6ur J . Riochem 204,1183-1189 (1992) c, FEBS 1992

Limited proteolysis as a probe of conformational changes in aspartate aminotransferase from Sulfolobus solfataricus M. 1. ARNONE'. L. BIROLO', M. GIAMBERINI', M. V. CUBELLIS', G. NITTI', G. SANNIA' and G. MARINO'

Dipartimento di Chimica Organica e Biologica, Universiti di Napoli, Italy Dipartimento Biotecnologie, Farmitalia Carlo Erba, Milano, Italy

(Received September 18/November 11, 1991) - EJB 91 1247

The analysis of conformational transitions using limited proteolysis was carried out on a hyperthermophilic aspartate aminotransferase isolated from the archaebdcterium Sulfolobus solfutaricus, in comparison with pig cytosolic aspartate aminotransferase, a thoroughly studied mesophilic aminotransferase which shares about 15% similarity with the archaebacterial protein.

Aspartate aminotransferase from S. solfataricus is cleaved at residue 28 by thermolysin and residues 32 and 33 by trypsin; analogously, pig heart cytosolic aspartate aminotransferase is cleaved at residues 19 and 25 [Iriarte, A., Hubert, E., Kraft, K. & Martinez-Carrion, M. (1984) J . Biof. Chem. 259, 723 - 7281 by trypsin. In the case of aspartate aminotransferase from S. soffataricus, proteolytic cleavages also result in transaminase inactivation thus indicating that both enzymes, although evol- utionarily distinct, possess a region involved in catalysis and well exposed to proteases which is similarly positioned in their primary structure. It has been reported that the binding of substrates induces a conformational transition in aspartate aminotransferases and protects the enzymes against proteolysis [Gehring, H. (1985) in Trunsaminases (Christen, P. & Metzler, D. E., eds) pp. 323-326, John Wiley & Sons, New York]. Aspartate aminotransferase from S. solfataricus is protected against proteolysis by substrates, but only at high temperatures (> 60°C). To explain this behaviour, the kinetics of inactivation caused by thermolysin were measured in the temperature range 25 - 75 "C. The Arrhenius plot of the proteolytic kinetic constants measured in the absence of substrates is not rectilinear, while the same plot of the constants measured in the presence of substrates is a straight line.

Limited proteolysis experiments suggest that aspartate aminotransferase from S . solfataricus undergoes a conformational transition induced by the binding of substrates. Another con formational transition which depends on temperature and occurs in the absence of substrates could explain the non-linear Arrhenius plot of the proteolytic kinetic constants. The latter conformational transition might also be related to the functioning of the archaebacterial aminotransferase since the Arrhenius plot of k,,, is non-linear as well.

It has been observed that proteolysis of native globular proteins usually occurs in inter-domain regions particularly at 'hinges and fringes' of the polypeptide chain (for reviews see Neurath, 1980 and 1986, and Bennett and Huber, 1984). In addition, the studies carried out by Fontana and co-workers (1986) emphasized the additional role of protein flexibility, that is segmental mobility, as an essential part of the proteo- lytic event. For such reasons, limited proteolysis can be used

Correspondence to G. Marino, Dipdrtimento di Chimica Organica e Biologica, Universiti di Napoli, Via Mezzocannone 16, 1-80134 Napoli, Italy

Ahhrevintiorzs. AspAT, aspartate aminotransferase; AspATSs, aspartatc aminotransferase from Sulfolohus solfrctnricus; eAspATp, pig cytosolic aspartate aminotransferase; mAspATc, chicken mito- chondrial aspartate aminotransferase.

Enzymes. 4-Aminobutyrate aminotransferase (EC 2.6.1.19); aspartate aminotransferase (EC 2.6.1.1); dihydroxyphenylalanine decarboxylase (EC 4.1 .I .28); ornithine aminotransferase (EC 2.6.1 .13); serine hydroxymethyltransferase (EC 2.1.2.1); thermolysin (EC 3.4.24.4); trypsin (EC 3.4.21.4); tyrosine aminotransferase (EC 2.6.1.5).

____

as an effective tool to study the conformational properties of a globular protein, as well as to probe conformational transitions followed by a slight perturbation of its native state.

In the work described here, limited proteolysis experiments were used to investigate the conformational states of a thermophilic aspartate aminotransferase (AspAT) in compari- son with those of a mesophilic counterpart. The study was carried out on aspartate aminotransferase from the archaebacterium Sulfolobus solfalaricus (AspATSs) as its amino acid sequence (Cubellis et al., 1989) and main catalytic and physico-chemical properties (Marino et al., 1988) are known. The mesophilic enzyme chosen was the cytosolic isoenzyme from pig heart (cAspATp), one of the most thoroughly studied aspartate aminotransferases (Christen and Metzler, 1985), for which the tridimensional structure is known (Arnone et al., 1982).

