TIBS 17 - JULY 1992

PHOSPHORYL TRANSFER ENZYMES form one of the largest classes of enzymes 1, yet their mechanisms of action are still not fully understood 2. This class includes soluble inorganic pyrophosphatases (PPases), which hydrolyse inorganic pyrophosphate (PP~) to inorganic phosphate (Pt). These enzymes, recently shown to be essential in both bacteria 3 and yeast 4, generally have high activity in the cytoplasm, and are important for controlling the level of PP~ in the cell. PP~ is formed principally as the product of the many biosynthetic reactions that utilize ATP. The very high levels of PP~ that would result from the absence of PPase would certainly be toxic, accounting for the essential nature of this enzyme. Thus, Klemme5 calculated that, in the absence of PPase, the internal concentration of PP~ would rise to 3 M in one hour in rapidly growing Escherichia coli cells (the actual concentration is of the order of 1 mM). Even much lower concentrations of PP~ would make it difficult, for example, to synthesize RNA or DNA, or to charge a tRNA, because of the reversibility of these reactions. The two best-studied soluble PPases, those from Saccharomyces cerevisiae (SCE1-PPase) and E. coil (ECO-PPase), each provide a 101° rate acceleration as compared with the hydrolysis of PPt in solution, and a detailed understanding of their catalytic mechanisms is important for the understanding of the catalytic mechanisms of the class of phosphoryl transfer enzymes as a wholeE Progress in structure and mechanism studies on these PPases, and in the determination of amino acid sequences for PPases from several other organisms (Fig. 1) now provide strong support for the suggestion that the catalytic mechanism and active site structure of many PPases is, with minor variation, evolutionarily highly conserved, as discussed below. Five of the eight PPases, the sequences of which are displayed in Fig. 1 (SCE1, SCE2, ECO, BVR and PS3), are known to be soluble enzymes, although B. $, Cooperman is at the Department of Chemistry, Universityof Pennsyl(,ania, Philadelphia, PA 19104, USA.A./~ Baykov is at the A. N. BelozerskyInstitute of Physico-Chemical Biology, MoscowState University, Moscow119899, Russia. R. Lahti is at the Department of Biochemistry, Universityof Turku, SF-20500 Turku, Finland.

262

Evolutionary conservation of the active site of soluble inorganic pyrophosphatase

Soluble inorganic pyrophosphatases (PPases) are essential enzymes that are important for controlling the cellular levels of inorganic pyrophosphate (PPi). Although prokaryotic and eukaryotic PPases differ substantially in amino acid sequence, recent evidence now demonstrates clearly that PPases throughout evolution show a remarkable level of conservation of both an extended active site structure, which has the character of a minimineral, and a catalytic mechanism. PPases require several (three or four) Mg 2+ ions at the active site for activity and many of the 15-17 fully conserved active site residues are directly involved in the binding of metal ions. Each of the eight microscopic rate constants that has been evaluated for the PPases from both Escherichia coli and Saccharomyces cerevisiae is quite similar in magnitude for the two enzymes, supporting the notion of a conserved mechanism.

there is evidence that the mitochondrial enzyme, SCE2, can also bind to membranes ~7. The other three PPases (KLA, SPO and ATH) have only been characterized via their DNA sequences, but the strong sequence homologies evident in Fig. 1 make it likely that they correspond to soluble enzymes as well. By contrast, Sarafian et al. 18 have recently cloned and sequenced a plant tonoplast PPase, strongly anchored in the membrane, that has little or no obvious relationship to the sequences of the above PPases. This suggests that PP~ hydrolysis is catalysed in one way for soluble PPases but quite differently for membrane-bound PPases. Further consideration of this point will require the acquisition of more detailed mechanistic and active site structure information for the tonoplast enzyme than is currently available.

Overall, only a total of 24 residues are conserved among all eight sequences. Strikingly, 15 of these residues are part of a group of 17 polar residues identified by X-ray crystallographic studies of SCE1-PPase as being at the active site (see below). Two others, G94 and Pl18, are clearly within the active site cavity and may also play an important role in catalysis, either in forming the structure of the active site cavity or through direct participation via their peptide bonds.

