Protein structure in the truncated (2/2 ... - Wiley Online Library

Download PDF
18 downloads 4608 Views 424KB Size Report
Comment
IUBMB Life, 59: 535–541, 2007 ... Some of the organisms hosting truncated Hbs, that are ..... domain coupled to a cofactor-free monooxygenase domain of.
IUBMB

Life, 59(8 – 9): 535 – 541, August – September 2007

Critical Review Protein Structure in the Truncated (2/2) Hemoglobin Family Alessandra Pesce1, Marco Nardini2, Mario Milani2 and Martino Bolognesi2 1 2

Department of Physics, CNR-INFM and Center for Excellence in Biomedical Research, University of Genova, Genova, Italy Department of Biomolecular Sciences and Biotechnology, CNR-INFM, University of Milano, Milano, Italy

Summary The discovery of protein sequences belonging to the widespread ‘truncated hemoglobin’ family has been followed in the last few years by extensive analyses of their three-dimensional structures. Truncated hemoglobins can be classified in three main groups, in light of their overall structural properties. The three groups adopt a 2-on-2 a-helical sandwich fold, based on four main a-helices of the classical 3-on-3 a-helical sandwich found in vertebrate and invertebrate globins. Each of the three groups displays sequence and structure specific features. Among these, a protein matrix tunnel system is typical of group I, a Trp residue at the G8 topological site is conserved in groups II and III, and residue TyrB10 is almost invariant in the three groups. Despite sequence variability in the heme distal site region, a strongly intertwined, but varied, network of hydrogen bonds stabilizes the heme ligand in the three protein groups. Fine mechanisms of ligand recognition and stabilization may vary based on groupspecific distal site residues and on differing ligand diffusion pathways to the heme. Taken together, the structural considerations here presented underline that ‘truncated hemoglobins’ result from careful editing of the 3-on-3 a-helical globin sandwich fold, rather than from simple ‘truncation’ events. Thus, ‘2/2Hb’ appears the most proper term to concisely address this protein family. IUBMB Life, 59: 535–541, 2007 Keywords

Truncated hemoglobin; globin fold; hemoglobin; hemoglobin evolution; heme/ligand recognition.

Abbreviations Hb, hemoglobin; Mb, myoglobin; 2/2Hb, 2-on-2 globin; Ce-2/2HbN, Chlamydomonas eugametos HbN; Pc-2/2HbN, Paramecium caudatum HbN; Mt-2/2HbN, Mycobacterium tuberculosis HbN; Ss-2/2HbN, Synechocystis sp. HbN; Mt-2/2HbO, Mycobacterium tuberculosis HbO; Bs-2/2HbO, Bacillus subtilis HbO; Cj-2/2HbP, Campylobacter jejuni HbP; Ml-2/2HbO, Mycobacterium leprae HbO; amino acid residues have been labeled based on the topological site they occupy within the globin fold. Received 16 January 2007; accepted 16 January 2007 Address correspondence to: Prof. Martino Bolognesi, Department of Biomolecular Sciences and Biotechnology, University of Milano, Via Celoria 26, I-20131 Milano, Italy. Tel: þ39 02 50314893. Fax: þ39 02 50314895. E-mail: [email protected] ISSN 1521-6543 print/ISSN 1521-6551 online Ó 2007 IUBMB DOI: 10.1080/15216540701225933

INTRODUCTION More than one hundred truncated hemoglobins were discovered in the last 10 – 15 years in bacteria, higher plants, and unicellular eukaryotes, and identified as a distinct protein phylogenetic group within the hemoglobin (Hb) super-family (1, 2). Truncated Hb primary structure is generally 20 – 40 residues shorter than mammalian Hbs, resulting in deletion or shortening of a-helices and in modified interhelical loops. Such structural modifications, relative to the ‘classical’ vertebrate globins, are carefully distributed throughout the whole hemoprotein polypeptide chain. Analyses of the increasing number of amino acid sequences have shown that this new globin family branches into three groups, designated I, II and III (also distinguished by the N, O, and P suffixes, respectively), the component proteins being orthologous within each group and paralogous across the groups (2). Sequence identity between proteins from different groups is low (20% overall identity), but may be higher than 80% within a given group. Group III displays a high level of internal sequence conservation; group I and group II can be further separated into two and four subgroups, respectively. From the evolutionary viewpoint the phylogenetic analysis suggests that group II HbO is the ancestral gene, whereas group I and group III genes would result from later duplication and transfer events (2). Some of the organisms hosting truncated Hbs, that are held to be of very ancient origin, are aggressive pathogenic bacteria, perform photosynthesis, fix nitrogen, or have distinctive metabolic capabilities. Truncated Hbs from more than one group have been shown to coexist in the same bacterium, indicating a diversification of their functions, that may include long-term ligand or substrate storage, nitric oxide (NO) detoxification, O2/NO sensing, redox reactions, and O2 delivery under hypoxic conditions (1, 2). For example, group I HbN from Nostoc sp., is thought to scavenge O2, thus protecting the nitrogen-fixation apparatus (3). Group I HbN promotes an efficient dioxygenase reaction, converting NO into nitrate, in Mycobacterium bovis BCG (4). A similar

