HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 24, 21898-21905, June 14, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/24/21898    most recent M111492200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Mito, M. Articles by Morimoto, H. Search for Related Content PubMed PubMed Citation Articles by Mito, M. Articles by Morimoto, H. Social Bookmarking                      What's this?

Crystal Structures of Deoxy- and Carbonmonoxyhemoglobin F1 from the Hagfish Eptatretus burgeri^*

Megumi Mito, Khoon Tee Chong^§, Gentaro Miyazaki, Shin-ichi Adachi^¶, Sam-Yong Park, Jeremy R. H. Tame^**, and Hideki Morimoto

From the  Division of Biophysical Engineering, Graduate School of Engineering Science, Osaka University, Toyonaka 560-8531 Osaka, Japan, the ^§ Institute for Protein Research, Osaka University, Suita 560-0871 Osaka, Japan, the ^¶ RIKEN Harima Institute/SPring8, 1-1-1 Kouto, Mikazuki-cho, Sayo, Hyogo 679-5148, Japan, and the  Protein Design Laboratory, Yokohama City University, Tsurumi, Suehiro 1-7-29, Kanagawa 230-0045, Japan

Received for publication, December 3, 2001, and in revised form, March 25, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Hagfish are extremely primitive jawless fish of disputed ancestry. Although generally classed with lampreys as cyclostomes ("round mouths"), it is clear that they diverged from them several hundred million years ago. The crystal structures of the deoxy and CO forms of hemoglobin from a hagfish (Eptatretus burgeri) have been solved at 1.6 and 2.1 Å, respectively. The deoxy crystal contains one dimer and two monomers in a unit cell, with the dimer being similar to that found in lamprey deoxy-Hb, but with a larger interface and different relative orientation of the partner chains. Ile(E11) and Gln(E7) obstruct ligand binding in the deoxy form and make room for ligands in the CO form, but no interaction path between the two hemes could be identified. The BGH core structure, which forms the ₁₁ interface of all vertebrate ₂₂ tetrameric Hbs, is conserved in hagfish and lamprey Hbs. It was shown previously that human and cartilaginous fish Hbs have independently evolved stereochemical mechanisms other than the movement of the proximal histidine to regulate ligand binding at the hemes. Our results therefore suggest that the formation of the ₂₂ tetramer using the BGH core and the mechanism of quaternary structure change evolved between the branching points of hagfish and lampreys from other vertebrates.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In human hemoglobin, the interactions between a ligand bound to a heme and the globin structure can be divided into proximal effects (interactions with the proximal histidine imidazole group and the heme upon ligand binding) and distal effects (interactions with the distal amino acids). The proximal effect moves the F helix and FG corner upon ligand binding and connects the heme to the structure change at the ₁₂ interface associated with the quaternary structure change. The distal effect is more important in the  subunit than in the  subunit (1). When the x-ray structures of two cartilaginous fish Hbs were solved (2, 3), the quaternary structure and its change upon ligand binding were preserved, as expected. The proximal effect linking the quaternary structure change to the hemes was also preserved, but the distal effect, especially the role of Val(E11), was altered. The Bohr proton sites and the 2,3-diphosphoglycerate-binding site were not conserved, indicating that stereochemical mechanisms other than the proximal effect have evolved independently in the different species.

Hbs from hagfish and lampreys are not of the ₂₂ tetramer type. The evolution of the tetramer, including the duplication and differentiation of the globin gene into the  and  genes, the formation of the subunit interfaces, and the mechanism of the quaternary structure change, took place within a comparatively short period between the branching points of hagfish and lampreys from cartilaginous fish. Some primordial globin structures may well have appeared before tetrameric hemoglobin that facilitated its development, and these might be found in modern hagfish and lamprey Hbs. In the case of lamprey Hbs, the dissociation of ligands promotes oligomerization of the protein. On the basis of a monomer structure of the cyanide-bound ferric form, Honzatko and Hendrickson (4) suggested that the subunit interfaces of the deoxy-Hb oligomer were the same as the ₁₂ interface of the ₂₂ tetramer. Recently, however, Heaslet and Royer (5) solved the x-ray structure of lamprey deoxy-Hb and found a dimer with an interface between the E helix and AB corner, totally different from any of the interfaces in the ₂₂ tetramer.

