U c F a b a A R R 2 A A K E H U O A S 1 h i h ( w p f T t B r p [ A s l c c 0 h International Journal of Biological Macromolecules 52 (2013) 340– 348 Contents lists available at SciVerse ScienceDirect International Journal of Biological Macromolecules jo u rn al hom epa ge: www.elsev ier .com/ locate / i jb iomac rea-induced unfolding of Glossoscolex paulistus hemoglobin, in oxy- and yanomet-forms: A dissociation model rancisco A.O. Carvalhoa, José Wilson P. Carvalhoa, Patrícia S. Santiagoa,b, Marcel Tabaka,∗ Instituto de Química de São Carlos, Universidade de São Paulo, SP, Brazil Universidade Estadual Paulista “Julio de Mesquita Filho”, Registro, São Paulo, SP, Brazil r t i c l e i n f o rticle history: eceived 30 July 2012 eceived in revised form 2 September 2012 ccepted 25 September 2012 vailable online 2 October 2012 eywords: xtracellular hemoglobin a b s t r a c t The urea effect on the giant extracellular hemoglobin of Glossoscolex paulistus (HbGp) stability was stud- ied by analytical ultracentrifugation (AUC) and small angle X-ray scattering (SAXS). AUC data show that the sedimentation coefficient distributions curves c (S), at 1.0 mol/L of urea, display a single peak at 57 S, associated to the undissociated protein. The increase in urea concentration, up to 4.0 mol/L, induces the appearance of smaller species, due to oligomeric dissociation. The sedimentation coefficients and molec- ular masses are 9.2 S and 204 kDa for the dodecamer (abcd)3, 5.5 S and 69 kDa for the tetramer (abcd), 4.1 S and 52 kDa for the trimer (abc) and 2.0 S and 17 kDa for the monomer d, respectively. SAXS data show initially a decrease in the I(0) values due to the oligomeric dissociation, and then, above 4.0 mol/L of bGp rea ligomeric dissociation UC AXS denaturant, for oxy-HbGp, and above 6.0 mol/L for cyanomet-HbGp, an increase in the maximum dimen- sion and gyration radius is observed, due to the unfolding process. According to AUC and SAXS data the HbGp unfolding is described by two phases: the first one, at low urea concentration, below 4.0 mol/L, characterizes the oligomeric dissociation, while the second one, at higher urea concentration, is associ- ated to the unfolding of dissociated species. Our results are complementary to a recent report based on spectroscopic observations. . Introduction Giant extracellular hemoglobins, also known as erythrocruorins, ave been investigated as a model of extreme complex- ty in oxygen-binding heme proteins [1,2]. These extracellular emoglobins are characterized by a very high molecular mass MM), a high resistance to oxidation and a high oligomeric stability hen subjected to conditions of stress such as high temperature, H variation, and addition of chemical agents, such as urea and sur- actants [3–6], as compared, for instance, with human hemoglobin. hese properties make them an interesting and important sys- em for investigation [2,7], including in biomedical applications. esides, a strong motivation to study these giant hemoglobins is elated to their potential use as blood substitutes. Studies have been erformed in the past for Lumbricus terrestris hemoglobin (HbLt 8]), and are presently underway to test and validate the use of renicola marina hemoglobin (HbAm) in this direction [9,10]. They eem to be very promising due to the lack of undesirable immuno- ogical reactions in tests with animals, explained by the absence of ell membranes as occurs with human hemoglobin in red blood ells [9,10]. Besides, the resistance to oxidation of extracellular ∗ Corresponding author. Fax: +55 16 3373 9982. E-mail address: marcel@sc.usp.br (M. Tabak). 141-8130 © 2012 Elsevier B.V. ttp://dx.doi.org/10.1016/j.ijbiomac.2012.09.023 Open access under the Elsevier OA license. © 2012 Elsevier B.V. hemoglobins, as noticed by their high redox stability, is also an advantage as compared to the use of human hemoglobin in this medical application. The extracellular hemogobin of Glossoscolex paulistus (HbGp) has a molecular mass of 3.6 MDa, determined recently by analyti- cal ultracentrifugation (AUC, [11]). This protein has an oligomeric structure, similar to the orthologous HbLt [12], composed by 144 globin chains, and 36 additional chains lacking the heme group, named linkers. Mass spectrometry and AUC studies suggest that HbGp has the same stoichiometry as HbLt, based on the Vino- gradov model, which assumes that the whole protein is composed by twelve protomers, constituted by a dodecamer of globin chains and a trimer of linkers, [(abcd)3L3] [1,4,13]. Here a, b, c and d are globin chains forming an asymmetric tetramer (abcd), composed of a disulfide bonded trimer (abc) and a monomeric subunit d. Three linker chains, L1, L2 and L3 complete the native protomer structure. HbLt presents a high structural similarity as compared to HbGp, and the identity between the monomeric subunits d is 57%. The HbLt oxygenation process was investigated both for the whole oligomeric structure as well as for several structural parts of the protein in different conditions of pH, presence of salts and with the Open access under the Elsevier OA license. pure isolated subunits [14]. According to the authors, the tetramer (abcd) has a high oxygen affinity, indistinguishable from the whole HbLt, at pH 6.8. However, the oxygen equilibrium of the trimer (abc) is characterized by a very low cooperativity and a maximum dx.doi.org/10.1016/j.ijbiomac.2012.09.023 http://www.sciencedirect.com/science/journal/01418130 http://www.elsevier.com/locate/ijbiomac mailto:marcel@sc.usp.br dx.doi.org/10.1016/j.ijbiomac.2012.09.023 http://www.elsevier.com/open-access/userlicense/1.0/ http://www.elsevier.com/open-access/userlicense/1.0/ of Bio H w c g s d t b s t [ g 3 u s t T c o s o c d s ( c s p t u t 2 2 R p t w 4 0 S [ r u fi 1 t w h s n ( a w c ε 2 X F.A.O. Carvalho et al. / International Journal ill coefficient of 1.35 [14], significantly lower as compared to the hole protein, while the isolated subunits a, b and c do not present ooperativity in oxygen binding. Overall, these studies for HbLt sug- est that (abcd)2 is the primary cooperative unit in the oligomeric tructure [14]. AUC studies of HbGp, at pH 10.0, show that the whole oligomer issociates into monomer subunit d, dimer of monomers d, d2, rimer abc, and tetramer abcd. However, for cyanomet-HbGp, esides the appearance of these species, some fraction of undis- ociated whole protein