I86A/C295A mutant secondary alcohol dehydrogenase from Thermoanaerobacter ethanolicus has broadened substrate specificity for aryl ketones

Article abstract of DOI:10.1016/j.abb.2016.08.002

Thermoanaerobacter ethanolicus secondary alcohol dehydrogenase (SADH) reduces aliphatic ketones according to Prelog's Rule, with binding pockets for small and large substituents. It was shown previously that the I86A mutant SADH reduces acetophenone, which is not a substrate of wild-type SADH, to give the anti-Prelog R-product (Musa, M. M.; Lott, N.; Laivenieks, M.; Watanabe, L.; Vieille, C.; Phillips, R. S. ChemCatChem 2009, 1, 89–93.). However, I86A SADH did not reduce aryl ketones with substituents larger than fluorine. We have now expanded the small pocket of the active site of I86A SADH by mutation of Cys-295 to alanine to allow reaction of substituted acetophenones. As predicted, the double mutant I86A/C295A SADH has broadened substrate specificity for meta-substituted, but not para-substituted, acetophenones. However, the increase of the substrate specificity of I86A/C295A SADH is accompanied by a decrease in the k_cat/K_m values of acetophenones, possibly due to the substrates fitting loosely inside the more open active site. Nevertheless, I86A/C295A SADH gives high conversions and very high enantiomeric excess of the anti-Prelog R-alcohols from the tested substrates.

Full text of DOI:10.1016/j.abb.2016.08.002

Accepted Manuscript
I86A/C295A mutant secondary alcohol dehydrogenase from Thermoanaerobacter
ethanolicus has broadened substrate specificity for aryl ketones
Christopher M. Nealon, Travis P. Welsh, Chang Sup Kim, Robert S. Phillips
PII:
S0003-9861(16)30274-0
10.1016/j.abb.2016.08.002
YABBI 7342
DOI:
Reference:
To appear in: Archives of Biochemistry and Biophysics
Received Date: 29 June 2016
Revised Date: 29 July 2016
Accepted Date: 1 August 2016
Please cite this article as: C.M. Nealon, T.P. Welsh, C.S. Kim, R.S. Phillips, I86A/C295A mutant
secondary alcohol dehydrogenase from Thermoanaerobacter ethanolicus has broadened substrate
specificity for aryl ketones, Archives of Biochemistry and Biophysics (2016), doi: 10.1016/
j.abb.2016.08.002.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
I86A/C295A Mutant Secondary Alcohol Dehydrogenase from
Thermoanaerobacter ethanolicus has Broadened Substrate Specificity for Aryl
Ketones
†
¤
†¦
‡
*,†,§
Christopher M. Nealon, Travis P. Welsh, Chang Sup Kim, and Robert S. Phillips
†
Department of Chemistry, University of Georgia, Athens, GA 30602, USA
‡
Division of Applied Chemistry and Biotechnology, Hanbat National University, South Korea
Department of Biochemistry and Molecular Biology, University of Georgia, Athens, GA 30602,
§
USA
¤
Present address: Department of Chemistry, United States Naval Academy, Annapolis, MD
2
1402
¦
Present address: Medical College of Georgia, Augusta University, Augusta, GA 30912
*
Email: plp@uga.edu
KEYWORDS: mutagenesis, substrate specificity, alcohol dehydrogenase, stereospecificity

ACCEPTED MANUSCRIPT
Abstract
Thermoanaerobacter ethanolicus secondary alcohol dehydrogenase (SADH) reduces aliphatic
ketones according to Prelog’s Rule, with binding pockets for small and large substituents. It was
shown previously that the I86A mutant SADH reduces acetophenone, which is not a substrate of
wild-type SADH, to give the anti-Prelog R-product (Musa, M. M.; Lott, N.; Laivenieks, M.;
Watanabe, L.; Vieille, C.; Phillips, R. S. ChemCatChem 2009, 1, 89-93.). However, I86A SADH
did not reduce aryl ketones with substituents larger than fluorine. We have now expanded the
small pocket of the active site of I86A SADH by mutation of Cys-295 to alanine to allow
reaction of substituted acetophenones. As predicted, the double mutant I86A/C295A SADH has
broadened substrate specificity for meta-substituted, but not para-substituted, acetophenones.
However, the increase of the substrate specificity of I86A/C295A SADH is accompanied by a
decrease in the k /K values of acetophenones, possibly due to the substrates fitting loosely
cat
m
inside the more open active site. Nevertheless, I86A/C295A SADH gives high conversions and
very high enantiomeric excess of the anti-Prelog R-alcohols from the tested substrates.
+
Abbreviations: ADH: alcohol dehydrogenase; NADP : nicotinamide adenine dinucleotide
phosphate; SADH: Thermoanaerobacter ethanolicus secondary ADH
Keywords: mutagenesis; substrate specificity; stereospecificity; alcohol dehydrogenase;
thermophilic