The evolutionary history of hagfish is a much disputed topic. Some sequence comparisons support the view that hagfish and lampreys belong in a single group, the cyclostomes, separate from jawed vertebrates (6, 7). Other morphological and sequence analyses consider hagfish and lampreys to be phylogenetically distant groups and do not class them together (8, 9). Hagfish are distinguished by their lack of bone or any trace of vertebrae, but other anatomical features support grouping them with lampreys. Whether or not the 45 living species of hagfish form a distinct monophyletic group (Myxiniformes) separate from all other craniates, it is clear that they diverged from lampreys (Petromyzontiformes) around 500 million years ago. Grouping hagfish and lampreys together has been supported by molecular studies of globin (10), but the similarities in these sequences may reflect a primitive form present before hagfish diverged (11).

Despite the apparent gulf of time separating them, Hbs from hagfish and lampreys show similar linkage relations between oligomerization and ligand binding (12, 13). According to Bannai et al. (14), Eptatretus burgeri has four Hb components designated F1, F2, F3, and F4. We have purified F1 and solved the x-ray structures of the CO and deoxy forms at 2.1 and 1.6 Å, respectively. In this study, we have analyzed the evolutionary changes that occurred in Hb around the branching point of vertebrates and invertebrates by comparing not only the structures, but also the structure changes upon CO binding.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Protein Purification-- Hagfish (E. burgeri) were captured in the sea near the Miura Peninsula in Japan. The hagfish studied by Bannai et al. (14) were captured in the same area. Blood was collected and lysed as described by Chong et al. (2). The hemolysate was equilibrated with 2 mM Tris-HCl (pH 8.5) and then deionized by passage through Amberlite MB-3. After dialysis against 5 mM Tris-HCl (pH 8.5), the hemolysate was fractionated on a Whatman DE32-cellulose column equilibrated with the same buffer as described (14). The Hb was eluted with a linear gradient of KCl from 0 to 0.1 M in 5 mM Tris-HCl (pH 8.0). Three peaks were obtained and checked by starch gel electrophoresis. According to Bannai et al. (14), there are four Hb components in this hagfish, designated F1, F2, F3, and F4. We have used the same nomenclature in this study. The results of our column chromatography and electrophoresis were very similar to those described in 1972 by Bannai et al. (14). The first eluted peak was F1, the second was F2 (with a slight contamination of F3), and the third was F3 and F4. The approximate ratio of the amounts of each peak (F1/F2/F3 + F4) was 2:1:3. The concentration of hagfish hemoglobin F1 was calculated by assuming the same millimolar extinction coefficient as human HbCO A, 14.5 mM/cm at 539 nm.

cDNA Sequence Analysis-- Three 5'-primers were used for PCR, two of them corresponding to residues 1-7 and 10-17 of Korean hagfish sequenced by Professor Yazawa (Hokkaido University of Education).¹ The third one corresponds to residues 49-55, which are well conserved among cyclostome Hbs. cDNA sequence analysis was carried out as described by Chong et al. (2). Only three cDNA sequences were identified. The primer corresponding to residues 10-17 of Korean hagfish picked up a sequence that is fully consistent with our 1.6-Å resolution structure of E. burgeri deoxy-Hb, the F1 component; the other two sequences have some differences.

Oxygen Equilibrium Measurement-- Oxygen equilibrium curves were measured as described by Imai et al. (15). Measurement conditions were 60 µM protein in 50 mM bis-Tris² and 50 mM Tris buffer containing 100 mM chloride at 25 °C. The pH value was adjusted with concentrated NaOH. To minimize the autoxidation of hemoglobin during measurements, catalase and superoxide dismutase were added to each sample (16, 17). Deoxygenation curves were used to determine the p₅₀ (partial pressure of oxygen at half-saturation) and Hill coefficient (n_H; the maximum slope of Hill plots of oxygen equilibrium curves).

Analytical Gel Chromatography-- A 230 × 8-mm Superdex 75 column (Amersham Biosciences) was employed for analytical gel chromatography with 100 mM bis-Tris-HCl (pH 6.5) at 25 °C, the same pH as that used for crystallization. In the case of deoxy-Hb, the same column was used under N₂ with 1 mM Na₂S₂O₄. 0.3-ml samples were loaded onto the column. The time difference between the elution peak of the sample and blue dextran (Amersham Biosciences) was determined at different starting concentrations and calibrated using horse myoglobin and cross-linked human Hb. For HbCO, the starting concentrations of Hb were 100, 300, and 630 µM. For deoxy-Hb, they were 10, 60, 450, and 900 µM. The sample volume of 0.3 ml is a compromise between the uncertainty of the concentration and the broadening of the elution pattern as well as the amount of the sample required for measurement at the higher concentration range. In the case of E. burgeri Hb, however, the elution volume altered very little with protein concentration.

