www.els-journal.com
Hernan Biava
Nediljko Budisa
Department of Biocatalysis,
Institute of Chemistry, Berlin
Institute of Technology/TU
Berlin, Berlin, Germany
Review
Evolution of fluorinated enzymes: An emerging
trend for biocatalyst stabilization
Nature uses remarkably limited sets of chemistries in its repertoire, especially when
compared to synthetic organic chemistry. This limits both the chemical and struc-
tural diversity that can ultimately be achieved with biocatalysis, unless the powers
of chemical synthesis are merged with biological systems by integrating nonnatural
synthetic chemistries into the protoplasma of living cells. Of particular interest, here
is the fluorous effect that has recently established the potential to generate enzymes
with an increased resistance toward both high temperature and organic solvents. For
these reasons, we are witnessing a rapid development of efficient methodologies for
the incorporation of fluorinated amino acids in protein synthesis, using both in vivo
and in vitro strategies. In this review, we highlight relevant and trendsetting results
in the design and engineering of stable fluorinated proteins and peptides along with
whole-cell biocatalysis as an economically attractive and convenient application with
exclusive focus on industrial biocatalysis. Finally, we envision new strategies to im-
prove current achievements and enable the field to progress far beyond the current
state-of-the-art.
Keywords: Biocatalysis / Enzyme stabilization / Fluorinated amino acids / Fluorination /
Fluorous effect
Received: July 20, 2013; revised: October 1, 2013; accepted: October 27, 2013
DOI: 10.1002/elsc.201300049
1 Introduction
Enzymes have been used for centuries in technological processes,
such as the production of alcohol by fermentation, or enzymatic
degradation of milk, but their relevance as biocatalysts in indus-
trial processes has increased dramatically in recent decades [1].
Enzymes, whether in their isolated form or as part of a whole cell
fermentation process, have proven to be a sustainable alternative
to traditional chemical catalysis for the effective conversion of
natural accessible starting materials to high value products, such
as biofuels, pharmaceuticals, agrochemicals, etc. Several factors,
such as unrivaled chemo-, enantio- and regioselectivity, ability
to function under mild conditions, biodegradability, and recy-
Correspondence: Prof. Nediljko Budisa ([email protected]
berlin.de), Department of Biocatalysis, Institute of Chemistry, Berlin
Institute of Technology/TU Berlin, M¨
uller-Breslau-Straße 10, D-
10623 Berlin, Germany
Abbreviations: AARS, aminoacyl-tRNA synthetases; APS,automated
peptide synthesis; CalB, lipase B from Candida Antarctica;FPhe,fluo-
rophenylalanine; FPro,fluoroproline;hFLeu, hexafluoroleucine; hFVal,
hexafluorovaline; SPI, selective pressure incorporation; tFIle,trifluo-
roisoleucine; tFLeu,trifluoroleucine;tFVal,trifluorovaline
clability are among the main reasons that have prompted the use
of enzymes in industrial biocatalysis [2, 3].
All these particular properties displayed by enzymes are a
consequence of their complex 3-D structure. In the native state,
the polypeptide chain adopts such a conformation that the hy-
drophobic residues locate themselves in the protein core, mini-
mizing contact with the polar solvent, while hydrophilic residues
are located on the surface. Several other interactions between
amino acid residues, such as hydrogen bonds, van der Waals’s
forces, salt bridges, etc., contribute to stabilize the final folded
state.
These interactions, which are ultimately responsible for the
high enzymatic reactivity and specificity, are however quite sen-
sitive to environmental conditions. Any change in the physi-
cal parameters of the aqueous media in which the enzyme has
been synthesized and folded can alter its native state and con-
sequently, enzymatic activity can be lost or seriously damaged.
This constitutes a clear disadvantage for many biotechnolog-
ical applications, where organic solvents or high temperature
are usually required for the solubilization of highly hydrophobic
substrates [4].
Nevertheless, exhaustive research in different fields of en-
zyme technology has already been conducted in an attempt to
overcome these present limitations. Several strategies include
340 C2013 The Authors. Engineering in Life Sciences published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and
distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
www.els-journal.com Eng. Life Sci. 2014,14, 340–351 www.biotecvisions.com
Figure 1. (A) Typical phase sep-
aration behavior for highly fluori-
nated compounds. (B) Effect of the
organic solvent addition (blue) to a
folded protein. Hydrophobic inter-
actions with the organic solvent in-
duce immediate disruption of the
protein native state. (C) Substitu-
tion by a fluorous core (pink) pre-
vents denaturation caused by or-
ganic solvent or high temperature.
immobilization to solid supports, chemical modification, crys-
tallization, aggregation, protein engineering techniques, and
site-directed mutagenesis, which have been widely reviewed
[5,6].
Recent attempts show that enzymatic stabilization could be
obtained by introducing fluorinated analogs of hydrophobic
amino acids into protein structures in order to imprint these
biological biopolymers with some of the properties that charac-
terize fluorinated compounds, such as high temperature resis-
tance, inertness, and extreme hydrophobicity [7–9].
In this review, we aim to present the basis behind this prospect
and some of the advances made so far in the development of flu-
orinated enzymes and peptides that are relevant for industrial
biocatalysis. In addition, possible new strategies to improve ex-
isting results will also be discussed. For an exhaustive summary
of all other possible applications of fluorinated peptides and pro-
teins beyond enzymatic stabilization, the reader is kindly invited
to inspect other reviews published elsewhere [10–12].
2 The fluorous effect
Despite the scarcity of naturally occurring organofluorine com-
pounds, chemists have been interested in their synthesis for a
long time, mostly due to their unusual physicochemical prop-
erties [13]. Extensively fluorinated molecules exhibit resistance
to high temperature and an unusual phase separation behav-
ior: they are simultaneously hydrophobic and lipophobic, and
they preferentially interact with fluorocarbon compounds. This
phenomenon is usually referred to as the fluorous effect [7–10].
The nature of the fluorous effect has remained unclear for
several years, but it has been recently explained in terms of the
small size and low polarizability of the fluorine atom, which
causes fluorinated compounds to exhibit a large apolar sur-
face area, hindering effective packing with other nonfluorinated
molecules [14–16]. These characteristics combined with the
chemical inertness of fluorinated compounds have been suc-
cessfully exploited to develop unique materials, such as nonstick
coatings (Teflon) [10].
When compared with those interactions operating during
protein folding in aqueous solution, the fluorous effect shows
noticeable similarities. Both phenomena are driven by the ten-
dency of a certain group of hydrophobic residues to interact
with cognates rather than the surrounding polar media. The
only intrinsic difference is that the fluorinated molecules are
still lipophobic, so the fluorous effect would manifest even in
organic solvents and no perturbation of the fluorinated phase is
expected.
This led to the idea that, if the hydrocarbon core of an enzyme
would be replaced by a fluorinated analog, the resulting fluori-
nated enzymatic variant should still fold into a structure similar
to that of the natural protein and consequently, enzymatic ac-
tivity should be retained. Interestingly, the fluorous effect also
predicts that fluorinated proteins should be resistant to denatu-
ration by addition of standard organic solvents, since the fluo-
rocarbon side chains are also lipophobic [7]. At the same time, a
greater resistance to thermal denaturation is expected according
to the properties of highly fluorinated compounds (Fig. 1).