536

PESCE ET AL.

function has been proposed for Mycobacterium tuberculosis where group I HbN would be crucial in protecting the mycobacterium from nitrosative stress in vivo (4, 5). In Campylobacter jejuni, a pathogenic agent in gastrointestinal disease, group III HbP may play a role in cell respiration (6, 7). Several three-dimensional structures belonging to the three truncated Hb groups have been recently characterized by means of X-ray crystallography and NMR: four group I HbNs from Chlamydomonas eugametos (Ce-2/2HbN) (8), Paramecium caudatum (Pc-2/2HbN) (8), M. tuberculosis (Mt-2/2HbN) (9) and Synechocystis sp. (Ss-2/2HbN) (10 – 12), four group II HbOs from M. tuberculosis (Mt-2/ 2HbO) (13), from Bacillus subtilis (Bs-2/2HbO) (14), from Thermobifida fusca (15), from Geobacillus stearothermophilus (16), and one group III HbP from C. jejuni (Cj-2/ 2HbP) (17).

THE TRUNCATED HB 2/2 GLOBIN FOLD The three-dimensional structures mentioned above for group I HbN, group II HbO, and group III HbP show that their fold is related, but simpler than the ‘classical’ globin fold (the so called 3-on-3 a-helical sandwich) typical of sperm whale myoglobin (Mb) (18). In such simpler protein fold the heme is hosted in a 2-on-2 a-helical sandwich (‘2/2 fold’) whose four a-helices match the B-, E-, G-, and H-helices of the classical 3-on-3 globin fold; the antiparallel B/E and G/H helix pairs are arranged in a sort of a-helical bundle (Fig. 1). As anticipated, such fold is the result of multiple and complex structure-editing events, distributed at several sites relative to the 3-on-3 classical globin fold, such that the term ‘2/2Hb’ presently appears as more precise than ‘truncated Hb’ to refer to this hemoprotein family.

Besides the reduced size, the 2/2Hb fold deviates from the 3/3 conventional globin fold for structural features, partly common to groups I, II, and III, partly specific for each group. Among the family-conserved structural features we notice a drastically shortened A-helix (fully deleted in group III), the absence of a D-helix, an extended polypeptide segment (preF), followed by a short F-helix (one helix turn in groups I and III) supporting the heme proximal HisF8 residue (8, 13, 17) (Fig. 1). The G- and H-helices generally match the globin fold topology, but may be much shorter or bent, as compared to sperm whale Mb (8). The CD-D region of the 2/2Hb fold is trimmed to about three residues, linking directly C- and E-helices (Figs. 1 and 2). However, a 3-7 amino acid insertion located between the C- and E-helices, is invariantly found in group III 2/2HbPs (2, 17). Such elongation of the CD region has implications on the span of the C- and E-helices and on the 310 helical character of helix C. Notably, after extensive three-dimensional structure comparisons, the overall fold of group III 2/2HbP appears equally different in its Ca trace from group I and group II (Fig. 2B). Three conserved Gly motifs have been suggested to support attainment and stabilization of the compact 2/2 fold in groups I and II; the Gly-motifs are located at the AB, at EF interhelical corners, and just before the short F-helix (the last being present also in group III). In group I and II 2/2Hbs the Gly-motifs are held to help locating the short A-helix onto the B- and E-helices, and supporting a properly structured heme crevice around the pre-F segment (8, 9). In group III Cj-2/ 2HbP the AB Gly-Gly motif is absent, and the short A-helix deleted. As a result, the protein residues preceding the B-helix extend towards the GH region, in a conformation opposite to that found in groups I and II. Additionally, one or more Gly residues are present in the EF region of group III 2/2HbPs, however a well defined Gly-Gly motif cannot be recognized, as

Figure 1. Topology of the 2/2Hb fold. A ribbon stereo view of the structural overlay between group I Ce-2/2HbN (dark-grey wide ribbon; PDB accession code 1DLY) and 3/3 sperm whale Mb (light-grey thin trace; PDB accession code 1EBC). For clarity, the heme group is displayed only for Ce-2/2HbN. The heme distal side crevice is on the right of the porphyrin ring, shown here approximately in an edge on view. a-helices are labelled according to the classical globin topological convention, A through H, starting from the N-terminus. Drawn with MOLSCRIPT (19).

PROTEIN STRUCTURE IN THE TRUNCATED HEMOGLOBIN FAMILY

537

interactions may also result from sequence-specific substitutions in the surrounding of the heme propionates. The hexacoordinate Synechocystis sp. 2/2HbN displays a covalent bond linking HisH16 and the 2-vinyl group of the heme, that may modulate the reactivity of the heme group (10, 11, 21, 22). Moreover, heme isomerism has been reported in some of the 2/2Hb crystal structures (23). The Fe-coordinated proximal HisF8 residue (that is the only residue 100% conserved within the globin super-family) shows mostly an azimuthal orientation (staggered relative to the heme-pyrrole nitrogen atoms) typical of an unstrained imidazole ring. Such orientation facilitates the in-plane location of the heme-Fe atom, increasing the covalency of the Fe-HisF8 bond (7, 24 – 27). It also supports fast distal site O2 association (28) and electron donation to the bound distal ligand (Fig. 3), consistent with a ferric-superoxide character of the heme Fe-O2 pair (25).

