- Email: [email protected]
l -Glutamate dehydrogenase from the Antarctic fish Chaenocephalus aceratus1
Biochimica et Biophysica Acta 1543 (2000) 11^23 www.elsevier.com/locate/bba
L-Glutamate
dehydrogenase from the Antarctic ¢sh Chaenocephalus aceratus1 Primary structure, function and thermodynamic characterisation: relationship with cold adaptation M. Antonietta Ciardiello, Laura Camardella, Vito Carratore, Guido di Prisco * Institute of Protein Biochemistry and Enzymology, C.N.R., Via Marconi 10, I-80125 Naples, Italy Received 12 May 2000; received in revised form 18 July 2000; accepted 25 July 2000
Abstract In order to study the molecular mechanisms of enzyme cold adaptation, direct amino acid sequence, catalytic features, thermal stability and thermodynamics of the reaction and of heat inactivation of L-glutamate dehydrogenase (GDH) from the liver of the Antarctic fish Chaenocephalus aceratus (suborder Notothenioidei, family Channichthyidae) were investigated. The enzyme shows dual coenzyme specificity, is inhibited by GTP and the forward reaction is activated by ADP and ATP. The complete primary structure of C. aceratus GDH has been established; it is the first amino acid sequence of a fish GDH to be described. In comparison with homologous mesophilic enzymes, the amino acid substitutions suggest a less compact molecular structure with a reduced number of salt bridges. Functional characterisation indicates efficient compensation of Q10 , achieved by increased kcat and modulation of S0:5 , which produce a catalytic efficiency at low temperature very similar to that of bovine GDH at its physiological temperature. The structural and functional characteristics are indicative of a high extent of protein flexibility. This property seems to find correspondence in the heat inactivation of Antarctic and bovine enzymes, which are inactivated at very similar temperature, but with different thermodynamics. ß 2000 Elsevier Science B.V. All rights reserved. Keywords: Glutamate dehydrogenase; Cold adaptation; Amino acid sequence; Kinetic analysis; Thermodynamic analysis; Antarctic ¢sh
1. Introduction Environmental factors, such as temperature, may have profound e¡ects on enzyme activity, regulation
* Corresponding author. Fax: +39-81-593-6689; E-mail: [email protected] 1 The amino acid sequence of C. aceratus GDH has been submitted to the SWISS-PROT Data Bank with accession number P82264.
and structure. At decreasing temperatures the rates of enzyme-catalysed reactions are reduced by the low heat content of the cellular environment (Q10 e¡ect), which a¡ects the higher orders of the protein structure and the interactions with low-molecular-mass ligands. Therefore, strategies of biochemical adaptation allowing low-body-temperature species to obtain an adequate metabolic £ux, o¡setting the Q10 e¡ect, are expected. The study of organisms living under extreme conditions may help to understand temperature adaptation. A particularly good experimental system for
0167-4838 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 3 8 ( 0 0 ) 0 0 1 8 6 - 2
BBAPRO 36261 8-11-00
12
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
investigating adaptation to extremely cold temperatures are ¢shes of the teleost suborder Notothenioidei [1]. Most members of this suborder are endemic to the Southern Ocean, and in this habitat they encounter extremely low and stable temperatures, namely, temperatures near the freezing point of seawater, 31.9³C. Because of the constant temperature of the Antarctic sea, marine organisms developed a high degree of stenothermality in parallel with cold adaptation. In general, Antarctic ¢sh do not survive at temperatures only few degrees higher than that of the habitat [2^4]. Furthermore, it was reported [3,5,6] that the kinetic properties of enzymes from stenothermal species are conserved in a narrow temperature range whereas those of enzymes from eurythermal species show conserved properties over wider temperature range, indicating that the degree of stenothermy or eurythermy of the whole organism can be mirrored at the biochemical level. Comparative studies of structure^function relationships in enzymes from Antarctic and mesophilic organisms may provide useful information on the molecular basis sustaining cold adaptation. Although fragmentary studies on enzymatic cold adaptation in Antarctic ¢sh have appeared in the literature, extensive characterisation of puri¢ed enzymes has been undertaken only during the recent years. Structural and functional characterisation has been reported for several proteases (trypsin, elastase and cathepsin D) [7^9], an ATPase (dynein) involved in the function of contractile/locomotory proteins [10] and two enzymes having a key role in metabolism (lactate dehydrogenase and glucose-6-phosphate dehydrogenase) [11,12]. In vertebrate cells, L-glutamate dehydrogenase (GDH; EC 1.4.1.3), primarily associated to the mitochondrial matrix, has a key regulatory function in cellular metabolism in controlling levels of ammonia and glutamate. It catalyses the reversible oxidative deamination of L-glutamate to K-ketoglutarate and ammonia through reduction of NAD or NADP . In order to investigate the structure^function relationships in enzymes from cold-adapted ¢sh, we have puri¢ed and characterised GDH from the liver of the ice¢sh Chaenocephalus aceratus (belonging to the haemoglobinless family Channichthyidae). When appropriate, the kinetics of C. aceratus GDH were
analysed by carrying out parallel experiments on bovine GDH, in order to overcome reported di¡erences in kinetic properties ascribed to the reaction condition used [13^17]. The observation that GDH is one of the slowest evolving proteins for which extensive data on functional properties and amino acid sequences are available prompted its choice to study enzyme evolutionary adaptation to temperature. A rate of 1.8 point mutations per 100 residues per 108 years was estimated for this protein [18]. This feature is highly valuable for the study of molecular cold adaptation, since functional properties acquired during evolution can be correlated with relatively few amino acid substitutions. 2. Experimental procedures Specimens of C. aceratus were collected in the vicinity of Palmer Station, Antarctica (63³25P S, 62³15P W), where the seawater temperature is 31.8³C. Fish were transported to Palmer Station where livers were harvested and frozen before further treatment. LGlutamate, soybean trypsin inhibitor and cyanogen bromide were from Sigma; K-ketoglutarate, NAD , NADP , NADH, NADPH, GTP, ADP, ATP, phenylmethylsulfonyl£uoride, proteases (Lys-C, Asp-N, Arg-C) and benzamidine were from Boehringer. All other reagents were of the highest purity commercially available. Standard activity assays were performed at 20³C by recording the reduction of NAD (forward reaction) or the oxidation of NADH (reverse reaction), at 340 nm in a Varian DMS 300 spectrophotometer. The reaction mixture for the forward reaction contained 100 mM potassium phosphate bu¡er (pH 7.5), 40 mM L-glutamate, and 2 mM NAD . The reaction was initiated by addition of enzyme (1^3 Wg) in a ¢nal volume of 1 ml. The mixture for the reverse reaction contained 100 mM potassium phosphate bu¡er (pH 7.5), 100 mM ammonium sulfate, 5 mM K-ketoglutarate, 100 WM NADH and 0.2^0.4 Wg of enzyme in a ¢nal volume of 1 ml. One activity unit is de¢ned as the amount of enzyme catalysing the reaction of 1 Wmol of substrate/min under the standard assay conditions.
BBAPRO 36261 8-11-00
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
Kinetic constants were determined in 100 mM potassium phosphate bu¡er (pH 7.5), using 0.01^40 mM glutamate, 0.01^5 mM NAD . In the temperature-dependence experiments, the bu¡er was equilibrated at the assay temperature and pH was adjusted to the desired value. The assay mixture was equilibrated at the required temperature before adding the enzyme. 2.1. Puri¢cation All steps were carried out at 0^4³C. 2.1.1. Extraction Frozen livers were homogenised with 8 volumes of 50 mM potassium phosphate bu¡er (pH 7.5), 0.5 mM EDTA, 50 Wg/ml phenylmethylsulfonyl£uoride, 20 Wg/ml benzamidine and 10 Wg/ml soybean trypsin inhibitor. The mixture was centrifuged at 37 000Ug for 45 min and the supernatant was recovered. 2.1.2. Fractionation The supernatant was brought to 40% ammonium sulfate saturation and centrifuged at 37 000Ug for 30 min. The pellet was discarded; the supernatant was brought to 50% saturation and the precipitate was collected. 2.1.3. Heat treatment The precipitate was redissolved in a minimal volume of 50 mM potassium phosphate bu¡er (pH 7.5), containing 100 mM ammonium sulfate, kept at 52³C for 2 min and rapidly cooled. The mixture was diluted with an equal volume of bu¡er and centrifuged for 60 min at 37 000Ug. The cloudy supernatant was brought to 30% ammonium sulfate saturation, left at 4³C overnight, and centrifuged for 60 min at 37 000Ug. Protein in the clear supernatant was precipitated at 60% ammonium sulfate saturation and GDH activity was recovered in the pellet. 2.1.4. Chromatographic separations GDH was further puri¢ed by FPLC (Pharmacia, Uppsala, Sweden) at 4³C, monitoring the e¥uent at 280 nm. The solution, dialysed against bu¡er A (5 mM potassium phosphate, pH 8.0), was loaded on a Mono Q HR5/5 or an HR10/10 column equilibrated in 100% bu¡er A, at a £ow rate of 1 ml/min
13
or 3 ml/min, respectively. Elution was carried out by increasing the concentration of bu¡er B (200 mM potassium phosphate, pH 8.0) from 0% to 50% in 60 min. The active fractions were pooled and protein was precipitated at 60% ammonium sulfate saturation. The precipitate was dissolved in a minimal volume of 50 mM potassium phosphate bu¡er (pH 7.5), and 200 Wl was loaded on a gel ¢ltration column Superose 6 HR 10/30, equilibrated with 50 mM phosphate bu¡er (pH 7.5) containing 150 mM NaCl; the £ow rate was 0.4 ml/min. The active fractions were pooled and protein was precipitated at 60% ammonium sulfate saturation. Following centrifugation, the enzyme was dissolved in 100 mM potassium phosphate (pH 7.5)/50% glycerol and stored at 320³C. 2.2. Sodium dodecyl sulfate^polyacrylamide gel electrophoresis (SDS^PAGE) SDS^PAGE was carried out in the Bio-Rad Mini Protean apparatus, using 7% polyacrylamide gel. Staining and rinsing were performed in 0.02% Coomassie R-250 Brilliant blue in 10% v/v acetic acid/ 25% v/v isopropanol, and in 10% v/v acetic acid/25% v/v methanol, respectively. 2.3. Isoelectrofocusing The isoelectric point of puri¢ed GDH was measured in polyacrylamide gel in the pH range 3^9, at 10³C, in a Phast-System apparatus (Pharmacia). 2.4. Molecular mass The molecular mass of native GDH was determined by gel ¢ltration on Superose 6 HR 10/30 equilibrated in 0.1 M Tris^HCl (pH 7.5), 0.2 M NaCl and calibrated with thyroglobulin (669 kDa), ferritin (450 kDa), bovine GDH (333 kDa), catalase (240 kDa), and aldolase (158 kDa). The mass was calculated by plotting log molecular mass versus (Ve /V0 ). 2.5. Amino acid sequencing Denaturation and alkylation of the sulfhydryl groups with 4-vinylpyridine was carried out as de-
BBAPRO 36261 8-11-00
14
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