is observed (17%), due to the fact that he cyanomet- form is more stable than oxy-HbGp, at pH 10.0 4]. Spectroscopic studies of HbGp in the presence of urea sug- ested that urea induces, at low concentrations, between 1.0 and .0 mol/L, the oligomeric dissociation, and further increase in rea concentration promotes the denaturation of the dissociated pecies [6]. Cyanomet- and oxy-HbGp forms are more stable owards denaturation, in the presence of urea, than met-HbGp. hus the order of stability in the presence of urea is given by: yanomet- > oxy > met-HbGp [6]. Assuming a high similarity of the denaturation process for both xy- and cyanomet-HbGp forms, described in the previous spectro- copic study, the focus of the present work is the characterization f the different species that exist in the solution, at different urea oncentrations, and the evaluation of the global HbGp oligomeric issociation model, in the presence of this denaturant agent. This tudy was performed using AUC and small angle X-ray scattering SAXS). The additional SAXS and AUC data, obtained in this work, ontribute to elucidate the species involvement and the oligomeric tructure of this system and the differences between the unfolding rocesses of the oxy- and cyanomet-HbGp forms. Previous spec- roscopic studies were able to give only a general view of the HbGp rea-induced unfolding, in three forms, and the species in the solu- ion were not characterized. . Materials and methods .1. Protein extraction and purification G. paulistus annelid is prevalent in sites near Piracicaba and io Claro cities in the state of São Paulo, Brazil. The HbGp was repared using freshly drawn blood from the worms. HbGp solu- ion was centrifuged at 2500 rpm for 15 min, at 4 ◦C. The sample as filtered (Mw cut-off 30 kDa) and centrifuged at 250,000 × g, at ◦C, for 3 h. The pellet was resuspended in a minimum amount of .1 mol/L Tris–HCl buffer, at pH 7.0. Chromatography at pH 7.0 in a ephadex G-200 column gave the samples used in the experiments 6,15–18]. All concentrations were determined spectrophotomet- ically in a UV-1601 PC spectrophotometer (Shimadzu, Japan), sing the appropriate molar absorption coefficients [6,16]. The nal protein concentration in our stock solution was in the range 5–40 mg/mL, in Tris–HCl 0.1 mol/L buffer, pH 7.0. To obtain he cyanomet-HbGp form, the met-HbGp was further incubated ith a 5-fold molar excess of potassium cyanide relative to the eme. Excess of reagents was removed by dialysis against the ame buffer for 3 h. The cyanomet-HbGp has a CN-ligand coordi- ated in the sixth coordination to the iron in the oxidized state Fe3+), while the oxy-HbGp has a molecule of O2 as the sixth lig- nd with the iron in the reduced state (Fe2+). All concentrations ere determined spectrophotometrically using the absorption oefficients ε415nm = 5.5 ± 0.8 (mg/mL)−1cm−1 for oxy-HbGp and 420nm = 4.8 ± 0.5 (mg/mL)−1cm−1 for cyanomet-HbGp [6,15]. .2. AUC experiments AUC experiments were performed in a Beckman Optima L-A analytical ultracentrifuge. Sedimentation velocity (SV) logical Macromolecules 52 (2013) 340– 348 341 experiments were carried out at oxy-HbGp concentrations from 100 up to 300 �g/mL, in 100 mmol/L Tris–HCl, pH 7.0, containing 50 mmol/L NaCl. The samples were dialyzed against the same buffer, and the dialysate was used as reference solution in all experiments. The samples were exposed to 0.0–6.0 mol/L of urea 2 h before the experiments. In each unfolding titration experiment the appropriate volume of buffer and stock urea were mixed to achieve the desired urea concentration (0.0–6.0 mol/L) in a final volume of 1.0 mL, prior to the addition of a constant aliquot of concentrated protein solution, to give the desired final protein concentration. The urea concentration at 25 ◦C was estimated by the empirical formula suggested by Pace [19]: [Urea] = 117.66(�n) + 29.753(�n)2 + 185.56(�n)3 (1) where �n is the difference between the refractive index of the urea solution and of the buffer. The SV experiments were performed at 20 ◦C, using a rotor speed between 15,000 and 40,000 rpm (An60Ti rotor), and the scan data acquisition was measured at 236 and 416 nm. Absorbance data were collected using a radial step size of 0.003 cm. The sedimenta- tion coefficient value is affected by temperature, viscosity (�) and density (�), so we calculated the standard sedimentation coeffi- cient at infinite dilution (0 mg/mL, s0 20,w). The Sednterp software was used to estimate the � and � values for each urea concen- tration [20,21]. The software SEDFIT (Version 12.52) [22,23] was applied in order to fit the absorbance versus cell radius raw data. This new version of software allows to estimate directly the val- ues of s20,w , presenting it as the x-axis in the c (S) distribution. The sedimentation coefficients (s20,w) were found as the maximum of the peaks of the c (S) curves and the s0 20,w were estimated by linear regression of the s20,w values measured at each concentra- tion. Sedimentation equilibrium (SE) experiments were carried out for the oxy-HbGp at 5.0 mol/L of urea, at protein concentrations in the range 100–300 �g/mL, in 100 mmol/L Tris–HCl, pH 7.0, contain- ing 50 mmol/L NaCl. The SE experiments were performed at 20 ◦C, using speeds of 15,000, 20,000 and 25,000 rpm with the An60Ti rotor, and the data were collected at 415 nm. 2.2.1. AUC data analysis Sedimentation coefficient distribution function, c (S), was obtained by using SEDFIT program (Version 12.52) and the “Con- tinuous c (S) Distribution” model. This software models the Lamm equation so as to discriminate the spreading of the sedimenta- tion boundary from diffusion [22,23]. The c (S) distributions were obtained, keeping fixed the Vbar at 0.733 mL/g, and leaving the frictional ratio, f/f0, as the regularization parameter of the fitting process. The viscosity (�) and density (�) for each urea concen- tration were fixed. Besides, the distribution regularization method used was based on the “maximum entropy method” with a con- fidence interval P = 0.85. The average value of rmsd in analyses of c (S) curves was 0.008 ± 0.002, while the worst value found was 0.024. The analyses of MM of the species present in solution in the SV experiments were made from the global fitting with the “Species Analysis” model of the SEDPHAT program (version 9.4). All param- eters were allowed to float freely and then the statistical analyses were performed. The statistic method used was the “Monte–Carlo non-linear regression” with, at least, 200 iterations and a confi- dence level of 0.68. The SE data analyses were also made globally with all sam- ples (100, 200 and 300 �g/mL) and all speeds (15,000, 20,000 and 25,000 rpm), in the presence of 5.0 mol/L of urea, using the SED- PHAT program [24,25]. As initial input data, the MM determined by mass spectrometry [13] and the s-value determined by SV analysis [4] were used, and both parameters were allowed to float freely. 