ACCEPTED MANUSCRIPT
1
. Introduction Two alcohol dehydrogenases, one with a specificity for primary alcohols, and a
secondary alcohol dehydrogenase, were purified from Thermoanaerobacter ethanolicus, which
was isolated from hot springs in Yellowstone National Park [1, 2]. The wild-type secondary
alcohol dehydrogenase (SADH) follows Prelog’s rule in the reduction of ketones, adding the R-
hydride of the NADPH to the re-face of the ketone [3]. The substrate specificity of wild-type
SADH is restricted to primarily linear aliphatic and alicyclic ketones [4-9]. Figure 1 shows the
active site of SADH from Thermoanaerobacter brockii (formerly Thermoanaerobium brockii
[10]), which is identical with the enzyme from T. ethanolicus [11]. Thus, a number of active site
residues have been mutated in order to broaden the substrate specificity of SADH and make it
useful in biocatalytic applications. These mutant SADHs, including S39T, C295A, W110A, and
I86A, were found to have altered substrate specificity and stereoselectivity [9, 12-14].
The I86A mutation of SADH not only expands substrate specificity to include
acetophenone, which is a very poor substrate for wild-type SADH, but also reverses the usual
preferred stereochemistry to produce the anti-Prelog R-product [14]. Unfortunately, I86A SADH
exhibits limited reactivity with substituted acetophenones, fluorine being the only tolerable
substituent found in the initial study. Homochiral 1-arylalkanols are useful intermediates in the
preparation of pharmaceuticals. Thus, it was of interest to expand the active site further to allow
the reduction of substituted acetophenones. The C295A mutant SADH was studied previously
by Heiss and coworkers, and was found to increase the size of the alkyl group which can bind in
the “small pocket” by one carbon atom [12]. Due to the proximity of Cys-295 to Ile-86 (Figure
1
), we predicted that having both mutations in the active site would expand the size of the small
pocket to allow binding and reaction of ring-substituted acetophenones with subsitutents larger
than fluorine. To test this hypothesis, we have now prepared the I86A/C295A double mutant of

ACCEPTED MANUSCRIPT
SADH, and we have found, as expected, that it has broadened specificity for substituted
acetophenones.
2
. Materials and Methods
2
.1 General Methods. Gas chromatography was performed with a Varian 3300 GC (Agilent
Tech; Santa Clara, CA) using a Supelco (Sigma Aldrich; St. Louis, MO) β-Dex 120 cyclodextrin
chiral column (30 m, 0.25 mm [i.d.], 0.25 µm film thickness) with He as the carrier gas and
equipped with a flame ionization detector. Kinetic experiments and assays were performed on a
Varian Cary 100 UV-visible spectrophotometer (Agilent Tech; Santa Clara, CA) equipped with a
1
13
Peltier thermoelectric temperature-controlled 12-cell holder. H and C NMR analyses were
collected on a Varian 400MHz spectrometer with CDCl as the solvent at room temperature with
3
tetramethylsilane or the solvent peak as the reference.
2
.2 Materials. Substrates were used as purchased from commercial suppliers with the exception
of 3’-bromoacetophenone, which was prepared by a published procedure [15]. Acetophenone
was purchased from Fisher Scientific (Waltham, MA). 2,4’-Difluoroacetophenone was bought
from Acros (Geel, Belgium). The 2-acetylpyridine was purchased from Pfaltz and Bauer
(
Waterbury, CT). The rest of the substrates tested were purchased from Sigma Aldrich (St.
Louis, MO). Acetonitrile was a product of Burdick and Jackson (Morristown, NJ). NADPH was
purchased from Acros (Geel, Belgium). The antibiotics, kanamycin and ampicillin, were bought
from Roche (Indianapolis, IN) and Fisher Biotech (Fair Lawn, NJ), respectively. Commercial
grade solvents were used without further purification. The RedA-agarose was obtained from
Sigma-Aldrich (St. Louis, MO).

