198

Biochimica et Biophysica/,eta, 10g0 ( 1991 ) 198-204 t ~ 1991 Elsevier Science Publishers B.V. All rights reser~ ~-dI}167-4838/91/$03.50

BBAPRO 34{131

The active site of

Sulfolobus soifataricus

aspartate aminotransferase Leila Birolo i, Maria Immacolata Arnone ~, Maria Vittoria Cubellis I, Giuseppina Andreotti ~, Gianpaolo Nitti -', Gennaro Marino i and Giovanni Sannia t Dipartimento di Chm~ica Organica c Biologica. Unit'ersitfi di Napoli "Federico I1'. Napoli (Italy) and 2 Dipartimento Biotecnologie. Farnlitalia Carlo Erba, Mi!ano (Italy) (Received t2 March 1091)

Key words: Asparta~c aminotran~fc ras~. ? , .aebaetcriam: Active site: Pyridoxal binding peptide: (S. solfataricus)

Aspartate aminotransferase from the archaebacteri~m Sulfolobus solfataricus binds pyridoxal 5' phosphate, via an aldimine bond, with Lys-241. This residue has been identified by reducing the enzyme in the pyridoxal form with sodium cyanoboro[SH]hydride and sequei:ei,lg ,~- specifically labeled peptic peptides. The aminoacidic sequence centered around the coenzyme binding site is higmi~, conserved between thermophilic aspartate aminotransferases and differs from that found in mesophilit iso~..~zymes. An alignment of aspartate aminotransferase from Sulfolobus solfataricus with mesophilic isoenzymes, attempted in spite of the low degree of similarity, was confirmed by the correspondence between pvridoxal 5' phosphate binding lesidues. Using this alignment it was possible to insert the archaebacterial aspartate aminotransferase into a subclass, subclass l, of pyridoxal 5' phosphate binding enzymes comprising mesophilic aspartate nminotransferases, tyrosine aminotransferases and histidinol phosphate aminotransferases. These enzymes sha'e t2 i~Jvariant amino acids most of which interact with the coenzyme or with the substrates. Some enzymes of subclass . and in particular aspartate aminotransferase from Sulfolob~zs solfataricus, lack a positively charged residue, corresponding to Arg-292, which in pig ¢ytosolie aspartate aminotransferase interacts with the distal carboxylate of the substrates (and determines the specificity towards dicarboxylic acids), it was confirmed that aspartate aminotransferase from Sulfolobus solfataric~ts does not possess any arginine residue exposed to chemical modifications responsible for the binding of a~-carboxylate of the substrates. Furthermore, it has been found that aspartate aminotransferase from Sulfolobus solfataricus is fairly active when alanine is used as substrate and that this activi~' is not affected by the presence of formate. The K M value of the thermophilic aspartate aminotransferase towards alanine is at least one order of magnitude lower than that of the mesophilic analogue enzymes.

Introduction Aspartate aminotransferases (AspATs) are pyridoxal 5' phosphate (PLP) dependent enzymes which

Abbreviations: AspAr, aspartate aminotransferase: AspATSs. aspartale aminotransferase from Sulfolobus solfataricus: AspATB, aspartare ammotransferase from Bacillus species YM-2; cAspATp, pig cytosolic aspartale aminotransferase; CSA, cy.steine-sulphinic acid; HPAT, histidinol phc:;pha~e aminotransferase: PLP, pyridoxal 5' phosphate; PMP. pyridoxamine 5' phosphate: TyrAT, tyrosine aminotransferasc Correslxmdence: G. Sannia. Dipartimento di Chimica Organica e Biologica, Universit'~ di Napoli. Via Mczzocannone, 16 80134 Napoli, Italy.

link the pathways of amino acid metabolism to the citric acid cycle [1]. Comparative studies of AspATs have made si~" cant contributions towards understanding t'.Je evolution and the structure-function relationships of this class of enzymes. Recently, attention has been focused on AspATs from thermophilic organisms. The enzyme has been isolated from a thermophilic Baci!!,~ species (YM-2) and extensive research is currently being carried out by Soda and co-workers [2,3]. In the authors' laboratory, aspartate aminotransferase was isolated from Sulfolobus solfataricus (AspATSs), a thermoaeidophilic archaebacterium. The main physico-chemical characteristics of this enzyme were determined [4] and particularly its thermophily and thermostability were studied. The temperature for optimum activity (100°C)

