Comparative Biochemistry and Physiology Part B 124 (1999) 147 – 155 www.elsevier.com/locate/cbpb The Antarctic toothfish (Dissostichus mawsoni ) lacks plasma albumin and utilises high density lipoprotein as its major palmitate binding protein Victoria J. Metcalf *, Stephen O. Brennan, Peter M. George Molecular Pathology Laboratory, Department of Pathology, Christchurch School of Medicine, Uni6ersity of Otago, Christchurch, New Zealand Received 23 November 1998; received in revised form 26 February 1999; accepted 3 March 1999 Abstract Plasma from the Antarctic toothfish, Dissostichus mawsoni, a member of the advanced teleost Nototheniidae family, was analysed. Agarose gel electrophoresis showed a major diffuse anionic protein that bound [14C]palmitic acid but not 63Ni2 + , and two more cationic proteins that bound 63Ni2 + but not palmitate. Oil Red O staining following cellulose acetate electrophoresis indicated that the palmitate binding protein was a lipoprotein. Two-dimensional electrophoresis showed that this palmitate binding band was composed of three proteins with Mr of 11, 30, and 42 kDa, without any trace of material at  65 kDa, the mass of albumin. N-terminal sequencing of the palmitate binding band gave a major sequence of DAAQPSQELR-, indicating a high degree of homology to apolipoprotein A-I (apo-AI), the major apolipoprotein of high density lipoprotein (HDL). N-terminal sequencing of the major nickel binding band produced a sequence with no homology to albumin. When ultracentrifugation was used to isolate the lipoproteins from Antarctic toothfish plasma, the palmitate binding protein was found solely in the lipoprotein fraction. In competitive binding experiments, added human albumin did not prevent palmitate binding to toothfish HDL. In conclusion, there is no evidence for albumin in Antarctic toothfish plasma and HDL assumes the role of fatty acid transport. © 1999 Elsevier Science Inc. All rights reserved. Keywords: Albumin; Antarctic toothfish; Apo-AI; Apo-AIV; Fatty acid; High density lipoprotein (HDL); Palmitate; Protein sequence 1. Introduction The waters around Antarctica are constantly at subzero temperatures and have been since the formation of the continental ice sheet 10– 15 mya (million years ago) [29]. The Antarctic Ocean is characterised by little temperature fluctuation, high water viscosity, and increased oxygen solubility. Despite this inhospitable marine environment, over 1% of the world’s fish species are found in Antarctic waters [19,20], with more than 50% of Antarctic fish species belonging to the Perciformes suborder Notothenioidei [4]. The Order Per- * Corresponding author. PO Box 4345, Christchurch; Tel.: +64-33640548; fax: + 64-3-3640545. E-mail address: victoria.metcalf@chmeds.ac.nz (V.J. Metcalf) ciformes was the most recent teleost order to evolve, making species of the suborder Notothenioidei among the most phyletically derived of fish. The formation of the Antarctic Convergence current resulted in the isolation of the southern ocean 23– 30 mya [21]. The fishes of the suborder Notothenioidei are thought to have evolved in situ as the Antarctic continent moved south and the surrounding ocean cooled. The notothenioids found in the Antarctic Ocean are characterised by the presence of antifreeze glycopeptides (AFGPs), while those species with habitats outside the Convergence current, such as the earliest notothenioids, the Bovichtidae family, do not possess AFGPs [4,20]. The AFGP genes evolved from a pancreatic trypsinogen gene [13], with the multiple isoforms collectively maintained at an extremely high plasma concentration of 30– 35 mg ml − 1 [15,16]. Another notable characteristic of the notothenioids is a decrease in 0305-0491/99/$ - see front matter © 1999 Elsevier Science Inc. All rights reserved. PII: S 0 3 0 5 - 0 4 9 1 ( 9 9 ) 0 0 0 5 1 - 6 148 V.J. Metcalf et al. / Comparati6e Biochemistry and Physiology, Part B 124 (1999) 147–155 blood viscosity due to a lowering of the hematocrit (ratio of red cell volume to total blood volume) compared to other teleost species [57]. The icefishes of the Channichthyidae family have further reduced blood viscosity through a complete absence of haemoglobin [26,50]. Notothenioids also lack a swim bladder [20]. The current hypothesis of notothenioid evolution, using mtDNA divergence rates, is that after the evolution of the Bovichtidae family 10– 15 mya, rapid speciation of the other families occurred [4,49]. This recent divergence time is consistent with the divergence time of 5 – 14 my between the trypsinogen and AFGP gene [13], and the estimated 10– 15 mya time frame of the freezing of the Antarctic Ocean [29]. The notothenioid genus Dissostichus is composed of just two species; D. mawsoni and D. eleginoides [3], with the former endemic to the seas around Antarctica and possessing AFGP, and the latter found outside the Antarctic Convergence and lacking AFGP [20]. The Antarctic toothfish, D. mawsoni, is a relatively inactive piscivorous epibenthic species that attains a length of 175 cm and a weight of 80 kg or more [20,25]. D. mawsoni maintains neutral buoyancy in the absence of a swim bladder by decreasing the extent of bone and scale mineralisation, substituting cartilage