scribed [19]. Five 1-mg aliquots were subjected to chemical and enzymatic cleavages. An aliquot was dissolved in 70% formic acid (0.2 ml) and incubated with CNBr (5 mg) for 20 h at 20³C in the dark; the second was dissolved in 70% HCOOH (0.5 ml) and incubated for 20 h at 42³C. The peptide mixtures were evaporated under nitrogen and diluted with water to remove the acid, avoiding complete drying. The third aliquot was resuspended in 25 mM Tris^ HCl, pH 8.5 (0.5 ml), 1 mM EDTA and incubated for 20 h at 37³C, after addition of Lys-C (15 Wg) in 50 mM Tricine, pH 8.0 (0.15 ml), 10 mM EDTA; the fourth was incubated for 18 h at 37³C in 0.5 ml of a mixture containing 90 mM Tris^HCl (pH 7.6), 8.5 mM CaCl2 , 5 mM DTT, 0.5 mM EDTA and Arg-C (10 Wg); the last was dissolved in 0.06 ml 60% acetonitrile and incubated at 37³C for 6 h after addition of 200 mM sodium phosphate bu¡er, pH 8.0 (0.12 ml) and Asp-N (16 Wg) in 10 mM Tris^ HCl, pH 7.5 (0.4 ml). These three peptide mixtures were processed without drying. Peptides were separated by reverse-phase HPLC on a Vydac C4 column (4.6U250 mm, 5 W), using a Gold Apparatus (Beckman Instruments, Camino Ramon, CA). Elution was accomplished by a multistep linear gradient of eluent B (0.08% tri£uoracetic acid in acetonitrile) in eluent A (0.1% tri£uoracetic acid) at a £ow rate of 1 ml/min. The eluate was monitored at 220 and 280 nm. The separated fragments were manually collected and sequenced. Co-eluting peptides were further separated by reverse-phase HPLC on a WRPC C2/C18 column (2.1 mmU100 mm, 3 W) using a SMART System (Pharmacia). Elution was performed using a linear gradient of the solvents indicated above, at a £ow rate of 0.2 ml/min and monitored at 214 and 280 nm. Automated repetitive Edman degradation was performed on a pulsed liquid-phase Sequencer model 477A, equipped with a 120A analyser for the online detection of phenylthiohydantoin amino acids or on a Procise Protein Sequencer model 492 (Applied Biosystems, Division of Perkin^Elmer, Foster City, CA). 2.6. Allosteric e¡ectors ADP, ATP and GTP were dissolved in phosphate bu¡er and the pH was adjusted to 7.5. Activation
and inhibition under standard assay conditions are expressed as: % activation
V a 3V 0 U100; V0
and % inhibition
V 0 3V i U100 V0
where V0 is the reaction rate in the absence of nucleotides, Va and Vi are the reaction rates in the presence of activator and inhibitor, respectively. 2.7. Thermostability and thermodynamic parameters of heat-inactivation GDH (0.02 mg/ml) was incubated in 0.1 M potassium phosphate bu¡er (pH 7.5) at di¡erent temperatures; at time intervals the residual activity was measured at 20³C under standard assay conditions. The energy of activation of the heat inactivation process (Ea ) was calculated from the slopes of the Arrhenius plots of the rate constants (which equal 3Ea /R), the other thermodynamic parameters according to the following equations [20]:
v H i E a 3RT; v G i 3RTln hki =TkB where h is the Planck's constant, kB the Boltzmann constant and ki the rate constant of inactivation at T;
v S i v H i 3v Gi =T 2.8. Computational procedures Kinetic and thermodynamic parameters and standard errors of the mean were calculated using the computer program Enz¢tter (Elsevier-Biosoft) based on non-linear regression data analysis. 3. Results 3.1. Puri¢cation, pI, molecular mass GDH was puri¢ed from the frozen liver of C.
BBAPRO 36261 8-11-00
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
15
Fig. 1. Amino acid sequence of C. aceratus GDH. Arrows indicate peptide fragments obtained by acidic (Ac) or cyanogen bromide (CB) treatment or by enzymatic digestions with Lys-C (K), Asp-N (D), Arg-C (R). Peptides from each digestion are numbered according to their order in the sequence.
aceratus following the procedure described in the Experimental Procedures section. The recovery of electrophoretically pure GDH, with a speci¢c activity of 35 U/mg, was approximately 0.5 mg/10 g of tissue. Isoelectrofocusing showed a single band with pI = 5.3. The molecular mass of native GDH was estimated by FPLC gel ¢ltration on a calibrated Superose-6
column, from which it co-eluted with bovine GDH (333.4 kDa). SDS^PAGE showed a single band. 3.2. Primary structure Direct automated Edman-degradation of the native protein yielded a single sequence and 30 amino acid residues were identi¢ed from the N-terminus.
BBAPRO 36261 8-11-00
16
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
Fig. 2. Multiple alignment of the amino acid sequence of C. aceratus GDH with the vertebrate homologous enzymes known so far. The sequences were aligned by using the PileUp program (GCG-Wisconsin). The rat, mouse, human, bovine and chicken sequences were obtained from the SWISS-PROT Data Bank, using the accession nos. P10860, P26443, P00367, P00366, P00368, respectively; Z in the chicken sequence indicates ambiguous identi¢cation of Gln/Glu. The residues common to all sequences are underlined and those substituted only in the C. aceratus enzyme are marked by crosses. Asterisks indicate the residues of the Gly pattern typical of the coenzyme fold.
The sequence obtained by aligning the peptides from ¢ve chemical and enzymatic digestions of the S-pyridylethylated protein is shown in Fig. 1. Only peptides necessary to elucidate the complete sequence are indicated. Peptides were aligned by homology with vertebrate GDH sequences and by overlapping fragments. The deduced amino acid sequence contains 504 residues and the calculated molecular mass is 55 401 Da, in good agreement with results from SDS^PAGE. The amino acid composition deduced from the se-
quence of C. aceratus GDH was compared with those of enzymes from ox and chicken [21,22] (the only two vertebrate amino acid sequences obtained by direct protein sequencing, and therefore corresponding to the mature form of the protein). Antarctic GDH shows small di¡erences, such as higher amounts of Ala and Asn and lower amounts of Ser and Leu. The overall ratio Arg/(Arg+Lys), reportedly increased in thermophilic enzymes and decreased in psychrophilic enzymes [23^25], as well as the Gly and Pro contents, reportedly changed in
BBAPRO 36261 8-11-00
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
17
Table 1 Sequence identity (%) of GDHs from di¡erent sources Species
E. coli
S. cerev.
C. symb. S. solf.
P. fur.
H. salin.
Chicken
Ox
Rat
Human
Mouse
C. aceratus Mouse Human Rat Ox Chicken H. salinarium P. furiosus S. solfataricus C. symbiosum S. cerevisiae
28 28 29 26 26 26 29 32 32 53 55
28 27 27 27 25 25 32 32 31 50
30 31 31 31 30 30 33 36 32
39 42 42 42 40 39 47
42 40 40 39 38 38
83 91 91 90 94
84 95 95 94
88 99 97
88 98
88
38 36 36 38 36 35 42 44
Amino acid sequences of mouse [28], human [29], rat [27], bovine [21], chicken [22], Halobacterium salinarium [31], Pyrococcus furiosus [32], Sulfolobus solfataricus [33], Clostridium symbiosum [34], Saccharomyces cerevisiae [35] and Escherichia coli [36] GDHs, accession nos. P26443, P00367, P10860, P00366, P00368, P29051, P80319, P80053, P24295, P07262, P00370, respectively, were taken from the SWISS-PROT Data Bank. Percent identity was calculated with the Best¢t program (GCG-Wisconsin).
cold-adapted enzymes [26], are unchanged in Antarctic GDH. The amino acid sequence of C. aceratus GDH aligned with the vertebrate enzymes reported so far is shown in Fig. 2. The N-terminal region of Antarctic GDH is three residues longer than that of the bovine enzyme, but has the same length as chicken GDH; the comparison with the rat, mouse and human N-terminal region is not possible since only the nucleotide sequences encoding for the enzyme precursors are available for these organisms [27^29]. It is known, in fact, that GDH is synthesised as a larger precursor, as well as other intracellular translocatable proteins, and then processed to the mature form [30]. High conservation is observed among the sequences. Forty-two positions, conserved in all other vertebrates, are substituted in the ¢sh sequence. These substitutions produce an increase of Ala (Arg, Gly, Ser, AspCAla), Asn (His, Arg, Asp, SerCAsn) and Gln (Gly, HisCGln), and a decrease of charged (Asp, Glu, Lys, Arg, His) and large hydrophobic (Leu, Trp) residues. In these substitutions the residue is replaced by Ala seven times and by Asn four times. Most of these substitutions occur in clusters. Gln87 and His88 of Antarctic GDH are conserved in the same positions in the rat, mouse and human sequences, but their positions are inverted in ox and chicken. The same inversions occur in His224, Gly225 and Asn231, Glu232; Gly203, Lys204 and Pro205 correspond to Lys203, Pro204
and Gly205 in ox and chicken. This peculiarity accounts for approximately 45% of the di¡erence between C. aceratus GDH and mouse, rat and human GDHs, on one hand, and chicken and ox GDHs, on the other. Table 1 shows the sequence identity among GDHs of C. aceratus, vertebrates and representative bacteria, archaebacteria and yeast. The sequence identity with the other vertebrates (especially rat, human and
Fig. 3. Activity as a function of pH. Experiments were carried out in 100 mM potassium phosphate bu¡er, at 5³C (F) and 20³C (R) with C. aceratus and at 20³C (b) with bovine GDH.
BBAPRO 36261 8-11-00
18
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
mouse) is very high; the identity with GDHs from archaebacteria, bacteria and yeast gradually decreases. 3.3. E¡ect of pH Maximal activity of the forward reaction catalysed by C. aceratus GDH was reached at pH 8.0 and remained almost constant up to pH 8.9, both at 5³C and 20³C (Fig. 3). Continuously increasing activity was observed with bovine GDH at 20³C. The kinetic analysis described below was performed at pH 7.5. 3.4. Coenzyme speci¢city C. aceratus GDH can utilise both NAD(H) and NADP(H). At 20³C the activity with NADP(H) was approximately 10% of that with NAD(H) (the kcat values were 527, 54, 8500, 835/min with NAD , NADP , NADH, NADPH, respectively). A markedly lower a¤nity for the phosphorylated coenzyme was also observed (the S0:5 values of NAD , NADP , NADH, NADPH were 0.24, 1.6, 0.012, 0.12 mM, respectively). 3.5. E¡ectors The in£uence of ADP, ATP and GTP on the catalytic activity of C. aceratus GDH was investigated at increasing nucleotide concentrations. The forward and the reverse reactions were 90% inhibited by 80 and 200 WM GTP, respectively. 0.1 mM ADP and 2 mM ATP did not a¡ect the reverse reaction but produced a twofold activation of the forward reaction. 3.6. Kinetic properties The kinetics of the forward reaction at varying substrate and coenzyme concentration showed nonlinear double-reciprocal plots, as already reported for bovine GDH [37,38]; data were better analysed using the Hill equation that showed marked negative cooperativity. The kinetics as a function of NAD concentration were carried out with 40 mM glutamate since at 5^40³C this concentration saturated Antarctic and bovine GDH. The a¤nity for coenzyme of bovine
Fig. 4. Activity as a function of temperature. The rates of the reaction catalysed by C. aceratus GDH were obtained under standard assay conditions (R) and at coenzyme concentrations yielding saturation at each tested temperature (O). The rates of bovine GDH (b) were obtained under standard assay conditions, in which saturation is reached at each temperature.
