Unique features of the hemoglobin system of the Antarctic notothenioid fish Gobionotothen gibberifrons Panagiotis Marinakis, Maurizio Tamburrini, Vito Carratore and Guido di Prisco Institute of Protein Biochemistry, CNR, Naples, Italy ThehemolysateoftheAntarcticteleostGobionotothen gibberifrons (family Nototheniidae) contains two hemoglo- bins (Hb 1 and Hb 2). The concentration of Hb 2 (15–20% of the total hemoglobin content) is higher than that found in most cold-adapted Notothenioidei. Unlike the other Antarctic species so far examined having two hemoglobins, Hb 1 and Hb 2 do not have globin chains in common. Therefore this hemoglobin system is made of four globins (two a-andtwob-chains). The complete amino-acid sequence of the two hemoglobins (Hb 1, a 1 2 b 1 2 ;Hb2,a 2 2 b 2 2 ) has been established. The two hemoglobins have different functional properties. Hb 2 has lower oxygen affinity than Hb 1, and higher sensitivity to the modulatory effect of organophosphates. They also differ thermodynamically, as shown by the effects on the oxygen-binding properties brought about by temperature variations. The oxygen- transport system of G. gibberifrons, with two functionally distinct hemoglobins, suggests that the two components may have distinct physiological roles, in relation with life style and the environmental conditions which the fish may have to face. The unique features of the oxygen-transport system of this species are reflected in the phylogeny of the hemoglobin amino-acid sequences, which are intermediate between those of other fish of the family Nototheniidae and of species of the more advanced family Bathydraconidae. Keywords: fish; Antarctic; hemoglobin; amino-acid sequence; oxygen binding. Organisms living in extreme environments, such as the Arctic and Antarctic sea waters, are exposed to strong constraints, among which temperature is often a driving factor [1–4]. Hemoglobin (Hb), a direct link between the exterior and body requirements, has thus experienced a major evolutionary pressure in these organisms to adapt its functional features at molecular/functional level. The search for correlation between fish hematology and the extreme conditions of the Antarctic environment leads to a study on the biochemistry of oxygen transport, centred on Hb molecular structure and oxygen-binding properties, taking the ecological constraints under con- sideration. In view of the role of temperature in modifying the oxygenation-deoxygenation cycle in respir- ing tissues, thermodynamic analysis deserves special attention. The largely dominant Antarctic suborder Notothenioidei is by far the most thoroughly characterized group of fish in the world. Thirty-five species (all bottom dwellers) of the 38 so far investigated were shown to have a single major Hb (Hb 1) and often a minor one (Hb 2, 5% of the total Hb content, generally having the b-chain in common with Hb 1) both displaying, with some exceptions, strong Bohr and Root effects [2,5]. Each of the remaining three species (Trematomus newnesi and Pagothenia borchgrevinki,two active cryopelagic species; Pleuragramma antarcticum,a pelagic, sluggish but migratory fish) all belonging to the family Nototheniidae, have a unique oxygen-transport system, and each system appears adjusted to the fish specific life style, substantially different from that of the sluggish benthic species. Compared with other Notothenioidei, the Antarctic teleost, Gobionotothen gibberifrons (family Nototheniidae), is endowed with novel hematological features. A detailed study of the oxygen-transport system is herewith reported. A preliminary communication on the Hb system of this species has appeared previously [6]. G. gibberifrons lives at a depth range between 5 and 750 m in the waters of northern Antarctic Peninsula and of the islands located