Three-dimensionalstructural studies on SCEI-PPase and ECO-PPase Arutyuflyun and colleagues at the Institute of Crystallography in Moscow have published an unrefined 3 ]~ structure of the native SCE1-PPase dimer ~s. [In addition, Voet (University of Pennsylvania) and colleagues are in the process of completing a 2.5 ]~ structure of Amino acid sequences of PPases SCE1-PPase.] From a mechanistic point The eight complete PPase sequences of view, the most significant result of aligned in Fig. 1 may be placed into two this work is the finding of an active site groups, one including all four yeast cavity within the enzyme structure PPases as well as the PPase from which contains binding sites for four bovine retinal rod (internal identity > metal ions (for both SCE1-PPase and 49%), and one including the two bac- ECO-PPase, Mg2~confers highest activity, terial PPases (40% identical) as well as although appreciable activity is also the PPase from Arabidopsis thaliana found in the presence of Zn 2~, Co2~ or (36% identical with ECO-PPase). The Mn2~)19, as well as for two Pis ( or alternaidentity between ECO-PPase and SCE1- tively, for CaPP~, a known competitive PPase is 27% (Ref. 14). inhibitor of the enzyme). Within this © 1992,ElsevierSciencePublishers, (UK) 0376-5067/92/$05.00

TIBS 17 -

JULY 1992

cavity there are a total of 17 polar residues that are at or near metal-ion- or pyrophosphate-binding sites, including 15 that are totally conserved evolutionarily. Chirgadze et al. TM have recently reported the crystal structure of an enzyme-Mn3-P~2 complex in which the Mn-Mn distances are as shown (Fig. 2; there is also a fourth, more distant, metal-ion-binding site) ~5. The Mn3 site, which has strong interactions with amino acid residues Dl15, D152 and, through water, with Y93, and the Mn2 site, which has strong interactions with Dl17, D120 and E48, appear to correspond to the two high-affinity divalent metal ion sites bound deep within the active site cavity. The Mn~ site, with a strong interaction only with E58, most likely corresponds to an Mn2÷ion that is bound with high affinity only in the presence of P, or pp20. All three manganese ions bind to phosphate in one site, while the second phosphate only binds to Mn3. Further refinement of this structure should make it possible to construct a detailed picture of substrate binding to the active site. In addition to its consistency with the sequence studies presented above, the identification of the cavity as the active site is also consistent with the results of solution studies on SCE1PPase. Thus (1) SCE1-PPase requires three divalent metal ions per subunit for activity, with a fourth divalent metal ion that can be bound but plays an inhibitory roleS9; (2) electron spin resonance (ESR) results on both the Mn(II)21,22 and Cu(ll) 23 enzymes demonstrate the mutual proximity of at least three divalent metal ions on a PPase subunit; (3) nuclear magnetic resonance (NMR) and electron paramagnetic resonance (EPR) studies have indicated that P~ bound in the higheraffinity site (site 1) strongly interacts with at least two bound metal ions, whereas interaction with metal ions is weaker for P~ bound in the lower-affinity site (site 2)24.2s; (4) K56 (Ref. 26), R78 (Ref. 27), and E150 (Ref. 28), all of which fall within the cavity, have been shown to be important for enzymatic activity by chemical modification studies; and (5) u3Cd-NMR29 and Cu(II) ESR23 results clearly indicate the presence of a large number of oxygen-containing ligands at the active site, consistent with the large numbers of Glu (four), Asp (five) and Tyr (three) identified as active site residues. ECO-PPase is a hexamer of identical subunits. Crystals have been obtained

20 SCEI KLA SPO SCE2 BVR ATH

"ECO

40

-T-YTTRQIGAKNTLEYKVY-IEKDGKPVSAFHDIPLYADKENNIFNMVVEIPRWTNA-K -S-YTTRQVGAKNSLDYKVY-IEKDGKPISAFHDIPLYADEANGIFNMVVEIPRWTNA-K -SEYTTREVGALNTLDYQVY-VEKNGTPISSWHDIPLYANAEKTILNMVVEIPRWTQA-K HRQFSTIQQGSKYTLGFKKYLTLLNGEVGSFFHDVPLDLNEHEKTVNMIVEVPRWTTG-K -SSFSSEERAAPFTLEYRVFLKNEKGQYISPFHDIPIYADKE--VFHMVVEVPRWSNA-K KGYAFPLRNPNVTLNERNFAAFTHRSAAAHPWHDLEIGPEAPTVFN-CAVEI-SKGGKVK SLLNVPAGKDLPEDIY-VVIEIPANADPIK AFENKIVE-AFIEIPTGSQN-K