for bone, and also by the deposition of large amounts of lipid in the form of triacylglycerols in its muscles [18,20]. Usually teleosts utilise wax esters to maintain neutral buoyancy [39,40,46], and it is possible that the choice of triacylglycerols in D. mawsoni may reflect dual use as both buoyancy lipids and as an energy reserve during the winter months [51]. Its hepatocytes also contain numerous lipid droplets, with the liver being unique in containing many perisinusoidal cells specialised for protein synthesis [17]. It is postulated that the product of these cells transports lipids in plasma. Due to the inhospitable environment D. mawsoni lives in and its requirement for large amounts of lipid to maintain neutral buoyancy, lipid and long-chain fatty acid (LCFA) transport is of major importance. Highly insoluble by themselves [54], LCFA are transported by albumin in most vertebrates [48]. Albumin is typically the major anionic protein in vertebrate plasma, and has roles in maintaining blood oncotic pressure as well as acting as a transport molecule for a variety of ligands. Albumin binds long-chain fatty acids freed from lipids through enzymatic action, resulting in the stimulation of lipoprotein lipase activity in adipose tissue [10], as well as stimulating LCAT (lecithin – cholesterol acyltransferase) activity [56]. The presence of albumin in the Salmonidae family has been demonstrated by cDNA sequencing in Atlantic salmon [9], and N-terminal sequencing of albumin proteins from chinook salmon and brown trout [36]. There has been doubt however, as to whether albumin is present in all teleost fish [14]. The plasma lipoproteins HDL (high density lipoprotein), LDL (low density lipoprotein) and VLDL (very low density lipoprotein) also play important roles in lipid transport and metabolism. The HDL apolipoprotein, apoAI (apolipoprotein A-I), like albumin, functions as an activator of LCAT [24]. HDL is found in low concentrations in mammals ( 2 mg ml − 1 in humans), but occurs in higher concentrations in the plasma of teleost fish (2.4 – 33 mg ml − 1) [12]. The wide range of HDL concentrations suggests that it may have varying importance as a fatty acid transporter in teleost species. It is the major plasma protein in carp [30], and HDL concentrations in rainbow trout were found to be very high at 15 mg ml − 1 [11,53]. The authors recently reported that New Zealand eels lack albumin in their plasma, and that HDL acts to transport LCFA in these species [37]. As in higher vertebrates, apo-AI and apo-AII (apolipoprotein A-II) are the major HDL apolipoproteins in teleost fish [2,11,37,53]. Thus, while albumin is the major LCFA transporter in salmonids, at least two eel species lack albumin and use HDL to transport LCFA. This raises the question, how do other teleost families transport LCFA? The Antarctic toothfish is a member of the most phyletically derived teleost order, the Perciformes, and like other Antarctic notothenioid species it has extremely high blood AFGP concentrations, has a high demand on lipid transport and metabolism, and possesses unusually large lipid muscle stores to maintain neutral buoyancy. In this study the authors report that the major acidic protein in D. mawsoni plasma binds palmitic acid with high affinity and that this protein is HDL and not albumin. 2. Materials and methods Freeze-dried plasma from the Antarctic toothfish, D. mawsoni, was reconstituted to its original volume with distilled water and was stored at −80°C until required. Agarose gel electrophoresis was performed in 1% agarose in 38 mM Tris, 46 mM Na-barbitone, 16 mM diethylbarbituric acid (pH 8.6) for 40 min [6], and 63Ni binding and autoradiography was performed as previously described [6,7]. Binding of [14C]palmitic acid and autoradiography was carried out at room temperature as in [36]. Reducing SDS-PAGE (sodium dodecyl sulfate-polyacrylamide electrophoresis) at pH 8.8 was performed in 12% polyacrylamide gels with a 3% stacking gel as V.J. Metcalf et al. / Comparati6e Biochemistry and Physiology, Part B 124 (1999) 147–155 per the method of Laemmli [31]. Samples were incubated in SDS reducing buffer; 4% SDS, 10% 2-mercaptoethanol, 125 mM Tris, 40% glycerol, 0.002% bromophenol blue at 100°C for 5 min before electrophoresis. Gels were stained with 0.1% Coomassie Brilliant Blue R. Two-dimensional electrophoresis was performed as in [36,37]. Plasma samples (2.6 ml) were electrophoresed as above on agarose gels, half of each lane was cut out and reduced with SDS reducing buffer for 15 min at room temperature. This gel strip was then placed on top of a 1.5 mm SDS-PAGE gel (12%) with 5 mm of stacking gel (3%) and electrophoresed and stained as above. Cellulose acetate electrophoresis was performed on plasma samples as described in [32] and [37]. Plates (76 × 90 mm) were wetted in 38 mM Tris, 46 mM Na-barbitone, 16 mM diethylbarbituric acid (pH 8.6) buffer before sample application. Electrophoresis was performed in this same buffer at 170 V for 15– 30 min. Plates were either stained in: (a) 0.1% amido black in 10% acetic acid in 50% methanol for 10 min, followed by destaining in 10% acetic acid in 50% methanol; or (b) 30 ml 1.4% (w/v) Oil Red O in methanol/7 ml 1 M NaOH for 2 