Lysine residues in human adenylate kinase are essential for interaction with adenine nucleotides as found by site-directed random mutagenesis †

 

 

 

 

 

Takanori Ayabe ň,Åò, Hitoshi Takenaka ň, Osamu Takenaka Åa , Michihiro Sumida , Seung Kyu Park ň, Toshio Onitsuka Åò, Koichiro Shibata Åò , Seiichi Uesugi ^, and Minoru Hamada ň,*

 

 

 

ňDepartment of Hygiene ,Åò The Second Department of Surgery,

Miyazaki Medical College, Miyazaki, 889-16, Japan, ÅaPrimate Research Institute, Kyoto University, Aichi, 484, Japan, Department of Second Medical Biochemistry, Ehime University, School of Medicine, Onsen-gun, Ehime, 791-02,  ^ Department of Bioengineering, School of Engineering, Yokohama National University, Yokohama, Kanagawa, 240, Japan

 

 

 

* Correspondence should be addressed to Minoru HAMADA

 

Address: Department of Hygiene, Miyazaki Medical College, Kiyotake, Miyazaki-gun, Miyazaki, 889-16, Japan.

email: tayabe@post.miyazaki-med.ac.jp

Phone: 81-985-85-0873

FAX: 81-985-85-5177

 

 

 

 

 

† Footnotes: This work was supported by a general grant and Grant-in-Aid (M. H. #900395-29) from the Ministry of Education, Science, Sports, and Culture of Japan.  

 

 

Running Title: Lysines in Adenylate Kinase Interact with Substrates

 

 

 

Abbreviations: ADP, adenosine 5'-diphosphate; AK, adenylate kinase; AKc, chicken adenylate kinase; AMP, adenosine 5'-monophosphate; ATP adenosine 5'-triphosphate; EDTA, ethylenediaminetetraacetate; hAK1, human adenylate kinase 1; NADH, nicotinamide adenine dinucleotide; NMR; nuclear magnetic resonance; Tris, 2-amino-2-(hydroxymethyl)-1,3-propanediol; WTAK, wild-type adenylate kinase.

 

 

 

 

Abstract: To elucidate the minimum requirement of amino acid residues for the active center in human adenylate kinase (hAK1), we carried out random site-directed mutagenesis of key lysine residues (K9, K21, K27, K31, K63, K131, and K194), which were conserved in mammalian AK1 species, with the pMEX8-hAK1 plasmid (Ayabe et al., 1996, Biochem. Mol. Biol. Int. 38, pp.373-381).  Twenty different mutants were obtained and analyzed by steady-state kinetics, and all muteins showed activity loss by Km and/or kcat effects on one or both of MgATP2- and AMP2-.  The results have led to the following conclusions.  (1) Lys9 would appear to interact with both MgATP2- and AMP2- but to a larger extent than with AMP2-. (2) Lys21 is likely to play a role in substrate-binding of both MgATP2- and AMP2- but more strongly affects MgATP2-.  (3) Lys27 and Lys131 would appear to play a functional role in catalysis by interacting strongly with MgATP2-.  (4) Lys31 would appear to interact with MgATP2- and AMP2- at the MgATP2- site.  (5) Lys63 would be more likely to interact with MgATP2- than with AMP2-. (6) Lys194 in the flanking C-terminal domain would appear to interact not only with MgATP2- but also with AMP2- at the MgATP2- site by stabilizing substrate-binding.  The loss of the positively charged e-amino group of lysine affects both affinity for the substrate and catalytic efficiency.  Hence, hydrophilic lysine residues in hAK1 would appear to be essential for substrate-enzyme interaction with the coordination of the some arginine residues, reported previously (Kim et al., 1990, Biochemistry 29, pp.1107-1111).

 

 

 

Adenylate kinase (AK) (EC 2.7.4.3) is a ubiquitous enzyme, abundant in the cytoplasm, which catalyzes the reaction: MgATP2- + AMP2-  MgADP- + ADP3-.  There are two distinct nucleotide-binding sites, one for metal-chelated MgATP2- or MgADP-, and the other for unchelated AMP2- or ADP3- (Noda 1973; Hamada et al.,1979).  AK contributes to homeostasis of the adenine nucleotide composition in the cell (Atkinston, 1977).  In vertebrates, three isozymes (AK1, AK2, AK3) have been characterized (Noda 1973; Khoo & Russel, 1972; Tomasselli et al. 1979).  AK1 is present in the cytosol, AK2 is found in the intermembrane space of the mitochondria and AK3 exists in the mitochondrial matrix.  In human genetic disorders, an aberrant AK isozyme has been specifically found in the serum of Duchenne dystrophic patients (Hamada et al., 1985).  An AK1 deficiency associated with a hemolytic anemia has been reported (Miwa et al., 1983; Matsuura et al., 1989; Toren et al., 1994), and the replacement of Arg-128 with Trp in AK1 was considered to be the cause of the enzyme deficiency. The same mutant prepared by site-directed mutagenesis resulted in reduced catalytic activity and decreased solubility (Matsuura et al., 1989).  Based on these findings, it seems reasonable to study the properties of various mutant AKs in order to better understand disorders due to such mutant enzymes. 