This initiative caught the attention of several research groups,
due to the fact that the expected properties imprinted by a fluori-
nated core in enzymes would defeat two of their major drawbacks
for biotechnological processes, that is, organic solvent and tem-
perature susceptibilities as well as finding new utilities in var-
ious protein-based biotechnologies, such as new biomaterials,
biosensors, and protein therapeutics.
3 Bio-incorporation of fluorinated amino
acids into proteins and peptides
The substitution of amino acids typically present in the hy-
drophobic protein core by their fluorinated surrogates was ini-
tially conceived, as an approach to preserve biological activity,
since those residues are usually not directly involved in the cat-
alytic mechanism but are mostly responsible for maintaining
the native structural state [9]. Leucine, isoleucine, valine, glycine
and, to less extent, tryptophan, phenylalanine, methionine and
C2013 The Authors. Engineering in Life Sciences published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 341
www.els-journal.com Eng. Life Sci. 2014,14, 340–351 www.biotecvisions.com
H2N COOH
CF3
Trifluoroethylglycine
tFeGly
H2N COOH
(n-Fluoro)phenylalanine
(n-F)Phe
n=1-4
Fn
H2N COOH
CF3
*
Trifluoroleucine
tFLeu
H2N COOH
CF3
CF3
Hexafluoroleucine
hFLeu
H2N COOH
Trifluoroisoleucine
tFIle
CF3
H2N COOH
CF3
*
Trifluorovaline
tFVal
H2N COOH
CF3
F3C
Hexafluorovaline
hFVal
H2N COOH
Trifluoromethionine
tFMet
SCF3
H2N COOH
4-trifluoromethylphenylalanine
(4-CF3)Phe
CF3
H2N COOH
(n-Fluoro)tryptophan
(n-F)Trp
n=1-4
NH
Fn
N
H
COOH
F
(4R)Fluoro-Proline
(4R-F)Pro
N
H
COOH
F
(4S)Fluoro-Proline
(4S-F)Pro
Figure 2. Fluorinated analogs of amino acids that have been incorporated into proteins and peptides. Abbreviations correspond to those
used in this review.
proline fulfill this requisite. Various fluorinated analogs of these
amino acids, varying the number and position of the fluorine
atoms, were incorporated into peptides and proteins (Fig. 2).
Some of these are commercially available, while others can be
synthesized in a relatively straightforward manner, as reported
elsewhere [17,18].
The incorporation of fluorinated amino acids into proteins
was at first a big challenge. The earliest successful reports were
accomplished in vivo by employing strains of bacteria that were
auxotrophic for the parental amino acid that is intended to
be replaced by its fluorinated analog. Under the stress of this
metabolic requirement to survive, and in presence of the flu-
orinated analog, some bacterial strains could incorporate it
into their proteins to a certain extent [19, 20]. This method,
which is usually called selective pressure incorporation (SPI), has
been successfully employed to incorporate trifluoromethionine,
4-fluoroproline, trifluoroleucine (tFLeu), and different regioi-
somers of fluorophenylalanine (FPhe) into proteins expressed
mostly in Escherichia coli, even though another host microor-
ganisms such as Saccharomyces cerevisiae or Pichia pastoris have
been used [21].
The main disadvantage of this methodology is that fluori-
nation is never 100% effective because of the presence of the
natural amino acid derived from cellular metabolism and the
requirement for the fluorinated analog to be recognized by
the aminoacyl-tRNA synthetases (AARS). These AARS have cer-
tain substrate tolerances and some analogs for their natural sub-
strates can be activated and charged, although to a much lesser
extent [22, 23]. For this reason, the intracellular concentration
of the fluorinated analog must be much higher than that of the
natural counterpart, which can be highly detrimental for the
host cell growth [24, 25].
This also implies that there is no control over which amino
acid positions will be substituted and supposedly all intrinsic
occurring amino acid locations will be replaced. In practice this
does not happen though due to the high toxicity of the fluo-
rinated analog, which prevents its presence at an early stage of
the cell culturing process and accordingly, bacterial strains must
be first proliferated with the full suite of natural amino acids.
When an acceptable cell density is reached (normally close to
the stationary phase), the cells are collected and transferred to
new media where the natural amino acid has been completely
depleted and replaced by the fluorinated analog (Fig. 3). Due to
the remaining intracellular concentration of the parental amino
acid, substitution efficiencies of 70–90% are typically found.
Alternatively, fluorinated amino acids can be incorporated in
vitro by automated peptide synthesis (APS) [26, 27]. This ap-
proach makes use of a solid support and appropriately protected
derivates of amino acids in order to synthesize a polypeptide
chain by successive cycles of activation and deprotection, chem-
ically mediated by suitable activators and deprotecting agents.
Despite the fact that precise control over polypeptide sequence
and substitution sites can be achieved, this technique is unfor-
tunately limited to polypeptide chains up to approximately 50
amino acids and results too expensive for generating a relatively
large amount of biocatalyst.
342 C2013 The Authors. Engineering in Life Sciences published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.els-journal.com Eng. Life Sci. 2014,14, 340–351 www.biotecvisions.com
Figure 3. Scheme for the se-
lective pressure incorporation
(SPI) method. Strong host aux-
otrophism and control of fer-
mentation conditions are crucial
for effective fluorinated amino
acid incorporation.
4 Effect of fluorination on enzymatic
stability and reactivity
The first reports of the incorporation and effect on enzymatic ac-
tivity of fluorinated amino acids into proteins in vivo were made
in the late 1950s, employing some bacterial strains that were able
to grow in presence of p-FPhe and incorporate this amino acid
into some of their proteins [20–28]. Boyer was even able to feed
rabbits with this fluorinated analog that was then incorporated
into proteins of muscle, blood, and liver [29]. Aldolase with a
replacement of 25% and glyceraldehyde 3-phosphate dehydro-
genase with a replacement of approximately 16% of their Phe
residues were crystallized and assayed enzymatic activities were
similar to the corresponding wild-type enzymes. No difference in
stability was found after heat denaturation of glyceraldehyde 3-
phosphate dehydrogenase, as one of the first proofs that protein
fluorination can preserve enzymatic activity.
The amino acid tFLeu was also successfully incorporated into
E. coli proteins by means of the SPI method by Rennert and
Anker in 1963 [19]. They initially cultivated the corresponding
auxotrophic bacteria in presence of both leucine and the toxic
tFLeu. When growing only in presence of the fluorinated amino
acid, all leucine residues in proteins were replaced by tFLeu.
No abnormal enzymatic activity was detected, but a slower cell
growth was observed.
After these first results, several amino acids have been incor-
porated in vivo with dissimilar efficiencies and other aims than
enzymatic stabilization for biocatalysis [10–12, 30]. It was not
before the late 1990s that the fluorous effect was envisioned and
subsequently explored [7].
Since then, fluorinated variants of several proteins with dif-
ferent sizes, varying α-helix/β-sheet contents and overall folds
were obtained and characterized in order to determinate to what
extent this possibility could be regarded as a plausible enzy-
matic stabilization methodology [9–12]. Annexin V, Barstar,
Green Fluorescent Protein, and Mini-IGFBP-5 are some of the
early enzymes used as models [31]. These studies clearly showed
that in some cases extensive fluorination results in an increased
C2013 The Authors. Engineering in Life Sciences published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 343
www.els-journal.com Eng. Life Sci. 2014,14, 340–351 www.biotecvisions.com
NC
a
d
a
d
a
d
C
d
a
d
a
d
a
N
NC
a
d
a
d
a
d
C
d
a
d
a
d
a
N
NC
a
d
a
d
a
d
C
d
a
d
a
d
a
N
NC
a
d
a
d
a
d
C
d
a
d
a
d
a
N
NC
a
d
a
d
a
d
C
d
a
d
a
d
a
N
NC
a
d
a
d
a
d
C
d
a
d
a
d
a
N
NC
a
d
a
d
a
d
C
d
a
d
a
d
a
N
α4F2 (6,24) α4F2 (10,20) α4F4 (10,13,17,20)
α4F2 (13,17) α4F3 (6,13,20)
α4F3 (10,17,24)
α4F6
Increasing stability per hFLeu residue
Figure 4. Thermodynamic stability of a series of α4proteins with different Leu (white) and hFLeu (blue) motifs, as determinated by
GuHCl-induced unfolding. Adapted with permission from [38] with permission. Copyright 2009 American Chemical Society.
stability with retention of biological activity while in other cases
a destabilization or misfolding was observed, with subsequent
loss of enzymatic activity. It was also shown that incorporation
efficiency of fluorinated derivatives was dependent on the fluo-
rine content in the amino acid analog. The higher the content
of fluorine, the lower the substitution level in the target protein
was accomplished. It is clearly obvious that the bulkiness of the
fluorine-containing residues plays an important role during the
accommodation of amino acid residues in the folded state and
therefore, reactivity [32–34].
This observation inspired many groups to try to comprehend
the packing effects caused by fluorination, and to this purpose,
peptides and small proteins synthesized by APS are particu-
larly enlightening, since they permit access to homogenous sam-
ples with complete control of fluorination sites and amino acid
sequence.
The main structural models used to study the packing effects
following fluorination were 4-helix bundles. This 3-D arrange-
ment is commonly present in proteins and it has been the topic
of extensive de novo design efforts [35, 36]. It has been demon-
strated that the hydrophobic core of a 4-helix bundle can be
built from contact interactions between hydrophobic residues
at the “a” and “d” positions of a helical heptad (abcdefg)nre-
peat. Substitution of this hydrophobic core by fluorocarbon
interactions should be, in principle, conservative of the 3-D
structure.
One of the first reports of protein stabilization through fluori-
nation made on parallel coiled-coil protein domain synthesized
by APS was simultaneously conducted by the groups of Kumar
and coworkers [8–37]. They used a de novo designed coiled-coil
domain from a GCN4-bZip eukaryotic transcriptional activator,
a small protein capable of targeting DNA binding, as a simple
model to study fluorous effect. After substitution of all leucine
and valine residues by tFLeu and trifluorovaline, respectively,
an increase in the melting point from 47°Cto61°C was ob-
served. Similarly, substitution of solely leucine residues by tFLeu
resulted in a value of 62°C and an increase of approximately
1 kcal/mol for the free energy of unfolding, as suggested by denat-
uration experiments. Remarkably, the fluorinated GCN4-bZip
could retain its capability to bind target DNA as another early
indication that protein fluorination is conservative of biological
activity.
Other important studies on a de novo designed antiparallel
4-α-helix bundle protein were conducted by Marsh and cowork-
ers [38,39]. The protein model, namely α4H, contains Leu at the
three a and three d positions of the heptad repeat. The authors
studied the physicochemical properties of a series of fluorinated
analogs for these tetramers by progressively increasing the num-
ber of incorporated hFLeu at various positions within the core.
All the fluorinated analogs, namely α4Fn, retained the structure
of the native α4H. They observed that the stability of the α4Fn
proteins increased with the number of hFLeu residues. For ex-
ample, the protein α4F6,inwhichalltheLeuresiduesarereplaced
by hFLeu, was 14.8 kcal/mol more stable than α4H. Interestingly,
they observed that the increased stability depends not only on
the number of hFLeu, but also on the position and pattern of the
substitutions (Fig. 4). The most stabilizing arrangement seems
to be an alternating pattern in which hFLeu is incorporated at a
positions and Leu at the d positions, or vice versa.
The results obtained with these model peptides clearly
demonstrated the usefulness of the fluorous effect in protein
stabilization and encouraged several research groups to achieve
fluorinated amino acid incorporation into highly relevant en-
zymes for biocatalysis.
5 Evolved methodologies for generating
fluorinated proteins
Due to the limitations of APS to generate highly pure proteins in
acost-effectivemanner,SPI is stillthemost promisingmethodol-
ogyformanufacturing scalablequantitiesof fluorinated enzymes
orproteins for industrial applications. Differentefforts have been
made in order to “evolve” in vivo incorporation of fluorinated
amino acids trying to overcome the disadvantages exhibited by
SPI method, such as low protein yields, heterogeneous samples,
and noncontrol of substitution sites [40].
One of the methodologies to improve the substitution of a
natural amino acid by a noncanonical surrogate consists of en-
gineering the translational machinery of the host cell. This ap-
proach is normally called residue-specific evolution since it results
in the global substitution of all natural amino acid occurring sites
by its analog [41].
344 C2013 The Authors. Engineering in Life Sciences published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.els-journal.com Eng. Life Sci. 2014,14, 340–351 www.biotecvisions.com
Nonsense supression
Site specific incorporation of
non canonical amino acid
Evolved enzyme
Sense codon reassigment
Residue specific incorporation of
non canonical amino acid
Evolved enzyme
Overexpressed
aaRS
UAG
Mischarged
supressor tRNA
Supressor tRNA
Mischarged
sense tRNA
sense tRNA
aaRS
Auxotrophic host
= canonical amino acid
= fluorinated amino acid
Ortogonal
aaRS / RNA pair
Figure 5. Scheme for the sense codon reassignment and nonsense suppression strategies for evolution of fluorinated proteins. It should be
noted that orthogonal pair does not necessary need auxotrophy of the host cell.
The first results in this direction came from the laboratory
of Tirrell and coworkers [42]. Based on the previous observa-
tion that tFLeu could be incorporated in vivo by E. coli with-
out any requisite to evolve the host cell but in the case of the
highly fluorinated hFLeu, normal activation levels by wild-type
leucyl-tRNA synthetase were not high enough to achieve protein
synthesis [32], they engineered extra copies of the E. coli leuS en-
dogenous gene under a constitutive promoter to overexpress the
tRNA synthetase and achieve effective incorporation. By this
mechanism, enzymatic activity of the corresponding Leucyl-
tRNA synthetase was increased by eight-fold compared to the
wild-type strain. This evolved E. coli mutant was able to incor-
porate in vivo hFLeu and the incorporation efficiency was tested
on a coil–coil dimer, reaching a 74% replacement of all Leu
residues. This substitution induced protein structure stabiliza-
tion, and the increment on the melting temperature was even
higher than that of the analog that incorporates tFLeu (76°C
and 67°C, respectively). Both were more stable than the wild-
type protein (54°C melting point). The same stability results
were obtained during denaturation studies with urea. Incor-
poration of 5,5,5-trifluoroisoleucine and trifluorovaline into a
DNA-binding bzip model protein and murine dihydrofolate re-
ductase was also achieved in evolved bacterial hosts with high
incorporation efficiency (85–92%) by overexpressing an amino
acyl tRNA synthetase. In both cases, fluorine incorporation re-
sulted in good protein yields, increased stability against temper-
ature denaturation and unchanged biological activity [43–45].