Figure 2. The 2/2Hb fold in groups I, II, and III. A ribbon stereo view of the structural overlay of the following protein pairs: group I Ce-2/2HbN (dark-grey trace; PDB accession code 1DLY) and (A) group II Mt-2/2HbO (light-grey trace; PDB accession code 1NGK), and (B) group III Cj-2/2HbP (light-grey trace; PDB accession code 2IG3). The orientation of the protein pairs is approximately the same of Fig. 1. Drawn with MOLSCRIPT (19).

for the other two groups. Nevertheless, this sequence variation does not show dramatic effects on the overall 2/2Hb fold of Cj-2/2HbP.

HEME PROXIMAL SITE IN 2/2HBS Inspection of the known 2/2Hb three-dimensional structures indicates that residues responsible for building the heme crevice, and for protecting/stabilizing the bound heme group, are located mostly at topological sites that are conserved throughout the 2/2Hb family (1, 2, 20). Hydrophobic residues at positions C6, C7, CD1, E14, F4, FG3, G8, and H11 provide an efficient network of van der Waals contacts to the heme in all 2/2 globins. Other protein-heme contacts may arise from residues located in regions of the 2/2 fold that vary in the three groups, such as the CD and FG segments, and the aminoterminal part of the H-helix. Stabilizing interactions are further provided by hydrogen bonds linking the heme and Thr/Tyr residues at sites E2, E5, and EF6. Salt bridges involve the heme propionates and positively charged residues located at site E10 in all 2/2 globin structures, at site F2 in group I 2/2HbN, at F7 in group II 2/2HbO (where F7 is invariantly an Arg residue), and in group III Cj-2/2HbP. Further salt bridge

LIGAND BINDING AT THE HEME DISTAL SITE The E-helix of 2/2Hbs falls very close to the heme, due to the shortened CD region, resulting in side chain crowding for amino acids at topological sites B10, CD1, E7, E11, E14, E15, and G8. Some of these residues are polar, allow the formation of a network of hydrogen bonds that provides efficient stabilization of the exogenous heme ligand, can support the rebinding kinetics of dissociated ligands and may favor oxygen chemistry in the heme crevice (29, 30). Distal site polarity is a common property in 2/2 globins, although different residues may contribute to the architecture of the heme ligand surroundings in the three groups. In group I 2/2HbNs the hydrogen bonded network involves mostly residues at B10, E7, and E11 topological sites (23). For example, in Mt-2/2HbN a direct TyrB10-O2 hydrogen bond occurs, stabilized by GlnE11 interacting with TyrB10 (9, 24, 26, 31). Raman spectroscopy indicates that in Mt-2/2HbN the hemeFe-bound O2, CO, and OH- ligands are stabilized by hydrogen bonding to TyrB10 (24, 26). In Pc-2/2HbN and Ce-2/2HbN residues TyrB10, GlnE7, and GlnE11 are involved in ligand binding (8) (Fig. 3). Notably, minor fluctuations (50.4 A˚) of each interacting residue from the conformation observed in the crystal structures, may extend/reduce the distal site hydrogen-bonded network to the heme-bound ligand. In each case, however, TyrB10 plays a pivotal role in ligand stabilization through a direct hydrogen bond to the heme ligand. Sequence analysis of group I 2/2HbNs reveals that TyrB10 is a strongly conserved residue, although rare substitution with His (e.g., in 2/2HbN from Nostoc sp.), or with a hydrophobic residue, occur (2). In the few 2/2Hbs that display a hydrophobic side chain at B10, large hydrophobic residues occur at E7 and E11 sites (2), apparently building a subgroup within group I 2/2HbNs. As mentioned above, Ss-2/2HbN displays bis-histidine heme hexacoordination, where HisF8 and HisE10 are the

538

PESCE ET AL.

Figure 3. Structure of the heme pocket in 2/2Hbs. Stereo view of the heme (dark grey colour), of the proximal HisF8 and of the distal residues involved in stabilizing the exogenous ligand in group I cyano-met 2/2HbN from C. eugametos, taken as a prototype example of the distal site hydrogen bonded network stabilizing the heme-bound ligand. The cyanide molecule is shown as a stick; hydrogen bonds among the distal site residues and the ligand are shown as dashed lines. Residues are labelled according to their topological sites only, as adopted throughout the text. Drawn with MOLSCRIPT (19).