PS3 CONSERVED

K 60

i00

80

SCEI

LEI-TKEETLNPIIQDTKKGKLRFVRNCFPHHGYIHNYGAFPQTWEDPNVSHPETK

.... A

KLA SPO

LEI-TKEEPLNPIIQDTKKGKLRFVRNCFPHHGYIHNYGAFPQTWEDPNESHPETK LEI-TKEATLNPIKQDTKKGKLRFVRNCFPHHGYIWNYGAFPQTYEDPNVVHPETK

.... A .... A

SCE2 BVR ATH ECO PS3 CONSERVED

FEI-SKELRFNPIVQDTKNGKLRFVNNIFPYHGYIHNYGAIPQTWEDPTIEHKLGKCDVAL MEIATKD-PLNPIKQDVKKGKLRYVANLFPYKGYIWNYGAIPQTWEDPGHNDKHTG .... C YEL-DKNS--GLIKVD ...... RVLYSSI---VYPHNYGFIPRTIC ............... YEI-DKES--GALFVD ...... RFMSTAM---FYPCNYGYINHTLS ............... YEF-DKER--GIFKLD ...... RVLYSPM---FYPAEYGYLQNTLA ...............

K

D

B

120 SCEI KLA SPO SCE2 BVR ATH ECO PS3 CONSERVED

X

XG

T

140

160

VGDNDP IDVLE IGET IAYTGQVKQVKALGIMALLDEGETDWKVIAID VGDNDP LDVLE IGEQVAYTGQVKQVKVLGVMALLDEGETDWKVIAID

I N D P L A P K L N D IE INDP LAPKLND IE

KGD SDP LDVCE IGEARGYTGQVKQVKVLGVMALLD EGETDWKVIVIDVNDP LAP KLND IE KGDNDPLDCCE IGSDVLEMGS IKKVKVLGSLALIDDGELDWKVIVIDVNOP LS S K I D D L E C G D N D P I D V C E I G S K V C A R G E I I R V K V L G I L A M I D E G E T D W K V IA I N V E D P D A A N Y N D IN - ED SDPMDVLV-MQEPVLTGSF LRARAIGLMPMIDQGEKDDK I IAVC-CAD-DPEFRHYR - LDGDPVDVLVPTPYPLQPGSVIRCRPVGVLKMTDEAGEDAKLVAVP-HSKLSKEYDH IK - LDGDP LD ILVITTNPPFPGCVIDTRVIGYLNMVDSGEEDAKL DDP

D

G

G

180

200

D

IGVP-VED--PRFDEVR

DK

220

SCEI KLA

DVEKYFPGLLRA-TNEWFRIYK-IPDGKPENQFAFSGEAKNKKYALDIIKETHDSWKQL DVEKHLPGLLRA-TNEWFRIYK-IPDGKPENQFAFSGEAKNKKYTLDVIRECNEAWKKLI

SPO SCE2 BVR

DVERHMPGLIRA-TNEWFRIYK-IPDGKPENSFAFSGECKNRKYAEEVVRECNEAWERLI KIEEYFPGILDT-TREWFRKYK-VPAGKPLNSFAFHEQYQNSNKTIQTIKKCHNSWKNLI

ATH ECO PS3 CONSERVED

DVKRLKPGYLEA-TVEWFRRYK-VPDGKPENEFAFNAEFKDKNFAIDIIESTHEYWRALV DIKE-LPPHRLAEIRRFFEDYKKNE-NKKVDVEAFLPAQAAIDAIKDSMSLYELTSKLAC DVND-LPELLKAQIAHFFEHYKDLEKGKWVKVEGWENAEAAKAEIVASFER-AKNK SIED-LPQHKLKEIAHFFERYKDLQ-GKRTEIQTWEGPEAAAKLIDECIARYNEQK P

F

YK

240 SCEI KLA SPO SCE2 BVR ATH

I

K 260

280

AGKSSDSKGIDLTNVTLPDTPTYS-KAASDAI---PPASLKADAPIDKSIDKWFFISGSV SGKSADAKKIDLTNTTLSDTATYSAEAASAVP---AANVLPDE-PIDKSIDKWFFISGSA TGKTDAKSDFSLVNVSVTGSVANDPSVSSTIP---PAQELAPAP-VDPSVHKWFYISGSPL SGSLQEKYD-NLPNTERAGNGV--TLEDSVKPPSQ