GDH increased with temperature, and was highest at 35³C. In contrast, the a¤nity of the Antarctic enzyme was similar at 5 and 20³C, decreased at higher temperatures and was ten times lower at 40³C (Table 2). In bovine GDH 5 mM NAD was saturating between 5³C and 35³C. In C. aceratus GDH the NAD concentration necessary to reach saturation was 2 mM between 5³C and 20³C, 6 mM at 30³C and 15 mM at 40³C. At higher concentrations, an inhibitory e¡ect of NAD was observed. The saturation curves as a function of glutamate showed a Hill coe¤cient constantly below 1 for both Antarctic and bovine GDH. In bovine GDH, S0:5 values were similar at 20 and 35³C, but lower at 5³C, whereas in C. aceratus GDH the values were similar at all tested temperatures. In Antarctic GDH, the activity as a function of glutamate concentration was also investigated at 2 mM NAD. A decrease of S0:5 with increasing temperature (100 times lower at 40³C) and an increase of the Hill coe¤cient ( s 1 at 40³C) (Table 2, values in parenthe-
BBAPRO 36261 8-11-00
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
19
Table 2 Temperature e¡ect on the kinetic parameters of the forward reaction catalysed by C. aceratus and bovine GDHs Source
Assay S0:5 glu temperature (³C) (mM)
nH
S0:5 NAD (mM)
nH
kcat (/min)
kcat /S0:5 glu kcat /S0:5 NAD (/min/Wmol) (/min/Wmol)
C. aceratus
5 12 20 30 40
1.90 þ 0.17 2.10 þ 0.09 1.30 þ 0.13 1.10 þ 0.17 (0.62 þ 0.05) 2.00 þ 0.35 (0.02 þ 0.007)
0.60 0.70 0.66 0.70 (0.91) 0.50 (1.40)
0.22 þ 0.04 n.d. 0.24 þ 0.05 0.78 þ 0.20 2.60 þ 0.66
0.46 n.d. 0.54 0.68 0.55
228 þ 3 346 þ 5 527 þ 13 855 þ 66 1290 þ 127
120 165 405 777 645
1036 n.d. 2196 849 129
Ox
5 20 35
0.87 þ 0.1 5.00 þ 1.0 4.6 þ 0.07
0.75 0.41 0.54
1.70 þ 0.03 0.70 þ 0.1 0.54 þ 0.1
0.52 0.62 0.75
56 þ 1 245 þ 8 761 þ 28
64 49 165
33 70 1014
Values in parentheses were obtained with 2 mM NAD at 30³C and 40³C. n.d., not determined.
ses) was observed. These results indicated that: (i), in Antarctic GDH, depending on the temperature, the concentration of NAD has a profound e¡ect on the a¤nity for L-glutamate and on kinetics (positive or negative co-operativity) of substrate binding; (ii) a variation of the properties of Antarctic GDH, not shown by bovine GDH, occurs above 20³C. kcat values were calculated from the mean of the Vmax obtained from the saturation curves as a function of the concentration of substrate and coenzyme. At low temperature the rates measured with the Antarctic enzyme were higher than those of bovine GDH. kcat /S0:5 for NAD showed that the catalytic e¤ciency of C. aceratus GDH decreased at increasing temperature; the opposite trend was observed with the bovine enzyme. At 5³C, the catalytic e¤ciency for NAD of Antarctic GDH was 30-fold that of the bovine enzyme. 3.7. Temperature dependence of enzymatic activity The activity of C. aceratus and bovine GDH was investigated as a function of temperature under standard assay conditions. In the range 5^25³C, the Ant-
arctic enzyme catalysed the forward reaction with higher velocity than that of bovine GDH (Fig. 4) and had an apparent temperature optimum at 30³C. Above this temperature the activity decreased, although GDH was not irreversibly inactivated, as shown by thermostability experiments (see below). Bovine GDH showed maximal activity at 50³C and irreversible inactivation occurred above this temperature. Although the velocity of bovine GDH corresponds to kcat in the whole temperature range examined, whereas that of the Antarctic enzyme corresponds to kcat only in the 5^20³C range (as stated before, above 20³C substrate saturation is not achieved), Fig. 4 indicates nonetheless that the activity of Antarctic GDH decreases at lower temperatures than the bovine enzyme. As shown in many other enzymes from polar organisms, this feature is likely to be linked to cold adaptation. The e¡ect of temperature on kcat , measured between 5³C and 40³C, was analysed using the Arrhenius equation (Fig. 5). The activation energy Ea of Antarctic GDH (8540 cal/mol), calculated from the slope of the Arrhenius plot, was lower than that of the bovine enzyme (14 662 cal/mol). The thermody-
Table 3 Thermodynamic activation parameters of the forward reaction Source
Temperature (³C)
C. aceratus
5 30
Ox
5 35
vH* (cal/mol)
vS* (e.u.)
vG* (cal/mol)
7988 7938
327.01 327.08
15496 16143
14110 14060
37.81 37.12
16281 16217
BBAPRO 36261 8-11-00
20
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
Table 4 Kinetic constants and thermodynamic parameters of the heat inactivation process Source
Temperature (³C)
t1=2 (min)
ki (/min)
C. aceratus
48 50 52 45 50 52 55
63.0 þ 2.2 13.6 þ 1.0 3.0 þ 0.1 182.5 þ 3.1 49.3 þ 1.5 16.0 þ 0.5 3.0 þ 0.1
0.011 þ 0.001 0.051 þ 0.002 0.230 þ 0.007 0.004 þ 0.001 0.014 þ 0.001 0.043 þ 0.006 0.230 þ 0.007
ox
namic parameters of activation of the forward reaction vH*, vS* and vG* were calculated [12] at 5³C, 30³C and 35³C, and are reported in Table 3. 3.8. Thermostability C. aceratus and bovine GDHs fully retained activity after 30-min incubation at 45³C; above this temperature irreversible heat inactivation occurred (Fig. 6). The inactivation of Antarctic GDH in the range 45^52³C is much faster. The kinetics of inactivation of both enzymes displayed exponentially decreasing residual activity. The ¢rst-order constants of inactivation (ki ) and the half-inactivation times (t1=2 ) are shown in Table 4. The inactivation constants were plotted using the Arrhenius equation (inset of Fig. 6) and the enthalpy variations of activation of the heat inactivation process (v H i ) were calculated from the slopes. The thermodynamic parameters v H i , v G i and v S i accompanying the process of heat-inactivation of both Antarctic and the bovine
Fig. 5. E¡ect of temperature on kcat of C. aceratus (R) and bovine (b) GDH (Arrhenius plot).
vHi * (cal/mol)
vGi * (cal/mol)
vSi * (cal/mol/K)
178000
19021
490
135000
19633
355
GDHs, at 52³C, are also shown in Table 4. The enthalpy and entropy variations of heat inactivation of Antarctic GDH were higher than those of the bovine enzyme. 4. Discussion This paper reports the ¢rst primary structure of a ¢sh GDH, extending the knowledge on vertebrate GDH amino acid sequences. Alignment with homologous enzymes allows inclusion of this protein in Family II of the hexameric class of GDHs [39]. As
Fig. 6. Time-course of heat inactivation. C. aceratus (continuous lines) and bovine (dashed lines) GDHs at 45³C (b), 50³C (F) and 52³C (R). Inset: temperature e¡ect on ¢rst-order rate constants of inactivation of C. aceratus (continuous line) and bovine (dashed line) GDH.
BBAPRO 36261 8-11-00
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
expected from an enzyme having a very slow rate of evolutionary change, alignment with GDHs of other vertebrates shows high identity, ranging between 83% and 88%. As expected, the 35 invariant residues in 21 sequences of hexameric GDH [39], which include essential Lys126 (bovine GDH numbering), are conserved in Antarctic GDH. Furthermore, Lys445, involved in GTP binding [40] and Gly274 and Glu275, involved in the binding of the adenine moiety of NAD [41], are conserved. Only 8% of the residues conserved in all other vertebrate sequences are substituted in C. aceratus GDH. Some of these substitutions occur in clusters and most of them are in the region interconnecting the elements of secondary structure of bovine GDH [42]. These substitutions produce an overall increase of small hydrophobic and hydrophilic residues and a decrease of large hydrophobic and charged amino acid residues. It is conceivable that the structural basis underlying the functional adaptation to cold of C. aceratus GDH resides in the (i) increase of hydrophobic residues with smaller side chains, which increase £exibility by decreasing the compactness of the molecular structure, and (ii) decrease of charged amino acid residues, which can reduce the number of salt bridges [7,43]. Similar to other vertebrate GDHs, C. aceratus GDH shows dual-coenzyme speci¢city; higher activity is observed with NAD(H). In contrast, bovine and other mammalian GDHs show the same activity level with either coenzyme. The preference for NAD(H) in C. aceratus GDH is shared with tuna [44] and dog¢sh [45] enzymes and is probably related to the speci¢c metabolism of ¢shes. Biosynthetic and catabolic roles have been attributed to NADP - and NAD -dependent GDHs, respectively. Therefore, an amphibolic role has been suggested for vertebrate GDHs showing similar activity with both coenzymes [46]; a preferential catabolic role can be hypothesised for ¢sh liver GDHs, namely glutamate deamination might play a more important physiological role than glutamate biosynthesis. In line with this hypothesis, the forward reaction shows higher sensitivity to allosteric e¡ectors, and therefore it is likely to be characterised by ¢ner metabolic regulation. Under standard conditions, the maximal activity of C. aceratus GDH is shifted towards low temperature, relative to the mammalian orthologue. Never-
21
theless, optimal activity at higher temperatures is observed by increasing the NAD concentration, necessary to reach saturation. Thus, this shift depends on the modi¢cation of ligand saturating concentration rather than heat inactivation. Therefore, caution is necessary when factors such as temperature optima are used as indicators of temperature adaptation. Kinetic characterisation revealed that the catalytic e¤ciency (kcat /S0:5 ), a measure of enzyme physiological e¤ciency [47], of Antarctic GDH, measured at low temperature is strikingly similar to that of bovine GDH at 35³C, indicating e¤cient functional compensation of the Q10 e¡ect. This result is achieved partially by increasing kcat and partially by modulating S0:5 for substrate and coenzyme. Coenzyme binding in C. aceratus GDH is a¡ected by temperature in an opposite way as compared to the bovine enzyme; NAD concentration seems to be an important parameter which, depending on temperature, can a¡ect the kinetic behaviour with glutamate of Antarctic GDH. A general rule on Km modi¢cation accompanying enzymatic cold adaptation appears not to be valid for all enzymes. Km values higher than those of homologous mesophilic enzymes are described in Antarctic ¢sh lactate dehydrogenase [11], but are neither observed in C. aceratus GDH (at least in S0:5 for NAD ), nor in G6PD from two other Antarctic ¢shes, Chionodraco hamatus and Dissostichus mawsoni [12]. The free energy of activation (vG*) is lower in the Antarctic enzyme, in line with reported results [12,48]. The higher negative entropy (vS*) may re£ect a condition in which a higher number of conformational states is accessible to the molecule [11], as well as a higher degree of structure ordering necessary to form the activated complex. Nevertheless, the conformational changes are obtained with less energy input (lower vH*), indicating higher protein £exibility which would allow more e¤cient catalysis in a cellular environment characterised by low heat and entropic content. In thermostability studies, measurements of unfolding by UV spectrophotometry and calorimetry were impaired by irreversible precipitation (data not shown). Heat inactivation followed ¢rst-order kinetics, allowing calculation of kinetic constants
BBAPRO 36261 8-11-00
22
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
and thermodynamic parameters. The validity of these parameters is supported by the observation of Fukushima et al. [49], who reported that time course of heat inactivation and protein unfolding of bovine GDH, monitored by £uorescence, were described by the same curve. Irreversible heat inactivation of ¢sh and bovine GDHs occurred at similar temperatures. However, Antarctic GDH was abruptly inactivated in a much narrower temperature range than bovine GDH; this behaviour had also been observed in glucose-6-phosphate dehydrogenase and glycogen phosphorylase b from Antarctic ¢sh [50]. As a consequence, the heat of activation of the inactivation process (v H i ) of the Antarctic enzyme was higher than that of bovine GDH. A comparative study of GDH inactivation as a function of the hydrostatic pressure showed a similar behaviour [51]. In fact, Antarctic and bovine GDH were inactivated at similar pressures, but the ¢sh enzyme was inactivated in a narrower pressure range. Therefore, under extreme conditions of either temperature or pressure, the GDH inactivation appears to follow a similar pattern. The higher protein £exibility of Antarctic GDH may result from a higher e¤ciency (or co-operativity) in the disruption, under extreme conditions of temperature (higher v H i ) or pressure (higher vVg ; [51]), or disruption/formation during catalysis (lower vH*, which may represent the main thermodynamic compensation) of some weak bonds involved in maintaining the molecular structure. Biochemical cold adaptation of Antarctic GDH is thus linked to the amino acid substitutions which have been observed in comparison with the vertebrate sequences, and most likely to only few of these substitutions. The primary structure of C. aceratus GDH, in conjunction with molecular models built on the basis of the recently reported bovine structure [30,52], and/or the possibility in the future to compare the crystallographic structure of Antarctic and mesophilic GDHs, will be important tools to shed further light on the molecular cold adaptation of this enzyme. To this aim, structure^function characterisation of GDHs from non-Antarctic teleosts and site-directed mutagenesis will be additional invaluable approaches.
Acknowledgements This research is in the framework of the Italian National Programme for Antarctic Research. The authors wish to gratefully acknowledge a stimulating exchange of ideas with Professor G.N. Somero.
References [1] J. Eastman, Antarctic Fish Biology, Academic Press, New York, 1993. [2] G.N. Somero, A.L. de Vries, Science 156 (1967) 257^258. [3] G.N. Somero, E. Dahlho¡, J.J. Lin, in: I.A. Johnston, A.F. Bennet (Eds.), Animals and Temperature, Cambridge University Press, Cambridge, UK, 1996, pp. 53^78. [4] J.A. Macdonald, J.C. Montgomery, R.M.G. Wells, Adv. Mar. Biol. 24 (1987) 321^387. [5] P.W. Hochachka, G.N. Somero, Biochemical Adaptation, Princeton University Press, Princeton, NJ, 1984. [6] G.N. Somero, in: G. di Prisco, B. Maresca, B. Tota (Eds.), Biology of Antarctic Fish, Springer, Berlin, 1991, pp. 232^ 247. [7] S. Genicot, F. Rentier-Delrue, D. Edwards, J. Van Beeumen, C. Gerday, Biochim. Biophys. Acta 1298 (1996) 45^57. [8] M. Aittaleb, R. Hubner, J. Lamotte-Brasseur, C. Gerday, Protein Eng. 10 (1997) 475^477. [9] C. Capasso, W.E. Lees, A. Capasso, R. Scudiero, V. Carginale, P. Kille, J. Kay, E. Parisi, Biochim. Biophys. Acta 1431 (1999) 64^73. [10] S.M. King, S.P. Marchese-Ragona, S.K. Parker, H.W. Detrich III, Biochemistry 36 (1997) 1306^1314. [11] P.A. Fields, G.N. Somero, Proc. Natl. Acad. Sci. USA 95 (1998) 11476^11481. [12] M.A. Ciardiello, L. Camardella, G. di Prisco, Biochim. Biophys. Acta 1250 (1995) 76^82. [13] G. di Prisco, H.J. Strecker, Eur. J. Biochem. 9 (1969) 507^ 511. [14] E.-H. Syed, P.C. Engel, Biochem. J. 271 (1990) 351^355. [15] S. Pazhanisamy, S.J. Maniscalco, N. Singh, H.F. Fisher, Biochemistry 33 (1994) 10381^10385. [16] N. Singh, H.F. Fisher, Arch. Biochem. Biophys. 312 (1994) 38^44. [17] X.-G. Wang, P.C. Engel, Biochemistry 34 (1995) 11417^ 11422. [18] A.C. Wilson, S.S. Carlson, J.W. Thomas, Annu. Rev. Biochem. 46 (1977) 573^639. [19] L. Camardella, V. Carratore, M.A. Ciardiello, G. Damonte, U. Benatti, A. De Flora, Blood 85 (1995) 264^267. [20] G. Di Matteo, G. Romeo, G. di Prisco, Arch. Biochem. Biophys. 200 (1980) 55^60.
BBAPRO 36261 8-11-00
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23 [21] J.H. Julliard, E.L. Smith, J. Biol. Chem. 254 (1979) 3427^ 3438. [22] K. Moon, D. Piszkiewicz, E.L. Smith, J. Biol. Chem. 248 (1973) 3093^3107. è .C. Fa¨gain, Biochim. Biophys. Acta 1252 (1995) 1^14. [23] O [24] F. Rentier-Delrue, S. Moyens, M. Lion, J.A. Martial, Gene 134 (1993) 137^138. [25] G. Feller, E. Narinx, J.L. Arpigny, M. Aittaleb, E. Baise, S. Genicot, C. Gerday, FEMS Microbiol. Rev. 18 (1996) 189^ 202. [26] A.O. Smala®s, E.S. Heimstad, A. Hordvik, P. Willasen, R. Male, Proteins Struct. Funct. Genet. 20 (1994) 149^166. [27] N. Amuro, K. Ooki, A. Ito, Y. Goto, T. Okazaki, Nucleic Acids Res. 17 (1989) 2356. [28] G. Tzimagiorgis, N.K. Moschonas, Biochim. Biophys. Acta 1089 (1991) 250^253. [29] Y. Nakatani, M. Schneider, C. Banner, E. Freese, Nucleic Acids Res. 16 (1988) 6237. [30] G. Mavrothalassitis, G. Tzimagiorgis, A. Mitsialis, V. Zannis, A. Plaitakis, J. Papamatheakis, N. Moschonas, Proc. Natl. Acad. Sci. USA 85 (1988) 3494^3498. [31] N. Benachenhou, G. Baldacci, Mol. Gen. Genet. 230 (1991) 345^352. [32] R.I. Eggen, A.C. Geerlind, K. Waldkotter, G. Antranikian, W.M. de Vos, Gene 132 (1993) 143^148. [33] B. Maras, V. Consalvi, R. Chiaraluce, L. Politi, M. De Rosa, F. Bossa, R. Scandurra, D. Barra, Eur. J. Biochem. 203 (1992) 81^87. [34] J.K. Teller, R.J. Smith, M.J. McPherson, P.C. Engel, J. Guest, Eur. J. Biochem. 206 (1992) 151^159.
23
[35] T. Nagasu, B.D. Hall, Gene 37 (1985) 247^253. [36] M.J. McPherson, J.C. Wootton, Nucleic Acids Res. 11 (1983) 5257^5266. [37] J.S. Barton, J.R. Fisher, Biochemistry 10 (1971) 577^585. [38] A.D. McCarthy, K.F. Tipton, Biochem. J. 230 (1985) 95^99. [39] N. Benachenhou-Lahfa, P. Forterre, B.J. Labedan, Mol. Evol. 36 (1993) 335^346. [40] M.T. Shoemaker, B.E. Haley, Biochemistry 32 (1993) 1883^ 1890. [41] H. Kim, B. Haley, Bioconjug. Chem. 2 (1991) 142^147. [42] P.E. Peterson, T.J. Smith, Structure 7 (1999) 769^781. [43] S. Davail, G. Feller, E. Narinx, C. Gerday, J. Biol. Chem. 269 (1994) 17448^17453. [44] F.M. Veronese, R. Bevilacqua, E. Boccu©, D.M. Brown, Biochim. Biophys. Acta 445 (1976) 1^13. [45] L. Corman, L.M. Prescott, N.O. Kaplan, J. Biol. Chem. 242 (1967) 1383^1390. [46] R.C. Hudson, R.M. Daniel, Comp. Biochem. Physiol. 106B (1993) 767^792. [47] M.R. Pollock, Biochem. J. 94 (1965) 666^675. [48] P.S. Low, J.L. Boda, G.N. Somero, Proc. Natl. Acad. Sci. USA 70 (1973) 430^432. [49] K. Fukushima, K. Matsumoto, M. Okawauchi, T. Inoue, R. Shimozawa, Biochim. Biophys. Acta 872 (1986) 42^49. [50] M.A. Ciardiello, L. Camardella, G. di Prisco, Cybium 21 (1997) 443^450. [51] M.A. Ciardiello, B. Schmitt, G. di Prisco, G. Herve¨, Mar. Biol. 134 (1999) 631^636. [52] P.E. Peterson, J. Pierce, T.J. Smith, J. Struct. Biol. 120 (1997) 73^77.