In the field of the aspartate aminotransferases and, more generally, amongst the pyridoxal-phosphate-dependent en- zymes there are a large number of known cases in which some selective tryptic cleavages in the NH2-terminal region occur:

1184

such is the case of mitochondrial (Sandmeier and Christen, 1980; Meer and Gehring, 1984) and cytosolic (Meer and Gehring, 1984; Iriarte et al., 1984) aspartate aminotrans- ferases, tyrosine aminotransferase (Hargrove and Granner, 1981), serine hydroxymethyltransferase (Schirch et al., 1986), 4-aminobutyrate aminotransferase (Kim et al., 1984), dihydroxyphenylalanine decarboxylase (Tancini et al., 1988) and ornithine aminotransferase (Simmaco et al., 1989).

As far as aspartate aminotransferase is concerned, both the structural and functional roles of the NH2-terminal region, where some selective tryptic cleavages occur, have been well studied using crystallographic, spectroscopic, immunological and calorimetric methods (Gehring, 1985; Picot et al., 1991). In brief, it has been shown that in homodimeric aspartate aminotransferases, the NH2-terminal segment of one subunit runs on the surface of the globular part of the same subunit leaving a few exposed peptide bonds, passes in front of the active-site cleft and ends in contact with the other subunit (Eichele et al., 1979). This chain segment, which actually bridges the two subunits, spans residues 1-32 of chicken mitochondrial aspartate aminotransferase (mAspATc) whose structure is commonly used as a reference. Although once the polypeptide backbone has been proteolytically cleaved, no tight links fasten the NH2-terminal peptide to the core enzyme, yet the enthalpy and the transition temperature of thermal denaturation of the truncated enzyme are lower than that of the native enzyme (Iriarte et al., 1984). This suggests that the NH2-terminal peptide contributes to the stability of aspartate aminotransferase.

Moreover, it has been demonstrated that the bridging seg- ment is involved in conformational changes which seem to be an essential feature of the mechanism of action of AspATs. In fact, the loss of the NH,-terminal peptide leads to a decrease of the enzymatic activity although the modified enzyme retains its enzymatic competence (that is, the K, values are roughly unchanged). In addition, the same inhibitors or substrates that induce a conformational change in the crystalline enzyme (Jansonius et al., 1984; Harutyunyan et al., 1984), also mark- edly delay proteolytic cleavage of the NH2-terminal peptide caused by trypsin.

As a conclusion of these studies, 'conceivable roles of the bridging segment in catalysis are those of a mediator between the structural rearrangements at the active site accompanying the covalency changes and more distant parts of the enzyme or of a syncatalytically adjustable barrier at the entrance to the active-site pocket' (Sandmeier and Christen, 1980).

The work carried out so far on aspartate aminotransferase from S. solfataricus has shown a very low degree of similarity (about 15%) between its amino acid sequence and that of mesophilic aspartate aminotransferases (Cubellis et al., 1989), whereas the degree of identity amongst this class of enzymes, including all the eukaryotic and eubacterial mesophilic AspATs known, is about 45% (Mehta et al., 1989). Despite this, AspATSs shares with its mesophilic counterparts a large number of catalytically and structurally significant features (Marino et al., 1988, 1991 ; Birolo et al., 1991 a, b).

On the other hand, it is worth recalling that AspATSs belongs to a restricted group of enzymes defined as hyper- thermophilic (Marino et al., 1991), as demonstrated by its extreme thermophilicity and thermostability ; in fact, the opti- mum temperature of AspATSs is 100°C and the apparent melting temperature is about 109°C (Arnone et al., 1988).

A hypothesis has been put forward suggesting that thermophilic enzymes are rigid molecules with respect to their mesophilic counterparts (Brock, 1985; Fontana, 1991),

enough to retain their native structure at high temperatures; however, this feature could reflect negatively on their catalytic efficiency. This seems to be the case of aspartate amino- transferases, where some flexibility of the protein structure appears to be essential for efficient catalysis. In fact, the melting temperature of AspATSs is 30 "C higher than that of cAspATp (Marino et al., 1991) but in the temperature range where both enzymes are stable the specific activity of thermophilic AspATSs is always lower than that of mesophilic cAspATp (see below, Fig. 6).

Comparative studies on the flexibility of these enzymes seem to be very useful to clarify the molecular mechanisms of the thermophilicity of AspAT from S. solfataricus.

This paper describes how the structure and the functioning of thermophilic AspATSs can be related to those of its mesophilic counterparts using limited proteolysis. Moreover, the temperature dependence of the sensitivity of AspATSs to proteolysis is analyzed.