Fe-atom ligands (21, 22, 32). Binding of an exogenous ligand to the heme distal site, requires the dissociation of HisE10 from the heme and remarkable conformational changes affecting the B- and E-helices (12, 33). Ce-2/2-HbN may also display a six-coordinate heme-Fe-atom, in the ferrous and ferric forms, under specific conditions (24, 34). Heme hexacoordination has also been reported for the ferrous derivative of Mycobacterium leprae HbO (Ml-2/2HbO) at neutral pH (35). In general terms, however, 2/2Hbs do not show a marked trend for bis-histidine heme hexacoordination. Specific residue substitutions characterize group II 2/ 2HbOs distal site environment relative to group I 2/2HbN. In Bs-2/2HbO cyano-met crystal structure TyrB10 is the residue directly hydrogen bonded to the ligand. GlnE11, TrpG8, and ThrE7 complete the distal site polar cage, with GlnE11 side chain and the TrpG8 indolic nitrogen atom hydrogen bonded to the heme ligand (14). The occurrence of TyrCD1, however, is sufficient to drastically modify the hydrogen bonding distal network in Mt-2/2HbO, where TyrCD1, not TyrB10, is the residue hydrogen bonding to the heme diatomic ligand (13). Further stabilizing polar interactions are provided in Mt-2/2HbO by TrpG8, whose indole NE1 atom is hydrogen bonded to the heme-bound ligand and to TyrCD1 OH; notably, residue E11 is Leu. Thus, the group II (and III) conserved TrpG8 residue can stabilize the heme ligand with one hydrogen bond, and may modulate the rate of ligand escape from the distal pocket (13, 14). In the Mt-2/2HbO case, residue TyrB10 may be firstly involved in stabilizing the orientation of CD1 and G8 residues. However, when residue CD1 is Phe, TyrB10 plays the role of

hydrogen bond donor to the ligand (36). A distal hydrogen bonded network matching that of Mt-2/2HbO is expected to occur in Ml-2/2HbO (35), and in other group II 2/2HbOs, bearing Tyr residues at B10 and CD1 sites (2). The crystal structure of Mt-2/2HbO has shown that the simultaneous presence of Tyr residues at B10 and CD1 sites has the potential for the formation of a covalent (iso-dityrosine like) bond between the two side chains (13). Such unusual covalent modification has been suggested to take place via encounter of the protein with an oxidative species (which triggers the formation of a Tyr-phenoxyl radical); no hints on the functional role played by iso-dityrosine in Mt-2/ 2HbO is however available. Structural and sequence analyses suggests that the nature of residues at CD1 and E11 sites are correlated in group II 2/2HbO. All 2/2HbOs that display a Tyr residue at position CD1 have a nonpolar residue at the E11 site (2, 13, 14). Conversely, when a non-hydrogen bonding Phe residue is found at CD1, a hydrogen bond donor is present at the E11 site (Gln or Ser). Alternatively, a His residue (a hydrogen bonding residue) at the topological position CD1 site is always matched by a hydrophobic E11 residue (Leu or Phe). Thus, one of the necessary hydrogen bonding elements involved in ligand stabilization is alternatively located at opposite edges of the heme distal cavity, either at the CD1 or at the E11 sites, but never simultaneously. TyrB10 and TrpG8 are invariant in group II and in group III 2/2Hbs. The recently reported X-ray structure of cyano-met Cj-2/2HbP has shed first light on the distal site hydrogen bonding network responsible for ligand binding in

PROTEIN STRUCTURE IN THE TRUNCATED HEMOGLOBIN FAMILY

group III (17). In Cj-2/2HbP the heme distal pocket residues TyrB10, PheCD1, HisE7, IleE11, PheE14, and TrpG8 surround the heme-bound ligand. Among these, HisE7 is the only residue fully conserved in group III (2). TyrB10 and TrpG8 are directly hydrogen bonded to the ligand, closely matching the ligand binding mode found in group II Bs-2/2HbO (14). Contrary to group II 2/2HbO, group III 2/2HbPs display an invariant Phe residue at position CD1 and a hydrophobic residue at position E11. In Cj-2/2HbP a water molecule, filling a small distal site cavity, matches the site occupied by GlnE11 OE1 atom in Bs-2/2HbO. However, such water molecule does not stabilize the bound ligand, contrary to the interaction observed for Bs-2/2Hb GlnE11 OE1 atom (14). The crystal structure shows that in group III Cj-2/2HbP the only hydrogen bonding residues involved in ligand stabilization are TyrB10 and TrpG8, remarkably excluding from such role the polar residue HisE7 (17). Indeed, residue HisE7 is observed in two alternate conformations, ‘open’ and ‘closed’ corresponding to the side chain pointing towards the solvent (away from the heme ligand) or towards the heme distal site, respectively. No polar contacts (neither to the heme ligand, nor to TyrB10) characterize the HisE7 closed conformation, at difference from the cyanide stabilization mode observed in sperm whale Mb, where the gating-residue HisE7 locks the heme-Fe(III)-bound cyanide in the distal site through hydrogen bonding (37). Such diverse role played by HisE7 in the two globins can be related to the different overall protein structure in the CD-D-E regions, resulting in diverging orientations of the HisE7 supporting E-helices in sperm whale Mb and in group III Cj-2/2HbP.

PROTEIN MATRIX CAVITIES IN 2/2HBS The 2/2Hb three-dimensional structures suggest that, within a very well conserved and simple fold, ligand diffusion to/from the heme may be based on very different routes. In group I a suitable protein cavity/tunnel system, connecting the protein surface to the heme distal site, has been clearly identified for several 2/2HbNs (31). The tunnel is composed of two roughly orthogonal branches in Ce-2/2HbN and Mt-2/2HbN. On one hand, a 20 A˚ long tunnel branch connects the heme distal site to the protein surface region nestled between the AB and GH corners. On the other, a path of about 8 A˚ connects an opening in the protein structure between G- and H-helices to the heme. Both tunnel branches display inner diameters of about 5 – 7 A˚, with an overall ligand accessible volume of about 330 – 360 A˚3. In Pc-2/2HbN the heme site may be linked to the solvent by three closely located cavities (overall volume of 180 A˚3), topologically distributed along the long tunnel branch described above (31). Residues lining the tunnel branches are hydrophobic and are substantially conserved throughout group I (2). PheE15, a well conserved residue, adopts two conformations in Mt-2/2HbN; one of these may block ligand access to the heme (9, 31).

539

The remarkably modified overall structure of hexacoordinated Ss-2/2HbN does not display a cavity/tunnel system typical of group I 2/2Hbs. An alternate heme distal site access through the exposed 8-methyl edge of the heme group and near the propionates, has been proposed for this Hb (10). Ligand rebinding kinetics have shown that the tunnel/cavity network in Ce-2/2HbN and Pc-2/2HbN may indeed act as CO storage site(s), whose filling strikingly affects ligand rebinding kinetics (30). Group II 2/2Hbs do not show an evident tunnel/cavity system connecting the protein surface to the heme distal pocket. In Mt-2/2HbO (13) and Bs-2/2HbO (14) the protein matrix tunnel typical of group I 2/2HbNs, is dramatically restricted, due to different relative orientations of the G- and H-helices and to an increased volume of side chains at topological sites B1, B5, G8, G9, G12, and H12. Thus, the long tunnel branch retains only two cavities in 2/2HbOs, both solvent inaccessible. The absence of a clear protein matrix tunnel in group II 2/2HbOs seems to be mirrored by the general presence of a small distal site E7 residue (2). Accessibility of diatomic ligands (such as O2, CO, and NO) to the 2/2HbO heme distal site would then be favoured by the small-apolar E7 residue, not hindering entrance to the heme distal cavity and supporting an E7 route entry path. Nevertheless, the presence of two protein matrix cavities still indicates that they might be required as ligand docking stations. Furthermore, both Mt-2/2HbO (13) and Bs-2/2HbO (14) display a shallow depression on the proximal side of the heme, mostly lined with conserved hydrophobic residues. Such depression provides partial solvent access to the heme C-pyrrole, and may serve as a docking site for a reaction partner, possibly having functional significance in heme redox chemistry (13). Contrary to what reported above for group I and group II 2/2Hbs (1, 31), inspection of group III Cj-2/2HbP structure shows no evident protein matrix tunnel/cavity system, as defined by a 1.4 A˚ radius probe. Thus, the ‘open’ and ‘closed’ conformations adopted by the group III fully conserved HisE7 residue have been proposed to underline an E7 hemedistal-site gating role for ligand diffusion to/from the heme (17).

CONCLUSIONS The 2/2Hb fold, as depicted by experimental determination of structures from members of all three 2/2Hb groups, stems from extensive and complex modifications of the ‘classical’ (3/3) globin fold, that are distributed throughout the whole protein molecule, not resulting from simple ‘truncation’ events. Key structural determinants of the 2/2 fold are a drastically reduced or fully absent A-helix, a short CD-D segment linking the C- and E-helices that forces the latter helix close to the heme distal face, a proximal F-helix largely replaced by an extended polypeptide loop, a C-terminal

540

PESCE ET AL.

H-helix highly variable in length and linearity. Residues of the heme crevice involved in van der Waals contacts to the porphyrin and in polar/electrostatic interactions with the heme propionates are generally well conserved in all 2/2 globins, although fine group-specific variations are evident (20). Residue TyrB10, nearly invariant in all 2/2Hbs, is primarily responsible for binding and stabilization of the exogenous heme ligand. The only significant variation to this pattern has been recorded in Mt-2/2HbO, where TyrCD1, instead of TyrB10, acts as the ligand hydrogen bonding residue. As a general consequence of the close location of the E-helix to the heme in 2/2 globins, the heme distal pocket is crowded with amino-acid side chains. Among these, the residues at key topological sites (usually E7, E11, and G8) can provide additional hydrogen bonds to the ligand, and to Tyr B10. Thus, the combination of distal polar interactions (that may vary somehow in the three groups) builds up 2/2Hb specific, ligation-sensitive, hydrogen bonding networks, responsible for ligand stabilization and for the rebinding kinetics of dissociated ligands. Two Gly-Gly sequence motifs, located at the AB and EF inter-helical hinges, and one single Gly residue, six amino acids downstream of the EF Gly-Gly motif, have been identified as fingerprints of the 2/2Hb fold in groups I and II. Recent data on Cj-2/2HbP three-dimensional structure (17) have shown that the AB Gly-Gly motif is completely absent, and the single pre-F Gly residue is present and conserved throughout group III. Considering that one or more Gly residues are scattered in group III EF region, although not identifying a prominent/conserved sequence motif, it must be concluded that two Gly-Gly sequence motifs are not essential, although virtually invariant in groups I and II, to preserve the 2/2Hb fold throughout the three groups. The protein tunnel detailed for several group I HbNs and the two-cavity system identified for group II 2/2HbOs do not appear to be a prominent structural feature of group III HbPs. In the absence of such cavity/tunnel system, alternative heme distal site access may be granted through other routes, for instance through the E7 gate, as proposed Cj-2/2HbP (17), or through an E7 open route, as suggested for Mt2/2HbO (13). Gene and protein structural analyses have shown that amino- and carboxy-termini structural extensions can be found coupled to a ‘classical’ 3-on-3 globin core, as observed in human cytoglobin (38 – 41). Such extensions can be of a respectful size, as observed in flavohemoglobins or in globincoupled sensors (42, 43). Extensive genome analyses have recently unravelled several new members of the 2/2Hb family. Sequence comparisons, at the light of our current structural knowledge on 2/2Hbs, suggest that also the 2/2Hb fold may occur as a stable structural core within more complex proteins, displaying significant amino- and carboxy-terminal extensions. A notable case in this context is provided by a chimaeric protein from Streptomyces avermitilis that hosts a 2/2Hb