AK has ten a-helices and five b-strands (Schulz et al., 1974).  The primary structures of some isoenzymes of AK have been well identified, but their tertiary structures have not been completely elucidated  (Schulz et al., 1986).  Extensive attempts have been made to identify the substrate binding sites of AK: by X-ray crystallography of porcine AKs (Sachsenheimer & Schulz, 1977; Dreusicke et al., 1988) and yeast AK (Egner et al., 1987), by NMR studies of isolated rabbit muscle AK (Fry et al., 1985; Fry et al., 1987; Fry et al., 1988) and the synthetic MgATP2- binding fragments of AK (Hamada et al., 1979), and by site-directed mutagenesis studies of chicken AK (AKc) (Tagaya et al., 1989; Yan et al., 1990; Tian et al., 1990; Yoneya et al., 1990; Yan & Tsai, 1991; Okajima et al., 1991) and human AK (hAK1) (Kim et al, 1990).  However, the active site for catalysis and important residues for substrate-binding have not been exactly determined.  To elucidate the AK structural model proposed by X-ray crystallography and NMR studies, interaction with substrates and specific residues should be tested by site-directed mutagenesis and steady-state kinetics.  In hAK1, the conserved arginine residues (R44, R97, R132, and R138) were substituted with alanine and the ATP site in the original X-ray model was reassigned to the AMP site; these arginine residues were determined to be essential for catalytic activity (Kim et al, 1990).  The reassignment of the AMP site in the revised X-ray model has also been suggested by NMR data (Smith & Mildvan, 1982) and by the data for the F86W mutant of AKe (Liang et al., 1991).  However, there have been remained to proceed detailed analysis of MgATP2- site and AMP2- one in an AK structural model.

In the present study, at first, to evaluate a mutant form of hAK1 as a model of enzyme deficiency disorders, and secondary, to elucidate the minimum requirement of amino acid residues at the active center in hAK1, as succeeding analysis of Arg-mutants in hAK1 (Kim et al., 1990|), we substituted the lysine residues conserved in mammalian AK, which were suggested to be essential to interact with the substrate by NMR and X-ray diffraction analyses (Fry et al., 1985, Fry et al., 1988, Dreusicke et al., 1988, Reinstein et al., 1990, Tian et al., 1990, Byeon et al., 1995).  As reported previously, Lys21, Lys27, and Lys194 were suggested to interact with the substrates (Hamada et al., 1979).  The reason why we chose lysine is that the changes from aliphatic positively charged and hydrophilic basic amino acid would be largely changed compared to those from the other amino acid of a neutral, hydrophobic, and non-charged side chain.  To define a structural AK model previously proposed by Kim et al (1990), we constructed the pMEX8-hAK1 plasmid (Ayabe et al., 1996) from the pAK vector as described previously (Kim et al., 1990).  We partially modified the method for random site-directed mutagenesis based on the phosphorothioate technique (Taylor et al., 1985, Nakamaye et al., 1986, Sayers et al., 1992) by synthesizing random antisense nucleotide primers containing three mismatched bases for a target codon in anticipation of random mutations at one target codon, which annealed as an oligonucleotide primer to the single DNA strand and was extended as a mutant strand.  In hAK1, a loss of the positively charged guanidino groups of the arginine residues decreased catalytic efficiency (Kim et al., 1990), and the positive charge should interact with the negatively charged phosphates of MgATP2- and AMP2-.  To elucidate the function of the positive e-amino groups of the hydrophilic lysine residues, we mutated Lys9, Lys21, Lys27, and Lys31 in the head domain, Lys63 and Lys131 in the middle region, and Lys194 in the tail domain of hAK1.  These mutants were analyzed by steady-state kinetics. 

 

Materials and Methods

Materials.

The plasmid pMEX8-hAK1 (Ayabe et al., 1996) was used for random site-directed mutagenesis.  The bacterial strain JM109 was purchased from TaKaRa Shuzo, Co., Ltd. (Tokyo, Japan).  TG1 and a SculptorTM in vitro Mutagenesis Kit were from Amersham LIFE SCIENCE (Buckinghamshire, England).  The Blue Sepharose CL-6B column, Superose 12 (HR 10/30), and the fast protein liquid chromatography system were purchased from Pharmacia Biotech, Inc. (Tokyo, Japan).  Adenine nucleoside mono- and triphosphates, AMP, ATP, and nicotinamide adenine dinucleotide (its reduced form, NADH) were purchased from Oriental Yeast Co. (Tokyo, Japan).  Pyruvate kinase (PK), phosphoenolpyruvate (PEP), and lactate dehydrogenase (LDH) were from Sigma Chemicals (St. Louis, MO, USA).  All other reagents were of analytical grade and were purchased from either Wako Pure Chemicals (Osaka, Japan) or Nacalai Tesque (Kyoto, Japan).

Purification of single strand pMEX8-hAK1 DNA.

