Fluorinated chloramphenicol acetyltransferase thermostability and activity profile: Improved thermostability by a single-isoleucine mutant

Article abstract of DOI:10.1016/j.bmcl.2007.07.107

A lysate-based thermostability and activity profile is described for chloramphenicol acetyltransferase (CAT) expressed in trifluoroleucine, T (CAT T). CAT and 13 single-isoleucine CAT mutants were expressed in medium supplemented with T and assayed for thermostability on cell lysates. Although fluorinated mutants, L82I T and L208I T, showed losses in thermostability, the L158I T fluorinated mutant demonstrated an enhanced thermostability relative to CAT T. Further characterization of L158I T suggested that T at position 158 contributed to a portion of the observed loss in thermostability upon global fluorination.

Full text of DOI:10.1016/j.bmcl.2007.07.107

Bioorganic & Medicinal Chemistry Letters 17 (2007) 5907–5911
Fluorinated chloramphenicol acetyltransferase thermostability
and activity profile: Improved thermostability
by a single-isoleucine mutant
a
a
a
Natalya Voloshchuk, Man Xia Lee, Wan Wen Zhu,
b
a,c,
*
Ismet Caglar Tanrikulu and Jin Kim Montclare
a
Department of Chemical and Biological Sciences, Polytechnic University, 6 Metrotech Center, Brooklyn, NY 11201, USA
b
Division of Chemistry and Chemical Engineering, California Institute of Technology, 1200 E. California Boulevard,
Pasadena, CA 91125, USA
c
Department of Biochemistry, SUNY Downstate Medical Center, 450 Clarkson Avenue, Brooklyn, NY 11203, USA
Received 9 June 2007; revised 25 July 2007; accepted 26 July 2007
Available online 29 August 2007
Abstract—A lysate-based thermostability and activity profile is described for chloramphenicol acetyltransferase (CAT) expressed in
trifluoroleucine, T (CAT T). CAT and 13 single-isoleucine CAT mutants were expressed in medium supplemented with T and
assayed for thermostability on cell lysates. Although fluorinated mutants, L82I T and L208I T, showed losses in thermostability,
the L158I T fluorinated mutant demonstrated an enhanced thermostability relative to CAT T. Further characterization of L158I
T suggested that T at position 158 contributed to a portion of the observed loss in thermostability upon global fluorination.
Published by Elsevier Ltd.
Non-natural amino acids have been used to create
defined mutations that are unique from the set of amino
acid monomers in naturally occurring proteins. Intro-
duction of a tailored side chain at one or more residues
has been used to explore the effects of particular hydro-
mutagenesis with residue-specific incorporation in which
each T was mutated into isoleucine. CAT and 13 single-
isoleucine CAT mutants were expressed in medium sup-
plemented with T and assayed for thermostability on
cell lysates. This was done to prevent the introduction
of T at specified positions. Here, we report the thermo-
stability profile of wild-type CAT and 13 single-isoleu-
cine CAT mutants expressed in the presence of T (and
leucine, L, as a control) by utilizing a microplate-based
lysate activity screen.
1
3
1
gen bonds on protein folding, to stabilize protein con-
2
formations, and to probe the catalytic mechanisms of
3
enzymes. Residue-specific incorporation of non-natural
amino acids allows for the global replacement of a nat-
4
,5
ural amino acid in vivo. This form of multi-site substi-
tution can be used to alter the overall properties of
6
,7
proteins. Substitution of leucine with trifluoroleucine
T) or hexafluoroleucine (H) in a coiled-coil protein
revealed an enhanced stability toward thermal and
The development of a robust screen that can be easily
monitored is critical for the identification of mutants
(
1
4
bearing enhanced or altered properties. We measured
the activity of Escherichia coli cell lysates from a 5 mL
growth at 412 nm in the presence of chloramphenicol,
8
–11
chemical denaturation.
cific incorporation of T into chloramphenicol acetyl-
By contrast, the residue-spe-
1
2
0
transferase (CAT) resulted in a loss of thermostability.
acetyl coenzyme A, and 5,5 -dithiobis(2-nitrobenzoic
acid) (DTNB) to measure the activity of crude protein
(Fig. 1a). For cells bearing CAT (positive control), the
lysates exhibited an increase in intensity at 412 nm,
which indicated catalytic activity. Cell lysates in the
absence of CAT (negative control) did not demonstrate
any increase in intensity above background.