BBAPRO 36261 8-11-00
L-Glutamate
dehydrogenase from the Antarctic ¢sh Chaenocephalus aceratus1 Primary structure, function and thermodynamic characterisation: relationship with cold adaptation M. Antonietta Ciardiello, Laura Camardella, Vito Carratore, Guido di Prisco * Institute of Protein Biochemistry and Enzymology, C.N.R., Via Marconi 10, I-80125 Naples, Italy Received 12 May 2000; received in revised form 18 July 2000; accepted 25 July 2000
Abstract In order to study the molecular mechanisms of enzyme cold adaptation, direct amino acid sequence, catalytic features, thermal stability and thermodynamics of the reaction and of heat inactivation of L-glutamate dehydrogenase (GDH) from the liver of the Antarctic fish Chaenocephalus aceratus (suborder Notothenioidei, family Channichthyidae) were investigated. The enzyme shows dual coenzyme specificity, is inhibited by GTP and the forward reaction is activated by ADP and ATP. The complete primary structure of C. aceratus GDH has been established; it is the first amino acid sequence of a fish GDH to be described. In comparison with homologous mesophilic enzymes, the amino acid substitutions suggest a less compact molecular structure with a reduced number of salt bridges. Functional characterisation indicates efficient compensation of Q10 , achieved by increased kcat and modulation of S0:5 , which produce a catalytic efficiency at low temperature very similar to that of bovine GDH at its physiological temperature. The structural and functional characteristics are indicative of a high extent of protein flexibility. This property seems to find correspondence in the heat inactivation of Antarctic and bovine enzymes, which are inactivated at very similar temperature, but with different thermodynamics. ß 2000 Elsevier Science B.V. All rights reserved. Keywords: Glutamate dehydrogenase; Cold adaptation; Amino acid sequence; Kinetic analysis; Thermodynamic analysis; Antarctic ¢sh
1. Introduction Environmental factors, such as temperature, may have profound e¡ects on enzyme activity, regulation
* Corresponding author. Fax: +39-81-593-6689; E-mail: [email protected] 1 The amino acid sequence of C. aceratus GDH has been submitted to the SWISS-PROT Data Bank with accession number P82264.
and structure. At decreasing temperatures the rates of enzyme-catalysed reactions are reduced by the low heat content of the cellular environment (Q10 e¡ect), which a¡ects the higher orders of the protein structure and the interactions with low-molecular-mass ligands. Therefore, strategies of biochemical adaptation allowing low-body-temperature species to obtain an adequate metabolic £ux, o¡setting the Q10 e¡ect, are expected. The study of organisms living under extreme conditions may help to understand temperature adaptation. A particularly good experimental system for
0167-4838 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 3 8 ( 0 0 ) 0 0 1 8 6 - 2
BBAPRO 36261 8-11-00
12
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
investigating adaptation to extremely cold temperatures are ¢shes of the teleost suborder Notothenioidei [1]. Most members of this suborder are endemic to the Southern Ocean, and in this habitat they encounter extremely low and stable temperatures, namely, temperatures near the freezing point of seawater, 31.9³C. Because of the constant temperature of the Antarctic sea, marine organisms developed a high degree of stenothermality in parallel with cold adaptation. In general, Antarctic ¢sh do not survive at temperatures only few degrees higher than that of the habitat [2^4]. Furthermore, it was reported [3,5,6] that the kinetic properties of enzymes from stenothermal species are conserved in a narrow temperature range whereas those of enzymes from eurythermal species show conserved properties over wider temperature range, indicating that the degree of stenothermy or eurythermy of the whole organism can be mirrored at the biochemical level. Comparative studies of structure^function relationships in enzymes from Antarctic and mesophilic organisms may provide useful information on the molecular basis sustaining cold adaptation. Although fragmentary studies on enzymatic cold adaptation in Antarctic ¢sh have appeared in the literature, extensive characterisation of puri¢ed enzymes has been undertaken only during the recent years. Structural and functional characterisation has been reported for several proteases (trypsin, elastase and cathepsin D) [7^9], an ATPase (dynein) involved in the function of contractile/locomotory proteins [10] and two enzymes having a key role in metabolism (lactate dehydrogenase and glucose-6-phosphate dehydrogenase) [11,12]. In vertebrate cells, L-glutamate dehydrogenase (GDH; EC 1.4.1.3), primarily associated to the mitochondrial matrix, has a key regulatory function in cellular metabolism in controlling levels of ammonia and glutamate. It catalyses the reversible oxidative deamination of L-glutamate to K-ketoglutarate and ammonia through reduction of NAD or NADP . In order to investigate the structure^function relationships in enzymes from cold-adapted ¢sh, we have puri¢ed and characterised GDH from the liver of the ice¢sh Chaenocephalus aceratus (belonging to the haemoglobinless family Channichthyidae). When appropriate, the kinetics of C. aceratus GDH were
analysed by carrying out parallel experiments on bovine GDH, in order to overcome reported di¡erences in kinetic properties ascribed to the reaction condition used [13^17]. The observation that GDH is one of the slowest evolving proteins for which extensive data on functional properties and amino acid sequences are available prompted its choice to study enzyme evolutionary adaptation to temperature. A rate of 1.8 point mutations per 100 residues per 108 years was estimated for this protein [18]. This feature is highly valuable for the study of molecular cold adaptation, since functional properties acquired during evolution can be correlated with relatively few amino acid substitutions. 2. Experimental procedures Specimens of C. aceratus were collected in the vicinity of Palmer Station, Antarctica (63³25P S, 62³15P W), where the seawater temperature is 31.8³C. Fish were transported to Palmer Station where livers were harvested and frozen before further treatment. LGlutamate, soybean trypsin inhibitor and cyanogen bromide were from Sigma; K-ketoglutarate, NAD , NADP , NADH, NADPH, GTP, ADP, ATP, phenylmethylsulfonyl£uoride, proteases (Lys-C, Asp-N, Arg-C) and benzamidine were from Boehringer. All other reagents were of the highest purity commercially available. Standard activity assays were performed at 20³C by recording the reduction of NAD (forward reaction) or the oxidation of NADH (reverse reaction), at 340 nm in a Varian DMS 300 spectrophotometer. The reaction mixture for the forward reaction contained 100 mM potassium phosphate bu¡er (pH 7.5), 40 mM L-glutamate, and 2 mM NAD . The reaction was initiated by addition of enzyme (1^3 Wg) in a ¢nal volume of 1 ml. The mixture for the reverse reaction contained 100 mM potassium phosphate bu¡er (pH 7.5), 100 mM ammonium sulfate, 5 mM K-ketoglutarate, 100 WM NADH and 0.2^0.4 Wg of enzyme in a ¢nal volume of 1 ml. One activity unit is de¢ned as the amount of enzyme catalysing the reaction of 1 Wmol of substrate/min under the standard assay conditions.
BBAPRO 36261 8-11-00
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
Kinetic constants were determined in 100 mM potassium phosphate bu¡er (pH 7.5), using 0.01^40 mM glutamate, 0.01^5 mM NAD . In the temperature-dependence experiments, the bu¡er was equilibrated at the assay temperature and pH was adjusted to the desired value. The assay mixture was equilibrated at the required temperature before adding the enzyme. 2.1. Puri¢cation All steps were carried out at 0^4³C. 2.1.1. Extraction Frozen livers were homogenised with 8 volumes of 50 mM potassium phosphate bu¡er (pH 7.5), 0.5 mM EDTA, 50 Wg/ml phenylmethylsulfonyl£uoride, 20 Wg/ml benzamidine and 10 Wg/ml soybean trypsin inhibitor. The mixture was centrifuged at 37 000Ug for 45 min and the supernatant was recovered. 2.1.2. Fractionation The supernatant was brought to 40% ammonium sulfate saturation and centrifuged at 37 000Ug for 30 min. The pellet was discarded; the supernatant was brought to 50% saturation and the precipitate was collected. 2.1.3. Heat treatment The precipitate was redissolved in a minimal volume of 50 mM potassium phosphate bu¡er (pH 7.5), containing 100 mM ammonium sulfate, kept at 52³C for 2 min and rapidly cooled. The mixture was diluted with an equal volume of bu¡er and centrifuged for 60 min at 37 000Ug. The cloudy supernatant was brought to 30% ammonium sulfate saturation, left at 4³C overnight, and centrifuged for 60 min at 37 000Ug. Protein in the clear supernatant was precipitated at 60% ammonium sulfate saturation and GDH activity was recovered in the pellet. 2.1.4. Chromatographic separations GDH was further puri¢ed by FPLC (Pharmacia, Uppsala, Sweden) at 4³C, monitoring the e¥uent at 280 nm. The solution, dialysed against bu¡er A (5 mM potassium phosphate, pH 8.0), was loaded on a Mono Q HR5/5 or an HR10/10 column equilibrated in 100% bu¡er A, at a £ow rate of 1 ml/min
13
or 3 ml/min, respectively. Elution was carried out by increasing the concentration of bu¡er B (200 mM potassium phosphate, pH 8.0) from 0% to 50% in 60 min. The active fractions were pooled and protein was precipitated at 60% ammonium sulfate saturation. The precipitate was dissolved in a minimal volume of 50 mM potassium phosphate bu¡er (pH 7.5), and 200 Wl was loaded on a gel ¢ltration column Superose 6 HR 10/30, equilibrated with 50 mM phosphate bu¡er (pH 7.5) containing 150 mM NaCl; the £ow rate was 0.4 ml/min. The active fractions were pooled and protein was precipitated at 60% ammonium sulfate saturation. Following centrifugation, the enzyme was dissolved in 100 mM potassium phosphate (pH 7.5)/50% glycerol and stored at 320³C. 2.2. Sodium dodecyl sulfate^polyacrylamide gel electrophoresis (SDS^PAGE) SDS^PAGE was carried out in the Bio-Rad Mini Protean apparatus, using 7% polyacrylamide gel. Staining and rinsing were performed in 0.02% Coomassie R-250 Brilliant blue in 10% v/v acetic acid/ 25% v/v isopropanol, and in 10% v/v acetic acid/25% v/v methanol, respectively. 2.3. Isoelectrofocusing The isoelectric point of puri¢ed GDH was measured in polyacrylamide gel in the pH range 3^9, at 10³C, in a Phast-System apparatus (Pharmacia). 2.4. Molecular mass The molecular mass of native GDH was determined by gel ¢ltration on Superose 6 HR 10/30 equilibrated in 0.1 M Tris^HCl (pH 7.5), 0.2 M NaCl and calibrated with thyroglobulin (669 kDa), ferritin (450 kDa), bovine GDH (333 kDa), catalase (240 kDa), and aldolase (158 kDa). The mass was calculated by plotting log molecular mass versus (Ve /V0 ). 2.5. Amino acid sequencing Denaturation and alkylation of the sulfhydryl groups with 4-vinylpyridine was carried out as de-
BBAPRO 36261 8-11-00
14
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