MATERIALS AND METHODS

Materials

Cells of Sulfolobus solfataricus (MT4) were kindly supplied by Dr Agata Gambacorta (Servizio Batteri Termofili, Istituto per la Chimica di Molecole di Interesse Biologico, Consiglio Nazionale delle Ricerche, Napoli). Thermolysin (from Bacil- lus thermoproteolyticus rokke) and trypsin (treated with N- tosyl-L-phenylalanine chloromethane) were purchased from Sigma; aprotinin was from Boehringer Mannheim. All other materials were of analytical grade.

Protein purification and assay

Aspartate aminotransferase from S. solfataricus (AspAT- SS) was purified in the pyridoxal form as previously described (Marino et al., 1988). Aspartate aminotransferase from pig heart (cAspATp) was purified as described by Porter et al. (1981).

AspATSs and cAspATp were routinely assayed at 6 0 T and 37°C respectively. The activity was measured, as pre- viously described (Marino et al., 1988) by monitoring the rate of increase in absorbance at 412 nm due to the addition of the enzyme to a reaction mixture (2 ml) containing 2 mM 2- oxoglutarate and 13 mM L-cysteine sulphinate, in 50 mM Tris/HCl pH 8.5, 0.1 mM EDTA and 0.15 mM 5,5'-dithio- bis(2-nitrobenzoic acid). The same spectrophotometric assay was employed to measure AspATSs activity at different tem- peratures ranging over 25 - 85 "C. At temperatures higher than 85°C a radiometric assay was used. The initial velocity of transamination was determined by extrapolation measuring the convertion of 2-0xo['~C]glutarate (2.5 mM, 0.51 mCi/ nmol) in 25 mM Tris/HCl pH 8.5, 0.1 mM EDTA, 13 mM L- cysteine sulphinate after different times of incubation with the enzyme. The reaction was stopped by diluting the samples 1 : 5 in 1 M HCOOH (Powell and Morrison, 1978). [14C]Glu was separated from the unreacted oxoacid on an anion-exchange resin, AGI x 8 Bio-Rad, equilibrated in 10 mM glutarate, 1 M HCOOH. The labeled product was washed off with 1 M HCOOH and radioactivity was measured using a Beckman LS-1701 scintillation counter.

Thermolysin was assayed at 35°C as described by Matsubara (1970) using casein as a substrate.

Protein concentrations were determined using the Bio-Rad protein assay system (Bradford, 1976).

1185

Tryptic hydrolysis

Digestions with trypsin were carried out in 50 mM Tris/ HCI pH 8.9 at 37°C using either AspATSs or cAspATp at concentrations of 1 mg/ml, with repetitive addition of trypsin (100 pg in 1 pl 0.001 M HCl every 30 min) to 1 ml amino- transferase solution. Digestions were stopped by addition of aprotinin (10 pg/ml final concentration).

Thermolytic hydrolysis

Digestions with thermolysin were carried out in 50 mM Tris/HCl, 2 mM CaC12, pH 8, at various temperatures ranging over 25 - 80"C, using 0.1 mg/ml or 1 mg/ml AspATSs. Ali- quots of thermolysin (1 : 20, final-mass ratio) were added every 30 min to AspATSs solutions. Hydrolysis was stopped by addition of 5 mM final concentration EDTA. In the cases of both thermolysin and trypsin, repetitive additions of relatively large doses of proteases were required due to their rapid autodigestion and thermal inactivation under the conditions used.

NH,-terminal analysis of electroblotted proteins

Samples of digested aspartate aminotransferase (10 pg) were electroblotted from sodium dodecyl sulphate/poly- acrylamide gels (12.5%) onto polyvinyldifluorobenzene mem- branes ('Immobilon', Millipore). Transfer was conducted as described by Yuen et al. (1986) in 10 mM 3-(cyclohexyl- amino)-I-propane-sulfonic acid, pH 11, 10% methanol at 5mA/cm2 for 1 h; transferred proteins were stained as de- scribed by Matsudaira (1987).

Automated Edman degradation was performed using an Applied Biosystems 470A gas-phase sequencer equipped with an on-line 120A phenylthiohydantoin amino acid analyzer and a 900A data module, following manufacturer's instruc- tions.

RESULTS

Limited proteolysis

AspATSs, when incubated at 37°C in the presence of tryp- sin, is cleaved after Lys32 and Lys33 and looses its activity. The time course for the hydrolysis followed by SDS gel electrophoresis shows that a protein band with an apparent molecular mass 4000 Da lower than that of native AspATSs appears as the aminotransferase activity decreases (Fig. 1 A). After a 330-min incubation at 37°C with trypsin, this band becomes the most abundant species and activity drops to 8%. Fig. 2 shows the NH2-terminal sequence of the truncated protein determined as described in Methods: the yields of the first step of Edman degradation reveal that 72% of the truncated protein derives from a cleavage after Lys33 and 18% derives from a cleavage after Lys32. Digestion with carboxypeptidase shows that the C-terminal sequence of the truncated form is the same as that of the native protein (data not shown). Taken together, these data demonstrate the for- mation of a 32/33 -402 AspATSs.