To deconvolute those fluorinated residues that contrib-
ute to the observed decrease in thermostability of the
fluorinated CAT (CAT T), we merged site-directed
Keywords: Non-natural amino acid; Trifluoroleucine; Mutagenesis;
Thermostability; Protein engineering.
*
Corresponding author. Tel.: +1 718 260 3679; fax: +1 718 260
676; e-mail: jmontcla@poly.edu
Single colonies bearing the CAT gene were grown and
expressed in microtiter plates to adapt the assay for
3
0960-894X/$ - see front matter Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2007.07.107

5908
N. Voloshchuk et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5907–5911
mutation would preserve the side-chain hydrophobicity
1
8
and surface area, creating minimal perturbation to
the structure and function of CAT T. By making such
a homologous substitution, we were able to probe the
effect of each T within the context of the fluorinated
CAT T. Although a single substitution approach would
not rule out the synergistic effects of multiple T interac-
tions, it would provide information as to which of the
fluorinated residues is critical to its function.
In a 96-well microtiter plate, six replicates of each mu-
tant were grown and expressed in 19 amino acids supple-
mented with T and with L as a control. The lysate
activity after exposure to elevated temperatures or RA
of CAT T and single-isoleucine mutants was measured
(
Fig. 2a). The lysate activity for each enzyme was also
assessed under ambient temperatures to determine
whether the differences in thermostability were due to
losses or improvements of initial activity (Fig. 2b). Ther-
mostability and activity of mutants expressed in medium
supplemented with T (fluorinated conditions) or with L
(conventional conditions) were normalized to CAT T or
CAT L, respectively. Ten of the 13 mutants showed sim-
ilar thermostabilities and initial activities when directly
compared to fluorinated or conventional CAT protein
Figure 1. (a) Activity assay on cell lysates. In the absence of the CAT
gene there is no observed color change above background, indicating
no activity, while in the presence of the CAT gene there is a distinct
color change upon formation of 5-thio-2-nitrobenzoic acid (TNB) that
can be monitored at 412 nm. Validation of CAT cell lysates grown in
the presence of (b) 20 natural amino acids and (c) 19 amino acids
supplemented with T.
(
Fig. 2a). Most of the T and L residues were permissive
to an isoleucine substitution, indicating that each of
these T or L residues does not contribute to thermosta-
bility or activity in either the fluorinated or conventional
contexts.
96-well format. The variability of the activities of a single
sequence CAT library was assessed to determine the
1
4
reproducibility of the screen. The coefficient of variance
for the 96-well plate bearing CAT expressed in medium
supplemented with the natural set of 20 amino acids
Two fluorinated mutants, L82I T and L208I T, demon-
strated a loss in thermostability relative to CAT T, sug-
gesting that a mutation into isoleucine at those positions
destabilized the enzyme (Fig. 2a). The activities at room
temperature of both mutants were compared to CAT T
to investigate whether the observed reduction in thermo-
stability was due to loss in initial activity. Fluorinated
mutant L82I T revealed a similar initial activity relative
to CAT T (Fig. 2b). In the case of L208I T, substitution
with isoleucine negatively affected the initial activity
with respect to CAT T, indicating that isoleucine at
position 208 possesses a deleterious effect on enzyme
function (Fig. 2b). The same trend was observed under
conventional 20 amino acid conditions.
(
CAT L) was calculated as 19.3% (Fig. 1b). Another single
sequence CAT library was expressed in medium supple-
mented with T (CAT T). A medium shift was performed
on the 96-well plate, where the wells were washed twice
with 0.9% NaCl before the addition of the new medium
bearing T to enhance the incorporation of the non-natural
amino acid. The coefficient of variance was 23.2% in the
presence of T (Fig. 1c). To evaluate the extent of T incor-
poration in the wells, all the lysates were pooled, purified,
and determined to possess 63% substitution by amino acid
analysis. Prior work had demonstrated that the
incorporation level leads to different extent of stability;
low levels such as 17% or 37% T substitution exhibited dis-
Remarkably, fluorinated mutant L158I T revealed a 1.8-
fold increase in thermostability relative to CAT T
(Fig. 2a). The same mutation expressed in medium bear-
ing 20 amino acids, L158I L, exhibited similar thermo-
stability relative to CAT L. Thus, the isoleucine
mutant only exhibited an effect in the context of the fluo-
rinated protein. The observed improvement in thermo-
stability of L158I T was not a result of initial activity
changes; L158I in conventional and fluorinated contexts
exhibited comparable initial activities to their wild-type
counterparts.