domain coupled to a cofactor-free monooxygenase domain of about 100 residues (44).

ACKNOWLEDGEMENTS Stimulating discussions with Profs. Paolo Ascenzi (University of Rome Tre, Italy), Michel Guertin (Laval University, Canada), and Luc Moens (Antwep University, Belgium) are fully acknowledged. This work was supported by MIUR FIRB Grant ‘Biologia Strutturale’ to M.B. M.M. is the recipient of a post-doctoral fellowship supported by NIH Grant 1-R01-AI052258 (2004-2008). M.B. is grateful to CIMAINA (University of Milano, Italy) for continuous support.

REFERENCES 1. Wittenberg, J. B., Bolognesi, M., Wittenberg, B. A., and Guertin, M. (2002) Truncated hemoglobins: a new family of hemoglobins widely distributed in bacteria, unicellular eukaryotes, and plants. J. Biol. Chem. 277, 871 – 874. 2. Vuletich, D. A., and Lecomte, J. T. (2006) A phylogenetic and structural analysis of truncated hemoglobins. J. Mol. Evol. 62, 196 – 210. 3. Hill, D. R., Belbin, T. J., Thorsteinsson, M. V., Bassam, D., Brass, S., Ernst, A., Boger, P., Paerl, H., Mulligan, M. E., and Potts, M. (1996) GlbN (cyanoglobin) is a peripheral membrane protein that is restricted to certain Nostoc spp. J. Bacteriol. 178, 6587 – 6598. 4. Ouellet, H., Ouellet, Y., Richard, C., Labarre, M., Wittenberg, B., Wittenberg, J., and Guertin, M. (2002) Truncated hemoglobin HbN protects Mycobacterium bovis from nitric oxide. Proc. Natl. Acad. Sci. USA 99, 5902 – 5907. 5. Pathania, R., Navani, N. K., Gardner, A. M., Gardner, P. R., and Dikshit, K. L. (2002) Nitric oxide scavenging and detoxification by the Mycobacterium tuberculosis haemoglobin, HbN in Escherichia coli. Mol. Microbiol. 45, 1303 – 1314. 6. Wainwright, L. M., Elvers, K. T., Park, S. F., and Poole, R. K. (2005) A truncated haemoglobin implicated in oxygen metabolism by microaerophilic food-borne pathogen Campylobacter jejuni. Microbiology 151, 4079 – 4091. 7. Wainwright, L. M., Wang, Y., Park, S. F., Yeh, S. R., and Poole, R. K. (2006) Purification and spectroscopic characterization of Ctb, a group III truncated hemoglobin implicated in oxygen metabolism in the food-borne pathogen Campylobacter jejuni. Biochemistry 45, 6003 – 6011. 8. Pesce, A., Couture, M., Dewilde, S., Guertin, M., Yamauchi, K., Ascenzi, P., Moens, L., and Bolognesi, M. (2000) A novel two-overtwo a-helical sandwich fold is characteristic of the truncated hemoglobin family. EMBO J. 19, 2424 – 2434. 9. Milani, M., Pesce, A., Ouellet, Y., Ascenzi, P., Guertin, M., and Bolognesi, M. (2001) Mycobacterium tuberculosis hemoglobin N displays a protein tunnel suited for O2 diffusion to the heme. EMBO J. 20, 3902 – 3909. 10. Falzone, C. J., Vu, B. C., Scott, N. L., and Lecomte, J. T. (2002) The solution structure of the recombinant hemoglobin from the cyanobacterium Synechocystis sp. PCC 6803 in its hemichrome state. J. Mol. Biol. 324, 1015 – 1029. 11. Hoy, J. A., Kundu, S., Trent, J. T. 3rd, Ramaswamy, S., and Hargrove, M. S. (2004) The crystal structure of Synechocystis hemoglobin with a covalent heme linkage. J. Biol. Chem. 279, 16535 – 16542.