199 and the apparent melting temperature (109°C) suggest this enzyme should be classified as a hyperthcrmophilic c,:~,,mc [5]. The primary structure of AspATSs was deduct.:t from the nucleotid;c sequence of the gcnc and was compared with that of other aminotransferascs [6]. Mehta et al. [7. 8] recognized amongst the PLP-dcpendent enzymes a subclass comprising mesophilic aspartate aminotransferases (AspATs), tyrosine aminotransferases (TyrATs) and histidinol phosphate aminotransfcrases (HPATs), which share 12 invariant a,qno acids. They include the residues which are mainly responsible for the binding of the coenzyme and in particular the lysine residue which covalently binds PLP. Moreover, they include the arginine residue which interacts with the a-carboxylate of the substrates, it is worth noting that another arginine residue (Arg-292, according to the numbering of pig cytosolic AspAT (cAspATp)) which interacts with the tocarboxylate of the substrates and is more exposed to the solvent than the previous one [9], was not included in the group of the 12 invariant amino acids [7]. In fact, this residue is conserved in AspATs and TyrATs but not in HPATs although these last enzymes are also active on dicarboxylic substrates such as a-ketoglutaric acid and glutamic acid. AspATSs contains only a few amino acids, about 15% to 25% of the total, identical to those of mcsophilic aminotransferases. The highest degree of identity was observed when the AspATSs sequence was aligned with that of rat liver TyrAT, and a tentative alignment was reported [6]. However, due to the low degree of similarity, any attempt to align the sequence of AspATSs to those of other aminotransferases should be experimentally confirmed; this goal can be achieved by identifying, in the primary structure of AspATSs. the chemically accessible residues present in the active site, such as the lysyl residue which binds PLP and the residue which should interact with the carboxylatc of the sabstrates. Materials and Methods

Materials Aspartate aminotra,:sferase from Sulfidobus solfataricus was purified as previously reported [4]. Cytosolic and mitochondrial isozymes from pig heart were purified as described by Porter et al. [10]. Pepsin and lactate deilydrogenase were purchased from Sigma. Sodium cyanoboro[3H]hydride (5.0 Ci/mmol) and 2oxo[14C]glutarate (56.5mCi/mmol) were obtained from Amersham. All the other substances used were pure chemicals obtained from various commercial sources.

Methods Enz3'matic assays. To determine the K M of AspATSs towards aspartic acid and alanine, the enzy-

matic activity was assayed lk~llowingthe trarsamination of 2-oxo[~C]glutaratc (2.25 mM, (!.51 mCi/mmol) in 25 mM Tris-HCI (pH 8.5), 0.05 mM EDTA at 60°C. The reaction was stopped by diluting the samples 1:5 in I M HCOOH [I I]. [HC]Glu was scparatcd from the unrcacted kctoacid on an anion-exchange resin. AGlx8 Biorad, equilibrated in 11~mM Glu, 1 M HCOOH. The labeled product was eluted with 1 M HCOOH and the radioactivity was measured using a Beckman LS-1701 scintillation counter. The same radiometric assay was also employed to assay AspAT activity towards cysteinesulphinic acid, alaninc and alanine plus 1 M formate in 50 mM sodium borate pH 8.0, (I.01 mM PMP. The K M of cytosolic AspAT towards alanine was measured in 0.2 M Tris-HCI, 4.5 mM 2-oxo-glutarate, using a coupled assay with lactate dehydrogenase at 25°C. AspAT activity was routinely assayed using cysteinesulphinate as the aminoacidic substratc as already described [4]: the same method was also used, at 25°C, in order to determine the K M of cytosolie and mitochondrial AspAT towards cysteinesulphinate. Isolation of the PLP bh~ding peptide. 5 mg of AspATSs (4.8 mg/ml, in 300 mM (Hepes), pH 7.2) were reduced at 0°C by addition of 14#1 of a solution of NaCNB[3H]3 (1 M in NaOH 0.01 M, 4.2- 10?cpm/#l, 4.2- 10mcpm/mmol); ihese and all consecutive procedures were performed in the dark to minimize the photodcstruction of the chromt, phore [12]. After incubation for 10 rain at 0°C, the reactic,1 was stopped by addition of a 1000 × molar excess act.'.,me; the reduced enzyme was completely inactive and the maximum absorption due to the coenzyme was shifted to a shorter wavelength. The labeled protein was then precipitated with trichloroacetic acid (5% final concentration). The precipitate was dissolved in 2 ml 6 M guanidine. After incubation for 45 rain at room temperature, the solution was dialyzed extensively against 10 mM HCI and then lyophilized. An aliquot (0.5 mg) was hydrolyzed, under vacuum, at 110°C lor 24 h in HCI 6 M, and the hydrolyzate was analyzed using a Beckman mod. 119 CL amino acid analyzer. The remaining lyophilized enzyme was resuspended in 0.4 ml 5¢~ formic acid and partially hydrolyzed by addition of 50#g pepsin. After incubation for 90 rain at 37°C with gentle stirring, the reaction mixture became clear and was lyophilized. Peptides were separated by HPLC using 9, Violet C s (250 × 4.1 mm, 5#, 300 A) column on a Kontron instrument equipped with a double wavelength detector (mod. 430) and a data processing system (mod. 450). Peptides produced by peptic hydrolysis of reduced AspATSs were separated with a 90 min linear gradient of acetonitrile (5-60% v/v) in the presence of 0.1% iv/v) trifluoroacetic acid at a flow rate of 1 ml/min.

