Identification of residues critical for activity of the wound-induced leucine aminopeptidase (LAP-A) of tomato Yong-Qiang Gu* and Linda L. Walling Department of Botany and Plant Sciences, University of California, Riverside, CA, USA The importance of two putative Zn 2+ -binding (Asp347, Glu429) and two catalytic (Arg431, Lys354) residues in the tomato leucine aminopeptidase (LAP-A) function was tested. T he i mpact of substitutions at these positions, corresponding to the bovine LAP residues A sp255, Glu334, Arg336, and Lys262, was e valuated in His 6 –LAP-A fusion proteins expressed in Escherichia coli. S ixty-five percent of the mutant H is 6 –LAP-A proteins were unstable or h ad complete or partial defects in hexamer assembly or stability. The activity of hexameric His 6 –LAP-As on Xaa-Leu and Leu-Xaa dipeptides was tested. Most substitutions of Lys354 (a catalytic residue) resulted in His 6 –LAP-As that cleaved d ipeptides at slower rates. The Glu429 mutants ( a Zn 2+ -binding residue) had more diverse phenotypes. Some mutations abolished activity and others retained partial o r complete activity. The E429D His 6 –LAP-A enz yme had K m and k cat values similar to the wild-type His 6 –LAP-A. One catalytic (Arg431) and one Zn-binding (Asp347) residue were essential f or His 6 –LAP-A activity, as most R431 and D347 mutant His 6 –LAP-As did not hydrolyze dipeptides. The R431K His 6 –LAP-A that retained the positive c harge had pa rtial ac tivity as reflected in the 4.8 -fold d ecrease in k cat . Surprisingly, while the D347E mutant (that retained a neg- ative charge a t position 347) was inactive, the D347R mutant that introduced a positive charge retained partial activity. A model to explain these d ata is proposed. Keywords: aminopeptidases; wounding; d efense response; proteolysis; exopeptidases. Aminopeptidases catalyze the hydrolysis of N-terminal amino-acid residues from peptides and proteins. B ased on peptide sequences and enzymatic properties, hexameric leucine aminopeptidases (LAP; EC 3.4.11.1) are present in plants, animals, and prokaryotes [1–4]. Most animals, prokaryotes and non-Solanaceous plants express a single class of LAP protomer [5–7]. In contrast, there are two classes of LAP enzymes (LAP-A and LAP-N) in tomato with distinct biochemical characters and responses to devel- opmental and environmental cues [4,7–11]. LAP-N has 55-kDa protomers w ith a neutral pI. LAP-N levels do not change during plant growth and development or in response to environmental stress (C. J. Tu & L. L. Walling, unpublished results, [ 7,9]). While LAP-N is detected in all plants examined, LAP-A is detected only in a subset of the Solanaceae (N. Ly & L. L . Walling, unpublished results, [7,12]). LapA mRNAs, proteins and activities are detected during floral and fruit development, but not in vegetative organs of the healthy plants [8,11]. LapA is expressed in leaves in response to several environmental stresses inclu- ding water deficit, salinity stress, caterpillar feeding, and mechanical wounding [8,13,14]. The changes in LapA RNA and protein levels after treatments with systemin, jasmonic acid, and abscisic acid implicate the wound octadecanoid pathway in LapA regulation [8,15,16]. The tomato LAP-A was biochemically characterized recently and compared to the porcine LAP and Escherichia coli LAP (XerB or PepA) [4,17]. The LAPs from the three kingdoms are similar, as all three LAP enzymes are hexameric metallopeptidases that have high temperature and pH optima and are inhibited by the potent amino- peptidase inhibitors amastatin and bestatin [3,4]. Analysis of amino acyl-p-nitroanilide and -b-naphthylamide chromo- genic substrates showed that all three enzymes efficiently hydrolyze substrates with N-terminal Leu, Arg, and Met residues [4]. Studies with dipeptides and tripeptides demon- strated that, in general, the plant, animal and prokaryotic LAPs prefer to hydrolyze nonpolar aliphatic (Leu, Val, Ile, and Ala), basic (Arg) and sulfur-containing (Met) residues [4,17]. P eptide substrates with N-terminal ( P1) Asp o r Gly residues were cleaved inefficiently. When substrates with N-terminal aromatic residues w ere evaluated, significant differences in the plant, E. coli , a nd porcine LAPs were noted [17]. Phe and Tyr in the P1 position of dipeptides promoted dipeptide cleavage by the porcine LAP and E. coli PepA, but reduced tomato LAP-A hydrolysis r ates by 65 and 83%, respectively. Reciprocally, the bulkier aromatic residue Trp impaired dipeptide cleavage by the porcine and prokaryotic enzymes more than the tomato LAP-A. The penultimate residue (P1¢) strongly influenced LAP-A hydrolysis of dipeptides and the magnitude of its effect was dependent on the P1 residue [17]. P1 ¢ Pro, Asp, Correspondence to L. L. Walling, Department of Botany and P lant Sciences, University of California, Riverside, CA 92521–0124, USA. Fax: + 909 787 4437, Tel.: + 909 787 4687, E-mail: lw alling@citrus.ucr.edu Abbreviations: LAP, leucine aminopeptidase; blLAP, bovine lens LAP; IPTG, isopropyl thio-b- D -galactoside; Leu-p-NA, L -leucine-p-nitroanilide; TNBS, 2,4,6-trinitrobenzene sulfonic acid; Xaa, variable amino-acid residue. *Presen t address : Crop Improvement and Utilization Research Unit, Western Regional Research Center, ARS, U SDA, Albany, CA 94710, USA. Note: web page available at http//cnas.ucr.edu/bps/walling.html (Received 29 A ugust 2001, revised 16 January 2002, accepted 18 January 2002) Eur. J. Biochem. 269, 1630–1640 (2002) Ó FEBS 2002 Lys and Gly slowed the hydrolysis rates of the tomato LAP-A, porcine LAP, and E. coli PepA markedly. An alysis of tripeptides showed that more diversity was tolerated in the P2¢ position [17]. Given the similar properties of the plant, animal and bacterial LAPs, it is no t surprising that there is substantial identity in the primary sequence o f these LAPs [3,13]. The bovine lens LAP (blLAP) is well-characterized biochemi- cally [1,3]. X-ray crystal structures of the native enzyme and the blLAP complexed with aminopeptidase inhibitors (amastatin or bestatin) and transition state analogues ( L -leucine phosphonic acid or L -leucinal) have been studied [3,18–22]. More recently, X-ray crystal structures of the native E. coli PepA were resolved [23,24]. Based on the bovine a nd E. coli data, several mechanisms for LAP action on substrates have been proposed and specific residues have been implicated in substrate and metal binding and i n catalysis. Each LAP subunit h as two domains. T he larger lobe of the LAP protomer binds two z inc ions in a nonequ ivalent fashion [25]. The Zn ions are located at the edge of an eight- stranded, saddle-shaped b shee t. In the blLAP, the freely exchanging zinc ion (Zn1; Zn488) is coordinated by the carboxylate oxygens of Asp255, Asp332 and Glu334, and the carbonyl oxygen of Asp332. The second zinc ion (Zn2; Zn489) is more deeply imbedded in the LAP protomer and exchanges divalent cations more slowly. The carboxylate oxygens of Asp255, Asp273, Glu334, and backbone amino group of Lys250 [1]. Analysis of X-ray crystal structures for the bovine and prokaryotic LAPs and a limited mutational