F G F a b a A R R A A K E G M E S S 1 h o a a m d o i a a u s i S A t i c c 1 d Process Biochemistry 46 (2011) 2144–2151 Contents lists available at SciVerse ScienceDirect Process Biochemistry jo u rn al hom epa ge: www .e lsev ier .com/ locate /procbio urther characterization of the subunits of the giant extracellular hemoglobin of lossoscolex paulistus (HbGp) by SDS-PAGE electrophoresis and MALDI-TOF-MS rancisco Adriano O. Carvalhoa, José Wilson P. Carvalhoa, Patrícia S. Santiagoa,b, Marcel Tabaka,∗ Instituto de Química de São Carlos, Universidade de São Paulo, São Carlos, SP, Brazil Universidade Estadual Paulista “Júlio de Mesquita Filho”, Campus experimental de Registro, SP, Brazil r t i c l e i n f o rticle history: eceived 7 June 2011 eceived in revised form 3 August 2011 ccepted 24 August 2011 vailable online 31 August 2011 eywords: a b s t r a c t Further characterization of hemoglobin of Glossoscolex paulistus (HbGp) subunits was performed based on SDS-PAGE, size exclusion chromatography (SEC) and MALDI-TOF-MS analysis. SDS-PAGE has shown a total of four linker chains, two quite intense and two of lower intensity. HbGp fractions (I–VI), obtained by size exclusion chromatography (SEC), from oligomeric dissociation at alkaline pH 9.6, were moni- tored. Fraction I is identical to the whole protein. The monomeric chains c, obtained from the trimer abc reduction, present four isoforms with MM 17,336 Da, 17,414 Da, 17,546 Da and 17,620 Da. Furthermore, xtracellular hemoglobin lossosocolex paulistus ALDI-TOF-MS lectrophoresis ubunits characterization the trimer subunit presents two isoforms, T1 and T2, with MM 51,200 ± 60 and 51,985 ± 50 Da, respec- tively. Based on SDS-PAGE, the linker chains seem to be distributed along the different fractions of the SEC chromatogram, appearing along the peaks corresponding to fractions I–V. The fraction IV contains, predominantly, trimers with some linkers contamination. The strong interaction of linker chains L with the trimers abc, makes it difficult to obtain these subunits in pure form. The monomer d in fraction VI in ag ubunits molecular masses appears to be quite pure, . Introduction Giant extracellular hemoglobins, also known as erythrocruorins, ave been investigated as a model of extreme complexity in xygen-binding heme proteins [1–3]. They are characterized by very high molecular mass (MM), and their oligomeric structure nd the crowded and protected heme environment are two of the ain factors responsible for the high redox stability. Superoxide imustase (SOD)-like intrinsic activity, observed for hemoglobins f Lumbricus terrestris (HbLt) and of Arenicola marina (HbAm), s another important factor [1,4]. These hemoglobins present highly cooperative oxygen binding and a peculiar behavior ssociated to their oligomeric dissociation into smaller sub- nits and possible rearrangement back into the native oligomeric tructure [5,6]. Moreover, a strong motivation to study these giant hemoglobins s related to their potential use in medicine as blood substitutes. tudies are presently underway to test and validate the use of . marina hemoglobin (HbAm) in this direction [4,7]. They seem o be very promising due to the lack of undesirable immunolog- cal reactions in tests with animals, explained by the absence of ell membranes as occurs with human hemoglobin in red blood ells [4,7]. Besides that, the resistance to oxidation of extracellular ∗ Corresponding author. Fax: +55 16 3373 9982. E-mail address: marcel@sc.usp.br (M. Tabak). 359-5113/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. oi:10.1016/j.procbio.2011.08.013 reement with previous studies. © 2011 Elsevier Ltd. All rights reserved. hemoglobins, as noticed by their redox stability, is also an advan- tage as compared to the use of human hemoglobin in this medical application. The quaternary structure of this class of hemoglobin is constituted by two superimposed hexagonal arrays of twelve spherical subunits, the hexagonal bilayer (HBL). The extracellu- lar hemoglobins are constituted of four subunits, containing heme group, chains a, b, c and d. The subunits a, b and c are linked by disulfide bonds forming the trimer (abc) and subunit d remains in the monomeric form [8]. Additional structural components named linker chains do not have the heme group. These subunits are prob- ably related to the maintenance of the stability of the oligomeric structure [4]. Recent partial characterization of oxy-HbGp (giant extracel- lular hemoglobin of Glossoscolex paulistus) molecular mass by matrix assisted laser desorption time-of-flight mass spectrometry (MALDI-TOF-MS) confirmed the similarity of its subunits to those of orthologous proteins of this class, mentioned above [9]. This characteristic multisubunit content confers to the whole protein a double-layered hexagonal oligomeric structure [10,11]. The MM of HbGp has been re-examined recently by analytical ultracentrifuga- tion (AUC), giving results of MM of approximately 3600 ± 100 kDa and 3700 ± 100 kDa for oxy- and cyanomet-forms, respectively, at pH 7.0 [12]. Studies based on AUC, dynamic light scattering (DLS) and fluorescence emission spectroscopy show that the oxy-HbGp at pH 7.0 remains in its full oligomeric state, while at pH 9.0 it is partially dissociated [12,13]. dx.doi.org/10.1016/j.procbio.2011.08.013 http://www.sciencedirect.com/science/journal/13595113 http://www.elsevier.com/locate/procbio mailto:marcel@sc.usp.br dx.doi.org/10.1016/j.procbio.2011.08.013 Bioche h i e i o p s t t b h 1 e t b [ t w s 1 t r i g c p t m h h u b a s m d t 2 2 t f s 3 a s s a c 0 f o e w ε f 2 T w t v l F.A.O. Carvalho et al. / Process It is worthy of notice that HbGp belongs to the same class of emoglobins as L. terrestris (HbLt), which is one of the most stud- ed hemoglobins in this group. Despite the fact that HbLt has been xtensively studied over the past 20 