Although this methodology allows a higher level of incorpo-
ration of the fluorinated amino acid into the tRNA and sub-
sequently improved yields for the fluorous protein, all existing
codons for the parental amino acid could be potentially recog-
nized by this mischarged tRNA, resulting in a general fluorina-
tion of the cellular proteome. This is mainly avoided by overex-
pressing the gene for the target protein whose transduction then
makes use of nearly all fluorinated analog, preserving cellular
viability. Still no control on substation sites can be accomplished
following this methodology.
As discussed above, studies with peptide models obtained by
APS led to the conclusion that a compromise should be achieved
linking fluorine content and enzyme stability and activity, in
view of the fact that for many folding motifs excessive fluori-
nation results in a destabilization of the tertiary structure. For
this reason, a desirable goal would be to achieve site-substitution
by in vivo procedures in order to be able to biosynthesize larger
amounts of fluorinated enzymes for biocatalysis with control on
substitution sites and fluorine content.
Thispossibilityhasbeen developedbyrecentadvancescreated
by synthetic biologists to reprogram the genetic code in order to
expand the number of amino acids that can be genetically en-
coded beyond those 21 considered canonical or standard amino
acids. This approach, known as nonsense suppression (Fig. 5),
consists in inserting a noncanonical amino acid during protein
synthesis in response to one of the stop codons by chemically
charging the corresponding suppressor tRNA [46].
For example, E. coli amber stop codon (UAG) can be intro-
duced at a desired location on a gene that codifies for a protein
of interest. If we expressed this gene into a cell-free expression
system in presence of the corresponding amber suppressor tRNA
charged with the desired noncanonical amino acid, the resulting
protein will incorporate this amino acid at the specific sites of
the polypeptide chains where the amber codon have been intro-
duced [47]. Generally, the expression of the new protein could
also be accomplished in vivo without dramatically affecting the
expression of other proteins, due to the scarcity of the amber
codon in the E. coli genome.
Nonsense suppression evolution effectively expanded the flex-
ibility for incorporation of fluorinated amino acids, but the
method has restrictions for big scale applications. Suppressor
aminoacyl-tRNA is a stoichiometrically consumed reagent and
its synthesis is troublesome and expensive.
C2013 The Authors. Engineering in Life Sciences published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 345
www.els-journal.com Eng. Life Sci. 2014,14, 340–351 www.biotecvisions.com
Further research to overcome this limitation was conducted;
efforts were focused on engineering an AARS/suppressor tRNA
pair by directed mutagenesis in order to make it capable of charg-
ing the suppressor tRNA with the noncanonical amino acid in
vivo, eliminating the need to synthesize this component. At the
same time, this engineered pair must be completely noncross
reactive and orthogonal to the endogenous AARS/tRNA pairs
within the host cell. Furter et al. used an engineered tRNA/AARS
pair, originally responsible for charging phenylalanine in
S. cerevisiae, for site-specific incorporation of p-fluorophenyl-
alanine in E. coli [48]. Similarly, a tRNA/tyrosyl-tRNA synthetase
pair from Methanococcus jannaschii, established through several
rounds of negative and positive screening from a mutant library
by the group of Schultz, was successfully employed for the incor-
poration of p-trifluoromethylphenylalanine in E. coli [49,50].
So far this technique is limited to one of the nonsense codons,
though efforts are under way to use sense codons (“codon re-
assignment”), and five base codon–anticodon pairs in order to
synthesize proteins incorporating multiple noncanonical amino
acids in vivo in one single experiment [51,52].
Other components of the protein synthesis machinery could
be also evolved to achieve efficient enzymatic fluorination. Mu-
tation on the ribosome is an alternative currently under early
investigation for improving noncanonical amino acid incorpo-
ration [53].
It is also important to mention that even though fluorination
without rational design can result in a partial loss of biological
activity, the artificial fluorinated enzymes can still be submitted
to directed evolution in order to recover full activity or even
enhance it [54,55]. This approach was used by Tirrell to recover
biological activity of chloramphenicol acyltransferase (CAT) af-
ter incorporation of tFLeu, which resulted in enhanced thermal
stability but partial activity loss. After screening a library of
error-prone PCR E. coli mutants auxotrophic for Leu, a fully ac-
tive fluorinated chloramphenicol acyltransferase with enhanced
thermal stability was identified [56].
6 Fluorination effect in stability and activity
of proteins and peptides with relevance to
biocatalysis
The industrial application of biotechnological routes for the gen-
eration of highly valuable chemicals has spread worldwide and
the discovery of novel biosystems capable of innovative chemical
transformation is a long-established research field under perma-
nent expansion. At the same time, finding new methodologies
for stabilizing these biocatalysts in order to permit their applica-
tion under nonnative reaction conditions represents a scientific
area with equivalent growth.
As discussed previously, introduction of fluorinated hy-
drophobic amino acids together with rational design constitutes
a new promising alternative for enzymatic stabilization against
the effect of organic solvents and high temperature, but this
possibility still needs to be fully explored and developed.
Despite our limited knowledge of how to achieve effective
stability by fluorination with minimum compromise on bio-
logical activity, some successful examples have been reported
of stabilization of enzymes with strong relevance for industrial
biocatalysis.
One of the first studies involving high-value biocatalysts was
conducted by Wiltschi on the residue specific fluorination of aro-
matic residues for lipase B from Candida Antarctica (CalB) [57]
This highly relevant industrial lipase has multiple applications,
including polymerizations, ring opening reactions of β-lactams,
resolutions of chiralalcohols and amines,modificationsof sugars
and related compounds, etc. [58, 59]. Wild-type CalB contains
319 amino acids, from which five are Trp residues, nine are Tyr,
and 11 correspond to Phe. A CalB expression vector inducible
by methanol and transformed into an auxotrophic P. pastoris
yeast as host cell was used for large-scale protein expression.
The introduction in parallel of the fluorinated amino acids 5-
fluorotryptophan, 3-fluorotyrosine, and 4-FPhe by SPI method
generated the corresponding fluorinated proteins with yields be-
tween 32 and 64% of the parental nonfluorinated lipase.
The MS analysis of the cell lysate after protein expression in
presence of the fluorinated amino acid shows the successful in-
corporation of all three analogs at multiple sites (Fig. 6A). The
incorporation was stochastic: a mixture of differentially fluori-
nated proteins was obtained and, even though complete sub-
stitution was observed in all three cases, the dominant species
in the mass spectra corresponded to those carrying three to
four replacements (Fig. 6B). Similar results were reported for
the incorporation of trifluoromethionine in P. pastoris and
S. cerevisae. As previously described, this seems to be a con-
sequence of the intracellular availability of the canonical amino
acid during fluorination in response to metabolic turnover.
Physicochemical characterization of the fluorinated proteins
by circular dichroism, far UV light absorption, and fluores-
cence showed some alterations of the secondary structure for
the fluorinated variants (Fig. 6C), but their behavior against
proteolysis by protein quinase K remained intact. Lipase activ-
ity assays showed retention of the enzymatic properties with a
similar profile of activity versus temperature but with a lower ef-
ficiency (Fig. 6D). Interestingly, the fluorinated variants retained
their lipase activity even after several months under storage at
4°C, against a substantial detriment observed for the wild-type
protein.