analysis of the E. coli PepA have suggested Lys262 and Arg336 have a role in catalysis [22,23]. The residues implicated in zinc binding and catalysis for blLAP are co nserved in the tomato LAP-A. For this reason, a site-directed mutagenesis strategy was pursued to charac- terize the active site in the tomato LAP-A. Four residues (Asp347, Glu429, Arg431, and Lys354) in the tomato LAP-A that corresponded to blLAP residues implicated in metal coordination (Asp255, Glu334) and catalysis (Arg336, Lys262) were targeted for study. Seven to nine substitutions were made at each position. The impact of each residue on LAP-A protein stability and LAP-A assembly into a hexamer were assessed. To determine if substitutions altered the ability of L AP-A to hydrolyze substrates with varying P 1 and P1¢ residues, the hydrolysis of 10 Xaa-Leu and 9 Leu-Xaa dipeptide substrates by wild- type and mutant LAP-A enzymes was determined. Meas- urements of K mand k cat for two muta nt His 6 –LAP-As (R431K and E335D) were compared with values for the wild-type His 6 –LAP-A. MATERIALS AND METHODS Site-directed mutagenesis The LapA1 cDNA clone (pBLapA-M) s pans the entire LapA1 coding region for the mature LAP-A1 peptide (corresponding to LAP-A residues 54–571) [4]. pBLapA-M was digested with BamHI and HindIII and the 1.6-kb LapA fragment was cloned into BamHI/HindIII-digested pUC119 (pULAP-M). Site-directed mutagenesis was performed according to Kunkel et al. [26] using the Muta-Gene Kit (Bio-Rad, Hercules, CA, USA). Most mutants were created by using LapA oligonucleotide primers, which replaced the codon of interest with random nucleotides (NNN) (Table 1) . Over 4 0 recombinant clones p er mutagenized residue were screened for m utations using LapA1 gene- specific primers and dideoxy-chain-termination DNA sequencing. A random distribution of mutations was not obtained. Therefo re, several specific oligonucleotide primers were designed to generate additional mutations of interest (Table 1). Overexpression and purification of wild-type His 6 –LAP-A and mutant His 6 –LAP-As The wild-type and mutant LAP-A1 proteins were over- expressed in E. coli as fusion proteins with six N-terminal Table 1. Oligonucleotides used to create substitutions at tomato LAP-A residues D347, K354, E429, and R431. Th e bovine and E. coli LAP residues implicated in catalysis and zinc ion binding are described by Stra ¨ ter et al. [22,24]. Corresp onding residues from the tomato LAP-A1 p rotein were identified. Oligonucleotides used for L apA1 mutagenesis are shown, where N ¼ A , T, G, or C. The mutagenized codon i s underlined. Predicted role in LAP LAP residue Bovine E. coli Tomato Oligonucleotide sequence Mutants Zn-binding D255 D275 D347 5¢ GGATTAACTTTT NNNAGTGGTGGCTAC 3¢ 5¢ GGATTAACTTTT CGCAGTGGTGGCTAC 3¢ 5¢ GGATTAACTTTT AACAGTGGTGGCTAC 3¢ D347A, D347I, D347V, D347S, D347G, D347Y, D347E D347R D347N Catalysis K262 K282 K354 5¢ GGCTACAACCTC NNNGTCGGAGCTCGT 3¢ K354M, K354G, K354T, K354W, K354N, K354E, K354R Zn-binding E334 E354 E429 5¢ CAATACTGATGCT NNNGGTAGGCTCACA 3¢ 5¢ CAATACTGATGCT GACGGTAGGCTCACA 3¢ 5¢ CAATACTGATGCT AGCGGTAGGCTCACA 3¢ E429A, E429V, E429G, E429W, E429Q, E429R E429D E429S Catalysis R336 R356 R431 5¢ TGATGCTGAGGGT NNNCTCACACTTGC 3¢ 5¢ TGATGCTGAGGGT CAGCTCACACTTGC 3¢ R431A, R431V, R431G, R431W, R431E, R431K R431Q Ó FEBS 2002 Site-directed mutagenesis of the tomato LAP-A (Eur. J. Biochem. 269) 1631 His residues. pQLapA-M expresses the mature LAP-A1 protein as a His 6 –LAP-A protein (wild-type His 6 –LAP-A) in E. coli and was previously described [4]. LapA cDNA inserts with mutations (above) were excised from t he pULapA-M clones by digestion with BamHI and HindIII, cloned into BamHI/HindIII-digested pQE11 (Qiagen, Chatsworth, CA) and transformed into E. coli JM109 to generate the pQ series of mutant His 6 –LAP-A clones. The methods to over-express and purify the wild-type and mutan t His 6 –LAP-A proteins by Ni/nitrilotriacetic acid resin columns (Qiagen) have been described [17]. A liquots (3 lL) of each 1-mL Ni-nitrilotriacetic acid column fraction were separated by SDS/PAGE and stained with Coomassie Brilliant Blue. Aliquots were also assayed for LAP ac tivity using t he chromogenic substrate L -Leu-p-nitroanilide (Leu-p-NA) [4]. His 6 –LAP-A multimeric complexes were evaluated by native PAGE. Enzyme activity assays The activities of the wild-type and mutant His 6 –LAP-A enzymes on peptide substrates were determined using the assay d escribed by Mikkonen [27] with a few modifications described by Gu & Walling [4]. The amino acids liberated by the LAP enzymes were quantitated with 2,4,6-trinitroben- zene sulfonic acid (TNBS; Sigma, St Louis, MO, USA) reagent containing cupric ions to block the reaction with the peptide amino groups [28]. The substrate solution contained 6.25 m M dipeptide in 42 m M sodium carbonate (pH 9 .2). In a typical assay, 100 lL of substrate solution was mixed with 10 lI of wild-type o r mutant His 6 –LAP-A enzyme (2.5 ngÆlL )1 )and15lLof5m M MnCl 2 . The mixture was incubated at 37 °C for 30 min, and the reaction was stopped by adding 3 mL of fresh TNBS reagent (65 m M sodium tetraborate, 0.8 m M cupric sulfate, 0.96 m M TNBS). After a 30-min incubation at room temperature, the absorbance of TNB/amino-acid conjugates was measured at A 420 .AstheA 420 of each TNB/amino-acid conjugate was distinct, standard curves for each amino acid were deter- mined using concentrations between 0.5 and 5.0 m M .A correlation coefficient was calculated for each peptide to determine t he moles of peptide hydrolyzed. Relative rates of hydrolysis were corrected for the percentage of enzyme in the active h exameric form, w hich was determined f rom scanned gel images and image analysis using ALPHAIMAGE software (Alpha Innotech Corporation, San Leandro, CA, USA). All dipeptides were in t he L -configuration. Dipeptides were ordered from Sigma Chemical Co. or Bachem (Torrance, CA, USA). Protein concentrations were deter- mined by the Bradford assay [29] with bovine serum albumin as a standard. To facilitate discussion of the specificity of the LAP enzymes, the nomenclature initially developed by Schechter and Berger [30] and more recently described b y B arrett [31] was u tilized. Aminopeptidases cleave between the N-terminal residue (P1) and the penul- timate residue (P1¢). The C-terminal residue in tripeptides is in the P2¢ position. The K m and k cat values of wildtype and mutant His 6 – LAP-As were determined from the initial rat es of hydrolysis of Leu-Gly at concentrations ranging from 0.5 to 8.0 m M at optimum pH (9.2). A nonlinear least-squares r egression of enzyme kinetics was used to deter mine K m and k cat [32]. The k cat value was based on (a) the masses o f the wild-type (55 6 90 Da), E429D (55 743 Da), and R431K (55 666 Da) His 6 –LAP-As, (b) the fact that all enzyme p reparations were greater than 95% pure, and (c) the percentage of His 6 – LAP-A in hexameric form. Enzyme purity and assembly (see above) was based on staining of SDS or native PAGE gels, respectively. SDS and native PAGE and immunoblot analyses SDS/PAGE was carried out according to the method of