Since AspATSs is an extremely thermophilic enzyme, with an optimum temperature of 100°C (Marino et al., 1991), it was thought it might be interesting to carry out limited proteolysis experiments at temperatures higher than 37 "C. The protease chosen was thermolysin since it is active within a broad range of temperatures (Matsubara, 1970). Fig. 1 B shows that

thermolysin inactivates AspATSs at 60 "C, producing a lower- molecular-mass species and the loss of enzymatic activity cor- responds to the relative abundance of the truncated protein. Independently of the duration of incubation with thermolysin, a single amino-terminal sequence, starting with Va128, was determined for the truncated protein (Fig. 2) . These exper- iments lead to the conclusion that both at 37°C and at 60°C, AspATSs possesses an exposed segment accessible to proteo- lytic hydrolysis, located in the region comprising Va128, Lys32 and Lys33.

Substrate protection

The occurrence of a conformational change induced by the binding of the substrates was unequivocally proved by X-ray crystallography both in cAspATp and in mAspATc (Jansonius et al., 1984; Harutyunydn et al., 1984). It has been demonstrated that this conformational change influences the susceptibility of AspATs to proteolytic hydrolysis. In agree- ment with this and with the results obtained by Meer and Gehring (1984), the tryptic inactivation rate of cAspATp was measured and found to be reduced sevenfold in the presence of saturated substrate concentrations (35 mM glutamate and 2 mM 2-oxoglutarate; data not shown). On the other hand, susceptibility of AspATSs to tryptic hydrolysis at 37 "C is not influenced by the presence of substrates. It should be noted, however, that the activity of AspATSs at 37°C is about 1.5% of that measured at 75 "C using 2-oxoglutarate and cysteine sulphmate as substrates (Marino et al., 1988), although K, values at 37°C do not significantly differ from those deter- mined at 75°C; they are, in fact, respectively 0.33 mM and 0.29 mM for the oxoacid and 2.90 mM and 2.88 mM for the amino acid. Proteolysis experiments in the presence and ab- sence of substrates were then performed at 37°C and 75°C using thermolysin. Fig. 3 shows that, at 37°C (Fig. 3A), sub- strates are not able to protect the enzyme from proteolysis, while at 75°C (Fig. 3B) the presence of substrates lowers the inactivation rate about fourfold. Therefore, AspATSs also behaves as do the other mesophilic AspATs, but the protective effect exerted by substrates is observed only at high tempera- tures.

Temperature dependence of limited proteolysis

The effect of temperature on substrate-induced protection of AspATSs was investigated by hydrolysing the protein with thermolysin at different temperatures in the 25 - 75 "C range. AspATSs was incubated with thermolysin at different tem- peratures and then assayed under standard conditions at 60 "C: at each temperature differently timed hydrolysis exper- iments were carried out in the presence or in the absence of substrates (35 mM glutamate and 2 mM 2-oxoglutarate). In order to obtain pseudo-first-order kinetics for the hydrolysis reaction, the activity of thermolysin had to be constant during incubation with AspATSs. It was empirically determined that this requirement could be met at all temperatures by adding the protease repetitively (see Methods for details). Moreover, during the experiment carried out at 60"C, the activity of thermolysin was directly monitored by a standard caseinolytic assay (Matsubara, 1970) and it was found to be roughly con- stant over time with differences never exceeding 20%. Re- sidual transaminase activities were calculated after incubation with the protease and the half-life of the enzyme was derived at each temperature. The reciprocal of the half-life is an apparent kinetic constant (kAH) which depends on temperature, and on

1186

act ivi ty (8) hydrolysis(%) 100% I100

- 80

~ 60

- 40

- 20

I ' 0

0 50 100 150 200 250 300 350 time (min)

100 100

80 80

60 60

40 4 0

20 2 0

0 0 0 20 4 0 60 80 100 120 140 160 180 200

time (min)

Fig. 1. SDS/PAGE of S. soljiituvicus aspartate aminotransferase (AspATSs) digested with trypsin and with thcrmolysin. Aliquots of AspATSs at different incubation times with trypsin at 37°C (A) or thermolysin at 60'T (B) were taken from the reaction mixture and hydrolysis was immediately stopped (see Methods). Samples were assayed for aspartate aminotransferase activity (at 6 0 ' Q subjected to SDS/PAGE and gels stained with Coomassie brillant blue. The bands with molecular masses near 29 kDa in A, and near 36 kDa in B are due to trypsin and thermolysin, respectively. In the lower part of the figurc, the percentages of aminotransferasc activity, ( A ) and the densitomctrically determined modified AspATSs. (7) are reported.

