Human m-NAD(P)-ME [mitochondrial NAD(P)+-dependent ME (malic enzyme)] is a homotetramer, which is allosterically activated by the binding of fumarate. The fumarate-binding site is located at the dimer interface of the NAD(P)-ME. In the present study, we decipher the functional role of the residue Lys57, which resides at the fumarate-binding site and dimer interface, and thus may be involved in the allosteric regulation and subunit–subunit interaction of the enzyme. In the present study, Lys57 is replaced with alanine, cysteine, serine and arginine residues. Site-directed mutagenesis and kinetic analysis strongly suggest that Lys57 is important for the fumarate-induced activation and quaternary structural organization of the enzyme. Lys57 mutant enzymes demonstrate a reduction of Km and an elevation of kcat following induction by fumarate binding, and also display a much higher maximal activation threshold than WT (wild-type), indicating that these Lys57 mutant enzymes have lower affinity for the effector fumarate. Furthermore, mutation of Lys57 in m-NAD(P)-ME causes the enzyme to become less active and lose co-operativity. It also increased K0.5,malate and decreased kcat values, indicating that the catalytic power of these mutant enzymes was significantly impaired following mutation of Lys57. Analytical ultracentrifugation analysis demonstrates that the K57A, K57S and K57C mutant enzymes dissociate predominantly into dimers, with some monomers present, whereas the K57R mutant forms a mixture of dimers and tetramers, with a small amount of the enzyme in monomeric form. The dimeric form of these Lys57 mutants, however, cannot be reconstituted into tetramers with the addition of fumarate. Modelling structures of the Lys57 mutant enzymes show that the hydrogen bond network in the dimer interface where Lys57 resides may be reduced compared with WT. Although the fumarate-induced activation effects are partially maintained in these Lys57 mutant enzymes, the mutant enzymes cannot be reconstituted into tetramers through fumarate binding and cannot recover their full enzymatic activity. In the present study, we demonstrate that the Lys57 residue plays dual functional roles in the structural integrity of the allosteric site and in the subunit–subunit interaction at the dimer interface of human m-NAD(P)-ME.

INTRODUCTION

ME (malic enzyme) is an important enzyme that catalyses the oxidative decarboxylation of malate to yield CO2 and pyruvate with the concomitant production of NAD(P)H [1,2]. A divalent metal ion (Mn2+ or Mg2+) is required for this enzymatic reaction. These enzymes are highly conserved between species with regard to protein sequences and structural topology [25]. In mammals, MEs can be divided into three isoforms according to their cofactor preference. They are c-NADP-ME (cytosolic NADP+-dependent ME) [57], m-NADP-ME (mitochondrial NADP+-dependent ME) [8], and m-NAD(P)-ME [mitochondrial NAD(P)+-dependent ME] [913]. m-NAD(P)-ME has dual coenzyme specificity, utilizing both NAD+ and NADP+ as coenzymes, although the preferred coenzyme under physiological conditions is NAD+ [2,14]. The physiological role of c-NADP-ME and m-NADP-ME isoforms is mainly to produce NADPH for fatty acid biosynthesis. The m-NAD(P)-ME isoform primarily generates NADH as the reducing equivalent in energy production and this enzyme may be associated with rapid tumour growth through the NADH and pyruvate products in glutaminolysis [1519].

The m-NAD(P)-ME isoform is distinct from the other two mammalian isoforms in that it is a regulatory enzyme with a multifaceted control system to moderate its catalytic activity [2022]. The enzyme displays a positive co-operative manner of binding the substrate L-malate, and it can be allosterically activated by fumarate binding [14,16,21,23]. Furthermore, it is inhibited by ATP through an active-site competition mechanism [2426]. This ATP inhibition may have evolved through the particular role of the enzyme in the pathways of malate and glutamine oxidation in tumour mitochondria [27,28].

The crystal structures of MEs reveal that they have a quaternary structure of a dimer of dimers (Figure 1A), and suggest that these enzymes may be a new class of oxidative decarboxylases with a novel backbone structure [2,4,2933]. The structures of human m-NAD(P)-ME reveal that there are two regulatory sites in addition to the active site (Figure 1A) [2,21,22,24]. One, called the exo-site, is situated at the tetramer interface, and is occupied by an NAD or ATP molecule. The other, located at the dimer interface, is occupied by fumarate [21]. In Ascaris suum m-NAD(P)-ME, a separate allosteric site is also found at the dimer interface [23,34,35]. Figure 1(B) shows the fumarate-binding site at the dimer interface. In the structure, fumarate interacts directly with Arg67 and Arg91. Mutation analysis confirmed that Arg67 and Arg91 are crucial for fumarate activation. However, both Arg67 and Arg91 are also conserved among non-allosteric ME isoforms. Therefore additional factors may participate in the mechanism of fumarate activation control. Besides interacting with fumarate, Arg67 is also ion-paired with Glu59. Earlier studies have shown that the E59L mutation totally abolishes the fumarate-induced enzyme activation, and demonstrated the remarkable effects of Glu59 on the allosteric activation of the enzyme [21].

Fumarate-binding site of human m-NAD(P)-ME

Figure 1
Fumarate-binding site of human m-NAD(P)-ME

(A) Tetramer of human m-NAD(P)-ME (PDB code 1PJ3). The active site in each subunit, the exo-site in the tetramer interface and the fumarate-binding site in the dimer interface are indicated as a ball-and-stick model. NAD+ in the active site is coloured in blue and that in the exo-site is coloured in light blue; fumarate is coloured in yellow. (B) Lys57 in the fumarate-binding pocket and in the dimer interface. Both (A) and (B) were produced using PyMOL (DeLano Scientific; http://pymol.sourceforge.net/). (C) Multiple sequence alignments of three clusters of ME isoforms around the fumarate-binding region in the dimer interface. Amino acid sequences of MEs were obtained by BLAST [45], and alignments were generated by ClustalW [46]. (C) was generated using the BioEdit sequence alignment editor program [47].

Figure 1
Fumarate-binding site of human m-NAD(P)-ME

(A) Tetramer of human m-NAD(P)-ME (PDB code 1PJ3). The active site in each subunit, the exo-site in the tetramer interface and the fumarate-binding site in the dimer interface are indicated as a ball-and-stick model. NAD+ in the active site is coloured in blue and that in the exo-site is coloured in light blue; fumarate is coloured in yellow. (B) Lys57 in the fumarate-binding pocket and in the dimer interface. Both (A) and (B) were produced using PyMOL (DeLano Scientific; http://pymol.sourceforge.net/). (C) Multiple sequence alignments of three clusters of ME isoforms around the fumarate-binding region in the dimer interface. Amino acid sequences of MEs were obtained by BLAST [45], and alignments were generated by ClustalW [46]. (C) was generated using the BioEdit sequence alignment editor program [47].