Further studies were conducted by Merkel et al. on an-
other highly relevant industrial lipase from Thermoanaerobacter
thermohydrosulfuricus [60]. This enzyme consists of 267 amino
acids, including six proline, 16 phenylalanine, and two trypto-
phan residues. In an attempt to study how global fluorination
affects structural integrity and enzymatic activity, simultane-
ous substitution of all mentioned residues by their fluorinated
analogs 4-fluoroproline, 4-FPhe, and 6-fluorotryptophan was
attempted in a single experiment. This results in 10% fluorina-
tion of the primary sequence. After expression protein in an E.
coli strain and SPI incorporation of the fluorinated analogs, the
fully fluorinated variant was obtained as the main species on the
MS profile, but protein production was achieved with a low yield
(11.5 mg/L culture compared to 47 mg/L for the nonfluorinated
protein). This lower production yield could be attributed to the
formation of inclusion bodies that lowers expression levels for
the fluorinated soluble variant.
CD spectroscopy showed no significant alteration of the
secondary structure as well as biological studies, which
346 C2013 The Authors. Engineering in Life Sciences published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.els-journal.com Eng. Life Sci. 2014,14, 340–351 www.biotecvisions.com
Figure 6. (A) Protein preparations of CalB expressed in the presence of the amino acids 5-fluorotryptophan (5FW), 3-fluorotyrosine (mFY),
and 4-FPhe (pFF). Wild-type CalB (MW =33 kDa) is indicated as WFY. Equal amounts of protein were loaded onto each lane. (B) ESI-
MS spectrum for the variant containing 5-fluorotryptophan (5FW). The peaks correspond to the congeners containing 1–5 5FW residues.
(C) Far-UV CD spectra for the fluorinated CalB variants. (C) Lipase activity of the variants (determinated by monitoring p-nitrophenylpalmitate
hydrolysis at 410 nm) after several months storage at 4°C compared to the wild type (WYF, set as 100%). Adapted from [57] with permission.
demonstrated the presence of lipase activity, but with a max-
imum that was only about 60% of the parental lipase. Inter-
estingly, the fluorinated variant showed an optimal temperature
value of approximately 60°C, 10°C lower than the wild-type pro-
tein. This constitutes a very desirable feature for the industrial
design of so-called “cold-wash” lipases.
A more systematic study on the same lipase was presented by
Hoesl et al. [61]. They incorporated fluorinated derivatives of
Phe, Pro, and Tyr at multiple positions and characterized their
structure, stability, and activity. The results indicated that for the
fluorinated Pro and Phe analogs, the protein yield was higher
than for the corresponding wild-type protein. However, none of
the congeners showed higher temperature resistance compared
to the parent lipase, even though the variant containing 3-FPhe
displayed a broader substrate tolerance than the wild type.
Other relevant enzymes for industrial biocatalysis are DNA
polymerases. These enzymes catalyze the synthesis of DNA
copies from a DNA template. Thermostable DNA poly-
merases have been applied for the detection and amplifica-
tion of DNA sequences of interest with the aim of identify-
ing microorganism contamination in food or pharmaceuticals,
virus detection, medical diagnosis, gene synthesis for molec-
ular biology applications, DNA sequencing, etc. [62–65]. All
these applications rely on the fidelity and stability of DNA
polymerases.
The group of Marx and coworker globally replaced 32 proline
residues on a 540 amino acid residue KlenTaq DNA polymerase
by (4R)-fluoroproline ((4R)-FPro) following the SPI method on
aE. coli auxotrophic for Proline [66]. According to the previous
example, in this case the yield of fluorinated protein was lower
compared to the wild type (0.2–0.5 vs. 8 mg/L). After MS-ESI
analysis, they found that about 92% of all Pro residues were re-
placed by (4R)-FPro. Interestingly, all the efforts for incorporat-
ing (4S)-fluoroproline failed. It has been shown previously that
C2013 The Authors. Engineering in Life Sciences published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 347
www.els-journal.com Eng. Life Sci. 2014,14, 340–351 www.biotecvisions.com
Figure 7. Chemical structures for assayed substrates for fluori-
nated S5 phosphotriesterase: paraoxon (1), chlorpyrifos (2), and
2-naphthylacetate (3). Adapted with permission from [74]. Copy-
right 2011 Wiley-VCH.
introduction of a fluorine atom in position 4 of Pro can strongly
affect the endo/exo conformation equilibrium of the proline ring
and this explains that usually only one of the stereoisomers of
4FPro can be tolerated as substitute for proline in certain pro-
tein structures [67,68]. As a consequence of fluorination, a small
decrease in thermal stability was detected after incubation for
several periods at 95°C. More importantly, no loss in fidelity,
activity, or sensitivity could be detected after PCR experiments
to amplify DNA fragments, compared to the parental wild-type
enzyme.
Organophosphate esters are highly toxic compounds, which
became widely dispersed on the Earth’s surface due to the in-
discriminate misuse of agricultural insecticides and during de-
velopment of chemical warfare agents. If ingested or adsorbed,
they cause respiratory failures and even death to humans and
other vertebrates [69]. Several approaches have been applied to
decontaminate the soil, such as incineration, acid or base hy-
drolysis, chemical oxidation, etc., but most of these methodolo-
gies are expensive or environmentally hazardous [70,71]. Stable
phosphoesterases constitute a promising biocatalytic alternative,
since they catalyze the hydrolysis of organophosphates esters into
harmless products. These enzymes can also find application in
the production of biofuels and the hydrolysis of other interesting
organic esters [72,73].
Baker et al. reported the improved refoldability of a phos-
photriesterase after residue-specific substitution of Phe by
p-FPhe [74]. This is a globular dimeric protein where dimer-
ization is controlled by particular contact interactions between
seven Phe residues. Introduction of fluorinated Phe should en-
hance the interaction force dimer formation (due to the in-
creased apolar surface area) and hence, stability. After 4FPhe
supplementation to an auxotrophic E. coli strain and protein iso-
lation and purification, 89.4% incorporation efficiency into the
phosphotriesterase was achieved, as determinated by MALDI-
TOF MS. Circular dichroism studies revealed that introduction
of 4FPhe improved refoldability after heating at 70°C. While
the wild-type phosphotriesterase exhibited a complete loss in
structure, the fluorinated variant still retained around 30% of
the native state. Determination of melting temperature for both
proteins showed a slightly improved stability, with melting tem-
peraturevalues about 1.3–2.5°Chigher infavorofthe fluorinated
variant.
The activity of the phosphotriesterases was tested against two
organophosphates and one ester as model substrates (Fig. 7).
Outstandingly, fluorination led to 5.8-fold higher turnover for
two of these substrates and a residual 40% activity after heating at
55°C was detected for the fluorinated phosphotriesterase, while
the wild-type variant retained only 13% of the activity after
analogous treatment (Table 1).
These selected examples clearly demonstrate that introduc-
tion of fluorinated amino acids constitute an alternative ap-
proach to traditional protein engineering for achieving stable
biocatalysts for industrial applications. Despite these being pre-
liminary results and none of these fluorinated biocatalysts have
been applied so far beyond laboratory experiments, we trust that
further investigations will indubitably lead to future technolog-
ical applications, at least in some cases.
7 Concluding remarks
The impact of biocatalysis has been remarkably amplified during
the last few decades, driven by a global necessity for transforming
our current production systems into more sustainable method-
ologies, but also as a result of the attractive advantages that
biosystems have to offer to the manufacture of fine chemicals.
The considerable advances made in this field by molecular biol-
ogist have remained so far limited to the possibilities offered by
the 20 natural amino acids.
Nowadays, the expansion of the pool of amino acids for
ribosomally mediated protein synthesis, in combination with
the latest genomic and proteomic methods, has made it pos-
sible to incorporate new properties into proteins and bioma-
terials, which will radically transform our traditional view of
biocatalysis.