Laemmli [33] using a 10% polyacrylamide resolving gel and a 4% stacking gel. Native P AGE was performed according to Gu et al. [ 9] using a 7.5% polyacrylamide resolving gel and a 4% stacking gel. Gels were stained with Coomassie Brilliant Blue in methanol/acetic acid/water (40 : 10 : 50, v/v/v) a nd destained in methanol/acetic acid/water (40 : 10 : 50, v/v/v). Immunoblots were p erformed as described by Gu et al. [ 9]. The tomato LAP-A polyclonal antiserum does not significantly cross react with E. coli proteins [4]. Sequence alignments The tomato LAP-A1 (LeLAP-A1; U50151) peptide sequence was previously reported [34] and was aligned w ith plant, prokaryotic and animal LAPs using the PILEUP program from the Wisconsin Genetics Group. The Solanum tuberosum (potato) LAP (StLAP; X67845) [12], Arabidopsis thaliana LAP ( AtLAP; P30184) [35], Petroselinum crispum (parsley) LAP (PcLAP; X99825), E. coli PepA (EcPepA; P11648) [36], Rickettsia prowazekii PepA (RpPepA; P27888) [37], human LAP (HsLAP; AAD17527), and bovine mature, kidney LAP (BtLAP; 1 LCPA) [22,38] were included in this study. The RASMOL program (Berkeley version) and the blLAP– leucinal complex (1lan) was used to determine distances between atoms in blLAP [22]. Distances from the C a atoms of Asp255, Arg336, Lys262 and Glu334 to side-chain residues, Zn ions or leucinal atoms were determined. In addition, the average distance between the C a of Glu and its carboxylate oxygen, C a of Asp a nd its carboxylate oxygen, C a of Arg and its side-chain amines or guanidinium, and C a of Lys and its side-chain amine was determined by measuring distances in the blLAP. These average distances were dete rmined by measuring atomic distances in 35 Glu residues, 23 A sp residues, 23 Arg r esidues, and 26 Lys residues in the blLAP-leucinal X-ray crystal structure [22]. The distances between the C a of Asp255 and Asp255 carboxylate oxygens, Zn1, Zn2, and C-terminal oxygens of leucinal were previously reported. RESULTS LAPs have a highly conserved carboxyl domains The leucine aminopeptidases of eukaryotic and p rokaryotic organisms shared a high degree of primary sequence identity in the carboxyl region. Alignment of C-terminal domain of eight LAPs is shown in Fig. 1. The region spanning the tomato LAP-A residues 323–571 had between 76 and 92% identity with the Arabidopsis, potato, and parsley LAPs. Identities with nonplant LAPs (bovine, human, E. coli ,and Rickettsia)rangedfrom41to46%.Thealignmentofeight 1632 Y Q. Gu and L. L. Walling (Eur. J. Biochem. 269) Ó FEBS 2002 LAPs in this 148-residue region showed a strict conservation of residues i mplicated in zinc-ion binding and catalysis (Fig. 1) [1]. Futhermore, extended regions of sequence identity were also seen. T he shading in Fig. 1 underesti- mates the degree of similarity between these enzymes, as only residues c onserved in a ll or 5/8 of the different LAPs were highlighted. Sixty-five of the 148 residues were identical in all e ight LAPs (43.9% identity). Most of these residues were clustered into nine regions of identity (I–IX). Inspec- tion of these regions showed that residues implicated in Zn 2+ binding and catalysis were imbedded in the highly conserved regions I, II and IV. Ten a dditional residues located in hydrophobic pockets and clefts have been postulated to be important in van der Waal interactions or hydrogen bonding with the aminopeptidase inhibitors, bestatin and amastatin [39,40]. These residues were located in regions II, IV, V, VII, VIII and IX. While many of these residues were invariant, some were changes were noncon- servation substitutions changing the charge or polarity of residues (Fig. 1). Finally, while regions III and VI were highly conserved, the role(s) of these conserved residues in LAP structure and/or function is not presently known. Over-expression and purification of wild-type and mutant His 6 –LAP-A proteins The c ompelling sequence identities of the bovine and prokaryotic LAPs suggested that the plant LAP may use a c atalytic mechanism similar t o that employed by the animal and bacterial LAPs [22,24]. Using site-directed mutagenesis, a series of LAP-A1 mutants were generated (Table 1). The tomato residues with putative roles in Zn 2+ binding (Asp347 and Glu429) and catalysis (Lys354 and Arg431) were substituted with residues that had the same or a different charge, were isosteric, or had different degrees of hydrophobicity or hydrophilicity. The tomato LAP-A1 residue designations are used throughout. The analogous bovine and E. coli LAP residues are identified b y subtract- ing 92 or 112, respectively, from the tomato LAP-A1 residue number. Previous studies showed that the FPLC-purified tomato LAP-A1 and the His 6 –LAP-A have similar biochemical properties [17]. Therefore, wild-type and 31 mutant LAP-A proteins were expressed in E. coli as His 6 /fusion proteins. After IPTG induction of bacterial cultures, total proteins were isolated, purified from soluble protein fractions using nickel column chromatography, and fractionated by SDS/ PAGE. Coomassie Blue-stained gels and immunoblot analyses using a tomato LAP-A antiserum [9] revealed that high levels of soluble His 6 –LAP-A proteins were produced in most bacterial lysates (data not shown). F or most His 6 – LAP-A clones, an abundant 55-kDa His 6 –LAP-A protein was isolated by affinity chromatography (Fig. 2). Four mutations of His 6 –LAP-A resulted in unstable His 6 –LAP-A proteins. In the K354T mutant strain, both the 55-kDa His 6 –LAP-A protein a nd a 40-kDa degradation product were detected (Fig. 2B). The replacement of Arg431 with glycine (R431G) influenced the stability of the His 6 –LAP-A proteins, because only a stable 28-kDa His 6 –LAP-A polypeptide was purified by nickel column chromatography (Fig. 2D). Finally, the D347A and R431E substitutions resulted in rapid ly degraded His 6 –LAP-A proteins that were not detected in E. coli crude lysates or after affinity chromatography (data not shown). Quaternary structure of wild-type LAP-A and mutant His 6 –LAP-As The plant, a nimal a nd prokaryotic LAPs have multimeric structures [9,41,42]. Similar to the bovine lens LAP, the tomato wound-induced LAP-A is a hexameric enzyme composed of six identical 55-kDa subunits [9] and the multimeric structure is critical for the enzyme activity [4]. To investigate the ability of mutant His 6 –LAP-As to assemble into hexamers, purified wild-type and 29 mutant His 6 –LAP- As were subjected to native-PAGE and gels were stained with Coomassie Blue (Fig. 3). Although the purified wild-type His 6 –LAP-A and the mutant His 6 –LAP-As displayed in Fig. 3 had intact 55-kDa protomers (Fig. 2), the ability of mutant H is 6 –LAP-As to Fig. 1. Comparison of plant, prokaryotic and a nimal LAP peptide sequences spanning residues proposed to be involved in Zn 2+ binding and catalysis. The deduced amino -acid se quen ces for t he tom ato L AP-A1 (Le; residues 323–571), potato LAP (St; residues 304–554), Arabidopsis LAP (At; residues 269–520), parsley LAP (Pc; residues 44–295); E. coli PepA (Ec; residues 251–503), R. prowazekii (Rp; residues 247–500), human LAP (Hs; residues 263–519), and mature bovine-kidney LAP (Bt; residues 231–487) are shown. Sequence accessions are liste d in Materials and methods. Gaps in aligned peptide sequences (dots) were introduced to maximize similarities. Residues identical in all six LAP s are highlighted in black; residues conserved in five of the eight LAPs areshadedingrey.ResiduespredictedtohavearoleinZnionbinding (m)andcatalysis(d) are in dicated [1,24]. Residues po stulated to in- teract bestatin or amastatin are located in a hyd rophobic pocket or cleft and are indicated in open circles ( s)[39]. Ó FEBS 2002 Site-directed mutagenesis of the tomato LAP-A (Eur. J. Biochem. 