Our previous work suggests that the electrostatic balance in the fumarate-binding pocket may be a crucial factor governing the regulatory mechanism of fumarate-induced activation [17]. In the present study, we further examine a cationic residue Lys57, which is ion-paired with Glu59 (Figure 1B). Both Lys57 and Glu59 uniquely exist in the m-NAD(P)-ME isoform, but not in the non-allosteric isoforms. Thus we investigated the functional role of Lys57 within the allosteric site. Also of interest is the fact that Lys57 is situated at the dimer interface. The side chain of Lys57 is not only ion-paired with Glu59, but also hydrogen-bonded with the Pro216 and Tyr218 residues from another subunit. Therefore we also examined the possible role of Lys57 in subunit–subunit interactions. Based on multiple sequence alignments, Lys57 is substituted to cysteine and serine residues, found in c-NADP-ME and m-NADP-ME respectively (Figure 1C). The K57A and K57R mutations are designed to examine the effect of positive charge on this residue. Detailed kinetic and analytical ultracentrifugation analysis shows that the Lys57-involved hydrogen-bond network at the dimer interface plays dual functional roles in the allosteric regulation and subunit–subunit interaction of the human m-NAD(P)-ME.

MATERIALS AND METHODS

Preparation of recombinant MEs

Detailed expression and purification protocols for human m-NAD(P)-ME have been described previously [4,9,29]. Briefly, the m-NAD(P)-ME was cloned into the expression vector (pRH281) under an inducible trp promoter system. The ampicillin-resistant vector was transformed into Escherichia coli BL21 cells for enzyme overexpression. Expression of m-NAD(P)-ME was induced with 50 μg/ml IAA (β-indole-3-acetic acid) and the cells were incubated at 25 °C overnight. After induction by IAA, cells were centrifuged at 6000 g at 4 °C for 15 min. The supernatant was removed and the cell pellets were resuspended in buffer A [3 mM MgCl2, 1 mM MnCl2, 2 mM 2-mercaptoethanol, 0.2 mM EDTA and 30 mM Tris/HCl (pH 7.4)]. After sonication to lyse the cells, the lysate was centrifuged (15000 g at 4 °C for 20 min) and the supernatant was collected for further purification. An anionic exchange, DEAE–Sepharose (Amersham Biosciences) column, followed by an ATP–agarose affinity chromatography (Sigma) column was utilized in the enzyme purification. The DEAE–Sepharose was equilibrated with buffer A and the supernatant was then added to the column. The lysate–DEAE was washed in a stepwise procedure (buffer A with steps of 20, 40, 60 and 80 mM NaCl) to minimize the association of unwanted proteins. Finally, human m-NAD(P)-ME was eluted with buffer B (buffer A with 100 mM NaCl). After dialysis with buffer A, the enzyme was then loaded into the ATP–agarose affinity chromatography, which was pre-equilibrated with buffer A. The enzyme was finally eluted using buffer C (buffer A with 4 mM NAD+). The purified enzyme was then buffer-exchanged and concentrated in 30 mM Tris/HCl (pH 7.4) with 2 mM 2-mercaptoethanol by a centrifugal filter device (Amicon Ultra-15, Millipore) with a molecular mass cut-off of 30 kDa. The enzyme purity was examined by SDS/PAGE, and the protein concentrations were estimated using the Bradford method [36].

Site-directed mutagenesis

Site-directed mutagenesis was carried out using the QuikChange® kit (Stratagene). The purified DNA of human m-NAD(P)-ME was used as a template, and the primers with desired codon changes were employed to change Lys57 to an alanine, cysteine, serine or arginine residue using a high-fidelity Pfu DNA polymerase in the PCR reaction. Primers spanning the mutation site were 25–45-mers, which is considered necessary for specific binding of template DNA. The synthetic oligonucleotides used in the site-directed mutagenesis experiments were: 5′-CTTCAAGGACTTCTACCTCCCGCGATAGAGACACAAGATATTC-3′ for K57A, 5′-CTTCAAGGACTTCTACCTCCCTGCATAGAGACACAAGATATTC-3′ for K57C, 5′-CTTCAAGGACTTCTACCTCCCTCTATAGAGACACAAGATATTC-3′ for K57S and 5′-CTTCAAGGACTTCTACCTCCCCGTATAGAGACACAAGATATTC-3′ for K57R. The nucleotides underlined and marked in bold indicate the mutation positions. The PCR reaction was performed at 95 °C for 30 s, 55 °C for 1 min, and 68 °C for 2 min/kb of plasmid length using Pfu DNA polymerase for 16 cycles. After 16–18 temperature cycles, mutated plasmids were created and staggered nicks were made. The vector templates were digested with the restriction enzyme DpnI and then the nicked DNA containing the desired mutations was transformed into E. coli XL-1 cells. All mutation sites were verified by automated sequencing.

Enzyme kinetic analysis

The enzymatic reaction of ME was measured by recording the NADH production. The reaction cocktail was composed of 50 mM Tris/HCl (pH 7.4), 40 mM malate, 2.0 mM NAD+ and 10 mM MgCl2 in a total volume of 1 ml in the absence or presence of fumarate. The absorbance at 340 nm at 30 °C was immediately recorded after the enzyme was added to the reaction mixture and monitored continuously in a Beckman DU 7500 spectrophotometer. A molar absorption coefficient of 6.22 mM−1·cm−1 for NADH was employed to calculate the initial velocities and kcat values. Apparent Michaelis constants of the substrate and coenzymes were determined by changing the concentration of one substrate (or coenzyme) in the region of its Km value and, at the same time, keeping the other components constant under the saturation condition.

The sigmoidal curves of [malate] against initial velocity were fitted into the Hill equation. Further analysis revealed the K0.5 value, the substrate concentration at half-maximal velocity, and the Hill coefficient (h), both of which were utilized to evaluate the degree of co-operativity (eqn 1):

 
formula
(1)

All data-fitting work was carried out using the Sigma Plot 10.0 program (Jandel).