10 20

V S L L D F N G N M S Q V T G E T T L L

30 4 0

Y K E I A R N V E K T K K I K I I D F G -. ~

t 4 Trypsyn

t-- - Thermolysin

5 0 6 0

I G Q P D L P T F K R I R D A A K E A L . . _ _ .

. . .

Fig. 2. N-terminal sequence of S. solfatuvicus aspartate aminotrans- ferase cleaved with trypsin and thermolysin. Arrows indicate sites of trypsin and thermolysin cleavage as determined by gas-phase sequcnce analysis of electroblotted protein bands after proteolytic digestion. The amino-tcrminal sequences of the truncated products determined are underlined.

the activation energy and the frequency factor of the proteo- lytic reaction as well as on the k,,, of transamination at 60'C. Table 1 shows the apparent constants of AspATSs inactivation rates in the presence (+) or in the absence (-) of substrates at different temperatures. It can be observed that the rates of inactivation both in the presence and in the absence of sub- strates increase with temperature and that the apparent kinetic constants measured in the presence of substrates are equal to or lower than those measured in the absence of substrates. The ratio obtained by dividing the half-life in the absence of SUbStrdteS by that measured in the presence of substrates depends on temperaturc; in the temperature range 25 -40°C

the ratio equals 1 , but it rapidly decreases as the temperature rises from 40°C to 75°C (Fig. 4). As preincubation with 2- oxoglutarate and glutamate does not influence the activity of AspATSs at 60' C and does not inactivate thermolysin, it can be concluded that the presence of substrates renders AspATSs less sensitive to the protease, possibly by inducing a confor- mational transition.

The Arrhenius plots of the apparent inactivation constants were fulfilled for both (+ or - substrates) hydrolysis kinetics (Fig. 5). While the Arrhenius plot in the presence of the substrates is linear, in the absence of substrates it is clearly non-rectilinear. In the latter case, the experimental data might be represented by two straight lines intersecting at about 60-C.

DISCUSSION

The experiments of limited proteolysis described in this paper show that AspATSs, in spite of a low degree of similarity, shares with mesophilic AspATs some relevant struc- tural and functional properties which can be summarized as follows.

a) Native AspATs, including the enzyme from Sdfo1ohu.s solfuturicus, are cleaved by proteases in their amino-terminal region. The peptide bonds sensitive to proteolysis in native AspATSs are at Lys32 and Lys33 for tryptic hydrolysis and at Val28 for thermolysin-catalyzed hydrolysis. It has been reported that cytosolic and mitochondria1 AspATs possess cleavage sites particularly exposed to tryptic hydrolysis in their amino-terminal region : these sites are Arg26 and Lys31

1187

B activ i ty ( % )

A activity 1%)

0 50 100 150 200 260 300 350 0 50 100 150 200 250 300 350

time (min) time (min)

Fig. 3. Inactivation of native aspartate aminotransferase from S. solfhiaricus after incubation with thermolysin. Digestions were performed at 37 'C (A) and 75°C (B) as dcscribcd in Materials and Methods. At the times indicated, aliquots were assayed for aminotransferase activity, at 60°C. ( * ) Enzyme in the pyridoxal form in the absence of substrate; (0) enzyme in the presence of 35 mM glutamate and 2 mM 2-oxoglutarate. The controls were kept a t 37°C and 75°C respectively without proteases and without substrates (A).

._____ Table 1. Inactivation rates of S. solfataricus aspartate aminotransferase. Inactivation rates of AspATSs in the absence (-) or presence (+) of

t1,2(-)/t1,2(+) 1 2

substrates (35 mM glutamate, 2 mM 2-oxoglutarate) by treatment with thermolysin, at various temperatures, are shown. Hydrolysis conditions are the same as indicated in Fig. 3. The inactivation rates are reported as the reciprocal of the half-lives (l / t l jz) calculatcd from a linear semilogarithmic plot of activity (%) versus time.

08 Temperature I 0 3 / t l j Z tl/Z ( - P I / z ( + I

25 1.5 1.5 1 31 2.0 2.0 1 37 2.4 2.4 1 42 2.7 3.6 0.75 49 3.3 4.9 0.68 50 3.8 5.9 0.65 58 4.0 1.7 0.52 65 5.3 13.3 0.39 67 5.2 14.1 0.35 71 6.2 20.8 0.30 15 7.5 34.5 0.22

in chicken and pig mitochondria1 AspAT (Sandmeier and Christen, 1980; Meer and Gehring, 1984), Arg25 and Arg31 in chicken cAspAT (Meer and Gehring, 1984) and LyslY and Arg25 in pig cytosolic AspAT (Iriarte et al., 1984). X-ray analysis of mesophilic AspATs revealed that the cleavage sites are located in a well exposed and potentially flexible stretch of amino acids (Eichele et al., 1979). Although analogously positioned in AspATs, these cleavage sites are not hydrolyzed at the same rate. In particular, the half-lives of AspATSs and of cAspATp, measured in the presence of trypsin at 37°C under the same experimental conditions, arc 113 min and 19 min, respectively (data not shown). Flexibility, more than exposure, could then determine the observed difference : in fact, the thermostability of proteins has been correlated both to their rigidity and to their resistance to proteolysis (Fontana et al., 1986; Daniel et al., 1982; Amelunxen and Murdoch, 1978).