A single colony of JM109/pMEX8-hAK1 (JM109 transformed with pMEX8-hAK1) was cultured in 10 ml of TYP medium (1.6% Tryptone, 1.6% yeast extract, 0.5% NaCl, 0.25% K2HPO4) containing 50 mg/ml of ampicillin at 37°C overnight until the A600 was approximately 0.5.  Helper phages (VCS-M13, Stratagene, La Jolla, CA, U.S.A.) were added to the culture medium at a multiplicity of infection between 10 and 20 (a phage : cell ratio of between 10 : 1 and 20 : 1) with 25 mg/ml of kanamycin.  The medium was cultured at 37°C with vigorous aeration overnight and was centrifuged at 3,000 X g for 20 min.  To the supernatant was added a solution of polyethyleneglycol 6000 and 2.5 M NaCl.  The mixture was allowed to stand at room temperature for 15 min.  After centrifugation of the solution, the pellet was resuspended with TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0), and single stranded DNA of pMEX8-hAK1 was extracted by TE-saturated phenol and chloroform according to the manufacturer's recommendations. 

Random site-directed mutagenesis of hAK1.

     The antisense primers were 5'-CGAAGATGATYXXAGTCTTCTTAAGC-3' for Lys9 residue, 5'-GCACTGGGTACCYXXGCCAGAACC-3' for Lys21, 5'- GCACGATYXXCTCGCACTGGG -3' for Lys27, 5'-GTGTAGCCGTAYXXCTGCACGATTTTC-3' for Lys31, 5'-CCAGCTGACCYXXTTCCATGATTTC-3' for Lys63, 5'-GTTTCGCCGCGYXXCAGCAGGCG-3' for Lys131, and 5'-CGAAGATGATYXXAGTCTTCTTAAGC-3' for Lys194.  These primers were synthesized with a DNA synthesizer (Applied Biosystems, Model 394) and phosphorylated with T4 Polynucleotide Kinase (WAKO Co., Tokyo).  The underlined letters, YXX, represent a target codon for random mutagenesis, where X was either A, G, C or T, while Y was either G or C.  Site-directed mutagenesis was carried out with a combination of the SculptorTM in vitro Mutagenesis Kit (Amersham LIFE SCIENCE, Buckinghamshire, England) using these site-specific primers annealed to the template single strand DNA of pMEX8-hAK1.  This method was based on the phosphorothioate technique (Taylor et al., 1985, Nakamaye et al., 1986, Sayers et al., 1992), that is, dCTP was used instead of dCTPaS during both the annealing and extension reaction of the randomly constructed oligonucleotides.

Screening of mutants by DNA cycle sequencing.

A homoduplex mutant DNA constructed by site-directed mutagenesis was transformed with competent cells (TG1) prepared according to the directions in the mutagenesis kit and was spread on an LB plate containing 50 mg/ml of ampicillin.  Single clones were cultured in 10 ml of LB medium overnight.  A double-stranded DNA of the plasmid was purified according to the manufacturer's instructions (Flex Prep Purification Kit, Pharmacia Biotech, Tokyo, Japan).  The mutant plasmid DNA was sequenced by the dideoxy method (Sanger et al., 1977).  Sequencing primers labeled with fluorescent isothiocyanate were the product of Japan Bioservices.  The DNA sequence of the forward primer was 5'-TGGAATTGTGAGCGGATAAC-3' and that of the reverse primer was 5'-AAAATCTTCTCTCATCCGCC-3'.  Polymerase chain reaction (PCR) was performed with the AmpliCycleTM Sequencing kit (Perkin Elmer, Branchburg, NJ, U.S.A.) with a DNA Thermal Cycler (Model PJ-480, Perkin Elmer Cetus) according to the partially modified protocol described as follows.  To enhance Taq DNA polymerase activity, AmpliTaqTM DNA Polymerase solution and Taq DNA Polymerase (Promega, Madison, WI, U.S.A) were mixed at a ratio of 9 : 1 as a cycling mixture.  Eight ml of a master mixture was made by mixing 2.8 mg of double-stranded DNA, 1 ml of dimethylsulfoxide (Sigma Chemical Co., St. Louis, MO, U.S.A.), and 2 pmol of the FITC-labeled sequence-primer.  This mixture was incubated at 95°C for 10 min and immediately cooled on ice.  Two ml of the cycling mixture and 0.2 unit of a Perfect Match Enhancer (Stratagene, La Jolla, CA, U.S.A.) were added to the cooled master mixture.  PCR conditions were initial denaturation at 95°C for 5 min, 20 cycles at 95°C for 30 s, 53°C for 30 s, and 72°C for 60 s, and 20 cycles at 95°C for 30 s and 72°C for 60 s.  PCR products were denatured by 95% formamide.  Electrophoresis of the PCR product was performed by autosequencer (Shimadzu, DSQ1, Kyoto, Japan).  DNA sequencing of the mutant plasmid was performed with forward and backward sequencing primers to avoid undesirable mutations in the entire hAK1 gene. 

Expression and purification of wild type AK (WTAK) and mutant AK.