Quaternary structure determination by analytical ultracentrifugation

Sedimentation velocity experiments were executed using a Beckman Optima XL-A analytical ultracentrifuge. Sample (400 μl) and buffer (420 μl) solutions were loaded into the double sector centrepiece individually and were housed in a Beckman An-50 Ti rotor. Experiments were carried out at 20 °C and a rotor speed of 42000 rev./min was used. Protein samples were monitored by UV absorbance at 280 nm in a continuous mode with a time interval of 480 s and a step size of 0.002 cm. Numerous scans at different time points were fitted to a continuous size-distribution model by the program SEDFIT [3741]. All size distributions were resolved on a confidence level of P=0.95, a best-fitted average anhydrous frictional ratio (f/f0), and a resolution N of 200 sedimentation coefficients between 0.1 and 20.0 S.

Molecular modelling

The template used for the homology modelling was the crystal structure of human m-NAD(P)-ME in a pentary complex with its natural substrate pyruvate, its cofactors NAD+, Mn2+ and the allosteric activator fumarate (PDB ID: 1PJ3) [33]. The tetrameric models of WT (wild-type) ME and four mutants (K57A, K57C, K57S and K57R) were created using MODELLER version 9.5 [42]. From 100 models generated from each modelling process, we selected five models with the best molpdf (molecular probability density function) scores. We then submitted the models to the PISA (protein interfaces, surfaces and assembly) service at the European Bioinformatics Institute (http://www.ebi.ac.uk/msd-srv/prot_int/pistart.html) [43] to measure the tetramer interface area and the number of potential hydrogen bonds across the interface.

RESULTS

Activating effect of fumarate on the human WT and Lys57 mutant m-NAD(P)-ME

The initial velocity of m-NAD(P)-ME measured under various concentrations of fumarate displayed a hyperbolic kinetic pattern (Figure 2). The WT enzyme reached a maximal activation of approx. 1.6-fold at the saturated fumarate concentration (Figure 2, ●), with an apparent activation constant (KA) of 0.20±0.03 mM (Table 1). The apparent KA values of K57R, K57A and K57S mutant enzymes were higher than that of WT by approx. 10-fold (Table 1). The enzyme activity of K57R can be stimulated by fumarate with an apparent KA value of 2.8±0.4 mM (Table 1), and its maximal enzyme activity was recovered to approx. 60% of WT (Figure 2, ■). The K57A and K57S enzymes had similar titration curves with fumarate (Figure 2, ○ and Δ respectively). The apparent KA values of K57A and K57S were 3.36±0.73 mM and 2.36±0.27 mM respectively (Table 1), and their enzyme activity with saturated fumarate was restored to approx. half of WT levels. The K57C enzyme demonstrated the lowest enzyme activity among these mutants (Figure 2, ▼). Although the K57C mutant enzyme can be stimulated by fumarate, the maximal enzyme activity was only one-third of WT levels and it had an apparent KA value (16.8±5.1 mM) much greater than that of WT by approx. 80-fold (Table 1). These preliminary results suggested that Lys57, although not a direct fumarate-binding ligand, can influence the rate of fumarate-induced activation and plays an important role in fumarate-binding affinity. These Lys57 mutant enzymes had a higher fumarate-induced activation threshold than WT, suggesting that they had lower affinity for the effector fumarate.

Table 1
Kinetic parameters for the human WT and Lys57 mutant m-NAD(P)-ME

Values are means±S.E.M. −, without fumarate; +, at saturated fumarate concentration (4 mM for WT; 20 mM for K57A, K57S and K57R; 65 mM for K57C).

WT/mutant Fumarate K0.5,malate* (mM) Km,NAD (mM) kcat (s−1kcat/K0.5,malateKm,NAD (mM−2·s−1Fold increase in kcat/K0.5,malateKm,NAD h Kact,app† (mM) 
WT − 12.0±0.8 1.57±0.30 225.4±13.1 11.96±2.52 1.8±0.1 0.20±0.03 
 3.7±0.5 0.53±0.06 283.7±9.2 144.76±25.93 1.0±0.1  
K57A − 211.7±35.8 1.44±0.15 59.3±1.2 0.19±0.04 0.015 1.0±0.2 3.36±0.73 
 8.6±0.6 0.57±0.06 114.5±1.6 23.36±2.97 0.16 1.1±0.1  
K57C − 78.9±9.3 7.51±2.08 52.7±5.8 0.09±0.03 0.008 1.0±0.1 16.8±5.1 
 32.0±1.5 0.63±0.14 125.2±4.9 6.21±1.43 0.04 1.0±0.1  
K57S − 173.0±13.7 3.05±0.24 59.4±1.6 0.11±0.01 0.009 1.1±0.6 2.36±0.27 
 9.7±0.5 1.62±0.38 99.1±1.8 6.31±1.52 0.04 1.0±0.1  
K57R − 93.7±12.3 4.09±0.73 68.5±3.1 0.18±0.04 0.015 1.1±0.2 2.82±0.40 
 5.0±0.5 1.29±0.20 161.3±3.4 25.01±4.64 0.17 1.2±0.1  
WT/mutant Fumarate K0.5,malate* (mM) Km,NAD (mM) kcat (s−1kcat/K0.5,malateKm,NAD (mM−2·s−1Fold increase in kcat/K0.5,malateKm,NAD h Kact,app† (mM) 
WT − 12.0±0.8 1.57±0.30 225.4±13.1 11.96±2.52 1.8±0.1 0.20±0.03 
 3.7±0.5 0.53±0.06 283.7±9.2 144.76±25.93 1.0±0.1  
K57A − 211.7±35.8 1.44±0.15 59.3±1.2 0.19±0.04 0.015 1.0±0.2 3.36±0.73 
 8.6±0.6 0.57±0.06 114.5±1.6 23.36±2.97 0.16 1.1±0.1  
K57C − 78.9±9.3 7.51±2.08 52.7±5.8 0.09±0.03 0.008 1.0±0.1 16.8±5.1 
 32.0±1.5 0.63±0.14 125.2±4.9 6.21±1.43 0.04 1.0±0.1  
K57S − 173.0±13.7 3.05±0.24 59.4±1.6 0.11±0.01 0.009 1.1±0.6 2.36±0.27 
 9.7±0.5 1.62±0.38 99.1±1.8 6.31±1.52 0.04 1.0±0.1  
K57R − 93.7±12.3 4.09±0.73 68.5±3.1 0.18±0.04 0.015 1.1±0.2 2.82±0.40 
 5.0±0.5 1.29±0.20 161.3±3.4 25.01±4.64 0.17 1.2±0.1  
*

Values derived from Figure 2.