200 Elution was monitored at 210 and 305 nm; the only fraction retarded by the column which absorbed at 305 nm and accounted for about 10~ of the total radioactivity iacorporated into the protein, was further purified, A 40 min gradient of acetonitrile (5-60% v / v ) in the presence of 10 mM potassium phosphate (pH 7.5) was used to develop the chromatograph at a flow rate of 1 ml/min. The fractions absorbing at 305 nm were lyophilized and subjected to automated Edman degradation. An Applied Biosystem equipment was used which consisted of a gas phase sequencer (mod 470A) and an HPLC PTH Analyzer (mod 120A) connected with a Microbore PTH C18 column. Chemical modification of arginyl residues. A s p A T (0.25 mg) was equilibrated against 50 mM sodium borate (pH 8.0) and incubated with 1.5 mM phenylglyoxal or with 10 mM cyelohexane-l,2-dione. Both modification reactions were carried out in the presence or absence of 300 mM glutamate, 30 mM 2-oxo-glutarate and were stopped, after 6(1 ntin, by dilution (1:101 with 511 mM sodium borate (pH 8.0).

peak was recovered from the analyzer column. The elution volume of this peak was intermediate between those of histidine and lysine and corresponded to that of N6-pyridoxyllysine [13] (data not shown). After being reduced with sodium cyanoboro[3H]hydride and denatured in 6 M guanidinium chloride, AspATSs was hydrolyzed with pepsin in 5% H C O O H at 37°C for 90 min. The resulting peptides were fractionatcd by reverse-phase H P L C and the absorbance monitored at 220 and 305 nm. A fraction retained by the column turned out to be radioactive and absorbed at 3(15 nm, thus indicating the presence of derivatized peptides. This fraction was [yophilized and rechromatographed using a different eluent system. On the basis of the elution profile, two peaks showing absorbance at 305 nm were collected and further analyzed. A homogeneous peptide corresponding to the first eluted peak was subjected to automated Edman degradation; its sequence is shown below (the picomol of PTH amino acids recovered after each step are shown under the corresponding residues)

Results

Phe-Ser-X:.x-Thr-Phe-Ser-Met-Thr-Gly-Trp-Arg-Leu (88) (401 (601 (32){ 16) (181 (121 (27) (12) (201 (3)

Coenzyme binding site

No PTH-amino acid was unequivocally identified at the third cycle of the Edman degradation; only PTHlysine was recovered, in a very low yield, in place of the derivatized amino acid. This might be interpreted on the basis of the possible destruction of pyridoxyllysine derivatives during the sequencing cycles or in the subsequent convertions of thiazolinones [13]. The second fraction eluted contained two peptides whose amino acid sequences were determined (data not shown). Sequences of all the three peptides obtained, partially overlapped, so that, by combining the sequence data available, it was possible to map the PLP-binding amino acid sequence in the region of the

In order to verify whether AspATSs, as do all other AspATs, binds the coenzyme via an aidimine bond, reduction with sodium cyanoboro[3H]hydride, was performed. The incubation of the enzyme in the pyridoxal form with this reagent resulted in the complete loss of activity and in a shift of the absorbance maximum centered at 335 nm to a lower wavelength, 305 nm. This suggests the reduction of an aldimine bond conjugated with the pyridine ring essential for catalytic activity. When the reduced enzyme was hydrolyzed with 6 M HCI and subjected to amino acid analysis, a radioactive