Table 1. Kinetic parameters for wild-type S5 phosphotriesterase (PTE) and its fluorinated variant (pFF-PTE).
Substrate 1 Substrate 2 Substrate 3
PTE KM[×10−4M] 0.0031 ±0.0007 2.6 ±0.6 1.5 ±0.4
kcat [s−1]4.3±0.1 8.3 ±0.6 3.3 ±0.8
kcat/KM[×104M−1s−1] 1400 ±300 3.3 ±0.8 2.3 ±0.2
pFF-PTE KM[×10−4M] 0.0075 ±0.0012 8.7 ±2.6 1.9 ±0.5
kcat [s−1] 29.7 ±0.8 7.8 ±0.6 20.2 ±0.4
kcat/KM[×104M−1s−1] 400 ±800 0.9 ±0.2 10.8 ±0.7
Adapted with permission from [74]. Copyright 2011 Wiley-VCH.
348 C2013 The Authors. Engineering in Life Sciences published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.els-journal.com Eng. Life Sci. 2014,14, 340–351 www.biotecvisions.com
In this sense, the discovery of the fluorous effect has been
an inspiration for uninterrupted efforts trying to design effi-
cient incorporation methodologies for fluorinated amino acids
and to understand how fluorination affects structure and
reactivity.
Encouraged by our partial success hitherto, the initial results
in the stabilization of highly relevant biocatalysts for industrial
applications stimulate even more the interest of the scientific
community to improve these outcomes and it is not inconceiv-
able to anticipate that we will be eventually able to generate
complete functional highly stable enzymes and even cells with a
high fluorine content in their proteomes.
But, as we approach this point, what will be the next direc-
tion on our way to fluorinated biocatalysts? From our discussion
here, it results obvious to envisage that the limiting steps for
future industrial application of fluorinated enzymes will be the
ability to produce fluorinated amino acids on a large-scale and
at reasonable price. So far, all these fluorinated analogs have to
be synthesized chemically and supplemented into the culture
medium during cell host growth. Would it be possible to evolve
the bacterial metabolism in order to biosynthesize the fluori-
nated amino acid in vivo?
The group of Schultz has been able to evolve an E. coli strain
that is capable of biosynthesizing p-aminophenylalanine and in-
corporatethisnoncanonicalamino acidintoproteinsinresponse
to amber stop codons [75]. It is clear that metabolic engineer-
ing is the most advantageous approach to generate noncanon-
ical amino acids of interest directly in the protein synthesizing
cell.
However, it should be kept in mind that metabolic engineer-
ing of fluorinated amino acids is, by far, a much larger challenge
due to the intrinsic toxicity of the fluorinated metabolites and
the lack of a biocatalyst capable of fluorine activation. The re-
cent isolation of a fluorination enzyme from the soil bacterium
Streptomyces cattleya constitutes one exception [76,77]. The flu-
orinase catalyzes the conversion of (S)-adenosyl-l-methionine
and inorganic fluoride ion to 5-fluoro-5-deoxyadenosine,
which is afterwards metabolically transformed into fluoroacetate
and 4-fluorothreonine. This is an exceptional natural mech-
anism revealing the inimitable potential of enzymes as cata-
lysts since chemical activation of fluoride anion is particularly
difficult.
Some of these fluorinated metabolites have been subse-
quently biotransformed into other highly valuable pharmaco-
logical compounds, such as salinosporamide [78] and fluori-
nated polyketides, employing polyketide synthase and fluoroac-
etate [79]. Chemo-enzymatic methoxy to fluorine substitution
has been accomplished as well [80].
Although there is still a long way to go until we could biosyn-
thesize the desired fluorinated amino acid while simultaneously
incorporating it into the target enzyme in vivo, this topic con-
stitutes a very promising research field that may have a deep
future impact on biotechnological processes and protein-based
technologies.
We believe the evolution of fluorinated proteins could indeed
transform our perception of industrial biocatalysis and several
studies are currently conducted in our group in order to gen-
erate highly stable fluorinated variants of valuable industrial
enzymes.
Practical application
Protein evolution in order to develop stable enzymes for
industrial applications constitutes a field in continuous ex-
pansion. Previously, ribosome-mediated protein synthesis
was restricted to a repertoire of 20 canonical amino acids.
Recently, the emerging field of synthetic biology has the po-
tential to enrich protein biosynthesis with synthetic amino
acids that carry new functionalities not developed during
the evolution of the living world.
Introduction of fluorinated amino acids into the hy-
drophobic core of enzymes imprints these biocatalysts with
some outstanding properties, such as resistance to denat-
uration caused by organic solvents or high temperatures
as a consequence of the so-called “fluorous effect.” These
chemical properties represent very valuable features for
industrial applications of enzymes.
In this review, we present some advances in our under-
standing of the fluorous effect and we discuss some litera-
ture examples showing promising practical relevance. We
anticipate that further research on this topic will allow us
to design a novel generation of uniquely stable biocatalysts
for industrial and therapeutic use.
The authors are kindly grateful to Alexander von Humboldt Foun-
dation and UniCat cluster of excellence at TU Berlin for financial
support and postdoctoral fellowship. They are in debt to Dr. Patrick
Durkin for the critical reading of this manuscript.
The authors have declared no conflict of interest.
8 References
[1] Vasic-Racki, D., History of industrial biotransformations—
dreams and reality, in: Liese, A., Seelbach, K., Wandrey, C.
(Eds.), Industrial Biotransformations, Wiley VCH, Weinheim
2006, pp. 1–36.
[2] Illanes, A., Introduction, in: Illanes, A. (Ed.), Enzyme
Biocatalysis—Principles and Applications, Springer, Dordrecht
2008, pp. 1–56.
[3] Davis, B., Boyer, V., Biocatalysis and enzymes in organic syn-
thesis. Nat. Prod. Rep. 2001, 18, 618–640.
[4] Gr¨
oger, H., Asano. Y., Introduction—principles and historical
landmarks of enzyme catalysis in organic synthesis, in: Drauz,
K., Gr¨
oger, H., May, O. (Eds.) Enzyme Catalysis in Organic
Synthesis, Wiley VCH, Weinheim 2012, pp. 3–42.
[5] ´
O’F´
ag´
ain, C., Enzyme stabilization-recent experimental
progress. Enzyme Microb. Technol. 2003, 33, 137–149.
[6] Moehlenbrock, M., Minteer, S., Introduction to the field of en-
zyme immobilization and stabilization, in: Minteer, S. (Ed.),
Enzyme Stabilization and Immobilization-Methods and Proto-
cols, Springer, Dordrecht 2011, pp. 1–8.
[7] Marsh, E., Towards the nonstick egg: Designing fluorous pro-
teins. Chem. Biol. 2000, 7, R153–R157.
C2013 The Authors. Engineering in Life Sciences published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 349
www.els-journal.com Eng. Life Sci. 2014,14, 340–351 www.biotecvisions.com
[8] Yoder, N., Kumar, K., Fluorinated amino acids in protein de-
sign. Chem. Rev. Soc. 2002, 31, 355–341.
[9] Buer, B., Marsh, E., Fluorine: A new element in protein design.
Prot. Sci. 2012, 21, 453–462.
[10] Cametti,M.,Crousse,B.,Metrangolo,P.,Milani,R.etal,The
fluorous effect in biomolecular applications. Chem.Soc.Rev.
2012, 41, 31–42.