h, followed by gentle rinsing and wiping using distilled water. Competitive palmitate-binding experiments were carried out at room temperature as follows. A mixture was made of Antarctic toothfish plasma (5 mg HDL protein) and 5 mg freeze-dried human albumin (purified in the laboratory by ion exchange chromatography). This was added to 1.25 ml of [14C]palmitic acid diluted to 5 mCi ml − 1, mixed, and then applied to an agarose gel, before electrophoresis and autoradiography as in [36]. Ultracentrifugation was performed using a variation [37] of the methods described for the isolation of rain- 149 bow trout lipoproteins [11,53]. Reconstituted Antarctic toothfish plasma (2 ml) and fresh human plasma (2 ml) were diluted to 5 ml with 1.015 g ml − 1 NaCl, and the density of the solution adjusted to 1.215 g ml − 1 with the addition of 1.6 g KBr. Samples (5 ml) were transferred to 6.5 ml ultracentrifuge tubes and spun in a Beckman TYP TFT 45.6 rotor in a Beckman L8-70M Ultracentrifuge at 37 000 rpm (175 000×g) for 17 h at 15°C. Remaining material was analysed for total protein (g l − 1), and cholesterol (mmol l − 1) on a Hitachi Boehringer Mannheim 717 Automatic Analyser. After ultracentrifugation tubes were cut into two fractions; a top fraction of 1.5 ml, and a bottom fraction of 3.5 ml. Each Antarctic toothfish fraction was re-diluted to 5 ml with 1.215 g ml − 1 KBr density solution and 100 ml of each was analysed as above for total protein and cholesterol. Fractions were then recentrifuged, and recut as before. The top fraction contained the purified lipoproteins, while the bottom fraction (infranatant) contained the plasma proteins. The second ultracentrifugation step was not performed on the human plasma control. Fractions were dialysed exhaustively at 4°C against 0.02% sodium-azide. Antarctic toothfish protein samples were transferred to ProBlott membrane directly from agarose gels, with N-terminal sequence analysis performed by the Protein Microchemistry Facility, University of Otago, using established methods [28]. 3. Results Reconstituted plasma from the Antarctic toothfish, D. mawsoni, was analysed using agarose gel electrophoresis (pH 8.6). Albumin is usually the predominant anionic plasma band under these conditions, as is the case with human plasma (Fig. 1A, lane 1). While plasma from D. mawsoni also had a major anionic band Fig. 1. Agarose gel electrophoresis (pH 8.6) showing (A) Coomassie-blue stained gel, (B) [14C]palmitic acid autoradiograph, and (C) 63Ni2 + autoradiograph of plasma samples and ultracentrifugation fractions. (A) and (B): lane 1, 3 ml of a 1 in 10 dilution of human plasma; lane 2, 4 ml human lipoprotein fraction; lane 3, 6 ml human infranatant fraction; lane 4, 1.3 ml (A) and 3 ml (B) D. mawsoni plasma; lane 5, 6 ml (A) and 9 ml (B) D. mawsoni lipoprotein fraction; lane 6, 10 ml (A) and 6 ml (B) D. mawsoni infranatant fraction. (C). Lane 1, 3 ml of a 1 in 10 dilution of human plasma; lane 2, 2 ml of human infranatant fraction; lane 3, 1.6 ml D. mawsoni plasma; lane 4, 6 ml D. mawsoni infranatant fraction. 150 V.J. Metcalf et al. / Comparati6e Biochemistry and Physiology, Part B 124 (1999) 147–155 (Fig. 1A, lane 4), this was considerably more acidic than human albumin, and other characterised albumins [36,44]. Unlike albumin, this acidic protein was diffuse, and showed variable mobility. On agarose gel electrophoresis lipoproteins appear as broad diffuse bands, suggesting that as for New Zealand eels [37] the major acidic protein in D. mawsoni plasma may be a lipoprotein. Amido black and Oil Red O staining following cellulose acetate electrophoreses (pH 8.6) confirmed that the major acidic toothfish protein was a lipoprotein (data not shown). The ability of a protein to bind palmitate has been used as a principal means of identifying albumin. The major acidic protein in D. mawsoni plasma bound [14C]palmitic acid with high affinity (Fig. 1B, lane 4), as did albumin in human plasma (Fig. 1B, lane 1). In each species palmitate bound only to a single plasma protein. Ultracentrifugation was employed to isolate lipoproteins from human and D. mawsoni plasma, and the resulting fractions were analysed for cholesterol and protein concentration. In both cases most (99%) of the cholesterol was found in the lipoprotein fraction, with the majority of the protein associated with the infranatant (data not shown). Agarose gel electrophoresis of the human plasma ultracentrifugation fractions showed that as expected, they contained both alpha (HDL) and beta (LDL/VLDL) lipoproteins (Fig. 1A, lane 2) which in the absence of albumin both bound [14C]palmitic acid (Fig. 1B, lane 2). The infranatant fraction contained the remainder of the plasma proteins, including albumin (Fig. 1A, lane 3), which was the only protein to bind palmitate in the infranatant (Fig. 1B, lane 3). The major band in the lipoprotein fraction of Antarctic toothfish plasma corresponded to the most acidic plasma protein (Fig. 1A, lane 5) and bound palmitate (Fig. 1B, lane 5), establishing it tentatively as HDL. Small amounts of protein more cationic than HDL were observed in the lipoprotein fraction, and may correspond to beta lipoproteins (Fig. 1A, lane 5). The infranatant fraction of Antarctic toothfish plasma shows diffuse palmitate binding spanning several plasma proteins (Fig. 1B, lane 6). When free [14C]palmitic acid was electrophoresed under identical conditions it migrated towards the anode in a very diffuse, erratic manner (data not shown) unlike the binding shown in Fig. 1B, lane 6. Binding in the infranatant may therefore be non-specific or low affinity binding to plasma proteins in the absence of a specific high affinity binding protein such as HDL or albumin. 