TABLE I

Apparent K~ for O'stemesulphinic acid (CSA), 2 o.to-ghaaratc (2oxo-Gh~L a.spartic acid and alanine AspATSs activity has been determined as described under Materials and Methods at 60 ° C i,~ the presence of different concentrations of each substraie. For comparison the apparent K M of AspAT from /5, c o / / a n d pig heart cytosol and mitochondria are reported. Apparent K M for CSA and Ala have also been determined, as described under Materials and Methods. for pig heart isozymes K M x 103 M

CSA 2oxo-Glu Asp Ala

S. solfataricus

E. coil

1.8 ~' (I.3 '~ 0.25 1411

N,D, 0.7 h 4.4 h N.D.

a Data reported by Marino et al. [4]. b Data reported by Mavrides and Orr [22]. ¢ Data reported by Martinez-Carrion el al. [23]. d Data reported by Eichele et al. [24].

cytosolic 26.3 (1.4 ~ 4.0 ~" > 1000

mitochondrial 7.1 0.7 J 0.5 d N.D.

201 T A B L E II Actiritv of Sulfolohus solfittaricus and pig (3'tc,iolic a.spartate aminotramfi,rasc.~ towards ahtnim, m the ahsem e and prese,wc t,[ fi,rmatc AspATSs and cAspATp were incubated in 51) mM sodium borate (pll 8.0). 0.1)1 mM pyrid(~xaminc ¢' phosph::tc. 2.25 mM 2-oxol '~(']ghm~ratc [0.51 m C i / m m o l ) at 61) ° C and 25 ° C. respectively. Different xubstratcs ~crc assayed: 2(1 mM cys|einc,,ulphinatc, 3(1t) mM ahminc, 3(HI mM alanine in the presence of I M sodium fl)rmalc. F(~r each substratc the nanomol of [ll(']ghitamalc produced per rain arc rcpo;tetl. Only in lile case of I mM aspartic acid the buffer was 25 mM Tris-II('l ~'pll 8.5) Substrate

AspATS~ (nmol/min)

cAspATp (nmol/min)

20 mM cysteine sulphinate

38.21) 39.1H1 5.18 3.04

24.311

I mM aspartic acid 300 mM alanine 300 mM alanine+ I M Na formate

AspATSs sequence spanning from residue 239 to residue 252 demonstrating that the PLP binding lysine corresponds to Lys-?41.

Substrates binding site AspATSs efficiently catalyzes the transamination of dicarboxylic substrates. In Table I, K M values towards different substrates are summarized and arc compared to those of mesophilic AspATs. Binding to AspATSs is not severely impaired if the substrate lacks a negatively charged group on the tocarbon; in fact, the K M of AspATSs towards alanine has been measured and turned out to be at least one order of magnitude lower than that tentatively determined for cAspATp. The activity of AspATSs towards alanine and cysteinesulphinic acid (CSA) was assayed following the transamination of 2-oxo['4C]glutarate; assuming 1(10% activity with CSA, the relative activity with the non-dicarboxylic substrate is 14%. This value is quite high if compared to the relative activity (It?b) of cAspATp towards alanine. The enzymes were assayed under the same conditions at the appropriate temperatures, 60°C and 25°C respectively (Table I1). More significant is the effect of formate on the transamination of alanine; in the presence of 1 M sodium formate, the activity of cAspATp is greatly enhanced (15 times) while that of AspATSs is not affected (Table !1). Many different experimont~ have been describe-,! which clearly demonstrate the involvement of an argi-

0.03 I).46

nine residue (Arg-292) in the binding of the tocarboxylate group of substrates in mesophilic AspATs [14, 15]. Arg-292 is exposed to the solvent and reacts with chemicals such as phenyiglyoxal or cyclohexane1,2-dione. The modification of this group in mesophilic AspATs drastically reduces the activity of the enzyme towards dicarboxylic substrates whilc it does not affect its activity towards alanine. Moreover, Arg-292 is protected from the action of chemical modifiers by dicarboxylic substrates such as 2-oxo-glutarate and glutamate [16-18]. In order to test the occurrence of an arginyl residue in AspATSs, functionally analogous to Arg-292 and equally exposed to chemical modification, incubation of AspATSs with 1.5 mM phenylglyoxal or with l0 mM cyclohexane-l,2-dione in the presence or in the absence of 30 mM 2-oxo-glutara:e and 300 mM glutamate, was performed. However, these experiments, carried out both at 25°C and at 60°C, had no effect on the enzymatic activity of AspATSs. Discussion

The residue of AspATSs which covalcntly binds pyridoxal phosphate (PLP) through the aldimine bond was identified. It is Lys-241 and corresponds functionally to the I ys-258 residue of cAspATp. The peptide sequence which includes Lys-241 can be compared to the PLP binding peptides of other eukaryotic and eubacterial AspATs (Fig. 1). For this :malysis, the

2~

255

Pig heart cytosol [25] Pig heart mitochondria E. coli Bacillus

...Q S F S ~ [26]

species ~M-2 [3] solfatarlcus

.Q S Y AIiK N M G L Y G E R V G A . . . .S S Y S K N F G L Y N E R V G A . . .

[27]

sulfolobus

26~

NFGLYNERVGN...

[6]

.N G V S

!

...N G F S ~

SHSMTGWRIGY... T F S M T G W R L G Y...

Fig. I. Comparison of sequences of pyridoxal binding peptides from various meso)hilic and thermophflic aspartate aminotransferases. The residues are numbered according to the sequence of pig cytosolie aspartate aminotransferase.

_0. p A r S s with the o t h e r m c s o p h i l i c A s p A T s is a b o u t 3()g: while it r c a c h c s a b o u t 75% wh¢~n A s p A T S s is c o m p a r c d with the o t h e r t h c r m o p h i l i c A s p A T from a

,~,"'~, e"~',.,~ o f a m i n o acids c o m p r i s e d b e t w e e n residue 237

and rcsiduc 252 o f A s p A T S s can bc c o n s i d e r e d . T h e d c g r c c o f identity in this region w h e n comp~tring As20 &Asp~Tp

- A P P S V F A ~ V P Q A Q P V L : : F K L ] A D F R E D P D P R K V - - - N L G V G A Y ~ T D D C ~ P W V

AspATSS

V S L L D F N C N H S Q V T G E T T L L Y K ~ / A R N V E K T K K I K ~ I D ~ G I C - -

AspATB

- - H K E i L A N R V K T L T P 3 T T L A I T A K A K E M K A Q G I D V I G L ~ A C - - - / P D

60

.~P;

80

CAspKTp

L P V V R K V E Q R

I

&spATSI

L P T F K R

Z RD-

- A A K [ A L

AspAT5

7NTp~?N

Z HD

A N D S S L N H E - - - Y L P Z

- .- A A ~

~QG

L G L A E F R T C A S R L A L G D D S

F T P ' Y T S A F G

D S H Q Q G Y T K Y T P S

100

GG

I D E L R E K

I ;~QYLN

L PAL

I I E I(F

KQR

120

cAspATp

V G O V O S L G G T G A L R Z

A~p~TSs

- V I V - T P G A K P A L F L - - - V F

AspATB

-

X Z V -

GV

GA

-

I.Y~

-

-

L F QV

T

P A L Q E K ~

P Y G T D V K K

K R DNQ

v

....

~ L N E G D E . . . .

I

L P D P S F ¥ $ YA

V ~ I

P I

P YWV

E VV

R S Y R Y W D T E K R G L D L Q C F L S D L E N A P

AspATB

V P V Y I E A T S - -

I Y A N L ) ~ W S R E E -

GFS

EQ-

I

D - V D D L Q S

NYKI

E F5 K1SK

T - A E Q L K N A

I

i

F V L H A C A H N

RTKH



PTC

T DP-

I V F N N P H N P T ~ ; T

T D K T K A V

220

I

I

K O I A ~ V H K R R F L F P F F D S A Y Q O Y A S G N L

AspATSs

K K I V D / S R D N K Z

N S P S N P T G M V Y T R E E L

AspATB

[ D ] [ A K I A L E ~

£-

-

KDAWA

IRYIrVB~.GF

E L F C A Q S F

L L S D E I Y D N F V - - - Y E - - G K H R S T L E D S D W R D F L

N ~ L IVS

D~:

ZYEK

26,0

LL-

-

-

Y N GAEHFS

1AQ

I

SE

¢AspATp

S K N I ' G L Y N E R V G N L T V V A K E P D S

AIpATSs

S K T F S H T G V R

LG-

AIpATB

S K S H S M T G W R

I

I V A K

I

I V I

N GV

-

I

300

L R V L S Q M E

K Z V R V T W S N P

PAQG-

R . . . . .

E I

Z Q K M G I

L A A N V Y T A P T S

. . . . .