ACCEPTED MANUSCRIPT
2
.3 Mutagenesis. The I86A plasmid, made as reported previously [14], was isolated from the host
E. coli DH5α cells. Using the Quikchange method, the pADHB25-I86A plasmid was further
mutated with the following forward and reverse primers, with the mutation indicated in bold. The
resulting plasmid was sequenced to confirm the presence of the C295A mutation.
C295A-F (ATAAAAGGCGGGCTAGCCCCCGGTGGACG)
C295A-R (TTTCCGCCCGATCGGGGGCCACCTGCAGA).
2
.4 Purification of Secondary Alcohol Dehydrogenase. The protein was expressed and purified
based on our previously published method [9]. Modifications made to this method include the
use of 50 mM Tris HCl, pH 8.0, buffer containing 5 mM DTT and 10 µM ZnCl , a 5 mL RedA-
2
agarose column, and optimization of the low salt (0.02 M NaClO ) and high salt (0.2 M NaClO )
4
4
wash and elution solutions. No further purification was needed after the RedA-agarose column
as polyacrylamide gel electrophoresis showed a single band. The pooled fractions were stored at
o
-
80 C.
2
.5 Enzyme Assays. The enzyme was assayed as previously described [14]. The enzyme assays
o
were performed at 50 C in triplicate. The enzyme activity was measured in 50 mM potassium
o
phosphate buffer (pH 6.5 at 50 C) with 0.4 mM NADPH to follow ketone reduction. The initial
velocity was recorded on a Varian Cary 100 UV/Vis spectrophotometer at 340 nm for 10 min by
monitoring NADPH consumption (ketone reduction). The enzyme was preincubated in 50 mM
o
buffer solutionwith 5 mM DTT and 10 µM ZnCl at 50 C in the UV/Vis sample compartment
2
for 10 min before addition to the assay mixtures. The substrate stock solutions were prepared in
acetonitrile. Each assay contained not more than 5% acetonitrile, since higher concentrations
were found to inhibit the enzyme activity. The kinetic data were fit to the Michaelis-Menten

ACCEPTED MANUSCRIPT
equation (Equation 1) with the HYPERO program of Cleland [16]. The enzyme concentrations
were determined from the A280 value of 0.82 for a 0.1% solution, and using the subunit molecular
weight of 37.7 kDa.
v = Vmax*[S]/(K + [S])
(1)
m
2
.6 Asymmetric Reduction with I86A and I86A/C295A. Reactions included in 1 mL, substrate
(
0.21 mmol), 1 mg NADPH, 0.5 mL Tris-HCl buffer (8.0 pH with 10 µM ZnCl ), 0.2 mL 2-
2
propanol, 1 mg DTT and I86A or I86A/C295A SADH (0.654 mg). The reactions were incubated
o
for 24 hr at 50 C in a shaker. After 24 hr, the reactions were quenched by addition of
dichloromethane. This organic extract could be used to quantify the percent conversions by gas
chromatography. The alcohol products were identified by comparison with the standards
obtained from NaBH reduction of the ketones. Acetylation of the reaction products with acetic
4
anhydride and pyridine was performed [17]. The enantiomeric excess of the products was found
by GC analysis after acetylation. The ee values were calculated from the peak areas of the R and
S-acetates, using Equation 2.
(Area(R)-Area(S))/(Area(R)+Area(S))*100
(2)
2
.7 Hammett-Taft calculations. The curve fitting tool (cftool) in MATLAB (Version R2014b) was
used to fit the three sets of data (Hammett, Taft, and either log(kcat/(kcat ) or
)
0
log((k /K )/(k /K ) )) to the equation z = a*x+b*y. Equation 3 fitting parameters: 1.013
cat
M
cat M 0
2
2
(
SSE), 0.1435 (R ), 0.02113 (Adj. R ), 0.3805 (RMSE). Equation 4 fitting parameters: 0.2519
2
2
(
SSE), 0.6082 (R ), 0.5522 (Adj. R ), 0.1897 (RMSE). Equation 3- a=1.531 (-2.595, 5.656)
b=1.312 (0.3748, 2.25). Equation 4- a=1.52 (-0.5461, 3.586) b=1.105 (0.6524, 1.558).

ACCEPTED MANUSCRIPT
3
. Results and Discussion. In our previous report on the substrate specificity of I86A SADH,
we did not evaluate kinetic parameters for all of the aromatic substrates [14]. As shown in Table
, I86A SADH has the highest specificity constant, k /K , with unsubstituted acetophenone,
1
cat
m
consistent with our previous results. Extension of the alkyl chain by one carbon to
propiophenone results in only a 2-fold decrease in k /K . However, with increasing size of
cat
m
substituents on the ring, the k /K value decreases dramatically. Interestingly, we found in this
cat
m
study that 3’-chloroacetophenone, which we had not tested previously, reacts with I86A mutant
SADH, although the kcat/K value is decreased by about 36-fold compared to acetophenone
Table 1). We had found previously that acetophenones with other substituents larger than
fluorine are not detectable as substrates for I86A SADH, which was confirmed in this study
Table 1). Another interesting result is that 2-acetylpyridine, 4-acetylpyridine and 3-
m
(
(
acetylthiophene react much better than 3-acetylpyridine and 2-acetylthiophene with I86A SADH.
Due to the similar overall size of these heterocyclic substrates, these differences must be due to
electronic effects rather than steric effects. The electron-withdrawing effects of the pyridine ring
on the carbonyl are strongest in 2- and 4-acetylpyridine, due to resonance contributions, while
the electron-donating effect of thiophene is weakest for 3-acetylthiophene. These observations
suggest that hydride transfer to the carbonyl is at least partially rate-determining for these
substrates and facilitated by making the carbonyl carbon more electron-deficient.
The aromatic ring of these substrates must bind in the “small pocket” of I86A SADH,
since the anti-Prelog R-enantiomer is the observed reduction product. We incorporated the
C295A mutation into the I86A mutant SADH to expand the small pocket of the active site even
further, and thus we hoped to broaden the scope of acetophenone substrates that can be reduced.
While the I86A/C295A double mutant SADH unexpectedly showed no detectable reduction with

ACCEPTED MANUSCRIPT
para-substituted acetophenones, meta-substitution was accepted, as shown in Table 1. Small
ortho-substituents are also tolerated, since 2’-methylacetophenone was found to be a slow
substrate (Table 1). It is interesting that the k /K value for unsubstituted acetophenone is
cat
m
reduced about 8-fold for I86A/C295A SADH compared to I86A SADH. This suggests that the
decrease in substrate specificity is gained at the expense of reaction efficiency. In contrast, 3’-
chloroacetophenone is about 3-fold better as a substrate for I86A/C295A SADH than for I86A
SADH. The K
m
values for all acetophenones with I86A/C295A SADH were generally around 8-
1
0 mM, with some exceptions, so the kcat value is most affected by substitution, as can be seen in
Table 1. The reactivity of the acetophenones also decreases with increasing size of the meta-
substituent, and the meta-substituent can be as large as iodo, methoxy or trifluoromethyl.
Since it appeared that both electronic and steric effects influence the reactivity of
acetophenone substrates, a dual-parameter Hammett-Taft analysis [18] was performed with the
kcat and kcat/K
m
data for I86A/C295A SADH in Table 1. The plots of the results are shown in
Figure 2. The three-dimensional fit to the general equation z=ax+by for each was as follows:
Log(k /(k ) )=1.531σ + 1.312E
S
(3)
(4)
cat
cat 0
Log((k /K )/(k /K ) )=1.52σ + 1.105E
S
cat
m
cat m 0
The sensitivity factors ρ and δ (a and b of the general equation above) define how sensitive the
kinetic parameters are to Hammett σ and Taft E , respectively. From each of the best fit planes,
S
there is similar impact from the electronic effects as well as the steric effects on the reaction.
The value of ρ is identical for both kcat and k /K , 1.53 and 1.52, respectively, in Equations 3
cat
m
and 4, showing that both kinetic parameters are affected similarly by electronic effects. This
suggests that the hydride transfer step for acetophenone reduction is partially rate-determining