63 Ni2 + binding is usually associated with mammalian albumins [45] (Fig. 1C, lane 1), but not fish albumins [36]. As expected the palmitate binding lipoprotein in Antarctic toothfish plasma failed to bind nickel. However, two more cathodal proteins bound 63Ni2 + ; one with high affinity(Fig. 1C, lane 3). Like human albumin Fig. 2. Two-dimensional electrophoresis of Antarctic toothfish plasma. First dimension, 1% agarose gel electrophoresis; second dimension, 12% SDS-PAGE. The first dimension agarose gel profile is shown aligned along the top, with the position of the acidic palmitate binding band (APBB) indicated. Left lane, molecular mass markers; right lane, 0.4 ml Antarctic toothfish plasma. (Fig. 1C, lanes 1 and 2) these nickel binding proteins segregated with the infranatant fraction in Antarctic toothfish (Fig. 1C, lane 4). Albumin typically has a mass of :65 kDa, and this characteristic Mr is often used to identify albumins. When two-dimensional electrophoresis was used to separate the components of Antarctic toothfish plasma the palmitate binding band was seen to be composed of at least three proteins, but as expected devoid of protein at the mass of albumin, 65 kDa (Fig. 2). The major protein component was at 30 kDa, with other minor proteins at :9, 11, and 42 kDa. This confirmed that HDL is the palmitate binding band in D. mawsoni plasma, as the major apolipoprotein in human HDL is apo-AI with a mass of 28 kDa [12]. The toothfish palmitate binding band was isolated from plasma by transfer to ProBlott from agarose gels, and subsequent N-terminal sequence analysis produced a major protein sequence of DAAQP(S)QEL(R)-, and a minor sequence of EVPSQ(L)(Q)HI(V)-. Comparison of the major sequence with apo-AI N-terminal sequences from a range of species including human and eels established its identity as apo-AI (Fig. 3A). The minor sequence exhibited homology to N-terminal apoAIV protein sequences from mammalian species (Fig. 3B). These sequences identify the palmitate binding protein in the toothfish as HDL. The N-terminal sequence of the major nickel binding band, ISLGLVNTshowed no homology to albumin from other species. In competitive binding experiments, [14C]palmitic acid was added to a mixture of toothfish plasma and V.J. Metcalf et al. / Comparati6e Biochemistry and Physiology, Part B 124 (1999) 147–155 151 Fig. 5. Twelve percent SDS-PAGE of ultracentrifuge fractions. Lane 1, molecular mass markers; lane 2, 0.4 ml Antarctic toothfish plasma; lane 3, 0.8 ml Antarctic toothfish lipoprotein fraction; lane 4, 6 ml human lipoprotein fraction. Fig. 3. Mature N-terminal sequences for Antarctic toothfish apo-AI (A) and apo-AIV (B) aligned with mature N-terminal sequences from other species [5,8,22,27,35,37,38,41– 43,47,52,55,58]. A dash indicates a position where a gap has been introduced in the sequence to maximise sequence identity between species; identical and chemically conserved residues are highlighted. purified human albumin, and the distribution of palmitate assessed by agarose gel electrophoresis and autoradiography. The presence of human albumin, which has a high affinity for palmitate, did not interfere with the binding of palmitate to HDL (Fig. 4, lanes 1 – 4). In- deed palmitate appeared to preferentially bind to the HDL. The apolipoprotein components of Antarctic toothfish HDL were examined by reducing SDS-PAGE (12%). The major apolipoprotein had a molecular mass of 30 kDa (Fig. 5, lane 3), and this protein is presumably apo-AI (28 kDa in humans (Fig. 5, lane 4) [12]). An 11 kDa apolipoprotein, presumably apo-AII, was present at an :1:1 ratio with apo-AI; in humans the apo-AII:apo-AI ratio is 1:20, while in eels it is : 1:5 [37]. A third apolipoprotein was present at :42 kDa, and this is probably apo-AIV, which has an Mr of 46 kDa in rat HDL [12]. The Antarctic toothfish lipoprotein fraction also contained a small amount of material at :55, 80, and \200 kDa. The latter material may correspond to apo-B, which is the major apolipoprotein of human LDL/VLDL and has an Mr of\ 250 kDa. 4. Discussion Fig. 4. Competitive binding of palmitate between Antarctic toothfish HDL and human albumin. Lanes 1 – 2, autoradiograph, 12 days film exposure; lanes 3 – 4, Coomassie stain. Lanes 1 and 3, 5 ml Antarctic toothfish plasma; lanes 2 and 4, 5 ml Antarctic toothfish plasma plus 0.75 ml human albumin at 5 mg ml − 1. The position of the human albumin band is indicated with an arrow. The major acidic plasma protein in Antarctic toothfish bound palmitate (Fig. 1A and B), and Oil Red O staining, ultracentrifugation, and cholesterol analysis showed that it was a lipoprotein. N-terminal sequencing established that two of the components of this band were the apolipoproteins