years, the issue of its true MM s still not fully understood. Daniel et al. have argued that the MM f HbLt could vary between 3.6 and 4.4 MDa [14,15]. They pro- ose a model for the whole protein, consisting of twelve equal tructures involving a dodecamer (abcd)3, and three linkers L3, ogether with twelve tetramers (abcd), in such a way that the pro- omer corresponding to the 1/12 of the whole oligomer is given y (abcd)3L3(abcd), or alternatively, (abcd)4L3 [14]. On the other and, Vinogradov et al. [16] have proposed a model for HbLt, where /12 of the whole molecule is given by (abcd)3L3, so that the differ- nce between the two models is the presence of twelve additional etramers in the former occupying the central part of the hexagonal ilayer. Recent work by Royer et al. on the crystal structure of HbLt 17,18] has suggested that the Vinogradov model is very consis- ent with the crystal structure. Besides that, very recent preliminary ork on the crystal structure of HbGp [19] has also suggested a trong similarity between HbGp and HbLt, both belonging to class , where the two hexagonal layers forming the bilayer are rotated o 16 degree one relative to another. Another interesting question, elevant to the understanding of the overall oligomeric subunit sto- chiometry, is the existence of several isoforms for some of the lobin chains. This has also been elusive in the description of the rystal structure of the whole assembly at atomic resolution. So, the recise characterization of the several globin and linker subunits hat constitute the native extracellular hemoglobins still remains a atter for further research. In this work, MALDI-TOF-MS analysis of the extracellular emoglobin of G. paulistus, HbGp, in the met- and cyanomet-forms, as been performed to further characterize the MM of the sub- nits of this protein, at pH 7.0. Moreover, all fractions obtained y size exclusion gel filtration chromatography (SEC) of oxy-HbGp t pH 9.6 were characterized by polyacrylamide gel electrophore- es (SDS-PAGE) and MALDI-TOF-MS. The detailed studies of the olecular masses of HbGp subunits can contribute to the eluci- ation of the stereochemistry and of the quaternary structure of his interesting HBL hemoglobin. . Materials and methods .1. Protein extraction and purification G. paulistus annelid is prevalent in sites near Piracicaba and Rio Claro cities in he state of São Paulo, Brazil. The HbGp was prepared using freshly drawn blood rom worms. HbGp solution was centrifuged at 2500 rpm, 15 min at 4 ◦C. Then the ample was filtered (Mw cut-off 30 kDa) and centrifuged at 250,000 × g at 4 ◦C during h. The pellet was resuspended in a minimum amount of 0.1 mol/L Tris–HCl buffer t pH 7.0. Chromatography at pH 7.0 in a Sephadex G-200 column furnished the amples used in the experiments [13,20,21]. All concentrations were determined pectrophotometrically in a UV-SHIMADZU 1601PC spectrophotometer, using the ppropriate molar absorption coefficients [13,22]. The final protein concentration orresponding to our stock solution was in the range of 15–30 mg/mL in Tris–HCl .1 mmol/L buffer, pH 7.0. In order to obtain the oxidized met-HbGp, an excess of 5- old potassium ferricyanide relative to heme is added followed by 2 h incubation. To btain the cyanomet-HbGp form, the met-HbGp was further incubated with a 5-fold xcess of potassium cyanide relative to the heme concentration. All concentrations ere determined spectrophotometrically using the molar extinction coefficients 415 nm = 5.5 ± 0.8 (mg/mL)−1 cm−1 for oxy-HbGp, ε402 nm = 4.1 ± 0.7 (mg/mL)−1 cm−1 or met-HbGp and ε420 nm = 4.8 ± 0.5 (mg/mL)−1 cm−1 for cyanomet-HbGp [13,23]. .2. Size exclusion chromatography (SEC) The size exclusion chromatography (SEC) was performed with a Superdex 200 M prep grade column coupled to an ÄKTA prime plus system (GE). The system as equilibrated with 100 mmol/L Tris–HCl buffer, pH 7.0 and flow rate 1 mL/min, he sample concentrations were approximately 18 mg/mL, and the elution sample olume was 1 mL. The elution solution was the same buffer used for column equi- ibration. The elution was monitored at a wavelength of 280 nm in the instrument, mistry 46 (2011) 2144–2151 2145 and, additionally, at 415 nm in a Shimadzu UV-1601PC spectrophotometer. The use of these two wavelengths is very convenient for monitoring hemoglobin samples, since at 280 nm a direct measurement of protein (absorption is due mainly to the aro- matic aminoacids) is made, while at 415 nm the absorption is due to the hemoglobin heme groups. In this way, simultaneous monitoring of protein and heme content can be achieved. Fractions of the subunits of the hemoglobin of G. paulistus were pooled together, stored at 4 ◦C, and characterized by electrophoresis SDS-PAGE. Samples for MALDI-TOF-MS were further dialyzed against a 5 mmol/L Tris–HCl buffer to reduce the salt concentration. All experiments were performed at 25 ◦C. 2.3. Gel electrophoresis The molecular masses of the protein subunits were determined by SDS-PAGE from the fractions collected from the SEC experiments. The percentage of acrylamide used was of 15% at pH 8.6. Samples were run with Tris–HCl 25 mmol/L and glycine 192 mmol/L buffer, pH 6.8, at a constant applied voltage of 140 V. Gels were stained with Coomassie Brillant R-250 from Bio-Rad. The electrophoresis experiments were always made in two plates, simultaneously, one without �-mercaptoethanol and the other one in the presence of the reducing agent. Molecular masses markers were used to determine the apparent molecular mass of the subunits and supplied by Page Ruler TM Unstained Protein Ladder of Fermentas. Relative molecular masses of separated proteins were estimated using protein standards run in parallel in the same gels. 2.4. Matrix-assisted laser desorption/ionization time-of-flight/mass spectrometry – MALDI-TOF-MS The protein fractions isolated by size exclusion chromatography (SEC) and oxy- , met- and cyanomet-integral HbGp oxidation forms, were subjected to analyses on a matrix-assisted laser desorption time-of-flight mass spectrometer. The pro- tein solutions were extensively dialyzed to reduce the buffer (salt) concentration to 5 mmol/L, followed by a further dilution to about 1–3 mg/mL, which is the protein concentration used in MALDI-TOF-MS experiments. The 3,5-dimethoxy-4-hydroxy-cinnamic acid (sinapinic acid) was obtained from Aldrich. 