11 I L L _ - ) - 0 20 30 4 0 50 60 7 0 80 90 100

temperature 'C Fig. 4. Temperature dependence of limited proteolysis. The ratio of half- lives in the absence (-) or presence (+) of substrates are plotted against temperature. Thc calculated values are those shown in Table 1.

b) The amino-terminal peptide is necessary for the integrity of the active site, both in mesophilic AspATs and in AspATSs, as its hydrolysis results in enzymatic inactivation: this obser- vation extends the analogies between thermophilic and mesophilic AspATs from the structure to the mechanism of action of these enzymes. It is worth mentioning that none of the amino-terminal residues of mesophilic AspATs interact strongly with the coenzyme or with the substrates, but they are critically involved in the conformational transition that accompanies the binding of substrates (Kirsch et al., 1984).

c) Two conformational states can be experienced by aspartate aminotransferases which differ in the exposure and/ or the flexibility of a peptide located in the amino-terminal

1188

4 t \

I P 285 2 9 295 3 ?I15 ? I 315 3 2 325 3 3 335 I/T x lo3 ( K 1 )

Fig. 5. Arrhenius plot of inactivation rates in the presence of thermolysin. The values of the apparent constant of inactivation rates (kAH) are calculated as the reciprocal of the half-lives shown in Table 1. The best fit of the In kA,, values versus l /T in the presence (+) or absence (-) of substrates are shown as a dotted line (-...) or full line (-) respectively.

[91 kJ/moll

0 :I,\-] 1 L _ - - 1 100 -,I 80 60

2 5 2 6 2 1 2 8 29 3 31 3 2 3 3 3 4 35 3 b

I t T x l o J ( K ' ) Fig. 6. Arrhenius plots for the catalytic constants of S. solfataricus and pig heart cytosolic aspartate aminotransferases. The kca, values for transaminase reaction were determined at different temperatures for AspATSs and cAspATp using a saturating concentration of L-cysteine sulphinate and 2-oxoglutarate as substrates. Different enzymatic as- says were employed depending on the temperature (see Methods for details): (+ ) values calculated assaying AspATSs radiometrically; (*) values calculated assaying AspATSs spectrophotometrically; (0) values calculated assaying cAspATp spectrophotometrically. The acti- vation energies calculated from the slopes of each linear segment are reported in brackets.

region. In the presence of substrates, all AspATs tested so far are cleaved more slowly than in the absence of substrates. Crystallographic analysis revealed that in the presence of sub- strates or dicarboxylic inhibitors, mesophilic AspATs assume a conformation, denoted as closed, which is more compact than the conformation, denoted as open, of the uncomplexed enzyme (Kirsch et al., 1984; Picot et al., 1991).

While carrying out experiments meant to draw analogies between mesophilic and thermophilic AspATs, an interesting feature emerged from the temperature dependence of the ap- parent rate constants of hydrolysis of AspATSs (In kAH versus 1/T, Fig. 5 ) . The plot obtained for the data collected in the presence of substrates is a straight line, while that obtained

for the data collected in the absence of substrates is clearly not linear. This effect could be due either to the catalyst (thermolysin) or to the substrate (AspATSs). The first hypoth- esis is improbable as AspATSs, used as a substrate in the presence of 2-oxoglutarate and glutamate, gives a linear plot and neither the oxoacid nor the amino acid affect the activity of thermolysin (data not shown). As the second hypothesis appears more likely, it follows that free AspATSs is a better substrate for thermolysin at high temperatures rather than at low temperatures.

A hypothesis can be put forward to suggest that free AspATSs undergoes a temperature-induced conformational transition and that the sensitivity to thermolysin is correlated to the flexibility of the protein.

It has previously been reported (Marino et al., 1988) that the Arrhenius plot of AspATSs k,,, is biphasic and its slope shows a discontinuity at about 60°C. This effect was more thoroughly investigated in the present study as shown in Fig. 6. The activation energy for transamination above 60'C is smaller than below 60°C, suggesting that the higher tem- perature induces a better catalytic fitness in the thermophilic enzyme. Amongst many other possibilities, this discontinuity could be due to a conformational change in AspATSs occur- ring at about 60°C.