ACCEPTED MANUSCRIPT
for both kcat and kcat/K
m
. The positive value of ρ indicates that electron-withdrawal increases
substrate reactivity, consistent with the results from acetylpyridines and acetylthiophenes
discussed above. The δ values, 1.312 and 1.105, are slightly less than the ρ values and not
significantly different. This suggests that the larger substituents may not allow the substrate to
adopt the optimal geometry for hydride transfer on binding to the active site.
Another comparison of the efficiency of the two mutant enzymes is through preparative
reactions, as we had done previously. Table 2 compares the previously published I86A SADH
data [14], with two additional substrates, and the new I86A/C295A SADH data. These reactions
contained 0.21 mmol of substrate, corresponding to 25.2 mg in the case of acetophenone. In
general, the preparative reaction results correlate well with the kinetic results in Table 1. As
expected, 3’-chloroacetophenone reacts much better with I86A/C295A SADH than with I86A
SADH, giving conversions of 23% and 0.6%, respectively. 2-Acetylpyridine and 4-
acetylpyridine gave fairly high conversion with I86A/C295A SADH, although slightly lower
conversion than the previously published results with I86A [14]. One interesting observation
was that the percent conversion with 2-acetylthiophene and 3-acetylthiophene rose with 2-
acetylthiophene with I86A/C295A SADH, while the percent conversion fell with 3-
acetylthiophene with I86A/C295A SADH. 3’-Aminoacetophenone, 3’-nitroacetophenone, 3’-
ethoxyacetophenone, and 4’-methylacetophenone did not show any detectable reduction product
with I86A/C295A SADH, and thus would be too large for the active site of I86A SADH. In all
cases where reaction was observed, we found the product to be the anti-Prelog R-alcohol with
>
99% ee.
Reetz et al. have also performed mutagenesis studies with SADH from T. brockii in order
to change the substrate specificity and stereoselectivity [19]. For the reduction of axially chiral

ACCEPTED MANUSCRIPT
cyclohexanones, which are reduced by wild-type SADH with good yield but modest
enantioselectivity (66% ee), they found that the W110T mutant SADH produced the R-alcohols
with high conversions and >92% ee. Conversely, the I86A mutant SADH discussed in this paper
provided the S-alcohols in some cases, with stereoselectivities ranging from modest to excellent.
Recently, they used triple code saturation mutagenesis to optimize the reduction of small
saturated heterocyclic ketones which are difficult to reduce [20]. The residues in the active site
targeted for mutagenesis included Ala-85, Ile-86, Trp-110, Leu-294, and Cys-295. After an
initial round of saturation mutagenesis of each of these sites, the best mutations were grouped
into two sites for a second round of mutagenesis. Site A (A85/I86/L294/C295) was mutated
using a V-N-L triple code, and site B (A85/I86/W110/L294) was mutated with a V-Q-L triple
code. A double mutant, I86N/C295N, was identified from the site A library that reduces the
ketones to the R-alcohols in with very high ee. A triple mutant, I86V/W110L/L294Q, was found
in the site B library that provide the S-alcohols in 95% ee.
Mutagenesis has been used to modify substrate specificity of other dehydrogenases. The
carbonyl reductase from Candida parapsilosis (CPCR2) contains a zinc ion, as does SADH, and
also has activity with medium chain alcohols. Jakoblinnert and coworkers expanded the active
site of CPCR2, and thus affected the substrate specificity [21]. Five residues were selected for
site-saturation mutagenesis (Leu-55, Pro-92, Gly-118, Leu-119, Leu-262). Of the active mutants
at those locations, only L119M gave activity that was greater than wild-type CPCR2. Even
though methionine and leucine have side chains of similar size, the lack of branching in the
methionine side chain provided a less constraining CPCR2 active site. Zhu et al. obtained a
crystal structure of the carbonyl reductase, and using the structure as a guide, prepared three

ACCEPTED MANUSCRIPT
mutants, F285A, W286A, and F285A/W286A [22]. All of these mutant enzymes showed a
reversal of stereochemistry in the reduction of aryl ketones, giving the anti-Prelog products.
Zhu and coworkers also studied Sporobolomyces salmonicolor carbonyl reductase
(SsCR), with the goal to improve the activity of wild-type SsCR towards para-substituted
acetophenones [23]. They focused their site-saturation mutagenesis on the Gln-245 residue as a
result of docking studies. The mutant enzymes with the greatest improvement were Q245H,
Q245P, and Q245L, all three of which interestingly switched the stereochemistry of the alcohol
product from R to S. They utilized site-saturation mutagenesis on the Met-242 residue to probe
the SsCR active site further, and found improved properties for some of the substrates tested
[23]. Also, M242Y, M242D, M242C and M242G mutations reversed the stereochemistry of the
alcohol product. Since both Met-242 and Gln-245 mutations alter the substrate specificity
separately, they opted to design a double mutant at those sites. M242L/Q245P and
M242L/Q245T showed the largest boost in ee [23]. Broadening the substrate scope to include
benzophenones, Li and coworkers found that mutagenesis at the Gln-245 residue improved
substrate specificity as well as providing the S-isomer [24]. Zhang and coworkers recently
reported a mutant SsCR study with β-aminoketones [25]. While there was a little activity with
M242F/Q245T and β-aminoacetophenone, a docking study followed by site-saturation
mutagenesis yielded useful mutations at Pro-170 and Leu-174. L174Y and L174W were the
single mutations with the greatest activity against β-aminoacetophenone. In addition, they
designed double mutant enzymes of Pro-170 and Leu-174, with the most interesting mutant
SsCRs being P170R/L174Y, P170R/L174W, and P170H/L174W. When each of these mutated
enzymes were tested against β-amino-2-acetylthiophene, only L174Y and P170R/L174Y gave
high ee for each of the two substrates [25].