1
2
tinct half-lives at 60 °C. Although this was lower than
the amount of incorporation obtained for large-scale
1
2
expression previously and below, comparisons were
made under the same degree of substitution.
To identify fluorinated mutants that differ in thermosta-
bility (enhanced or diminished) from CAT T, the screen
must be sensitive enough to observe small changes in
thermostability that result from a single amino acid sub-
1
5
stitution. Here, thermostability is a measure of the
residual activity (RA) after incubation at elevated tem-
peratures normalized to the activity at ambient temper-
To quantify the enhancement in thermostability, mutant
L158I was expressed in medium supplemented with T or
L, purified, and characterized. The activity and thermo-
stability of the mutants were compared to CAT T and
CAT L, respectively. For L158I T and CAT T, mass
1
6,17
atures.
Thirteen single-leucine to isoleucine mutants
of CAT were generated to investigate the contribution
of each T toward the thermostability of CAT T as de-
fined above. We reasoned that T (or L) to isoleucine

N. Voloshchuk et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5907–5911
5909
Figure 2. (a) Plot of the RA of CAT and isoleucine mutants. (b) Plot of the activity of CAT and isoleucine mutants at room temperature. Gray bars
signify enzymes expressed in 20 amino acids while black bars signify enzymes expressed in 19 amino acids supplemented with T. Asterisks highlight
thermostability or activity differences that are statistically different from wild-type CAT. Data represent an average of six trials. Error bars denote
standard deviation.
spectrometry and amino acid analysis demonstrated an
8
2% and 83% T substitution, respectively (Supporting
information). Quantitative levels of incorporation were
not achieved as a result of the trace amounts of natural
1
2
leucine liberated by degradation of cellular proteins.
The 82–83% T incorporation was sufficiently high to
avoid the possibility for the synthesis of wild-type pro-
tein that could account for the improved thermostability
observed. Furthermore, since the comparison of the
fluorinated proteins is made on proteins with nearly
identical levels of T incorporation, the enhancements
observed can be attributed to the protein. The specific
activities of CAT and L158I reveal that the catalytic
efficiencies were essentially unchanged under both fluo-
rinated and conventional conditions (Table 1). A 2.4-
Figure 3. Plots of RA of purified proteins. CAT T is black, L158I T is
red, CAT L is gray and L158I L is pink. Data represent an average of
three trials. Error bars denote standard deviation.
fold increase in K was observed for L158I mutant when
m
compared to the CAT bearing T or L. Since the increase
in Km was accompanied by a concomitant increase in
kcat, the overall values of kcat/Km for the conventional
and fluorinated proteins were similar, indicating that
activity was not altered drastically by the isoleucine sub-
stitution and upon fluorination.
increase in RA relative to CAT T (Table 1 and Fig. 3,
Supporting information). In the context of the 20 natu-
ral amino acids, L158I L demonstrated a slight decrease
in RA relative to CAT L (Table 1 and Fig. 3, Supporting
information).
The thermostability data on the purified proteins as
1
7
measured by RA were highly consistent when com-
pared to the data obtained from the lysate assay. Under
fluorinated conditions, L158I T demonstrated a 1.5-fold
By modeling the crystal structure of CAT-I with that of
1
9,20
the CAT-III chloramphenicol
complex, we can iden-
tify the isoleucine mutants and their positions relative to
chloramphenicol (Fig. 4). First, we consider the two mu-
tants that resulted in reduced thermostability: L82I T
and L208I T. Structural analysis reveals that both muta-
tions point away from the chloramphenicol-binding
pocket and interact with other hydrophobic residues.
L82 is positioned on sheet D and is stabilized by a set
of hydrophobic residues presented by V11, W16, M75,
and M77. L208 is located on helix 5 and is stabilized
by all the interior leucines, L39, L63, L66, L187, and
L205 or in the case of the fluorinated enzyme, interior
TFLs (Fig. 4). The interactions of these residues may
be significant in maintaining the proper packing of the
CAT structure critical for catalysis. Previous work
Table 1. Kinetic parameters and thermostability
4
ꢀ1
ꢀ1
a
RA (%)
Protein
K
m
(lM)
kcat/K
m
(10 lM min
)
CAT L
CAT T
L158I L
L158I T
13.3 ± 2.0
11.8 ± 0.7
32.2 ± 2.9
28.3 ± 4.9
6.0 ± 0.9
5.5 ± 0.3
4.5 ± 0.4
5.0 ± 0.9
94.9
34.6
89.0
51.7
Kinetic measurements are the average of three experiments (±standard
deviation).
a
Residual activity at 55 °C. Values of RAs are calculated from an
average of three experiments (standard deviation <5%).

5910
N. Voloshchuk et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5907–5911
,8–11
2
23,24
of the analog,
with some recent exceptions.
By
using this approach, we can now gain insight into the
effects of the non-natural amino acid on the protein
function and pinpoint which non-natural residues are
critical to function. Because the screen developed here
provides a functional profile of each non-natural amino
acid side chain, it can be used to improve the under-
standing of non-natural proteins even when the struc-
tures of such proteins are not yet determined.
Acknowledgments
Figure 4. Model of E. coli CAT-I trimer bound to chloramphenicol
red). Two of the mutations that result in thermostability loss, L82I
We thank the Othmer Institute, Weschler Award, NIH
(GM 62523 and 5F32 GM67375-2), ACS-Fluorine
Chemistry Division (MXL), and the Beckman Institute
at the California Institute of Technology for financial
support. We are grateful to Professors David Tirrell,
Frances Arnold, Kent Kirshenbaum, and Paramjit
Arora for use of equipment and reagents for this work.
(
and L208I, are depicted in yellow and red, respectively. The mutation
that leads to an enhancement in thermostability, L158I, is represented
in gray. Residues highlighted in blue are the leucine/TFL residues of
CAT that when mutated into isoleucine does not significantly alter the
thermostability or activity.
points to the importance of the C-terminal helix for
2
1
CAT activity. Deletion of 9 residues within helix 5 of
CAT-I (CATD9) resulted in an inactivate enzyme due
Supplementary data
2
1,22
to misfolding or structural destabilization.
Although
Supplementary data associated with this article can
be found, in the online version, at doi:10.1016/j.bmcl.
2007.07.107.
isoleucine presented similar hydrophobic interactions
with leucine or TFL, the difference in the configuration
of the side chains could create destabilizing contacts,
leading to the observed decrease in thermostability.
References and notes
Next, we consider L158I T, which resulted in an enhanced
themostability. According to our model, L158 is posi-
1. Ellman, J. A.; Mendel, D.; Schultz, P. G. Science 1992,
255, 197.
˚
tioned approximately 4 A from the substrate (Fig. 4).
2
. Hodges, J. A.; Raines, R. T. J. Am. Chem. Soc. 2003, 125,
262.
The leucine is located on the outer region on sheet H
and is stabilized by the predominantly hydrophobic inter-
actions with N34, V160, F166, and V170 (Fig. 4). Substi-
tution of TFL at 158 with isoleucine improves the
thermostability of the fluorinated enzyme, suggesting that
the TFL contributes negatively to thermostability.
Because it is located on the outside and faces away from
the core, the presence of TFL at this position may result
9
3
. Yee, C. S.; Chang, M. C. Y.; Ge, J.; Nocera, D. G.;
Stubbe, J. J. Am. Chem. Soc. 2003, 125, 10506.
4
5
. Kiick, K. L.; Tirrell, D. A. Tetrahedron 2000, 56, 9487.
. Budisa, N.; Minks, C.; Alefelder, S.; Wenger, W.; Dong,
F. M.; Moroder, L.; Huber, R. FASEB J. 1999, 13, 41.
. Yoder, N. C.; Kumar, K. Chem. Soc. Rev. 2002, 31, 335.
6
7. Budisa, N. Chembiochem 2004, 5, 1176.
8. Tang, Y.; Tirrell, D. A. J. Am. Chem. Soc. 2001, 123,
11089.
1
2
in misfolding, leading to instability. Here, isoleucine
provides a compensatory mutation that allows for the
recovery of a portion of the thermostability that is lost
upon global fluorination.
9
. Tang, Y.; Ghirlanda, G.; Petka, W. A.; Nakajima, T.;
DeGrado, W. F.; Tirrell, D. A. Angew. Chem., Int. Ed.
Engl. 2001, 40, 1494.
1
1
1
0. Bilgicer, B.; Xing, X.; Kumar, K. J. Am. Chem. Soc. 2001,
23, 11815.
1. Bilgicer, B.; Kumar, K. Proc. Natl. Acad. Sci. U.S.A.
004, 101, 15324.
2. Panchenko, T.; Zhu, W. W.; Montclare, J. K. Biotech.
Bioeng. 2006, 94, 921.
These results confirm that a mutation under two compo-
sitionally different proteins leads to different outcomes.
Moreover, this demonstrates that a single isoleucine
mutation can improve the properties of the fluorinated
protein. Since the isoleucine mutation in the context of
fluorination led to an enhanced thermostability, the data
suggest that the observed loss in thermostability of CAT T
was due in part by T at position 158. The T residue at 158
appears to be critical to thermostability and a simple sub-
stitution to isoleucine can improve the thermotolerance of
the fluorinated protein, possibly by the prevention of mis-
1
2
13. Single-isoleucine mutants of the CAT gene were generated
via QuikChange (Stratagene: La Jolla, CA) mutagenesis of
0
pCCCAT using the following primers: L39I 5 -cctataacca
0
0
gaccgttcagatcgatattacggcc-3 ; L45I 5 -gctggatattacggccttt
0
0
atcaagaccgtaaag-3 ; L63I 5 -ccggcctttattcacattatcgcccgc
0
0
0
ctgatg-3 ; L66I 5 -cacaatcttgcccgcatcatgaatgctcatccgg-3 ;
0
0
L82I 5 -ggcaatgaaagacggtgagatcgtgatatgggatag-3 ; L105I
0
1
2
folding. In order to truly understand the mode by which
this substitution is improving the thermostability, three-
dimensional structural studies are needed.
0
0
5
ttccggcagtttatccacatatattcgcaag-3 ; L131I 5 -ggcgtgttacggt
-ctgaaacgttttcatcgatctggagtgaataccacg-3 ; L117I 5 -cgat
0
0
0
0
gaaaacatcgcctatttccc-3 ; L158I 5 -caccagttttgatatcaacgtgg
0
0
ccaatatggac-3 ; L184I 5 -cgcaaggcgacaaggtgatcatgccgctg
0 0
Previously, functional proteins bearing non-natural ami-
no acids have been limited to the appropriate placement
gcgattc-3 ; L187I 5 -ggcgacaaggtgctgatgccgatcgcgattcagg-
0
0
0
3 ; L205I 5 -ccatgtcggcagaatgatcaatgaattacaacag-3 ; L208I

N. Voloshchuk et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5907–5911
5911
0
0
5
-ggcagaatgcttatgaaatccaacagtactgcgatg-3
and their
17. Turner, S. L.; Ford, G. C.; Mountain, A.; Moir, A. Prot.
Eng. 1992, 5, 535.
18. Roseman, M. A. J. Mol. Biol. 1988, 201, 621.
19. Leslie, A. G. W.; Liddell, J. M.; Shaw, W. V. J. Mol. Biol.
1986, 188, 283.
20. Leslie, A. G. W. J. Mol. Biol. 1990, 213, 167.
21. Robben, J.; Vanderschueren, J.; Volckaert, G. J. Biol.
Chem. 1993, 268, 24555.
22. VanderSchueren, J.; Robben, J.; Goossens, K.; Heremans,
K.; Volckaert, G. J. Mol. Biol. 1996, 256, 87820.
23. Kajihara, D.; Hohsaka, T.; Sisido, M. Prot. Eng. Des. Sel.
2005, 18, 273.
complements. The integrity of each clone was confirmed
by DNA sequencing.
4. Salazar, O.; Sun, L.. In Methods in Molecular Biology:
Directed Enzyme Evolution—Screening and Selection
Methods; Arnold, F. H., Georgiou, G., Eds.; Humana
Press: Totowa, NJ, 2003; Vol. 230, pp 85–97.
1
1
1
5. Zhao, H. M.; Arnold, F. H. Prot. Eng. 1999, 12, 47.
6. Cell lysates were heated in 96-well format at 60 °C for
3
0 min and cooled to 4 °C using a thermocycler (PTC-200,
MJ Research). To determine outliers of activity and
thermostability, a one-tailed, two sample t-test that
assumed equal variance was applied with a level of
significance of 0.005.
24. Montclare, J. K.; Tirrell, D. A. Angew. Chem., Int. Ed.
2006, 45, 4518.