north-east. G. gibberifrons has Hb 1 and Hb 2. The complete amino- acid sequence of the two components has been established, and the regulation of oxygen binding by pH, allosteric effectors (chloride and organophosphates) and temperature has been investigated. Materials and methods Toyopearl Super Q-650S was from TosoHaas (Laboratory Service Analytical); trypsin (EC 3.4.21.4) treated with L -1-tosylamide-2-phenylethylchloromethylketone from Cooper Biomedical; dithiothreitol from Fluka; Tris, bis- Tris, Hepes, Mes, 4-vinylpyridine and IHP from Sigma; sequanal-grade reagents from Applied Biosystems; Correspondence to G. di Prisco, Institute of Protein Biochemistry, CNR, Via Marconi 12, I-80125 Naples, Italy. Fax: + 39 0815936689; Tel.: + 39 0817257242; E-mail: diprisco@dafne.ibpe.na.cnr.it Abbreviations: Hb, haemoglobin; OPA, o-phthalaldehyde; IHP, inositol hexakisphosphate; P 50 , partial pressure of oxygen required to achieve Hb half saturation. Note: The protein sequence data reported in this paper will appear in the SWISS-PROT and TrEMBL knowledgebase under the accession numbers P83611 (a 1 chain), P83612 (b 1 chain), P83613 (a 2 chain) and P83614 (b 2 chain). (Received 7 July 2003, revised 1 August 2003, accepted 7 August 2003) Eur. J. Biochem. 270, 3981–3987 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03786.x HPLC-grade acetonitrile from Laboratory-Scan Analytical. All other reagents were of the highest purity commercially available. Specimens of G. gibberifrons were caught at Dallmann Bay and Low Island (63°25¢S, 62°15¢W), onboard the research vessels R/V Hero and R/S Polar Duke. Blood samples were drawn from the caudal vein by means of heparinized syringes. Hemolysates were prepared as des- cribed [7]. Stripping of endogenous ligands was carried out by running aliquots through a small column of Dowex AG 501 X8 (D), a mixed bed ion-exchange resin. Separation of Hbs was achieved by FPLC anion- exchange chromatography on a Toyopearl Super Q-650S column (1.5 · 10 cm). Elution was carried out with a gradient from 0 to 50% of buffer B (500 m M Tris/HCl, pH 7.6) in buffer A (10 m M Tris/HCl, pH 7.6) in 75 min. The flow rate was 0.5 mLÆmin )1 ;theabsorbancewas measured at 546 nm. The Hb-containing pooled fractions were dialysed against 10 m M Hepes pH 7.7. All steps were carried out at 0–5 °C. No oxidation was spectrophoto- metrically detectable. Hb solutions were stored in small aliquots at )80 °C until use. Globin separation was accomplished by reverse-phase HPLC of purified Hbs on a lBondapak C 18 column (Waters, 3.9 · 300 mm). Elution was carried out with a gradient from 0 to 100% of eluent B (60% acetonitrile) in eluent A (45% acetonitrile, containing 0.3% trifluoroacetic acid) in 32.5 min. The flow rate was 1 mLÆmin )1 ;the absorbance was followed at 280 nm. Alkylation of the sulfhydryl groups with 4-vinylpyridine, deacetylation of the a-chain N-terminus and tryptic diges- tions were carried out as described [8,9]. Tryptic peptides were purified by reverse-phase HPLC on a lBondapak-C 18 column (Waters, 3.9 · 300 mm). Clea- vage of Asp-Pro bonds was performed on polybrene-coated glass-fibre filters in 70% (v/v) formic acid, for 24 h at 42 °C [10]. Asp-Pro-cleaved globins were treated with OPA before sequencing [11] in order to block the non-Pro N-terminus and reduce the background. Sequencing was performed with an Applied Biosystems Procise 492 automatic sequencer, equipped with on-line detection of phenylthiohydantoin amino acids. ThemolecularmassoftheS-pyridylethylated a-and b-chains and of peptides of less than 10 kDa was measured by MALDI-TOF mass spectrometry on a PerSeptive Biosystems Voyager-DE Biospectrometry Workstation. Analyses were performed on premixed solutions prepared by diluting samples (final concentra- tion, 5 pmolÆlL )1 ) in four volumes