3 of Bio A t S l 2 0 7 0 t e o a B r w o c m h p q l 2 e t s m t w T s c G a a I t H f n f p o s o t p c m b p p t t 42 F.A.O. Carvalho et al. / International Journal fter the global fitting, statistical analyses were performed using he “Monte–Carlo non-linear regression” method provided by the EDPHAT with, at least, 200 iterations, and the same confidence evel of the SV experiments. .3. SAXS experiments The protein concentrations used in these experiments were .5 and 3.0 mg/mL, in 30 mmol/L acetate–phosphate buffer, at pH .0, in the presence of urea, in the concentration range from .0 to 8.0 mol/L. Oxy-HbGp, cyanomet-HbGp, and pure buffer (in he absence and presence of different urea concentrations) were xposed for 90 s and the scattered X-rays intensity was monitored n an image plate detector. SAXS experiments were performed t the National Synchrontron Light Laboratory (LNLS), Campinas, razil, using a sample to detector distance of 731 mm and an X- ay wavelength, �, of 1.608 Å. Scattering curves were recorded ithin the range 0.01–0.25 Å−1. According to the sampling the- rem [26], the maximum dimension of scattering particle that ould be observed was circa 628 Å (Dmax = 2�/qmin). The experi- ental intensities were corrected for sample attenuation, detector omogeneity, and blank scattering, both in the absence and in the resence of urea. The scattering vector amplitude q is defined as = 4� sin �/�, 2� being the scattering angle and � the X-ray wave- ength of 1.608 Å [26]. .3.1. SAXS data analysis Small angle X-ray scattering technique involves diffraction vents from electrons and provides data about the protein struc- ure at low resolution in solution [27,28]. In general, proteins are tudied in high positive solute–solvent contrasts because they are ore electron-dense than water and other buffer [27]. The scat- ering contributions associated to the presence of urea in solution ere subtracted from the HbGp SAXS curves before the analysis. herefore, it was necessary to measure a scattering curve for each olution containing the pure buffer and different experimental urea oncentrations. In this paper, the radius of gyration Rg, was determined by uinier approximation [26] and using the GNOM software [29]. For monodisperse solution of globular macromolecules the Guinier pproximation is defined by Eq. (2) [26]: (q → 0) = I(0) exp(−(R2 gq2)/3) (2) Thus, I(0) and Rg can be obtained from the y-axis intercept and he slope of the linear plot of (lnI(q) versus q2), respectively [26]. owever, the range over which the Guinier approximation is valid or each measured scattering curve must be considered. In the absence of interference effects, a Fourier transform con- ects the scattering intensity I(q) to the pair distance distribution unction, p(r), which is related to the probability of finding a air of small elements at a distance r within the entire volume f the scattering particle [26,30]. The p(r) curve represents the tructure in the real space and corresponds to the distribution f all the interatomic distances r within the macromolecule. In his work, we made use of the GNOM program [29] to obtain (r), Rg and Dmax values by fits of the experimental scattering urves. Another important analysis in the study of denaturation of acromolecules by SAXS is the Kratky representation, which can e defined as the plot of q2 · I(q) versus q [28,30]. Folded globular roteins have a prominent peak at low angles, whereas unfolded roteins display a continuous increase in q2 · I(q) with q [28]. Thus, he Kratky representation is a good indicator of the protein struc- ure in solution. logical Macromolecules 52 (2013) 340– 348 2.4. Dynamic light scattering (DLS) experiments The instrument Zetasizer Nano ZS (Malvern, UK) was used on the light scattering measurements for particle size determination. This instrument allows dynamic light scattering measurements incorporating noninvasive backscattering (NIBS) optics. A He–Ne laser has been used as a light source with wavelength � = 633 nm. The intensity of light scattered at an angle of 173◦ is measured by an avalanche photodiode. The solutions were placed in the ther- mostated sample chamber that maintained the sample stabilized at 25 ◦C, with an accuracy of 0.1 ◦C. At each sample seven measure- ments are performed and for each one a number of measurements is taken (normally around ten) to obtain an adequate statistics. The experiments were performed using oxy- and cyanomet-HbGp concentrations of 3.0 mg/mL at pH 7.0 in the phosphate 30 mmol/L buffer, in the urea concentration range from 0.0 to 8.0 mol/L. Samples were prepared following identical procedures as those used for SAXS. All DLS experiments were performed using a 45 �L cuvette from Hellma (Hellma GmbH, Germany) with 3 mm × 3 mm dimensions. 3. Results and discussions 3.1. AUC data In Fig. 1 the c (S) distributions are shown for oxy-HbGp in the presence of three urea concentrations. At 1.0 mol/L of urea, a sin- gle peak is observed, with sedimentation coefficient, s20,w, around 57 S that can be assigned to the whole protein (Fig. 1A). In Table 1 the s0 20,w values are shown for all urea concentrations used in our experiments. This single s0 20,w value is consistent with the s0 20,w observed previously for oxy-HbGp, in the absence of urea, at pH 7.0 [11]. Moreover, other hydrodynamic properties, shown in Table 1, such as the MM of 3600 ± 100 kDa, and the Stokes radius (RS) of 13.5 ± 0.2 nm, suggest that the species present at 1.0 mol/L of urea is indeed undissociated oxy-HbGp. The RS value estimated for oxy- HbGp by AUC is very consistent with the hydrodynamic diameter value (Dh) of 27 ± 1 nm, determined independently by DLS studies [3]. The MM values shown in Table 1 were obtained by the global analysis of the sedimentation velocity data (SV) using the SEDPHAT software (version 9.4). The increase of urea concentration induces the appearance of the contribution of other species in solution. At 2.0 mol/L of urea (Table 1) the contribution of two species is observed: the first one, with s0 20,w 10.3 ± 0.3 S and MM of 205 ± 5 kDa (3 ± 1%, Table 1), is associated to the dodecamer (abcd)3 and the second species is assigned to the whole protein (97 ± 1%, Table 1). The RS value found for the dodecamer varies between 6.2 and 5.1 nm (Table 1) and is consistent with the hydrodynamic diameter of 11 nm obtained by DLS experiments on the isolated dodecamer (data not shown). Our results suggest that oxy-HbGp is very stable at 2.0 mol/L of denatu- rant since the contribution of dissociated species in solution is quite small, reaching only 3%. In this context, the work of Krebs et al. [31], focusing the proper- ties of the dodecamer of HbLt, has shown that this subunit obtained from urea-induced dissociation of the whole protein consisted of a concentration-dependent equilibrium of three species with sed- imentation coefficients of 8.5–9.4 S, 3.6–4.4 S and 1.9 S. In a recent study of oxy- and cyanomet-HbGp forms, at pH 10.0 [4], no signifi- cant contribution of the dodecameric species (abcd)3 in the solution was observed. This observation is, probably, due to a significant decrease of the HbGp stability in alkaline pH, above pH 9.0. In Fig. 1B the c (S) distributions are shown for oxy-HbGp, in the presence of 3.0 mol/L of urea, at pH 7.0, monitored at 415 nm. Three species appear in the distribution, where the smallest MM species is not observed at 1.0 and 2.0 mol/L of urea. This third species with F.A.O. Carvalho et al. / International Journal of Biological Macromolecules 52 (2013) 340– 348 343 Fig. 1. Continuous sedimentation coefficient distributions of oxy-HbGp in 100 mmol/L Tris–HCl, pH 7.0, 50 mmol/L NaCl. The panel displays the c (S) fittings for HbGp concentrations of 100, 200 and 300 �g/mL. The baselines for different protein concentrations are shifted along the ordinate axis to emphasize their differences. The s20,w for each concentration was determined as the maximum of Gaussian curves. Absorbance monitored at 415 nm. (A) Urea at 1.0 mol/L, (B) at 3.0 mol/L, and (C) at 6.0 mol/L. Table 1 Hydrodynamic properties, s0 20,w in (S), MM in (kDa) and RS in (nm), for the oxy-HbGp, in the presence of indicated urea concentrations, at pH 7.0 and 20 ◦C, obtained from sedimentation velocity (SV) data in the SEDFIT and SEDPHAT softwares. The Area corresponds to the species contribution (%) as measured by the corresponding area of c (S) peaks. [Urea] (mol/L) Properties Species observed Monomer, d Trimer, abc Tetramer, abcd Dodecamer, (abcd)3 Whole protein 0.0 s0 20,w – – – – 58.6 ± 0.6 MM – – – – 3620 ± 80 RS – – – – 13.7 ± 0.1 Area – – – – 100 1.0 s0 20,w – – – – 57.3 ± 0.3 MM – – – – 3600 ± 100 RS – – – – 13.5 ± 0.2 Area – – – – 100 2.0 s0 20,w – – – 10.3 ± 0.3 57.3 ± 0.3 MM – – – 205 ± 5 3600 ± 100 RS – – – 6.2 ± 0.3 13.3 ± 0.4 Area – – – 3 ± 1 97 ± 1 3.0 s0 20,w – – 6.1 ± 0.8 9.2 ± 0.6 58.5 ± 0.6 MM – – 72 ± 6 203 ± 16 3500 ± 200 RS – – 4.2 ± 0.3 6.1 ± 0.4 13.9 ± 0.3 Area – – 20 ± 4 15 ± 2 65 ± 7 4.0 s0 20,w 2.0 ± 0.6 3.8 ± 0.4 5.5 ± 0.4 9.4 ± 0.6 57.4 ± 0.6 MM 17 ± 1 51 ± 1 70 ± 5 204 ± 3 3600 ± 50 RS 2.5 ± 0.2 3.2 ± 0.2 3.9 ± 0.3 5.1 ± 0.4 13.2 ± 0.2 Area 25 ± 6 24 ± 5 8 ± 2 17 ± 6 26 ± 3 5.0 s0 20,w 2.1 ± 0.6 4.1 ± 0.6 5.8 ± 0.5 9.3 ± 0.7 – MM 17 ± 2 53 ± 2 67 ± 4 202 ± 8 – MMa 16.8 ± 0.7a 52 ± 3a 69 ± 2a 206 ± 3a – RS 2.4 ± 0.1 3.1 ± 0.2 3.8 ± 0.3 5.3 ± 0.6 – Area 46 ± 2 37 ± 5 6 ± 1 11 ± 4 – 6.0 s0 20,w 2.1 ± 0.6 3.8 ± 0.5 5.3 ± 0.8 9.1 ± 0.8 – MM 18 ± 2 51 ± 4 71 ± 5 206 ± 12 – RS 2.7 ± 0.5 3.5 ± 0.4 4.3 ± 0.2 5.5 ± 0.4 – Area 40 ± 2 49 ± 4 7 ± 1 4 ± 1 – HbLt s0 20,w b 1.83 ± 0.01 3.85 ± 0.02 4.73 ± 0.02 9.38 ± 0.05 – MMc 16.6 53 69.6 208.8 3600 a Molecular masses obtained by global fit of the equilibrium sedimentation data. b Prediction of the s20,w for the HbLt subunits by the HydroPro software [24]. c Prediction of the MM for the HbLt subunits by the Sednterp software using the aminoacid sequence [9,20]. 344 F.A.O. Carvalho et al. / International Journal of Biological Macromolecules 52 (2013) 340– 348 F the 30 c and (D t nd (D) s a i v b v t d t o t r o O i t o s g a a u u e c c T m r p t a c t [ ig. 2. (A) and (C) Small angle X-ray scattering curves of oxy-HbGp 3.0 mg/mL, in orrespond to the experimental data and the lines to the fits by GNOM software. (B) o the scattering curves in the (A) and (C), respectively. The symbols shown in (B) a 0 20,w of 6.1 ± 0.8 S and MM of 72 ± 6 kDa is assigned to the tetramer bcd (Fig. 1B, Table 1). Furthermore, taking into account the lim- tations of the technique and the complexity of the samples, the alues of MM for the tetramer (abcd) obtained from our SV data, y global analysis, are quite reasonable, as compared to the MM alue of 68 kDa determined by MALDI-TOF-MS [13,32]. In Table 1 he relative percentages for the three species in equilibrium are isplayed: they are, respectively, 20 ± 4, 15 ± 2 and 65 ± 7%, for the etramer, dodecamer and whole protein. Therefore, the increase f urea concentration promotes the protein oligomeric dissocia- ion, even though, the larger contribution of the solution species emains the whole protein (65 ± 7%), suggesting that, at 3.0 mol/L f urea, oxy-HbGp is still quite stable towards urea denaturation. ur results suggest that the dissociation process induced by urea s similar, in some aspects, to alkaline dissociation [4]. However, he urea-induced HbGp oligomeric dissociation, up to 4.0 mol/L f denaturant, is characterized by slight changes in the secondary tructure, monitored in the peptide region, as well as, in the heme roup region. Previous studies have shown that only slight alter- tions are observed in the ellipticity signal and in the Soret band bsorbance [6]. Thus, the HbGp oligomeric dissociation induced by rea corresponds to the formation of an intermediate state in the nfolding process, quite similar to the native state. At 4.0 mol/L of urea two additional new species are observed in quilibrium, as shown in Table 1. According to the sedimentation oefficient s0 20,w and molecular mass MM values, these new species an be attributed to the polypeptide chain d and the trimer abc. he values of s0 20,w and MM are 2.0 ± 0.6 S and 17 ± 1 kDa for the onomer d, and 3.8 ± 0.4 S and 51 ± 1 kDa for the trimer (abc), espectively (Table 1). The contribution of the undissociated rotein decreases significantly to 26 ± 3%, which is comparable o the contributions of the two dissociated species, the monomer nd the trimer (around 25% each, Table 1). The sedimentation oefficients and MM values obtained for the monomer d and he trimer (abc) are similar to those found in our previous work 4,13,31]. Further increase of urea concentration, up to 5.0 mol/L, mmol/L acetate–phosphate buffer at pH 7.0, in the presence of urea. The symbols ) Distance distribution functions p(r) obtained from GNOM software corresponding have the same meaning as for (A) and (C). leads to the complete protein dissociation into smaller species (Table 1). Percentage species contributions are 46 ± 2, 37 ± 5, 6 ± 1 and 11 ± 4%, for the monomer d, trimer (abc), tetramer (abcd) and dodecamer (abcd)3, respectively. The fact that, in the distributions curves, the contributions of the linkers chains, L1, L2 and L3 are not observed, is, probably, due to the overlap of the contributions of these polypeptide chains with other species, being, for this reason, unresolved. In a recent paper, describing the characterization of HbGp species by SDS-PAGE electrophoresis and MALDI-TOF-MS, it is suggested that a linker chain is strongly bound to the trimer (abc) [32]. Besides, AUC measurements of isolated trimeric frac- tions show that 80% of the solution composition corresponds to the trimer (abc), and the remaining 20% are due to the trimer, associated to a linker chain (abc + L, data not shown, [32]). The predictions of the s0 20,w for the linkers, based on the aminoacid sequences of the known HbLt chains, give values very close to the sedimentation coefficients obtained for the monomeric subunit d [32]. Thus, it is quite possible that the contributions from these polypeptide chains overlap with the contribution of the monomer d and trimer (abc) and, for this reason, the contribution of isolated linker chains is not observed. Besides, there are some addi- tional points that make it difficult the characterization of the linkers in solution, such as their asymmetrical shape as compared to the globins chains, and their arrangement in the oligomeric structure, with a localization and orientation inside the oligomeric struc- ture, more protected from the solvent. Our rigid body estimates for HbLt species suggest that the linkers, in the monomeric form, are expected to have hydrodynamic parameters close to those for our monomeric fraction, due to the high asymmetry of the linkers [32]. An initial analysis of SE data, for oxy-HbGp at pH 7.0, in the pres- ence of 5.0 mol/L of urea, was performed by the SEDPHAT software using separate data sets corresponding to each individual protein concentration, but in a multi-speed mode. Based on these fits the MM for the species present in the solution were determined, and they are consistent with the data obtained by MALDI-TOF-MS [32], as well as, with the corresponding SV data (see Table 1). F.A.O. Carvalho et al. / International Journal of Biological Macromolecules 52 (2013) 340– 348 345 Table 2 SAXS and DLS parameters, for the extracellular hemoglobin of G. paulistus, in the oxy- and cyanomet- forms, at 3.0 mg/mL, in the presence of indicated urea concentrations. HbGp forms Urea (mol/L) Dh (nm)a PDIa Rg (Å)b Rg (Å)c Dmax (Å)b I(0) (a.u.)b Oxy- 0.0 27 ± 1 0.002 106 ± 1 105 ± 10 300 0.26 ± 0.01 2.0 29 ± 1 0.003 106 ± 1 107 ± 9 300 0.19 ± 0.01 4.0 31 ± 1 0.035 106 ± 1 108 ± 8 300 0.10 ± 0.02 5.0 30 ± 1 0.13 118 ± 2 115 ± 12 500 0.019 ± 0.005 6.0 15 ± 1 0.25 169 ± 4 135 ± 17 620 0.012 ± 0.001 7.0 14 ± 1 0.27 203 ± 6 201 ± 40 740 0.011 ± 0.001 Cyanomet- 0.0 27 ± 1 0.02 107 ± 1 107 ± 8 300 0.20 ± 0.01 2.0 28 ± 1 0.005 106 ± 1 108 ± 7 300 0.16 ± 0.01 4.0 31 ± 1 0.02 107 ± 1 107 ± 7 300 0.14 ± 0.02 6.0 26 ± 1 0.23 106 ± 1 145 ± 20 350 0.034 ± 0.002 7.0 14 ± 1 0.31 160 ± 2 166 ± 40 550 0.011 ± 0.001 8.0 14 ± 1 0.32 189 ± 3 170 ± 30 600 0.009 ± 0.002 a Parameters obtained by DLS experiments. D is the average hydrated diameter and PDI the polydispersity index. articl t i m t c s o e p d p 3 3 c c a f o a v i u o p i ( D u t t u w o a l p o o t i c D t h b Parameters obtained by GNOM software. Rg is the radius of gyration, Dmax the p c Radius of gyration obtained by Guinier approximation. At 6.0 mol/L of urea (Fig. 1C), the observed species are, essen- ially, the same found at 5.0 mol/L. However, a significant decrease n the contribution of the dodecamer (abcd)3 is observed, and the onomer d and the trimer (abc) are the predominant species, con- ributing both a total of 89% (Table 1). Thus, the increase of urea oncentration induces further dissociation of the larger species in olution. At higher concentrations, urea induces the denaturation f the dissociated species, as monitored in a recent study by sev- ral spectroscopic techniques [6]. Differently from the studies at H 10.0 [4], where the contribution of the dimer of monomers d, 2, with s0 20,w value around 2.7 S, was noticed, our results in the resence of urea suggest that this species is not observed. .2. SAXS and DLS data Fig. 2 A and C shows the SAXS curves obtained for oxy-HbGp .0 mg/mL, at pH 7.0, 20 ◦C, as a function of the increase of urea oncentration. SAXS curves, in the absence of urea, present three haracteristics shoulders, centered at q values of 0.04 Å−1, 0.07 Å−1 nd 0.12 Å−1 (see Fig. 2A). The corresponding distance distribution unctions p(r) and parameters obtained by GNOM software, for the xy-HbGp, as a function of urea concentration, are shown in Fig. 2B nd D, and in Table 2. In the absence of urea, oxy-HbGp presents Rg, Dmax and I(0) alues, respectively, of 106 ± 1 Å, 300 Å, and 0.26 ± 0.01, which are n agreement with previous SAXS studies [16]. In the presence of rea, in the range from 1.0 to 4.0 mol/L, no significant changes are bserved in the values of Dmax and Rg, characterizing the scattering article dimensions (Table 2). However, a decrease in the scatter- ng intensity at low q (Fig. 2A and C), I(0), from 0.26 to 0.10 a.u. Table 2) is noticed; this decrease and the p(r) curves (Fig. 2B and ), suggest that oxy-HbGp undergoes partial dissociation, in this rea concentration range. The decrease of intensity is associated o the formation of smaller species in the solution, as observed in he AUC data (see Table 1). The changes in p(r) curves and I(0) val- es are relatively