.......... IPPEVQKWYYV

TKKT-DGKGISCMNTTVSESPFQCDPDAAKAIVDALPPPCESACTIPTDVDKWFHHQKN NANEETSPFPFLPVCLDITEAAFYTTCMLDKISIGAFNFVMLIRKHC

Figure 1 Alignment of PPase sequences. Numbering is for the sequence of SCE1-PPase7. Other sequences shown are for PPases from Kluyveromyces lactis 8 (KLA), S. pombe 9 (SPO), the mitochondria of S. cerevisiae 4 (SCE2), bovine retina (BVR)1°, Arabidopsis thaliana ~ (ATH), E. coil 12 (ECO), and the thermophilic bacterium PS-313. Conserved residues are indicated. Those underlined have been placed at the active site of SCE1-PPase~4-16.

(A. Goldman, pers. commun.) that diffract to 2.8A and their structure is currently being solved. It will be interesting to see how closely the overall structures of ECO-PPase and SCE1PPase overlap in view of the conservation of putative active site residues between the two enzymes.

Site-specific mutagenesis of ECO-PPase Site-specific mutagenesis of the relevant amino acids in ECO-PPase has demonstrated that the conserved activesite residues are in fact important for enzymatic activity (Table 1). Of the 17 polar active site residues identified in the X-ray structure of SCE1-PPase,

263

TIBS 17 - JULY 1992

mutation of all but three O'51, El01 and K148) of the identical residues in ECOPPase leads to substantial losses of activity, even whe~ the mutations are conservative (D~E, K-*R). Interestingly, and supporting the notion that it is largely or exclusively only the conserved residues that are important for enzymatic activity, Y-*F mutation of each of five tyrosines in ECOPPase that are not conserved in other PPases (Y16, Y30, Y57, Y77 and Yl17) leads to little or no loss (0-20%) in enzymatic activity. By contrast, modest to large losses (36-93%) of activity are seen on Y--*Fmutation of the conserved residues, Y55, Y141 and YS1. Similarly, none of the five histidine residues in ECO-PPase is conserved, and mutation of each to glutamine yields three variants (H60Q, Hll0Q and Hll9Q) that retain full activity, a fourth (H136Q) showing modest loss of activity (to 70%), while the fifth (H140Q) shows a substantial loss in activity (to 25%). However, even in this case, the nonconservative nature of the mutation

catalysis. Parallel mutagenesis studies on SCE1-PPase are also underway.

Mn2 \ 4.2 A

3.5 A/

SCE1-PPase catalysis in the presence

M n 3 - - 5.3 A - - - M n l Rgure 2 Mn2+-Mn2+ distances in the Mn3Pi2 complex of SCE1-PPase16.

makes it clear that H140 has no essential role in catalysis. The data in Table I provide a measure of maximal velocity at a single pH value. Other experiments, currently underway, are directed at fuller characterization of the most interesting variants listed in Table I, with respect both to catalytic properties and to enzyme structure. The results of such experiments should lead to a clearer understanding of the roles of the most important active-site residues in PPase

Table I. Enzymatic activity of variant ECO-PPases Aligned residue@ SCE1-PPase

ECO-PPase

E48 K56 d E58 R78d Y89 Y93 Dl15 Dl17 D120 D147

E20 K29d E31 R43 Y51 Y55 D65 D67 D70 D97

E148 E150d D152

E98 GIO0 D102

K154

K104

Y192 K193 K198

ECO-PPase variants

Relative PPase activity (%) of ECO-PPasevariantsb

Y141 K142 K148

E20D K29R E31D R43K Y51F Y55F D65E D67E D70E D97E D97V E98V EIOID e DIO2E DIO2V KIO4R K1041 Y141F K142R K148R

16 2 6 10 64 7 6 1 0 22 0 33 110 3 0 3 0 22 17 100

Non¢onserved Tyrin ECO-PPase

F43 L57 A95 T127 L168

Y16 Y30 Y57 Y77 Yl17

Y16F Y30F Y57F Y77F Y117F

100 80 100 100 100

Other conserved ~ polar residues ~ -

K61 D71

K34 D42

K34R

17

Putative active site residuesc

-

~_

a See Rg. 1. b Cloned wild-type ECO-PPasehas 100°£ activity. Values presented correspond essentially to relative/

Evolutionary conservation of the active site of soluble inorganic pyrophosphatase.

Soluble inorganic pyrophosphatases (PPases) are essential enzymes that are important for controlling the cellular levels of inorganic pyrophosphate (P...
623KB Sizes 0 Downloads 0 Views