Crystallization and Data Collection-- Both deoxy and CO crystals were grown at 27 °C by batch crystallization under an atmosphere of nitrogen or carbon monoxide, respectively, using screw-top vials. Deoxy crystals were obtained with 19% polyethylene glycol 4000, 10% glycerol, 50 mM bis-Tris-HCl (pH 6.5), 20 µM 2-mercaptoethanol, 100 mM NaCl, and 1.5% protein. Data were collected from a cryo-cooled crystal at RIKEN beam line-2 (BL44B2) of the SPring8 Synchrotron in Harima, Japan (18), using a MARCCD detector. The data were processed with MOSFLM (19) and scaled with SCALA (20). The CO crystals were obtained with 21% polyethylene glycol 4000, 10% glycerol, 50 mM bis-Tris-HCl (pH 6.5), 20 µM 2-mercaptoethanol, 100 mM NaCl, and 1.5% protein. Data were collected at BL44B2 at cryogenic temperature, processed with DENZO, and scaled with SCALEPACK (21).

Structure Determination and Refinement-- Both structures were solved by molecular replacement using X-PLOR (22). The initial model for the CO structure was lamprey deoxy-Hb monomer mutated to the hagfish amino acid sequence (5). Using data between 15 and 5 Å, the orientations and positions of four molecules in an asymmetric unit were found and improved by rigid-body refinement. The refined CO monomer structure was used as a search model for the deoxy structure. The initial models were refined with a simulated annealing protocol (23) using a bulk solvent correction (24) with data between 20.0- and 2.1-Å resolution (CO form) or 20.0- and 1.6-Å resolution (deoxy form). After each cycle of refinement, the models were manually adjusted using the program TURBO. In the CO form, the electron density map corresponding to the residues between positions 54 and 61 is comparatively weak. As described above, there are relatively large structural differences among the four subunits, and the quality of the electron density in this region differs from one molecule to another. As the connectivity of the 2F_o  F_c electron density map of the main chain is maintained at 1.2 in some of the subunits even in the worst region (residues 58-60), we have modeled these residues in all subunits. In the deoxy structure, there were no difficulties in modeling this region. Isolated electron density greater then 3 in F_o  F_c maps and 1.3 in 2F_o  F_c maps was modeled as water molecules if the locations were sterically reasonable. Crystallographic and refinement data are shown in Table I.

View this table: [in this window] [in a new window]   Table I Data collection and structure refinement statistics Rmerge = hi|I(h,i)  I(h)|/hi|I(h,i)|, where I(h,i) is the intensity value of the ith measurement of h and I(h) is the corresponding mean value of I(h,i) for all i measurements; the summation is over the reflections with I/(I) > 1.0. Rfree is the R factor calculated for 5% of reflections that were randomly selected and were excluded from the refinement. Rfactor = Fo|  |Fc/|Fo|, where Fo is the observed structure factor and Fc is that calculated from the model.

Comparison of Hb Structures-- The coordinates of different Hbs were obtained from the Protein Data Bank: human deoxy-Hb A, code 1bz0; human deoxy-Hb ₄, code 1cbm; human HbCO ₄, code 1cbl; shark deoxy-Hb, code 1gcv; shark HbCO, code 1gccw; sperm whale (Physeter catodon) deoxymyoglobin, code 1bzp; sperm whale MbCO, code 1bzr; sea lamprey (Petromyzon marinus) deoxy-Hb, code 3lhb; sea lamprey ferricyano-Hb, code 2lhb; blood clam (Scapharca inaequivalvis) deoxy-Hb, code 4hbi; blood clam HbCO, code 3hbi; marine blood worm (Glycera dibranchiata) deoxy-Hb, code 2hbg; and marine blood worm HbCO, code 1hbg. The high-resolution model of human HbCO A has been refined by Park et al.³