of matrix, namely (a) 10 mgÆmL )1 sinapinic acid in 30% acetonitrile containing 0.3% trifluoroacetic acid (v/v/v; for globin analysis), and (b) 10 mgÆmL )1 a-cyano-4-hydroxycinnamic acid in 60% acetonitrile containing 0.3% trifluoroacetic acid (v/v/v; for peptide analysis). Oxygen-saturation curves were determined as described [8]. Oxygen equilibria were measured at 5 °Cand10°C, in 100 m M Hepes buffer (pH range 6.0–8.0) prepared at the temperature of the oxygen-binding measurements. The final Hb concentration was 0.5–1.0 m M on a heme basis. An average standard deviation of ± 3% for values of P 50 was calculated. Experiments were performed in duplicate. To measure stepwise oxygen saturation, a modified gas diffu- sion chamber (Eschweiler) was used, coupled to cascaded Wo ¨ sthoff pumps for mixing pure nitrogen with air [12,13]. Absorbance variations between deoxygenated and oxygen- ated Hb were measured at 436 nm with an Eppendorf spectrophotometer model 1101 M. pH values were meas- ured at the end of each experiment with a Radiometer BMS Mk2 thermostatted electrode. Sensitivity to chloride was assessed by adding NaCl to a final concentration of 100 m M . The effects of IHP were measured at a final ligand concentration of 3 m M , namely a large excess over the concentration of tetrameric Hb. Oxygen affinity (denoted by P 50 ) and cooperativity (n H )werecalculatedfromthe linearized Hill plot of log S/(1-S) vs. log P O2 at half saturation; S denotes fractional oxygen saturation. The amplitude of the Bohr effect is given by the Bohr coefficient, / ¼ Dlog P 50 /DpH. The overall oxygenation enthalpy change DH (kcalÆmol )1 ;1kcal¼ 4.184 kJ), corrected for the heat of oxygen solubilization ()3kcalÆmol )1 ), was calculated by the integrated van’t Hoff equation, DH ¼ ) 4.574[(T 1 T 2 )/ (T 1 –T 2 )]Dlogp 50 /1000. Results Hb and globin purification Cellulose acetate electrophoresis showed that the hemo- lysate of G. gibberifrons contains two Hbs (Hb 1 and Hb 2), accounting for 80–85% and 15–20%, respectively, of the total Hb content. The two Hbs were purified by ion- exchange chromatography on a Super Q ToyoPearl column (Fig. 1). Hb 2 often appeared to be contaminated by Hb 1; a second chromatography on the same column yielded pure Hb 2 (not shown). Reverse-phase HPLC of the hemolysate showed two major and two minor peaks (Fig. 2). HPLC of Hb 1 showed two peaks, whose elution times corresponded to those of the major peaks of the hemolysate; Hb 2 showed two peaks having elution times corresponding to those of the minor peaks. The molecular mass values (Da), obtained by MALDI-TOF mass spectrometry, were 15 597 and Fig. 1. Ion-exchange chromatography of G. gibberifrons hemolysate on a Toyopearl column. Details are given in Materials and methods. 3982 P. Marinakis et al.(Eur. J. Biochem. 270) Ó FEBS 2003 16 097 (a 1 and b 1 chains, respectively, of Hb 1), and 15 800 and 16 420 (a 2 and b 2 of Hb 2). Amino-acid sequencing The primary structure of the globins was established by sequencing intact proteins, internal regions obtained after specific hydrolysis of Asp-Pro bonds, and tryptic peptides purified by reverse-phase HPLC. a-Chains Direct sequencing of intact a-chains was unsuccessful, suggesting that (as in all Antarctic fish Hbs examined so far) the N-terminal residue is blocked, therefore not available to direct Edman degradation. The molecular masses of the N-terminal tryptic peptides of a 1 and a 2 , measured by MALDI-TOF mass spectrometry, were 43 Da higher than those found after sequencing the unblocked peptides, thus confirming that the a-chain N-terminus is acetylated. Following cleavage of the Asp-Pro bonds, sequencing proceeded from Pro96 to Asp127 in a 1 , and from Pro96 to Lys140 in a 2 . In a 1 , cleavage at Lys5 and at Arg93 by trypsin was not complete, therefore peptides T1–T2 and T12–T13 were also found. The peptide bond after Lys116 was not cleaved at all. Also the peptide bonds after Lys61 and Lys62 were not completely cleaved, generating peptides T10a and T10b which coeluted. Four additional peptide pairs (T1 and T1– T2; T3 and T16; T6 and T13; T7 and T12–T13) coeluted from the column; however, sequences were unambiguosly established on the basis of their different amounts. In a 2 , trypsin failed to cleave the peptide bond after Lys7. Two peptide pairs (T1 and T10; T5 and T9) coeluted; again, sequences were established on the basis of their different amounts. Sequence 101–140 was established only after Asp- Pro cleavage. Figure 3A,B shows the complete amino-acid sequences of the a chains. Tryptic peptides were aligned on the basis of sequence homologies with known globin sequences, and with the sequences obtained following Asp-Pro cleavage. Each chain is made of 142 residues. The molecular masses, calculated from the sequence, are 15 605 and 15,790, for a 1 and a 2 , respectively. b-Chains Direct sequencing proceeded for 20 and 31 residues for the b-chain of Hb 1 (b 1 )andHb2(b 2 ), respectively. After cleavage of the Asp–Pro bond, the internal sequences from Pro100 to Leu134 in b 1 , and from Pro100 to Lys132 in b 2 , were established. In b 1 , T2 and T11 were not found, and their sequence was directly established from the N-terminus of the intact globin (T2) and from the internal sequence obtained after cleavage of the Asp-Pro bond (T11). T10 and T12 coeluted in the same chromatographic peak, and their sequence was established on the basis of their different amount. In b 2 , T2 and T6 coeluted in the same peak. Being the sequence of T2 already known from the N-terminus, the sequence of T6 was established by difference analysis. Figure 3C,D reports the complete sequences of the b chains. Tryptic peptides were aligned as described for the a chains, and with the sequences obtained from the N-terminus. Each chain is made of 146 residues. The molecular masses, calculated from the sequence, are 16 081 and 16 400 for b 1 and b 2 , respectively. Oxygen binding Functional studies were carried out on Hb 1 and Hb 2, determining the oxygen-binding curves as a function of pH in the temperature range 5–10 °C, in the absence and presence of allosteric effectors (Fig. 4 and Table 1). In the pH range examined, the oxygen affinity of Hb 2 was lower than that of Hb 1. All samples displayed the alkaline Bohr effect, slightly enhanced by chloride and, to a higher extent, by organophosphate. The latter had a very strong effect at alkaline pH values, especially in Hb 2, which, although apparently reducing the amplitude of the Bohr coefficient (/) in the pH range examined, is indicative of a stronger overall Bohr effect. In all samples oxygen binding was cooperative above pH 6.5 in the absence of effectors. Chloride and, to a higher extent, phosphate, enhanced the decrease in oxygen-binding cooperativity brought about by increasing proton concentration. In fact, IHP virtually abolished cooperativity from pH 7.5 downwards. All samples displayed the Root effect, which was maximal in the pH range 6.5–7.5 (Fig. 5). Oxygen-satura- tion curves were determined at atmospheric pressure. In the Fig. 2. Reverse-phase HPLC of G. gibberifrons hemolysate, Hb 1 and Hb 2 (A, B and C, respectively). Details are given in Materials and methods. Ó FEBS 2003 The hemoglobins of G. gibberifrons (Eur. J. Biochem. 270) 3983 Fig. 3. Amino-acid sequences of a 1 , a 2 , b 1 and b 2 globin chains (A, B, C and D, respectively). The tryptic peptides (T) and the sequence portions elucidated by automated Edman degradation from the N-terminus and after cleavage of an Asp-Pro bond are indicated below the sequences. 3984 P. Marinakis et al.(Eur. J. Biochem. 270) Ó FEBS 2003 absence of IHP, complete saturation was achieved at pH 7.5 with Hb 1 and Hb 2. At pH 6.0, the saturation of Hb 1 and Hb 2 was 73 and 56%, respectively. In the presence of IHP, maximal saturation was achieved at pH 8.0 in Hb 2, and at pH 7.5 in Hb 1. At pH 6.0, the saturation of Hb 1 and Hb 2 was 58 and 47%, respectively. In all samples, IHP shifted the inflexion of the curve (corresponding to maximal Root effect) towards more alkaline pH. Temperature variations had different effects on the oxygen binding of Hb 1 and Hb 2, both in the absence and presence of allosteric effectors (Fig. 6). Compared to Hb 1, Hb 2 showed more exothermic DH values at acidic pH, whereas lower values were measured at pH 8.0. In the presence of chloride, the DH values of Hb 2 were higher (in absolute value) than those of Hb 1 in the entire pH range. Oxygen-binding studies were also carried out on intact erythrocytes and stripped hemolysate (not shown), which showed intermediate functional properties between those of Hb 1 and Hb 2. Erythrocytes contain endogenous organo- phosphates; consequently, the curves are similar to those obtained with the stripped hemolysate in the presence of effectors. Discussion In the Antarctic suborder Notothenioidei, most species of the family Nototheniidae have one major and one minor Hb (Hb 1 and Hb 2, 95% and 5% of the total, respectively) [2, 4,14]. The two Hbs have the b-chain in common, with the exception of those of Cygnodraco mawsoni [15] which share the a-chain. Therefore, Nototheniidae (and all Notothe- nioidei) are generally characterized by reduced Hb multi- plicity compared to many teleosts of temperate waters [2]. In T. newnesi, P. antarcticum and P. borchgrevinki higher multiplicity was observed [16–18], but these species are pelagic and migratory, differing in life style from the other notothenioids, which are in general sluggish, benthic fish. G. gibberifrons is also a sluggish, benthic nototheniid. Not much more is known about its life style. However, in comparison with all other benthic nototheniids, it has distinct and novel hematological features. The blood has the highest level (approx. 15–20%) of the minor component Hb 2 ever found in Antarctic Notothenioidei; unlike other nototheniids, in which Hb 2 tends to disappear in adults (e.g. T. bernacchii and D. mawsoni only have Hb 2 in juveniles), the level of Hb 2 does not decrease in adult fish. Moreover, unlike all other species, the two Hbs of G. gibberifrons do not have any globin in common. An Hb system where two components are made by four chains is a unique case among Notothenioidei. Four chains instead of three might well be a feature of speciation in the pathway of evolution of the suborder. Fig. 4. Oxygen-equilibrium isotherms (Bohr effect) and subunit coop- erativity, as a function of pH, of Hb 1 (A, B, respectively) and Hb 2 (C, D, respectively). Experiments were carried out at 5 °C in 100 m M Hepes or Mes buffers, in the absence of effectors (s), in the presence of 100 m M NaCl (m), and of 100 m M NaCl, 3 m M IHP (j). Table 1. Bohr coefficients (/), calculated from the oxygen-binding curves determined in the absence of effectors and in the presence of 100 m M NaCl or 100 m M NaCl and 3 m M IHP. No effectors NaCl (100 m M ) NaCl/IHP (100/3 m M ) Hb 1 5 °C )0.64 )0.69 )0.62 10 °C )0.57 )0.68 )0.61 Hb 2 5 °C )0.75 )0.77 )0.43 10 °C )0.89 )0.77 )0.61 Fig. 5. Oxygen-saturation curves at atmospheric pressure (Root effect) ofHb1(A)andHb2(B).Experiments were carried out at 2 °C, in 100 m M Tris/HCl or bisTris/HCl buffers, in the absence (s)and presence (d)of3m M IHP. Ó FEBS 2003 The hemoglobins of G. gibberifrons (Eur. J. Biochem. 270) 3985 Although from a morphological point of view G. gibber- ifrons evidently belongs to the family Nototheniidae [19–21], Tokita et al. [22] have reported dendrograms based on genetic distances obtained from two-dimensional gel elec- trophoresis of total protein constituents of cardiac muscle. In these dendrograms the nototheniid G. gibberifrons appears more closely related to species belonging to the most phyletically advanced notothenioid families Bathydraconidae and Channichthyidae, than to the clade composed by three other nototheniid species. As far as red- blooded species are concerned, this interesting conclusion is not fully supported by our results on Hb sequences (Table 2). In fact, the sequence identities do not indicate clear grouping of G. gibberifrons either with species of the same family (Trematomus bernacchii and T. newnesi)orof Bathydraconidae (Gymnodraco acuticeps and C. mawsoni). However, the lack of a clear relationship with the Hb sequences of other nototheniid species suggests that in this case the evolution of the oxygen-transport system has occurred in response to special needs of this species, as shown by the intermediate position taken by G. gibberifrons Hbs in phylogenetic trees [23]. In line with other Antarctic Hbs, the sequence identity between the a chains of the two Hbs of G. gibberifrons is 68%; it is 70% between the b-chains. In summary, the a-chain and the extra b-chain of Hb 2 have much higher sequence identity with minor than with major Hbs of other Antarctic species. Thus the latter extra chain also groups with minor Hbs. In Antarctic fish Hbs, Aspb94, which in human HbA establishes a strong ionic bond with Hisb146, important for the Bohr-effect mechanism [24], is generally conservatively replaced by Glu. Moreover, Vala1 is always replaced by Ac- Ser, which cancels the contribution of the N-terminus to the Bohr effect. These substitutions are also found in the two Hbs of G. gibberifrons, characterized by the Bohr and Root effects. However, these Hbs show significantly distinct Bohr coefficients and amplitude of Root effect; in particular, Hb 2 has lower oxygen affinity than Hb 1 in the pH range examined, and phosphate modulation of the affinity is stronger in Hb 2 than Hb 1. These differences might be due to other substitutions in the primary structure. For instance, in Hb 2 it is worth noting that position b82, which is part of the phosphate binding site [25,26], is occupied by Lys, whereas in Hb 1 the latter residue is replaced by Ala. This substitution may well account for the lower effect of organophosphates in the latter Hb. Temperature dependence, which is governed by the associated overall enthalpy change, is an important feature of the reaction of Hbs with oxygen. Heat absorption and release can be considered physiologically relevant modula- ting factors, similar to hetero and homotropic ligands. The two Hbs of G. gibberifrons also differ in thermodynamic behaviour. Hb 1 is less sensitive to temperature variations than Hb 2 which, in turn, shows strong variations of enthalpy change especially at pH below 7.5, depending on chloride and/or phosphate. In particular, Hb 2 shows a progressive increase of the exothermic enthalphy change as a function of proton concentration. This feature has never been reported in fish Hbs. Chloride virtually abolishes this exothermic change by providing a strong endothermic contribution to oxygen binding. Although a molecular inter- pretation is hard to find, this differential thermodynamic Fig. 6. Oxygenation enthalpy of Hb 1 (A) and Hb 2 (B). DH values were calculated in the temperature range 5–10 °C from the oxygen- binding data reported in Fig. 4 and Table 1. Experimental conditions were: 100 m M Hepes or Mes buffers, in the absence of effectors (s), in the presence of 100 m M NaCl (m), and of 100 m M NaCl, 3 m M IHP (j). Table 2. Sequence identity (%) between a- and b-chains of G. gibb erifrons and of some other Antarctic fish Hbs. T. bernacchii Hb C, P. borchgrevinki Hb 0 and C. mawsoni Hb 2 are minor components. P. borchgrevinki Hb 0 and Hb 1, in addition to C. mawsoni Hb 1 and Hb 2, share the a-chain. T. bern. Hb 1 T. bern. Hb C P. borch. Hb 1 P. borch. Hb 0 G. acut. Hb C. maws. Hb 1 C. maws. Hb 2 a-chains G. gibberifrons Hb 1 94 – 90 90 89 90 90 G. gibberifrons Hb 2 69 – 65 65 66 68 68 b-chains G. gibberifrons Hb196719171838867 G. gibberifrons Hb268927091666587 3986 P. Marinakis et al.