo- nents (s i > 3.8 ns) to the total fluorescence signifi- cantly increased from the native LamA to the thermally denatured at 110 °C LamA, to the heat unfolded LamA, to the chemically treated partially unfolded protein. This order is consistent with the increased solvent exposure of the tryptophans in the heat-treated samples as observed in the steady-state fluorescence spectra. There is a deviation from the order in the case of LamA in the presence of 8 m GdnHCl (Fig. 5). This could be due to the significant contribution from completely exposed tryptophans (s 5 > 7 ns) of the heat unfolded protein that is absent in the GdnHCl partially unfolded LamA. However, we should bear in mind that steady-state measurements provide an intensity-weighted, time-averaged descrip- tion of the fluorophore emission and, hence, are pro- portional not to the most populated state but to the state that emits most. This trait of steady-state emis- sion and the fact that specific interactions may elude time-resolved fluorescence detection and, thus, conceal a part of the interpretation are additional reasons for the discrepancy. An analysis of the time-resolved anisotropy in terms of protein conformer-lifetime assignments was also attempted. The rapidly relaxing component in native LamA, / 1 , can be ascribed to flexibility of the indole ring in the protein matrix or other local dynamic events of the tryptophans which cause very fast de- polarization. Upon heat and chemical treatment, the tryptophans rotate more freely as a result of rearrange- ments in the protein matrix around the fluorophore(s). This is shown in the fractional contribution b 1 of the short correlation time and the calculated rotation angle of the tryptophans in Table 2. The presence of many tryptophans distributed over the protein backbone is advantageous for the calcula- tion of the rotational diffusion of a protein in solution. The rotational properties depend on the orientation of the dipoles relative to the main symmetry axis and, hence, a large number of fluorophores ensures that all orientations are sampled and the pristine rotational correlation time is determined by the anisotropy decay [36]. After thermal denaturation at 110 °C, the long lived component slightly increased to 19.39 ns. In the completely and partially unfolded states the intra- molecular interactions and internal structural con- straints are loosened or lost and, hence, large parts of the polypeptide chain become solvent exposed. Therefore, the rotational freedom of the tryptophans substantially increases and the system loses anisotropy much faster (Fig. 6; curves c and d). In these cases, the size of LamA could not be determined from the CD, fluorescence and anisotropy of LamA S. Koutsopoulos et al . 5492 FEBS Journal 272 (2005) 5484–5496 ª 2005 FEBS parameters recovered due to hydration of internal pro- tein segments resulting in largely expanded conforma- tions. The medium correlation time of 3.8 ns that was observed in the completely unfolded LamA corres- ponds to tryptophans trapped locally that just lose anisotropy faster than the tryptophans in the native state (Table 3). The medium-lived component (2.1 ns) observed in the guanidine-treated partially unfolded LamA emerged at the expense of the shortest pico- second correlation lifetime. Motions with correlation times ranging from 1 to 3 ns describe segmental back- bone fluctuations of the polypeptide chain [37,38]. These motions are important when the protein integ- rity is disrupted and the protein backbone is solvated and more flexible. Data from CD, fluorescence spectroscopy (steady- state, time-resolved and ANS binding), and anisotropy were used to probe conformational features of LamA before and after heat or chemical treatment. It was suggested that upon heating at 110 °C, the local micro- environment of the tryptophans resembles but it is not identical to that of the native state. It is likely that this state represents a structurally disturbed or locally unfolded state rather than completely unfolded. The structural elements may be maintained by a mechan- ism involving specific local and long-range interactions, some of which are native-like [39–43]. The