269) 1633 form stable hexamers varied markedly (Fig. 3). Similar to previous studies, the wild-type His 6 –LAP-A assembled into a 357-kDa hexamer in E. coli and His 6 –LAP-A monomers were not detected after nickel column chromatography (Figs 2 and 3). Over 65% o f the His 6 –LAP-A mutants exhibited some impairment in hexamer assembly o r stabil- ity. Eleven of the His 6 –LAP-A substitution mutants (E429W, E429V E429D, E429S, D437G, D347R, K354R, R431V, R431Q, R431W and R431A) assembled into stable hexameric complexes. Two of the Asp347 mutant His 6 – LAP-As (D347Y and D347E) and five of the Lys354 mutan t His 6 –LAP-As (K354N, K354T, K354W, K354G, and K354M) exhibited more complex staining patterns. These mutant Hi s 6 –LAP-As were able t o assemble into hexamers, however, faster migrating complexes contribu- ting to different percentages of the purified His 6 –LAP-A protein were also visualized. These additional bands may represent dissociation products or assembly intermediates of the hexameric His 6 –LAP-A. The remaining 10 mutant His 6 –LAP-As (E429G, E429Q, E429R, E429A, D347I, D347S, D347V, D347N, K354E and R431G) did not form stable hexamers, for only the fast-migrating dissociation products were observed in native-PAGE gels (Fig. 3). Activity of wild-type and mutant His 6 –LAP-A enzymes on Xaa-Leu dipeptides The mutan t His 6 –LAP-As that did not assemble into hexamers (Fig. 3) had < 0.5% activity on the Leu-Leu dipeptide compared t o the wild-type His 6 –LAP-A (data not shown). These data supported previous observations that only the hexameric tomato LAP-A is functional [4]. The remaining 19 mutant H is 6 –LAP-As that a ssembled into hexameric enzymes were useful for examining the impact of substitutions of active site residues on His 6 –LAP-A activity and specificity. The ability o f wild-type and mutant His 6 – LAP-As to hydrolyze peptides with 10 different N-terminal (P1) residues was evaluated by measuring the rate of Xaa-Leu p eptide hydrolysis (Table 2). The wild-type His 6 –LAP-A hydrolyzed Xaa-Leu substrates at different rates ranging from 46 (Tyr-Leu) to 419 lmolÆmin )1 Æ mg protein )1 (Leu-Leu) as previo usly observ ed [17]. Although all 19 mutants were analyzed, data for two or three representative D347, K354, E429, and R431 mutants are presented (Table 2). The aspartic acid residue at position 347 has a postulated role in coordinating both Zn1 and Zn2 (Table 1, [40]). Substitution of Asp347 with a s imilarly charged residue (glutamic acid; D347E) or a small nonpolar residue (glycine; D347G) abolished His 6 –LAP-A activity on all 10 Xaa-Leu dipeptides (Table 2). When Asp347 was r eplaced with the Fig. 2. SDS/PAGE fractionation of wild-type and mutant His 6 –LAP-A proteins over-expressed in E. coli. Total p roteins w ere isolated from E. c oli strains after IPTG induction. Proteins were fractionated by SDS/PAGE and stained with Coomassie Blue. E. coli strains expre s- sing the wild-type His 6 –LAP-A (Panel C) or mutant His 6 –LAP-A proteins with substitutions for the residues D347 (Panel A), K354 (Panel B), E429 (Panel C), and R431 (Panel D) are displayed. The His 6 –LAP-A protein sizes were determined by marker proteins (in kDa)runinaparallellane. Fig. 3. Native PAGE fractionation of wild-type and mutant His 6 –LAP- A proteins over-ex pressed in E. coli. Total proteins were isolated from E. c oli strains after IPTG induction. Proteins were fractionated by native-PAGE and stained with Coomassie Blue. E. coli strains expres- sing the wild-type or mutant His 6 –LAP-A proteins with substitutions for t he residues D347 (Panel A), K354 (P anel B), E429 (Panel C), and R431 (Panel D) are displayed. The mass of the wild-type His 6 –LAP-A was previously determined [17] and is shown at the left of each panel (357 kDa). 1634 Y Q. Gu and L. L. Walling (Eur. J. Biochem. 269) Ó FEBS 2002 positively charged arginine (D347R), the impact on LAP activity was variable. The D347R mutation abolished hydrolysis of the Trp-Leu, Pro-Leu and Tyr-Leu substrates. Surprisingly, between 0.6% (Phe-Leu) and 5.9% (Thr-Leu) of wild-type His 6 –LAP-A activity was observed with o ther Xaa-Leu dipeptide substrates. Glu429 is also proposed to have a role in coordination of both Zn1 and Zn2 in each LAP subunit ( Table 1 and [18]). E429 mutants had three phenotypes. Replacement of Glu429 with tryptophan (E429W) abolished His 6 –LAP-A activity on all Xaa-Leu dipeptides tested (Table 2). In contrast, t he E429V (Table 2) and E429S (data not shown) mutants retained partial activity on all Xaa-Leu dipeptides ranging from 0.7 to 2.7% of the wild-type His 6 –LAP-A activity. Finally, t he replace ment o f glutamic a cid with aspartic acid (E429D) yielded a His 6 –LAP-A enz yme that retained greater than 95% of its activity on the Leu-Leu peptide. Rates of hydrolysis of other Xaa-Leu dipeptides were not impaired or were greater than 79% of the wild- type ac tivity. Fig. 4. Model for the tomato LAP-A active site. The model of the tomato LAP-A active site is based on mechanisms proposed for the bovine LAP [22] a nd E. coli PepA [24]. Amino-acid residu e coordi- nates for the b ovine L AP and tomato LAP-A (in parentheses) a re shown. Zn1 is coordinated by ca rbonyl oxygens of Glu334 ()429), Asp255 ()347), and Asp332 ()427; very weakly), a backbone carbonyl of Asp322, and a water molecule. Zn2 is coo rdinated by the carbonyl oxygens of Glu334 ()429), Asp255 ()347), Asp273 ()366), the back- bone amino g roup of L ys250 ( )342), and a water molecule. The amino terminus of the P1 r esidue is teth ered by Z n2 and a c arboxyl oxygen of Asp273 ()366), while the p eptide’s carboxylate oxygen interacts with Zn1 and side-chain amine group of Lys262 ()354). The P1¢ backbone amino group interacts w ith the b ackbone carbonyl of Leu360 ()455), while a bicarbonate is mo de led i nto the active site, there is not known if water, bic arbonate or a bihydroxide ion a cts as t he general base in LAP-A. The bicarbonate may interact with the water spanning Zn1 and Zn2, and residues from Gly335 ()430), Arg336 ()431), and Leu360 ()445). Amino-acid residues of LAP and Zn ions are shaded for contrast. Bond lengths are not to scale. Table 2. Activity o f wild-type and m utant His 6 –LAP-A proteins on Xaa-Leu dipeptides. The r ate of Xaa-Leu dipeptide hydrolysis was determined and e xpressed as 100% of wild-type His 6 –LAP-A activity (± standard d eviation). T he hydrolysis rates for the wild-typ e H is 6 –LAP-A are shown in lmolÆmin )1 Æmg protein )1 . Rates indicated