ACCEPTED MANUSCRIPT
Kavanagh and coworkers studied the active site of Candida tenuis xylose reductase
CtXR) [26], which was later found to react with α-ketoesters [27]. Through modeling studies,
(
the Trp-23 appeared to conflict with the α-carbonyl of the substrates. The tryptophan was
replaced by phenylalanine and then tyrosine using site-directed mutagenesis. The activity was
increased for each of these mutants when compared with wild-type CtXR against the various α-
ketoesters [27]. Kratzer and researchers studied the kinetics of acetophenone and other assorted
ketones reduction with W23F, W23Y and wild type CtXR [28]. While the mutations decreased
the activity with most of the ketones, oxopantoyl lactone and ethyl benzoylformate interestingly
increased in activity of four to five-fold and two to three-fold, respectively. Due to the mixed
kinetic results, the researchers rationalized that the Trp-23 has a role in selective binding and the
specificity of the enzyme [28].
4
. Conclusions. Taken together, our present results and the previously published results
demonstrate that rational mutagenesis and directed evolution are practical approaches to broaden
substrate specificity of alcohol dehydrogenases for reduction of a wide range of non-
physiological substrates. The wild-type SADH active site cannot accommodate a phenyl moiety
in the small pocket, since wild-type SADH shows negligible activity with acetophenone in our
assays [14]. Thus, the observation that only a single mutation (I86A) could open the small pocket
4
-1 -1
enough for a phenyl ring to fit inside and give a substrate with a k /K value of >10 M s
cat
m
(Table 1) is remarkable. The “small pocket” of wild-type SADH allows binding of substituents
only as large as propyl or isopropyl [4, 6, 12]. The I86A mutation removes three carbons from
the small pocket, and now the substrate of I86A SADH can have a six-carbon phenyl substituent.
Hence, one can consider the active site of these enzymes essentially as a rigid scaffold, the shape
of which can be tailored by mutagenesis to accommodate a wide range of substrates. We have

ACCEPTED MANUSCRIPT
now shown that incorporation of an additional mutation (C295A) of a neighboring residue in the
I86A SADH active site has expanded the pocket to allow reaction of m-substituted aryl ketones
(
Figure 3). The substituents that are tolerated are sterically similar to or smaller than the sulfur
atom that was removed by mutagenesis. It is noteworthy that in increasing the size of the active
site through multiple mutations, the efficiency of the enzyme decreased even as the substrate
pool was able to increase. This could be because the substrate has more freedom of motion
within the expanded active site. In conclusion, this I86A/C295A double mutation provides a
starting point for further expansion of the SADH active site to allow reaction of both m- and p-
substituted aromatic ketones.
Acknowledgments. This research did not receive any specific grant from funding agencies in
the public, commercial, or not-for-profit sectors.
Supporting Information Available. Gas chromatographic data, Hammett-Taft plot data and
kinetic assay data.
References
1
2
3
4
. Zeikus, J. G., Hegge, P. W. and Anderson, M. A. (1979) Arch. Microbiol., 122, 41-48.
. Wiegel, J. and Ljungdahl, L. G. (1981) Arch. Microbiol., 128, 343-348.
. Prelog, V. (1964) Pure Appl. Chem., 9, 119-130.
. Keinan, E., Hafeli, E. K., Seth, K. K. and Lamed, R., (1986) J. Am. Chem. Soc., 108, 162-
1
69.
. Pham, V. T., Phillips, R. S. and Ljungdahl, L. G. (1989) J. Am. Chem. Soc., 111, 1935-
936.
. Pham, V. T. and Phillips, R. S. (1990) J. Am. Chem. Soc., 112, 3629-3632.
5
6
1