2-Mercaptoethanol, cytochrome c (from Bovine Heart) and bovine serum albumin (BSA) were obtained from Sigma. The laser desorption matrix mate- rial (sinapinic acid) was dissolved in 0.5% trifluoroacetic acid, 50% acetonitrile:water. Samples of the whole native HbGp and the isolated fractions of HbGp were mixed in ratios of 1:5, 1:10 and 1:20 (v/v) with a saturated solution of sinapinic acid [9]. The greater dilution with the matrix was generally made for the whole pro- tein sample which was more concentrated (3.0 mg/mL of protein). One microliter of this mixture (sample/matrix) was spotted onto a MALDI plate and analyzed in MALDI-TOF-MS. Analysis were performed, randomly, in linear, positive ion mode in an Etthan MALDI-TOF mass spectrometer (Amersham Bioscience) using an accelera- tion voltage of 20 kV. Each spot was analyzed twice, accumulating spectra composed of approximately 200 laser shots in total and the resulting spectra were analyzed by Ettan Evaluation software (Amersham Bioscience). The instrument was calibrated using external cytochrome c and BSA standards. The reported molecular masses were obtained as averages of several individual experiments and taking into account the values obtained for single protonated and double protonated molecular species. 3. Results and discussions 3.1. Size exclusion chromatography (SEC) and electrophoresis – SDS-PAGE In Fig. 1A the chromatogram of a SEC experiment for oxy-HbGp 18 mg/mL, incubated at pH 9.6 for 15 h, followed by re-acidulation back to pH 7.0 prior to application to the column, is shown. Chro- matography was performed at pH 7.0. Several species are observed in equilibrium, and in the chromatogram six fractions (indicated by I–VI) were selected and pooled together for further analysis. Detailed assignment of these fractions can only be made based on SDS-PAGE and MALDI-TOF-MS data described later on. However, on the basis of our previous work [9], and on the ratio of heme to protein content, as monitored by the absorbance ratio A415/A280 (Fig. 1B), it is possible to make some associations. The first frac- tion I corresponds to a protein fraction very close to the whole un-dissociated native HbGp eluting in the dead volume, the fourth fraction IV is, predominantly, due to the contribution of the trimer abc, and the sixth fraction VI corresponds to the monomeric subunit d in a very pure state as observed in previous MALDI-TOF-MS [9] and ultracentrifugation studies [24]. The remaining fractions, II, III and V, have been subjected to further studies by other techniques to 2146 F.A.O. Carvalho et al. / Process Biochemistry 46 (2011) 2144–2151 Fig. 1. (A) Superdex 200 gel filtration chromatograms of oxy-HbGp 18.0 mg/mL, in Tris–HCl 0.1 mol/L, pH 7.0, at 25 ◦C. Detection wavelengths are 280 nm (circles) and 4 ( w i q o 2 o n m e m o F t L F t T c a t s e F o t ( a r w c I c c a a t Fig. 2. SDS-polyacrylamide gel electrophoresis of fractions recovered from gel fil- tration in Superdex 200 column of oxy-HbGp. The gel concentration was 15% in 25 mmol/L Tris–HCl, 192 mmol/L glycine, pH 8.3, and stained with Coomassie Blue R-250. Molecular masses of standard proteins were of 200, 50, 40, 30, 25, 20, 15 and 10 kDa and are indicated. The slot (S), (W) corresponds to the standard masses and whole protein (HbGp), respectively, in gels (A) and (B). (A) Gel electrophore- 15 nm (squares). (B) Ratio of absorbances at 415 nm and 280 nm. Roman numerals I–VI) correspond to different species obtained for alkaline dissociation of HbGp, hich were pooled and concentrated for further analysis. mprove our analysis. It is worthy of mentioning, that fraction V is uite rich in linker subunits, as will be seen below. This is expected n the basis of the very low ratio of absorbances at 415 nm and 80 nm (Fig. 1B), implying lower heme content. In fact, the choice f this fraction was mainly motivated to evaluate the content and ature of the linker chains. All fractions I–VI from the chromatogram in Fig. 1A were onitored by sodium dodecyl sulfate (SDS)-polyacrylamide gel lectrophoresis (PAGE), in the absence or presence of �- ercaptoethanol. Fig. 2A shows the results of the electrophoresis f HbGp fractions obtained from the SEC experiment described in ig. 1A, in the absence of the reducing agent. The first lane con- ains the protein standards of MM in the range from 10 to 200 kDa. anes 2–3 and 6–7 correspond, respectively, to fractions I and IV in ig. 1A. They share a common band at 47.8 ± 0.5 kDa (see Table 1) hat is assigned to the trimer abc, observed in previous MALDI- OF-MS experiments with a mass around 52 kDa [9]. Fraction I also ontains contributions from linkers (32.4 ± 0.3 and 27.3 ± 0.5 kDa) nd monomers (12.3 ± 0.3 kDa, see Table 1). It is worthy of notice, hat fraction I is, probably, the whole HbGp oligomer. This can be een by comparison of fraction I (lanes 2–3) with the whole HbGp letrophoresis profile shown in the last lane (denoted as W) of ig. 2A (see also the last column of Table 1). The fraction IV, on the ther hand, contains a significant amount of trimers abc and, addi- ionally, a small amount of a different linker chain at 24.1 ± 0.3 kDa see Table 1). Its content is, predominantly, associated to the trimer bc. The fraction V in the chromatogram in Fig. 1A shows several elatively weak peaks in the MM range between 20 and 30 kDa, hich are, probably, associated