D I

I

L A S H S T S N

O - Y - A A G N A

1 Y V N C F

EVKAQT

280

Y-

PEQW

L F S P N D V

240

cAspATp

I

K L A G C

200 •

KP

~ L L G ~

S Y P E Q v

180

KDI

£E K PNE

K D T P V Y V S S P T W E N ~ D G V F T T A G F N P S D E

16~

ASpATSs

LE¥

140

GAE~'LARWYNGTNN

K H V LYT

cAspATp

. . . . .

320

NAMTD

AR

VA

F V Q K A R V -

P T T A S Q Y A A

I

-

-

340

cAKpATp

R T L S D P E L F H E W T G N V K T M A D R 1 L S H R S E L ~ A R L E - - A L K T P G T W N H I T D Q I G

AspATSs

K A F - - - D T F D E V N ~ H V ~ L F K K R R D V H Y D E ~ T F V K G V E V S K P N G A F Y H F P N V S K

AspAT8

EAY--NGPQDSVEEHRKAFESRLETI~'PKLShIPGFKVVKPQGAFYLLPDVBE

360

380

CAIp~Tp

N F S F T G L N P I ~ Q V E Y -

AspATSs

Z L K T S O F D V K S L A I - K L

~SpATB

A A Q X T G F A S V D E F A S A L L T ~ A N V A V I P G B G F G - - - A P B T I R I S Y A - T 3 L N L I E

-

L 1VEKH I

IYL

£ E K G V V T I

LFSG

. . . . . . . . . . .

R I

N HC

G L T T KN

L'~

P G E V F P L N Z G K E F L R L S F A - V N E E V I K

400

¢kspkTp

Y V A T S I J4 [

AsIpATBa

E O J ¢ K I R I~ • A g 0 H N N S R

ASpATB

IC It

I

A V T K I Q

Z R X D R r

V K

Fig. 2. Sequence ali~mcnt, The amino acid sequence of S. solfamric,.~ aspartate aminotransferase (AspATSs~ [6] is aligned with that of cy.tos~4ic aspartate amim,:ransferase trom pig heart (cAspATp) [25] and thai of Ihermophilic aspartale aminotransferase from a Bacillus speck's (AspATB) [31. Tire residues are numbered according to the sequence of cAspATp, lnvariant amino acids of aminotransferases, as identified b3 Mehta et al. 17] are marked by an asterisk.

203

Bacillus species (AspATB) [3]; it is intercsting to notc that a similar average dcgrcc of identity (aboul 75";~) is obtained when the scqucnces of mesophilic AspATs in this region arc comparcd. The degree of identity drops when entire scqucnccs are considered, being about 21F~ between AspATSs and mesophilic AspATs, 36% amongst thermophilic AspATs and about 45% amongst mcsophilic AspATs. Therefore, two classes of AspATs, one grouping mesophilic enzymcs and the other grouping tl:crmophilic cnzymes, can bc dcfincd: in each class of AspATs the sequence of thc PLP binding pcptidc appears to be highly conserved. The comparison describcd above was based on the alignment of AspATSs with the other aminotransferases performed using a profile analy.,;is program [ 19]. Mehta et al. [7] were able to group mesophilic AspATs, TyrATs and HPATs into a subclass (subclass 1) among PLP binding enzymes and to recognize 12 inwiriant amino acids common to these enzymes. For the sake of simplicity, Fig. 2 shows only the alignment of AspATSs with a representative member of mesophilic AspATs, namely cAspATp which was chosen sincc the numbering of this enzyme is generally used to identify rclewmt residues in aminotransferases. According to this analysis, the twelve residues which arc invariant in AspATs, Tyr~,Fs and HPATs and are characteristic of subclass 1, are conserved in AspATSs and in particular Lys-241 in AspATSs corresponds to Lys-258 in cAspATp. The experiments described in this paper demonstrated that Lys-241 is indeed the PLP-binding residue. This result strengthens the physical reliability of the alignment and excludes the possibility that major artcfacts might have been introduced due to the low degree of identity between AspATSs and mcsophilic aminotransferases (18-25%). The sequence of AspATB is included in Fig. 2 following the alignment with AspATSs reported by Sung et al. [3]; this protein also shares thc 12 invariant amino acid residues discussed abovc. The proposed alignment makes it possible to forcsec some of thc properties of AspATSs; for example, it can be predicted that the archaebactcrial enzyme lacks an arginyl residue which corresponds to Arg-292. This hypothesis can bc experimentally tested since Arg-292, together with Lys-258, are the only residues in the mesophilic AspAT active sites whose function was reliably assessed by chemical modification experiments [20]. The experiments, performed with phcnylglyoxal and cyclohexane-l,2-dione, exclude the existence of an exposed arginyl residue in AspATSs which can be functionally related to Arg-292. It is worth noting that AspATB also lacks an arginine residue which corresponds to Arg-292 [3] (see also Fig. 2). Yet, AspATSs is able to efficiently bind dicarboxylic substrates, i.c. 2-oxo-glutarate, cysteinesulphinate and aspartate, as is indicated by the K M values. Hov,'ever, the negatively