The buried surface area between intermolecular interfaces (half the difference in solvent accessibility between the independent and assembled partner molecules) was calculated using AREAIMOL, part of the CCP4 suite (20). To compare Hbs, the coordinates were superimposed either on the heme or on appropriate parts of helices. The heme frame contains 28 atoms: 4 nitrogen and 20 carbon atoms of the porphyrin ring and 4 methyl carbons. To compare human and hagfish Hbs, we defined a helix frame consisting of the central regions of the helices, A5-A15, B8-B16, C1-C7, CD1, E7-E18, F1-F9, G4-G14, and H8-H17 in human - and -globins. F1-F9 and H8-H17 in human Hb correspond to F7-F15 and H1-H10 in hagfish and lamprey Hbs (see Table II), respectively. The BGH frame was originally defined by Baldwin and Chothia (25) as the conserved structural core of human Hb at the ₁₁ interface. For Glycera Hb and sperm whale Mb, the sequences were aligned according to Lesk and Chothia (26), and for blood clam, according to Royer et al. (27).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Overall Structure-- Although according to Bannai et al. (14) E. burgeri Hb has four components, we were able to identify only three cDNAs, including F1. We have determined the sequence of F1, which is consistent with the crystal structure. It has 146 amino acid residues, 2 less than Hb III from another hagfish, Myxine glutinosa (Table II). The sequence identity to M. glutinosa Hb III is 61%; that to the  and  chains of human Hb is ~20%; and that to lamprey (P. marinus) Hb V is ~40%. The EF corner and G and H helices are shorter than their counterparts in human Hb by 2, 3, and 7 residues, respectively; the F helix is longer by 6 residues, starting earlier in the sequence. In common with lamprey Hb V (28), hagfish Hb has a long N-terminal tail (NA region) with 11 residues attached to the A helix and a 10-residue deletion from the tail of the G helix to the start of the H helix. The NA region is only 2 or 3 residues long in all other vertebrate Hbs. Sea cucumber (Caudina arenicola) and blood clam (S. inaequivalvis) Hbs also have long N-terminal tails, but they are dissimilar to hagfish and lamprey Hbs in both primary and tertiary structure (29, 30). F1 has only a short D helix (residues 61-63 inclusive), whereas lamprey Hbs have a D helix with 7 residues (28).

View this table: [in this window] [in a new window]   Table II Sequence alignment of F1 with hemoglobins from the hagfish M. glutinosa (Hbs I, IIA, and III) and lamprey P. marinus (Hb V) and human - and -globins The residue numbers are those of hagfish Hb F1. Helices are indicated by name (A-H); the helix designations above the sequences are those of E. burgeri Hb F1, and those below are those of the human  chain.

The notable mutations among F1 heme contact residues are Gln⁷¹(E7), Ile⁷⁵(E11), and Phe¹⁰⁷(FG3), which replace well conserved His, Val, and Leu residues, respectively, among vertebrate ₂₂-type Hbs. In M. glutinosa Hb III, E7 is also Gln; but in lampreys, it is His. Ile⁷⁵(E11) and Phe¹⁰⁷(FG3) are common to all known sequences of lamprey and hagfish Hbs (13).

Properties in Solution-- The average molecular mass of F1 was found by analytical gel filtration to be 19 kDa in the CO form and 23 kDa in the deoxy form (data not shown). The oxygen equilibrium characteristics are presented in Table III. The hagfish hemolysate showed lower oxygen affinity than the human Hb hemolysate, and purified F1 showed lower oxygen affinity still. Although for the hemolysate the apparent Hill coefficient (n_H) is 1.0, for F1, the maximum n_H is ~1.3 at pH 7.4, which shows that a small but significant cooperativity is present in this Hb. The Bohr effect is weak. Bannai et al. (14) reported oxygen equilibrium properties of the same Hb under different conditions (neutral pH in 0.1 M phosphate buffer at 22 °C); and therefore, it is not possible to compare their results with ours directly, but they also showed the Bohr effect to be comparatively small and that F1 had lower oxygen affinity than the hemolysate. The oxygen affinity of M. glutinosa Hb seems to be a little higher (12).

Crystal Packing-- The overall structure of hagfish hemoglobin F1 is very similar to that of lamprey Hb V. In the deoxy crystal, the asymmetric unit consists of two F1 monomers. One of them, designated d1, forms a large contact (855 Å²) with a symmetry-related partner in the unit cell (d1'); and the other one, designated d2, has a smaller contact (330 Å²) with d1 using a different patch of the surface. d2 has only minor contacts with its partner in the same unit cell (Fig. 1). No compact tetramers are formed; instead, the unit cell apparently consists of a dimer and two monomers of F1. F1 does not strongly self-associate in solution (14), in common with Hbs from the M. glutinosa hagfish (12).

View larger version (42K): [in this window] [in a new window]   Fig. 1.   Stereo view of crystal packing in the E. burgeri deoxy-Hb F1 crystal. The asymmetric unit consists of molecules d1 and d2. d1' and d2' are symmetry-related molecules in the same unit cell. The b-c plane of the crystal is in the plane of the paper with the b axis vertical.