to the linker chains and some small ontamination of monomers d and trimers abc. Fractions II and II are shown, respectively, in lanes 4 and 5. Since their protein oncentrations are lower, as compared to the other peaks in the hromatogram of Fig. 1A, the intensity of the corresponding bands re smaller. Finally, the fraction VI is, probably, the most intense nd pure band in the chromatogram of Fig. 1A, corresponding to he monomer d with a MM of 12.3 ± 0.3 kDa (Table 1). sis in the absence of �-mercaptoethanol and (B) with the reducing agent. Roman numerals correspond to the different species obtained for alkaline dissociation and gel filtration as shown in Fig. 1A. In Fig. 2B the results of electrophoresis for the samples with �- mercaptoethanol are presented. Here again, as in Fig. 2A, the lanes 2–3 and 6–7 correspond to fractions I and IV, respectively, in the chromatogram of Fig. 1A. For fraction I the band at 47.8 ± 0.5 kDa is not observed in this case (see Table 1, lower line marked either with an asterisk or with the absence of numbers), indicating that the trimer disulfide bonds were reduced. Besides that, three addi- tional bands are observed in the gel above the 12.3 kDa monomeric band (Table 1, lower line marked either with an asterisk or with the absence of numbers). They are due to the contribution of the monomeric subunits a, b and c produced upon disulfide bond reduction of the trimer. This reduction of the trimer is also clearly observed for the fraction IV in the gel (lanes 6–7). The contribu- tions of the monomeric subunits a, c and b, correspond to MM of 16.2 ± 0.2 kDa, 14.9 ± 0.1 kDa and 14.3 ± 0.3 kDa, respectively (Table 1, lower lines marked with an asterisk). The fraction V (lanes 8 and 9, Fig. 2B) presents several subunits, such as the trimer abc and its monomers obtained upon reduction, and the monomer d, with similar masses as those for the fraction IV (Table 1). However, besides the linker at 28.8 ± 0.5 kDa that is almost coincident with one of the two linkers (at 32.4 ± 0.3 and 27.3 ± 0.5 kDa), observed for the whole native protein, two addi- tional linkers are detected at 26.5 and 23.7 kDa (see Fig. 2B and Table 1). It is important to notice, that in the lanes 8 and 9 (Fig. 2A and B) four bands are noticed, that could be associated to dif- ferent linker chains. The most intense, the second larger MM, at 27.3 kDa, seems to correspond to the lower MM linker band F.A.O. Carvalho et al. / Process Biochemistry 46 (2011) 2144–2151 2147 Table 1 Molecular masses (kDa) of Glossoscolex paulistus hemoglobin (HbGp) subunits, obtained from SDS-PAGE electrophoresis in the absence and presence of �-mercaptoethanol. Subunits Fractions (masses (kDa)) I II III IV V VI Wb Trimer (abc) 47.8 ± 0.5 48.3 ± 0.4 48.3 ± 0.4 47.8 ± 0.5 47.8 ± 0.5 – 49.1 ± 0.6 – – – – – – – Linker 32.4 ± 0.3 32.7 ± 0.3 – – – – 33.1 ± 0.5 32.5 ± 1.2a – – – – – 34.5 ± 1.0a 27.3 ± 0.5 27.2 ± 0.4 27.2 ± 0.4 – 28.8 ± 0.5 – 27.5 ± 0.5 30.9 ± 0.4a 30.4 ± 0.4a 30.4 ± 0.4a 30.1 ± 0.3a 30.9 ± 0.4a – 31.6 ± 0.5a – – – – 26.5 ± 0.5 – – – – – 26.2 ± 0.2a 25.9 ± 0.2a – – – – 24.1 ± 0.3 23.7 ± 0.5 – – – – – 24.1 ± 0.3a 23.1 ± 0.2a – Monomer (a) – – – – – – – 16.2 ± 0.1a 16.1 ± 0.2a 16.1 ± 0.1a 16.2 ± 0.2a 15.9 ± 0.6a – 15.5 ± 0.7a Monomer (c) – – – – – – – 14.9 ± 0.2a 14.8 ± 0.2a 14.8 ± 0.2a 14.9 ± 0.1a 15.0 ± 0.2a 15.0 ± 0.2a 14.1 ± 0.4a Monomer (b) – – – – – – – 14.4 ± 0.4a 14.0 ± 0.4a 14.0 ± 0.4a 14.3 ± 0.3a 13.6 ± 0.4a 13.3 ± 0.5a Monomer (d) 12.3 ± 0.3 12.6 ± 0.3 – – 12.3 ± 0.2 12.3 ± 0.3 11.7 ± 0.5 12.3 ± 0.7a 11.6 ± 0.7a – – 12.3 ± 0.7a 12.3 ± 0.7a 11.4 ± 0.3a The roman numerals correspond to the peaks (fractions) obtained through gel filtration as shown in Fig. 1A. y an a o s e a r S t e e g t i m s s o i v p t a F t e f o w a T H d t p P b t w 3.2.1. Whole protein In Fig. 3 the MALDI-TOF-MS spectrum of a solution of cyanomet- HbGp, at pH 7.0, using sinapinic acid as a matrix in the positive ion Fig. 3. MALDI-TOF-MS spectrum of cyanomet-HbGp (A) at pH 7.0 without �- a The samples in the presence of �-mercaptoethanol. The errors were obtained b b This column corresponds to the whole native HbGp. bserved for the whole protein and fraction I. However, it is not afe to infer only based on SDS-PAGE electrophoresis, which are xactly the bands corresponding to the linker chains L1, L2, L3 nd L4 that are expected for HbGp, in total analogy with the eported data for HbLt [5,6]. The reason for this is the fact that DS-PAGE electrophoresis results can be considered only quali- atively, not as quantitative data. The mass values estimated by lectrophoresis are very dependent on the protein migration prop- rties through the gel used in SDS-PAGE. The migration in the el is strongly dependent upon the protein hydrodynamic proper- ies, shape and surface charge characteristics. The main difference n subunits MM based on the electrophoresis is observed for the onomers and trimers. Any factor that affects the shape of the ubunits will interfere with their migration in the electrophore- is gel. For this reason, all the molecular masses estimates based n SDS-PAGE eletrophoresis have only a qualitative meaning, serv- ng mostly for comparisons. Besides that, SDS-PAGE is certainly ery important for the assignment of the SEC chromatographic eaks. The fraction V of the chromatogram in Fig. 1A is expected o present a high amount of linker subunits due to its low bsorption at 415 nm associated to a low heme content (see ig. 1B). However, it seems that the linker subunits are dis- ributed differently and spread over the different fractions luted in the chromatography of Fig. 1A. Some of them pre- er to remain associated either to a fraction close to the whole ligomer (fractions I and II in Fig. 1A), while some are eluted ith the trimer in the chromatographic process (fractions IV nd V in Fig. 1A), as observed by SDS-PAGE (Fig. 2A and able 