apo-AI and apo-AIV, identifying it as HDL (Fig. 3). LDL and VLDL were at low or negligible levels in D. mawsoni plasma, which is similar to the situation in New Zealand eels, rainbow trout, and carp where HDL is also present at high levels, and LDL/VLDL are at very low or non-de- 152 V.J. Metcalf et al. / Comparati6e Biochemistry and Physiology, Part B 124 (1999) 147–155 tectable levels [11,30,37,53]. While a major nickel binding protein was present in Antarctic toothfish plasma, it was not albumin, as determined by its N-terminal protein sequence, and lack of palmitate binding. Albumin appears to be absent from the plasma of D. mawsoni, as is also the case with New Zealand eels and probably carp [1,37]. The high negative charge on toothfish HDL is unusual among vertebrate species. HDL runs just cathodal to albumin in human and other mammalian plasmas, and at a position equivalent to that of human albumin in New Zealand eels [37]. This indicates then that Antarctic toothfish HDL may differ considerably in its structure or composition as compared with HDL from other vertebrates. Alternatively, the high negative charge may be a spurious result of using reconstituted lyophilised plasma. The major apolipoprotein in D. mawsoni plasma was apo-AI (Fig. 5) with a molecular weight of 30 kDa. The other major apolipoprotein with a mass of 11 kDa is presumably apo-AII (Fig. 5), and a minor apolipoprotein with a mass of 42 kDa is likely to correspond to apo-AIV (Fig. 5). Both rainbow trout and eel HDL have similar apolipoprotein compositions [11,37,53]. The molecular masses of all three apolipoproteins in D. mawsoni HDL show an apparent post-ultracentrifugation increase (Fig. 5, lanes 2 and 3). This probably reflects the irreversible structural and compositional changes in HDL that are known to be induced by repeated ultracentrifugation [12]. In plasma the apoAII:apo-AI ratio is : 1:5, whilst in the lipoprotein fraction it is : 1:1 (Fig. 5, lanes 2 and 3). The relative amount of apo-AIV decreases following ultracentrifugation (Fig. 5, lanes 2 and 3). In eel HDL the apoAII:apo-AI ratio was 1:5 [37], while both a 1:5 and a 1:1 ratio have been observed following bioaffinity of carp plasma [1,2]. Loss of apo-AI is known to occur during ultracentrifugation of human HDL [23], but this does not explain the apo-AIV decrease, unless it has also been lost during ultracentrifugation. The major protein sequence from D. mawsoni HDL shows homology to apo-AI from other species, and in particular, other teleosts (Fig. 3). The Ala at residue 2, and Pro at residue 6 are unique to teleosts. The Antarctic toothfish is a more recently evolved teleost than both eels and salmonids. The Pro – Ala– Pro sequence insertion from residues 3 – 5 in the primitive eels is absent in the more recently evolved salmonids, although the Antarctic toothfish have an Ala at residue 4 and a Gln at residue 5, which could result from a point mutation of Pro. The Ser at residue 7 is conserved in all species except chicken, while the Gln at residue 8 appears to be restricted to lower vertebrates and mouse. A charged residue, Arg is found at residue 11 in D. mawsoni, and although dissimilar to the Asp or His found in other teleosts, is similar to the Arg or Lys found in higher vertebrates. It appears as if two point mutations have occurred at residues 9 (Glu) and 10 (Leu) in D. mawsoni, as the residues in other teleosts are Leu– Glu. The minor HDL protein sequence from Antarctic toothfish shows homology to mammalian apo-AIV sequences, with differences at 3 residues (3, Pro/Thr/Ser; 8, Gln/Ala; 9, His/Asn) all explainable by point mutations. This is the first apo-AIV sequence determined for a non-mammalian species. Although more apo-AII than apo-AIV is present in D. mawsoni HDL, apo-AII in several species including turkey, rat and human has a blocked N-terminus [12], probably explaining why a N-terminal apo-AII sequence was not detected in Antarctic toothfish HDL. Palmitate appears to bind preferentially to Antarctic toothfish HDL over added albumin. It appears that rather than just binding LCFAs in the absence of albumin, HDL in D. mawsoni has had its high affinity selected for during evolution. Like New Zealand eels, the situation in D. mawsoni is analogous to analbuminaemia in humans. People suffering from analbuminaemia have circulating levels of albumin of B 1 mg ml − 1 instead of the normal range of 35– 45 mg ml − 1. A greatly increased beta lipoprotein concentration in these individuals results in altered lipid and fatty acid metabolism [45], while Antarctic toothfish and eels possess low levels of beta lipoproteins but high HDL levels [37]. In these fish species and analbuminaemic humans, lipoprotein levels are increased, maintaining LCFA transport. It appears that albumin expression has been lost in at least three teleost species (Antarctic toothfish, two New Zealand eel species, and probably carp). Eels were among the first teleosts to evolve, followed by the orders containing salmonids and carps, with the Antarctic notothenioid suborder being the most recent teleost group. The holostean origin of teleosts is now thought to be polyphyletic, meaning the ancestral eel arose from holosteans independently, and separate