Sandbox Wabash 02 Fumarase

Determination of the True Active Site of Fumarase C from E. Coli.

Fumarase C from E. coli is an enzyme homologous to the cytosolic and mitochondrial enzymes found in eukaryotic cells. It catalyzes the hydration/dehydration reaction between the metabolites L-malate and fumarate. Eukaryotic fumarases have been studied extensively by classical kinetic and isotope labeling methods (Hill & Teipel, 1971; Cleveland, 1977). Fumarase from E. coli is less well characterized. Two basic groups are thought to be involved in the overall catalytic process. The first (B1) is responsible for the removal of a proton from the C3 position of L-malate, resulting in a carbanion stabilized by an acid-carboxylate intermediate formed at C4. The last stage of the catalytic process is thought to involve a second basic group on the protein labeled B2. In the direction of fumarate production, this basic group would be protonated and the removal of the -OH from C2 results in the formation of a water molecule. The proton at B1 in Figure 1 has unusual properties and is believed to be removed as the next substrate molecule binds (Rose et al., 1992). The crystallographic studies of wild-type fumarase indicated that the enzyme had an unusual subunit arrangement with a core of 20 alpha-helices, 5 from each of the subunits (Weaver et al., 1995). Subsequent crystallographic studies with several inhibitors including pyromellitic acid and beta-trimethylsilyl maleate produced some unexpected results. Although both are related to the normal substrate, they are bound to different sites. The original tungstate site, a heavy atom derivative, was also the binding site of the inhibitors citrate and pyromellitic acid. This was labeled the A-site and is comprised of atoms from three of the four subunits (Weaver & Banaszak, 1996). The second site contained L-malate in the native crystals and beta-trimethylsilyl maleate in another crystallographic study (Weaver & Banaszak, 1996), and was labeled the B-site and was formed by atoms from a single subunit. The results led to the dilemma as to which of the two sites was the active site. ^[1]

The discovery that fumarase harbored two adjacent anion binding sites led to the dilemma as to which site was the active site (Weaver & Banaszak, 1996). The H129N and H188N mutants were generated to resolve the two site problem. The fact that fumarase is only active as a tetramer provided strong initial support for the A-site being the active site because it has components from three subunits (Weaver et al., 1995). It was suspected that the A-site was the active site because of the observation that no active monomeric form of fumarase has ever been described, and the A-site was formed by residues from three of the four subunits. Because the biochemical data suggested that a histidine side chain was one of the bases participating in the catalytic reaction (Brant et al.. 1963), testing whether H129 or H188 affected catalytic activity appeared to offer a way of resolving the two site dilemma. Evidence as to which of the two ligand binding sites was indeed the catalytic site should be obtainable by mutating the histidines at the two different sites. If the A-site was the active site, changing H188 should dramatically affect the catalytic activity. Conversely, if the B-site was the active site then a mutation at H129 should affect catalysis. ^[2]

As evident from the calculated specific activities for wild-type fumarase and the H129N and the H188N mutants, little effect is observed in changing H129 into an asparagine residue as both wild type and H129N have about the same specific activity. However, the H188N mutation dramatically affects the catalytic reaction. The specific activity of the H188N is approximately 200 times less than that of wild-type enzyme or H129N. These results prove that the active site of fumarase is near H188 or the A-site. The significance of the B-site to the catalytic reaction is unknown. X-ray crystal structures were determined to see what effects the mutations had on local conformation at both the A- and B-sites. ^[3]

Structure of the Active Site of Fumarase.

A number of stereochemical factors describing the two sites were examined in the wild-type crystal structures (Weaver & Banaszak, 1996). Site A was in a relatively deep pit removed from bulk solvent. It also contained an unusually bound water, although there was no obvious way of linking this directly to the catalytic process. One of the side chains interacting with this water molecule is H188. In the crystallographic coordinates of the wild-type enzyme, the water molecule forms a short hydrogen bond, 2.5 A, with the imidazole ring of H188. The side chain of H188 is also within hydrogen bonding distance to an oxygen atom of bound citrate or pyromellitic acid. The B-site is closer to the surface of the enzyme (Weaver & Banaszak, 1996). There are three principal interactions between the ligand and wild-type fumarase at the B-site, and the A- and B-sites are linked by residues 131 to 140 in a single subunit. Main chain hydrogen bonds between the oxygen atoms of the bound ligand and main chain -NHs of D132 and N131 on the N-terminal end of the pi-helix are important to stabilization at the B-site. Oxygen atoms of the other carboxylate of the ligand at the B-site are hydrogen bonded to R126-NE and H129-NDI. The hydrogen bonds between side chain atoms of N135 and N103, and between N103 and S140 form an indirect connection between the B- and the A-site. H129 is the only basic group close to a ligand bound at the B-site. ^[4]