The trend of the k,,, Arrhenius plot is not peculiar to AspATSs but is observed for many thermophilic enzymes, although the temperature at which the discontinuity occurs differs (see, for example, the cases described by Bryant and Adams, 1989; Cocco et al., 1988; Ferracin et al., 1989; Fusek et al., 1990; Hecht et al., 1989; Livingston et al., 1987; Plant et al., 1988; Sugimoto and Nosoh, 1971; Wrba et al., 1990).

In the case of AspATSs, a correspondence between the plots of In k,,, and In kAH versus 1/T is observed as both display a discontinuity at about 60°C. Both plots derive from the application of the Arrhenius equation in simple form to complex reactions. Each reaction can be influenced by a set of several parameters and the two sets are quite different. For this reason, the observed correspondence seems to suggest that a common explanation must be found and supports the hypothesis that a temperature-induced conformational change can occur in AspATSs.

The experiments described in this paper suggest that con- formational transitions occur in AspATSs induced by sub- strates and by temperature. Minor changes which do not affect the overall folding of the protein (see for example Picot et al., 1991) can account for the conformational transitions de- scribed above. For direct proof of these transitions, either the resolution of the X-ray structure of the conformers has to be obtained, or a properly targetted spectroscopic measurement has to be carried out.

The authors are indebted to Dr A. Gambacorta, Head of the Unit Batteri Termofili (ICMIB, CNR, Arc0 Felice, Italy) who kindly provided the biomass; we would also like to thank Dr B. Valsasina for her help in sequence analysis. Grants were obtained from the Ministero della Pubblica Istruzione, Ministero dell'Universita e Ricerca Scientifica, Universita di Napoli and Consiglio Nazionale delle Ricerche (Progetto Finalizzato Biotecnologie e Biostrumentazione) . The skilful assistance of Ms M. E. Lisboa is gratefully acknowledged.

REFERENCES Amelunxen, R. E. & Murdoch, A. L. (1978) Crit. Rev. Microbial. 6,

343 - 393. Arnone, A,, Briley, P. D., Rogers, P. H., Hyde, C. C., Metzler, C.

M. & Metzler, D. E. (1982) in Molecular structure und hiological

1189

activity (Griffen, J. F. & Daux, W. L., eds) pp. 57-74, Elsevier/ North Holland, Ncw York.

Arnone, M. I., Cubellis, M. V., Nitti, G., Sannia, G. & Marino, G. (1 988) ZtaL J. Biochem. 37,347 - 348.

Bennett, W. S. & Huber, R. (1984) CRC Crit. Rev. Biochem. IS, 291 - 384.

Birolo, L., Arnone, M. I. , Cubellis, M. V., Manno, G. & Sannia, G. (1991 a) in Vitamin B6 and carbonyl catalysis (Wada, H., Soda, K., Fukui, T. & Kagamiyama, H., eds) pp. 21 1 - 214, Pcrgamon Press, Oxford.

Birolo, L., Arnone, M. I., Cubellis, M. V., Andreotti, G., Nitti, G., Marino, C. & Sannia, G. (1991 b) Biochim. Biophys. Actu 1080, 198 - 204.

Bradford, M. M. (1976) Anal. Biochem. 72,248-254. Brock, T. D. (1985) Science 230,132- 138. Bryant, F. 0. & Adams, M. W. (1989) J . Biol. Chem. 264, 5070-

5079. Christen, P. & Metzler, D. E. (eds) (1985) Transaminases, John

Wiley & Sons, New York. Cocco, D., Rinaldi, A., Savini, I., Cooper, J. M. & Bannister, J. V.

(1988) Eur. J . Biochem. 174, 267-271. Cubellis, M. V., Rozzo, C., Nitti, G., Arnone, M. I., Marino, G. &

Sannia, G. (1989) Eur. J . Biochem. 186, 375 - 381. Daniel, R. M., Cowan, D. A., Morgan, H. W. & Curran, M. P. (1982)

Biochem. J . 207, 641 - 644. Eichclc, G., Ford, G. C., Glor, M., Jansonius, J. N., Mavrides, C. &

Christen, P. (1979) J. Mol. Biol. 133, 161 - 180. Ferracin, A., Annicchiarico, M., Coscarella, A., Teichner, A. &

Dell’Agata, M. (1989) Comp. Biochem. Physiol. 94B, 435 -443. Fontana, A. (3991) in Lifp under extreme conditions: biochemical

adaptation (di Prisco, G., ed.) pp. 89 - 113, Springer-Verlag, Berlin, Heidelberg.

Fontana, A,, Fassina, G., Vita, C., Dalzoppo. D., Zamai, M. & Zambonin, M. (1 986) Biochemistry 25, 1847 - 1851.

