Catalytic mechanism and kinetics of malate dehydrogenase Open Access

Malate dehydrogenase (MDH, EC 1.1.1.37) catalyzes the reversible conversion of L-malate ((S)-malate) to oxaloacetate, accompanied by the conversion of NAD⁺ to NADH + H⁺, with the addition of the pro-R hydrogen to NAD⁺ (Figure 1) [1,2]. This enzyme is highly conserved across all kingdoms of life. Eukaryotes contain cytosolic MDH as well as isozymes that are specifically transported to organelles, including mitochondria, chloroplasts, peroxisomes, and glyoxysomes. Although the size of MDH varies among organisms and organelles, MDH is approximately 36 kDa and most commonly functions as a dimer. For NADH produced by cytosolic MDH, reducing equivalents in the form of L-malate allows for the net import of NADH into the mitochondria via the L-malate-aspartate shuttle to drive the production of ATP. Mitochondrial MDH catalyzes an essential step in producing oxaloacetate in the citric acid cycle, in addition to NADH for making ATP.

The active site of MDH features a crucial His 177-Asp 150 pair [3] (this review uses the amino acid numbering for Escherichia coli MDH, Uniprot P61889). His 177 forms a hydrogen bond to the hydroxyl group of L-malate, facilitating its conversion to oxaloacetate by removing the hydroxyl proton (Figure 2). L-Malate is oriented in the active site by forming salt bridges between its C1 and C4 carboxylate and Arg 153 and Arg 81, respectively. As described below in more detail, the two substrates bind, and then a flexible loop bearing Arg 81 and a third essential Arg residue, Arg 87, encloses the active site before the reaction proceeds.

The observed equilibrium constant for MDH greatly favors the formation of L-malate and NAD⁺. Under near-physiological conditions (38°C, pH 7.0, ionic strength I = 0.25, 1 mM Mg²⁺), $Keq'=2.9×10-5$⁠, corresponding to a ΔG = +27 kJ/mol [4]. The endergonic reaction catalyzed by MDH is facilitated within the mitochondrial cellular environment by coupling with downstream metabolic pathways. NADH is utilized in the electron transport chain to produce ATP, and oxaloacetate is a critical intermediate in the citric acid cycle. The enzyme citrate synthase is energetically favored and consumes the produced oxaloacetate, driving the MDH-catalyzed reaction forward. In laboratory settings, the exergonic reverse reaction (conversion of oxaloacetate to L-malate) is often studied.

As described below, the L-malate oxidation reaction is more favorable at high pH than the oxaloacetate reduction reaction. K_eq increases with rising pH levels, reaching $Keq'=9.1×10-4$ at pH 9 [5]. Despite this increase, the equilibrium still greatly favors oxaloacetate reduction, with the equilibrium percentage of oxaloacetate decreasing slightly from 99.8% at pH 7 to 97.1% at pH 9. Computational modeling of the MDH-catalyzed oxidation of L-malate suggests two distinct steps, each characterized by its own transition states. Initially, His 177 of the E. coli MDH deprotonates the hydroxyl group of L-malate with a calculated activation energy ΔG^‡ of +29 kJ/mol (Figure 3). Subsequently, the hydride anion from the C2 of L-malate can transfer to NAD⁺, representing the rate-limiting catalytic step with a calculated ΔG^‡ of +63 kJ/mol [6].

The catalytic dyad and one substrate binding residue, Arg 153, provide initial contacts with L-malate/oxaloacetate, followed by further interactions with a mobile loop between β4 and α4 (residues 76-88), located near the L-malate/oxaloacetate binding site. This loop includes Arg 81 and Arg 87, which interact with the C4 carboxylate and stabilize the anionic transition state, respectively [7]. The structure of MDH from the archaeon Metallosphaera sedula was determined in the presence and absence of NAD⁺/L-malate [8]. The active site loop is disordered with high crystallographic temperature factors when no substrate is present and becomes more ordered when the substrate binds (Figure 4). In the structure of the enzyme from M. sedula with and without NAD⁺, the loop is most open in the apoenzyme. When NAD⁺ binds, the loop partially closes, and then completely closes when L-malate binds. On average, the loop moves 6.8 Å between the most open and most closed forms. Molecular dynamics simulations of Thermus thermophilus MDH revealed increased mobility of the active site loop at higher simulated temperatures, while the mobility of other loops in the enzyme did not change with temperature [9].