1). Observing Fig. 2A and B, it is clearly seen that the migration of bGp subunits, in the presence of �-mercaptoethanol, is somewhat ifferent, through the gel, producing, in general, a less resolved pat- ern. This could be associated to some additional unfolding of the rotein induced by the reducing agent as compared to the SDS- AGE performed in its absence. For this reason, a small difference etween the masses presented in Table 1 for the same species, in he absence and the presence of reducing agent (lower lines marked ith an asterisk), is observed. verage of data obtained from several runs. 3.2. MALDI-TOF-MS mercaptoethanol. The insert shows the expanded ordinate, corresponding to the intensity in arbitrary units, highlighting the trimer and linkers peaks; (B) the expanded region for the mono-protonated monomer d+ from 15,000 to 19,000 Da; and (C) the expanded region for the double-protonated monomer d2+ from 8000 to 9000 Da. d1, d2, d3, and d4 correspond to monomer d isoforms. 2148 F.A.O. Carvalho et al. / Process Biochemistry 46 (2011) 2144–2151 Table 2 Molecular masses (Da) of Glossoscolex paulistus hemoglobin (HbGp) subunits, at different indicated iron oxidation states, obtained from MALDI-TOF-MS. Subunits Oxy-HbGp Oxy-HbGpa Cyanomet-HbGp Cyanomet-HbGpa Met-HbGp Met-HbGpa d1 16,370 16,360 16,350 ± 15 16,345 ± 20 16,335 ± 20 16,360 ± 4 d2 16,415 ± 10 16,445 ± 20 16,410 ± 15 16,440 ± 30 16,412 ± 20 16,495 ± 20 d3 16,650 ± 40 16,620 16,615 ± 30 16,620 16,630 ± 10 – d4 16,850 ± 40 16,820 16,835 ± 35 16,820 16,795 ± 50 – b – 16,480 – 16,500 – 16,460 ± 40 c1 – 17,330 ± 15 – 17,320 ± 5 – 17,385 ± 10 c2 – 17,410 – 17,413 ± 20 – 17,410 ± 40 c3 – 17,546 – 17,548 ± 10 – 17,558 ± 15 c4 – 17,620 – 17,620 ± 5 – – a – 18,245 ± 20 – 18,245 ± 20 – 18,267 ± 20 L1 25,780 ± 30 25,870 ± 50 25,860 ± 30 25,870 ± 15 25,817 ± 20 25,940 ± 100 L2 26,750 ± 80 26,720 ± 80 26,815 ± 20 26,870 ± 60 26,885 ± 40 26,875 ± 100 2db – 33,870 – 33,710 – 33,985 ± 100 L4 – – – – – – L3 or 2dc 32,870 32,900 32,850 32,910 32,865 ± 80 32,890 ± 80 2c2 – 34,700 – 34,620 – 34,635 ± 20 T1 (abc) 51,200 – 51,140 – 51,220 ± 100 – T2 (abc) 51,985 51,630 51,940 51,400 52,025 ± 100 51,915 ± 70 abcd 68,400 69,280 68,300 68,270 68,275 ± 60 69,675 ± 35 a Samples in the presence of �-mercaptoethanol. m d4. m d2. m p 1 t p d M 1 s I a p a M c M w o m 6 t T b o p o l i i t p w s O p e o r a 1 2 3 monomer species (2d4) and (3) dimers of monomers (2c). In Fig. 4B and C the expanded regions for mono-protonated and double pro- tonated monomers are shown, respectively. These figures show Fig. 4. MALDI-TOF-MS spectrum of cyanomet-HbGp (A) at pH 7.0 with �- mercaptoethanol. The insert shows the expanded ordinate, corresponding to the intensity in arbitrary units, highlighting the trimer and linkers peaks; (B) the b The value of mass corresponds very closely to the dimer of the monomer isofor c The value of mass corresponds very closely to the dimer of the monomer isofor ode, is shown without �-mercaptoethanol. In agreement with revious studies [9], Fig. 3A shows an intense peak centered around 6.3 kDa, corresponding to the monomer d subunit, which is consis- ent with its relatively easy ionization. Besides that, characteristic eaks due to the trimer abc, linkers chains, and tetramer abcd, with ifferent ionization degrees, are also noticed in the spectrum. The M observed for d+ subunits were 16,350 ± 15 Da, 16,410 ± 15 Da, 6,615 ± 30 Da and 16,835 ± 35 Da, corresponding, respectively, to ingle protonation of isoforms d1, d2, d3, and d4 (Fig. 3B and Table 2). n Fig. 3C the monomer d2+ isoforms, double protonated, are shown nd a higher definition of the individual peaks is observed as com- ared to Fig. 3B. The monomer isoform masses in Table 2 are an verage of the values obtained from the spectra in Fig. 3B and C. The M of 25,860 ± 30 Da and 26,815 ± 20 Da are attributed to linker hains L1 and L2 mono-protonated species (Table 2). The observed M values for the peptide chains of cyanomet-HbGp in the present ork are in agreement with the values observed previously for xy-HbGp [9,24]. As can be noticed from Fig. 3A and Table 2, the trimer abc asses of 51,140 Da and 51,940 Da, and the tetramer abcd mass of 8,300 Da, were also observed, implying the existence of, at least, wo isoforms for the trimeric subunit, all of them mono-protonated. he dimer of monomers, (2d2)+, with a mass of 32,850 Da is, proba- ly, superposed to the L3 linker. The fourth linker (L4) is not clearly bserved by MALDI-TOF-MS data, suggesting that this subunit is, robably, less abundant and the contribution of this species is not bserved [5,6]. The linker at 25,860 ± 30 Da could also be over- apped with the double protonated trimer (abc)2+ [5,6]. The small ntensity of the tetrameric subunit peak is, probably, associated to ts low ionization, even smaller than that for the trimer [9]. The rimer subunit isoforms are associated to the variations in com- osition of chains c, which are constituted by several isoforms, as ill be shown below. Literature reports regarding HbLt have also hown that chains a are constituted by different isoforms [5,6]. ur present results show that in the case of HbGp the subunits resenting several isoforms are d and c chains. The MALDI-TOF-MS spectrum for cyanomet-HbGp in the pres- nce of �-mercaptoethanol is shown in Fig. 4. Analysis of the inset f Fig. 4A, focusing the range from 30,000 to 38,000 Da, shows three esolved peaks with the following masses: (1) 32,910, (2) 33,710, nd (3) 34,620 Da. Considering their half value, masses of 16,405, 6,855 and 17,320 Da are obtained. Based on the masses given in Table 2, these peaks could correspond to the following species: (1) dimer of monomers species, (2d ), or linker L , (2) dimers of expanded region for the mono-protonated monomer d+ from 15,000 to 19,000 Da; and (C) the expanded region for the double-protonated