In the CO form, the asymmetric unit of the crystal consists of four F1 monomers, designated c1-c4 (Fig. 2). In this crystal form, there are no pairs of closely fitting monomers related to each other by 2-fold symmetry, but the contact surface areas between monomer molecules are comparatively large (31). The largest is that between c3 of one asymmetric unit and c2 of another (c2" in Fig. 2), with a contact area of 494 Å². Hb forms a variety of crystal types containing 6-12 monomers in an asymmetric unit. In the case of lamprey deoxy-Hb V, the asymmetric unit contains six dimers arranged around two approximate 3-fold screw axes into hexamers (5), so the crystal packing of both lamprey and hagfish Hbs is rather unusual.

View larger version (58K): [in this window] [in a new window]   Fig. 2.   Stereo view of the crystal packing in the HbCO F1 crystal. The asymmetric unit consists of four molecules: c1-c4. c1' and c2' are obtained by a translation of c1 and c2 by one cell length along the c axis. c1"-c4" are obtained by screw symmetry along the b axis. Regarding c1-c4 as a single tetramer, they can be superimposed on c4, c3, c2', and c1' by rigid-body transformation with a r.m.s. deviation of 1.3 Å.

There are relatively large differences between d1 and d2 in the deoxy form and between c1, c2, c3, and c4 in the CO form compared with the differences between crystallographically nonequivalent  or  subunits in human Hb A (32). Comparing monomers in the CO form, the r.m.s. deviation of all the main chain atoms varies from 0.25 Å (c1-c4) to 0.42 Å (c1-c3). The main chain r.m.s. deviation between d1 and d2 in the deoxy form is 0.54 Å. Most of these differences probably arise from the large number of non-crystallographic contacts with neighboring molecules. In this study, c1 and d1 have been used as canonical structures to draw figures and to make tables.

Comparison between the Deoxy-Hb Dimers of Lampreys and Hagfish-- Hagfish deoxy-Hb F1 and lamprey deoxy-Hb V form dimers using a similar interface, which includes the AB corner and the N terminus of the E helix. In the hagfish dimer, the distance between the two F helices is shorter than found in lamprey Hb (Fig. 3), and the contact area is bigger (855 versus 478 Å²) owing to the participation of the E helix C-terminal end and the F helix in the interface. A number of main contact positions of lamprey Hb, including Tyr(A16), Tyr(B2), Glu(B3), and Asn(E13), are also found in the hagfish Hb dimer interface, but E2, E5, E6, and E9 are mutated to Pro, Lys, His, and Val, respectively (Table II).

View larger version (54K): [in this window] [in a new window]   Fig. 3.   Comparison of the E. burgeri deoxy-Hb F1 dimer (A) with the lamprey Hb V dimer (B) (Protein Data Bank code 3lhb). The E and F helices and AB corners are indicated. Lamprey deoxy-Hb has a more open conformation in the upper half of the interface, and the F helices do not contribute to dimer formation.

Comparing hagfish Hb F1 and lamprey Hb V, only a few amino acid changes seem to be responsible for the different orientation of the partner chains in the deoxy dimer. In hagfish Hb, Tyr²⁷(A16) forms a hydrogen bond with Glu⁶⁷(E3). This Glu becomes Asp in lamprey Hb, too short to reach the tyrosine. Asn⁷⁷(E13) in hagfish Hb is preserved in lamprey Hb (as Asn⁷⁹ due to a 2-residue insertion around residue 60); but in the former, it hydrogen-bonds to His⁷⁰(E6). Lamprey Hb has tryptophan at position 72 (equivalent to hagfish His⁷⁰), so the pattern of intersubunit bonds is quite different. This histidine-to-tryptophan mutation may be more important than any other in altering the dimer packing, as the tryptophan is too bulky to fit into the hagfish interface. It pushes Asn⁷⁹(E13) of the partner chain ~6 Å away from the position of the equivalent Asn⁷⁷(E13) in the hagfish dimer (Fig. 4). Asn(E13) lies close to the dimer axis in the hagfish structure, with the two equivalent side chains lying within 4 Å of each other. In the lamprey dimer, Asn(E13) is >12 Å from its equivalent in the partner chain, and Trp⁷²(E6) instead comes very close to the dimer axis. Hagfish Hb F1 also has a pair of symmetrical hydrogen bonds across the dimer interface formed between Lys⁹⁴(F5) and Asp⁹⁸(F9), whereas the lamprey protein has methionine instead of lysine at position 94 (Table II). However, these bonds cannot form without the F helices of the partner chains being brought closer together than they are in the lamprey Hb dimer. The principal changes that rotate one chain relative to the other seem to be the mutations in the E helix that create interactions with the AB corner not found in the lamprey protein. This rotates the chains, bringing the F helices together. If one subunit of the lamprey deoxy dimer is superposed on one subunit of the hagfish deoxy dimer, it can be seen that the partner chains in each dimer are shifted relative to each other by an almost pure rotation of 29° with a screw translation of <0.6 Å (Fig. 4). This rotation axis lies at 73° to the rotation axis relating the partner chains of the hagfish dimer. From the structure, the hagfish dimer seems rather more stable than the lamprey dimer. However, cooperativity of oxygen binding will obviously depend on the change in oxygen affinity between the monomer and dimer forms as well as the stability of the dimer in the absence of heme ligands.