(Eur. J. Biochem. 270) Ó FEBS 2003 regulation of the oxygenation/deoxygenation cycle (that may play a significant role in keeping the internal tempera- ture constant) may be an important adaptive tool. This paper reports some novel features of the oxygen transport of a notothenioid fish species. The results suggest that G. gibberifrons Hb 2 cannot merely be considered an evolutionary remnant, as in other Antarctic Notothenioidei [27]. The functional differences suggest that Hb 2, rather than being a vestigial or larval remnant, may indeed have a physiological role; the two Hbs of this cold-adapted teleost might be used alternatively to face special needs in relation with life style and different environmental conditions (e.g. temperature fluctuations during migration) requiring fine regulation of oxygen binding. Finally, although in an organism biosynthesis of higher amounts of an additional Hb can be easily accomplished and may be considered a short-time response to environmental changes, preservation of the role of the gene duplication which has produced an additional chain is a physiologically complex long-term response, and may well be considered an evolutionarily important adaptation. Acknowledgements This study is in the framework of the Italian National Programme for Antarctic Research. References 1. di Prisco, G. & Tamburrini, M. (1992) The hemoglobins of marine and freshwater fish: the search for correlations with physiological adaptation. Comp. Biochem. Physiol. 102B, 661–671. 2. di Prisco, G. (1998) Molecular adaptations in Antarctic fish hemoglobins. In Fishes of Antarctica. A Biological Overview (di Prisco, G., Pisano, E. & Clarke, A., eds), pp. 339–353. Springer, New York. 3. di Prisco, G., Tamburrini, M. & D’Avino, R. (1998) Oxygen- transport systems in extreme environments: multiplicity and structure/function relationship in hemoglobins of Antarctic fish. In Cold Ocean Physiology (Po ¨ rtner, H.O. & Playle, R., eds), pp. 143–165. Cambridge University Press, Cambridge, UK. 4. Tamburrini, M. & di Prisco, G. (2000) Oxygen-transport system and mode of life in Antarctic fish. In Hemoglobin Function in Vertebrates. Molecular Adaptation in Extreme and Temperate Environments (di Prisco, G., Giardina, B. & Weber, R.E., eds), pp. 51–59. Springer-Verlag, Italy. 5. di Prisco, G., D’Avino, R. & Tamburrini, M. (1999) Structure and function of hemoglobins from Antarctic organisms: the search for correlations with adaptive evolution. In Cold-Adapted Organisms. Ecology, Physiology, Enzymology and Molecular Biology (Margesin, R. & Schinner, F., eds), pp. 239–253. Springer-Verlag, Berlin, Germany. 6. di Prisco, G. (1988) A study of hemoglobin in Antarctic fishes: purification and characterisation of hemoglobin from four species. Comp. Biochem. Phisiol. 90B, 631–637. 7. D’Avino, R. & di Prisco, G. (1988) Antarctic fish hemoglobin: an outline of the molecular structure and oxygen binding properties. I. Molecular structure. Comp.Biochem.Physiol.90B, 579–584. 8. D’Avino, R. & di Prisco, G. (1989) Hemoglobin from the Ant- arctic fish Notothenia coriiceps neglecta. 1. Purification and char- acterisation. Eur. J. Biochem. 179, 699–705. 9. Tamburrini, M., Brancaccio, A., Ippoliti, R. & di Prisco, G. (1992) The amino acid sequence and oxygen-binding properties of single hemoglobin of the cold-adapted Antarctic teleost Gymnodraco acuticeps. Arch. Biochem. Biophys. 