ACCEPTED MANUSCRIPT
7
. Zheng, C. S., Pham, V. T. and Phillips, R. S. (1992) Bio. Med. Chem. Letts., 2, 619-622.
8
9
. Zheng, C., Pham, V. T. and Phillips, R. S. (1994) Catalysis Today, 22, 607-620.
. Tripp, A. E., Burdette, D. S., Zeikus, J. G. and Phillips, R. S. (1998) J. Am. Chem. Soc.,
1
20, 5137-5141.
0. Lee, Y. E., Jain, M. K., Lee, C., & Zeikus, J. G. (1993) Int. J. Syst. Evol. Microbiol., 43,
1-51.
1. Nealon, C. M., Musa, M. M., Patel, J. M. and Phillips, R. S. (2015) ACS Catal., 5, 2100-
114.
2. Heiss, C., Laivenieks, M., Zeikus, J. G. and Phillips, R. S. (2001) Bioorg. Med. Chem., 9,
659-1666.
1
1
1
1
1
4
2
1
3. Musa, M. M., Ziegelmann-Fjeld, K. I., Vieille, C., Zeikus, J. G. and Phillips, R. S. (2007)
J. Org. Chem., 72, 30-34.
4. Musa, M. M., Lott, N., Laivenieks, M., Watanabe, L., Vieille, C. and Phillips, R. S.
(2009) ChemCatChem, 1, 89-93.
1
1
1
1
5. Pearson, D. E., Pope, H. W. and Hargrove, W. W. (1960) Org. Syn., 40, 7.
6. Ghanem, A. and Schuring, V. (2003) Tetrahedron: Asymmetry, 14, 57-62.
7. Cleland, W. W. (1979). Methods Enzymol., 63, 103-138.
8. Taft, R. W., Jr., (1965) in “Steric Effects in Organic Chemistry”, Newman, M. S., Ed.,
John Wiley & Sons, Inc., New York, N. Y., , p. 565.
1
2
9. Agudo, R., Roiban, G. D. and Reetz, M. T. (2012) J. Am. Chem. Soc, 135, 1665-1668.
0. Sun, Z., Lonsdale, R., Ilie, A., Li, G., Zhou, J. and Reetz, M. T. (2016) ACS Catal., 6,
1
598-1605.

ACCEPTED MANUSCRIPT
2
2
2
1. Jakoblinnert, A., Wachtmeister, J., Schukur, L., Shivange, A. V., Bocola, M., Ansorge-
Schumacher, M. B. and Schwaneberg, U. (2013) Protein Eng., Des. Sel., 26, 291-298.
2. Zhu, D., Yang, Y., Majkowicz, S., Pan, T. H., Kantardjieff, K. and Hua, L. (2008) Org.
Lett., 10, 525-528.
3. Li, H., Yang, Y., Zhu, D., Hua, L. and Kantardjieff, K. (2010) J. Org. Chem., 75, 7559-
7
564.
2
2
4. Li, H., Zhu, D., Hua, L. and Biehl, E. R. (2009) Adv. Synth. Catal., 351, 583−588.
5. Zhang, D.,Chen, X., Chi, J., Feng, J., Wu, Q. and Zhu, D. (2015) ACS Catal., 5, 2452-
2
457.
6. Kavanagh, K. L., Klimacek, M., Nidetzky, B. and Wilson, D. K. (2002) Biochemistry, 41,
785-8795.
2
8
2
2
7. Kratzer, R. and Nidetzky, B. (2007) Chem. Commun., 1047-1049.
8. Kratzer, R., Leitgeb, S., Wilson, D. K. and Nidetzky, B. (2006) Biochem. J., 393, 51-58.

ACCEPTED MANUSCRIPT
Figure Legends
+
Figure 1. Stereoview of TbSADH, with residues of interest labeled. NADP shown in stick-
form and zinc as a cyan sphere. This image was prepared with Pymol (The PyMOL Molecular
Graphics System, Version 1.7 Schrödinger, LLC) using the PDB file (1YKF).
Figure 2. I86A/C295A SADH Hammett-Taft Plot. Top: Plot of substitutent effects on kcat.
Bottom: Plot of substituent effects of k /K .
cat
m
Figure 3. Cartoon showing the binding of meta-substituted acetophenone in the small pocket of
I86A/C295A SADH.