View larger version (37K): [in this window] [in a new window]   Fig. 4.   Superposition, shown in stereo, of the deoxy dimers of lamprey Hb V and Hb F1. The proteins are overlapped on the main chain atoms of the helices in one subunit, shown on the right. The partner chains on the left adopt quite different positions due to changes in the contact residues in the E helix. The only side chains shown are E6 on the right and E13 on the left. In hagfish Hb F1 (shown in blue), His70(E6) hydrogen-bonds to Asn77(E13). Replacing His70(E6) in hagfish Hb with Trp72(E6) in lamprey Hb (shown in red) forces Asn79(E13) of the partner chain into a position lying flat against the indole ring. The different interactions between E6 and E13 across the dimer interface appear to be the main reason for the observed relative rotation of the partner chains. Despite the highly conserved tertiary structure, the quaternary structure is therefore significantly altered. Comparing main chain atoms of the BGH frame, the r.m.s. difference between the lamprey deoxy monomer and the F1 monomer is 0.61 Å.

The self-association of different Hb components from M. glutinosa has been studied using analytical ultracentrifugation by Fago et al. (33). These experiments showed that Hbs I and III do not self-associate, but that Hb IIA forms dimers with itself, with Hb I, and with Hb III. Both Hbs I and IIA from M. glutinosa have Trp at E6, and both also lack Lys(F5) and Asp(F9). Hb III has leucine at E6 and alanine at E3. Because Hb IIA has methionine (instead of tyrosine) at A16, Asp (instead of Glu) at E3, and Gln (instead of Asn) at E13 and has Trp at the key E6 position, it seems unlikely that any of the dimers formed among the M. glutinosa Hbs will resemble the F1 dimer.

Both sea cucumber and blood clam Hbs also have a dimer interface using the E and F helices (29, 30). Although the contact region overlaps that found in hagfish Hb, the direction of the 2-fold symmetry axis is very different. The evolutionary advantage of a dimer interface that permits ready communication between hemes has resulted in more than one dimer form associated through the E and F helices that can sense the presence of ligand at the heme (34).

Tertiary Structure Comparison with Vertebrate Globins-- Comparison of vertebrate deoxyglobins including F1 shows that the similarity (in terms of the r.m.s. deviation by residue) between the hagfish and lamprey Hbs is clearly greater than that between human - and -globins, although the level of amino acid sequence identity is about the same for both pairs (~40%). The ₁₁ subunit interface of vertebrate tetrameric Hb is formed by the B, G, and H helices, which are conserved among vertebrate Hbs (2, 3, 35, 36) and which form a relatively rigid core that moves little upon ligand binding (25). Table IV presents the r.m.s. deviations of all the helix core residues that indicate that the BGH core structure is relatively fixed among different deoxyglobins, including hagfish Hb. Deoxy-Hbs were used in Table IV because better structures are available for them, but similar results were obtained when CO forms were compared. It can be concluded that the BGH core structure evolved before the branching point of the hagfishes, before it began to serve as a subunit interface of the ₂₂ tetramer.

Structure Changes at the Heme upon Ligand Binding-- Fig. 5 shows a stereo view of the heme region of the deoxy and CO forms of hagfish Hb superposed by least-squares fitting the helical core residues. E11 is Ile rather than Val in all known sequences of hagfish and lamprey Hbs (13) and may be one of the reasons for their low oxygen affinity (Table III). E7 is His in most vertebrate Hbs, including lamprey Hb. Both Ile(E11) and His(E7) occupy positions obstructing CO binding in the deoxy form and move away in the CO form by a concerted tilting of the heme and displacement of the residues, just as for the  subunit of human Hb (1). The movements of residues on the proximal side are relatively small.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Stereoscopic comparison of the heme environments of E. burgeri deoxy-Hb and HbCO F1. The two structures have been superposed using the helix frame. The boldface lines show the CO form, and the thin lines show the deoxy form. In the deoxy form, the distal residues Ile(E11) and Gln(E7) sterically obstruct ligand binding.