charged groups located on the to-carbon atoms of these m,,:[cc~:[c~, might n,~! be primarily inv,~!vcd in !he rcc~gnition of substratcs by AspATSs. As a minter of fact, the activity of this enzyme towards alaninc, which is surprisingly high, is not affected by the presence of fiwmalc. As opposed to this, foro~,.ttc enhances the vlaninc amiv-transferay.~: ~.|C::.',ri[y of cAspATp, possibly filling the subsitc normally occcupicd by the distal carboxylatcs of the 'qibslratcs [21]. The experiments described in thb; paper support the hypothesis that thcrmophilic AspATS.~ behmgs to subcla~:, i of thc PLP-dcpcndcnt enzymes which comprisc mesophilic AspATs, I~rATs and HPATs. AspATSs, in fact, shares with 'he othcr aminotransferases the 12 invariant amino acids identified by Mchta et al, [7] and marked in Fig. 2 by an asterisk. These findings suggest that the tridimensional structure is conserved in subclass 1 ~f the aminotranstcrases and that crystallographic structures of mcsophilic AspA Is can bc used as models for the study of those enzymes, such as AspATSs, whose X-ray analysis is still in fieri. Moreover, the results described in this paper indicate that AspATSs possesses in its active site a subsite for the binding of tocarboxylatc, different from that of mesophilic AspATs and possibly similar to that of the other thcrmophilic AspAT and HPATs. Therefore, it can be hypothesised that the specificity towards diearboxylic substratcs was acquired by aminotransferascs in different ways throughout their evolution.

Acknowledgements The authors are indebted to Dr. A. Gambacorld, Head of the Unit "Battcri termofili" of the ICMIB of the CNR, Arco Felice, Italy. who kindly and generously provided us with the biomass. The authors gratefully acknowledge Professor P. Christen and Dr. P.K. Mchta 'or kindly providing thc alignmcnt of AspATSs. Grants wcrc obtained from the Ministero della Pubblica Istruzionc, Ministero dcii'Universith e Ricerca Scicntifica, Univcrsitfi di Napoli and Consiglio Nazionalc dclle Ricerche (Pro~etto Finalizzato Biotecnologic c Biostrumcntazionc). The skillful assistance of Ms. M. E. Lisboa is gratefully acknowledged.

References I Christen. P. and M ~zlcr, ¢3 [5. (1985) Transaminases, John Wiley and S,ms. New York. 2 Sung, M,II., TaniTawa. K., Tanaka, H., Kuramitsu, S., Kagamiyama, tl. and Soda, K. (1990)J. Bacleriol. 172, 1345-1351. 3 Sung. M - [ I . T'¢-i;,~',,:~ I~ T~*naka, tt., Kuramitsu, S., Kagamiyama, H,. Hirotsu. K., Okamoto. A,, Higuchi. T. and Soda, K (1991)J. Biol. ('hem.. in press. 4 Marino, G.. Nini, G.. Arnone, M.I.. Sannia, G., Gambacorta, A. and De Rosa. M. (1988)J. Biol. Chem. 263. 12305-123(Y4.