PROTEIN STRUCTURE IN THE TRUNCATED HEMOGLOBIN FAMILY

12. Trent, J. T. 3rd., Kundu, S., Hoy, J. A., and Hargrove, M. S. (2004) Crystallographic analysis of synechocystis cyanoglobin reveals the structural changes accompanying ligand binding in a hexacoordinate hemoglobin. J. Mol. Biol. 341, 1097 – 1108. 13. Milani, M., Savard, P. Y., Ouellet, H., Ascenzi, P., Guertin, M., and Bolognesi, M. (2003) A TyrCD1/TrpG8 hydrogen bond network and a TyrB10-TyrCD1 covalent link shape the heme distal site of Mycobacterium tuberculosis hemoglobin O. Proc. Natl. Acad. Sci. USA 100, 5766 – 5771. 14. Giangiacomo, L., Ilari, A., Boffi, A., Morea, V., and Chiancone, E. (2005) The truncated oxygen-avid hemoglobin from Bacillus subtilis: x-ray structure and ligand binding properties. J. Biol. Chem. 280, 9192 – 9202. 15. Bonamore, A., Ilari, A., Giangiacomo, L., Bellelli, A., Morea,V., and Boffi, A. (2005) A novel thermostable hemoglobin from the actinobacterium Thermobifida fusca. FEBS J. 272, 4189 – 4201. 16. Ilari, A., Kjelgaard, P., von Wachenfeldt, C., Catacchio, B., Chiancone, E., and Boffi, A. (2007) Crystal structure and ligand binding properties of the truncated hemoglobin from Geobacillus stearothermophilus. Arch. Biochem. Biophys. 457, 85 – 94. 17. Nardini, M., Pesce, A., Labarre, M., Richard, C., Bolli, A., Ascenzi, P., Guertin, M., and Bolognesi, M. (2006) Structural determinants in the group III truncated hemoglobin from Campylobacter jejuni. J. Biol. Chem. 281, 37803 – 37812. 18. Perutz, M. F. (1979) Regulation of oxygen affinity of hemoglobin: influence of structure of the globin on the heme iron. Annu. Rev. Biochem. 48, 327 – 386. 19. Kraulis, P. J. (1991). MOLSCRIPT: a program to produce both detailed and schematic plots of protein structures. J. Appl. Crystallogr. 24, 946 – 950. 20. Nardini, M., Pesce, A., Milani, M., and Bolognesi, M. (2007) Protein fold and structure in the truncated (2/2) globin family. Gene in press. 21. Couture, M., Das, T. K., Savard, P. Y., Ouellet, Y., Wittenberg, J. B., Wittenberg, B. A., Rousseau, D. L., and Guertin, M. (2000) Structural investigations of the hemoglobin of the cyanobacterium Synechocystis PCC6803 reveal a unique distal heme pocket. Eur. J. Biochem. 267, 4770 – 4780. 22. Scott, N. L., Falzone, C. J., Vuletich, D. A., Zhao, J., Bryant, D. A., and Lecomte, J. T. (2002) Truncated hemoglobin from the cyanobacterium Synechococcus sp. PCC 7002: evidence for hexacoordination and covalent adduct formation in the ferric recombinant protein. Biochemistry 41, 6902 – 6910. 23. Milani, M., Pesce, A., Nardini, M., Ouellet, H., Ouellet, Y., Dewilde, S., Bocedi, A., Ascenzi, P., Guertin, M., Moens, L., Friedman, J. M., Wittenberg, J. B., and Bolognesi, M. (2005) Structural bases for heme binding and diatomic ligand recognition in truncated hemoglobins. J. Inorg. Biochem. 99, 97 – 109. 24. Couture, M., Yeh, S. R., Wittenberg, B. A., Wittenberg, J. B., Ouellet, Y., Rousseau, D. L., and Guertin, M. (1999) A cooperative oxygen-binding hemoglobin from Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. USA 96, 11223 – 11228. 25. Das, T. K., Couture, M., Ouellet, Y., Guertin, M., and Rousseau, D. L. (2001) Simultaneous observation of the O-O and Fe-O2 stretching modes in oxyhemoglobins. Proc. Natl. Acad. Sci. USA 98, 479 – 484. 26. Yeh, S. R., Couture, M., Ouellet, Y., Guertin, M., and Rousseau, D. L. (2000) A cooperative oxygen binding hemoglobin from Mycobacterium tuberculosis. Stabilization of heme ligands by a distal tyrosine residue. J. Biol. Chem. 275, 1679 – 1684. 27. Samuni, U., Ouellet, Y., Guertin, M., Friedman, J. M., and Yeh, S. R. (2004) The absence of proximal strain in the truncated hemoglobins from Mycobacterium tuberculosis. J. Am. Chem. Soc. 126, 2682 – 2683. 28. Bolognesi, M., Bordo, D., Rizzi, M., Tarricone, C., and Ascenzi, P. (1997) Nonvertebrate hemoglobins: structural bases for reactivity. Prog. Biophys. Mol. Biol. 68, 29 – 68.