Four key parameters describing the link between ligand binding and tertiary structure change are given in Table V. Myoglobin is included because its branching point is near that of hagfish and lamprey Hbs (10). On the distal side, the positions and movements of His(E7) and Val(E11) (Gln and Ile, respectively, in hagfish) are rather variable among different animals, more so for E11, but with no apparent pattern. This probably indicates that the E helix is relatively easy to move and also reflects the different role of E11 in different globins. The movement of iron into the heme plane upon ligand binding and the associated movement of the proximal His imidazole to a more perpendicular position (Table V) are very marked in vertebrate ₂₂-type Hbs, but small or insignificant in hagfish Hb and myoglobin. In tetrameric Hbs, these proximal movements (and associated shifts of the F helix) correlate with the large shift at the ₁₂ interface upon quaternary structure change. No significant movement of the F helix is observed in hagfish Hb and myoglobin. In human Hb ₄, which shows no quaternary structure change upon ligation, the difference around the F helix is not observed, although the movement of the proximal His imidazole is apparent (Table V) (37, 38). In this case, it seems that the F helix is linked more to the quaternary structure change than to heme ligation. Heaslet and Royer (39) have recently examined the structure changes in lamprey Hb by exposing deoxy crystals to carbon monoxide. The electron density maps show that the distal histidine appears to be highly flexible and cannot be modeled in a single conformation, unlike the clear density seen for this residue in the deoxy and ligated monomer structures. The changes seen by Heaslet and Royer (5, 39) show that ligand affinity is controlled by distal effects, principally movements of the E helix and distal histidine, just as in hagfish Hb F1.

View this table: [in this window] [in a new window]   Table V Key parameters describing the changes in heme geometry upon ligand binding Fe-Pn, the distance between iron and the center of the four porphyrin nitrogen atoms. This parameter represents the heme doming upon ligand binding. F8[C--N-3]-[C-N-1], the difference of two distances representing a tilting of the imidazole of the proximal His. One is between C- of the imidazole of the proximal His (F8 in human Hb and shark Hb and Mb and F14 in hagfish Hb) and nitrogen of pyrole 3 of porphyrin, and the other is between C- of the imidazole of the proximal His and nitrogen of pyrole 1. This tilting links the shift of the F helix toward the FG corner upon ligand binding. Val(E11) C-2-O(CO), the distance between Val(E11) C-2 (Ile C- for hagfish) and the oxygen atom of the ligand CO. For the deoxy form, the ligand has been inserted into the model by overlapping the CO form of the same protein on the heme atoms. His(E7) N--O(CO), the distance between His(E7) N- (Gln O-1 for hagfish) and the oxygen atom of the ligand CO, treating the deoxy form as for Val(E11) C-2.

Cooperative Oxygen Binding in Cyclostome Hbs-- F1 shows small but significant cooperativity in oxygen binding. Like human hemoglobin, deoxy-Hb F1 crystals crack when exposed to the air, indicating that some intermolecular interactions (not necessarily the interaction between subunits of the dimer) are oxygen-linked. Because the distal Ile(E11) and Gln(E7) obstruct oxygen binding in the deoxy form, then if the dimer is more stable in the deoxy form than in the oxy form, as suggested by the crystal packing described in this study, then oxygen binding must be cooperative.

In lamprey Hb V, dimer formation displaces the E helix, pushing the distal His closer to the ligand-binding site and reducing the oxygen affinity of the deoxy conformation (5). The most marked conformational change takes place at Trp⁷²(E6). As discussed earlier, this residue is mutated to His in hagfish Hb, but similar distal residue movements take place upon ligand binding in both proteins (Fig. 5). Overall, the deviations of the proximal side chains are smaller than those of distal residues, and there is no apparent connection between the movement of the proximal His and the hydrogen bonds between Lys(F5) and Asp(F9) at the deoxy dimer interface. The only apparent link between heme ligation and dimerization involves a heme propionate. Ligation flattens the heme, pressing against Phe¹⁰⁷ on the proximal side and moving it ~0.5 Å across the face of the proximal histidine side chain. In turn, the benzene ring pushes against Lys¹⁰², breaking the salt bridge that this residue forms with the heme propionate in the deoxy form. This group lies within 4 Å of Glu⁸¹ on the partner chain of the dimer, so breaking the bond with Lys¹⁰² destabilizes the dimer in the liganded form. The 2F_o  F_c electron density map of the heme region of the molecule in the deoxy form is shown in Fig. 6.

View larger version (52K): [in this window] [in a new window]   Fig. 6.   2Fo  Fc electron density map showing the heme of deoxy-Hb F1. The map is contoured at 1.3 and clearly shows the absence of ligand bound to the heme. The resolution of the data is high enough to model the side chain conformations of residues around the heme unambiguously.