Values derived from Figure 1.

Activating effect of fumarate on the human WT and Lys57 mutant m-NAD(P)-ME

Figure 2
Activating effect of fumarate on the human WT and Lys57 mutant m-NAD(P)-ME

The assay mixture contained ME (1.5 μg), 40 mM malate, 10 mM MgCl2 and 2 mM NAD+ with various fumarate concentrations as indicated. ●, WT enzyme; ○, K57A enzyme; ▼, K57C enzyme; ◃, K57S enzyme; ■, K57R enzyme.

Figure 2
Activating effect of fumarate on the human WT and Lys57 mutant m-NAD(P)-ME

The assay mixture contained ME (1.5 μg), 40 mM malate, 10 mM MgCl2 and 2 mM NAD+ with various fumarate concentrations as indicated. ●, WT enzyme; ○, K57A enzyme; ▼, K57C enzyme; ◃, K57S enzyme; ■, K57R enzyme.

Kinetic parameters of the human WT and Lys57 mutant m-NAD(P)-ME

Since the Lys57 mutant enzymes required more fumarate to achieve their maximal enzyme activities, the kinetic parameters of these mutant enzymes were determined using the respective saturated fumarate concentration. Table 1 shows the kinetic parameters of WT and Lys57 mutant enzymes, determined with or without fumarate. In the absence of fumarate, the Lys57 mutant enzymes displayed a K0.5,malate value which was greater than WT by approx. 6–17-fold. Furthermore, the kcat value of these mutants was much less than WT, only 20∼30% of WT levels. The specificity constant (kcat/K0.5,malateKm,NAD) of the Lys57 mutant enzymes was only 1% of WT levels, demonstrating a much decreased catalytic efficiency. These results indicated that the enzyme was significantly impaired in its catalytic power by mutation of Lys57, thus these Lys57 mutants exhibited a less active enzyme form.

Fumarate was able to partially activate these Lys57 mutant enzymes. In the presence of fumarate, the high Km,NAD and K0.5,malate values of Lys57 mutant enzymes were reduced and the low kcat values were elevated (Table 1). For K57A, K57S and K57R, the K0.5,malate value can be reduced to a level similar to WT, except for the K57C mutant. The K0.5,malate value of K57C, although reduced approx. 2.5-fold by the addition of fumarate, is still 10-fold higher than WT. Meanwhile, the kcat values of these Lys57 mutants, although elevated by the addition of fumarate, were still lower than WT by 1.8–3-fold. The kcat/K0.5,malateKm,NAD values of the K57A and K57R enzymes were elevated to 16% of WT levels, but the K57C and K57S enzymes were still only 4% of the WT enzyme levels (Table 1), demonstrating that the catalytic efficiencies of these Lys57 mutants was still much less than WT. The key fact from these kinetic data is that the Lys57 mutant enzymes have unusually large K0.5,malate values which can be decreased to a level similar to that of WT by the addition of fumarate; however, the 50% fumarate saturation levels for these mutants is much higher than WT. These kinetic results, when taken together, implied that not only the allosteric, but also the catalytic site of the enzyme, were influenced by the replacement of Lys57. A conformational change which is unfavourable for catalysis and allosteric activation of the enzyme may occur in these Lys57 mutants.

Dependence curve of malate against initial velocity in the presence of fumarate for the human WT and Lys57 mutant m-NAD(P)-ME

The human m-NAD(P)-ME acts in a co-operative fashion upon binding of malate. The initial rates of the WT enzyme measured under various concentrations of malate demonstrated sigmoidal kinetics (Figure 3, ●). The co-operativity of malate binding, however, can be abolished by fumarate. In the presence of fumarate, the curve changed from sigmoidal to hyperbolic (Figure 3, ▼), and the h value was extensively reduced from 1.8 to 1 (Table 1). For these Lys57 mutant enzymes, the co-operativity of malate binding seen in WT almost fully vanished (Figures 3B–3E, ●), indicating that these mutants had become non-co-operative enzymes. The h value of these mutants was 1.0 both with and without fumarate (Table 1).

Dependence curve of malate against initial velocity in the presence of fumarate for the human WT and Lys57 mutant m-NAD(P)-ME

Figure 3
Dependence curve of malate against initial velocity in the presence of fumarate for the human WT and Lys57 mutant m-NAD(P)-ME

The assay mixture contained ME (1.5 μg), 10 mM MgCl2 and 2 mM NAD+ with various malate concentrations. ●, Without fumarate; ○, with low concentrations of fumarate; ▼, with high concentrations of fumarate. Fumarate concentrations used in the experiments for these mutants are as indicated in the Figure.

Figure 3
Dependence curve of malate against initial velocity in the presence of fumarate for the human WT and Lys57 mutant m-NAD(P)-ME

The assay mixture contained ME (1.5 μg), 10 mM MgCl2 and 2 mM NAD+ with various malate concentrations. ●, Without fumarate; ○, with low concentrations of fumarate; ▼, with high concentrations of fumarate. Fumarate concentrations used in the experiments for these mutants are as indicated in the Figure.

In addition to the non-co-operative binding of malate, these Lys57 mutant enzymes, in the absence of fumarate, showed very low enzyme activities and required a very high malate concentration to reach their Vmax (Figures 3B–3E, ●). However, in saturated fumarate, these Lys57 mutants could still be greatly stimulated by fumarate and required less malate to reach their maximal velocities (Figures 3B–3E, ▼).

Quaternary structure of the human WT and Lys57 mutant m-NAD(P)-ME

Lys57 is located at the dimer interface and is hydrogen-bonded to Pro216 and Tyr218 from the other subunit. Kinetic analysis indicated that the co-operativity of malate binding was totally lost in these Lys57 mutant enzymes. Since the tetramer organization may be involved in the co-operative behaviour of this enzyme [2], we used analytical ultracentrifugation to examine possible changes in the quaternary structure of these Lys57 mutants. Figure 4 shows the continuous sedimentation coefficient distribution of WT and Lys57 mutants. The sedimentation coefficients of 6.5 S and 9.0 S, corresponding to the molecular masses of 124 and 248 kDa, were the dimeric and tetrameric forms of the protein respectively. The WT m-NAD(P)-ME exists in a dimer–tetramer equilibrium (Figure 4A), and this dimer is possibly an AB dimer rather than an AD dimer, since the contacts in the tetramer interface are weaker than those in the dimer interface (Figure 1A) [2]. In contrast, the K57A, K57C and K57S mutants display predominantly a dimeric quaternary structure with a small amount of monomers, tetramers and polymers (Figures 4B–4D), indicating that the mutation of Lys57 leads to a disruption of the quaternary structure at the dimer interface, leading to the dissociation of the enzyme. The dimeric form in the Lys57 mutants, however, might be an AD dimer rather than an AB dimer, and the AB dimer seen in WT may be dissociated into monomers in these Lys57 mutants. The quaternary structure of the K57R enzyme, similar to that of WT, also exists in a dimer–tetramer equilibrium (Figure 4E), suggesting that the positive-charge on residue 57 may play a role in maintaining the quaternary structure organization in the dimer interface of the enzyme.