ACCEPTED MANUSCRIPT
Figure 1

ACCEPTED MANUSCRIPT
Figure 2

ACCEPTED MANUSCRIPT
Figure 3

ACCEPTED MANUSCRIPT
Table 1. I86A and I86A/C295A SADH Kinetic Parameters
O
OH
R2
SADH
R1
R2
R1
NADPH
NADP+
Substrate
I86A/C295A
I86A
1
2
-1
-1
R
R
k
cat (s )
k
cat/K
m
1 -1
k
cat (s )
kcat/K
m
(M s )
-
-1 -1
(M s )
CH
3
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
3
3
3
3
3
3
3
3
3
3
3
3
35+6
1080+110 1.01+0.16
1330+340
C H
6
15.8+1.6
12.3+2.4
6.5+1.1
1900+220
8.2+0.6
14900+3100
410+90
<1
5
3
3
3
3
2
2
3
4
2
3
’-ClC
6 5
H
1300+250 0.66+0.09
’-BrC H5
650+110
180+28
160+30
18+3
<0.01
<0.01
6
’-CH
3
C
6
H
5
5
2.0+0.33
0.61+0.06
0.18+0.04
2.31+0.23
<1
’-IC H
<0.01
<1
6
5
’-CH
3
C
6
H
<0.01
<1
-pyridyl
-pyridyl
-pyridyl
-thienyl
-thienyl
260+30
630+14
2400+400
260+60
350+60
1500+150
230+60
210+40
6.9+1.6
0.47+0.15
1.7+0.1
1.7+0.2
14.9+4.2
6.5+5.5
2.2+0.7
<0.01
1600+260
200+55
1050+101
510+85
4300+730
8100+1700
620+140
<1
a
N/A
13.1+1.9
1.5+0.2
a
N/A
C
6
H
5
CH
2
CH
3
23.3+5.5
2.9+0.7
2
3
,4-F C H
CH
3
3
2
6
3
’-CH
3
OC
6
H
5
CH
0.08+0.02

ACCEPTED MANUSCRIPT
3
’-CF
3
C
6
H
5
CH
3
0.04+0.008
25+10
<0.01
<1
a
Substrate inhibition precluded determination of kcat.

ACCEPTED MANUSCRIPT
Table 2. I86A and I86A/C295 SADH GC Assays
O
OH
SADH
R1
R2
R1
R2
NADPH
NADP+
O
OH
SADH
Substrate
I86A/C295A SADH
I86A SADH
Conv. (%)
ee (%)
R/S
Conv. ee
R/S
(
%)
(%)
1
2
R
R
a
a
a
C H
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
3
3
3
3
3
3
3
3
3
3
3
3
3
68
23
8
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
R
R
R
R
R
R
R
R
R
R
R
R
R
R
47
98
R
6
5
3
3
3
3
2
3
3
2
3
4
2
3
’-ClC H
0.6
>99
R
-
6
5
’-BrC
6
H
5
<0.1
<0.1
<0.1
<0.1
<0.1
-
-
-
-
-
-
’-IC H
4
-
6
5
’-CH
3
C
6
H
5
23
<1
5
-
’-CH C H
-
3
6
5
’-CH
3
OC
6
H
5
-
’-CF C H
5
3
<0.1
-
3
6
a
a
a
-pyridyl
-pyridyl
-pyridyl
-thienyl
-thienyl
53
75
98
27
39
31
>99 >99
R
a
a
a
a
46
>99
R
a
a
>99 >99
R
2.52 >99
R
a
a
a
a
76
60
>99
>99
R
a
a
C H
CH
2
CH
R
6
5
3

ACCEPTED MANUSCRIPT
a
a
a
2
,4-F
2
C
6
H
3
CH
3
4
>99
R
33
>99
R
a
From Musa et al. [14]

ACCEPTED MANUSCRIPT
•
•
•
•
•
•
I86A SADH reduces acetophenone to R-1-phenylethanol, but not for substituents other than F
I86A/C295A SADH was designed to accept a wider range of acetophenone substrates
I86A/C295A SADH has lower efficiency but broader substrate range
I86A/C295A SADH reduces meta-substituted acetophenones, but not para
All reduction products have >99% ee R-stereochemistry
Both steric and electronic effects affect substrate activity of I86A/C295A SADH