scribed [19]. Five 1-mg aliquots were subjected to chemical and enzymatic cleavages. An aliquot was dissolved in 70% formic acid (0.2 ml) and incubated with CNBr (5 mg) for 20 h at 20³C in the dark; the second was dissolved in 70% HCOOH (0.5 ml) and incubated for 20 h at 42³C. The peptide mixtures were evaporated under nitrogen and diluted with water to remove the acid, avoiding complete drying. The third aliquot was resuspended in 25 mM Tris^ HCl, pH 8.5 (0.5 ml), 1 mM EDTA and incubated for 20 h at 37³C, after addition of Lys-C (15 Wg) in 50 mM Tricine, pH 8.0 (0.15 ml), 10 mM EDTA; the fourth was incubated for 18 h at 37³C in 0.5 ml of a mixture containing 90 mM Tris^HCl (pH 7.6), 8.5 mM CaCl2 , 5 mM DTT, 0.5 mM EDTA and Arg-C (10 Wg); the last was dissolved in 0.06 ml 60% acetonitrile and incubated at 37³C for 6 h after addition of 200 mM sodium phosphate bu¡er, pH 8.0 (0.12 ml) and Asp-N (16 Wg) in 10 mM Tris^ HCl, pH 7.5 (0.4 ml). These three peptide mixtures were processed without drying. Peptides were separated by reverse-phase HPLC on a Vydac C4 column (4.6U250 mm, 5 W), using a Gold Apparatus (Beckman Instruments, Camino Ramon, CA). Elution was accomplished by a multistep linear gradient of eluent B (0.08% tri£uoracetic acid in acetonitrile) in eluent A (0.1% tri£uoracetic acid) at a £ow rate of 1 ml/min. The eluate was monitored at 220 and 280 nm. The separated fragments were manually collected and sequenced. Co-eluting peptides were further separated by reverse-phase HPLC on a WRPC C2/C18 column (2.1 mmU100 mm, 3 W) using a SMART System (Pharmacia). Elution was performed using a linear gradient of the solvents indicated above, at a £ow rate of 0.2 ml/min and monitored at 214 and 280 nm. Automated repetitive Edman degradation was performed on a pulsed liquid-phase Sequencer model 477A, equipped with a 120A analyser for the online detection of phenylthiohydantoin amino acids or on a Procise Protein Sequencer model 492 (Applied Biosystems, Division of Perkin^Elmer, Foster City, CA). 2.6. Allosteric e¡ectors ADP, ATP and GTP were dissolved in phosphate bu¡er and the pH was adjusted to 7.5. Activation
and inhibition under standard assay conditions are expressed as: % activation
V a 3V 0 U100; V0
and % inhibition
V 0 3V i U100 V0
where V0 is the reaction rate in the absence of nucleotides, Va and Vi are the reaction rates in the presence of activator and inhibitor, respectively. 2.7. Thermostability and thermodynamic parameters of heat-inactivation GDH (0.02 mg/ml) was incubated in 0.1 M potassium phosphate bu¡er (pH 7.5) at di¡erent temperatures; at time intervals the residual activity was measured at 20³C under standard assay conditions. The energy of activation of the heat inactivation process (Ea ) was calculated from the slopes of the Arrhenius plots of the rate constants (which equal 3Ea /R), the other thermodynamic parameters according to the following equations [20]:
v H i E a 3RT; v G i 3RTln hki =TkB where h is the Planck's constant, kB the Boltzmann constant and ki the rate constant of inactivation at T;
v S i v H i 3v Gi =T 2.8. Computational procedures Kinetic and thermodynamic parameters and standard errors of the mean were calculated using the computer program Enz¢tter (Elsevier-Biosoft) based on non-linear regression data analysis. 3. Results 3.1. Puri¢cation, pI, molecular mass GDH was puri¢ed from the frozen liver of C.
BBAPRO 36261 8-11-00
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
15
Fig. 1. Amino acid sequence of C. aceratus GDH. Arrows indicate peptide fragments obtained by acidic (Ac) or cyanogen bromide (CB) treatment or by enzymatic digestions with Lys-C (K), Asp-N (D), Arg-C (R). Peptides from each digestion are numbered according to their order in the sequence.
aceratus following the procedure described in the Experimental Procedures section. The recovery of electrophoretically pure GDH, with a speci¢c activity of 35 U/mg, was approximately 0.5 mg/10 g of tissue. Isoelectrofocusing showed a single band with pI = 5.3. The molecular mass of native GDH was estimated by FPLC gel ¢ltration on a calibrated Superose-6
column, from which it co-eluted with bovine GDH (333.4 kDa). SDS^PAGE showed a single band. 3.2. Primary structure Direct automated Edman-degradation of the native protein yielded a single sequence and 30 amino acid residues were identi¢ed from the N-terminus.
BBAPRO 36261 8-11-00
16
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
Fig. 2. Multiple alignment of the amino acid sequence of C. aceratus GDH with the vertebrate homologous enzymes known so far. The sequences were aligned by using the PileUp program (GCG-Wisconsin). The rat, mouse, human, bovine and chicken sequences were obtained from the SWISS-PROT Data Bank, using the accession nos. P10860, P26443, P00367, P00366, P00368, respectively; Z in the chicken sequence indicates ambiguous identi¢cation of Gln/Glu. The residues common to all sequences are underlined and those substituted only in the C. aceratus enzyme are marked by crosses. Asterisks indicate the residues of the Gly pattern typical of the coenzyme fold.
The sequence obtained by aligning the peptides from ¢ve chemical and enzymatic digestions of the S-pyridylethylated protein is shown in Fig. 1. Only peptides necessary to elucidate the complete sequence are indicated. Peptides were aligned by homology with vertebrate GDH sequences and by overlapping fragments. The deduced amino acid sequence contains 504 residues and the calculated molecular mass is 55 401 Da, in good agreement with results from SDS^PAGE. The amino acid composition deduced from the se-
quence of C. aceratus GDH was compared with those of enzymes from ox and chicken [21,22] (the only two vertebrate amino acid sequences obtained by direct protein sequencing, and therefore corresponding to the mature form of the protein). Antarctic GDH shows small di¡erences, such as higher amounts of Ala and Asn and lower amounts of Ser and Leu. The overall ratio Arg/(Arg+Lys), reportedly increased in thermophilic enzymes and decreased in psychrophilic enzymes [23^25], as well as the Gly and Pro contents, reportedly changed in
BBAPRO 36261 8-11-00
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
17
Table 1 Sequence identity (%) of GDHs from di¡erent sources Species
E. coli
S. cerev.
C. symb. S. solf.
P. fur.
H. salin.
Chicken
Ox
Rat
Human
Mouse
C. aceratus Mouse Human Rat Ox Chicken H. salinarium P. furiosus S. solfataricus C. symbiosum S. cerevisiae
28 28 29 26 26 26 29 32 32 53 55
28 27 27 27 25 25 32 32 31 50
30 31 31 31 30 30 33 36 32
39 42 42 42 40 39 47
42 40 40 39 38 38
83 91 91 90 94
84 95 95 94
88 99 97
88 98
88
38 36 36 38 36 35 42 44
Amino acid sequences of mouse [28], human [29], rat [27], bovine [21], chicken [22], Halobacterium salinarium [31], Pyrococcus furiosus [32], Sulfolobus solfataricus [33], Clostridium symbiosum [34], Saccharomyces cerevisiae [35] and Escherichia coli [36] GDHs, accession nos. P26443, P00367, P10860, P00366, P00368, P29051, P80319, P80053, P24295, P07262, P00370, respectively, were taken from the SWISS-PROT Data Bank. Percent identity was calculated with the Best¢t program (GCG-Wisconsin).
cold-adapted enzymes [26], are unchanged in Antarctic GDH. The amino acid sequence of C. aceratus GDH aligned with the vertebrate enzymes reported so far is shown in Fig. 2. The N-terminal region of Antarctic GDH is three residues longer than that of the bovine enzyme, but has the same length as chicken GDH; the comparison with the rat, mouse and human N-terminal region is not possible since only the nucleotide sequences encoding for the enzyme precursors are available for these organisms [27^29]. It is known, in fact, that GDH is synthesised as a larger precursor, as well as other intracellular translocatable proteins, and then processed to the mature form [30]. High conservation is observed among the sequences. Forty-two positions, conserved in all other vertebrates, are substituted in the ¢sh sequence. These substitutions produce an increase of Ala (Arg, Gly, Ser, AspCAla), Asn (His, Arg, Asp, SerCAsn) and Gln (Gly, HisCGln), and a decrease of charged (Asp, Glu, Lys, Arg, His) and large hydrophobic (Leu, Trp) residues. In these substitutions the residue is replaced by Ala seven times and by Asn four times. Most of these substitutions occur in clusters. Gln87 and His88 of Antarctic GDH are conserved in the same positions in the rat, mouse and human sequences, but their positions are inverted in ox and chicken. The same inversions occur in His224, Gly225 and Asn231, Glu232; Gly203, Lys204 and Pro205 correspond to Lys203, Pro204
and Gly205 in ox and chicken. This peculiarity accounts for approximately 45% of the di¡erence between C. aceratus GDH and mouse, rat and human GDHs, on one hand, and chicken and ox GDHs, on the other. Table 1 shows the sequence identity among GDHs of C. aceratus, vertebrates and representative bacteria, archaebacteria and yeast. The sequence identity with the other vertebrates (especially rat, human and
Fig. 3. Activity as a function of pH. Experiments were carried out in 100 mM potassium phosphate bu¡er, at 5³C (F) and 20³C (R) with C. aceratus and at 20³C (b) with bovine GDH.
BBAPRO 36261 8-11-00
18
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
mouse) is very high; the identity with GDHs from archaebacteria, bacteria and yeast gradually decreases. 3.3. E¡ect of pH Maximal activity of the forward reaction catalysed by C. aceratus GDH was reached at pH 8.0 and remained almost constant up to pH 8.9, both at 5³C and 20³C (Fig. 3). Continuously increasing activity was observed with bovine GDH at 20³C. The kinetic analysis described below was performed at pH 7.5. 3.4. Coenzyme speci¢city C. aceratus GDH can utilise both NAD(H) and NADP(H). At 20³C the activity with NADP(H) was approximately 10% of that with NAD(H) (the kcat values were 527, 54, 8500, 835/min with NAD , NADP , NADH, NADPH, respectively). A markedly lower a¤nity for the phosphorylated coenzyme was also observed (the S0:5 values of NAD , NADP , NADH, NADPH were 0.24, 1.6, 0.012, 0.12 mM, respectively). 3.5. E¡ectors The in£uence of ADP, ATP and GTP on the catalytic activity of C. aceratus GDH was investigated at increasing nucleotide concentrations. The forward and the reverse reactions were 90% inhibited by 80 and 200 WM GTP, respectively. 0.1 mM ADP and 2 mM ATP did not a¡ect the reverse reaction but produced a twofold activation of the forward reaction. 3.6. Kinetic properties The kinetics of the forward reaction at varying substrate and coenzyme concentration showed nonlinear double-reciprocal plots, as already reported for bovine GDH [37,38]; data were better analysed using the Hill equation that showed marked negative cooperativity. The kinetics as a function of NAD concentration were carried out with 40 mM glutamate since at 5^40³C this concentration saturated Antarctic and bovine GDH. The a¤nity for coenzyme of bovine
Fig. 4. Activity as a function of temperature. The rates of the reaction catalysed by C. aceratus GDH were obtained under standard assay conditions (R) and at coenzyme concentrations yielding saturation at each tested temperature (O). The rates of bovine GDH (b) were obtained under standard assay conditions, in which saturation is reached at each temperature.