Continuous sedimentation coefficient distribution of the human WT and Lys57 mutant m-NAD(P)-ME

Figure 4
Continuous sedimentation coefficient distribution of the human WT and Lys57 mutant m-NAD(P)-ME

The enzyme concentration used in the experiment was 0.3 mg/ml in 50 mM Tris/HCl buffer (pH 7.4) at 20 °C. The solid and broken lines represent the enzyme without or with fumarate (4 mM) respectively. (A) WT enzyme; (B) K57A enzyme; (C) K57C enzyme; (D) K57S enzyme; (E) K57R enzyme. D, dimer; T, tetramer.

Figure 4
Continuous sedimentation coefficient distribution of the human WT and Lys57 mutant m-NAD(P)-ME

The enzyme concentration used in the experiment was 0.3 mg/ml in 50 mM Tris/HCl buffer (pH 7.4) at 20 °C. The solid and broken lines represent the enzyme without or with fumarate (4 mM) respectively. (A) WT enzyme; (B) K57A enzyme; (C) K57C enzyme; (D) K57S enzyme; (E) K57R enzyme. D, dimer; T, tetramer.

Fumarate can assist in reconstituting the dimeric enzymes into tetramers, suggesting that fumarate may be involved in the tetramer reorganization of the enzyme and a function that may facilitate enzyme catalysis [2]. The dimer–tetramer equilibrium of the WT enzyme was shifted following addition of fumarate (Figure 4A, broken line). The K57A, K57C and K57S mutants occurred as dimers, but could not be reconstituted into tetramers like WT (results not shown). Although the K57R enzyme is in dimer–tetramer equilibrium with very few monomers, it could not be reconstituted from dimers into tetramers by addition of fumarate (Figure 4E, broken lines). These results suggest that the tetramer organization was truly perturbed in these mutants.

Modeling of the Lys57 mutant m-NAD(P)-ME

To assess the difference in the dimer interface between the WT and Lys57 mutant structures, we built tetrameric models for K57A, K57C, K57S and K57R mutants by using the comparative modelling software MODELLER [42]. The models were based on the crystal structure of human m-NAD(P)-ME enzyme in a pentary complex with its natural product pyruvate, its cofactors NAD+, Mn2+ and its allosteric activator fumarate (PDB ID: 1PJ3) [33]. MODELLER employs methods of conjugate gradients and molecular dynamics with simulated annealing to optimize the possible side-chain conformation in the protein–protein interface. We also used a similar procedure to create the WT models. Subsequently, we compared the WT models with the mutant models by calculating the tetramer interface area and the number of potential hydrogen bonds across the interface (Table 2). Mutation of Lys57 to alanine, cysteine or serine caused an interface area loss between monomers A and B (IAB), and between monomers C and D (ICD). The K57R mutation did not lead to the area loss in IAB or in ICD. Interestingly, there was no significant difference in the interface between monomers A and D (IAD), or the interface between monomers B and C (IBC) when we compared the mutant models with the WT models. Mutation of Lys57 to alanine, cysteine or serine also reduced the number of hydrogen bonds across the interfaces IAB and ICD. However, the overall influence of the K57R mutation on the number of intermolecular hydrogen bonds across IAB and ICD was inconclusive, as we observed an increase of intermolecular hydrogen bonds in IAB, but a decrease in ICD.

Table 2
Dimer and tetramer interface of the human WT and Lys57 mutant m-NAD(P)-ME

Values are means±S.E.M.

WT/mutant Interface Interface area (Å2)† Number of hydrogen bonds 
WT IAB2030.1±44.1 16.8±1.6 
 ICD 2072.8±28.3 17.2±1.1 
 IAD 899.0±22.8 10.6±1.7 
 IBC 862.3±22.8 10.4±0.5 
K57A IAB 1964.1±40.2 13.2±1.5 
 ICD 1995.2±22.4 13.0±2.2 
 IAD 884.1±36.0 9.4±1.1 
 IBC 854.6±23.2 9.4±1.5 
K57C IAB 2011.5±15.2 15.2±2.2 
 ICD 2037.8±44.3 14.4±1.5 
 IAD 898.2±17.0 10.2±1.8 
 IBC 860.4±9.5 10.0±1.4 
K57S IAB 2016.5±43.1 15.6±2.1 
 ICD 1985.9±42.7 14.4±2.1 
 IAD 890.5±27.2 10.2±2.8 
 IBC 853.2±8.4 9.4±1.1 
K57R IAB 2063.9±26.0 18.4±2.1 
 ICD 2068.7±50.8 15.2±3.1 
 IAD 889.4±32.2 10.2±1.8 
 IBC 850.1±21.9 10.0±1.0 
WT/mutant Interface Interface area (Å2)† Number of hydrogen bonds 
WT IAB2030.1±44.1 16.8±1.6 
 ICD 2072.8±28.3 17.2±1.1 
 IAD 899.0±22.8 10.6±1.7 
 IBC 862.3±22.8 10.4±0.5 
K57A IAB 1964.1±40.2 13.2±1.5 
 ICD 1995.2±22.4 13.0±2.2 
 IAD 884.1±36.0 9.4±1.1 
 IBC 854.6±23.2 9.4±1.5 
K57C IAB 2011.5±15.2 15.2±2.2 
 ICD 2037.8±44.3 14.4±1.5 
 IAD 898.2±17.0 10.2±1.8 
 IBC 860.4±9.5 10.0±1.4 
K57S IAB 2016.5±43.1 15.6±2.1 
 ICD 1985.9±42.7 14.4±2.1 
 IAD 890.5±27.2 10.2±2.8 
 IBC 853.2±8.4 9.4±1.1 
K57R IAB 2063.9±26.0 18.4±2.1 
 ICD 2068.7±50.8 15.2±3.1 
 IAD 889.4±32.2 10.2±1.8 
 IBC 850.1±21.9 10.0±1.0 
*

For the abbreviation used in this column, IXY means the interface between monomers X and Y in a tetrameric malic enzyme.