[11] Salwiczek, M., Nyakatura, E., Gerling, U., Ye, S. et al., Fluo-
rinated amino acids: Compatibility with native protein struc-
tures and effects on protein–protein interactions. Chem. Soc.
Rev. 2012, 41, 2135–2171.
[12] Merkel, L., Budisa, N., Organic fluorine as a polypeptide build-
ing element: In vivo expression of fluorinated peptides, pro-
teins and proteomes. Org. Biomol. Chem. 2012, 10, 7241–7261.
[13] Begu´
e, J.-P., Bonnet Delpon, D. (Eds.), Bioorganic and Medic-
inal Chemistry of Fluorine, Wiley & Sons, New Jersey 1993,
pp. 1–22.
[14] Biffinger, J., Kim, H., DiMagno, S., The polar hydropho-
bicity of fluorinated compounds. Chembiochem 2004, 5,
622–627.
[15] Dalvi, V., Rossky, P., Molecular origins of fluorocarbon
hydrophobicity. Proc. Natl. Acad. Sci. USA 2010, 31, 13603–
13607.
[16] Buer, B., Levin, B., Marsh, E., Influence of fluorination on the
thermodynamic of protein folding. J. Am. Chem. Soc. 2012,
134, 13027–13034.
[17] Qiu, X.-L., Meng, W.-D., Qing, F.-L., Synthesis of fluorinated
amino acids. Tetrahedron 2004, 60, 6711–6745.
[18] Qiu, X.-L., Qing, F.-L., Recent advances in the synthesis of
fluorinated amino acids. Eur. J. Org. Chem. 2011, 2011, 3261–
3278.
[19] Rennert, O., Anker, H., On the Incorporation of 5,5,5-
trifluoroleucine into proteins of E. coli.Biochemistry 1963, 2,
471–476.
[20] Conway, T., Lansford, E., Shive. W., Influence of phenylalanine
analogues upon bacterial accumulation and incorporation of
phenylalanine. J. Bacteriol. 1963, 85, 141–149.
[21] Minks, C., Alefelder, S., Moroder, L., Huber, R. et al., Towards
new protein engineering: In vivo building and folding of pro-
tein shuttles for drug delivery and targeting by the selective
pressure incorporation (SPI) method. Tetrahedron 2000, 56,
9431–9442.
[22] Hartman, M., Josephson, K., Szostak, J., Enzymatic aminoacy-
lation of tRNA with unnatural amino acids. Proc. Natl. Acad.
Sci. USA 2006, 103, 4356–4361.
[23] Hussain, T., Kamarthapu, V., Kruparani, S., Deshmukh, M.
et al., Mechanistic insights into cognate substrate discrimina-
tion during proofreading in translation. Proc. Natl. Acad. Sci.
USA 2010, 107, 22117–22121.
[24] Kim, H.-W., Perez, J., Ferguson, S., Campbell, I. The specific
incorporation of labeled aromatic amino acids into proteins
through growth of bacteria in the presence of glyphosate. FEBS
Lett. 1990, 272, 34–36.
[25] Filler, R., Saha, R. Fluorine in medicinal chemistry: A century
of progress and a 60-year retrospective of selected highlights.
Future Med. Chem. 2009, 1, 777–791.
[26] Chang, W., White, P., Basic principles, in: Chang, W., White, P.
(Eds.),Fmoc SolidPhase Peptide Synthesis:A Practical Approach,
Oxford University Press, Oxford 2000, pp. 9–40.
[27] Amblard, M., Fehrentz, J., Martinez, J., Subra, G., Methods
and protocols of modern solid phase peptide synthesis. Mol.
Biotechnol. 2006, 33, 239–254.
[28] Yoshida, A., Studies on the mechanism of protein synthesis: In-
corporation of p-fluorophenylalanine into α-amilase of bacil-
lus subtilis. Biochim. Biophys. Acta 1960, 41, 89–98.
[29] Westhead, E., Boyer, P., The incorporation of p-
fluorophenylalanine into some rabbit enzymes and other pro-
teins. Biochim. Biophys. Acta 1961, 54, 145–156.
[30] Richmond, M., The effect of amino acid analogues on growth
and protein synthesis in microorganism. Bacteriol. Rev. 1962,
26, 398–420.
[31] Budisa, N., Pipitone, O., Siwanowicz, I., Rubini, M. et al.,
Efforts towards the design of “teflon” proteins: In vivo transla-
tionwith trifluorinatedleucineand methionineanalogs. Chem.
Biodivers. 2004, 1, 1465–1474.
[32] Tang, Y., Ghirlanda, Y., Petka, W., Nakajima, T. et al., Fluori-
nated coiled-coil proteins prepared in vivo display enhanced
thermal and chemical stability. Angew. Chem. Int. Ed. 2001, 40,
1494–1496.
[33] O’Hagan, D., Rzepab, H., Some influences of fluorine in bioor-
ganic chemistry. Chem. Commun. 1997, 645–642.
[34] J¨
ackel, C., Salwiczek, M., Koksch, B., Fluorine in a native pro-
tein environment—how the spatial demand and polarity of
fluoroalkyl groups affect protein folding. Angew. Chem. Int.
Ed. 2006, 45, 4198–4203.
[35] Lehmann, A., Saven, J., Computational design of four-helix
bundle proteins that bind nonbiological cofactors. Biotechnol.
Prog. 2008, 24, 74–79.
[36]Hill,R.,Raleigh,D.,Lombardi,A.,DeGrado,W.,Denovo
design of helical bundles as models for understanding protein
folding and function. Acc. Chem. Res. 2000, 33, 745–754.
[37] Tang, Y., Ghirlanda, G., Vaidehi, N., Kua, J. et al., Stabiliza-
tion of coiled-coil peptide domains by introduction of trifluo-
roleucine. Biochemistry 2001, 40, 2790–2796.
[38] Buer, B., Salud-Bea, R., Hashimi, H., Marsh, E., Engineering
protein stability and specificity using fluorous amino acids:
The importance of the packing effects. Biochemistry 2009, 48,
10810–10817.
[39]Lee,K.,Lee,H.-Y.,Slutsky,M.,Anderson,J.etal.,Fluorous
effect in proteins: De novo design and characterization of
afour-α-helix bundle protein containing hexafluoroleucine.
Biochemistry 2004, 43, 16277–16284.
[40] Link, A., Mock, M., Tirrell, D., Noncanonical amino acid in
protein engineering. Curr. Opin. Biotechnol. 2003, 14, 603–609.
[41] Johnson, J., Lu, Y., Van Deventer, J., Tirrell, D., Residue-specific
incorporation of noncanonical amino acids into proteins: Re-
cent developments and applications. Curr.Opin.Chem.Biol.
2010, 14, 774–780.
[42] Tang, Y., Tirrell, D., Biosynthesis of a highly stable coiled-coil
protein containing hexafluoroleucine in an engineered bacte-
rial host. J. Am. Chem. Soc. 2001, 123, 11089–11090.
[43] Wang, P., tang, Y., Tirrell, D., Incorporation of trifluo-
roisoleucine into proteins in vivo. J. Am. Chem. Soc. 2003,
125, 6900–6906.
[44] Wang, P., Fichera, A., Kumar, K., Tirrell, D., Alternative trans-
lations of a single RNA message: An identity switch of (2S,3R)-
4,4,4-trifuorovaline and isoleucine codons. Angew. Chem. Int.
Ed. 2004, 43, 3664–3666.