292, 295–302. 10. Landon, M. (1977) Cleavage at aspartyl-prolyl bonds. In Methods Enzymology,Vol.47(Hirs,C.H.W.&Timasheff,S.N.,eds),pp. 145–149. Academic Press, New York, USA. 11. Brauer, A.W., Oman, C.L. & Margolies, M.N. (1984) Use of o-phthalaldehyde to reduce background during automated Edman degradation. Anal. Biochem. 137, 134–142. 12. Sick, H. & Gersonde, K. (1969) Method for continuous registra- tion of O 2 -binding curves of hemoproteins by means of a diffusion chamber. Anal. Biochem. 32, 362–376. 13. Lykkeboe, G., Johansen, K. & Maloy, G.M.O. (1975) Functional properties of hemoglobin in the teleost Tilapia grahami. J. Comp. Physiol. 104, 1–11. 14. di Prisco, G. (1997) Physiological and biochemical adaptations in fish to a cold marine environment. In Proceedings of the SCAR 6th Biol. Symposium, Venice (Antarctic Communities: Species, Structure and Survival) (Battaglia, B., Valencia, J. & Walton, D.W.H., eds), pp. 251–260. Cambridge University Press, Cambridge, UK. 15. Caruso,C.,Rutigliano,B.,Romano,M.&diPrisco,G.(1991) The hemoglobins of the cold-adapted Antarctic teleost Cygno- draco mawsoni. Biochim. Biophys. Acta 1078, 273–282. 16. D’Avino, R., Caruso, C., Tamburrini, M., Romano, M., Ruti- gliano, B., Polverino de Laureto, P., Camardella, L., Carratore, V. & di Prisco, G. (1994) Molecular characterization of the functionally distinct hemoglobins of the Antarctic fish Tremato- mus newnesi. J. Biol. Chem. 269, 9675–9681. 17. Tamburrini, M., D’Avino, R., Fago, A., Carratore, V., Kunz- mann, A. & di Prisco, G. (1996) The unique hemoglobin system of Pleuragramma antarcticum, an Antarctic migratory teleost. Structure and function of the three components. J. Biol. Chem. 271, 23780–23785. 18. Riccio, A., Tamburrini, M., Carratore, V. & di Prisco, G. (2000) Functionally distinct haemoglobins of the cryopelagic Antarctic teleost Pagothenia borchgrevinki. J. Fish Biol. 57A, 20–32. 19. Iwami, T. (1985) Osteology and relationships of the family Channichthyidae. Mem. Natl. Institute Polar Res. Tokyo Series E. 36, 1–69. 20. Gon, O. & Heemstra, P.C., eds (1990) Fishes of the Southern Ocean. Smith Institute of Ichthyology, South Africa. 21. Balushkin, A.V. (1992) Classification, phylogenetic relationships, and origins of the families of the suborder Notothenioidei (Perciformes). J. Ichthyol. 32, 90–110. 22. Tokita, M., Ishii, S., Iwami, T. & Miyazaki, J I. (2002) Phylo- genetic analysis of Antarctic notothenioid fishes based on two- dimensional gel electrophoresis. Polar Biol. 25, 163–168. 23. Verde, C., Carratore, V., Riccio, A., Tamburrini, M., Parisi, E. & di Prisco, G. (2002) The functionally distinct hemoglobins of the Arctic spotted wolffish Anarhichas minor. J. Biol. Chem. 277, 36312–36320. 24. Perutz, M.F. & Brunori, M. (1982) Stereochemistry of cooperative effects in fish and amphibian haemoglobins. Nature 299, 421–426. 25. Arnone, A. (1972) X-ray diffraction study of binding of 2,3- diphosphoglycerate to human deoxyhaemoglobin. Nature 237, 146–149. 26. Arnone, A. & Perutz, M.F. (1974) Structure of inositol hexa- phosphate-human deoxyhaemoglobin complex. Nature 249, 34–36. 27. di Prisco, G., D’Avino, R., Caruso, C., Tamburrini, M., Cam- ardella, L., Rutigliano, B., Carratore, V. & Romano, M. (1991) The biochemistry of oxygen transport in red-blooded Antarctic fish. In Biology of Antarctic Fish (diPrisco,G.,Maresca,B.& Tota, B., eds), pp. 263–281. Springer-Verlag, New York, USA. Ó FEBS 2003 The hemoglobins of G. gibberifrons (Eur. J. Biochem. 270) 3987 . Unique features of the hemoglobin system of the Antarctic notothenioid fish Gobionotothen gibberifrons Panagiotis Marinakis,. phylogeny of the hemoglobin amino-acid sequences, which are intermediate between those of other fish of the family Nototheniidae and of species of the more