Presentation of Mandelate Racemase enzyme

Biological context:
Mandelate Racemase (E.C. 5.1.2.2.; MR)^3.3 from the soil bacteria Pseudomonas putida catalyzes the interconversion of the enantiomers of mandelic acid[217] (see figure 2.1).

3cm

Experimentally it has been studied as a paradigm for enzymes which catalyze rapid carbon-hydrogen bond cleavage. Early studies suggested that the mandelate pathway of P. putida consists of five enzymes that facilitate the conversion of both (R) and (S)-mandelate to benzoate, which is subsequently converted to acetyl coenzyme A (CoA) and succinyl-CoA.

Why Mandelate Racemase?:
Among all the racemases family MR is the most studied enzyme. The following reasons justify the choice:
1. It is a cofactor independent inducible enzyme that can be obtained in large amounts by fermentation from Pseudomonas Putida
2. Immobilization leads to an enhanced activity and facilitates its recovery
3. The catalytic efficiency of this enzyme is exceptional (turnover frequency $\sim$1000 s$^{-1}$)
4. It promotes a reaction that is almost impossible by chemical means
These are probably the reasons why MR was the first of the racemases to have an elucidated structure by X-ray spectroscopy.

X-ray:
The Mandelate Racemase X-ray structure and several of its mutational variants are available at reasonable high resolution. This is an essential experimental data that gives us the opportunity to start a theoretical study.

Figure 2.1: Scheme of Mandelate Racemase natural reaction

In its crystal structure MR is an octamer of 422 symmetry. Every subunit is composed of two major structural domains. An N-terminal $\alpha + \beta$ domain and a central parallel $\alpha/\beta$ barrel, there is also a third, smaller, irregular C-terminal domain (see figure 2.2).

Figure 2.2: Stereoview of a ribbon diagram showing the three dimensional structure of MR backbone. In ball & stick the mandelate substrate is bound in the active site. This figure was prepared by VMD [218]

.

Common evolution ancestors:
The molecular weight of enzyme's subunit is 38 570 and its secondary, tertiary and quaternary structures are strikingly similar to those of muconate lactonizing enzyme and galactonate dehydratase. This similarity indicates a common evolution ancestor to perform the necessary chemical task of abstracting protons $\alpha$ to a carboxylate that have relatively high $pK_a$ values.

Ion dependent:
Wild Mandelate Racemase is Mg$^{2+}$ dependent. Some studies suggest other less effective divalent metal ions such as Co$^{2+}$, Ni$^{2+}$, Mn$^{2+}$ and Fe$^{+2}$[217]. This required divalent metal ion was found to be tightly bound to MR and close to the bound mandelate, suggesting that the metal ion helps to the deprotonation reaction withdrawing the excess of electron density. In figure 2.3 we can see the ligand sphere of the divalent cation.

Many substrates, inhibitors and inactivators:
Many possible substrates can bind the active site, or even racemize through Mandelate Racemase. Besides its natural substrate, the enzymatic racemization of other $\alpha$-hydroxy carbonyl compounds can take place. In table 2.1 a brief description of the most studied substrates is given.

Table: Different enzymatic activity of some substrates that bind to Mandelate Racemase active site

Substrate		Activity
2*Mandelic acid	2*	natural substrate
		$k_{cat}=350s^{-1}$[214]
2*Vinyl-glycolate	2*	active substrate
		$k_{cat}=250s^{-1}$[219]
2*Propargyl-glycolate	2*	active and partially
		inactivator $k_{cat}=79$[220]
2*Mandelic acid amide	2*	active substrate
		(15% of mandelate)[221]
2*(R)-$\alpha$-PhenylGlycidate	2*	inhibitor(used in Xray) [222]
2*Benzylphosphonate	2*	inhibitor[223]
2*(Hetero)-aryl-substituted	2*	active substrates[224]

The design of substrates gives mechanistic information:
The design of alternative substrates has helped to find out experimentally the important residues in the active site for the binding process. For example, the $\alpha$-OH group seems to be crucial for the binding and orto-substituted phenyl ring provokes a remarkable steric hindrance[224]. Since the amide derivative of mandelate, although with a significant lower rate, can racemize [221], the presence of a carboxyl or negative charged oxygen on the substrate does not seem to be essential for the binding^3.4.