monomer d2+ from 8000 to 9000 Da. d1, d2, d3, and d4 correspond to monomer d isoforms. c1, c2, c3, and c4 correspond to monomer c isoforms, obtained upon reduction of trimer abc disulfide bonds. F.A.O. Carvalho et al. / Process Biochemistry 46 (2011) 2144–2151 2149 Fig. 5. MALDI-TOF-MS spectrum of met-HbGp (A) at pH 7.0 without �- mercaptoethanol. The insert shows the expanded ordinate, corresponding to the intensity in arbitrary units, highlighting the trimer and linkers peaks; (B) the expanded region for the mono-protonated monomer d+ from 15,000 to 19,000 Da; a 9 c M a c F t a a s i o m w H 1 m a s a i c i w g i i c p f Fig. 6. MALDI-TOF-MS spectrum of met-HbGp (A) at pH 7.0 with �- mercaptoethanol. The insert shows the expanded ordinate, corresponding to the intensity in arbitrary units, highlighting the trimer and linkers peaks; (B) the expanded region for the mono-protonated monomer d+ from 15,000 to 19,000 Da; and (C) the expanded region for the double-protonated monomer d2+ from 8000 nd (C) the expanded region for the double-protonated monomer d2+ from 8000 to 000 Da. d1, d2, d3, and d4 correspond to monomer d isoforms. learly the multiple isoforms for monomeric chains d and c. The M of subunits c are 17,320 ± 5 Da, 17,413 ± 20 Da, 17,548 ± 10 Da nd 17,620 ± 5 Da (Table 2), corresponding to isoforms c1, c2, c3, and 4, respectively. The MM values of subunit d isoforms are shown in ig. 3B and C, and the values of the other chains b and a, are, respec- ively, 16,500 and 18,245 ± 20 Da (Table 2). Since monomeric chain has a low intensity due to an inefficient ionization, apparently single mass value is obtained. On the other hand, the chain b is uperposed with the chains d and here again a single unique chain s observed. Figs. 5 and 6 show, respectively, the MALDI-TOF-MS spectra f met-HbGp 3.0 mg/mL, in the absence and the presence of �- ercaptoethanol. The general behavior observed for this species as similar to that for the other two forms, oxy- and cyanomet- bGp. However, the peaks resolution, in the region between 5,000 Da and 19,000 Da, for the met-form, in the presence of �- ercaptoethanol, is lower as compared to cyanomet-HbGp (Fig. 6B nd C). In summary, our results suggest that HbGp subunits are very imilar in mass to HbLt, but the chains c and d of HbGp display total of four isoforms, while in HbLt the chains having several soforms are d and a. Our present results correspond to a further haracterization of HbGp subunit masses. However, more detailed nformation is still necessary, especially regarding the linker chains, hich are not so easily isolated in the pure form as seen both in the el filtration and eletrophoretic experiments. Besides that, work is n progress to obtain the subunits sequences. This is certainly an mportant and interesting matter for future work. Overall, globin hains of HbGp are quite similar to those of HbLt, as expected, resenting several isoforms. Their relevance for the hemoglobin unctioning remains another interesting matter for future research. to 9500 Da. d1, d2, d3, and d4 correspond to monomer d isoforms. c1, c2, c3, and c4 correspond to monomer c isoforms, obtained upon reduction of trimer abc disulfide bonds. 3.2.2. Isolated protein fractions In Table 3 the masses corresponding to each protein fraction, obtained from the gel filtration of the whole oxy-HbGp exposed to pH 9.6 (Fig. 1A), are shown. They can be compared to the first col- umn in Table 2, which corresponds to the whole native oxy-HbGp. The fraction I displays a similar composition as the whole protein, suggesting the preservation of the native HbGp structure. The con- tribution of the tetrameric species (abcd)+ is not observed in this fraction, and this could be associated to the smaller sample con- centration as compared to the native HbGp. The presence of a third linker, L3, which is superposed to the dimer of monomers, 2d, is also not clearly observed. Moreover, the masses of the monomeric and trimer isoforms are very similar. It is worthy of notice, that for the two linker chains observed in fraction I, the masses are higher than those reported for the native protein (around 100 Da). Fraction II was analyzed, in the presence and the absence of �- mercaptoethanol. For this fraction, the addition of the reducing agent leads to some broadening of the monomeric peak in the spec- trum precluding the resolution of the four isoforms (see Table 2 for the whole oxy-HbGp). Only the more intense d2 isoform gave a precise mass value (16,470 ± 14 Da). The mass values of the two linker chains L1 and L2 seem to be also somewhat higher (200 Da) in the presence of the reducing agent. The mass of the tetramer also presents the same trend (1 kDa). Based on the masses presented in Table 3, fractions II and III are quite similar and seem to be formed by the same subunits as the whole native protein. The fraction IV is expected to be rich in trimers abc. As can be seen from Table 3 it is quite pure, presenting only three peaks: one at 16,430 Da, corresponding to monomer isoform d2, some con- 2150 F.A.O. Carvalho et al. / Process Biochemistry 46 (2011) 2144–2151 Table 3 Molecular masses (Da) of Glossoscolex paulistus hemoglobin (HbGp) subunits, obtained from MALDI-TOF-MS, for the fractions obtained in the gel filtration chromatography as shown in Fig. 1A. Subunits Fractions (masses in Da) I II IIa III IV VI d1 16,350 16,332 ± 21 – 16,334 ± 23 – 16,330 ± 10 d2 16,423 ± 18 16,406 ± 19 16,470 ± 14 16,410 ± 14 16,430 16,405 ± 17 d3 16,638 ± 25 16,605 ± 37 – 16,627 ± 19 – 16,627 ± 15 d4 16,860 16,825 ± 20 – 16,832 ± 33 – 16,826 ± 11 b – – – – – – c1 – – 17,365 ± 35 – – – c2 – – – – –– c3 – – 17,556 ± 20 – – c4 – – 17,750 ± 50 – – a – – 18,265 ± 35 – – – L1 25,970 ± 25 25,897 ± 84 26,115 ± 50 25,940 25,820 – L2 26,860 ± 60 26,797 ± 26 27,017 ± 4 26,630 ± 90 26,620 ± 100 – 2db – – 34,045 – – 33,885 ± 200 L4 – – – – – – L3 or 2dc – 33,270 ± 100 33,190 33,180 – 32,740 ± 200 2c2 – – 34,820 – – – 3d 49,700 T1 (abc) 51,230 ± 30 51,200 ± 45 – 51,080 ± 45 – – T2 (abc) 51,940 ± 60 51,950 ± 40 51,590 51,875 ± 100 52,270 – abcd – 68,280 69,260 68,260 – – The roman numerals correspond to the peaks (fractions) obtained through gel filtration as shown in Fig. 1A. a Samples in the presence of �-mercaptoethanol. b The value of mass corresponds very closely to the dimer of the monomer isoform d4. c The value of mass corresponds very closely to the dimer of the monomer isoform d2. Fig. 7. MALDI-TOF-MS spectrum of fraction VI, obtained from native HbGp in the oxy-form, through chromatography on a Superdex 200 column as shown in Fig. 1, and corresponding to the pure monomer d: (A) whole MS spectrum; the insert shows the expanded ordinate, corresponding to the intensity in arbitrary units, highlight- ing the trimer and linkers peaks; (B) the expanded region of masses from 16,000 to 18,500 Da for the mono-protonated monomer d+; and (C) the expanded region from 8000 to 9000 Da for the double-protonated monomer d2+. d1, d2, d3, and d4 correspond to monomer d isoforms. tamination at 25,820 Da, 26,620 ± 100 Da, which can be assigned to linker chains L1 and L2, respectively. The highest contribution in this fraction is the peak at 52,270 Da, assigned to the trimer abc. This peak is quite intense and the two isoforms are not well resolved. In our previous work [9], it was noticed that the monomers are easily ionized, while the trimers do not behave in the same way. In fact, the fraction associated to the trimers is not very stable under conditions of low salt concentration, required for MALDI-TOF experiments, and tend to precipitate upon long standings. Our results regarding the fraction IV are consistent with gel filtration data as well as the electrophoresis analysis. Moreover, no contributions of tetramer abcd and dimer of monomers 2d are observed for this fraction. The contamination of fraction IV with linkers is observed in the gel presented in Fig. 2A, and supports the observation of the strong interaction of linkers with the abc trimer. Our data show that fraction VI corresponds to pure monomer d, where only the contributions of its several mono-protonated iso- forms, in the mass range from 15,000 Da to 19,000 Da, dimers of monomers in the range between 30,000 and 35,000 Da, and trimers of monomers around 49,700 Da (Fig. 7), are observed, suggesting that peak VI in Fig. 1A corresponds to monomer d in a very pure form. These results are also consistent with the observed SDS-PAGE electrophoresis data (see Fig. 2A). The fraction V, expected to be quite abundant in several linker chains, was very diluted and its low concentration was inadequate for mass spectrometry analysis. 4. Conclusions In order to evaluate the potential usefulness of such a giant protein such as HbGp for biomedical applications, a detailed knowl- edge of its constituent subunits is very relevant. In the present study, further characterization of G. paulistus hemoglobin was performed, through the use of size exclusion chromatography, SDS-PAGE electrophoreses and MALDI-TOF-MS. The gel filtration chromatography of HbGp, submitted to alkaline oligomeric dis- sociation, is capable to produce the separation of several protein subunits. The electrophoresis analysis of the fractions produced in Bioche S u t c e h g i t m s O t a t c a i p o H i H A M S a t R t a d fi t F g f c B R [ [ [ [ [ [ [ [ [ [ [ [ [ [ Spectroscopic studies of the met form of the extracellular hemoglobin from F.A.O. Carvalho et al. / Process EC experiments, indicate that HbGp is composed by many sub- nits. Moreover, our present results show an important progress in he HbGp subunits characterization, especially regarding the diffi- ulties encountered to separate completely the linker chains by SEC xperiments. It appears clear from our data that the linker chains ave a very high affinity for some globin chains, especially the lobin trimers abc. On the other hand, the monomer d (VI) fraction s indeed quite pure, while the fraction rich in trimer (IV) is con- aminated by the presence of linker chains and a small amount of onomers. Although, the remaining fractions present a mixture of ubunits, a detailed characterization was made for these subunits. ur new MALDI-TOF-MS analysis performed in this work showed hat the monomer c presents four isoforms and the trimer, abc, is lso characterized by two isoforms, T1 and T2. In addition, the con- ribution of the dimer of monomers was observed for two globin hains, d and c, and the presence of the tetramer, abcd, with MM t 68,400 kDa was detected. Finally, HbGp hemoglobin, at different ron oxidation states, is very similar regarding its subunits com- osition. Our present studies are consistent with literature reports n several other extracellular hemoglobins, such as the HbLt and bAm. We believe our results represent a nice contribution and an mportant and necessary step in the complete characterization of bGp oligomeric structure. cknowledgments The authors are grateful to Prof. Júlio C. Borges from Biologia olecular e Bioquímica Laboratory (BMB), Instituto de Química de ão Carlos, Universidade de São Paulo, São Carlos, Brazil, for making vailable the gel electrophoreses facility used in the initial part of his work. The authors are indebted to M.Sc. José Fernando Bachega uggiero for many discussions and interest in this work, related o the determination of subunits structure and sequences for the tomic resolution of the HbGp crystal structure. Thanks are also ue to the Brazilian agencies FAPESP, CNPq, and CAPES for partial nancial support. P.S. Santiago is grateful to FAPESP for postdoc- oral grants. F.A.O. Carvalho is the recipient of a PhD. grant from APESP (2009/17261-6). J.W.P. Carvalho is the recipient of a PhD. rant from FAPESP (2010/09719-0). M. Tabak is grateful to CNPq or a research grant. Thanks are also due to Mr. Ézer Biazin for effi- ient support in the sample preparations and SEC experiments. Ézer iazin is grateful to CNPq for a technical support grant. eferences [1] Weber RE, Vinogradov SN. Nonvertebrate hemoglobins: functions and molec- ular adaptations. Physiol Rev 2001;81:570–611. [2] Weber RE, Zhu H, Gorr TA, Flores JF, Knapp JE. Allosteric hemoglobin assembly: diversity and similarity. J Biol Chem 2005;280:27477–80. [3] Vinogradov SN. The stoichiometry of the four linker subunits of Lum- bricus terrestris hemoglobin suggests an asymmetric distribution. Micron 2004;35:127–9. [4] Rousselot M, Delpy E, La Rochelle CD, Lagente V, Pirow R, Rees J, et al. Arenicola marina extracellular hemoglobin: a new promising blood substitute. Biotech- nol J 2006;1:333–45. [ mistry 46 (2011) 2144–2151 2151 [5] Zhu H, Ownby DW, Riggs CK, Nolasco NJ, Stoops JK, Riggs AF. Assembly of the gigantic hemoglobin of the earthworm Lumbricus terrestris—roles of subunit equilibria, non-globin linker chains, and valence of the heme iron. J Biol Chem 1996;271:30007–21. [6] Fushitani K, Riggs AF. The extracellular hemoglobin of the earthworm, Lum- bricus terrestris—oxygenation properties of isolated chains, trimer, and a reassociated product. J Biol Chem 1991;266:10275–81. [7] Harnois T, Rousselot M, Rogniaux H, Zal F. High-level production of recom- binant Arenicola marina globin chains in Escherichia coli: a new generation of blood substitute. Artif Cell Blood Sub 2009;37:106–16. [8] Royer WE, Strand K, Van Heel M, Hendrickson WA. Structural hierar- chy in erythrocruorin, the giant respiratory assemblage of annelids. PNAS 2000;97:7107–11. [9] Oliveira MS, Moreira LM, Tabak M. Partial characterization of giant extracel- lular hemoglobin of Glossoscolex paulistus: a MALDI-TOF-MS study. Int J Biol Macromol 2007;40:429–36. 10] Zal F, Green BN, Martineau P, Lallier FH, Toulmond A, Vinogradov SN, et al. Polypeptide chain composition diversity of hexagonal-bilayer hemoglobins within a single family of annelids, the alvinellidae. Eur J Biochem 2000;267:5227–36. 11] Meirelles NC, Oliveira B, de Paula E, Marangoni S, Rennebeck GM. Erythro- cruorin of Glossoscolex paulistus (oligochaeta, glossoscolecidae)—dissociation at alkaline pH and its ligand properties as revealed by chemical, immunochemical and electron-microscopy studies. Comp Biochem Physiol 1987;88A:337–79. 12] Carvalho FAO, Santiago PS, Borges JC, Tabak M. On the molecular mass of the extracellular hemoglobin of Glossoscolex paulistus: analytical ultracentrifuga- tion reexamination. Anal Biochem 2008;385:257–63. 13] Santiago PS, Moura F, Moreira LM, Domingues MM, Santos NC, Tabak M. Dynamic light scattering and optical absorption spectroscopy study of pH and temperature stabilities of the extracellular hemoglobin of Glossoscolex paulis- tus. Biophys J 2008;94:2228–40. 14] Daniel E, Lustig A, David MM, Tsfadia Y. Towards a resolution of the long-standing controversy regarding the molecular mass of extracellular ery- throcruorin of the earthworm Lumbricus terrestris. Biochim Biophys Acta 2003;1649:1–15. 15] Ownby DW, Zhu H, Schneider K, Beavis RC, Chait BT, Riggs AF. The extracellular hemoglobin of the earthworm, Lumbricus-terrestris—determination of subunit stoichiometry. J Biol Chem 1993;268:13539–47. 16] Vinogradov SN, Lugo S, Mainwaring MG, Kapp OH, Crewe AV. Bracelet pro- tein: a quaternary structure proposed for the giant extracellular hemoglobin of Lumbricus terrestris. Proc Natl Acad Sci USA 1986;83:8034–8. 17] Royer WE, Sharma H, Strand K, Knapp JE, Bhyravbhatla B. Lumbricus erythro- cruorin at 3.5 angstrom resolution: architecture of a megadalton respiratory complex. Structure 2006;14:1167–77. 18] Royer WE, Strand K, Van Heel M, Hendrickson WA. Structural hierarchy in ery- throcruorin, the giant respiratory assemblage of annelids. Proc Natl Acad Sci 2000;97:7107–11. 19] Bachega JFR, Bleicher L, Horjales ER, Santiago PS, Garratt RC, Tabak M. Crystallization and preliminary structural analysis of the giant haemoglobin from Glossoscolex paulistus at 3.2 angstrom. J Synchrotron Radiat 2011;18: 24–8. 20] Santiago PS, Moreira LM, de Almeida EV, Tabak M. Giant extracellular Glos- soscolex paulistus hemoglobin (HbGp) upon interaction with cethyltrimethy- lammonium chloride (CTAC) and sodium dodecyl sulphate (SDS) surfactants: dissociation of oligomeric structure and autoxidation. Biochim Biophys Acta 2007;1770:506–17. 21] Poli AL, Moreira LM, Tabak M, Imasato H. SDS (sodium dodecyl sulfate) effect on the autoxidation of the Glossoscolex paulistus giant extracellular hemoglobin: kinetic studies at pH 7.0 and 9.0. Colloids Surf B 2006;52:96–104. 22] Agustinho SCM, Tinto MH, Perussi JR, Tabak M, Imasato H. Fluorescence studies of extracellular hemoglobin of Glossoscolex paulistus in met form obtained from sephadex gel filtration. Comp Biochem Physiol 1998;118A:171–81. 23] Agustinho SCM, Tinto MH, Imasato H, Tominaga TT, Perussi JR, Tabak M. Glossoscolex paulistus. Biochim Biophys Acta 1996;1298:148–58. 24] Carvalho FAO, Santiago PS, Borges JC, Tabak M. Molecular masses and sedi- mentation coefficients of extracellular hemoglobin of Glossoscolex paulistus: alkaline oligomeric dissociation. Int J Biol Macromol 2011;48:183–93. Further characterization of the subunits of the giant extracellular hemoglobin of Glossoscolex paulistus (HbGp) by SDS-PAG... 1 Introduction 2 Materials and methods 2.1 Protein extraction and purification 2.2 Size exclusion chromatography (SEC) 2.3 Gel electrophoresis 2.4 Matrix-assisted laser desorption/ionization time-of-flight/mass spectrometry – MALDI-TOF-MS 3 Results and discussions 3.1 Size exclusion chromatography (SEC) and electrophoresis – SDS-PAGE 3.2 MALDI-TOF-MS 3.2.1 Whole protein 3.2.2 Isolated protein fractions 4 Conclusions Acknowledgments References