541

29. Hiner, A. N., Raven, E. L., Thorneley, R. N., Garcia-Canovas, F., and Rodriguez-Lopez, J. N. (2002) Mechanisms of compound I formation in heme peroxidases. J. Inorg. Biochem. 91, 27 – 34. 30. Samuni, U., Dankster, D., Ray, A., Wittenberg, J. B., Wittenberg, B. A., Dewilde, S., Moens, L., Ouellet, Y., Guertin, M., and Friedman, J. (2003) Kinetics modulation in the carbonmonoxy derivatives of truncated hemoglobins. The role of distal heme pocket residues and extended apolar tunnel. J. Biol. Chem. 278, 27241 – 27250. 31. Milani, M., Ouellet, Y., Ouellet, H., Guertin, M., Boffi, A., Antonini, G., Bocedi, A., Mattu, M., Bolognesi, M., and Ascenzi, P. (2004) Cyanide binding to truncated hemoglobins: a crystallographic and kinetic study. Biochemistry 43, 5213 – 5221. 32. Lecomte, J. T., Scott, N. L., Vu, B. C., and Falzone, C. J. (2001) Binding of ferric heme by the recombinant globin from the cyanobacterium Synechocystis sp. PCC 6803. Biochemistry 40, 6541 – 6552. 33. Vu, B. C., Nothnagel, H. J., Vuletich, D. A., Falzone, C. J., and Lecomte, J. T. (2004) Cyanide binding to hexacoordinate cyanobacterial hemoglobins: hydrogen-bonding network and heme pocket rearrangement in ferric H117A Synechocystis hemoglobin. Biochemistry 43, 12622 – 12633. 34. Couture, M., and Guertin, M. (1996) Purification and spectroscopic characterization of a recombinant chloroplastic haemoglobin from the green alga Chlamydomonas eugametos. Eur. J. Biochem. 242, 779 – 787. 35. Visca, P., Fabozzi, G., Petrucca, A., Ciaccio, C., Coletta, M., De Sanctis, G., Bolognesi, M., Milani, M., and Ascenzi, P. (2002) The truncated hemoglobin from Mycobacterium leprae. Biochem. Biophys. Res. Commun. 294, 1064 – 1070. 36. Ouellet, H., Juszczak, L., Dantsker, D., Samuni, U., Ouellet, Y. H., Savard, P. Y., Wittenberg, J. B., Wittenberg, B. A., Friedman, J. M., and Guertin, M. (2003) Reactions of Mycobacterium tuberculosis truncated hemoglobin O with ligands reveal a novel ligand-inclusive hydrogen bond network. Biochemistry 42, 5764 – 5774. 37. Bolognesi, M., Rosano, C., Losso, R., Borassi, A., Rizzi, M., Wittenberg, J. B., Boffi, A., and Ascenzi, P. (1999) Cyanide binding to Lucina pectinata hemoglobin I and to sperm whale myoglobin: an X-ray crystallographic study. Biophys. J. 77, 1093 – 1099. 38. Ascenzi, P., Bocedi, A., de Sanctis, D., Pesce, A., Bolognesi, M., Marden, M. C., Dewilde, S., Moens, L., Hankeln, T., and Burmester, T., (2004) Neuroglobin and cytoglobin: two new entries in the hemoglobin superfamily. Biochem. Mol. Biol. Educ. 32, 305 – 313. 39. Hankeln, T., Ebner, B., Fuchs, C., Gerlach, F., Haberkamp, M., Laufs, T. L., Roesner, A., Schmidt, M., Weich, B., Wystub, S., Saaler-Reinhardt, S., Reuss, S., Bolognesi, M., de Sanctis, D., Marden, M. C., Kiger, L., Moens, L., Dewilde, S., Nevo, E., Avivi, A., Weber, R. E., Fago, A., and Burmester, T. (2005) Neuroglobin and cytoglobin in search of their role in the vertebrate globin family. J. Inorg. Biochem. 99, 110 – 119. 40. Pesce, A., Bolognesi, M., Bocedi, A., Ascenzi, P., Dewilde, S., Moens, L., Hankeln, T., and Burmester, T. (2002) Neuroglobin and cytoglobin. Fresh blood for the vertebrate globin family. EMBO Rep. 3, 1146 – 1151. 41. de Sanctis, D., Dewilde, S., Pesce, A., Moens, L., Ascenzi, P., Hankeln, T., Burmester, T., and Bolognesi, M. (2004) Crystal structure of cytoglobin: the fourth globin type discovered in man displays heme hexa-coordination. J. Mol. Biol. 336, 917 – 927. 42. Ermler, U., Siddiqui, R. A., Cramm, R., and Friedrich, B. (1995) Crystal structure of the flavohemoglobin from Alcaligenes eutrophus at 1.75 A˚ resolution. EMBO J. 14, 6067 – 6077. 43. Freitas, T. A., Hou, S., and Alam, M. (2003) The diversity of globin coupled sensors. FEBS Lett. 552, 99 – 104. 44. Bonamore, A., Attili, A., Arenghi, F., Catacchio, B., Chiancone, E., Morea, V., and Boffi, A. (2007) A novel chimera: the ‘‘truncated hemoglobin-antibiotic monooxygenase’’ from Streptomyces avermitilis. Gene in press.