1 Å=0.1 nm.

DISCUSSION

Human m-NAD(P)-ME is characterized as an allosteric enzyme, which differentiates it from the other two isoforms. Fumarate has been recognized as the allosteric activator for human m-NAD(P)-ME by reducing the Km values of the substrates [13,21,23,28]. Structural analysis indicates that the allosteric site is located at the dimer interface of the enzyme and the binding network of fumarate in this site has been identified [21]. In the fumarate-binding pocket, two arginine residues, Arg67 and Arg91, are the direct ligands for fumarate. An anionic residue, Glu59, which is ion-paired with Arg67 in this site, is also important for the fumarate activation (Figure 5A). Our recent studies have shown that Glu59 is involved, not only in the allosteric regulation, but also the subunit–subunit interactions of the enzyme [44]. An interface cationic residue, Lys57, which is ion-paired with Glu59, also forms a hydrogen bond network with Pro216 and Tyr218 from another subunit (Figure 5A). In the present study, we explored the functional role of Lys57 in the allosteric site and in the dimer interface of the enzyme.

Hydrogen-bond network in the fumarate-binding site of the human WT and Lys57 mutant m-NAD(P)-ME

Figure 5
Hydrogen-bond network in the fumarate-binding site of the human WT and Lys57 mutant m-NAD(P)-ME

A and B subunits are represented as green and cyan ribbons respectively. (A) WT m-NAD(P)-ME; (B) K57A m-NAD(P)-ME; (C) K57C m-NAD(P)-ME; (D) K57S m-NAD(P)-ME; (E) K57R m-NAD(P)-ME. The modelled structures of Lys57 mutants (BE) were created using MODELLER version 9.5 [42]. These Figures were generated using PyMOL (DeLano Scientific; http://pymol.sourceforge.net/). The broken lines represent the polar contacts between fumarate and amino acid residues.

Figure 5
Hydrogen-bond network in the fumarate-binding site of the human WT and Lys57 mutant m-NAD(P)-ME

A and B subunits are represented as green and cyan ribbons respectively. (A) WT m-NAD(P)-ME; (B) K57A m-NAD(P)-ME; (C) K57C m-NAD(P)-ME; (D) K57S m-NAD(P)-ME; (E) K57R m-NAD(P)-ME. The modelled structures of Lys57 mutants (BE) were created using MODELLER version 9.5 [42]. These Figures were generated using PyMOL (DeLano Scientific; http://pymol.sourceforge.net/). The broken lines represent the polar contacts between fumarate and amino acid residues.

The Lys57-involved hydrogen-bond network is important for the structural integrity of the allosteric site

Site-directed mutagenesis and kinetic analysis strongly suggest that Lys57 has profound effects on the enzymatic allosteric regulation and on the enzyme catalysis. The Lys57 mutant enzymes display a much higher threshold for their maximal activation compared with WT, indicating that the conformation of the allosteric site is significantly changed in these mutants and that the binding affinity for fumarate may be significantly decreased. Computer modelling of the Lys57 mutants suggests that the hydrogen-bond network in the fumarate-binding site may be partially abolished (Figures 5B–5E) which may lead to conformational changes that are disadvantageous to fumarate-induced activation.

Furthermore, mutation of Lys57 in m-NAD(P)-ME causes the enzyme to become a less-active enzyme with a very large K0.5,malate, small kcat value and low catalytic efficiency (kcat/K0.5,malateKm,NAD) (Table 1), implying that the conformation of the active site has been considerably changed. This may result from the loss of the hydrogen-bond network in the fumarate-binding site, which may be detrimental to the active-site conformation of the enzyme. Despite the fact that these structural changes in the Lys57 mutants may be partially readjusted by addition of fumarate [which leads to a reduction in K0.5,malate and an elevation in kcat values and enzyme activity (Table 1 and Figures 1 and 3)], these Lys57 mutants are still less active than the WT enzyme, highlighting the importance of the Lys57-involved binding network. Indeed, mutation of the corresponding residue in c-NADP-ME (S57K mutant) did not cause any change in kinetic properties or in the quaternary structure organization (results not shown). Thus, Lys57 has an unique role in the m-NAD(P)-ME isoform.

The Lys57-involved hydrogen-bond network is important for the subunit–subunit interaction at the dimer interface

In addition to the fumarate-induced activation, Lys57 may be involved in the subunit–subunit interactions. Indeed, we have shown that the quaternary structure of Lys57 mutants has been significantly changed (Figure 4). Modelling structures of the Lys57 mutant enzymes demonstrate that the Lys57-involved subunit–subunit interactions in the dimer interface of the WT enzyme, including the interface area and the number of hydrogen bonds, may be reduced (Table 2 and Figures 5B–5E). Even though the Lys57-involved hydrogen-bond network between subunit A and B was changed, these mutant enzymes still partially retained the ability to be activated by fumarate. Nevertheless, these Lys57 mutants cannot be reconstituted into tetramers by the addition of fumarate and cannot recover their full enzymatic activity. Furthermore, perturbation of the tetramer organization between subunits A and B truly caused the enzyme to lose its co-operativity. Conservation of the positive charge on this residue (K57R mutant; Figure 4E) may be helpful in maintaining the quaternary structure of the enzyme, but the co-operativity of the enzyme is still abolished. Thus these Lys57 mutants no longer display a sigmoidal kinetic behaviour, the interactions between subunits are perturbed and the co-operativity is abolished. Taken together, these results suggest that the Lys57-involved hydrogen-bond network at the dimer interface of the enzyme is crucial for the quaternary structural organization of the enzyme, and for enzyme regulation.

In conclusion, we suggest that Lys57 has dual functional roles in the allosteric site and at the dimer interface of the human m-NAD(P)-ME. Lys57, in the fumarate-binding site, contributes to the conformational integrity of the allosteric site. In addition, at the dimer interface, the Lys57-involved hydrogen-bond network stabilizes the subunit–subunit interaction.