GDH increased with temperature, and was highest at 35³C. In contrast, the a¤nity of the Antarctic enzyme was similar at 5 and 20³C, decreased at higher temperatures and was ten times lower at 40³C (Table 2). In bovine GDH 5 mM NAD was saturating between 5³C and 35³C. In C. aceratus GDH the NAD concentration necessary to reach saturation was 2 mM between 5³C and 20³C, 6 mM at 30³C and 15 mM at 40³C. At higher concentrations, an inhibitory e¡ect of NAD was observed. The saturation curves as a function of glutamate showed a Hill coe¤cient constantly below 1 for both Antarctic and bovine GDH. In bovine GDH, S0:5 values were similar at 20 and 35³C, but lower at 5³C, whereas in C. aceratus GDH the values were similar at all tested temperatures. In Antarctic GDH, the activity as a function of glutamate concentration was also investigated at 2 mM NAD. A decrease of S0:5 with increasing temperature (100 times lower at 40³C) and an increase of the Hill coe¤cient ( s 1 at 40³C) (Table 2, values in parenthe-
BBAPRO 36261 8-11-00
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
19
Table 2 Temperature e¡ect on the kinetic parameters of the forward reaction catalysed by C. aceratus and bovine GDHs Source
Assay S0:5 glu temperature (³C) (mM)
nH
S0:5 NAD (mM)
nH
kcat (/min)
kcat /S0:5 glu kcat /S0:5 NAD (/min/Wmol) (/min/Wmol)
C. aceratus
5 12 20 30 40
1.90 þ 0.17 2.10 þ 0.09 1.30 þ 0.13 1.10 þ 0.17 (0.62 þ 0.05) 2.00 þ 0.35 (0.02 þ 0.007)
0.60 0.70 0.66 0.70 (0.91) 0.50 (1.40)
0.22 þ 0.04 n.d. 0.24 þ 0.05 0.78 þ 0.20 2.60 þ 0.66
0.46 n.d. 0.54 0.68 0.55
228 þ 3 346 þ 5 527 þ 13 855 þ 66 1290 þ 127
120 165 405 777 645
1036 n.d. 2196 849 129
Ox
5 20 35
0.87 þ 0.1 5.00 þ 1.0 4.6 þ 0.07
0.75 0.41 0.54
1.70 þ 0.03 0.70 þ 0.1 0.54 þ 0.1
0.52 0.62 0.75
56 þ 1 245 þ 8 761 þ 28
64 49 165
33 70 1014
Values in parentheses were obtained with 2 mM NAD at 30³C and 40³C. n.d., not determined.
ses) was observed. These results indicated that: (i), in Antarctic GDH, depending on the temperature, the concentration of NAD has a profound e¡ect on the a¤nity for L-glutamate and on kinetics (positive or negative co-operativity) of substrate binding; (ii) a variation of the properties of Antarctic GDH, not shown by bovine GDH, occurs above 20³C. kcat values were calculated from the mean of the Vmax obtained from the saturation curves as a function of the concentration of substrate and coenzyme. At low temperature the rates measured with the Antarctic enzyme were higher than those of bovine GDH. kcat /S0:5 for NAD showed that the catalytic e¤ciency of C. aceratus GDH decreased at increasing temperature; the opposite trend was observed with the bovine enzyme. At 5³C, the catalytic e¤ciency for NAD of Antarctic GDH was 30-fold that of the bovine enzyme. 3.7. Temperature dependence of enzymatic activity The activity of C. aceratus and bovine GDH was investigated as a function of temperature under standard assay conditions. In the range 5^25³C, the Ant-
arctic enzyme catalysed the forward reaction with higher velocity than that of bovine GDH (Fig. 4) and had an apparent temperature optimum at 30³C. Above this temperature the activity decreased, although GDH was not irreversibly inactivated, as shown by thermostability experiments (see below). Bovine GDH showed maximal activity at 50³C and irreversible inactivation occurred above this temperature. Although the velocity of bovine GDH corresponds to kcat in the whole temperature range examined, whereas that of the Antarctic enzyme corresponds to kcat only in the 5^20³C range (as stated before, above 20³C substrate saturation is not achieved), Fig. 4 indicates nonetheless that the activity of Antarctic GDH decreases at lower temperatures than the bovine enzyme. As shown in many other enzymes from polar organisms, this feature is likely to be linked to cold adaptation. The e¡ect of temperature on kcat , measured between 5³C and 40³C, was analysed using the Arrhenius equation (Fig. 5). The activation energy Ea of Antarctic GDH (8540 cal/mol), calculated from the slope of the Arrhenius plot, was lower than that of the bovine enzyme (14 662 cal/mol). The thermody-
Table 3 Thermodynamic activation parameters of the forward reaction Source
Temperature (³C)
C. aceratus
5 30
Ox
5 35
vH* (cal/mol)
vS* (e.u.)
vG* (cal/mol)
7988 7938
327.01 327.08
15496 16143
14110 14060
37.81 37.12
16281 16217
BBAPRO 36261 8-11-00
20
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
Table 4 Kinetic constants and thermodynamic parameters of the heat inactivation process Source
Temperature (³C)
t1=2 (min)
ki (/min)
C. aceratus
48 50 52 45 50 52 55
63.0 þ 2.2 13.6 þ 1.0 3.0 þ 0.1 182.5 þ 3.1 49.3 þ 1.5 16.0 þ 0.5 3.0 þ 0.1
0.011 þ 0.001 0.051 þ 0.002 0.230 þ 0.007 0.004 þ 0.001 0.014 þ 0.001 0.043 þ 0.006 0.230 þ 0.007
ox
namic parameters of activation of the forward reaction vH*, vS* and vG* were calculated [12] at 5³C, 30³C and 35³C, and are reported in Table 3. 3.8. Thermostability C. aceratus and bovine GDHs fully retained activity after 30-min incubation at 45³C; above this temperature irreversible heat inactivation occurred (Fig. 6). The inactivation of Antarctic GDH in the range 45^52³C is much faster. The kinetics of inactivation of both enzymes displayed exponentially decreasing residual activity. The ¢rst-order constants of inactivation (ki ) and the half-inactivation times (t1=2 ) are shown in Table 4. The inactivation constants were plotted using the Arrhenius equation (inset of Fig. 6) and the enthalpy variations of activation of the heat inactivation process (v H i ) were calculated from the slopes. The thermodynamic parameters v H i , v G i and v S i accompanying the process of heat-inactivation of both Antarctic and the bovine
Fig. 5. E¡ect of temperature on kcat of C. aceratus (R) and bovine (b) GDH (Arrhenius plot).
vHi * (cal/mol)
vGi * (cal/mol)
vSi * (cal/mol/K)
178000
19021
490
135000
19633
355
GDHs, at 52³C, are also shown in Table 4. The enthalpy and entropy variations of heat inactivation of Antarctic GDH were higher than those of the bovine enzyme. 4. Discussion This paper reports the ¢rst primary structure of a ¢sh GDH, extending the knowledge on vertebrate GDH amino acid sequences. Alignment with homologous enzymes allows inclusion of this protein in Family II of the hexameric class of GDHs [39]. As
Fig. 6. Time-course of heat inactivation. C. aceratus (continuous lines) and bovine (dashed lines) GDHs at 45³C (b), 50³C (F) and 52³C (R). Inset: temperature e¡ect on ¢rst-order rate constants of inactivation of C. aceratus (continuous line) and bovine (dashed line) GDH.