JM109/pMEX8-hAK1 (wild type AK) and TG1/pMEX8-mutant-hAK1 were separately cultured in 10 ml of LB medium containing 50 mg/ml of ampicillin overnight at 37°C and were transferred into 250 ml of LB medium.  After 1 hr culture, isopropyl-b-D-thio-galactopyranoside (IPTG) was added at a final concentration of 1 mM, and growth in the medium was continued for 16 hr under the same conditions.  E. coli cells were centrifuged at 5,000 X g for 20 min, and the pellet was suspended in 10 ml of a standard buffer [20 mM Tris-HCl, 1 mM EDTA, 0.1 mM dithiothreitol (DTT), pH 7.4].  All chromatographic steps were carried out at 4°C in the chromatochamber.  The suspension of E. coli cells was disrupted with an ultrasonicator (Model 250 sonifier, Branson Ultrasonics Co., Danbury, CT, U.S.A.) at 20 kHz and 20 W for 3 min on ice.  The homogenate was centrifuged at 12,000 X g for 20 min at 4°C.  The supernatant was subjected to affinity chromatography on Blue Sepharose CL-6B (ø 1 X 5 cm) which had been equilibrated with the standard buffer and eluted with a NaCl gradient (0 M to 1 M NaCl in the standard buffer) at a velocity of 0.5 ml/min.  AK protein was determined by 12.5% polyacrylamide gel electrophoresis in the presence of sodium dodecylsulfate (SDS-PAGE) to confirm a single AK protein (Sambrook et al., 1989), and the AK fraction was concentrated by centrifuging with Centriplus-10 (Amicon, Inc., Tokyo, Japan)  The concentrated AK (0.5 ml) sample was loaded onto a Superose 12 column (ø 1 X 30 cm) and eluted with imidazole buffer (5 mM imidazole-HCl, 1 mM EDTA, 0.1 mM DTT, pH 6.9) at a velocity of 0.5 ml/min.  Each fraction was also determined by 12.5% SDS-PAGE to confirm a single band of AK protein. 

Kinetic analysis of forward reaction of AK.

Enzyme activity was assayed in the forward direction by adding various amounts of MgSO4, ATP, and AMP.  The initial velocity of the forward reaction was measured by observing the absorbance change at 340 nm with a Cary 2290 spectrophotometer (Varian, Mulgrave, Australia) [for NADH, in a coupled-enzyme assay in the presence of pyruvate kinase and lactate dehydrogenase to monitor ADP formation at 25°C as previously described] (Hamada & Kuby, 1978).  The forward reaction mixture in a total of 1 ml contained 75 mM triethanolamine hydrochloride (pH 7.4), 120 mM KCl, 0.2 mM NADH, 0.3 mM PEP, 0.3 mg/ml bovine serum albumin (BSA), 10 units of LDH, 5 units of PK, and 1.0 mM MgSO4, and the various concentrations of MgATP2- and AMP2- were as follows: five combinations selected from 0.05, 0.1, 0.2, 0.5, 1.0, 2.0, 3.0, 5.0 mM MgATP2- and at a fixed concentration of 2 mM AMP2- were used in the determination of the apparent Michaelis constants (Km) for MgATP2-.  Five combinations selected from 0.05, 0.1, 0.2, 0.5, 1.0, 2.0, 3.0, 5.0 mM AMP2- and a fixed concentration of 2.0 mM MgATP2- were used in the determination of the apparent Km for AMP2-.  The reaction was initiated by the addition of 10 ml of the recombinant hAK1 solution diluted to desired concentrations.  The AK sample was diluted with a buffer (5 mM imidazole-HCl, 1 mM EDTA, 0.1 mM DTT, 1% BSA, pH 7.4).  Km and Vmax were estimated using a double-reciprocal plot (Lineweaver & Burk, 1934), and kcat was calculated by dividing Vmax by the total amount of enzyme (Et) present in the reaction mixture.  One unit of enzyme activity is defined as the amount of enzyme that produces 1 mmol of ADP in one minute.      

Others.

     The concentration of protein was determined by the method of Lowry (Lowry et al., 1951).  Concentrations of adenine nucleotides and NADH were spectrophotometrically determined by using millimolar extinction coefficients of 15.4 and 6.22, respectively.

 

Results

Site-directed mutagenesis of Lys residues and purification of wild type AK (WTAK) and mutant AKs.

The results of site-directed random mutagenesis of pMEX8-hAK1 are summarized in Table 1.  Twenty-six mutants at seven target residues were determined with backward and forward sequencing primers, and the efficiency of the random mutagenesis was averaged to 27.7% (26 mutants/94 colonies).  For the K27I (ATA) mutant, the third position of codon 27 was replaced by 'A', although the third base was set as 'G' or 'C' in the annealing primer.  This could be due to either contamination of 'A' during primer synthesis by the DNA synthesizer or by an unexpected point mutation during mutagenesis.  The mutant pMEX8-hAK1 plasmid was expressed, and the mutant AK enzymes were purified to homogeneity by column chromatography using Blue Sepharose and Superose 12.  These enzymes possessed basically the same chromatographic elution patterns as the WTAK.  However, both K27V and K131P mutants could not be purified after disruption of the expressed cells due to their insolubility in the standard buffer described in Methods.  All other mutant enzymes showed single bands with the same mobility as that of WTAK on 12.5% SDS-PAGE (data not shown).  The expressed yields and the specific activities of the mutants were low compared to those of WTAK (Table 1).

Kinetic parameters of mutant AKs.    

     Steady-state kinetic data for WTAK and each mutant in the forward reaction are shown in Table 2.

(1) Properties of K9-mutants.  The Km values for K9P were slightly decreased (0.2-fold) for MgATP2- and AMP2- compared to those of WTAK, however, the decrease in kcat  was comparable (9.3%) among the other K9-mutants.  The kcat of K9F was decreased to 0.3% with an unchanged Km for both substrates.  With K9L there was a 6.8-fold increase in Km for AMP2-, kcat  was decreased to 0.4% and the kcat/Km value for AMP2- was dramatically decreased to 6.5 X 10-2%.  The Km values of K9T showed more than 10-fold increases for MgATP2- and AMP2-, and the kcat was reduced to 1.6%.  These results strongly suggest that K9-mutants interact with both MgATP2- and AMP2- and have a larger effect on AMP2- than on MgATP2-.  

(2) Properties of Lys21-mutant.  The Km values of K21P for MgATP2- and AMP2- were markedly increased, 19.4-fold and 13.8-fold, respectively.  The kcat  value was decreased to 5.0%.  The results for K21P suggested interaction with both MgATP2- and AMP2-, however, the mutant more strongly affected MgATP2-.

(3) Properties of Lys27-mutants.  The Km values for AMP2- of K27R, K27L, and K27I mutants were not changed (1.1- to 1.5-fold), while those for MgATP2- were moderately increased (3.6- to 8.9-fold).  The kcat  values were decreased greater than 2.0%, and the kcat/Km values were significantly decreased.  The K27-mutants appeared to interact with MgATP2- to a greater extent than with AMP2- and showed correspondingly decreased catalytic efficiency.  It is surprising that for K27R, although the positive charge was conserved, a decrease in kcat (to 1.7%) was observed despite an unchanged Km.

(4) Properties of K31-mutants.  The Km values of K31I were unchanged for MgATP2- and AMP2-, 0.9- and 1.4-fold, respectively, while the kcat  value decreased to 1.9%.  K31F showed an increased Km for MgATP2- (12.1-fold), and the kcat/Km values for MgATP2- were decreased to 9.5 X 10-2%  In contrast, the Km value of K31S dramatically increased (23.6-fold) for AMP2-, and the kcat/Km for AMP2- showed a large decrease to 2.3 X 10-2%.  K31I did not significantly affect the affinity for either substrate, although K31F strongly affected MgATP2- and K31S greatly affected AMP2-.  K31-mutants appeared to interact with MgATP2- and AMP2- and to be essential for catalysis.

(5) Properties of K63-mutant.  The Km values of K63F increased 8.1-fold for MgATP2- and were unchanged (0.8-fold) for AMP2-.  The kcat decreased to 0.8%, and the kcat/Km value for MgATP2- was decreased to 9.5 X 10-2%.  The K63-mutant interacted with MgATP2- and showed decreased catalytic efficiency.

(6) Properties of K131-mutants.  Results for K131A showed a dramatic increase in Km for MgATP2- (16.9-fold), a decrease by 0.4% for kcat  and a large decrease in kcat/Km for MgATP2- (2.6 X 10-2%).  The Km of K131F showed a small change (3.2-fold for MgATP2- and 1.6-fold for AMP2-), but the kcat values were decreased to 0.4%.  K131-mutants were suggested to interact greatly with MgATP2- in catalysis.

(7) Properties of K194-mutants.  The Km value of K194S for MgATP2- was moderately increased (14.5-fold), while that for AMP2- was not changed (3.4-fold), and the kcat  was not significantly changed to 37.3%.  In the case of K194L, the Km showed a moderate increase of 9.6-fold for MgATP2- but decreased 0.1-fold for AMP2-, and the kcat slightly decreased to 13.7%.  The kcat/Km of K194L for MgATP2- was markedly decreased to 1.4%, however, that for AMP2- was strikingly increased catalytic efficiency for AMP2- (229.4%).  On the other hand, despite the fact that the K194V mutant was similar to K194L, the Km value of K194V was greatly increased (20.8-fold) for AMP2- and unchanged (0.9-fold) for MgATP2-.  The kcat  decreased to 1.4%.  The kcat/Km of K194V for AMP2- was markedly decreased (7.1 X 10-2%) in spite of a small decrease in catalytic efficiency for MgATP2- (1.5%).  K194I showed a small decrease of 0.2-fold in Km values for MgATP2- and 0.6-fold for AMP2-, and the kcat was decreased to 2.4%.  K194P showed the largest decrease (0.8%) in kcat among all the K194-mutants.  K194-mutants interacted not only with MgATP2- but also with AMP2-.

 

Discussion

X-ray structural studies indicate that AK protein has ten a-helices and five b-strands (Schulz, et al., 1974), and that the active center cleft opens to some extent (Figure 1).  In a previous study of hAK1, replacement of the arginine residues at R44, R97, R132, and R138 with alanine resulted in decreased catalytic efficiency (Kim et al., 1990).  The loss of the positively charged guanidino groups of the arginine residues would inhibit catalytic efficiency, and the arginine residue was suggested to be essential for catalysis by interacting with the negative charge of the phosphates of adenine nucleotides.  In hAK1, site-directed mutagenesis has not been performed except Arg-residues.  In the present study, to elucidate the minimum requirement for amino acid residues at the active center in hAK1, and to elucidate an interaction with substrates and specific residues, we substituted the positively charged e-amino groups of Lys9, Lys21, Lys27, Lys31, Lys63, Lys131, and Lys194 residues by random site-directed mutagenesis.  These lysine residues were well conserved in mammalian species.  At first, we partially modified the site-directed mutagenesis strategy based on the phosphorothioate technique (Taylor et al., 1985, Nakamaye et al., 1986, Sayers et al., 1992).  In anticipation of random mutations, the antisense primers were synthesized to contain a three-base-mismatched codon of YXX (Y=G or C; X=A, G, C, or T) for a target amino acid, and the constructed mutant homoduplex DNA was annealed to the single strand DNA of pMEX8-hAK1.  Twenty-two different mutant pMEX8-hAK1 plasmids were confirmed by screening from 94 colonies, and the efficiency of our site-specific mutagenesis averaged 23.4% (22/94).  We have not analyzed the secondary structures of all mutants with circular dichroism spectra because these lysine residues might not be directly involved in the secondary structures, as reported in a previous paper, in the case of replacement of arginine with alanine very large differences were not displayed (Kim et al., 1990).  However, as a control of Lys-mutant, circular dichroism studies should be shortly performed on several kinds of Lys-mutants.  K27V and K131P mutants could not be analyzed because they were insoluble in the standard buffer described in Materials and Methods, which cause might be considered due to the helical change as observed in other proteins.  The conformation or hydrophobicity of these insoluble mutants might have been altered upon folding, however, as it was not solublized, which mutants should be detected by optical rotational dichroism (ORD) studies.  In human genetic disorders associated with nonspherocytic hemolytic anemia, an AK1 deficiency in erythrocytes has been reported (Miwa et al., 1983; Matsuura et al., 1989; Toren et al., 1994).  The nucleotide sequence of the patient's AK1 gene resulted in a replacement of Arg-128 with Trp, and the properties of the mutant AK expressed in vitro showed reduced catalytic activity as well as decreased solubility in AKc (Matsuura et al., 1989).  A new mutant enzyme constructed by site-directed mutagenesis may be useful to study the pathophysiological mechanism of disorders due to a mutant enzyme derived from a point mutation.  The affinity for substrates changed and/or the catalytic efficiency of Lys-muteins were variously decreased and in spite of a decreased solubility, which event could explain a part of the cause of hemolytic anemia derived from AK1-mutant. 

As an assumption, we interpreted the results of steady-state kinetics of each Lys-mutein.  K9-mutants affected the affinity for both MgATP2- and AMP2- but more strongly interacted with AMP2-, which decreased catalytic efficiency.  Lys9 would be essential for catalysis by interacting with MgATP2- and AMP2-.  The glycine-rich loop plays an important role in catalysis (Fry et al., 1988; Mildvan & Fry, 1987; Reinstein et al., 1988).  The K21P could make the flexible P-loop bend in any direction, which could affect the substrate-binding site, reducing the affinity for both substrates and decreasing the kcat value.  The Lys21 residue would be essential for catalysis and substrate-binding.  Lys21 was proposed to be close to the MgATP binding site (Fry et al., 1985), and the e-amino group was suggested to be close to the subsite for the g-phosphate of ATP (Tagaya et al., 1987).  According to an NMR study of K21M in AKc, Lys21 may play a key structural role (Tian et al., 1990).  Based on NMR analyses of K21R and K21A, Lys21 would orient the triphosphate chain of MgATP to the proper conformation required for catalysis (Byeon et al., 1995).  K27-mutants showed moderately decreased affinity for MgATP2- and decreased catalytic efficiency.  From our data obtained, in fact, Lys27 would appear to play a functional role in catalysis at the MgATP2- site.  The K27R mutant, in spite of the change of an e-amino group to a guanidino group retaining a positive charge, showed a largely reduced kcat.  The catalytic function of Lys27 could not be replaced by a conservative residue arginine, and the e-amino group was presumed to be necessary for phosphoryl binding and transfer.  Based on NMR studies of K27M in AKc (Tian et al., 1990), Lys27 was nonessential for either catalysis (including substrate binding) or structure; however, the discrepancies between this report and our data might be due in part to the difference in experimental conditions using different species and types of mutations in the side chains.  The roles of these same mutations might differ between AKc and hAK1.  This should be tested using the K27M mutant in hAK1.  A study of K31-mutants revealed that K31F showed a decreased affinity for MgATP2- but K31S showed reduced affinity for AMP2-.  Lys31 was proposed to be at or near bound MgATP2- (Fry et al., 1988); however, the result affected both MgATP2- and AMP2-.  When the enzyme forms a ternary complex by substrate-binding, the side chain of Lys31 at the MgATP2- site might permit the AMP2- substrate to interact with the phosphate of AMP2- for catalysis.

Lys63 is denoted on the right site close to the AMP2- site (Fig. 1).  K63F showed a greatly reduced affinity for MgATP2- and decreased catalytic efficiency.  By photoaffinity labeling analysis, the amino acid domain (Gly64 to Arg77) was proposed to constitute the neighborhood identifiable with AMP binding during binary-to-ternary complex formation (Pai et al., 1992).  When two substrates (MgATP2- and AMP2-) bind with the enzyme and form a ternary complex, some conformational change could occur which allows the side chain of Lys63 access to the MgATP2- substrate.  The 120-133 helical region exists at the top of the AK model (Fig. 1).  K131-mutants showed moderately decreased affinity for MgATP2- and decreased catalytic efficiency.  Based on kinetic analyses of R132A in hAK1 (Kim et al., 1990), Arg132 next to Lys131 would appear to be essential for catalysis by interacting strongly with MgATP2- and only partially with AMP2-.  According to NMR studies of R132M in AKc, Arg132 was important for transition-state stabilization (Yan & Tsai, 1991).  The residue 131 to 141 segment was observed to move in substrate holding, which correlates especially with ATP-binding (Schulz et al., 1990).  Lys131 would be essential for catalysis by interacting with MgATP2-.   

The flanking a-helices of the C-terminal domain are denoted on the left side of the proposed AK model, and Lys194 is in the tail section (Fig. 1).  AKc consists of 193 amino acid residues (Kishi et al., 1986), and hAK1 consists of an additional lysine at the 194 position.  The Leu190 residue in AKc appears to participate in components of the MgATP-binding domain (Yoneya et al., 1990).  In hAK1, K194S showed a moderately decreased affinity for MgATP2-.  However, similar mutations in K194V and K194L affected the affinity for MgATP2- or AMP2- differently and decreased catalytic efficiency.  Different hydrophobic changes such as valine and leucine at position 194 may affect the binding of MgATP2- and AMP2- substrates, and the mutants may cause a conformational change of the C-terminal domain by altering its hydrophobicity.  The C-terminal domain might appear to move and to access and interact with the AMP2- substrate forming a ternary complex.  Lys194 would appear to interact not only with MgATP2- but also with AMP2-, possibly indirectly, and thus the flanking a-helices of the C-terminal domain would be essential for catalysis by stabilizing the transition state and holding the substrates and Lys194 could not be replaced by other amino acids.

In a brief summary, steady-state kinetic studies above have led to the following conclusions.  (i) Lys9 would be essential for catalysis by interacting with both MgATP2- and AMP2- but by affecting AMP2- more strongly. (ii) Lys21 would be essential for catalysis by interacting with both MgATP2- and AMP2- but by affecting MgATP2- more strongly.  (iii) Lys27 and Lys131 would appear to play a functional role in catalysis by interacting strongly with MgATP2-.  (iv) Lys31 would appear to interact with MgATP2- and AMP2- at the MgATP2- site and be essential for catalysis.  (v) Lys63 is likely to interact to a much greater extent with MgATP2-.  (vi) Lys194 in the flanking a-helices of the C-terminal domain would interact not only with MgATP2- but also with AMP2- at the MgATP2- site by holding both substrates.  The loss of the e-amino group of lysine (Lys9, Lys21, Lys27, Lys31, Lys63, Lys131, Lys194) resulted in decreased or unchanged affinity for the substrates and reduced catalytic efficiency.  The positive charge of lysine is not replaced in nature and may interact with the negatively charged phosphates of MgATP2- and AMP2-, which in turn may orient the phosphate chains of the two substrates to the proper conformation required for catalysis.  We would predict insertion of important lysine residues as shown in a new AK structural model (Fig. 1).  These lysine residues would take part in a phosphoryl transfer reaction between MgATP2- and AMP2-, in cooperation with the other active site residues, as in the case of arginine residues (Kim et al., 1990).  The AK structural model proposed by X-ray crystallographic studies is obtained in crystal with substrate-inhibitory analogs, the structural data from NMR studies are confirmed in solution.  A new model from the present data by steady-state kinetics (Fig. 1) might stand likely for more dynamic and physiological model in solution than which X-ray crystallographic and NMR studies indicate.  These discrepancy would be expected to be solved in the near future.  To elucidate the structures of ternary complexes of substrates with mutant AK, NMR studies or detailed X-ray crystallographic studies would be required, and circular dichroism and ORD studies must be searched on the structural change for our Lys-mutants.  Our future efforts will be directed toward the other key residues required for the physiological function of to define a interaction with substrates and critical amino acid residues.

 

Acknowledgment

We wish to thank Professor Dr. Albert S. Mildvan (The Johns Hopkins University, School of Medicine) for his kind reading of the manuscript and helpful criticism.

 

References

Atkinson, D. E. (1977) in Cellular Energy Metabolism and Its Regulation, pp. 85-107, Academic Press, Orlando, FL

Ayabe, T., Takenaka, H., Takenaka, O., Takenaka, A., Nagahama, H., Maruyama, H., Yamamoto, A., Nagata, M., Koga, Y., Sumida, M., & Hamada, M. (1996) Biochem. Mol. Biol. Int. 38, 373-381. 

Byeon, I.-J. L., Shi, Z., & Tsai, M-D. (1995). Biochemistry 34, 3172-3182.

Dreusicke, D., Karplus, A., & Schulz, G. E. (1988) J. Mol. Biol. 199, 359-371.

Egner, U., Tomasselli, A. G., & Schulz, G. E. (1987) J. Mol. Biol. 195, 649-658.

Fry, D. C., Byler, D. M., Susi, H., Brown, E. M., Kuby, S. A., & Mildvan, A. S. (1988) Biochemistry 27, 3588-3598.

Fry, D. C., Kuby, S. A., & Mildvan, A. S. (1985) Biochemistry 24, 4680-4694.

Fry, D. C., Kuby, S. A., & Mildvan, A. S. (1987) Biochemistry 26, 1645-1655.

Hamada, M., & Kuby, S. A. (1978) Arch. Biochem. Biophys. 190, 772-779.

Hamada, M., Palmieri, R. H., Russell, G. A., & Kuby, S. A. (1979) Arch. Biochem. Biophys. 195, 155-177.

Hamada, M., Sumida, M., Kurokawa, Y., Sunayashiki-Kusuzaki, K., Okuda H., Watanabe, T., & Kuby, S. A. (1985) J. Biol. Chem. 260, 11595-11602.

Khoo, J. C., & Russel, P. J. (1972) Biochem. Biophys. Acta 268, 98-101.

Kim, H. J., Nishikawa, S., Tokutomi, Y., Takenaka, H., Hamada, M., Kuby, S. A., & Uesugi, S. (1990) Biochemistry 29, 1107-1111.

Kishi, F., Maruyama, M., Tanizawa, Y., & Nakazawa, A. (1986) J. Biol. Chem. 261, 2942-2945.

Liang, P., Jr., G. N. P., & Glaser, M. (1991) Proteins:Struct., Funct., Genet. 9, 28-36.

Lineweaver, H., & Burk, D. (1934) J. Am. Chem. Soc. 56, 658-666.

Lowry, O. H., Rosebrough, N. J., Farr, A. L., & Randall, R. J. (1951) J. Biol. Chem. 193, 265-275.

Matsuura, S., Igarashi, M., Tanizawa, Y., Yamada, M., Kishi, F., Kajii, T., Fujii, H., Miwa, S., Sakurai, M., & Nakazawa, A. (1989) J. Biol. Chem. 264, 10148-10155.

Mildvan, A. S., & Fry, D. C. (1987) Adv. Enzymol. Relat. Areas Mol. Biol. 58, 241-313.

Miwa, S., Fujii, H., Tani, K., Takahashi, K., Tanizawa, T., & Igarashi, T. (1983) Am. J. Hematol. 14, 325-333.

Nakamaye, K., & Ecstein, F. (1986) Nucl. Acids Res. 14, 9679-9698.

Noda, L. (1973) The Enzymes, (Boyer, P. D., ed.) Vol. 8, pp. 279-305, Academic Press, Orlando, FL

Okajima, T., Tanizawa, K., Yoneya, T., & Fukui, T. (1991) J. Biol. Chem. 266, 11442-11447.

Pai, P. K., Ma, Z., & Coleman, P. S. (1992) J. Biol. Chem. 267, 25003-25009.

Reinstein, J., Brune, M., & Wittinghofer, A. (1988) Biochemistry 27, 4712-4720.

Sachsenheimer, W., & Schulz, G. E. (1977) J. Mol. Biol. 114, 23-36.

Sambrook, J., Fritsch, E. F., & Maniatis, T. (1989) Molecular Cloning. A Laboratory Manual. second ed., Chapter 6.

Sanger, F., Nicklen, S., & Coullson, A. R. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 5463-5467.

Sayers, J. R., Krekel, C., & Eckstein, F. (1992) Biotechniques 13, 592-596.

Schulz, G. E., Elzinga, M., Marx, F., & Schirmer, R. H. (1974) Nature 250, 120-123.

Schulz, G. E., Müller, C. W., & Diederichs, K. (1990) J. Mol. Biol. 213, 627-630.

Schulz, G. E., Schiltz, E., Tomasselli, A. G., Frank, R., Brune, M., Wittinghofer, A., & Schirmer, R. H. (1986) Eur. J. Biochem. 161, 127-132.

Smith, G. M., & Mildvan, A. S. (1982) Biochemistry 24, 6119-6123.

Tagaya, M., Yagami, T., & Fukui, T. (1987) J. Biol. Chem. 262, 8257-8261.

Tagaya, M., Yagami, T., Noumi, T., Futai, M., Kishi, F., Nakazawa, A., & Fukui, T. (1989) J. Biol. Chem. 264, 990-994.

Taylor, J. W., Ott, J., & Eckstein, F. (1985) Nucl. Acids Res. 13, 8764-8785.

Tian, G., Yan, H., Jiang, R-T., Kishi, F., Nakazawa, A., & Tsai, M-D. (1990) Biochemistry 29, 4296-4304.

Tomasselli, A. G., Schirmer, R. H., & Noda, L. (1979) Eur. J. Biochem. 93, 257-262.

Toren, A., Brok-Simoni,  F., Ben-Bassat, I., Holtzman, F., Mandel, M., Neumann, Y., & Ramot, B. (1994) Br. J. Hematology 87, 376-380.

Yan, H., Dahnke, T., Zhou, B., Nakazawa, A., & Tsai, M-D. (1990) Biochemistry 29, 10956-10964.

Yan, H., & Tsai., M-D. (1991) Biochemistry 30, 5539-5546.

Yoneya, T., Okajima, T., Tagaya, M., Tanizawa, K., & Fukui, T. (1990) J. Biol. Chem. 265, 21488-21493.

 

 

 

 

 

Legends to Figure

 

 

Fig. 1

The side chains of Lys9, Lys21, Lys27, Lys31, Lys63, Lys131, and Lys194 are depicted in a proposed model of the binding sites for AMP and MgATP in adenylate kinase by Kim et al. (1990).  These lysine residues could take part in a phosphoryl transfer reaction between MgATP2- and AMP2-, in cooperation with the other active site residues, such as arginine residues (Kim et al., 1990).