Fusek, M., Lin, X. & Tang, J. (1990) J . Biol. Chem. 265, 1496- 1501. Gehring, H. (1985) in Transaminases (Christen, P. & Metzler, D. E.

eds) pp. 323 - 326, John Wiley & Sons, New York. Hargrove, J . L. & Granner, D. K. (1981) J . Biol. Chem. 256, 8012-

8017. Harutyunyan, E. G., Malashkevich, V. N., Kochkina, V. M. &

Torchinsky, Yu. M. (1984) in Chemical and biological aspects of vitamin B6 catalysis (Evangelopoulos, A. E., ed.) vol. B, pp. 205 - 212, Alan R. Liss, New York.

Hecht, K., Wrba, A. & Jaenicke, R. (1989) Eur. J . Biochem. 183,69- 14.

Iriarte, A., Hubert, E., Kraft, K. & Martinez-Carrion, M. (1984) J . B id . C’hem. 259, 723 -728.

Jansonius, J. N., Eichele, G., Ford, G. C., Kirsch, J. F., Picot, D., Thaller, C. & Vincent, M. G. (1984) in Chemical and biological uspects of vitamin B6 catalysis (Evangelopoulos, A. E., ed.) vol. B, pp. 195-203, Alan R. Liss, New York.

Kim, D. S., Churchich, J. E. & Schiltz, E. (1984) in Chemical and biological aspects of vitamin B6 catalysis (Evangelopoulos, A. E., ed.) vol. B, pp. 161 -167, Alan R. Liss, New York.

Kirsch, J. F., Eichele, G. , Ford, G . C., Vincent, M. G. & Jansonius, J. N. (1984) J . Mol. Biol. 174,491425.

Livingston, D. J.. Fox, J. A., Orme-Johnson, W. H. & Walsh, C. T. (1 987) Biochemistry 26.4228-4237.

Marino, G., Nitti, G., Arnone, M. I., Sannia, G., Gambacorta, G. & De Rosa, M. (1988) J . Biol. Chem. 263, 12305-12309.

Marino, G., Arnone, M. I., Birolo, L., Cubellis, M. V., Nitti, G., Pucci, P., Zappacosta, F. & Sannia, G. (1991) in Vitamin B, und carbonyl catalysis (Wada, H., Soda, K., Fukui, T. & Kagamiyama. H., eds) pp. 43 - 53, Pergamon Press, Oxford.

Matsubara, H. (1970) Methods Enzymol. 19,642-645. Matsudaira, P. (1987) J . Biol. Chem. 262, 10035- 10038. Meer, L. & Gehring, H. (1984) in Chemical and biological aspects of

vitamin B6 catalysis (Evangelopoulos, A. E., ed.) vol. R, pp. 277- 283, Alan R. Liss, New York.

Mehta, P. K., Hale, T. 1. & Christen, P. (1989) Eur. J . Biochem. 186,

Neurdth, H. (1980) in Protein folding (Jaenicke, R., ed.) p. 501, Elsevier/North Holland Biochemical Press, Amsterdam. New York.

249 - 253.

Neurath, H. (1986) G e m . Scripta 27B, 221 -229. Picot, D., Sandmeier, E., Thaller, C., Vincent, M. G., Christen, P. &

Jansonius, J . N. (1991) Eur. J . Biochem. 196, 329-341. Plant, A. R., Parratt, S., Daniel, R. M. & Morgan, H. W. (1988)

Biochem. J . 255, 865 - 868. Porter, P. B., Barra, D., Bossa, F., Cantalupo, G., Doonan, S., Mac-

tini, F., Sheehan, D. & Wilkinson, S. M. (1981) Conzp. Biochem. Physiol. 6YB, 737 - 746.

Powell, J . T. & Morrison, J. F. (1978) Eur. J . Biochem. 87, 391 -400. Sandmeier, E. & Christen, P. (1980) J. Biol. Chem. 255, 10284-

10289. Schirch, L., Gavilancs, F., Peterson, D., Bullis, B., Barra, D. & Bossa,

F. (1986) Eur. J . Biochem. 161,45-49. Simmaco, M., Carr, L., Bossd, F., Barrd, D., Basford, J . M. & John,

R. A. (1989) J . Bid . Chem. 264,7473-7476. Sugimoto, S . & Nosoh, Y. (1971) Biochim. Biophys. Acta 235. 210-

221. Tancini, B., Dominici, P., Simmaco, M., Schinina, M. E., Barra. D. &

Voltattorni, C . B. (1988) Arch. Biochem. Biophys. 260, 569- 576.

Wrba, A., Jdenicke, R., Huber, R. & Stetter, K. 0. (1990) Eur. J . Biochem. 188, 195-201.

Yuen, S., Hunkapiller, M. W., Wilson, K. J. & Yuan, P. M. (1986) Applied Biosystems User Bulletin 25.