Bradykinetic alcohol dehydrogenases make yeast fitter for growth in the presence of allyl alcohol

Abstract

Previous studies showed that fitter yeast (Saccharomyces cerevisiae) that can grow by fermenting glucose in the presence of allyl alcohol, which is oxidized by alcohol dehydrogenase I (ADH1) to toxic acrolein, had mutations in the ADH1 gene that led to decreased ADH activity. These yeast may grow more slowly due to slower reduction of acetaldehyde and a higher NADH/NAD(+) ratio, which should decrease the oxidation of allyl alcohol. We determined steady-state kinetic constants for three yeast ADHs with new site-directed substitutions and examined the correlation between catalytic efficiency and growth on selective media of yeast expressing six different ADHs. The H15R substitution (a test for electrostatic effects) is on the surface of ADH and has small effects on the kinetics. The H44R substitution (affecting interactions with the coenzyme pyrophosphate) was previously shown to decrease affinity for coenzymes 2-4-fold and turnover numbers (V/Et) by 4-6-fold. The W82R substitution is distant from the active site, but decreases turnover numbers by 5-6-fold, perhaps by effects on protein dynamics. The E67Q substitution near the catalytic zinc was shown previously to increase the Michaelis constant for acetaldehyde and to decrease turnover for ethanol oxidation. The W54R substitution, in the substrate binding site, increases kinetic constants (Ks, by >10-fold) while decreasing turnover numbers by 2-7-fold. Growth of yeast expressing the different ADHs on YPD plates (yeast extract, peptone and dextrose) plus antimycin to require fermentation, was positively correlated with the log of catalytic efficiency for the sequential bi reaction (V1/KiaKb=KeqV2/KpKiq, varying over 4 orders of magnitude, adjusted for different levels of ADH expression) in the order: WT≈H15R>H44R>W82R>E67Q>W54R. Growth on YPD plus 10mM allyl alcohol was inversely correlated with catalytic efficiency. The fitter yeast are "bradytrophs" (slow growing) because the ADHs have decreased catalytic efficiency.

Fig. 1

Stereo view of one subunit of the tetrameric yeast alcohol dehydrogenase showing the locations of the amino acid residues that were substituted. NAD, TFE (trifluoroethanol) and the catalytic zinc atom are shown. The four substituted arginine residues are in magenta near the residue number. Residue 67 is a glutamic acid, which is behind the catalytic zinc and was substituted with a glutamine in a previous study [9]. The structure of the yeast ADH was determined at 2.4 Å (PDB ID: 2HCY).

Fig. 2

Native, non-denaturing, polyacrylamide gel electrophoresis of purified, mutated ADHs. The full gel is shown with the origin at the top. The dashed line indicates the position of wild-type enzyme, and numbers at the dye front indicate the expected change in charge due to the substitutions. The gel was 6% acrylamide (12 cm high, 14 cm wide, 1 mm thick) and used the high pH discontinuous buffer system, which separates at pH 9.5 [39]. The high pH is required to separate the wild-type enzyme (with histidine residues) from the forms that retain an additional positive charge above pH 9 (as indicated on the figure). Commercial yeast ADH, which comes from Saccharomyces carlsbergensis, has 2 additional negative charges [18]. The E67Q enzyme also migrates with more positive charge [9]. About 1 unit of enzyme activity in a loading buffer with 10% sucrose, 0.05% Bromophenol Blue, and 0.2 mg/ml of -lactalbumin (to stabilize enzyme activity) was loaded into each lane. The gel was run for one hour at 25 mA, and then for 3 h at 200 V at 4 ^oC to avoid overheating. The samples were stained for activity (0.1 M Tris-HCl, pH 8, 0.25 M ethanol, 1 mM NAD⁺, 0.024 mg/ml phenazine methosulfate, 0.4 mg/ml nitroblue tetrazolium) at room temperature and destained in water. (The mutated enzymes can also be separated on a thin 1% agarose gel in 50 mM sodium glycine buffer, pH 9.0, run at 100 V and 2 mA for 1.5 h at 5 °C.)

Fig. 3

Model for the substitution in the W82R enzyme. The arginine residue can overlay most atoms of the tryptophan residue and might be stabilized by electrostatic interactions with Asp-86. However, the hydrophobic interactions with His-138 and Leu-33 within the site may be diminished.

Fig. 4

Model for the substitution in the W54R enzyme. A sterically acceptable substitution of Trp-54 could overlay most atoms and place the guanidino group of Arg-54 into the substrate binding site near the trifluoromethyl group of TFE. However, burying the positively-charged group may be energetically unacceptable, and the structure may undergo considerable changes.

Fig. 5

Relative growth rates of transformed yeast on selective media. Yeast strains were grown to OD600/cm of 4-5 in SC-Leu media at 30 °C with shaking at 200 rpm and then diluted to different OD levels before spotting 5 μl of culture on SC-Leu, YPD + 1 μg/ml antimycin A, YPD + 1 mM allyl alcohol, and YPD + 10 mM allyl alcohol agar plates. Plates were incubated at 30 °C for 3 days with cell growth recorded each day. All cultures grew on YPD, but none grew on minimal media lacking tryptophan.