from the origin of other teleost orders. Salmonids possess albumin indicating that loss of albumin and the adoption of HDL as the LCFA transport protein may have happened at least two and possibly three times during teleost evolution; in the eel line after divergence from the rest of teleosts, in the carp line, and possibly again in the notothenioid line. Lack of albumin expression in these teleosts may be the result of gene deletion, mutational inactivation, or permanent down-regulation. Alternatively if all teleosts are derived from ancestral eels, a more parsimonious explanation would be that loss of albumin expression occurred once in the ancestral eel. Thus, the absence of albumin in carp and notothenioids would be due to a phyletic effect, rather than a specific adaptation. Albumin expression in salmonids could then be explained by mutational reac- V.J. Metcalf et al. / Comparati6e Biochemistry and Physiology, Part B 124 (1999) 147–155 tivation of the gene or gene up-regulation sometime after the divergence of salmonids from other teleosts. It has been proposed that loss of albumin in eels and the adoption of HDL in fatty acid transport may be an adaptation to a predominantly freshwater life [37]. There must be different evolutionary reasons governing the loss of albumin and use of HDL as the fatty acid transporter in D. mawsoni when compared with eels. Loss of gene expression can occur in a short evolutionary period if environmental pressure is significant, and the notothenioid suborder shows examples of this. The AFGP family is thought to have evolved in situ in Antarctic notothenioids as the continent moved south and freezing of the ocean occurred. Notothenioid species which lack AFGPs are those of the earliest family, the Bovichtidae, as well as three species including D. eleginoides [20]. All these species have habitats outside the subzero waters, and in the latter three species the AFGP absence is a secondary more recent loss. The Antarctic notothenioids have decreased the number of haemoglobin forms in the last 15 my, again through gene deletion. One family, the icefishes, have lost haemoglobin entirely as a method of reducing blood viscosity to combat the subzero water temperatures [57]. The kinematic viscosity of seawater doubles as the temperature decreases from 20 to 2°C [34]. The loss of albumin in D. mawsoni may be a means of reducing blood viscosity, and thus will have occurred in this species in the last 5 – 15 my. Albumin is typically found in high concentrations in plasma, :40 mg ml − 1 in mammalian species and 10– 15 mg ml − 1 in teleosts [45]. Antarctic notothenioid blood with AFGPs at high plasma concentrations of 30– 35 mg ml − 1 in plasma will already be viscous without the addition of albumin. In D. mawsoni, the HDL is not present at a level comparable with that of albumin in other teleosts, thereby reducing viscosity. This may enable this and potentially also other notothenioid species to survive better at low temperatures. At subzero temperatures the metabolism of lipids and the transport of fatty acids are important processes in Antarctic fish. Large muscular lipid depositions are also essential in maintaining neutral buoyancy in D. mawsoni, and are possibly used as an energy store in winter [18,20,51]. Lipids are stored in the form of triacylglycerols [18], with the major fatty acids being long chain (palmitic and oleic acid). In addition, intracellular lipid droplets in the liver and muscle may serve as an oxygen store enhancing oxygen diffusion. As cold environmental temperatures affect metabolism, this oxygen store may enhance rates of ATP production and metabolite diffusion [20]. The use of HDL as the LCFA transporter of choice rather than albumin in D. mawsoni may reflect an enhanced ability to transport fatty acids at low temperatures, thus ensuring that neutral buoyancy, energy stores, and metabolism are all maintained. 153 It is known that several Antarctic fish enzymes are more efficient at low temperatures than their temperate counterparts [33]. The binding of LCFA to HDL is possibly more efficient at low temperatures than the binding of LCFA to albumin. Regardless of the reasons for this unusual adaptation, it is worthwhile investigating other notothenioid fish, as well as other teleosts to see whether use of HDL to transport LCFA and the loss of albumin are widespread among other species. In particular, the sister species to D. mawsoni, D. eleginoides, which was originally thought to have been cold-adapted but now lives outside the subzero water temperature zone, and has lost AFGP, warrants study. Acknowledgements The authors thank Dr Craig Marshall, Department of Biochemistry, University of Otago for donating freeze-dried Dissostichus mawsoni plasma, Dr Michael Hubbard and Diana Carne of the Protein Microchemistry Facility (University of Otago) for performing the N-terminal sequencing; Trevor Walmsley, Sharon Corsbie, and Marianne Hunt, of the Lipid Laboratory, Canterbury Health Laboratories, for help in performing the ultracentrifugation and Dr Bill Davison for helpful comments. Victoria Metcalf also wishes to acknowledge the following sponsors for her recent trip to Antarctica: Enderby Trust; Christchurch School of Medicine; Canterbury Medical Research Foundation; New Zealand Society of Biochemistry and Molecular Biology; Dr Stephen Brennan; Lena and Lawrie Metcalf; Christchurch Sunrise Rotary Club; David Gilbert; Annies Wine Bar and Restaurant; Waimari Lions Club; Christchurch Garden City Lions Club; Scott Base staff; Kathmandu Ltd; and Earth Sea and Sky Equipment. References [1] Amthauer R, Concha M, Villanueva J, Krauskopf M. Interaction of Cibacron Blue and anilinonaphthalenesulphonate with lipoproteins provides a new means for simple isolation of these plasma proteins. Biochem Biophys Res Comm 1988;154(2):752– 7. [2] Amthauer R, Villanueva J, Vera MI, Concha M, Krauskopf M. Characterization of the major plasma apolipoproteins of the high density lipoprotein in the carp (Cyprinus carpio). Comp Biochem Physiol 1989;92B:787– 93. [3] Andersen NC. Genera and subfamilies of the family Nototheniidae (Pisces, Perciformes) from the Antarctic and Subantarctic. Steenstrupia 1984;10(1):1 – 34. [4] Bargelloni L, Ritchie PA, Patarnello T, Battaglia B, Lambert DM, Meyer A. Molecular evolution at subzero temperatures: mitochondrial and nuclear phylogenies of fishes from Antarctica (Suborder Notothenioidei), and the evolution of antifreeze glycopeptides. Evol Mol Biol 1994;11(6):854– 63. [5] Boguski MS, Birkenmeier EH, Elshourbagy NA, Taylor JM, Gordon JI. Evolution of the lipoproteins. Structure of the rat 154 [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] V.J. Metcalf et al. / Comparati6e Biochemistry and Physiology, Part B 124 (1999) 147–155 apo-A-IV gene and its relationship to the human genes for apo-A-I, C-III, and E, J. Biol Chem 1986;261(14):6398– 407. Brennan SO, George PM, Peach RJ. Characterisation of a slow component of normal human serum albumin. Clin Chem Acta 1988;176:179 – 84. Brennan SO, Owen MC, Boswell DR, Lewis JH, Carell RW. Circulating proalbumin associated with a variant proteinase inhibitor. Biochim Biophys Acta 1984;802(1):24– 8. Byrnes L, Luo CC, Li WH, Yang CY, Chan L. Chicken apolipoprotein AI: cDNA sequence tissue expression and evolution. Biochem Biophys Res Comm 1987;148:485– 92. Byrnes L, Gannon F. Atlantic Salmon (Salmo salar) serum albumin: cDNA sequence, evolution, and tissue expression. DNA Cell Biol 1990;9:647 – 55. Campbell J, Martucci AD, Green GR. Plasma albumin as an acceptor of free fatty acids. Biochem J 1964;93:183 – 9. Chapman MJ, Goldstein S, Mills GL, Leger C. Distribution and characterization of the serum lipoproteins and their apoproteins in the rainbow trout (Salmo gairdnerii ). Biochemistry 1978;17 (21):4455– 64. Chapman MJ. Animal lipoproteins: chemistry, structure and comparative aspects. J Lipid Res 1980;21:789 – 853. Chen L, DeVries AL, Cheng C-HC. Evolution of antifreeze glycoprotein gene from a trypsinogen gene in Antarctic notothenioid fish. Proc Natl Acad Sci 1997;94:3811– 6. Davidson WS, Bartlett SE, Birt TP, Birt VL, Green JM. Palmitate-binding serum albumin-like proteins in salmonids. FEBS Lett 1988;233:299 – 302. DeVries AL. The role of antifreeze glycopeptides and peptides in the freezing avoidance of Antarctic fishes. Comp Biochem Phsyiol 1988;90B:611– 21. DeVries AL. Biological antifreeze agents in coldwater fishes. Comp Biochem Phsyiol 1982;73A:627 – 40. Eastman JT, DeVries AL. Hepatic ultrastructural specialization in Antarctic fishes. Cell Tissue Res 1981;219:489– 96. Eastman JT, DeVries AL. Buoyancy adaptations in a swim-bladderless Antarctic fish. J Morphol 1981;167:91 – 102. Eastman JT. Evolution and diversification of Antarctic notothenioid fishes. Am Zool 1991;31:93 – 109. Eastman JT. Antarctic Fish Biology: Evolution in a Unique Environment. San Diego: Academic Press, 1993:156– 219;276 – 313. Elliot DH. Physical geography- geographical evolution. In: Bonner WN, Walton DWH, editors. Key Environments: Antarctica. Oxford: Pergamon, 1985:39 – 61. Elshourbagy NA, Walker DW, Boguski MS, Gordon JI, Taylor JM. Structure and expression of the human apolipoprotein A-IV gene. J Biol Chem 1987;261(5):1998– 2002. Fainaru M, Glangeaud MC, Eisenberg S. Radioimmunoassay of human high density lipoprotein apoprotein A-I. Biochim Biophys Acta 1975;386:432 – 43. Fielding CJ, Shore VG, Fielding PE. A protein cofactor of lecithin: cholesterol acyltransferase. Biochem Biophys Res Commun 1972;46(4):1493– 8. Gon O, Heemstra PC, editors. Fishes of the Southern Ocean. Grahamstown, South Africa: J.L.B Smith Institute of Ichthyology, 1990:285 – 9. Hamoir G. Biochemical adaptation of the muscles of the Channichthyidae to their lack of haemoglobin and myoglobin. Comp Biochem Physiol 1988;65B:199– 206. Hixson JE, Kammerer CM, Mott GE, Britten ML, Birnbaum S, Powers PK, Vandeberg JL. Baboon apolipoprotein A-IV. Identification of Lys76 “ Glu that distinguishes two common isoforms and detection of length polymorphisms at the carboxyl terminus. J Biol Chem 1993;268(21):15667– 73. Hubbard MJ. Calbindin28kDa and calmodulin are hyperabundant in rat enamel dental cells. Identification of the protein [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] phosphatase calcineurin as a principal calmodulin target and of a secretion– related role for calbindin28kDa. Eur J Biochem 1995;230:68 – 79. Kennett JP. Cenzoic evolution of