small due to the significant contribution of the hole protein scattering in this urea concentration range. More- ver, above 4.0 mol/L, the SAXS curves undergo significant changes: dramatic decrease of I(0), as well as, the loss in shoulders reso- ution are observed, suggesting the beginning of the denaturation rocess (Fig. 2A and C). Furthermore, at 5.0 mol/L of urea, increases f Rg from 106 ± 1 to 118 ± 2 Å and Dmax from 300 to 500 Å are bserved (Table 2). These changes, at 5.0 mol/L, can be associated o the oligomeric dissociation/denaturation process, as suggested n previous spectroscopic studies [6]. At 6.0 and 7.0 mol/L of urea, omplete denaturation of HbGp species takes place, with Rg and max values of 169 ± 4 and 203 ± 6 Å, and 620 and 740 Å, respec- ively (Table 2). e maximum dimension and I(0) the scattering intensity at q = 0. Therefore, above 4.0 mol/L, oxy-HbGp is extensively dissoci- ated and denatured, as can be observed from the p(r) profiles shown in Fig. 2B. The distance distribution function is character- ized by a two-species population, one with smaller dimensions (Rg around 50 Å) and a second one, characterized by a broad distri- bution with Rg values higher than 150 Å. The smaller dimension species could be associated to the monomer and trimer species, while the species with higher dimensions could be assigned as the dodecamer, tetramer and one half of whole protein in denatured form, as observed in previous work [16]. Previous studies by optical absorption, fluorescence emission, and CD show that both forms, oxy- and cyanoment-HbGp have a similar denaturation process. However, the oxidized form dis- played a higher stability. Thus, as the intermediate oligomers are the same for the two forms, only the SAXS study was made to compare the stability of the oxy- and cyanomet-HbGp forms. The cyanomet-HbGp, in the presence of urea, has a similar dissociation and denaturation behavior as the oxy- form. How- ever, cyanomet-HbGp is clearly more stable than the oxy- form, since it remains intact up to 5.0 mol/L of urea (see Fig. 3A). At 6.0 mol/L of denaturant, a decrease of I(0) intensity occurs in the SAXS curves, as well as, the shoulders present a decrease of intensity and loss of resolution (Fig. 3C), consistent with the beginning of the oligomeric dissociation process [6]. The changes of the SAXS curves for cyanomet-HbGp are considerably smaller as compared with the oxy-form, at this urea concentration (see Fig. 2C and 3C). Moreover, for cyanomet-HbGp, the Dmax val- ues start to increase from 6.0 mol/L, while the Rg and I(0) values already present some changes at lower urea concentrations. The Rg values, given by GNOM approximation, begin to increase at 7.0 mol/L of urea, from 106 ± 1 to 160 ± 2 Å. Furthermore, the I(0) values show a 4-fold reduction (from 0.14 to 0.034 a.u.), probably, due to the oligomeric dissociation together with the beginning of the denaturation of the dissociated species in solution. However, the Rg values obtained by Guinier approximation, start to change significantly from 6.0 mol/L of urea, when an Rg increase from 107 ± 7 to 145 ± 20 Å is observed (Table 2). Therefore, urea-induced cyanomet-HbGp unfolding, occurs between 6.0 and 7.0 mol/L, con- sistent with previous results based on spectroscopic studies [6]. The p(r) distributions show that cyanomet-HbGp presents a sim- ilar behavior as oxy-HbGp, with significant changes shifted to higher urea concentrations (between 6.0 and 7.0 mol/L, Fig. 3D). The broader distribution of larger size species for cyanomet-HbGp suggests that, in this case, the oligomeric dissociation and denatur- ation processes, induced by urea, are more complex as compared to oxy-HbGp. This could be due to the higher resistance to unfolding of the whole protein and dissociated species of the oxidized form as compared to oxy-HbGp. 346 F.A.O. Carvalho et al. / International Journal of Biological Macromolecules 52 (2013) 340– 348 F L, in 3 c and (D t D) hav i b t o 4 c o ( t T c i s s T R ig. 3. (A) and (C) Small angle X-ray scattering curves of cyanomet-HbGp 3.0 mg/m orrespond to the experimental data and the lines to the fits by GNOM software. (B) o the scattering curves in (A) and (C), respectively. The symbols shown in (B) and ( Complementary studies were also performed for the urea- nduced unfolding using a protein concentration of 0.5 mg/mL, for oth oxy- and cyanomet-HbGp, at pH 7.0. At low protein concen- ration, the oxy- and cyanomet-HbGp are less stable in the presence f urea, since the two forms undergo changes above 3.0 and .0 mol/L of urea, respectively. As described above, at higher con- entration, the changes on Rg, Dmax and I(0) parameters take place nly above 4.0 mol/L and 6.0 mol/L, for both forms, respectively Tables 2 and 3). The decrease of protein concentration, from 3.0 o 0.5 mg/mL promotes the oligomeric dissociation process earlier. herefore, HbGp undergoes oligomeric dissociation at smaller urea oncentrations (Tables 2 and 3). The effect of protein concentration n the dissociation process was observed previously by DLS and ize exclusion chromatography [16]. AUC data were obtained for maller protein concentrations, in the range from 100 to 300 �g/mL able 3 g, Dmax and I(0) values for the extracellular hemoglobin of G. paulistus, in oxy- and cyano Urea (mol/L) HbGp forms Rg (Å)a 0.0 Oxy 107 ± 1 2.0 106 ± 1 3.0 107 ± 1 4.0 120 ± 2 5.0 215 ± 5 5.5 225 ± 6 0.0 Cyanomet 108 ± 1 2.0 110 ± 1 3.0 111 ± 1 4.0 111 ± 1 5.0 129 ± 2 5.5 181 ± 4 a Parameters obtained by GNOM software. Rg is the radius of gyration, Dmax the particl b Radius of gyration obtained by Guinier approximation. 0 mmol/L acetate–phosphate buffer at pH 7.0, in the presence of urea. The symbols ) Distance distribution functions p(r) obtained from GNOM software corresponding e the same meaning as for (A) and (C). and no significant differences in the c (S) distributions were observed, probably, due to the small protein concentration range. The corresponding Kratky’s plots are shown in Fig. 4A and B. At 0.0 mol/L of urea, for oxy- and cyanomet-HbGp forms, characteristic peaks of folded globular proteins at low angles are observed, suggesting that the HbGp is totally folded. The increase in urea concentration induces a progressive peak inten- sity decrease, due to the oligomeric dissociation. Furthermore, at 5.0 and 7.0 mol/L of urea, for oxy- and cyanomet-HbGp, respec- tively, the complete peak disappearance is observed, indicating the protein denaturation. Additional DLS experiments, performed for oxy- and cyanomet- HbGp, in the presence of urea, suggest that both oxidation forms are stable up to 4.0 mol/L, since the polydispersity index of the solution remains constant (Table 2), indicating mono-disperse met- forms at 0.5 mg/mL, in the presence of urea. Rg (Å)b Dmax (Å)a I(0) (a.u.)a 108 ± 10 300 0.047 ± 0.002 109 ± 8 300 0.021 ± 0.001 109 ± 9 300 0.025 ± 0.002 124 ± 14 450 0.006 ± 0.002 175 ± 12 680 0.004 ± 0.001 175 ± 30 750 0.005 ± 0.002 110 ± 6 300 0.041 ± 0.001 113 ± 11 300 0.032 ± 0.001 117 ± 10 300 0.035 ± 0.003 125 ± 15 300 0.023 ± 0.001 142 ± 18 450 0.014 ± 0.001 170 ± 30 620 0.010 ± 0.002 e maximum dimension and I(0) the scattering intensity at q = 0. F.A.O. Carvalho et al. / International Journal of Bio F p p p i T c 4 3 a s t p c c t a H issue for HbGp future research is the correlation of structural data F s ig. 4. Kratky plots for HbGp 3.0 mg/mL, in 30 mmol/L acetate–phosphate buffer at H 7.0, in the presence of urea. (A) Oxy-HbGp and (B) cyanomet-HbGp. rotein solutions. At larger urea concentrations, an increase of the olydispersity index of the scattering particles size distributions s observed, associated to the dissociation/denaturation process. he results of SAXS and DLS experiments are quite consistent, onfirming HbGp unfolding at higher urea concentrations, above .0 mol/L. .3. Unfolding model for HbGp in the presence of urea Our previous study of HbGp, in the oxy- and cyanomet- forms, t pH 7.0, suggests that the HbGp oligomeric stoichiometry is imilar to that proposed by Vinogradov [33] for HbLt. According o Vinogradov’s model the structure of HbLt is composed of 12 rotomers, composed by a globin dodecamer and three linker hains L, (abcd)3L3 [33]. A recent preliminary analysis of HbGp rystallographic structure, at 3.15 Å resolution [34], suggests that he stoichiometry, and the hierarchical levels of HbGp are the same s for HbLt. Thus, this model was used to describe the unfolding of bGp, in the presence of urea. Moreover, recent results on the urea ig. 5. Urea-induced oligomeric dissociation model for HbGp, at pH 7.0, based on the AUC ubunits was adapted from the PDB bank under ID code 2GTL [7] and corresponds to the logical Macromolecules 52 (2013) 340– 348 347 effect upon three HbGp forms, monitored by several spectroscopic techniques [6], will be used for the description of this model. Although the electronic density of the HbGp crystal is known, the crystallographic structure is not totally resolved, since only the polypeptide monomeric chain d has the primary sequence estab- lished. In the future, with the sequencing of all HbGp polypeptide chains, the crystallographic structure will be described in detail. Therefore, a complete description of the HbGp unfolding model, in the presence of urea, discussing the oligomer interface in relation to the association equilibrium, the oligomeric intermediates and the oxygen cooperativity, will only be possible with the complete resolution of the crystallographic structure. According to our experimental results, the HbGp unfolding pro- cess, in the presence of urea, is composed by two phases as shown in the schematic model in Fig. 5. In the first one the oligomeric dissociation occurs, producing several smaller species, followed by the second one, where the dissociated species undergo denaturation. HbGp oligomeric dis- sociation, in the presence of urea, can be described by a model, where the whole protein 12x(abcd)3L3 dissociates into the dode- camer (abcd)3, at 2.0 mol/L of denaturant agent. At 3.0 mol/L, the dodecamer (abcd)3 undergoes further partial dissociation into the tetramer (abcd, Fig. 5). Increasing further the urea concentration to 4.0 mol/L, the tetramer (abcd) starts to dissociate into the smaller species, trimer (abc) and monomer d. At concentrations, above 6.0 mol/L of urea, the dissociated species undergo denaturation. Our present AUC data shows clearly the partial HbGp oligomeric disso- ciation, at urea concentrations in the range from 2.0 to 4.0 mol/L. Previous work based on spectroscopic studies has shown that, in this concentration range, no significant changes in the protein sec- ondary structure and heme groups are observed [6]. Thus, the possible intermediate state in the unfolding process is assigned to the partial oligomeric dissociation of the whole protein into several dissociated species in the equilibrium, maintaining its properties very close to the original native state. Moreover, our present SAXS data show that the oxy- and cyanomet- forms undergo denaturation, at urea concentrations above 4.0 and 6.0 mol/L, respectively. The unfolding of the species is characterized by an increase of the values for Dmax and Rg (Table 2). The decrease in the I(0) values also suggests oligomeric disso- ciation. Besides, SAXS studies indicate a greater stability of the cyanomet-HbGp upon denaturation by urea, consistent with pre- vious work, based on spectroscopic methods [6], and both HbGp oxidation forms follow the two-phases oligomeric dissociation model proposed in this paper (Fig. 5). At this point it is worthy of mention the work on the subunits organization of HbLt linked to oxygen binding [14]. An important for subunits equilibrium in solution, under different conditions, with the main function which is oxygen binding and release. The study by Riggs mentioned above [14] suggested a special role for and SAXS data. The crystallographic structure used in this scheme for the different HbLt structure, which is similar to that of HbGp. 3 of Bio a i t c d s r i p 4 s s r c a a s o r p o o t s t c u p d s H o A L p C u C r v M d [ [ [ [ [ [ [ [ [ [ [ [ [ [ [ [ [ [ [ [ [ [ [ [ 48 F.A.O. Carvalho et al. / International Journal n octamer (dimer of tetramer, (abcd)2), which was found incorrect n subsequent structural work by Royer group [35], which showed hat the dodecamer is the main subunit in the oligomeric disso- iation of the whole protein. Unfortunately, since no oxygenation ata for HbGp subunits is presently available and since the crystal tructure still needs some work before it will be available at high esolution, we cannot at the moment, to analyze in more detail the ssue of subunits contacts. As noticed above, this is an interesting oint for future research. . Conclusions