as < 0.5% of the wild-type His 6 –LAP-A values were at the limit of detection. Relative activities of D347E, K354M, and R431K d ata were corrected for percentage enzyme that was in a hexameric form (see Materials and methods); all other enzymes were 100% hexamer. His 6 –LAP-A enzyme % Wild-type His 6 –LAP-A activity on dipeptides Leu-Leu Ala-Leu Val-Leu Phe-Leu Trp-Leu Pro-Leu Met-Leu Thr-Leu Tyr-Leu His-Leu Wild-type 100 100 100 100 100 100 100 100 100 100 419.3 ± 30.9 320.6 ± 20.5 281.6 ± 8.9 131.3 ± 15.2 214.3 ± 6.7 153.4 ± 6.2 345.3 ± 9.5 261.3 ± 22 45.7 ± 5.9 150.9 ± 9.0 D347G <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 D347E <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 D347R 3.0 ± 1.0 1.0 ± 0.5 1.9 ± 0.2 0.6 ± 0.4 <0.5 <0.5 1.8 ± 0.5 5.9 ± 1.0 <0.5 0.7 ± 0.1 K354M 3.1 ± 0.6 <0.5 0.9 ± 0.3 0.9 ± 1.0 1.0 ± 0.3 <0.5 1.9 ± 0.4 7.4 ± 0.3 1.2 ± 0.3 1.8 ± 0.6 K354R 4.0 ± 0.8 1.8 ± 0.3 0.9 ± 0.5 2.0 ± 0.7 2.0 ± 0.7 <0.5 2.8 ± 0.5 7.8 ± 1.2 1.8 ± 0.8 1.6 ± 0.2 E429W <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 E429V 1.7 ± 0.7 2.6 ± 0.6 2.7 ± 0.3 1.1 ± 0.4 0.9 ± 0.7 1.5 ± 0.5 0.7 ± 0.3 0.9 ± 0.4 1.0 ± 0.2 2.0 ± 0.2 E429D 95.6 ± 1.9 87.1 ± 10.2 101.3 ± 6.9 84.6 ± 9.3 91.6 ± 9.6 86.2 ± 15 94.6 ± 10.2 105.7 ± 9.4 78.5 ± 5.7 81.4 ± 12 R431A <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 R431K 5.6 ± 0.8 2.6 ± 0.6 8.3 ± 0.2 5.0 ± 0.7 7.7 ± 1.0 1.3 ± 0.4 4.9 ± 0.4 11.3 ± 1.4 7.4 ± 0.8 4.9 ± 0.2 Ó FEBS 2002 Site-directed mutagenesis of the tomato LAP-A (Eur. J. Biochem. 269) 1635 The tomato LAP-A1 residues Lys354 and Arg431 may facilitate catalysis of substrates and/or stabilization of the gem-diolate reaction intermediate (Table 1, [22]). With the exception of the K354R mutant, most of the K354 substitution mutants had an impaired ability to form hexamers (Fig. 3B). However, none of the mutations completely abolished His 6 –LAP-A a ctivity on all Xaa-Leu substrates. For example, the K354M mutation abolished activity on Pro-Leu and Ala-Leu, but other Xaa-Leu dipeptide substrates were hydrolyzed at detectable rates 0.9– 7.4% of the wild-type His 6 –LAP-A levels) (Table 2). The replacement of Lys354 with the similarly charged Arg residue (K354R) c reated an enzyme w ith activities only slightly above the K354M mutant activities (Table 2). The Arg431 d ata contrasted with t he Lys354 mutant analyses. Replacement o f Arg431 with alanine (R431A) (Table 2) and o ther residues (Gly, Trp, Val, Gln) (data not shown) abolished LAP activity. However, substitution of Arg431 with lysine (R431K) created an enzyme that retained partial activity on L eu-Leu (5.6% of w ild-type His 6 –LAP-A activity). Similar hydrolysis rates were noted with other Xaa-Leu dipeptide substrates (Table 2). Activity of wild-type and mutant His 6 –LAP-A enzymes on Leu-Xaa dipeptides To determine if mutation o f active site residues a ltered the hydrolysis of peptides with different P1¢ residues, the activity of mutant His 6 –LAP-A enzymes on nine Leu-Xaa dipeptide substrates were tested (Table 3). The rates o f w ild-type His 6 –LAP-A hydrolysis of these peptides v aried 45-fold (Table 3). The impact of mutations in the residues (E429 and D 347) implicated in coordinating Zn 2+ on hydrolysis of Leu-Xaa peptides was similar to that seen for the Xaa- Leu peptides. The D347G, D347E, and E429W enzymes were inactive on all Leu-Xaa dipeptide substrates tested, while E429V exhibited a diminished but significant activity on each substrate (Table 3). The D347R enzyme retained the ability to hydrolyze eight of the nine Leu-Xaa substrates at significantly higher rates than other D347 mutants. However, D347R did not hydrolyze the Leu-Arg signifi- cantly. R etention of the negative charge at position 429 (E429D) created enzymes with near wild-type activity on Leu-Xaa peptides (Table 3). The hydrolysis of Leu-Asp was most strongly impaired in the E429D mutant. R431K enzyme that retained the positive charge at the catalytic site retained partial activity (3.6–13%) on Leu-Xaa peptides (Table 3). In contrast, the R431A His 6 –LAP-A was inactive. Unlike Arg431, the active site residue Lys354 accommodated several substitution s. All K 354 mutant enzymes had residual activity on the Leu-Xaa dipeptides (Table 3; data not shown). Relative to t he other L eu-Xaa dipeptides, a ll K354 mutant His 6 –LAP-A enzymes had an enhanced rate of cleavage of the Leu-Tyr peptide. For example, the K354R and K354M enzymes retained 14 and 11% wild-type His 6 –LAP-A activity levels, respectively (Table 3). Kinetics of hydrolysis of wild-type and mutant His 6 –LAP-A enzymes The kinetic properties o f E429D, R431K and wild-type His 6 –LAP-A enzymes using Leu-Gly as a substrate were Table 3. Activity of wild-type and mutant His 6 –LAP-A prote ins o n L eu-Xa a dipe ptides . T he r ate o f L eu-X aa dip eptide h yd rolysi s w as de termined and data i s e xpress ed a t 1 00% o f w ild-type His 6 –LAP-A activity (± SD). Hydrolysis rates for the wild-type His 6 –LAP-A are shown in lmolÆmin )1 Æmg protein )1 . Rates indicated as <0.5% of the wild-type His 6 –LAP-A were at the limit of detection. Relative activities of D347E, K354M, and R431K data were corrected for percentage enzyme that was in a hexameric form (see Materials and methods); all other enzymes were 100% hexamer. His 6 –LAP-A enzyme % Wild-type His 6 –LAP-A activity Leu-Leu Leu-Phe Leu-Met Leu-Ser Leu-Gly Leu-Tyr Leu-Asn Leu-Asp Leu-Arg Wild-type 100 100 100 100 100 100 100 100 100 419.3 ± 30.9 440.9 ± 29.5 325.4 ± 20.9 150.4 ± 19.3 160.0 ± 5.6 200.8 ± 10.2 141.6 ± 7.8 9.8 ± 2.1 112.3 ± 6.9 D347G <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 D347E <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 0.9 ± 0.2 <0.5 D347R 3.0 ± 1.0 2.7 ± 0.5 5.3 ± 0.3 2.8 ± 1.2 0.8 ± 0.3 5.3 ± 0.6 1.4 ± 1.0 0.7 ± 0.3 <0.5 K354M 3.1 ± 0.6 2.7 ± 1.3 1.8 ± 0.4 <0.5 1.6 ± 0.3 10.7 ± 1.0 1.5 ± 0.3 1.2 ± 0.2 1.3 ± 0.9 K354R 4.0 ± 0.8 5.7 ± 0.2 2.3 ± 0.4 2.2 ± 0.4 1.2 ± 0.6 14.2 ± 1.5 3.7 ± 0.9 0.7 ± 0.1 2.9 ± 0.9 E429W <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 0.6 ± 0.2 <0.5 E429V 1.7 ± 0.7 1.1 ± 0.3 2.9 ± 0.1 1.3 ± 0.1 1.0 ± 1.0 4.5 ± 0.4 4.4 ± 0.6 0.8 ± 0.2 4.0 ± 0.2 E429D 95.6 ± 1.9 107.0 ± 4.5 87.7 ± 6.2 89.9 ± 2.4 83.6 ± 5.9 92.5 ± 7.0 84.6 ± 4.1 52.7 ± 7.1 92.4 ± 5.1 R431A <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 R431K 5.6 ± 0.8 3.6 ± 0.4 6.0 ± 0.5 4.1 ± 0.2 12.5 ± 0.8 11.1 ± 1.7 11.0 ± 1.0 7.5 ± 1.0 6.2 ± 0.6 1636 Y Q. Gu and L. L. Walling (Eur. J. Biochem. 269) Ó FEBS 2002 determined. The E429D enzyme hydrolyzed both Xaa-Leu and L eu-Xaa pept ides at 52–107% of wild-type H is 6 –LAP-A rates (Tables 2,3). Consistent with these observations, the E429D His 6 –LAP-A and wild-type His 6 –LAP-A had a similar K m values of 2.0 and 1.8 m M , respectively. In addition, the turnover constant (k cat ) was similar for the E429D His 6 –LAP-A (108 s )1 ) and wild-type His 6 –LAP-A enzymes (120 s )1 ). Unlike other Arg431 substitution mutants, the R431K enzyme retained low levels of activity on all dipeptides with 13% of the w ild-type activity on Leu-G ly. This change in hydrolytic activity was correlated with a 2.9-fold change in the K m of the R431K His 6 –LAP-A (5.1 m M )relativeto wild-type His 6 –LAP-A. In addition, the R431K enzyme had a fourfold reduction in the turnover k cat to 30 s )1 . DISCUSSION Intensive biochemical and structural studies of the bovine lens LAP a nd E. coli PepA provided a solid foundation for the study into the residues critical of the activity of the wound-induced tomato LAP-A1 [6,22,24]. Similar to the animal and prokaryotic enzymes, LAP-A1 is a homo- hexameric enzyme [9,17]. The alignment of 148 residues from the C-terminal domain of seven different LAPs with the tomato LAP-A1 showed a high degree of s equence identity. In addition to the strict conservation of residues implicated in Zn ion binding and catalysis [1], more extended regions of sequence identity were seen (regions I–IX). The residues implicated in Zn 2+ binding, catalysis and interactions with aminopeptidase inhibitors were located in seven of these regions (I, II, IV, V, VII, VIII, and IX) [39,40]. The remaining conserved domains (regions III and VI) make unknown contributions to structure and/or function of the LAP enzymes. The compelling sequence identities in LAP enzymes suggested that the tomato LAP-A may u se a c atalytic mechanism similar to that employed by the animal and bacterial LAPs [22,24]. To test this hypothesis, four residues in the tomato LAP-A1 protein with a potential roles in Zn i on binding (Asp347 and Glu429) and catalysis (Lys354 and Arg431) were subjected to site-directed mut- agenesis. These r esidues correlated to the bovine lens LAP (blLAP) residues Asp255, Glu334, L ys262, a nd Arg3 36 (Table 1) . Thirty-one mutant enzymes were purified and analyzed. A large proportion (65%) of the mutant enzymes was inherently unstable or had partial or complete defects in hexamer asse mbly or stability. With the se H is 6 –LAP-A preparations, no His 6 –LAP-A activity was detected. These data furth er supported the observation th at the hexameric structure of LAP-A was essential for its enzymatic activity [4,9]. Several of the Asp347 mutants (D347Y and D347E) and Lys354 mutants (K354N, K354T, K354W, K354G, and K354M) assembled hexameric His 6 –LAP-A enzymes, however, faster migrating complexes were also visualized. The faster migrating bands may represent a dissociation product or an assembly intermediate of hexameric His 6 – LAP-A, as the quantity of the His 6 –LAP-A hexamer was reduced in the strains expressing these His 6 –LAP-A forms. These data stress that documentation of the ability of mutant LAP protomers to assemble into a hexameric structure is critical prior to evaluating impact of a mutation on activity levels. Inspection of the active site residue replacements that caused changes in the conformation of the tomato L AP-A enzyme did not reveal any correlation with the n ature of t he residue in the mutant His 6 –LAP-A enzyme. X-ray structures of the unliganded blLAP and E. coli PepA, and of the blLAP complexed with bestatin, amastatin, L -leucine phosphonic acid, and L -leucinal have been resolved [18–24,39]. These studies also identified the residues important for LAP catalysis and coordination of the two zinc io ns in the a ctive site. The i mpact of the mutations on the tomato LAP enzyme must be viewed in the context o f these data. The most recent mechanism proposed for blLAP and E. coli PepA action is briefly reviewed using the residue designations for the blLAP (Table 1, [22,2 4]); the tomato LAP-A1 residues studied appear in parentheses. A model of the reactive site of the tomato LAP based on the bovine LAP and E. coli PepA appears in Fig. 4. When a peptide substrate is bound by blLAP, both Zn1 (Zn488) and Zn2 (Zn489) are bound by six ligands to form imperfect, octahedral coordination geometries. Zn2 is coordinated by carboxyl oxygens from Glu334 (Glu429), Asp273 (Asp366), and Asp255 (Asp347), the backbone amino group of Lys250 (Lys342), and the N-terminal amine of the P1 residue of the peptide substrate. A water molecule that bridges Zn1 and Zn2 is the sixth ligand. In addition to the zinc-bridging water molecule and the carbonyl oxygen of the P1 residue, Zn1 is bound by the two carboxyl oxygens of Asp25 5 (Asp347) and a car boxyl ox ygen of Glu334 (Glu429). T he interact ions of Zn1 and the carboxyl oxygen of Asp332 (Asp427) are weak. The peptide is further stabilized at the active site by the interaction of t he N-terminal amine of t he P1 residue with a c arboxylate oxygen of Asp273 (Asp366), the carbonyl oxygen of the P1 residue with the side-chain amine of Lys262 (Lys354), and the backbone amino group of the P 1¢ residue with the backbone carbonyl oxygen of Leu360 (Leu455). There is no proteineous residue in the blLAP active site that can act as a nucleophile. Therefore, the zinc-bridging water is proposed to have this function. Early studies suggested a role for waters and the bihydroxide ion in the general base mechanism of blLAP [21,24]. The recent identification of bicarbonate in the active site of the native E. coli PepA and blLAP and the influence of bicarbonate on PepA hydrolysis of a chromogenic substrate [24] suggests that bicarbonate may act as a g eneral base. Bicarbonate (or bihydroxide ion) may accept a proton from the zinc- bridging water and shuttle t his p roton t o the leaving group (the N-terminal P1¢ amine) [24]. While the mechanism for generating the nucleophilic hydroxide from the zinc- bridging water differs in the models (with bicarbonate or three waters at the active site), the subsequent steps t o form and resolve the gem-diolate intermediate are similar [22,24]. The nucleophilic hydroxide ion attacks the carbon at the scissile bond to form a gem-diolate intermediate. This intermediate is stabilized by continued interaction with Zn1, Zn2, residues f rom Lys262 ( Lys354), Asp273 (Asp366), Leu360 (Leu455), and the bicarbonate ion ( or bihydroxide ion). If analogous to blLAP, the tomato Arg431 (blLAP Arg336) will bind the bicarbonate ion at the active site [24]. Arg431 was essential for His 6 –LAP-A activity on dipeptide Ó FEBS 2002 Site-directed mutagenesis of the tomato LAP-A (Eur. J. Biochem. 269) 1637 substrates. Only the R431K mutant that retained the positive charge at t he active site had partial activity. These observations were supported by kinetic measurements using the Leu-Gly dipeptide. The K m of the R431K enzyme was threefold higher and the k cat decreased fourfold rela tive to the wild-type enzyme. The analogous mutation in the E. coli PepA (R356K) reduced the turnover number by 15-fold [24]. These data suggested that side-chain amine of Lys was positioned to partially substitute for the functions of the Arg guanidinium n itrogen and/or terminal amines. This less optimal Lys side chain interaction with bicarbonate may influence t he ability to e xtract a proton f rom t he Zn-coordinated water, which would i nfluence k cat . The tomato His 6 –LAP-A Arg431 mutant data com- pared favorably to studies with additional E. coli PepA Arg356 mutants [24]. PepA was inactivated by the R356E mutation but comparisons with the analogous tomato mutant could not be made b ecause the R431E His 6 –LAP- A protein did not accumulate in E. coli. While the R356A PepA had measurable activity on the chromogenic Leu-p- NA substrate, the R431A mutation abolished His 6 –LAP-A activity on all 19 dipeptides. As hydrolysis was not evident when peptides were used at 5.6 m M , kinetic analysis of the tomato R 431A His 6 –LAP-A was not possible. However because the residual activity of the R431A was < 0.5% and hydrolysis rates of Leu-Leu a nd Leu-Gly by the wild - type tomato enzyme are known [17], the k cat for the R431A His 6 –LAP-A could be no greater than 0.5 s )1 .Thisis similar to the value measured for the E. coli PepA R356A mutation (0.227/ s). The tomato LAP-A1 Lys354 (blLAP Lys262; PepA Lys282) is the second residue implicated in catalysis. This residue was not essential for the activity o f the tomato His 6 –LAP-A. Most substitutions for Lys354 created His 6 – LAP-A e nzymes with partial activity o n dipeptides. Significantly, the K354R mutant that retained a positive charge at position 354 did not have an enhanced His 6 – LAP-A activity relative to other K354 substitution mu- tants. These data suggested that the longer side chain of Arg d id not provide proper positioning of the guanidinium nitrogen or the t erminal amines relative to the P1 carboxyl oxygen of the peptide substrate. Stra ¨ ter et al. evaluated a single mutation in analogous site of the E. coli PepA (K282A) [24]. This mutation impacted both K m and k cat consistent with the role of Lys282 in PepA (blLAP Lys262) in binding of the carboxyl oxygen of the P1 residue and stabilization of the ge m-diolate transition state. The Ala substitution mutant (K354A) of the tomato His 6 –LAP-A was not evaluated in the tomato LAP-A1 series of mutations. The tomato residues Glu429 and Asp347 residues may coordinate both Zn1 and Zn2 residues [14,16]. While evidence for a n alteration in Zn 2+ binding was not provided here, it was clear that substitution of these acidic residues had a debilitating effect on LAP-A activity. Substitutions o f the tomato Glu429 (blLAP Gly334) impaired, but did not abolish, hydrolysis of Xaa-Leu and Leu-Xaa peptides. The substitutions at Glu429 eliminated one of th e hydrogen bonds that participated in the octahedral coordination geometry for Zn2 and Zn1. Perhaps the other hydrogen bonds were strong enough to retain the positioning of Zn1 and Zn2, the nucleophilic hydroxide, and substrate. Retention of charge at position 429 (E429D) produced a highly active His 6 –LAP-A enzyme suggesting that despite the differences in side chain length, the Asp and Glu at position 429 were appropriately positioned to promote near wild-type function. This is c onsistent with the fact t hat the Glu429 mutant (E429D) had similar K m and k cat values when compared to the wild-type His 6 –LAP-A enzyme. Similar mutants in the E. coli PepA or blLAP have yet to be evaluated. Asp347 (blLAP Asp255) was also implicated in Zn 2+ binding and stabilization o f the gem-diolate interme diate (14,16). The tomato Asp347 was essential for LAP activity because mutations abolished hydrolysis of all dipeptides. The inactivity o f the D347E His 6 –LAP-A showed that the additional carbon in the Glu side chain may have positioned the Glu carboxyl oxygens too close to Zn1 and Zn2, thereby preventing the octahedral coordination geometry normally associated with Asp347 Zn1 and Zn2 interactions [22]. This would indirectly influence the bin ding o f t he zinc-bridging water and/or ability of Z n2 and Zn1 to bind the N-terminal amine and carbonyl oxygens of the P1 residue of the substrate, respectively. T he replacement of Asp347 may more sev erely alter functions associated wit h Zn1. This hypothesis is b ased on the fact in the D431 mutants, the Zn2 would retain five of six productive coordination and Zn1 would be coordinated by only four strong and one potential, weak interaction. These data w ere also consistent with the surprising observation that rep lacement of Asp347 with Arg produced aHis 6 –LAP-A enzyme (D347R) with residual activity on many dipeptides. Given the long side chain of Arg, it is unlikely that the Arg side-chain guanidinium or the terminal amines would be positioned correctly to coordinate with Zn1 or Zn2. However, it is possible that the Arg residue could be oriented to productively interact with the P 1 carbonyl oxygen of the substrate and substitute for Zn1 function. This is supported by bond distances deduced from the blLAP-leucinal crystal structure [22]. In blLAP, the C a of Asp255 was approximately 6.64 and 6.54 A ˚ from the carbonyl oxygens of leucinal (a transition state analogue) and the average distance from the C a in Arg to its side-chain amines was 6.36 A ˚ . Collectively, these data indicate that the tomato LAP- A1 utilizes a reaction mechanism similar to that employed by the bovine LAP and E. coli PepA. Contin- ued evaluation of the t omato L AP-A1 m utants should allow m any newly emergent questions to be addressed. The surprising residual activity in the D347R His 6 – LAP-A is of particular interest. Future studies will evaluate the orientation of Arg in position 347 and determine if this basic residue is correctly positioned to bypass the role of Zn1 in substrate binding and catalysis. In addition, it will be important to address if the D347R His 6 –LAP-A enzymes and other s ubstitutions that alter the Zn1 and Zn2 binding residues influence Zn 2+ content of e ach LAP protomer. ACKNOWLEDGEMENTS We thank Dr M. F. Dunn, W. S. Chao, C . J. Tu, M. Matsui, and other Walling laboratory members for helpful conversations and Dr B. Hyman for use of h is sonic ator. This work was supported b y a National Science Foundation Grant IBN-9318260 and IBN-0077862 (to L. L. W.). 1638 Y Q. Gu and L. L. Walling (Eur. J. Biochem. 269) Ó FEBS 2002 REFERENCES 1. Kim, H. & Lipscomb, W.N. (1994) Structure and mechanism of bovine lens leucine aminopeptidase. Adv. Enzymol. Rel. Areas Mol. Biol. 68, 153–213. 2. Walling, L.L. & Gu, Y Q. (1996) Plant aminopeptidases: occur- rence, function and characterization. In Aminopeptidases (Taylor, A., ed.), pp. 174–219. R.G. Landes Co., Austin, TX, USA. 3. Taylor, A., Sanford, D. & N owell, T. (1996) Structure and func- tion o f bovine lens amino peptidase and comparison with homo- logous aminopeptidases. In Aminopeptidases (Taylor, A., ed.), pp. 21–67. R.G. Landes Co., Austin, TX, USA. 4. Gu, Y Q., Holzer, F.M. & Walling, L .L. (1999) Over-expression, purification a n d b ioc hemical characterization of t h e wound- induced leucine aminope ptidase of tom ato. Eur J. Biochem. 263, 726–735. 5. Melbye, S.W. & Carpenter, F.H. (1971) Leucine aminopeptidase (bovine lens). Stability and size o f subunits. J. Biol. Chem. 246, 2459–2463. 6. Taylor, A. (1996) Aminopeptidases R.G. Landes Co., Austin, TX, USA. 7. Chao, W.S., Pautot, V., Holzer, F.M. & Walling, L.L. (2000) Leucine am inopept idases: the ubiquity of LAP-N and the s peci- ficity of LAP-A. Planta 210, 563–573. 8. Chao, W.S., Gu, Y Q., Pautot, V., B ray, E.A. & Walling, L.L. (1999) Leucine aminopeptidase RNAs, proteins and activities increase in response to water deficit, salinity a nd the wound signals – systemin, methyl jasmonate, and abscisic acid. Plant Physiol. 120, 979–992. 9. Gu, Y.Q., Paut ot, V., Holzer, F.M. & Walling, L.L. (1996) A complex array of proteins r elated to the multimeric leucine ami- nopeptidase of tomato. Plant P hysiol. 110, 1257–1266. 10. Carrera, E. & Prat, S. (1998) Expression of the Arabidopsis abi1–1 mutant allele inhibits proteinase inhibitor wound-induction in tomato. Plant J. 15, 765–771. 11. Milligan, S.B. & Gasser, C.S. (1995) Nature and regulation of pistil-expressed genes in tomato. Plant Mol. Biol. 28, 691–711. 12. Hildmann, T., Ebneth, M., Pen ˜ a-Cortes, H., Sanchez-Serrano, J.J.,Willmitzer,L.&Prat,S.(1992)Generalrolesofabscisicand jasmonic acids i n gene activation a s a result of mechanical wounding. Plant Cell. 4, 1157–1170. 13. Pautot, V., Holzer, F.M., Reisch, B. & Walling, L.L. (1993) Leucine aminopeptidase: An inducible component of the d efense response in Lycopersicon esculentum (tomato). Proc. Natl Acad. Sci. USA 90, 9906–9910. 14. Pautot, V., H olzer, F. M., Chaufaux, J. & Walling, L.L. (2001) The induction of tomato leucine aminopeptidase genes (LapA ) after Pseudomonas syringae pv. tomato infection is primarily a wound response triggered b y coronatine. Mol. Plan t-Microbe I nt. 14 , 214– 224. 15. Ruiz-Rivero, O.J. & Prat, S. (1998) A -308 deletion of the tomato LAP promoters is able to direct flower-specific and MeJA-induced expression in transgenic plants. Plant Mol. Biol. 36, 639–648. 16. Schaller, A., Bergey, D.R. & Ryan, C.A. (1995) Induction of wound response g ene s in to mato leaves by bestatin, an in hibitor o f aminopeptidases. Plant Cell. 7, 1893–1898. 17. Gu, Y Q. & Walling, L.L. (2000) Specificity of the wound- induced leucine aminopeptidase (LAP-A) of t omato: activity on dipeptide and tripeptide substrates. Eur J. Bio chem. 267, 1178– 1187. 18. Burley, S.K., David, P.R., Taylor, A. & Lipscomb, W.N. (1990) Molecular s tructure of leucine aminop eptidase at 2.7-A ˚ resolution. Proc. Natl Acad. Sci. USA 87, 6878–6882. 19. Burley, S.K., David, P.R., Sweet, R.M., Taylor, A. & Lipscomb, W.N. (1992) Structure determination and refinement of bovine lens leucine aminopeptidase and its complex with bestatin. J. Mol. Biol. 224, 113–140. 20. Kim, H., Burley, S.K. & Lipscomb, W .N. (1993) Re-refinement of the X-ray crystal structure of b ovine lens leucine aminopeptidase complexed with bestatin. J. Mol. Biol. 230, 722–724. 21. Stra ¨ ter, N. & Lipscomb, W.N . (1995) Transition state analogue L -leucinephosphonic acid bound to bovine lens leucine amino- peptidase: X-ray structure at 1.65 A ˚ resolution in a new crystal form. Biochemistry. 34, 9200–9210. 22. Stra ¨ ter, N. & Lipscomb, W.N. (1995) Two-metal ion mechanism of bovine lens leucine aminopeptidase: active site solvent structure and binding mode of L -leucinal, a gem-diolate transition state analogue, by X-ray cry stallography. Biochemistry 34, 147 92– 14800. 23. Stra ¨ ter, N., Sherratt, D.J. & Colloms, S.D. (1999) X-ray structure of aminopeptidase A from Escherichia coli and a model for the nucleoprotein complex in Xer site-specific recombination. EMB O J. 18, 4513–4522. 24. Stra ¨ ter, N., Sun, L., Kantrowitz, E.R. & Lipscomb, W.N. (1999) A bicarbonate ion as a general base in the mechanism of peptide hydrolysis by dizinc leucine aminopeptidase. Proc. Natl Acad. Sci. USA 96, 11151–11155. 25. van Wart, H.E. & Lin, S.H. (1981) Metal binding stoichiometry and mechanism of metal ion modulation of the activity of porcine kidney leucine amin opeptidase . Bioche mi stry 20 , 5682–5689. 26. Kunkel, T.A., B ebenek, K. & McClary, J. (1991) Efficient site- directed mutagenesis using uracil-containing DNA. Methods Enzymol. 204, 125–139. 27. Mikkonen, A. (1992) Purification and characterization of leucine aminopeptidase from kidney bean cotyledons. Physiol. Plant. 84, 393–398. 28. Binkley, F., Leibach, F. & King, N. (1968) A new method of peptidase assay and the separation of three leucylglycinases of renal tissues. Arch. Biochem. Biophys. 128, 397–405. 29. Ramagli, L.S. & Rodriguez, L.V. (1985) Quantitation of microgram amounts of protein in two-dimensional polyacrylamide gel electrophoresis sample buffer. Electrophoresis 6, 559–563. 30. Schechter, I. & Berger, A . (1967) On the size of the active site in proteases. I. Papain. Biochem. Biophys. Res. Comm. 27, 157–162. 31. Barrett, A.J. (1994) Classification o f peptidases. Methods E nzymol. 244, 1–15. 32. Kuzmic, P. (1996) Program DYNAFIT for t he analysis of enzy me kinetic data: application to HIV proteinase. An al. Biochem. 237, 260–273. 33. Laemmli, U.K. (1970) Cleavage of structural proteins during assembly of the head of bacteriophage T4. Nature 227, 680–685. 34. Gu, Y.Q., Chao, W.S. & Walling, L.L. (1996) Localization and post-translational processing of the wound-induced leucine ami- nopeptidase proteins o f tomato. J. Biol. Chem. 271, 25880–25887. 35. Bartling, D. & Weiler, E.W. (1992) Leucine aminopeptidase from Arabidopsis thaliana – molecular evidence for a phylogenetically conserved enzyme of protein turnover in higher plants. Eur. J. Biochem. 205, 425–431. 36. Stirling, C.J., Colloms, S.D., Collins, J.F., Szatmari, G. & Sherratt, D.J. (1989) xerB,anEscherichia coli ge ne required fo r plasmid ColE1 site-specific recomb ination, is identical to pepA, encoding aminopeptidase A, a protein with substantial s imilarity to bovine lens leucine amino peptidase. EMBO J. 8, 1623–1627. 37. Wood, D.O., Solomon, M.J. & Speed, R.R. (1993) Character- ization o f the Rickettsia prowazekii pepA ge ne encoding leucine aminopeptidase. J. Bacteriol. 17 5, 159–165. 38. Wallner, B.P., Hession, C., Tizard, R., Frey, A.Z., Zuliani, A., Mura, C., Jahngen-Hodge, J. & Taylor, A. (1993) Isolation of bovine kidney leucine aminopeptidase cDNA comparison with lens enzyme a nd tissue-specific expression of two mRNAs. Bio- chemistry 32, 9296–9301. 39. Burley, S.K., David, P.R. & Lipscomb, W.N. (1991) Leucine aminopeptidase – bestatin inhibition and a model for Ó FEBS 2002 Site-directed mutagenesis of the tomato LAP-A (Eur. J. Biochem. 269) 1639 [...]...1640 Y. -Q Gu and L L Walling (Eur J Biochem 269) enzyme-catalyzed peptide hydrolysis Proc Natl Acad Sci USA 88, 6916–6920 40 Kim, H & Lipscomb, W.N (1993) X-ray crystallographic determination of structure of bovine lens leucine aminopeptidase complexed with amastatin formulation of a catalytic mechanism featuring a gem-diolate transition state Biochemistry 32, 8465– 8478 Ó FEBS 2002... transition state Biochemistry 32, 8465– 8478 Ó FEBS 2002 41 Carpenter, F.H & Vahl, J.M (1973) Leucine aminopeptidase (bovine lens) Mechanism of activation by Mg2+ and Mn2+ of the zinc metalloenzyme, amino acid composition, and sulfhydryl content J Biol Chem 248, 294–304 42 Vogt, V.M (1970) Purification and properties of an aminopeptidase from Escherichia coli J Biol Chem 245, 4760–4769 . foundation for the study into the residues critical of the activity of the wound-induced tomato LAP-A1 [6,22,24]. Similar to the animal and prokaryotic enzymes,. Identification of residues critical for activity of the wound-induced leucine aminopeptidase (LAP-A) of tomato Yong-Qiang Gu* and Linda