Values for phi and psi angles were plotted and compared with normally allowed torsional regions (Laskowski et al., 1993). Out of a total of 911 residues, only F356 fell outside of the allowed region. It is surrounded by a number of additional hydrophobic residues from three of the four subunits within the tetramer. F356b belongs to a sharp turn with the side chain pointing into a hydrophobic pocket that lies behind the active site. It is near to one of the molecular dyads and therefore close to both W297c and W297d. It is also positioned near L358b, which is close to the active site water molecule and it has van der Waal contacts with R186c and H 188c, both of which are considered part of the active site. Other residues within van der Waal contact of F356b are from the c-subunit including L298c, I306c, and L189c. Although attention was drawn to this phenylalanine by its unusual phi and psi angles, it is clearly a pivotal residue at the juncture of subunits near the active site. ^[5]

The H129N mutation was made to characterize the functional significance of a dicarboxylic acid binding site we have labeled the B-site. The B-site is formed from a single subunit of the tetramer and includes atoms from residues R126, H129, N131, and D132 (Weaver & Banaszak, 1996). H129 is the only potential side chain that could serve as one of the catalytic bases in the B-site (Brant et al., 1963). The H129N mutation had little effect on catalytic activity, confirming the active site to be site A. The crystal structure of H129N showed that the mutated protein had essentially the same conformation as the wild type but appeared to dramatically reduce binding of ligands at the B-site. ^[6]

Removal of the carboxylate binding site is due to a rotation of N129 about x, compared to the imidazole side chain position in the native structure. N129 is stabilized in this new position by forming bifurcated hydrogen bonds between its OD1 atom and the backbone nitrogen atoms of N131 and D132. These two backbone atoms form important hydrogen bonds to the O1 and O2 atoms of the dicarboxylic acid at the B-site in the native enzyme but are blocked by internal H-bonds in the H 129N mutant. The carboxyamide side chain of N129 in the mutant behaves like the carboxylate of ligand bound at site B. ^[7]

To further confirm that dicarboxylic acids will no longer bind to the B-site due to the steric hindrance imposed by N129, an H129N crystal was soaked in a solution containing P-trimethylsilyl maleate (TMSM) under the exact conditions used to generate the complex reported by Weaver and Banaszak (1996). No binding of TMSM was visible in the resulting maps. This result not only proved unambiguously that the A-site was the active site, but also that the H129N mutation dramatically reduced binding at the B-site. The A-site or active site in the crystal structure of H129N was unchanged by the mutation at H129. ^[8]

The mutation at the A-site, H188N, had a dramatic effect on the fumarase activity and suggested that this histidine side chain may have a direct role in the catalytic mechanism in H188N. The absence of the histidine side chain effectively reduces binding of citrate so that it is missing in the electron density maps. Another observation with respect to the H188N mutation is that the active site water molecule (W-26) is still present although shifted by approximately 0.70 A compared to the crystal structure of the wild-type enzyme. In the crystal structure of H188N, W-26 makes four hydrogen bonds to protein atoms. The replacement asparagine side chain, N188, forms two hydrogen bonds to W-26, one of which is rather short- 2.59 A. The N188 side chain seems able to mimic that of H188, leading to the previously observed short hydrogen bond between the side chain at 188 and W-26. S98b and N141b form the other two hydrogen bonds to the active site water. T100-OGI, which is one of the reported atoms interacting with the water in the crystal structure of native fumarase is no longer within hydrogen bonding distance. In the wild type, the active site water W-26 is hydrogen bonded to five different atoms. There are four protein ligands from two subunits including residues T1OOb, S98b, N141b. and H188c. The other interaction is only formed when a competitive inhibitor is present. With bound citrate ion, the water interaction involves the carboxylate at the C1 position (Weaver & Banaszak, 1996). In the case of the H129N structure, W-26 maintains all five of the interactions. ^[9]

^[10]