Friday, March 24, 2006

Crystallization and Preliminary Crystallographic Studies of Saccharomyces Cerevisiae Alcohol Dehydrogenase I

Yeast alcohol dehydrogenase was one of the first enzymes to be purified and crystallized. Many papers describe the kinetic, chemical and physical properties of the enzyme. The amino acid and gene sequences of the constitutive isoenzyme I have been determined. In yeast, this isoenzyme functions in glycolysis to reduce acetaldehyde. The enzyme is a tetramer of molecular mass 150,000 Da and binds one NAD co-enzyme per subunit. Dissociation into subunits inactivates the enzyme. it is most active on ethanol and less active on longer, branched or cyclic alcohols. Structure/function relationships have been probed by site-directed mutagenesis. More complete understanding of the substrate specificity, mechanism and quaternary structure requires knowledge of three-dimensional structure.

The structures of the horse liver alcohol dehydrogenase in various complexes and crystal forms have been determined. The liver enzyme is a dimer, and probably functions to oxidize ethanol and other alcohols encountered in the environment, as it has a broad specificity. Since
the liver and yeast enzymes are homologous, with 25% sequence identity, molecular modeling of the yeast enzyme can approximate the structure of one subunit, but not yet the quaternary arrangement.

Although the yeast enzyme has been crystallized from ammonium sulfate, the crystals are thin plates or needles that are not suited for X-ray crystallography. Furthermore, the enzyme purified from commercial yeast may be heterogeneous. We have purified a single form of enzymefrom yeast expressing the gene on a multicopy plasmid.

The enzyme was crystallized by the hanging drop method by vapor diffusion. A 10 to 20 ul drop containing about 12 mg/ml of enzyme in 100 mM sodium N-tris(hydroxymethyl)methyl-3-aminopropanesulfonate buffer (pH 8.2 to 8.6), 0.25 mM EDTA, 2mM NAD+, 0.1 M 2,2,2-trifluroethanol and 6% polyethylene glycol 4000, was placed over a 0.5ml reservoir with 12 to 16% polyethylene glycol 4000 in 0.1 M trifluoroethanol. trifluroethanol binds tightly (Kd = 2.8 mM) to the enzyme-NAD+ complex and should mimic the complex with ethanol. Hexagonal crystals about 0.5mm wide and 0.2 mm thick formed after a few weeks or months at 5 degree Celsius.

Numerous attempts were made to find other suitable conditions. Co-crystallization with NADH produced micro crystals. Crystals also formed at pH near 7, and with sodium N-tris(hydroxymethyl)methyl-2-aminoethanesulfonate or sodium phosphate, but they did not diffract well. At 25 degrees Celsius, crystals were generally smaller. A site directed mutant enzyme with His51 substituted with Ser II, with 24 differences in amino acid residues as compared to isoenzyme I also crystallized from polyethylene glycol 6000 at a pH of about 8. Ammonium sulfate and2-methyl-2,4-pentaediol precipitated the enzyme.

Crystals were transferred to an artificial medium with buffer, 2mM NAD+, 0.1 M trifluoroethanol and a polyethylene glycol concentration 5 to 10% higher than the reservoir before mounting in capillaries and being warmed slowly to 20 degrees Celsius. Low angle screen less precession photographs fit space group p622 and provided estimates of the cell dimensions. X-ray data were collected using a Rigaku AFC6R diffractometer (operating at 40 kV and 50 mA) fitted with a San Diego Multiwire Systems area detector or on a Rigaku R-Axis II image plate detector system. The diffraction limit was about 3.2 A. Results from several crystals gave unit cell dimensions of a=b=147.9 A, c=69.1A. The crystals were relatively stable, losing about 40% intensity in two days. Data from different orientation gave an R(merge) value of 7.7% at low resolution (4 A).

The assumption of a single subunit (37,500 Da) in the asymmetric unit gives a Vm value of 2.9 A^3/Da, which is within the range of commonly found values. (The Vm values for the natural horse enzyme and its complexes are 2.4) This requires that the tetrameric molecule must be centered at the point of 222 symmetry, in space group P622.

It is interesting that hexagonal leaflets of crystalline alcohol dehydrogenase form in the cells of Saccharomyces cerevisiae or S. carlsbergensis under certain groth conditions. Electron microscopy and low-angle X-ray scattering show that the molecules form a hexagonal array with the four subunits of about 4.3 nm diameter arranged tetrahedrally. These results were used to assign the crystals to space group P312, with a = 15.8nm, c=7.4nm, and six asymmetric units (6 tetrahedral molecules) per cell. This would give a relatively tight packing density of 1.8 A^3/Da.

Rotation and translation searches carried out with the programs MERLOT, X-PLOR, GLRF and ALMN in the CCP4 suite with the subunit or domains of horse liver alcohol dehydrogenase have failed to provide a solution consistent with the packing requirements in the unit cell.Work is underway to prepare heavy-atom derivatives.

No comments: