Redesign of cosubstrate specificity and identification of important residues for substrate binding to hChAT

Article abstract of DOI:10.1021/bi1007996

In eukaryotes, choline acetyltransferase (ChAT) catalyzes the reversible formation of the neurotransmitter acetylcholine from choline and acetyl-CoA. ChAT belongs to a family of CoA-dependent enzymes that also includes the carnitine acyltransferases CrAT, CrOT, and CPTs. In contrast to CrOT and CPTs that are very active toward medium- and long-chain acyl-CoAs, respectively, CrAT and ChAT display activity toward only short-chain acyl-CoAs. We recently demonstrated the substrate and cosubstrate promiscuity of the wild-type human ChAT (hChAT). To extend the flexibility of this enzyme, we have generated a series of single, double, and triple hChAT mutants. Here we report the conversion of hChAT into choline octanoyltransferase (ChOT) and choline palmitoyltransferase (ChPT). The E337 and C550 residues (numbering from hChAT) were previously shown to dictate the acyl-CoA cosubstrate specificity in the carnitine series. Here we identify and demonstrate the importance of C551, in addition to E337 and C550, in contributing to the acyl-CoA specificity of hChAT. We also show that either C550 or C551 needs to be present for the transfer of medium- and long-chain acyl-CoAs by hChAT. By exploring the potential expansion of the tunnel on the substrate side, we demonstrate that residues M84, Y436, and Y552 play a critical role in binding and holding the choline substrate in the ChAT active site.

Full text of DOI:10.1021/bi1007996

Biochemistry 2010, 49, 6219–6227 6219
DOI: 10.1021/bi1007996
Redesign of Cosubstrate Specificity and Identification of Important Residues for Substrate
Binding to hChAT†
Keith D. Green,§ Vanessa R. Porter,‡,§ Yaru Zhang,§, and Sylvie Garneau-Tsodikova*,‡,§,
‡Department of Medicinal Chemistry, §Life Sciences Institute, 210 Washtenaw Avenue, and Chemical Biology Doctoral Program,
University of Michigan, Ann Arbor, Michigan 48109-2216
Received May 19, 2010; Revised Manuscript Received June 15, 2010
ABSTRACT: In eukaryotes, choline acetyltransferase (ChAT) catalyzes the reversible formation of the
neurotransmitter acetylcholine from choline and acetyl-CoA. ChAT belongs to a family of CoA-dependent
enzymes that also includes the carnitine acyltransferases CrAT, CrOT, and CPTs. In contrast to CrOT and
CPTs that are very active toward medium- and long-chain acyl-CoAs, respectively, CrAT and ChAT display
activity toward only short-chain acyl-CoAs. We recently demonstrated the substrate and cosubstrate
promiscuity of the wild-type human ChAT (hChAT). To extend the flexibility of this enzyme, we have
generated a series of single, double, and triple hChAT mutants. Here we report the conversion of hChAT into
choline octanoyltransferase (ChOT) and choline palmitoyltransferase (ChPT). The E337 and C550 residues
(numbering from hChAT) were previously shown to dictate the acyl-CoA cosubstrate specificity in the
carnitine series. Here we identify and demonstrate the importance of C551, in addition to E337 and C550, in
contributing to the acyl-CoA specificity of hChAT. We also show that either C550 or C551 needs to be present
for the transfer of medium- and long-chain acyl-CoAs by hChAT. By exploring the potential expansion of the
tunnel on the substrate side, we demonstrate that residues M84, Y436, and Y552 play a critical role in binding
and holding the choline substrate in the ChAT active site.
Choline acetyltransferase (ChAT)1 catalyzes the reversi-
ble transfer of the acetyl moiety of acetyl-CoA to choline for
the formation of the neurotransmitter acetylcholine (ACh)
(Figure 1A). ACh is widely distributed in the central and
peripheral nervous systems as well as in non-nervous tissues
and thus plays an important role in a variety of brain and basic
cellular functions. A number of syndromes and neurodegenera-
tive disorders, including Alzheimer’s disease (AD) (1), amyo-
trophic lateral sclerosis (2), schizophrenia (3), Huntington’s
disease (HD) (4), sudden infant death syndrome (5), and Rett
syndrome (6), have been correlated with diminished ACh levels
associated with decreased ChAT activity.
ChAT belongs to a family of CoA-dependent acyltransferases
that also includes dihydrolipoyl transacetylase, chloramphenicol
acetyltransferase (CAT), carnitine acetyltransferase (CrAT),
carnitine octanoyltransferase (CrOT), and carnitine palmitoyl-
transferases (CPTI and CPTII) (7). The carnitine acyltransferases
are enzymes involved in the β-oxidation of fatty acids that
catalyze a reaction similar to that catalyzed by ChAT but utilize
carnitine instead of choline as their substrate and differ in the
F
(ChAT) and (B) carnitine acyltransferases (CrAT, CrOT, and CPT).
IGURE 1: Reactions catalyzed by (A) choline acetyltransferase
length (varying from 2 to 16 carbons) of their acyl-chain co-
substrate selectivity (Figure 1B). Carnitine and choline differ only
in their substituents at C1, where a carboxymethyl group of car-
nitine replaces a hydrogen atom in choline. Short-, medium-, and
long-chain fatty acids are favored by CrAT, CrOT, and CPTs,
respectively. A decrease in CrAT activity has been associated
with AD (8), several vascular diseases (9), and ataxic encephalo-
pathy (10).
The crystal structures of rat ChAT (rChAT) (11, 12), human
ChAT (hChAT) (13), human CrAT (hCrAT) (14), mouse CrAT
(mCrAT) (15), mouse CrOT (mCrOT) (16), two mutants of
mCrAT (17), and rat CPTII (rCPTII) (18) have been reported.
All these structures are strikingly similar and reveal a common
†This work was supported by start-up funds from the Life Sciences
Institute and the College of Pharmacy at the University of Michigan (to
S.G.-T.). V.R.P. was supported by a Summer Institute Rackham Merit
Fellowship (University of Michigan).
*To whom correspondence should be addressed. E-mail: sylviegt@
umich.edu. Phone: (734) 615-2736. Fax: (734) 615-5521.
1Abbreviations: CoA, coenzyme A; CrAT, carnitine acetyltransferase;
CrOT, carnitine octanoyltransferase; CPT, carnitine palmitoyltransfer-
ase; ChAT, choline acetyltransferase; ChOT, choline octanoyltransferase;
ChPT, choline palmitoyltransferase; EDTA, ethylenediaminetetraacetic
acid; LB, Luria-Bertani broth; SDS-PAGE, sodium dodecyl sulfate-
polyacrylamide gel electrophoresis; Tris-HCl, tris(hydroxymethyl)ami-
nomethane hydrochloride.
r
2010 American Chemical Society
Published on Web 06/18/2010
pubs.acs.org/Biochemistry

6220 Biochemistry, Vol. 49, No. 29, 2010
Green et al.
F
IGURE 2: Sequence alignment of human ChAT, CrAT, CrOT, and L-CPTI. White letters with a red background denote residues that are fully
conserved, whereas red letters denote residues that are highly conserved. Colored circles indicate positions mutated in this study. Yellow circles
indicate large residues in the channel on the choline side that we mutated to increase the size of the channel. Turquoise circles indicate mutations
made in the CrAT series also made in this study. Dark blue circles indicate residues that we mutated to the residues in CrOT or L-CPTI. The green
circle indicates the histidine residue located in the middle of the active site of the enzymes. Orange circles indicate residues mutated by others to
convert ChAT into CrAT. The double green line at the right of the top section indicates a gap in the sequences where no mutations where made.
active site located at the interface of the two domains of the
enzyme. The active site contains a histidine residue [H324 in
hChAT (Figure 2)] required for catalysis positioned in the center
of a catalytic tunnel with the choline substrate and CoA cosub-
strate lying on opposite sides of the tunnel (19-21). These
structures not only provided insights into the mechanism of
acyl-chain transfer by these enzymes but also highlighted the
important and significant structural differences in their choline/
carnitine as well as in their acyl moiety binding regions. A
structural model of liver CPTI (L-CPTI) based on the CrAT
crystal structure along with kinetic analyses of L-CPTI mutants
demonstrated the essential catalytic role of D477 and D567
[corresponding to D328 and D414 of hChAT, respectively
(Figure 2)], located near the carnitine and the CoA, respec-
tively (22).
In addition to the structural data, reports on the redesign of
CrAT into CrOT (23) and CPT (24) indicated that D356 and
M564 residues in rCrAT [corresponding to E337 and C550 in
hChAT, respectively (Figure 2)] are critical for dictating the
length of the acyl-CoA chains transferred by these enzymes.
When compared to the wild-type enzyme, CrAT-M564G
and CrAT-M564A mutants displayed higher activity toward
medium- to long-chain acyl-CoAs (6-14 carbons) and lower
activity toward the natural cosubstrate acetyl-CoA (23). The
introduction of an additional mutation site into the CrAT-
D356A/M564G double mutant increased the activity toward
the long-chain acyl-CoAs (10-18 carbons) (24).
M84 residue could potentially increase the size of the tunnel to
allow promiscuity for choline substrates with a larger group on
their quaternary ammonium center (Figure 3A). A combination
of protein alignment data (Figure 2) and structural data
(Figure 3A) also suggested that targeting residue Y436 or Y552
could achieve the same purpose. In this study, we interrogate the
role that three amino acid residues (M84, Y436, and Y552) play
in binding and locking of choline and its analogues into the active
site tunnel of hChAT, using single-point mutations.
Herein, we also report the conversion, by site-directed muta-
genesis, of hChAT into choline octanoyltransferase (ChOT) and
choline palmitoyltransferase (ChPT). In addition to the pre-
viously reported E337 and C550 residues (numbering of hChAT)
in the CrAT series, we have identified a new amino acid residue
(C551) that contributes to the acyl-CoA specificity in hChAT.
The hChAT-C551A mutant showed a decrease in activity toward
the natural cosubstrate acetyl-CoA but displayed an overall
increase in specificity toward all other short-, medium-, and
long-chain acyl-CoAs tested. The double hChAT-E337A/C551A
mutant exhibited a preference for medium- and long-chain acyl-
CoA cosubstrates when compared to the single hChAT-C551A
mutant. Interestingly, our work reveals a completely different
cosubstrate specificity profile for the hChAT-E337A/C550G,
hChAT-C550A, and hChAT-C550G mutants compared to those
of their counterparts in the CrAT series. Furthermore, by pre-
paration of a variety of double and triple mutants, we show that
the presence of at least one of the cysteine residues (C550 or
C551) is required for hChAT specificity toward medium- and
long-chain acyl-CoAs. We also demonstrated that mutation of
valine 333 to alanine increases the activity of the enzyme with the
natural cosubstrate acetyl-CoA.
A study of the conversion of rChAT into CrAT showed that
mutating ChAT residues V449, D450, N451, and N504 to T, E,
T, and R, respsectively (Figure 2), accommodated the additional
carboxymethyl group of carnitine (25). In addition to the
electrostatic and steric factors that govern the ChAT substrate
specificity, a close analysis of the hChAT crystal structure (13) in
conjunction with our recent report on the substrate and cosub-
strate promiscuity of hChAT (26) suggested that mutation of the
MATERIALS AND METHODS
Bacterial Strains, Plasmids, Materials, and Instrumenta-
tion. The pProEX-HTa plasmid containing the hChAT gene

Article
Biochemistry, Vol. 49, No. 29, 2010 6221
F
IGURE 3: Crystal structure of hChAT (Protein Data Bank entries 2FY3 and 2FY4) showing at the top of both panels the tunnels on (A) the
choline side and (B) the CoA side. The bottom of both panels focuses on the residues mutated in this study (displayed as sticks and highlighted in
dark red on the choline side and in pink on the CoA side). Choline and CoA are colored yellow. The sulfur atom to which the various acyl chains
are tethered in thioester linkages to CoA is indicated by a black arrow in panel B. Adaptive Poisson-Boltzmann Solver (APBS) (45) was used as a
tool within PyMOL (46) to calculate charges. Settings -12 and þ12 were used for negative and positive electrostatic potential, respectively.
was generously provided by B. Shilton (University of Western
Ontario, London, ON). Chemically competent Escherichia coli
TOP10 and BL21(DE3) cell strains were purchased from Invi-
trogen (Carlsbad, CA). Restriction endonucleases, T4 DNA
ligase, and Phusion DNA polymerase were bought from New
England BioLabs (NEB, Ipswich, MA). DNA primers for
polymerase chain reaction (PCR) were obtained from Integrated
DNA Technologies (IDT, Coralville, IA). Int-pET19b-pps con-
taining an N-terminal decahistidine tag separated from the gene
by a precision protease cleavage site was generously provided by
T. Biswas (Univeristy of Michigan) (27). DNA sequencing was
performed at the University of Michigan DNA Sequencing Core.
All CoA derivatives, aminocholine, choline, and buffer compo-
nents were purchased from Sigma-Aldrich (Milwaukee, WI) and
used without any further purification. We previously reported the
synthesis and characterization of all noncommercially available
choline derivatives utilized in this study (Figure S1 of the
Supporting Information) (26). Fast protein liquid chromatogra-
phy (FPLC) was performed as the last protein purification step
on a Bio-Rad (Hercules, CA) BioLogic DuoFlow system using
a HighPrep 26/60 Sephacryl S-200 high-resolution column
(GE Healthcare, Piscataway, NJ). Spectrophotometric assays
were performed in a high-throughput fashion using a Spectra-
Max M5 microplate reader.
DNA polymerase as described by NEB. Wild-type (wt) hChAT
was constructed in Int-pET19b-pps to serve as a control. Mutants
were constructed using the SOE method (28). In the first round
of PCR, the sequences downstream and upstream of the muta-
tion(s) were separately amplified using the hChAT-pProEX-HTa
plasmid as a template in conjunction with the 50 primer for the
mutant with the 30 primer for hChAT and the 50 primer for hChAT
with the 30 primer for the mutant, respectively (Table S1). The
resulting amplified PCR fragments were gel-purified and sub-
jected to a second round of PCR using the forward and reverse
primers for hChAT (Table S1). The newly amplified fragments
were then digested with NdeI and XhoI and subcloned into the
linearized Int-pET19b-pps vector via the corresponding NdeI and
XhoI restriction sites to give the single and double mutants
M84A, E337A, E337Y, C550G, C550A, C551A, C550G/C551A,
C550A/C551A, V333A, V542A, Y436A, and Y552A. To con-
struct the double and triple mutants V333A/E337A, E337A/
C550G, E337A/C550A, E337A/C551A, E337A/C550G/C551A,
E337A/C550A/C551A, E337Y/C550G, E337Y/C550A, E337Y/
C551A, E337Y/C550G/C551A, and E337Y/C550A/C551A, the
newly made V333A, E337A, and E337Y mutants were utilized as
templates. All expression clones were characterized by DNA
sequencing (University of Michigan DNA Sequencing Core).
Overproduction and Purification of hChAT and hChAT
Mutants. All proteins were expressed in E. coli BL21(DE3)
grown in LB medium (2 ꢀ 1 L) supplemented with ampicillin
(100 μg/mL), flash-frozen, and stored at -80 °C as previously
reported (Figure S2 of the Supporting Information) (26). Protein
Preparation of hChAT and hChAT Mutant Overexpres-
sion Constructs. The primers used for the amplification of
hChAT and mutant hChAT genes are listed in Table S1 of the
Supporting Information. PCRs were conducted using Phusion

6222 Biochemistry, Vol. 49, No. 29, 2010
Green et al.
Table 1: Kinetic Parameters of hChAT-wt and hChAT-C551A Determined with Aminocholinea
enzyme (vector)
CoA derivative
Km (μM)
kcat (s-1
)
kcat/Km (μM-1 s-1
)
hChAT-wt (pProEX-HTa)b
acetyl-CoA
19.4 ( 1.1c
12.7 ( 5.2
20.0 ( 4.7
60.8 ( 6.1
13.7 ( 3.2
4.39 ( 0.47
1.990 ( 0.012
0.197 ( 0.065
0.613 ( 0.028
7.07 ( 0.31
1.18 ( 0.06
0.226 ( 0.003
0.103
0.016
0.031
0.116
0.086
0.052
n-propionyl-CoA
n-propionyl-CoA
acetyl-CoA
hChAT-wt (Int-pET19b-pps)
hChAT-C551A (Int-pET19b-pps)
n-propionyl-CoA
n-butyryl-CoA
aKinetic parameters were determined at pH 7.5. bThe kinetic parameters for hChAT-wt (pProEX-HTa) were previously reported (26). cAll errors were
established from at least three independent trials of kinetic experiments.
concentrations were determined using a NanoDrop ND-1000
spectrophotometer (Thermo Scientific).
Kinetic Characterization of hChAT-wt and the hChAT-
C551A Mutant. We previously reported the expression, pur-
ification, and characterization of hChAT utilizing an hChAT-
pProEX-HTa plasmid (26). In this study, we cloned the hChAT
gene into the Int-pET19b-pps vector. Prior to construction of the
hChAT mutant clones using the Int-pET19b-pps vector, we
determined kinetic parameters for hChAT-wt (Int-pET19b-pps)
by UV-vis assays by holding the concentration of hChAT at a
final concentration of 60 nM and by varying the concentration of
aminocholine and the different cosubstrates and compared them
to our previously reported data (Table 1). Aminocholine was
used as it was previously found to be a better substrate for
hChAT-wt than choline. We found the Km value for n-propionyl-
CoA for hChAT-wt expressed from Int-pET19b-pps (20.0 (
4.7 μM) to be in good agreement with the Km measured using
hChAT-wt produced from pProEX-HTa (12.7 ( 5.2 μM).
The catalytic turnover (kcat) of 0.613 ( 0.028 s-1 and catalytic
efficiency (kcat/Km) of 0.0310 μM-1 s-1 were found to be 3- and
2-fold higher for hChAT-wt expressed from Int-pET19b-pps,
respectively. We therefore decided to construct all hChAT
mutants using the Int-pET19b-pps vector. As we identified
C551 as a novel important residue for dictating the length of
the acyl chain accepted by hChAT, we also tested the kinetic
parameters of the hChAT-C551A mutant (Table 1). With
hChAT-C551A and acetyl-CoA, a Km of 60.8 ( 6.1 μM and a
catalytic turnover of 7.07 ( 0.31 s-1 resulting in a catalytic
efficiency 0.116 μM-1 s-1 were determined. The C551A mutant
kinetic parameters were also determined for n-propionyl-CoA
and n-butyryl-CoA. Here, Km values of 13.7 ( 3.2 and 4.39 (
0.47 μM and catalytic turnover values of 1.18 ( 0.06 and 0.226 (
0.003 s-1 resulted in catalytic efficiencies of 0.086 and 0.052
μM-1 s-1, respectively.
Identification of Three Essential Residues for Choline
Binding to hChAT. We have recently demonstrated the ability
of hChAT to accept choline derivatives containing a variety of
quaternary ammonium centers and a hydroxyl or an amino
functionality separated by two methylene groups (26). Cronin
showed that ChAT can be converted to CrAT (25), suggesting the
ability of ChAT to be mutated to accommodate larger choline
derivatives. It is known that in hChAT, the entrance to the
substrate tunnel is much wider than that required to accommo-
date choline and that the tunnel narrows near the choline-binding
site, resulting in specificity for choline analogues in which only
one of the three methyl groups of the quaternary ammonium
center is modified. A close analysis of the hChAT crystal struc-
ture showed that residues M84, Y436, and Y552 are found in the
narrow part of the substrate tunnel (Figure 3A). With the idea
of expanding the promiscuity of hChAT toward choline analo-
gues with larger quaternary ammonium centers, we generated
hChAT mutants M84A, Y436A, and Y552A. When tested with
Characterization of hChAT and Its Mutants. Spectro-
photometric assays for determining the substrate specificity of
the hChAT mutants M84A, Y436A, and Y552A monitored at
324 nm the production of 4-thiopyridone from the reaction of
4,40-dithiodipyridine (DTDP) with the CoA formed by acylation
of a variety of choline analogues (Figure S1) (29). The reaction
mixtures (200 μL) containing hChAT (60 nM), DTDP (2 mM),
and a choline analogue (2 mM) were in HEPES buffer (50 mM,
pH 7.5 adjusted at room temperature). Reactions were initiated
with acetyl-CoA (100 μM) and mixtures incubated at 25 °C for
30 min taking measurements every 30 s in 96-well plates.
Similarly, spectrophotometric assays were performed to deter-
mine the cosubstrate specificity of hChAT-wt and the rest of the
hChAT mutants. In these cases, the reaction mixtures (100 μL)
contained hChAT (60 nM), DTDP (2 mM), and aminocholine
(2 mM) in HEPES buffer (50 mM, pH 7.5 adjusted at room
temperature), and reactions were initiated with CoA derivatives
(100 μM).
For the determination of kinetic parameters for the CoA deri-
vatives accepted by hChAT-wt and hChAT mutants, the concen-
tration of the enzymes was 60 nM, the concentration of the amino-
choline substrate was kept constant at 5 mM, and the concentration
of the CoA derivatives was varied from 0 to 1 mM. All determina-
tions of Km and kcat were conducted in triplicate. Parameters were
calculated using Kaleidograph curve fitting software.
RESULTS
Heterologous Expression and Purification of hChAT
and Its Mutants. hChAT-wt and all hChAT mutants were
heterologously expressed in E. coli as N-terminally His10-tagged
proteins to establish their substrate and cosubstrate specificity
profiles. Enzymes used in activity assays were purified to homo-
geneity by Ni(II)-NTA affinity chromatography and flash-frozen
for storage at -80 °C. Enzymes used in kinetic analysis were
subjected to an additional purification step by size exclusion
chromatography prior to being stored. As judged by size exclu-
sion chromatography, hChAT and its mutants appear to be
purified in monomeric form. Protein yields (in milligrams per liter
of culture) were 0.61 (wt), 1.80 (M84A), 0.66 (E337A), 0.89
(E337Y), 1.33 (C550G), 0.18 (C550A), 0.46 (C551A), 1.19
(C550G/C551A), 1.15 (C550A/C551A), 0.66 (E337A/C550G),
0.44 (E337A/C550A), 0.88 (E337A/C551A), 0.06 (E337A/C550G/
C551A), 16.2 (E337A/C550A/C551A), 0.66 (E337Y/C550G),
3.38 (E337Y/C550A), 1.19 (E337Y/C551A), 0.49 (E337Y/C550G/
C551A), 12.9 (E337Y/C550A/C551A), 1.43 (Y436A), 0.97 (Y552A),
2.31 (V333A), 1.53 (V542A), and 3.78 (V333A/E337A) (Figure S2
of the Supporting Information).

Article
Biochemistry, Vol. 49, No. 29, 2010 6223
acetyl-CoA, both M84A and Y552A were found to be completely
inactive toward choline and all its analogues tested (Figure S3 of
the Supporting Information). With acetyl-CoA as a cosubstrate,
aminocholine proved to be a weak hChAT-Y436A substrate
(24% activity compared to that of hChAT-wt) (data not shown).
These results demonstrate the importance of the M84, Y436, and
Y552 residues in holding the choline molecule in the hChAT
active site.
Alteration of hChAT Activity and Cosubstrate Specifi-
city by Site-Directed Mutagenesis. Because of the previous
successes in converting CrAT into CrOT and CPT, we predicted
that the cosubstrate specificity profile of hChAT could be
modified to those of ChOT and ChPT to accommodate medium-
and long-chain acyl-CoAs. To confirm this hypothesis, we
mutated E337 to alanine to mimic one of the double mutations
in the CrAT to CPT alteration (Figure 2) (24). Similarly, the same
residue was mutated to tyrosine, because its counterpart in CrOT
is a tyrosine. C550 was mutated to glycine and alanine to emu-
late the previous studies mutating CrAT into CPT and CrOT,
respectively (23). As the counterpart amino acid residues in CrOT
and CPT are smaller in size than the cysteine at position 551,
C551 was mutated to alanine to compensate for the larger size of
incoming CoAs. Combinations of all three mutations resulted
in the double and triple mutants, which we believed would
allow for promiscuity toward larger CoA cosubstrates. V333 and
V542 located on the R-helix adjacent to the CoA binding site
(Figure 3B) were also mutated to alanine with the thought that
these mutations would allow the longer CoA derivatives to fit
adjacent to the helix.
To determine the activity of the hChAT mutants, we tested the
enzymes with the substrate aminocholine and several CoA co-
substrates, including acetyl-, n-propionyl-, n-butyryl-, hexanoyl-,
octanoyl-, decanoyl-, lauroyl-, and palmitoyl-CoA. The activities
of all hChAT mutants and CoA derivatives were normalized to
the rate of the reaction of acetyl-CoA with aminocholine and
recombinant wild-type hChAT. Figure 4 and Figure S3 of
the Supporting Information show a comparison of the activity of
the hChAT mutants with the eight CoAs tested. hChAT mu-
tants E337A, E337Y, C550G/C551A, C550G, E337A/C550G/
C551A, E337A/C550A/C551A, E337Y/C550G, E337Y/C550G/
C551A, E337Y/C550A/C551A, V542A, and V333A/E337A
(Figure S3E-O) showed no or significantly reduced activity with
all CoA derivatives. The wild-type hChAT enzyme showed activity
with acetyl-CoA (100%) and n-propionyl-CoA (28%), whereas all
other CoAs had <8% activity (Figure 4A). The E337Y/C551A
and C550A/C551A hChAT mutants had 30 and 27% activity with
acetyl-CoA, respectively (Figure S3C,D). hChAT-E337Y/C551A
displayed low activity with all remaining CoAs, while hChAT-
C550A/C551A showed activity with n-propionyl-CoA (18%)
and n-butyryl-CoA (12%) and <5% activity with the remaining
CoAs. hChAT-E337A/C550A (Figure S3B) showed activity with
acetyl-CoA (12%) and <6% activity with the other CoAs.
Interestingly, hChAT-V333A displayed an increase in activity
with acetyl-CoA (134%) and activity with n-propionyl-CoA
(17%), but <5% activity was observed for the remaining CoAs
(Figure S3A). hChAT-C551A displayed activity with acetyl-CoA
(68%), n-propionyl-CoA (59%), n-butyryl-CoA (22%), and hex-
anoyl-CoA (12%), but <4% activity with the longer CoAs
(Figure 4B). Mutating hChAT residues E337 and C551 to alanine
resulted in activity with acetyl-CoA (27%), n-propionyl-CoA
(40%), n-butyryl-CoA (42%), hexanoyl-CoA (14%), octanoyl-
CoA (22%), and decanoyl-CoA (18%), and <5% activity with
F
IGURE 4: Histograms of hChAT mutants comparing the rates of
different CoA derivatives, normalized to the rate of the wild-type
enzyme with acetyl-CoA, for each mutant: (A) wild-type hChAT, (B)
hChAT-C551A, (C) hChAT-E337A/C551A, (D) hChAT-C550A,
(E) hChAT-E337Y/C550A, and (F) hChAT-E337A/C550G. The
number 2 denotes acetyl-CoA, the number 3 n-propionyl-CoA, the
number 4 n-butyryl-CoA, the number 6 hexanoyl-CoA, the number 8
octanoyl-CoA, the number 10 decanoyl-CoA, the number 12 lauroyl-
CoA, and the number 16 palmitoyl-CoA.
lauroyl- and palmitoyl-CoA (Figure 4C). Mutation of C550 to
alanine resulted in activity with acetyl-CoA (52%), n-propionyl-
CoA (34%), n-butyryl-CoA (18%), hexanoyl-CoA (10%), octa-
noyl-CoA (23%), decanoyl-CoA (13%), lauroyl-CoA (17%),
and palmitoyl-CoA (20%) (Figure 4D). The mutant with the
best overall activity was found to be hChAT-E337Y/C550A,
with activity with acetyl-CoA (211%), n-propionyl-CoA (96%),
n-butyryl-CoA (59%), hexanoyl-CoA (29%), octanoyl-CoA

6224 Biochemistry, Vol. 49, No. 29, 2010
Green et al.
(35%), decanoyl-CoA (26%), and lauroyl- and palmitoyl-CoA
(24%) (Figure 4E). Finally, even though it displays low activity
(14% with octanoyl-CoA), the mutant hChAT-E337A/C550G
was found to be highly specific toward octanoyl-CoA as all other
CoAs tested had <1% activity (Figure 4F).
V333A enzyme displayed a 1.3-fold increase in activity. Interest-
ingly, the doubly mutated hChAT-E337Y/C550A showed an
unexpected 2.1-fold increase in activity. All other mutant enzymes
showed a decrease in activity when reacted with aminocholine and
acetyl-CoA. In contrast to what we predicted, mutation of valine
542 to alanine (V542A) resulted in a significant decrease in
activity with all CoA derivatives tested (Figure 5A and Figure
S3N of the Supporting Information). hChAT-E337A exhibited a
4-fold reduction in activity, and E337Y exhibited a 2.5-fold
reduction. The C550G mutant exhibited a 10-fold reduction in
activity, and when the same residue was mutated to an alanine
(C550A), a 2-fold reduction was observed. hChAT-C551A ex-
hibited a 1.5-fold reduction in activity.
Even though the hChAT-E337Y/C550A mutant was found to
be the most active when reacted with acetyl-CoA, it was observed
that double mutations generally resulted in a larger decrease in
activity with acetyl-CoA than single mutations. In fact, an
additional mutation to hChAT-V333A, naturally more active
than hChAT-wt, led to the hChAT-V333A/E337A mutant dis-
playing a 2.5-fold rate reduction. As expected, changing both
glutamate 337 to alanine and cysteine 550 to alanine resulted in a
10-fold reduction in activity, a greater reduction than that
observed for the corresponding single mutants hChAT-E337A
and hChAT-C550A. Making the E337A/C550G mutations
nearly abolished all activity with acetyl-CoA. Mutation of both
cysteine residues resulted in 3.5-fold (C550A/C551A) and 7.7-
fold (C550G/C551A) decreases in activity. Interestingly, in
constrast to hChAT-E337Y/C550A that displays high activity
with acetyl-CoA, the hChAT-E337A/C551A, -E337Y/C550G,
and -E337Y/C551A mutants exhibited 3.7-, 50-, and 3.5-fold
decreases in activity, respectively. All triple mutants resulted in an
at least 20-fold reduction in activity. These results not surpris-
ingly indicate that increasing the overall size of the hChAT tunnel
on the cosubstrate side generally leads to a decrease in the affinity
for the natural acetyl-CoA cosubstrate.
However, as expected, several of the double hChAT mutants,
including E337A/C550G, E337A/C551A, and E337Y/C550A,
were found to nicely accommodate at least one of the larger CoA
derivatives that were tested (Figure 5B,C). Most notable among
these mutants is hChAT-E337Y/C550A, for which a 2.1-100-
fold increase in activity is observed for the transfer of all short-,
medium-, and long-chain acyl-CoAs tested. Not only is hChAT-
E337Y/C550A more active with the natural acetyl-CoA cosub-
strate than hChAT-wt (2.1-fold increase in activity), it is also the
most active enzyme generated against all CoA derivatives tested.
Indeed, with hChAT-E337Y/C550A, 3.4-, 7-, 7-, 57-, 130-, 120-,
and 24-fold increases in activity are observed during the transfers
of the n-propionyl, n-butyryl, hexanoyl, octanoyl, decanoyl,
lauroyl, and palmitoyl chains, respectively. This truly unique
mutant exhibiting high cosubstrate flexibility can therefore be
characterized as a ChAT, ChOT, and ChPT enzyme. Also worth
mentioning is hChAT-E337A/C550G. This double mutant ori-
ginally designed to mimic its previously reported counterpart in
the carnitine series, CrAT-D356A/M564G, displaying activity
toward a variety of medium- and long-chain acyl-CoAs (6-18
carbons) (24), was unexpectedly found to be a truly inflexible
ChOT (Figure 5B,C). Indeed, hChAT-E337A/C550G shows
minute activity with all CoA deriviatives tested with the exception
of octanoyl-CoA, where a 23-fold increase in activity is observed
when compared to that of hChAT-wt subjected to the same
reaction conditions. The discrepancy in the cosubstrate specifi-
city profile between the carnitine and choline series also holds
DISCUSSION
Promiscuous Enzymes as Tools for Scaffold Diversifica-
tion. Enzymes with diverse biological activities are found to be
naturally promiscuous. Many studies on expanding and/or
taking advantage of the substrate and cosubstrate promiscuity
of enzymes to generate new drugs have been reported (30-34).
Garneau-Tsodikova and co-workers took advantage of the
natural aminoglycoside substrate and coenzyme A cosubstrate
promiscuity of aminoglycoside acetyltransferases to chemoenzy-
matically generate novel mono-, homodi-, and heterodi-N-
acylated aminoglycosides (35, 36). Similarly, by using the sub-
strate promiscuity of the arylalkyamine N-acetyltransferase
(AANAT), Cole and co-workers developed cell-permeable AA-
NAT acetyltransferase inhibitors (37). By revealing the substrate
promiscuity of 1-deoxy-D-xylulose 5-phosphate (DXP) synthase,
Brammer and Freel Meyers demonstrated the potential of this
enzyme as a biocatalyst for the generation of novel R-hydroxy
ketones (38). Glycosyltransferases, key enzymes that catalyze the
attachment of a sugar moiety to an aglycone during the biosynth-
esis of many natural products, also display considerable flex-
ibility toward the sugar donor and/or the acceptor molecule
(39-42). However, as exemplified by the glycopeptide glycosyl-
transferases GtfC and GtfD, not all glycosyltransferases exhibit
relaxed substrate specificity (43). By means of directed evolution,
Thorson and co-workers expanded the promiscuity of glycosyl-
transferase enzymes (44).
In this study, we have focused on expanding the coenzyme A
cosubstrate promiscuity of the hChAT enzyme by site-directed
mutagenesis to generate choline octanoyltransferase (ChOT) and
choline palmitoyltransferase (ChPT) enzymes capable of trans-
ferring medium and long acyl chains to choline and its analogues,
respectively. We have also aimed to generate a ChAT enzyme
capable of transferring acyl chains to a variety of choline
analogues.
Conversion of hChAT into ChOT and ChPT. Substantial
efforts dedicated to the conversion of CrAT into CrOT (23) and
CPT (24) have identified D356 and M564 [corresponding to E337
and C550 in hChAT (Figure 2)] as important residues for fatty
acyl chain length specificity in rCrAT. We hypothesized that
similar mutations could lead to the conversion of hChAT into
ChOT and ChPT. A close analysis of the hChAT crystal structure
suggested that C551, V333, and V542 could also potentially be
critical in dictating the length of the acyl chain to be enzymatically
transferred to choline and its analogues (Figure 3B). Our investi-
gation began via overexpression of a series of N-terminally His10-
tagged ChAT mutants in soluble form (Figure S2 of the Support-
ing Information). To effectively compare the activity of each
enzyme, all enzymes were first tested by incubation of aminocho-
line with the natural cosubstrate acetyl-CoA (Figure 5A). All
activities are normalized to the rate of the acetyl-CoA, hChAT-
wt, and aminocholine reaction. Aminocholine was used in place
of the natural choline substrate as it was previously shown to be a
better substrate for hChAT-wt (26). Two mutants were found to
give an increase in hChAT activity when reacted with acetyl-CoA
and aminocholine. As predicted, the singly mutated hChAT-

Article
Biochemistry, Vol. 49, No. 29, 2010 6225
F
IGURE 5: Histograms comparing the normalized rates of (A) acetyl-CoA with all hChAT mutants, (B) n-propionyl-CoA, n-butyryl-CoA, and
hexanoyl-CoA, and (C) octanoyl-CoA, decanoyl-CoA, lauroyl-CoA, and palmitoyl-CoA with selected mutants. All rates are normalized to the
rate of hChAT-wt with acetyl-CoA.
true for the single mutants hChAT-C550G and -C550A prepared
as mimics of the CrAT-M564G and M564A enzymes, respec-
tively (23). In contrast to the M564G mutation in CrAT that
results in an increase in activity with medium- and long-chain
acyl-CoAs (4-16 carbons), the corresponding C550G mutation
in hChAT does not lead to its conversion to either ChOT or
ChPT (Figure S3G of the Supporting Information). However,
mutation of this amino acid residue to alanine (C550A) results in
conversion of the enzyme to ChPT with some observable ChOT
activity. Interestingly, in the carnitine series, even though it
displays cosubstrate promiscuity toward medium- and long-
chain acyl-CoAs, CrAT-M564A is found to be less active than
CrAT-M564G.
In this study, in addition to E337 and C550, we have identified
C551 as a key residue involved in dictating hChAT cosubstrate
specificity. The C551A single mutation results in an overall
increase in ChOT and ChPT activity. More specifically, 2-, 3-,
3-, 7-, 20-, 10-, and 5-fold increases in activity are observed during
the transfers of the n-propionyl, n-butyryl, hexanoyl, octanoyl,
decanoyl, lauroyl, and palmitoyl chains, respectively, by hChAT-
C551A (Figure 5B,C). Kinetic characterization of this mutant
also reveals its increased catalytic efficiency with respect to
transferring the n-propionyl and n-butyryl chain to aminocholine
(Table 1). The hChAT-E337A/C551A double mutant actually
displays better ChOT and ChPT activity than hChAT-C551A.
Not shown in Figure 5, but worthy of mention, are hChAT-
E337A/C550A and hChAT-E337Y/C551A (Figure S3B,D of the
Supporting Information). hChAT-E337Y/C551A shows reduced
activity with the shorter CoA derivatives (acetyl-CoA, n-propio-
nyl-CoA, and n-butyryl-CoA) but has an approximately 4-fold
increase in activity with the longer CoA derivatives (Figure S3D).
Enzymatic reactions with hChAT-E337A/C550A result in de-
creased rates with all CoAs except palmitoyl-CoAs, where a 5-fold
increase is observed (Figure S3B). The low ChOT and ChPT
activity of the hChAT-E337A/C550A and -E337Y/C551A mu-
tants is somewhat surprising and indicates that the effects resulting
from each mutation are not additive. Interestingly, the C550A/
C551A mutation gives no significant increase in rate with either
n-propionyl-CoA or n-butyryl-CoA, corroborating the nonaddi-
tivity of the mutations (Figure S3C). Simply put, the additive
effects of single and double mutants cannot easily be predicted.
Further structural studies will be required to fully understand the
rules governing the hChAT cosubstrate selectivity.
Importance of M84, Y436, and Y552 for Choline Binding
to hChAT. Having generated hChAT mutants able to catalyze
the transfer of a variety of acyl chains to aminocholine, we sought
to expand the substrate promiscuity of this enzyme. A close look
at the hChAT structure reveals a large tunnel entrance that
narrows near the choline-binding site. M84, Y436, and Y552 are
located in the narrow part of the hChAT substrate tunnel near

6226 Biochemistry, Vol. 49, No. 29, 2010
Green et al.
the choline quaternary ammonium center (Figure 3A). We
hypothesized that mutation of these residues to alanine could
widen the narrow part of the tunnel and potentially allow the
enzyme to utilize choline analogues with larger quaternary
ammonium centers for the production of novel acetylcholine
analogues. Unfortunately, mutations at these positions mostly
abolish activity toward choline and all choline analogues tested
(Figure S1 of the Supporting Information). The characterization
of these constructs provides experimental support to validate the
role of these three amino acid residues in binding and keeping the
choline substrate in the hChAT active site tunnel.
In summary, we have presented evidence that, as in the
carnitine series (conversion of CrAT to CrOT and CPT), the
cosubstrate promiscuity of hChAT can be expanded from short-
to medium-chain (ChOT) and long-chain (ChPT) acyl-CoAs. In
addition to the previously mentioned E337 and C550 residues, we
have identified a novel amino acid residue, C551, critical to
dictating chain length specificity in hChAT. We have demon-
strated that the presence of at least one of two cysteine residues,
C550 or C551, is necessary for chain length specificity in hChAT.
We have also identified M84, Y436, and Y552 as critical amino
acid residues for binding and locking of choline and its analogues
into the active site tunnel of the hChAT enzyme. In conjunction
with our previously reported characterization of the substrate
promiscuity of hChAT-wt (26), this work sets the stage for the
production of libraries of novel acetylcholine analogues as
potential cholinesterase inhibitors.
7. Jogl, G., Hsiao, Y. S., and Tong, L. (2004) Structure and function of
carnitine acyltransferases. Ann. N.Y. Acad. Sci. 1033, 17–29.
8. Makar, T. K., Cooper, A. J., Tofel-Grehl, B., Thaler, H. T., and Blass,
J. P. (1995) Carnitine, carnitine acetyltransferase, and glutathione in
Alzheimer brain. Neurochem. Res. 20, 705–711.
9. Melegh, B., Seress, L., Bedekovics, T., Kispal, G., Sumegi, B.,
Trombitas, K., and Mehes, K. (1999) Muscle carnitine acetyltransfer-
ase and carnitine deficiency in a case of mitochondrial encephalomyo-
pathy. J. Inherited Metab. Dis. 22, 827–838.
10. DiDonato, S., Rimoldi, M., Moise, A., Bertagnoglio, B., and Uziel,
G. (1979) Fatal ataxic encephalopathy and carnitine acetyltransferase
deficiency: A functional defect of pyruvate oxidation? Neurology 29,
1578–1583.
11. Cai, Y., Cronin, C. N., Engel, A. G., Ohno, K., Hersh, L. B., and
Rodgers, D. W. (2004) Choline acetyltransferase structure reveals
distribution of mutations that cause motor disorders. EMBO J. 23,
2047–2058.
12. Govindasamy, L., Pedersen, B., Lian, W., Kukar, T., Gu, Y., Jin, S.,
Agbandje-McKenna, M., Wu, D., and McKenna, R. (2004) Structur-
al insights and functional implications of choline acetyltransferase.
J. Struct. Biol. 148, 226–235.
13. Kim, A. R., Rylett, R. J., and Shilton, B. H. (2006) Substrate binding
and catalytic mechanism of human choline acetyltransferase. Bio-
chemistry 45, 14621–14631.
14. Wu, D., Govindasamy, L., Lian, W., Gu, Y., Kukar, T., Agbandje-
McKenna, M., and McKenna, R. (2003) Structure of human carnitine
acetyltransferase. Molecular basis for fatty acyl transfer. J. Biol.
Chem. 278, 13159–13165.
15. Jogl, G., and Tong, L. (2003) Crystal structure of carnitine acetyl-
transferase and implications for the catalytic mechanism and fatty
acid transport. Cell 112, 113–122.
16. Jogl, G., Hsiao, Y. S., and Tong, L. (2005) Crystal structure of mouse
carnitine octanoyltransferase and molecular determinants of sub-
strate selectivity. J. Biol. Chem. 280, 738–744.
17. Hsiao, Y. S., Jogl, G., and Tong, L. (2004) Structural and biochemical
studies of the substrate selectivity of carnitine acetyltransferase.
J. Biol. Chem. 279, 31584–31589.
ACKNOWLEDGMENT
18. Hsiao, Y. S., Jogl, G., Esser, V., and Tong, L. (2006) Crystal structure
of rat carnitine palmitoyltransferase II (CPT-II). Biochem. Biophys.
Res. Commun. 346, 974–980.
19. Brown, N. F., Anderson, R. C., Caplan, S. L., Foster, D. W., and
McGarry, J. D. (1994) Catalytically important domains of rat carni-
tine palmitoyltransferase II as determined by site-directed mutagen-
esis and chemical modification. Evidence for a critical histidine
residue. J. Biol. Chem. 269, 19157–19162.
Prof. Brian H. Shilton (University of Western Ontario) is
acknowledged for the generous gift of the hChAT-containing
plasmid. We thank Dr. Tapan Biswas (University of Michigan)
for help in generating Figure 3 as well as for the gift of the Int-
pET19b-pps vector. We thank Dr. Mi Hee Lim for the use of
equipment.
20. Morillas, M., Clotet, J., Rubi, B., Serra, D., Asins, G., Arino, J., and
Hegardt, F. G. (2000) Identification of the two histidine residues
responsible for the inhibition by malonyl-CoA in peroxisomal carni-
tine octanoyltransferase from rat liver. FEBS Lett. 466, 183–186.
21. Morillas, M., Gomez-Puertas, P., Roca, R., Serra, D., Asins, G.,
Valencia, A., and Hegardt, F. G. (2001) Structural model of the
catalytic core of carnitine palmitoyltransferase I and carnitine octa-
noyltransferase (COT): Mutation of CPT I histidine 473 and alanine
381 and COT alanine 238 impairs the catalytic activity. J. Biol. Chem.
276, 45001–45008.
SUPPORTING INFORMATION AVAILABLE
A table of primers used in this study (Table S1), structures
of substrates and cosubstrates tested with the hChAT-M84A,
-Y436A, and -Y552A mutants (Figure S1), gel containing Ni(II)-
NTA-purified proteins (Figure S2), histograms of hChAT mu-
tants comparing the rates of different CoA derivatives (Figure
S3), and representative examples of kinetic curves (Figure S4).
This material is available free of charge via the Internet at http://
pubs.acs.org.
22. Morillas, M., Lopez-Vinas, E., Valencia, A., Serra, D., Gomez-
Puertas, P., Hegardt, F. G., and Asins, G. (2004) Structural model
of carnitine palmitoyltransferase I based on the carnitine acetyltrans-
ferase crystal. Biochem. J. 379, 777–784.
23. Cordente, A. G., Lopez-Vinas, E., Vazquez, M. I., Swiegers, J. H.,
Pretorius, I. S., Gomez-Puertas, P., Hegardt, F. G., Asins, G., and
Serra, D. (2004) Redesign of carnitine acetyltransferase specificity by
protein engineering. J. Biol. Chem. 279, 33899–33908.
REFERENCES
1. Whitehouse, P. J. (1998) The cholinergic deficit in Alzheimer’s disease.
J. Clin. Psychiatry 59 (Suppl. 13), 19–22.
24. Cordente, A. G., Lopez-Vinas, E., Vazquez, M. I., Gomez-Puertas, P.,
Asins, G., Serra, D., and Hegardt, F. G. (2006) Mutagenesis of
specific amino acids converts carnitine acetyltransferase into carnitine
palmitoyltransferase. Biochemistry 45, 6133–6141.