interaction of LamA with 8 m GdnHCl resulted in significant structural changes but not in complete unfolding. The protein was partially unfolded with characteristics clearly distinct from the completely unfolded confor- mation obtained after incubation at 150 °C. Experimental procedures Purification of LamA, treatment and chemicals The gene encoding LamA (sequence deposited in GenBank: accession No. AF013169) was isolated from P. furiosus and after cloning it was overexpressed in Escherichia coli BL21 (DE3) using the T7 expression system [3]. Further purifica- tion was achieved by size exclusion chromatography in a Superdex 200 column (Amersham Pharmacia, Piscataway, NJ, USA). Pure LamA was stored at 4 °C in 0.01 m sodium phosphate buffer at pH 7.0. The protein concentra- tion was routinely determined by the absorption at 280 nm. Controlled heat treatment of LamA was carried out in a VP-DSC calorimeter (MicroCal Inc., Northampton, MA, USA). The heating rate was 1 °CÆmin )1 and after reaching 110 °C the sample was allowed to cool down to room tem- perature and used for further analyses. Heat incubation for 30 min at 150 °C was performed in a temperature con- trolled oil bath using thick-walled glass tubes with a lid capable of withstanding the vapor pressure of water. Chem- ical denaturation was studied in the presence of extra pure fluorescence-free GdnHCl (Merck, Rahway, NJ, USA). The GdnHCl solutions were prepared according to Pace et al. [44] and the concentration was determined by measuring their refractive index. LamA was allowed to interact with GdnHCl overnight at 20 °C. Circular dichroism Far- and near-UV CD spectra of 0.25 mgÆmL )1 LamA in 1 mm and 1 cm quartz cuvettes, respectively, were recorded in a JASCO J-715 (Tokyo, Japan) spectrophotometer equipped with a temperature controller (JASCO PTC 348 WI) which was set at 20 °C. Measurements were also performed at temperatures up to 110 °C in a closed metal-caged quartz cuvette under pressure to prevent eva- poration of water. The CD spectra referring to LamA after heat incubation at 150 °C were obtained from sam- ples which had been previously heated and then cooled to room temperature. The spectrophotometer was calibrated with a standard ammonium D-10-camphorsulphonate solution. The scan rate was 100 nmÆmin )1 , with 0.1 nm resolution, and 0.25 s response time. Spectra of LamA before and after heat or chemical treatment resulted from accumulation of 32 scans that were subsequently aver- aged. Blank spectra of buffer without protein, obtained at identical conditions, were subtracted. Data analysis was performed by fitting the acquired spectra with reference spectra using the contin program, which is based on nonlinear regression fitting algorithms without constraints (ridge-regression analysis) [45,46]. This program gives a much better estimate of b-sheets and turns than simple multiple linear regression [47]. An average molar mass of 115 Da per amino acid residue was used for calculating the ellipticity, h. Steady-state fluorescence spectroscopy Fluorescence emission was measured by a Varian Cary Eclipse spectrophotometer (Variam, Palo Alto, CA, USA). Unless otherwise indicated, all measurements were carried out at 20 °C using quartz cuvettes of 1 cm path length. Emission spectra of 0.025 mgÆmL )1 LamA were recorded in the range 300–400 nm on excitation at 300 nm. The excita- tion and emission slit widths were set at 5.0 and 2.5 nm, respectively. All spectra were corrected for the background emission of water. Spectra of samples containing GdnHCl were corrected using as reference the buffer solution with the same concentration of GdnHCl. Binding of ANS was studied between 400 and 600 nm on excitation at 380 nm. Fluorescence spectra of 0.1 mgÆmL )1 LamA in the presence of 50 lm of ANS were recorded at 20 °C before and after heat and chemical treatment. S. Koutsopoulos et al. CD, fluorescence and anisotropy of LamA FEBS Journal 272 (2005) 5484–5496 ª 2005 FEBS 5493 [...]