AUC and SAXS data analysis allowed the characterization of everal species present in the equilibrium, and to evaluate the dis- ociation/denaturation processes in the presence of urea. Our AUC data suggest that the species observed in the equilib- ium are strongly dependent on the urea concentration. At low urea oncentrations, below 4.0 mol/L, the larger species (whole protein nd dodecamer) are predominant, and the increase of chaotropic gent induces the formation of smaller species in the solution, uch as trimer (abc) and monomer (d). The s0 20,w and MM values btained by AUC for all species are quite close to literature values eported for HbLt, as well as, to those found for HbGp, at alkaline H. SAXS data show that the cyanomet-HbGp is more stable than xy-HbGp, and both forms undergo unfolding above 4.0–5.0 mol/L f urea. These results are complementary to previous studies on he effect of urea upon three HbGp forms, monitored by various pectroscopic techniques [6], suggesting that HbGp undergoes par- ial oligomeric dissociation followed by denaturation at higher urea oncentrations. In summary, the HbGp unfolding process, in the presence of rea, is composed of two phases. The first one is associated to the rotein oligomeric dissociation, while the second one to the species enaturation. Thus, the several dissociated species observed in olution, as a function of urea concentration, are assigned to the bGp intermediate state in the unfolding process, observed from ur previous spectroscopic study. cknowledgments The authors are grateful to the Spectroscopy and Calorimetry aboratory of the National Biosciences Laboratory (LNBio, Cam- inas, Brazil) and National Synchrotron Light Laboratory (LNLS, ampinas, Brazil) for making available the AUC and SAXS facilities sed in this work. Thanks are due to the Brazilian agencies FAPESP, NPq, and CAPES for partial financial support. F. A. O. Carvalho is a ecipient of a Ph.D. grant from FAPESP (2009/17261-6). J.W.P. Car- alho is a recipient of a Ph.D. grant from FAPESP (2010/09719-0). . Tabak is grateful to CNPq for a research grant. Thanks are also ue to Mr. Ézer Biazin for efficient support in sample preparations. [ [ logical Macromolecules 52 (2013) 340– 348 References [1] S.N. Vinogradov, Micron 35 (2004) 127–129. [2] R.E. Weber, S.N. Vinogradov, Physiological Reviews 81 (2001) 569–628. [3] P.S. Santiago, J.W.P. Carvalho, M.M. Domingues, N.C. Santos, M. Tabak, Biophys- ical Chemistry 152 (2010) 128–138. [4] F.A.O. Carvalho, P.S. Santiago, J.C. Borges, M. Tabak, International Journal of Biological Macromolecules 48 (2011) 183–193. [5] P.S. Santiago, L.M. Moreira, E.V. Almeida, M. Tabak, Biochimica et Biophysica Acta 1770 (2007) 506–517. [6] F.A.O. Carvalho, P.S. Santiago, M. Tabak, Archives of Biochemistry and Bio- physics 519 (2012) 46–58. [7] W.E. Royer, H. Zhu, T.A. Gorr, J.F. Flores, J.E. Knapp, Journal of Biological Chem- istry 280 (2005) 27477–27480. [8] R.E. Hirsch, L.A. Jelicks, B.A. Wittenberg, D.K. Kaul, H.L. Shear, J.P. Harrington, Artificial Cell Blood Substitute 25 (1997) 429–444. [9] M. Rousselot, E. Delpy, R.C. Drieu La, V. Lagente, R. Pirow, J.F. Rees, A. Hagege, D. Le Guen, S. Hourdez, F. Zal, Biotechnology Journal 1 (2006) 333–345. 10] T. Harnois, M. Rousselot, H. Rogniaux, F. Zal, Artificial Cell Blood Substitute 37 (2009) 106–116. 11] F.A.O. Carvalho, P.S. Santiago, J.C. Borges, M. Tabak, Analytical Biochemistry 385 (2008) 257–263. 12] W.E. Royer, K. Strand, M. van Heel, W.A. Hendrickson, Proceedings of the National Academy of Science of the United States of America 97 (2000) 7107–7111. 13] M.S. Oliveira, L.M. Moreira, M. Tabak, International Journal of Biological Macro- molecules 40 (2007) 429–436. 14] K. Fushitani, A.F. Riggs, Journal of Biological Chemistry 266 (1991) 10275–10281. 15] P.S. Santiago, F. Moura, L.M. Moreira, M.M. Domingues, N.C. Santos, M. Tabak, Biophysical Journal 94 (2008) 2228–2240. 16] J.W.P. Carvalho, P.S. Santiago, T. Batista, C.E.G. Salmon, L.R.S. Barbosa, R. Itri, M. Tabak, Biophysical Chemistry 163–164 (2012) 44–55. 17] A.L. Poli, L.M. Moreira, M. Tabak, H. Imasato, Colloids and Surface B 52 (2006) 96–104. 18] S.C.M. Agustinho, M.H. Tinto, J.R. Perussi, M. Tabak, H. Imasato, Comparative Biochemistry and Physiology 118A (1998) 171–181. 19] C.N. Pace, Methods in Enzymology 31 (1986) 266–280. 20] T.M. Laue, Current Opinion in Structural Biology 11 (2001) 579–583. 21] J. Lebowitz, M.S. Lewis, P. Schuck, Protein Science 11 (2002) 2067–2079. 22] P. Schuck, Biophysical Journal 78 (2000) 1606–1619. 23] P. Schuck, M.A. Perugini, N.R. Gonzales, G.J. Howlett, D. Schubert, Biophysical Journal 82 (2002) 1096–1111. 24] J. Vistica, J. Dam, A. Balbo, E. Yikilmaz, R.A. Mariuzza, T.A. Rouault, P. Schuck, Analytical Biochemistry 15 (2004) 234–256. 25] P. Schuck, Analytical Biochemistry 320 (2003) 104–124. 26] A. Guinier, G. Fournet, Small Angle Scattering of X-ray, Wiley, New York, 1955. 27] S.J. Perkins, R. Nan, K. Li, S. Khan, Y. Abe, Methods 54 (2011) 181–199. 28] H.D.T. Mertens, D.I. Svergun, Journal of Structural Biology 172 (2010) 128–141. 29] D.I. Svergun, Journal of Applied Crystallography 24 (1991) 485–492. 30] O. Glatter, O. Kratky, Small Angle X-ray Scattering, Academic Press, London, 1982. 31] A. Krebs, A.R. Kuchumov, P.K. Sharma, E.H. Braswell, P. Zipper, R.E. Weber, G. Chottard, S.N. Vinogradov, Journal of Biological Chemistry 271 (1996) 18695–18704. 32] F.A.O. Carvalho, J.W.P. Carvalho, P.S. Santiago, M. Tabak, Process Biochemistry 46 (2011) 2144–2151. 33] S.N. Vinogradov, S.D. Lugo, M.G. Mainwaring, O.H. Kapp, A.V. Crewe, Proceedings of the National Academy of Science of the United States of America 83 (1986) 8034–8038. 34] J.F.R. Bachega, L. Bleicher, E.R. Horjales, P.S. Santiago, R.C. Garratt, M. Tabak, Journal of Synchrotron Radiation 18 (2011) 24–28. 35] W.E. Royer, H. Sharma, K. Strand, J.E. Knapp, B. Bhyravbhatla, Structure 14 (2006) 1167–1177. Urea-induced unfolding of Glossoscolex paulistus hemoglobin, in oxy- and cyanomet-forms: A dissociation model 1 Introduction 2 Materials and methods 2.1 Protein extraction and purification 2.2 AUC experiments 2.2.1 AUC data analysis 2.3 SAXS experiments 2.3.1 SAXS data analysis 2.4 Dynamic light scattering (DLS) experiments 3 Results and discussions 3.1 AUC data 3.2 SAXS and DLS data 3.3 Unfolding model for HbGp in the presence of urea 4 Conclusions Acknowledgments References