BBAPRO 36261 8-11-00
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
expected from an enzyme having a very slow rate of evolutionary change, alignment with GDHs of other vertebrates shows high identity, ranging between 83% and 88%. As expected, the 35 invariant residues in 21 sequences of hexameric GDH [39], which include essential Lys126 (bovine GDH numbering), are conserved in Antarctic GDH. Furthermore, Lys445, involved in GTP binding [40] and Gly274 and Glu275, involved in the binding of the adenine moiety of NAD [41], are conserved. Only 8% of the residues conserved in all other vertebrate sequences are substituted in C. aceratus GDH. Some of these substitutions occur in clusters and most of them are in the region interconnecting the elements of secondary structure of bovine GDH [42]. These substitutions produce an overall increase of small hydrophobic and hydrophilic residues and a decrease of large hydrophobic and charged amino acid residues. It is conceivable that the structural basis underlying the functional adaptation to cold of C. aceratus GDH resides in the (i) increase of hydrophobic residues with smaller side chains, which increase £exibility by decreasing the compactness of the molecular structure, and (ii) decrease of charged amino acid residues, which can reduce the number of salt bridges [7,43]. Similar to other vertebrate GDHs, C. aceratus GDH shows dual-coenzyme speci¢city; higher activity is observed with NAD(H). In contrast, bovine and other mammalian GDHs show the same activity level with either coenzyme. The preference for NAD(H) in C. aceratus GDH is shared with tuna [44] and dog¢sh [45] enzymes and is probably related to the speci¢c metabolism of ¢shes. Biosynthetic and catabolic roles have been attributed to NADP - and NAD -dependent GDHs, respectively. Therefore, an amphibolic role has been suggested for vertebrate GDHs showing similar activity with both coenzymes [46]; a preferential catabolic role can be hypothesised for ¢sh liver GDHs, namely glutamate deamination might play a more important physiological role than glutamate biosynthesis. In line with this hypothesis, the forward reaction shows higher sensitivity to allosteric e¡ectors, and therefore it is likely to be characterised by ¢ner metabolic regulation. Under standard conditions, the maximal activity of C. aceratus GDH is shifted towards low temperature, relative to the mammalian orthologue. Never-
21
theless, optimal activity at higher temperatures is observed by increasing the NAD concentration, necessary to reach saturation. Thus, this shift depends on the modi¢cation of ligand saturating concentration rather than heat inactivation. Therefore, caution is necessary when factors such as temperature optima are used as indicators of temperature adaptation. Kinetic characterisation revealed that the catalytic e¤ciency (kcat /S0:5 ), a measure of enzyme physiological e¤ciency [47], of Antarctic GDH, measured at low temperature is strikingly similar to that of bovine GDH at 35³C, indicating e¤cient functional compensation of the Q10 e¡ect. This result is achieved partially by increasing kcat and partially by modulating S0:5 for substrate and coenzyme. Coenzyme binding in C. aceratus GDH is a¡ected by temperature in an opposite way as compared to the bovine enzyme; NAD concentration seems to be an important parameter which, depending on temperature, can a¡ect the kinetic behaviour with glutamate of Antarctic GDH. A general rule on Km modi¢cation accompanying enzymatic cold adaptation appears not to be valid for all enzymes. Km values higher than those of homologous mesophilic enzymes are described in Antarctic ¢sh lactate dehydrogenase [11], but are neither observed in C. aceratus GDH (at least in S0:5 for NAD ), nor in G6PD from two other Antarctic ¢shes, Chionodraco hamatus and Dissostichus mawsoni [12]. The free energy of activation (vG*) is lower in the Antarctic enzyme, in line with reported results [12,48]. The higher negative entropy (vS*) may re£ect a condition in which a higher number of conformational states is accessible to the molecule [11], as well as a higher degree of structure ordering necessary to form the activated complex. Nevertheless, the conformational changes are obtained with less energy input (lower vH*), indicating higher protein £exibility which would allow more e¤cient catalysis in a cellular environment characterised by low heat and entropic content. In thermostability studies, measurements of unfolding by UV spectrophotometry and calorimetry were impaired by irreversible precipitation (data not shown). Heat inactivation followed ¢rst-order kinetics, allowing calculation of kinetic constants
BBAPRO 36261 8-11-00
22
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23
and thermodynamic parameters. The validity of these parameters is supported by the observation of Fukushima et al. [49], who reported that time course of heat inactivation and protein unfolding of bovine GDH, monitored by £uorescence, were described by the same curve. Irreversible heat inactivation of ¢sh and bovine GDHs occurred at similar temperatures. However, Antarctic GDH was abruptly inactivated in a much narrower temperature range than bovine GDH; this behaviour had also been observed in glucose-6-phosphate dehydrogenase and glycogen phosphorylase b from Antarctic ¢sh [50]. As a consequence, the heat of activation of the inactivation process (v H i ) of the Antarctic enzyme was higher than that of bovine GDH. A comparative study of GDH inactivation as a function of the hydrostatic pressure showed a similar behaviour [51]. In fact, Antarctic and bovine GDH were inactivated at similar pressures, but the ¢sh enzyme was inactivated in a narrower pressure range. Therefore, under extreme conditions of either temperature or pressure, the GDH inactivation appears to follow a similar pattern. The higher protein £exibility of Antarctic GDH may result from a higher e¤ciency (or co-operativity) in the disruption, under extreme conditions of temperature (higher v H i ) or pressure (higher vVg ; [51]), or disruption/formation during catalysis (lower vH*, which may represent the main thermodynamic compensation) of some weak bonds involved in maintaining the molecular structure. Biochemical cold adaptation of Antarctic GDH is thus linked to the amino acid substitutions which have been observed in comparison with the vertebrate sequences, and most likely to only few of these substitutions. The primary structure of C. aceratus GDH, in conjunction with molecular models built on the basis of the recently reported bovine structure [30,52], and/or the possibility in the future to compare the crystallographic structure of Antarctic and mesophilic GDHs, will be important tools to shed further light on the molecular cold adaptation of this enzyme. To this aim, structure^function characterisation of GDHs from non-Antarctic teleosts and site-directed mutagenesis will be additional invaluable approaches.
Acknowledgements This research is in the framework of the Italian National Programme for Antarctic Research. The authors wish to gratefully acknowledge a stimulating exchange of ideas with Professor G.N. Somero.
References [1] J. Eastman, Antarctic Fish Biology, Academic Press, New York, 1993. [2] G.N. Somero, A.L. de Vries, Science 156 (1967) 257^258. [3] G.N. Somero, E. Dahlho¡, J.J. Lin, in: I.A. Johnston, A.F. Bennet (Eds.), Animals and Temperature, Cambridge University Press, Cambridge, UK, 1996, pp. 53^78. [4] J.A. Macdonald, J.C. Montgomery, R.M.G. Wells, Adv. Mar. Biol. 24 (1987) 321^387. [5] P.W. Hochachka, G.N. Somero, Biochemical Adaptation, Princeton University Press, Princeton, NJ, 1984. [6] G.N. Somero, in: G. di Prisco, B. Maresca, B. Tota (Eds.), Biology of Antarctic Fish, Springer, Berlin, 1991, pp. 232^ 247. [7] S. Genicot, F. Rentier-Delrue, D. Edwards, J. Van Beeumen, C. Gerday, Biochim. Biophys. Acta 1298 (1996) 45^57. [8] M. Aittaleb, R. Hubner, J. Lamotte-Brasseur, C. Gerday, Protein Eng. 10 (1997) 475^477. [9] C. Capasso, W.E. Lees, A. Capasso, R. Scudiero, V. Carginale, P. Kille, J. Kay, E. Parisi, Biochim. Biophys. Acta 1431 (1999) 64^73. [10] S.M. King, S.P. Marchese-Ragona, S.K. Parker, H.W. Detrich III, Biochemistry 36 (1997) 1306^1314. [11] P.A. Fields, G.N. Somero, Proc. Natl. Acad. Sci. USA 95 (1998) 11476^11481. [12] M.A. Ciardiello, L. Camardella, G. di Prisco, Biochim. Biophys. Acta 1250 (1995) 76^82. [13] G. di Prisco, H.J. Strecker, Eur. J. Biochem. 9 (1969) 507^ 511. [14] E.-H. Syed, P.C. Engel, Biochem. J. 271 (1990) 351^355. [15] S. Pazhanisamy, S.J. Maniscalco, N. Singh, H.F. Fisher, Biochemistry 33 (1994) 10381^10385. [16] N. Singh, H.F. Fisher, Arch. Biochem. Biophys. 312 (1994) 38^44. [17] X.-G. Wang, P.C. Engel, Biochemistry 34 (1995) 11417^ 11422. [18] A.C. Wilson, S.S. Carlson, J.W. Thomas, Annu. Rev. Biochem. 46 (1977) 573^639. [19] L. Camardella, V. Carratore, M.A. Ciardiello, G. Damonte, U. Benatti, A. De Flora, Blood 85 (1995) 264^267. [20] G. Di Matteo, G. Romeo, G. di Prisco, Arch. Biochem. Biophys. 200 (1980) 55^60.
BBAPRO 36261 8-11-00
M.A. Ciardiello et al. / Biochimica et Biophysica Acta 1543 (2000) 11^23 [21] J.H. Julliard, E.L. Smith, J. Biol. Chem. 254 (1979) 3427^ 3438. [22] K. Moon, D. Piszkiewicz, E.L. Smith, J. Biol. Chem. 248 (1973) 3093^3107. è .C. Fa¨gain, Biochim. Biophys. Acta 1252 (1995) 1^14. [23] O [24] F. Rentier-Delrue, S. Moyens, M. Lion, J.A. Martial, Gene 134 (1993) 137^138. [25] G. Feller, E. Narinx, J.L. Arpigny, M. Aittaleb, E. Baise, S. Genicot, C. Gerday, FEMS Microbiol. Rev. 18 (1996) 189^ 202. [26] A.O. Smala®s, E.S. Heimstad, A. Hordvik, P. Willasen, R. Male, Proteins Struct. Funct. Genet. 20 (1994) 149^166. [27] N. Amuro, K. Ooki, A. Ito, Y. Goto, T. Okazaki, Nucleic Acids Res. 17 (1989) 2356. [28] G. Tzimagiorgis, N.K. Moschonas, Biochim. Biophys. Acta 1089 (1991) 250^253. [29] Y. Nakatani, M. Schneider, C. Banner, E. Freese, Nucleic Acids Res. 16 (1988) 6237. [30] G. Mavrothalassitis, G. Tzimagiorgis, A. Mitsialis, V. Zannis, A. Plaitakis, J. Papamatheakis, N. Moschonas, Proc. Natl. Acad. Sci. USA 85 (1988) 3494^3498. [31] N. Benachenhou, G. Baldacci, Mol. Gen. Genet. 230 (1991) 345^352. [32] R.I. Eggen, A.C. Geerlind, K. Waldkotter, G. Antranikian, W.M. de Vos, Gene 132 (1993) 143^148. [33] B. Maras, V. Consalvi, R. Chiaraluce, L. Politi, M. De Rosa, F. Bossa, R. Scandurra, D. Barra, Eur. J. Biochem. 203 (1992) 81^87. [34] J.K. Teller, R.J. Smith, M.J. McPherson, P.C. Engel, J. Guest, Eur. J. Biochem. 206 (1992) 151^159.
23
[35] T. Nagasu, B.D. Hall, Gene 37 (1985) 247^253. [36] M.J. McPherson, J.C. Wootton, Nucleic Acids Res. 11 (1983) 5257^5266. [37] J.S. Barton, J.R. Fisher, Biochemistry 10 (1971) 577^585. [38] A.D. McCarthy, K.F. Tipton, Biochem. J. 230 (1985) 95^99. [39] N. Benachenhou-Lahfa, P. Forterre, B.J. Labedan, Mol. Evol. 36 (1993) 335^346. [40] M.T. Shoemaker, B.E. Haley, Biochemistry 32 (1993) 1883^ 1890. [41] H. Kim, B. Haley, Bioconjug. Chem. 2 (1991) 142^147. [42] P.E. Peterson, T.J. Smith, Structure 7 (1999) 769^781. [43] S. Davail, G. Feller, E. Narinx, C. Gerday, J. Biol. Chem. 269 (1994) 17448^17453. [44] F.M. Veronese, R. Bevilacqua, E. Boccu©, D.M. Brown, Biochim. Biophys. Acta 445 (1976) 1^13. [45] L. Corman, L.M. Prescott, N.O. Kaplan, J. Biol. Chem. 242 (1967) 1383^1390. [46] R.C. Hudson, R.M. Daniel, Comp. Biochem. Physiol. 106B (1993) 767^792. [47] M.R. Pollock, Biochem. J. 94 (1965) 666^675. [48] P.S. Low, J.L. Boda, G.N. Somero, Proc. Natl. Acad. Sci. USA 70 (1973) 430^432. [49] K. Fukushima, K. Matsumoto, M. Okawauchi, T. Inoue, R. Shimozawa, Biochim. Biophys. Acta 872 (1986) 42^49. [50] M.A. Ciardiello, L. Camardella, G. di Prisco, Cybium 21 (1997) 443^450. [51] M.A. Ciardiello, B. Schmitt, G. di Prisco, G. Herve¨, Mar. Biol. 134 (1999) 631^636. [52] P.E. Peterson, J. Pierce, T.J. Smith, J. Struct. Biol. 120 (1997) 73^77.
BBAPRO 36261 8-11-00