25. Cronin, C. N. (1998) Redesign of choline acetyltransferase specificity
by protein engineering. J. Biol. Chem. 273, 24465–24469.
26. Green, K. D., Fridman, M., and Garneau-Tsodikova, S. (2009)
hChAT: A tool for the chemoenzymatic generation of potential
acetyl/butyrylcholinesterase inhibitors. ChemBioChem 10, 2191–
2194.
2. Oda, Y. (1999) Choline acetyltransferase: The structure, distribution
and pathologic changes in the central nervous system. Pathol. Int. 49,
921–937.
3. Holt, D. J., Bachus, S. E., Hyde, T. M., Wittie, M., Herman, M. M.,
Vangel, M., Saper, C. B., and Kleinman, J. E. (2005) Reduced density
of cholinergic interneurons in the ventral striatum in schizophrenia:
An in situ hybridization study. Biol. Psychiatry 58, 408–416.
4. Massouh, M., Wallman, M. J., Pourcher, E., and Parent, A. (2008)
The fate of the large striatal interneurons expressing calretinin in
Huntington’s disease. Neurosci. Res. 62, 216–224.
5. Mallard, C., Tolcos, M., Leditschke, J., Campbell, P., and Rees, S.
(1999) Reduction in choline acetyltransferase immunoreactivity but
not muscarinic-m2 receptor immunoreactivity in the brainstem of
SIDS infants. J. Neuropathol. Exp. Neurol. 58, 255–264.
6. Jellinger, K. A. (2003) Rett Syndrome: An update. J. Neural Transm.
110, 681–701.
27. Biswas, T., and Tsodikov, O. V. (2008) Hexameric ring structure of
the N-terminal domain of Mycobacterium tuberculosis DnaB helicase.
FEBS J. 275, 3064–3071.
28. Ho, S. N., Hunt, H. D., Horton, R. M., Pullen, J. K., and Pease, L. R.
(1989) Site-directed mutagenesis by overlap extension using the
polymerase chain reaction. Gene 77, 51–59.

Article
Biochemistry, Vol. 49, No. 29, 2010 6227
29. Magalhaes, M. L., and Blanchard, J. S. (2005) The kinetic mechanism
of AAC3-IV aminoglycoside acetyltransferase from Escherichia coli.
Biochemistry 44, 16275–16283.
38. Brammer, L. A., and Meyers, C. F. (2009) Revealing substrate
promiscuity of 1-deoxy-D-xylulose 5-phosphate synthase. Org. Lett.
11, 4748–4751.
30. Fridman, M., Balibar, C. J., Lupoli, T., Kahne, D., Walsh, C. T., and
Garneau-Tsodikova, S. (2007) Chemoenzymatic formation of novel
aminocoumarin antibiotics by the enzymes CouN1 and CouN7.
Biochemistry 46, 8462–8471.
31. Freel Meyers, C. L., Oberthur, M., Heide, L., Kahne, D., and Walsh,
C. T. (2004) Assembly of dimeric variants of coumermycins by
tandem action of the four biosynthetic enzymes CouL, CouM, CouP,
and NovN. Biochemistry 43, 15022–15036.
39. Harle, J., and Bechthold, A. (2009) Chapter 12. The power of
glycosyltransferases to generate bioactive natural compounds. Meth-
ods Enzymol. 458, 309–333.
40. Williams, G. J., Goff, R. D., Zhang, C., and Thorson, J. S. (2008)
Optimizing glycosyltransferase specificity via “hot spot” saturation
mutagenesis presents a catalyst for novobiocin glycorandomization.
Chem. Biol. 15, 393–401.
41. Zhang, C., Albermann, C., Fu, X., Peters, N. R., Chisholm, J. D.,
Zhang, G., Gilbert, E. J., Wang, P. G., Van Vranken, D. L., and
Thorson, J. S. (2006) RebG- and RebM-catalyzed indolocarbazole
diversification. ChemBioChem 7, 795–804.
32. Steffan, N., and Li, S. M. (2009) Increasing structure diversity of
prenylated diketopiperazine derivatives by using a 4-dimethylallyl-
tryptophan synthase. Arch. Microbiol. 191, 461–466.
33. Liu, X., and Walsh, C. T. (2009) Characterization of cyclo-acetoace-
tyl-L-tryptophan dimethylallyltransferase in cyclopiazonic acid bio-
42. Faridmoayer, A., Fentabil, M. A., Haurat, M. F., Yi, W., Woodward,
R., Wang, P. G., and Feldman, M. F. (2008) Extreme substrate
promiscuity of the Neisseria oligosaccharyl transferase involved in
protein O-glycosylation. J. Biol. Chem. 283, 34596–34604.
43. Oberthur, M., Leimkuhler, C., Kruger, R. G., Lu, W., Walsh, C. T.,
and Kahne, D. (2005) A systematic investigation of the synthetic
utility of glycopeptide glycosyltransferases. J. Am. Chem. Soc. 127,
10747–10752.
synthesis: Substrate promiscuity and site directed mutagenesis studies.
Biochemistry 48, 11032–11044.
34. Monsan, P., Remaud-Simeon, M., and Andre, I. (2010) Transgluco-
sidases as efficient tools for oligosaccharide and glucoconjugate
synthesis. Curr. Opin. Microbiol. 13, 293–300.
35. Green, K. D., Chen, W., Houghton, J. L., Fridman, M., and Garneau-
Tsodikova, S. (2010) Exploring the substrate promiscuity of drug-
modifying enzymes for the chemoenzymatic generation of N-acylated
aminoglycosides. ChemBioChem 11, 119–126.
44. Williams, G. J., Zhang, C., and Thorson, J. S. (2007) Expanding the
promiscuity of a natural-product glycosyltransferase by directed
evolution. Nat. Chem. Biol. 3, 657–662.
36. Houghton, J. L., Green, K. D., Chen, W., and Garneau-Tsodikova, S.
(2010) The future of aminoglycosides: The end or renaissance?
ChemBioChem 11, 880–902.
37. Zheng, W., Scheibner, K. A., Ho, A. K., and Cole, P. A. (2001)
Mechanistic studies on the alkyltransferase activity of serotonin
N-acetyltransferase. Chem. Biol. 8, 379–389.
45. Baker, N. A., Sept, D., Joseph, S., Holst, M. J., and McCammon,
J. A. (2001) Electrostatics of nanosystems: Application to microtu-
bules and the ribosome. Proc. Natl. Acad. Sci. U.S.A. 98, 10037–
10041.
46. DeLano, W. L. (2008) The PyMOL Molecular Graphics System,
DeLano Scientific LLC, San Carlos, CA.