Abbreviations

     
  • c-NADP-ME

    cytosolic NADP+-dependent malic enzyme

  •  
  • IAA

    β-indole-3-acetic acid

  •  
  • ME

    malic enzyme

  •  
  • m-NADP-ME

    mitochondrial NADP+-dependent ME

  •  
  • m-NAD(P)-ME

    mitochondrial NAD(P)+-dependent ME

  •  
  • WT

    wild-type

FUNDING

This work was supported by the National Science Council, R.O.C. [grant number NSC-96-2311-B-005-005 (to H.-C.H.)], and partly by the Ministry of Education, Taiwan, R.O.C. under the ATU (Aiming for Top University) plan.

References

References
1
Cleland
W. W.
Chemical mechanism of malic enzyme as determined by isotope effects and alternate substrates
Protein Pept. Lett.
2000
, vol. 
7
 (pg. 
305
-
312
)
2
Chang
G. G.
Tong
L.
Structure and function of malic enzymes, a new class of oxidative decarboxylases
Biochemistry
2003
, vol. 
42
 (pg. 
12721
-
12733
)
3
Frenkel
R.
Regulation and physiological functions of malic enzymes
Curr. Top. Cell. Regul.
1975
, vol. 
9
 (pg. 
157
-
181
)
4
Xu
Y.
Bhargava
G.
Wu
H.
Loeber
G.
Tong
L.
Crystal structure of human mitochondrial NAD(P)+-dependent malic enzyme: a new class of oxidative decarboxylases
Structure
1999
, vol. 
7
 (pg. 
877
-
889
)
5
Loeber
G.
Dworkin
M. B.
Infante
A.
Ahorn
H.
Characterization of cytosolic malic enzyme in human tumor cells
FEBS Lett.
1994
, vol. 
344
 (pg. 
181
-
186
)
6
Chang
G. G.
Wang
J. K.
Huang
T. M.
Lee
H. J.
Chou
W. Y.
Meng
C. L.
Purification and characterization of the cytosolic NADP+-dependent malic enzyme from human breast cancer cell line
Eur. J. Biochem.
1991
, vol. 
202
 (pg. 
681
-
688
)
7
Chang
G. G.
Huang
T. M.
Chang
T. C.
Reversible dissociation of the catalytically active subunits of pigeon liver malic enzyme
Biochem. J.
1988
, vol. 
254
 (pg. 
123
-
130
)
8
Loeber
G.
Maurer-Fogy
I.
Schwendenwein
R.
Purification, cDNA cloning, and heterologous expression of the human mitochondrial NADP+-dependent malic enzyme
Biochem. J.
1994
, vol. 
304
 (pg. 
687
-
692
)
9
Loeber
G.
Infante
A. A.
Maurer-Fogy
I.
Krystek
E.
Dworkin
M. B.
Human NAD+-dependent mitochondrial malic enzyme
J. Biol. Chem.
1991
, vol. 
266
 (pg. 
3016
-
3021
)
10
Mandella
R. D.
Sauer
L. A. T.
The mitochondrial malic enzymes I. Submitochondrial localization and purification and properties of the NAD(P)+-dependent enzyme from adrenal cortex
J. Biol. Chem.
1975
, vol. 
250
 (pg. 
5877
-
5884
)
11
Sauer
L. A.
An NAD+ and NAD(P)+-dependent malic enzyme with regulatory properties in rat liver and adrenal cortex mitochondrial fractions
Biochem. Biophys. Res. Commun.
1973
, vol. 
50
 (pg. 
524
-
531
)
12
Teller
J. K.
Fahien
L. A.
Davis
J. W.
Kinetic and regulation of hepatoma mitochondrial NAD(P) malic enzyme
J. Biol. Chem.
1992
, vol. 
267
 (pg. 
10423
-
10432
)
13
Zdnierowicz
S.
Swierczynski
J.
Selewski
L.
Purification and properties of the NAD(P)+-dependent malic enzyme from human placental mitochondria
Biochem. Med. Metab. Biol.
1988
, vol. 
39
 (pg. 
208
-
216
)
14
Hsieh
J. Y.
Liu
G. Y.
Chang
G. G.
Hung
H. C.
Determinants of the dual cofactor specificity and substrate cooperativity of the human mitochondrial NAD(P)+-dependent malic enzyme: functional roles of glutamine 362
J. Biol. Chem.
2006
, vol. 
281
 (pg. 
23237
-
23245
)
15
Sanz
N.
Diez-Fernandez
D.
Valverde
A. M.
Lorenzo
M.
Benito
M.
Cascales
M.
Malic enzyme and glucose 6-phosphate dehydrogenase gene expression increases in rat liver cirrhogenesis
Br. J. Cancer
1997
, vol. 
75
 (pg. 
487
-
492
)
16
Moreadith
R. W.
Lehninger
A. L.
Purification, kinetic behavior, and regulation of NAD(P)+ malic enzyme of tumor mitochondria
J. Biol. Chem.
1984
, vol. 
259
 (pg. 
6222
-
6227
)
17
Fahien
L. A.
Teller
J. K.
Glutamate-malate metabolism in liver mitochondria
J. Biol. Chem.
1992
, vol. 
267
 (pg. 
10411
-
10422
)
18
Baggetto
L. G.
Deviant energetic metabolism of glycolytic cancer cells
Biochimie
1992
, vol. 
74
 (pg. 
959
-
974
)
19
Mckeehan
W. L.
Glycolysis, glutaminolysis, and cell proliferation
Cell Biol. Int. Rep.
1982
, vol. 
6
 (pg. 
635
-
650
)
20
Landsperger
W. J.
Harris
B. G.
NAD+-malic enzyme. Regulatory properties of the enzyme from Ascaris suum
J. Biol. Chem.
1976
, vol. 
251
 (pg. 
3599
-
3602
)
21
Yang
Z
Lanks
C. W.
Tong
L.
Molecular mechanism for the regulation of human mitochondrial NAD(P)+-dependent malic enzyme by ATP and fumarate
Structure
2002
, vol. 
10
 (pg. 
951
-
960
)
22
Hung
H. C.
Kuo
M. W.
Chang
G. G.
Liu
G. Y.
Characterization of the functional role of allosteric site Asp102 in the regulatory mechanism of human mitochondrial NAD(P)+-dependent malic enzyme
Biochem. J.
2005
, vol. 
392
 (pg. 
39
-
45
)
23
Karsten
W. E.
Pais
J. E.
Rao
G. S.
Harris
B. G.
Cook
P. F.
Ascaris suum NAD-malic enzyme is activated by L-malate and fumarate binding to separate allosteric site
Biochemistry