This substrate spectrum also gives information about the possible mechanism, for example, electron-donating phenyl substituents position enhance the enzyme activity, which means that a negative charged stabilization is needed for the racemization[226]. While phenyl substituents on para and meta positions bind the active site, orto-analogues do not bind due to steric limitations. An aromatic system must be present in $\beta$ position being the vinyl-glycolate the minimal conjugated system. In the absence of $\pi$-electrons in this position, such as for lactate, no racemization occurs.

Vinylglycolate has also been found to be an excellent substrate of Mandelate Racemase, with a value of $k_{cat}$ somewhat lower, but comparable to that of mandelate. In addition, propargylglycolate has been determined to be a moderately good substrate for racemization, with a $k_{cat}$ value of about 10% relative to mandelate. The case of propargylglycolate has been found to be specially interesting because it is also an irreversible inhibitor, with a partition ratio of racemization/inactivation of about 17 000[220]. These two alternative substrates along with mandelate have been used in this thesis (section 2.3, page ) to study its racemization process. This study will provide a mechanistic insight that could explain the tendency of the reaction kinetics.

The active site:
In figure 2.3 a schematic picture of the active site deduced from x-ray spectroscopy it is shown.

Figure 2.3: Schematic representation of the active site of Mandelate Racemase. When X=phenyl we have the natural substrate in S configuration

Other residues are not displayed in figure 2.3 for clarity reasons. Aspartic270 has been proposed to form a dyad with Histidine297 to increase the proton-donor capacity. Asparagine197 seems to interact with the $\alpha$-hydroxyl group to stabilize the transition state racemization. Both mutagenesis substitution experiments (D270N and N197A) provoke a reduction in catalytic efficiency of the enzyme.

Proposed mechanism and structure-function relationships:
Many experimental studies have been published to elucidate the Mandelate Racemase mechanism.

• Kinetic measurements (reference [214])
• Isotopic effects (references [214,222])
• Site-directed mutagenesis experiments: K166R [227], H297N [228], E317Q [214], D270N [229] and N197A [223]

The main conclusions of these experiments are:
Lys166 and His297 role:
Mandelate Racemase proceeds by a two-base mechanism. The $\epsilon$-amino group of Lys166 is the general base catalyst that abstracts the $\alpha$-proton from (S)-mandelate. Whereas the doubly protonated imidazole group of His297 acts as the general base catalyst that removes the corresponding $\alpha$-proton from (R)-mandelate. The conjugate acids of Lys166 and His297 serve as the proton donors in the formation of (S) and (R)-mandelate, respectively.

Stabilization by Lys164, Glu317 and Mg: Lysine164, Glutamic317 and magnesium cation help to the ligand binding at the first stage of the reaction and to the electron density withdrawing to stabilize the enolic intermediate when proton abstraction takes place.

Intermediate:
The experimental results suggest that the mechanism takes place through an enolic intermediate. The term enolic intermediate is preferred rather than carbanion, enolate or enol to avoid specifying the extent to which the proton is transferred from the general acid catalyst (Glu317 or Lys164) to the oxygen atom of the intermediate. In any case, any isolation of this intermediate has not been reported. At the end of the next chapter we will propose that such intermediate is a transition state rather than a stable species.

Moreover, Glutamic317 has been suggested to form a so called Low Barrier Hydrogen Bond (LBHB)[230] to explain the overstabilization of the intermediate by the enolate anion formation.[215,225] However Guthrie and Kluger[231] suggest that the electrostatic stabilization in combination with a reduction in medium polarity may be sufficient to stabilize the unstable species formed during the catalysis. Actually, if the mandelic acid amide, as we said above, was racemized at an acceptable rate[221] we can conclude that the formation of a LBHB, more difficult in the amide derivative, is not essential for the chemical step.

The possible explanation of pseudosymmetry: The $\epsilon-$ammonium group of Lys166 has a $pK_a$ that is evidently lowered by about 4 $pK$[217] units by electrostatic effects of the active site, while the $pK_a$ of the His297 is a more nearly normal value for a histidine imidazole group. This lowering of Lys166 makes both $pK_a$ closer to each other and this can account for the "pseudosymmetry" in the racemases reaction pointed out at the beginning of this section.