Antarctic glaciation, the circum-Antarctic Ocean and their impact on global paleoceanography. J Geophys Res 1977;82:3843 – 60. Krauskopf M, Amthauer R, Araya A, Concha M, Leon G, Rios L, Vera MI, Villanueva J. Temperature acclimatization of the carp. Cellular and molecular aspects of the compensatory response. Arch Biol Med Exp 1988;21:151 – 7. Laemmli UK. Cleavage of structural proteins during the assembly of the head of the bacteriophage T4. Nature 1970;227:680 – 5. Lehman H, Huntsman RG. Man’s Haemoglobin’s. Amsterdam: North Holland Publishing Company, 1966:272 – 3. Macdonald JA, Montgomery JC, Wells RMG. Comparative physiology of Antarctic fishes. In: Baxter JHS, Southward AJ, editors. Advances in Marine Biology, vol. 24. London: Academic Press, 1987:321 – 88. Macdonald JA, Wells RMG. Viscosity of body fluids from Antarctic notothenioid fish. In: di Prisco G, Maresca B, Tota B, editors. Biology of Antarctic Fish. Berlin: Springer-Verlag, 1991:163 – 78. Mahley RW, Weisgraber KH, Innerarity T, Brewer HB, Assmann G. Swine lipoproteins and atherosclerosis. Changes in the plasma lipoproteins and apoproteins induced by cholesterol feeding. Biochemistry 1975;14:2817 – 23. Metcalf V, Brennan S, Chambers G, George P. The albumins of chinook salmon (Oncorhynchus tshawytscha) and brown trout (Salmo trutta) appear to lack a propeptide. Arch Biochem Biophys 1998;350(2):239– 44. Metcalf VJ, Brennan SO, Chambers G, George PM. High Density Lipoprotein (HDL), and not albumin, is the major palmitate binding protein in New Zealand Long-finned (Anguilla dieffenbachii ) and Short-finned (Anguilla australis schmidtii ) eel plasma. Biochim Biophys Acta 1999;1429:467– 75. Nakai T, Whayne TF, Tang J. The amino- and carboxyl-terminal sequences of canine apolipoprotein A-I. FEBS Lett 1976;64:409 – 11. Neighbors MA, Nafpaktitis BG. Lipid compositions, water contents, swimbladder morphologies and buoyancies of nineteen species of midwater fishes (18 myctophids and 1 neoscopelid). Mar Biol 1982;66:207 – 15. Nevenzel JC. Occurrence function and biosynthesis of wax esters in marine organisms. Lipids 1970;5:308 – 19. O’Huigin C, Chan L, Li WH. Cloning and sequencing of bovine apolipoprotein A-I cDNA and molecular evolution of apolipoproteins A-I and B-100. Mol Biol Evol 1990;7(4):327– 39. Osada J, Pocovi M, Nicolosi RJ, Schaefer EJ, Ordovas JM. Nucleotide sequences of the Macaca fascicularis apolipoprotein C-III and A-IV genes. Biochim Biophys Acta 1993;1172(3):335– 9. Pan TC, Hao QL, Yamin TT, Dai PH, Chen BS, Chen SL, Kroon PA, Chao YS. Rabbit apolipoprotein A-I mRNA and gene. Evidence that rabbit apolipoprotein A-I is synthesised in the intestine but not in the liver. Eur J Biochem 1987;170:99 – 104. Peters T Jr, Davidson LK. Isolation and properties of a fatty acid-binding protein from the Pacific lamprey (Lampetra tridentata). Comp Biochem Physiol 1991;99B:619– 23. Peters T Jr. All About Albumin: Biochemistry, Genetics, and Medical Applications. San Diego: Academic Press, 1996:1 – 166; 234 – 43. Phleger CF, Grigor MR. Role of wax esters in determining buoyancy in Hoplostethus atlanticus (Beryciformes: Trachichthyidae). Mar Biol 1990;105:229 – 33. Powell R, Higgins DG, Wolff J, Byrnes L, Stack M, Sharp PM, Gannon F. The salmon gene encoding apolipoprotein A-I; V.J. Metcalf et al. / Comparati6e Biochemistry and Physiology, Part B 124 (1999) 147–155 [48] [49] [50] [51] [52] [53] cDNA sequence tissue expression and evolution. Gene 1991;104:155 – 61. Richieri GV, Anel A, Kleinfeld AM. Interactions of long-chain fatty acids and albumin: determination of free fatty acid levels using the fluorescent probe ADIFAB. Biochemistry 1993;32:7574 – 80. Ritchie PA, Lavoué S, Lecointre G. Molecular phylogenetics and the evolution of Antarctic Notothenioid fishes. Comp Biochem Physiol 1997;118A:1009– 25. Ruud JT. Vertebrates without erythrocytes and blood pigment. Nature 1954;173:848 – 50. Sargent JR, Henderson RJ, Toche DR. The lipids. In: Halver JE, editor. Fish Nutrition. San Diego: Academic Press, 1989:153 – 218. Sharpe CR, Sidoli A, Shelley CS, Lucero MA, Shoulders CC, Baralle FE. Human apolipoproteins AI, AII, CII and CIII. cDNA sequences and mRNA abundance. Nucl Acids Res 1984;12:3917 – 32. Skinner ER, Rogie A. The isolation and partial characterization [54] [55] [56] [57] [58] of serum lipoproteins of the rainbow trout. Biochem J 1978;173:507 – 20. Spector AA, Fletcher JE. Transport of fatty acid in the circulation. In: Dietschy JM, Gotto AM, Ontko JA, editors. Disturbances in Lipid and Lipoprotein Metabolism. Baltimore: Waverley Press, 1978:229 – 50. Stoffel W, Mueller R, Binczek E, Hoffmann K. Mouse apolipoprotein A-IV. cDNA-derived primary structure, gene organisation and complete nucleotide sequence. Biol Chem Hoppe-Seyler 1992;373(4):187– 93. Tove SB. The esterolytic activity of serum albumin. Biochim Biophys Acta 1962;57:230 – 5. Wells RMG, Ashby MD, Duncan SJ, MacDonald JA. Comparative study of the erythrocytes and haemoglobins in nothotheniid fishes from Antarctica. J Fish Biol 1980;17:517 – 27. Williams SC, Bruckheimer SM, Lusis AJ, LeBoeuf RC, Kinniburgh AJ. Mouse apolipoprotein A-IV gene: nucleotide sequence and induction by a high-lipid diet. Mol Cell Biol 1986;6(11):3807– 14. . . 155