204 :Mar!:zt, (].. Arr~mc. M 1. Biroh~. l... ('uhclli',, M.V., Nilli. ( ; , l)ut'~'i. P., ZilppacoMJ 1:. anti %il|]nlLi. ( i (tt,~(,~l) in Vi!amin 116 ,rod carbt,n~l cal.d~sis (Wad,~, II. S,~d;~. K Vukm. T. :mJ Kag:tmi~.;mla, I1.. ¢ds.}. t'~crgamon Pro "~,. i0 pro'-,', ,5 ('ul'~clli,,. M.V., 1,(o,'.~. ('.. Nittl I,;.. ,\rmm,," M.I.. Marin~, (,. al:tl Sanr~ia. (i '~l~S0t [:UT I. l:~iochcm IN,5,375-381. "7 Mchla, P.g... l!alc, F l ,ind ('hr~Icm P. I IUNUl ['Ul I. Igi,~chcm. IS~,. 24~ 253. N Mcl'da, P.K. and ('hri~Icn, P. (19tH) in Vitamin 1]'5 and c.~rbontl ~.;:*:dysis IWaJa. tl Soda. K. Fukm. T. and I':agamiyama, I1.. eds.L I'cr?:~mo.~i P~css in pre~s. ~) Ford. (_i.G.. Eicheh:. G. and J.~llsoRiir,. J.N. (19Ntl) Proc. Nail. Acad. Sci. USA. 7~. 2%9 25'5"~ 10 Porler, P.EI.. Barra. 1), I-~),,sa. |-"_ (';inlaltlp~,~. (;., [)t~,~llan. S.. M~rtim. F., Shcehan, l). and \Vilkin~on, S.M, ( l ~ g l ) ('crop. Biochem. Physiol. hgl'k 737-7.~6. II PowLll. J.T. and Morrison, J.F. (1978) Eur. J. Biochem. 87, 3qI1-4INf. 12 Rilehey, J.M., (iiblmns. I and Schachman. II.K. (1977) Huwhemislr3' 16. 45S4- 45~). 13 "lanizawa, K.. Masu, Y.. ,\,,mu. 3_ i'am~ka~ il. ~.nd ,%,,d,., K. (19N91J. Biol. Chem. 261, 2445 2.14t). 14 Jans~mius, J N F_ichele, G., Ford, G.C., Kirsch, J.F,. PicoL D,. Thaller. ('., Vincent, M(;., Gehring, tt. aml ('hrislen, P. ( 1~i~44~,i~ ('heroical and lilologi.:al Aspec::; of Vitamin Bf, ('atalysis. Parl 1'1 (Evan~eh;poulos. A.E., ed.), pp. It15-21)3. Alan R. I,is~,. Ne'0, York. 15 Kirseh. J.K.. l£i~helc. G.. Ford. G.('.. Vincent, M.G...lans~mius.

I N.. Gehring. tl. and Christen. P. (t984) J. Mol. lliol. 174. 497-525. In Samlmcicr. F. and Chriqen. t'. (1982)J. Bi,~l. Chore, 257. 674517 R;ord,m, J.F. and Scandvrra, R. (19751 Biochim Bi¢,phys. Res. ('ommun, h6. 417-424. IN Gilberl. II.F. and O'Lcary. M.II. (197~1 Bi¢~chim. Biophys. Rcs. (',~mmun. h7, 19N-2i12. 19 (?rih~ov. M., Mcl,achlan, A.D. ,rod Eisenher~, D. (19N7) Pn~:. Nail. Acad. Sci. USA 84. 4355- 4358. 211 Gehring, M. (19851 in Transaminases (Christen. P. and Melzler, I).E., eds.L pp. 317-321, John Wiley and S~ms, New York. 21 Saier. M.II.. J r and Jenkins. W.T. (19671 J. Biol. Chem. 243, II11 I11~. 22 Mavrides, C and Orr, W. (1975) J. Biol. ('hem. 2511, 4128-4133. 2.1 Martinez-Carrion, M., Turano, C., Riva, F. and Fas~lla P. (19671 .I. Biol. Chem. 242. 1428-1430. 24 Eichele. G., Fo~d, G.C., G h m M., ,lansonius. J.N., Mavrides, C. :rod Christen. P. (197,])J. Mol Biol. i33, 161-180.

25 Qvchinnikov. Yu. A.. Egorov. C.A.. Aidanova. N.A., Feigina. M. Yu.. L~pkin, ~¢.M., Abdulaev. N.G.. Grishnin. E.V., Kiselev. A.P., Modyamw, NN.. Braunstein, A.E.. Polyanovsky. O.I.. and Nosikov, V.V. (1973) FEBS Letl. 29, 31-34. 2h Kagamiyama. 11., Sakakibara, R., Tanase. S., M(}rino, Y., and Wada, I I. ( I nNl') J. Biol. ('hem. 255, 6153-6159. 27 Kuramilsu, S.. ()kuno. S., O~awa. T.. Ogawa,. !. and Kagamiyama. II. (1985) J. Biochem. (T,.~kyo) ~,~7. !259 1262.

The active site of Sulfolobus solfataricus aspartate aminotransferase.

Aspartate aminotransferase from the archaebacterium Sulfolobus solfataricus binds pyridoxal 5' phosphate, via an aldimine bond, with Lys-241. This res...
527KB Sizes 0 Downloads 0 Views