350 C2013 The Authors. Engineering in Life Sciences published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.els-journal.com Eng. Life Sci. 2014,14, 340–351 www.biotecvisions.com
[45] Son, S., Tanrikulu, C., Tirrell, D., Stabilization of bzip pep-
tides through incorporation of fluorinated aliphatic residues.
Chembiochem. 2006, 7, 1251–1257.
[46] Hecht, S., Alford, B., Kuroda, Y., Kitano, S., Chemical aminoa-
cylation of tRNA’s. J. Biol. Chem. 1978, 253, 4517–4520.
[47] Noren, C., Anthony-Cahill, S., Griffith, M., Schultz, P., A gen-
eral method for site-specific incorporation of unnatural amino
acids into proteins. Science 1989, 244, 182–188.
[48] Furter, R., Expansion of the genetic code: Site-directed p-
fluoro-phenylalanine incorporation in Escherichia coli.Prot.
Sci. 1998, 7, 419–426.
[49] Jackson, J., Duffy, S., Hess, K., Mehl, R., Improving nature’s
enzyme active site with genetically encoded unnatural amino
acids. J. Am. Chem. Soc. 2006, 128, 11124–11127.
[50] Jackson, J., Hammill, J., Mehl, R., Site-specific incorporation
of a 19F-amino acid into proteins as an NMR probe for char-
acterizing protein structure and reactivity. J. Am. Chem. Soc.
2007, 129, 1160–1166.
[51] Goltermann, L., Larsen, M., Banerjee, R., Joerger, R. et al.,
Protein evolution via amino acid and codon elimination. PLoS
One 2010, 5, e10104.
[52] Antonczak, A., Morris, J., Tippmann, E., Advances in the
mechanism and understanding of site-selective noncanonical
amino acid incorporation. Curr. Opin. Struct. Biol. 2011, 21,
481–487.
[53] Liu, W., Huang, Y., Translation systems for incorporating non-
canonical amino acids into proteins including modified ribo-
somes with a truncated L11 protein. US Patent 20120237971,
2012.
[54] Brustad, E., Arnold, F., Optimizing non-natural protein func-
tion with directed evolution. Curr. Opin. Chem. Biol. 2011, 15,
201–210.
[55] Wang, M., Sia, T., Zhao, H., Biocatalyst development by di-
rected evolution. Bioresour. Technol. 2012, 115, 117–125.
[56] Montclare, J., Tirrell, D., Evolving proteins of novel composi-
tion. Angew. Chem. Int. Ed. 2006, 45, 4518–4521.
[57] Budisa, N., Wenger, W., Wiltschi, B., Residue-specific global
fluorination of Candida Antarctica lipase B in Pichia pastoris.
Mol. Biosyst. 2010, 6, 1630–1639.
[58] Kapoor, M., Gupta, M., Lipase promiscuity and its biochemical
applications. Process Biochem. 2012, 47, 555–569.
[59] Marciello, M., Filicez, M., Palomo, J., Different strategies to
enhance the activity of lipase catalysts. Catal. Sci. Technol. 2012,
2, 1531–1543.
[60] Merkel, L., Schauer, M., Antranikian, G., Budisa, N., Parallel
incorporation of different fluorinated amino acids. On the way
to “Teflon” proteins. Chembiochem 2010, 11, 1505–1507.
[61] Hoesl,M.,Acevedo-Rocha,C.,Nehring,S.,Royter,M.etal.,
Lipase congeners designed by genetic code engineering. Chem-
CatChem. 2011, 3, 213–221.
[62] Hamilton, S., Farchaus, J., Davis, M., DNA polymerases
as engines for biotechnology. Biotechniques 2001, 31,
370–383.
[63] Kranaster, R, Marx, A., Engineered DNA polymerases in
biotechnology. ChemBioChem 2010, 11, 2077–2084.
[64] Batista, E., Santos, D., Chies, J., Basso, L., Use of nucleic acids
as standards for detection and quantification of microbes for
diagnosis of infections in food and cosmetic industries. PCT Int.
Appl. WO 2008122097, 2008.
[65] Moreel, J., Method for detection and identification of microbes
using hypotonic gels and PCR for potential use in diagnostic
and industrial applications. PCT Int. Appl. WO 2009090322,
2009.
[66] Holzberger, B., Marx, A., Replacing 32 proline residues
by noncanonical amino acids results in a highly ac-
tive DNA polymerase. J. Am. Chem. Soc. 2010, 132,
15708–15713.
[67] Hodges, J., Raines, R., Stereoelectronic effects on collagen sta-
bility: The dichotomy of 4-fluoroproline diastereomer. J. Am.
Chem. Soc. 2003, 125, 9262–9263.
[68] Holzberger, B., Obeid, S., Welte, W., Diederichs, K. et al. Struc-
tural insights into the potential of 4-fluoroproline to modulate
biophysical properties of proteins. Chem. Sci. 2012, 3, 2924–
2931.
[69] Povey, A., Gene-environmental interactions and organophos-
phate toxicity. Toxicology 2010, 278, 294–304.
[70] Wagner, G., Yang, Y-C. Rapid nucleophilic/oxidative decon-
tamination of chemical warfare agents. Ind. Eng. Chem. Res.
2002, 41, 1925–1928.
[71] Riede, U., Kleffner, H., Seubert, B., Process and alkali humi-
nates for decontaminating pesticides and chemical warfare agents
on surfaces and solid objects. European Patent Application
EP499696, 1992.
[72] Singh, B., Organophosphorus-degrading bacteria: Ecology
and industrial applications. Nat. Rev. Microbiol. 2009, 7,
156–164.
[73] Mandrich, L., Merone, L., Manco, G., Hyperthermophilic
phosphotriesterases/lactonases for the environment and hu-
man health. Environ. Technol. 2010, 31, 1115–1127.
[74] Baker, P., Montclare, J., Enhanced refoldability and thermoac-
tivity of fluorinated Phosphotriesterase. Chembiochem 2011,
12, 1845–1848.
[75] Mehl, R., Anderson, C., Santoro, S., Wang, L. et al., Generation
of a bacterium with a 21 amino acid genetic code. J. Am. Chem.
Soc. 2003, 125, 935–939.
[76] Zhu,X.,Robinson,D.,McEwan,A.,O’Hagan,D.etal.,Mecha-
nism of enzymatic fluorination in Streptomyces Cattleya.J. Am.
Chem. Soc. 2007, 129, 14597–14604.
[77] Cobb, S., Deng, H., McEwan, A., Naismith, J. et al.,
Substrate specificity in enzymatic fluorination. The
fluorinase from Streptomyces cattleya accepts 2-
deoxyadenosine substrates. Org. Biomol. Chem., 2006, 4,
1458–1460.
[78] Eust´
aquio, A., O’Hagan, D., Mooore, B., Engineering fluo-
rometabolite production: Fluorinase expression in salinispora
tropica yields fluorosalinosporamide. J. Nat. Proc. 2010, 73,
378–382.
[79] Walker, M., Thuronyi, B., Charkoudian, L., Lowry, B. et al.,
Expanding the fluorine chemistry of living systems using engi-
neered polyketide synthase pathways. Science 2013, 341, 1089–
1094.
[80] Rentmeister, A., Arnold, F., Fasan, R., Chemo-enzymatic fluo-
rination of unactivated organic compounds. Nat. Chem. Biol.
2009, 5, 26–28.
C2013 The Authors. Engineering in Life Sciences published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 351