The loop conformations of the MDH enzymes from Chlorobium vibrioforme and Chloroflexus aurantiacus, adapted to mesophilic and thermophilic environments, respectively, were studied by molecular dynamics simulations [10]. The MDH from the thermophile exhibited more salt bridges, leading to the stabilization of the active site loop by residues even in the absence of substrate. This suggests that the loop is less mobile in the thermophile.

MDH displays a high level of substrate discrimination, showing a six order of magnitude difference in the reduction rate of oxaloacetate compared to pyruvate. This specificity arises from charge balancing between substrates and the active site of MDH [11]. The activated complex formed upon oxaloacetate and NADH binding includes a positively charged imidazolium ion acting as a general acid, along with two negative charges contributed by each carboxylate of oxaloacetate, resulting in a net negative charge of -1. When oxaloacetate binds, the two loop Arg residues (81 and 87) are driven into close proximity to each other and the C4 carboxylate of oxaloacetate. As pyruvate lacks this complementary negative charge of oxaloacetate, charge repulsion prevents a catalytically productive conformation, leading to the 10⁶-fold discrimination [11].

Given the critical role of the active site loop in substrate binding, it is noteworthy that Arg 81 emerges as a key determinant for substrate specificity. In contrast, the related lactate dehydrogenase (LDH) has a Gln residue at the corresponding position. LDH has detectable activity with oxaloacetate, while MDH has comparatively little detectable activity with pyruvate. Mutating the Gln of LDH to an Arg at the corresponding position is critical in creating an LDH that is more specific for oxaloacetate than pyruvate [12]. Conversely, mutating this arginine to glutamine (R81Q) in MDH results in higher specificity for pyruvate, but significantly lower activity than wild-type MDH with oxaloacetate. However, this activity can be improved with other loop mutations; notably, the mutation of the nearby Ala 80 to proline results in a four-fold increase in k_cat/K_M [13]. Due to the critical role of loop residues in conferring specificity and the tendency for enzymes with loop mutations to retain some activity, the loop has become a subject of study in Course-based Undergraduate Research Experiences [14].

Furthermore, other mutations in MDH have been found to impact its activity. For instance, the R102Q mutation induces loop closure in E. coli MDH, enhancing thermal stability [15]. Mutation of the Arg 153 residue that binds to the C1 carboxylate of the substrate to Cys (R153C) results in a mutant with diminished catalytic activity, although K_M values for L-malate and oxaloacetate remain unaffected [16]. Some mutations, such as Q11G and N119G in Plasmodium falciparum MDH, lead to the loss of critical hydrogen bonding interactions with NADH and an increased K_M for oxaloacetate compared with the wild-type [17].

Human MDH2 shows at least three orders of magnitude greater activity with oxaloacetate compared with the longer diacid substrate, α-ketoglutarate. Despite this reduced activity, MDH2 may contribute to the production of 2-hydroxyglutarate in cancer cells. Notably, as the pH is lowered from 7.4 to 6.8, commonly associated with the hypoxic conditions of cells, the generation of 2-hydroxyglutarate by MDH increases while oxaloacetate reduction decreases over the same pH change [18].

MDH generally prefers NAD⁺ over NADP⁺, although NADP⁺-dependent MDH enzymes do exist in chloroplasts [19]. Cosubstrate specificity is determined by interactions in the β2-loop-α3 region (residues 34-38 in E. coli MDH). The cytoplasmic Thermus thermophilus (UniProt ID P10584, formerly Thermus flavus) MDH efficiently reduces oxaloacetate using both NADH and NADPH, with a 22-fold higher value of k_cat/K_M with NADH (8.5 × 10⁷ M⁻¹ s⁻¹ and 3.9 × 10⁶ M⁻¹ s⁻¹, respectively), along with a 14-fold lower K_M value for NADH [20].

Mutation of the β2-loop-α3 region in T. thermophilus (amino acids 42-48) to the corresponding residues in chloroplastic Zea mays MDH (UniProt ID P15719, amino acids 127-133) leads to a 7-fold increase in efficiency using NADPH as a cosubstrate and a 71-fold decrease when using NADH. Part of the decrease in the K_M for NADPH utilization may be due to the mutation of a Gln to Arg causing ionic attraction with the phosphate of NADPH. However, mutation of other residues in the loop improves the relative efficiency of NADPH-dependent reduction of oxaloacetate [20]. In Bacillus subtilis, substituting these same residues converts the wild-type MDH, with a 200-fold preference for NADH, into a mutant enzyme with a 9-fold preference for NADPH. Additionally, this mutant displayed a lower K_M value, higher k_cat, and higher k_cat/K_M values for NADPH than wild-type NADPH-dependent MDHs from T. flavus and Corynebacterium glutamicum [21]. The ability to alter the coenzyme specificity of B. subtilis MDH from NADH to NADPH opens up new possibilities for utilizing this enzyme in industrial processes that require NADPH-dependent reactions. Understanding the impact of mutations on protein-ligand binding affinity and kinetics is crucial, and computational methods offer valuable insights into MDH mutational dynamics to predict these effects [22].

The Rossman fold found in MDH is common to many nucleotide-binding enzymes, including S-adenosyl methionine (SAM)-dependent enzymes. Although E. coli MDH has no detectable affinity for SAM, removing three amino acids (ΔAla9, ΔGly10, ΔGln14) from the V/IxGxxGxxG motif found in the β1-loop-α1 region of E. coli MDH to create instead a V/IGxGxG motif, typically found in SAM binding Rossman folds, leads to an MDH that binds SAM (K_d = 13 µM) and has no observable binding to NAD⁺ [23]. These deletions shorten the α1 helix and flatten the loop, expanding the active site to bind SAM.

Steady-state enzyme kinetics data are available for both the forward and reverse reactions of many MDH isozymes from various organisms (Table 1). Older experiments often used racemic mixtures of L- and D-malate, and the table indicates when the L enantiomer was utilized. The typical reaction conditions for assays with oxaloacetate included 30–100 mM Tris or phosphate buffer, pH 7–8, with 0–100 mM KCl or NaCl at 30–37°C. As described in this review, enzyme assays using L-malate as a substrate were frequently conducted at a higher pH. For most enzymes, the K_M for oxaloacetate is lower than for L-malate, and the K_M for NADH is lower than for NAD⁺, which is reasonable given the lower cellular concentrations of oxaloacetate and NADH compared to L-malate and NAD⁺, respectively [24]. The k_cat values for L-malate oxidation were much smaller than for oxaloacetate reduction, with MDH from P. falciparum exhibiting the highest k_cat for L-malate oxidation [25].

When kinetic constants were available for both the cytoplasmic and mitochondrial isozymes in the same organism, the K_M values for oxaloacetate were often similar. However, the K_M for NADH varied, with higher values observed for the mitochondrial isozyme in most cases, as seen in Caenorhabditis elegans [36]. This observation is consistent with the higher NADH concentration found in the mitochondria (reviewed in [42]). The chloroplastic MDH from Arabidopsis thaliana also had a higher K_Mfor NADH compared to the cytosolic enzyme [34,35]. The K_M values for NADH from different steady-state kinetics experiments with Sus scrofa isoforms were more varied [30,32]. Detailed time-course measurements and kinetic models for both cytosolic and mitochondrial S. scrofa MDH isoforms predicted higher activity for cytosolic MDH at lower NADH concentrations [43], suggesting that the cytosolic MDH also has a lower K_M for NADH. This comprehensive understanding of MDH kinetics across diverse cellular contexts underscores the adaptability and crucial role of MDH in cellular metabolism.

MDH follows an ordered Bi-Bi kinetic mechanism characterized by the sequential binding of NAD⁺ and then L-malate, followed by the release of oxaloacetate and then NADH [5,43–46]. The kinetic evidence supports a compulsory order reaction mechanism, and some studies suggest that an enzyme isomerization is involved [47–49]. Lodola et al. found that dimeric MDH has two independent active sites that can function simultaneously [50].

Generally, MDH activity has an optimal temperature range between 30 and 40°C. Deviations from this range are somewhat related to the organism’s physiological temperature range. Thermostability also varies among MDH isoforms, with cytosolic isoforms being more stable than mitochondrial isoforms, such as in C. elegans [36], S. scrofa [28,51,52], and Sphyraena idiastes [30]. However, these enzymes from eukaryotic species are notably less stable than MDH from the mesophile Streptomyces coelicolor and the thermophilic bacterium T. thermophilus. S. coelicolor MDH remains active at 50°C but rapidly inactivates above this temperature, while T. thermophilus MDH retains activity at 90°C for 60 min [20,39,40,51]. A larger number of salt bridges between subunits contributes to thermal stability [28,53]. However, thermal stability is a complex trait influenced by multiple factors, including electrostatic interactions, hydrophobic interactions, hydrogen bonding, and overall protein folding [28,53–55].