... water was also tested for artificial luminescence For obtaining a dynamic instrumental response of the setup, the single-exponential fluorescence decay of paraterphenyl was measured in a mixture of cyclohexane and CCl4 in a 50:50% (v ⁄ v) ratio To avoid and correct eventual temporal shifts experimental data consisted of repeating sequences of measurements of the parallel and perpendicular polarized emission... total anisotropy decay The anisotropy decay parameters were estimated by simultaneous analysis of I||(t) and I^(t) The goodness -of- fit was judged by the value of parameter, v2, and inspection of the residual function graphs for each data set In all cases the v2 factor was close to unity, and the weighted residuals and the autocorrelation of the residuals were uniformly distributed around zero indicating... Crystal structure of Bacillus licheniformis 1,3–1, ˚ 4-b-d-glucan 4-glucanohydrolase at 1.8 A resolution FEBS Lett 374, 221–224 10 Michel G, Chantalat L, Duee E, Barbeyron T, Henrissat B, Kloareg B & Dideberg O (2001) The kappa-carrageenase of P carrageenovora features a tunnel-shaped active site: a novel insight in the evolution of clan-B glycoside hydrolases Structure 9, 513–525 11 Koutsopoulos S, van... Main chain and side chain dynamics of peptides in liquid solution from C-13 NMR: Melittin as a model peptide Biochemistry 36, 1678–1688 39 Wong KB, Freund SM & Fersht AR (1996) Cold denaturation of barstar: H-1, N-15 and C-13 NMR assignment and characterisation of residual structure J Mol Biol 259, 805–818 40 Mok YK, Kay CM, Kay LE & Forman-Kay JD (1999) NOE data demonstrating a compact unfolded state... 7228–7238 Kawahara K & Tanford C (1966) Viscosity and density of aqueous solutions of urea and guanidine hydrochloride J Biol Chem 241, 3228–3232 Ruggiero AJ, Todd DC & Flemming GR (1990) Subpicosecond fluorescence anisotropy studies of tryptophan in water J Am Chem Soc 112, 1003–1014 Callis PR (1997) L-1 (a) and L-1(b) transitions of tryptophan: Applications of theory and experimental observations to... Crowhurst KA, Tollinger M & Forman-Kay JD (2002) Cooperative interactions and a non-native buried Trp in the unfolded state of an SH3 domain J Mol Biol 322, 163–178 44 Pace CN, Shirley BA & Thomson JA (1990) Protein Structure A Practical Approach (Creighton TE, ed.), chapter 13 IRL Press, Oxford 45 Venyaminov SYu, Baikalov IA, Shen ZM, Wu C-SC & Yang JT (1993) Circular dichroic analysis of denatured proteins... fluorescence of proteins Methods Enzymol 278, 113–150 Yamamoto Y & Tanaka J (1972) Polarized absorptionspectra of crystals of indole and its related compounds Bull Chem Soc Jpn 45, 1362–1366 Turoverov KK, Biktashev AG, Khaitlina SYu & Kuznetzova IM (1999) The structure and dynamics of partially folded actin Biochemistry 38, 6261–6269 Eftink MR, Wasylewsky Z & Chiron CA (1987) Phase-resolved spectral measurements... the sample housing a Glan-laser polarizer was mounted, optimizing the already vertical polarization of the input light beam Between sample and photomultiplier a single fast lens, an interference filter (348.8 nm, Dk ¼ 5.4 nm), a computer controlled rotating sheet-type polarizer and a second single fast lens were placed, focusing the fluorescence on the photomultiplier cathode All polarizers were carefully... Biophys Acta 140, 37–44 Steinberg IZ (1971) Long-range nonradiative transfer of electronic excitation energy in proteins and polypeptides Annu Rev Biochem 40, 83–114 Griep MA & McHenry CS (1990) Dissociation of the DNA polymerase-III holoenzyme beta-2 subunits is accompanied by conformational change at distal cysteines-333 J Biol Chem 265, 20356–20363 Ewbank JJ, Creighton TE, Hayer-Hartl MK & Ulrich Hartl... with a pulse picker which decreased the repetition rate of the excitation pulses to 3.8 · 106 pulses per second The maximum pulse energy was a few pJ, the emission wavelength 295 nm and the pulse duration 3 ps The fluorescence was collected at an angle of 90° with respect to the direction of the excitation light beam Extreme care was taken to avoid artefacts from depolarization effects At the front of . conformation. The aniso- tropy of native LamA decays slower relative to that after heat and chemical treatment. Data analysis revealed two rotational correlation. from all emitting groups. Nevertheless, valuable information can be obtained from analyses of the conformational states of LamA upon heat treatment and