2003
, vol. 
42
 (pg. 
9712
-
9721
)
24
Hsu
W. C.
Hung
H. C.
Tong
L.
Chang
G. G.
Dual functional roles of ATP in the human mitochondrial malic enzyme
Biochemistry
2004
, vol. 
43
 (pg. 
7382
-
7390
)
25
Hung
H. C.
Chien
Y. C.
Hsieh
J. Y.
Chang
G. G.
Liu
G. Y.
The functional roles of ATP binding residues in the catalytic site of human mitochondrial NAD(P)+-dependent malic enzyme
Biochemistry
2005
, vol. 
44
 (pg. 
12737
-
12745
)
26
Hsieh
J. Y.
Liu
G. Y.
Hung
H. C.
Influential factor contributing to the isoform-specific inhibition by ATP of human mitochondrial NAD(P)+-dependent malic enzyme: functional roles of the nucleotide binding site Lys346
FEBS J.
2008
, vol. 
275
 (pg. 
5383
-
5392
)
27
Moreadith
R. W.
Lehninger
A. L.
The pathways of glutamate and glutamine oxidation by tumor cell mitochondria. Role of mitochondrial NAD(P)+-dependent malic enzyme
J. Biol. Chem.
1984
, vol. 
259
 (pg. 
6215
-
6221
)
28
Sauer
L. A.
Dauchy
R. T.
Nagel
W. O.
Morris
H. P.
Mitochondrial malic enzymes. Mitochondrial NAD(P)+-dependent malic enzyme activity and malate-dependent pyruvate formation are progression-linked in Morris hepatomas
J. Biol. Chem.
1980
, vol. 
255
 (pg. 
3844
-
3848
)
29
Bhargava
G.
Mui
S.
Pav
S.
Wu
H.
Loeber
G.
Tong
L.
Preliminary crystallographic studies of human mitochondrial NAD(P)+-dependent malic enzyme
J. Struct. Biol.
1999
, vol. 
127
 (pg. 
72
-
75
)
30
Yang
Z.
Floyd
D. L.
Loeber
G.
Tong
L.
Structure of a closed form of human malic enzyme and implications for catalytic mechanism
Nat. Struct. Biol.
2000
, vol. 
7
 (pg. 
251
-
257
)
31
Yang
Z.
Zhang
H.
Hung
H. C.
Kuo
C. C.
Tsai
L. C.
Yuan
H. S.
Chou
W. Y.
Chang
G. G.
Tong
L.
Structural studies of the pigeon cytosolic NADP+-dependent malic enzyme
Protein Sci.
2002
, vol. 
11
 (pg. 
332
-
341
)
32
Coleman
D. E.
Rao
G. S.
Goldsmith
E. J.
Cook
P. F.
Harris
B. G.
Crystal structure of the malic enzyme from Ascaris suum complexed with nicotinamide adenine dinucleotide at 2.3 Å resolution
Biochemistry
2002
, vol. 
41
 (pg. 
6928
-
6938
)
33
Tao
X.
Yang
Z.
Tong
L.
Crystal structures of substrate complexes of malic enzyme and insights into the catalytic mechanism
Structure
2003
, vol. 
11
 (pg. 
1141
-
1150
)
34
Lai
C. J.
Harris
B. G.
Cook
P. F.
Mechanism of activation of the NAD-malic enzyme from Ascaris suum by fumarate
Arch. Biochem. Biophys.
1992
, vol. 
299
 (pg. 
214
-
219
)
35
Rao
G. S. J.
Coleman
D. E.
Karsten
W. E.
Cook
P. F.
Harris
B. G.
Crystallographic studies on Ascaris suum NAD-malic enzyme bound to reduced cofactor and identification of an effector site
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
38051
-
38058
)
36
Bradford
M. M.
A rapid and sensitive method for the quantitation of microgram quantities of protein, utilizing the principle of protein-dye binding
Anal. Biochem.
1976
, vol. 
72
 (pg. 
248
-
254
)
37
Schuck
P.
Size-distribution analysis of macromolecules by sedimentation velocity ultracentrifugation and Lamm equation modeling
Biophys. J.
2000
, vol. 
78
 (pg. 
1606
-
1619
)
38
Lebowitz
J.
Lewis
M. S.
Schuck
P.
Modern analytical ultracentrifugation in protein science: a tutorial review
Protein Sci.
2002
, vol. 
11
 (pg. 
2067
-
2079
)
39
Schuck
P.
On the analysis of protein self-association by sedimentation velocity analytical ultracentrifugation
Anal. Biochem.
2003
, vol. 
320
 (pg. 
104
-
124
)
40
Schuck
P.
Perugini
M. A.
Gonzales
N. R.
Howlett
G. J.
Schubert
D.
Size-distribution analysis of proteins by analytical ultracentrifugation: strategies and application to model systems
Biophys. J.
2002
, vol. 
82
 (pg. 
1096
-
1111
)
41
Brown
P. H.
Schuck
P.
Macromolecular size-and-shape distributions by sedimentation velocity analytical ultracentrifugation
Biophys. J.
2006
, vol. 
90
 (pg. 
4651
-
4661
)
42
Sali
A.
Blundell
T. L.
Comparative protein modelling by satisfaction of spatial restraints
J. Mol. Biol.
1993
, vol. 
234
 (pg. 
779
-
815
)
43
Krissinel
E.
Henrick
K.
Inference of macromolecular assemblies from crystalline state
J. Mol. Biol.
2007
, vol. 
372
 (pg. 
774
-
797
)
44
Hsieh
J. Y.
Chiang
Y. H.
Chang
K. Y.
Hung
H. C.
Functional role of fumarate site Glu59 involved in allosteric regulation and subunit-subunit interaction of human mitochondrial NAD(P)+-dependent malic enzyme
FEBS J.
2008
, vol. 
276
 (pg. 
983
-
994
)
45
Altschul
S. F.
Boguski
M. S.
Gish
W.
Wootton
J. C.
Issues in searching molecular sequence databases
Nat. Genet.
1994
, vol. 
6
 (pg. 
119
-
129
)
46
Higgins
D.
Thompson
J.
Gibson
T.
Thompson
J. D.
Higgins
D. G.
Gibson
T. J.
CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties, and weight matrix choice
Nucleic Acids Res.
1994
, vol. 
22
 (pg. 
4673
-
4680
)
47
Hall
T. A.
BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT
Nucleic Acids Symp. Ser.
1999
, vol. 
41
 (pg. 
95
-
98
)