Optimal enzyme activity is achieved when MDH reduces oxaloacetate near pH 8, and activity is generally high from pH ∼7–8.5 [5,21,36,56]. Interestingly, the optimal pH for some thermostable MDHs varies with temperature [39,40]. The L-malate oxidation rate is higher at pH > 8 [25,39,43,56]. The pH-dependency of MDH kinetic behavior involves additional ionic states and still follows the ordered Bi-Bi mechanism but is modified to account for enzyme protonation states [5,43,57]. The mitochondrial porcine MDH binds a proton on a His residue upon binding to NADH [58], and a His is deprotonated before L-malate binds [1]. This is consistent with the active site base, His 177, having the appropriate protonation state in each catalytic direction. Also, the interaction of NAD⁺ with MDH was dependent on the protonation of two unidentified residues, with pK_a values of ∼6 and 7–8 [5,43].

Excessive oxaloacetate inhibits cytoplasmic and mitochondrial MDH, but at concentrations too high to influence enzyme activity in the cell [29,59]. Interestingly, oxaloacetate may also inhibit NAD⁺ binding, potentially with greater potency than its impact on NADH [47]. L-malate activates MDH at a site that is different from the active site [31,46]. Additionally, citrate is an allosteric regulator of MDH, and the influence of citrate on MDH activity is intricate and multifaceted. Citrate-activated mitochondrial MDH (mMDH) promotes the oxidation of L-malate [48], and this may be another mechanism to promote this energetically unfavorable reaction. However, citrate inhibited oxaloacetate reduction by mMDH and in both directions of the reaction by cytoplasmic MDH (cMDH) [48]. Interestingly, it was later found that citrate only activated mMDH when L-malate and NAD⁺ concentrations were elevated (2.5–10 mM and 1–5 mM, respectively), while citrate inhibited L-malate oxidation at low concentrations of L-malate or NAD⁺ [49]. All three effectors (oxaloacetate, L-malate, and citrate) bind to the same putative allosteric site [48].

Human mMDH is an allosteric enzyme that interacts cooperatively with L-malate, with tetramers showing higher activity than dimers [60]. When using NADH as a cofactor, human mMDH is activated by fumarate that binds at the dimer interface at a site ∼30 Å from the active site [60]. The enzyme is also inhibited by ATP, which binds to a site at the tetramer interface (also called the exosite) where ATP/ADP and NAD⁺/NADH can bind. The binding of ATP leads to the dimer being more stable than the tetramer, causing inhibition [61]. ADP has a similar but much weaker effect. The binding of NAD⁺ to the tetramer interface leads to the tetramer being more stable and increased activity. Other mammalian MDH enzymes (cMDH and mMDH-NADP⁺) have not been shown to be similarly regulated.

• In vivo, the reaction catalyzed by malate dehydrogenase mainly proceeds in the direction of L-malate oxidation, but in the lab, it is easier to study the reaction in the spontaneous reverse direction.

• MDH is present in all known organisms, and there are different isozymes in organelles.

• Multiple investigations of the flexible active site loop and other residues have shown that MDH is a good model system for investigating the basis of substrate specificity.

• Regulation of MDH is complex, and the enzyme is affected allosterically by several molecules.

• There is little published work on human MDH, which may be a fruitful area of future study.

The authors declare that there are no competing interests associated with the manuscript.

ΔG^‡

activation energy

A. thaliana

Arabidopsis thaliana

ATP

adenosine triphosphate

C. elegans

Caenorhabditis elegans

C. vulgaris

Citrullus vulgaris

chloro

chloroplastic

cMDH

cytoplasmic malate dehydrogenase

cyto

cytoplasmic

E. coli

Escherichia coli

glyox

glyoxosomal

H. marismortui

Haloarcula marismortui

H. sapiens

Homo sapiens

kDa

kilodalton

M. sedula

Metallosphaera sedula

MDH

malate dehydrogenase

mito

mitochondrial

mMDH

mitochondrial malate dehydrogenase

NAD⁺

nicotinamide adenine dinucleotide

NADH

dihydronicotinamide adenine dinucleotide

NADP⁺

nicotinamide adenine dinucleotide phosphate

NADPH

dihydronicotinamide adenine dinucleotide phosphate

OAA

oxaloacetate

P. falciparum

Plasmodium falciparum

S. coelicolor

Streptomyces coelicolor

S. scrofa

Sus scrofa

SAM

S-adenosyl methionine

T. thermophilus

Thermus thermophilus (formerly Thermus flavus)