The lamprey deoxy-Hb dimer has a comparatively small subunit contact area of 478 Å² (5), which may not completely explain the linkage properties of subunit assembly and ligand binding (40) because oligomers larger than dimers have been observed (41, 42). In the case of hagfish deoxy hemolysate, Bannai et al. (14) found Hb with a sedimentation coefficient corresponding to tetramers. Moreover, they showed by ultracentrifugation that a 1:1 mixture of F3 and F4 was tetrameric. More studies of cyclostome Hbs under conditions that stabilize higher molecular mass species are necessary to characterize cooperative oxygen binding in these Hbs.

Heterotropic Effects-- Lamprey and hagfish Hbs show markedly different heterotropic effects, with the former lacking regulation by organic phosphates and the latter showing a very weak Bohr effect. M. glutinosa hemolysate shows a slightly greater Bohr effect in the absence of organic phosphates and chloride, both of which favor dissociation (12, 13). In contrast, lamprey Hb has a Bohr effect as strong as that found in Root effect Hbs of teleost fish (43); mutating Glu⁷⁵ to Gln cuts this Bohr effect in half (44). Glu⁷⁵ lies very close to its symmetry mate and Glu³¹ in the lamprey deoxy-Hb dimer. Because hagfish Hb F1 has a valine residue at the equivalent position, it is expected that its Bohr effect will be small.

The oxygen affinity of hagfish (M. glutinosa) hemolysate is strongly lowered by bicarbonate ions (45), an unusual property that is shared with crocodile Hbs (46). The structure of F1 does not suggest an obvious binding site, although this may be because other hagfish Hb components are responsible for the effect.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

As noted by Fago and Weber (13), the two main differences between the red cell of hagfish and higher vertebrates are (i) the inability of the Hbs to form stable tetramers and (ii) the lack of an active anion exchanger in the red cell membrane. This leads to high concentrations of bicarbonate inside the red cell, which may be transported to the gills with the help of Hb, whose oxygen affinity is linked to bicarbonate binding. The low metabolic rate of hagfish does not seem to require a highly cooperative oxygen carrier, and our structures confirm that F1 is unable to form monomer-monomer interactions of the kind found in higher vertebrate Hbs. The long N-terminal extension and the deletion between the G and H helices make both ₁₁- and ₁₂-type interactions impossible (4). It has been suggested that hagfish Hbs represent an intermediate between invertebrate and vertebrate Hbs (47), but we see no evidence of direct heme-heme interaction as found in the Hb of the clam S. inaequivalvis (27). From the structures described in this study, it appears that the BGH core structure of the globin fold evolved before the branching point on the evolutionary tree where hagfish diverged from vertebrates. The ₂₂ tetramer, with its two quaternary structures, arose between the branching points of hagfish and lampreys from cartilaginous fish. The movement of the F helix commonly found upon ligand binding to vertebrate ₂₂-type Hb is not seen in F1 and appears to be linked more firmly to quaternary structure change than to the ligation state of the heme. The stereochemical mechanisms regulating heme and non-heme ligand binding have clearly evolved independently among hagfish and lampreys and other vertebrates since their divergence. The BGH core structure was apparently present in ancestral globins before this occurred, but the switching function of the proximal histidine appears to have arisen later. The structural basis of the heterotropic effects of protons, chloride, and bicarbonate ions in hagfish Hb remains unclear, which is perhaps not surprising given the number of mutant forms that have been studied to achieve our present understanding of these effects in human Hb.

	ACKNOWLEDGEMENTS

We thank Dr. Angela Fago and Prof. Roy Weber for extremely helpful comments on the manuscript.

	FOOTNOTES

^* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The atomic coordinates and the structure factors (code 1IT2 (deoxy-Hb) and 1IT3 (HbCO)) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).

^** To whom correspondence should be addressed. Tel.: 81-45-508-7228; Fax: 81-45-508-7366; E-mail: jtame@tsurumi.yokohama-cu.ac.jp.

Published, JBC Papers in Press, March 28, 2002, DOI 10.1074/jbc.M111492200

¹ Y. Yazawa, personal communication.

³ S.-Y. Park, J. R. H. Tame, S. Adachi, and N. Shibayama, manuscript in preparation.

	ABBREVIATIONS

The abbreviations used are: bis-Tris, 2-[bis(2-hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol; r.m.s., root mean square.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				G. Muller, A. Fago, and R. E. Weber Water regulates oxygen binding in hagfish (Myxine glutinosa) hemoglobin J. Exp. Biol., April 15, 2003; 206(8): 1389 - 1395. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF)