Site-Specific Incorporation of Selenocysteine Using an Expanded Genetic Code and Palladium-Mediated Chemical Deprotection

Article abstract of DOI:10.1021/jacs.8b04603

Selenoproteins containing the 21^st amino acid selenocysteine (Sec) exist in all three kingdoms of life and play essential roles in human health and development. The distinct low pK_a, high reactivity, and redox property of Sec also afford unique routes to protein modification and engineering. However, natural Sec incorporation requires idiosyncratic translational machineries that are dedicated to Sec and species-dependent, which makes it challenging to recombinantly prepare selenoproteins with high Sec specificity. As a consequence, the function of half of human selenoproteins remains unclear, and Sec-based protein manipulation has been greatly hampered. Here we report a new general method enabling the site-specific incorporation of Sec into proteins in E. coli. An orthogonal tRNA^Pyl-ASecRS was evolved to specifically incorporate Se-allyl selenocysteine (ASec) in response to the amber codon, and the incorporated ASec was converted to Sec in high efficiency through palladium-mediated cleavage under mild conditions compatible with proteins and cells. This approach completely obviates the natural Sec-dedicated factors, thus allowing various selenoproteins, regardless of Sec position and species source, to be prepared with high Sec specificity and enzyme activity, as shown by the preparation of human thioredoxin and glutathione peroxidase 1. Sec-selective labeling in the presence of Cys was also demonstrated on the surface of live E. coli cells. The tRNA^Pyl-ASecRS pair was further used in mammalian cells to incorporate ASec, which was converted into Sec by palladium catalyst in cellulo. This robust and versatile method should greatly facilitate the study of diverse natural selenoproteins and the engineering of proteins in general via site-specific introduction of Sec.

Full text of DOI:10.1021/jacs.8b04603

Article
pubs.acs.org/JACS
Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
Site-Specific Incorporation of Selenocysteine Using an Expanded
Genetic Code and Palladium-Mediated Chemical Deprotection
Jun Liu,† Feng Zheng,‡,⊥ Rujin Cheng,§,⊥ Shanshan Li,†,∥ Sharon Rozovsky,§ Qian Wang,‡
and Lei Wang*,†
†University of California, San Francisco, Department of Pharmaceutical Chemistry, 555 Mission Bay Boulevard South, San
Francisco, California 94158, United States
‡Hangzhou Research Institute of Technical Institute of Physics and Chemistry, Chinese Academy of Sciences, Hangzhou 310018,
China
§University of Delaware, Department of Chemistry and Biochemistry, Newark, Delaware 19716, United States
∥Department of Chemistry and Center for Therapeutics and Diagnostics, Georgia State University, Atlanta, Georgia 30302, United
States
S
* Supporting Information
ABSTRACT: Selenoproteins containing the 21st amino acid selenocys-
teine (Sec) exist in all three kingdoms of life and play essential roles in
human health and development. The distinct low pKa, high reactivity,
and redox property of Sec also afford unique routes to protein
modification and engineering. However, natural Sec incorporation
requires idiosyncratic translational machineries that are dedicated to Sec
and species-dependent, which makes it challenging to recombinantly
prepare selenoproteins with high Sec specificity. As a consequence, the
function of half of human selenoproteins remains unclear, and Sec-based protein manipulation has been greatly hampered. Here
we report a new general method enabling the site-specific incorporation of Sec into proteins in E. coli. An orthogonal tRNAPyl-
ASecRS was evolved to specifically incorporate Se-allyl selenocysteine (ASec) in response to the amber codon, and the
incorporated ASec was converted to Sec in high efficiency through palladium-mediated cleavage under mild conditions
compatible with proteins and cells. This approach completely obviates the natural Sec-dedicated factors, thus allowing various
selenoproteins, regardless of Sec position and species source, to be prepared with high Sec specificity and enzyme activity, as
shown by the preparation of human thioredoxin and glutathione peroxidase 1. Sec-selective labeling in the presence of Cys was
also demonstrated on the surface of live E. coli cells. The tRNAPyl-ASecRS pair was further used in mammalian cells to
incorporate ASec, which was converted into Sec by palladium catalyst in cellulo. This robust and versatile method should greatly
facilitate the study of diverse natural selenoproteins and the engineering of proteins in general via site-specific introduction of
Sec.
chemical ligation, promotion of protein oxidative folding, and
enhancement of enzyme activity.4−19 However, detailed
investigation of native selenoproteins and exploitation of Sec
for general proteins have been hampered by the difficulty of
procuring selenoproteins with high Sec specificity.
Natural incorporation of Sec into proteins involves multiple
dedicated factors using a complicated mechanism distinct from
that for the 20 common amino acids (Figure 1a).20−24 First,
Sec has no cognate aminoacyl-tRNA synthetase and is
biosynthesized on the tRNASec. tRNASec is initially charged
with serine by SerRS to form Ser-tRNASec. The charged serine
is then converted into Sec by selenocysteine synthase (SelA) in
bacteria, while in archaea and eukaryotes serine on the tRNASec
is first converted into o-phosphoserine by a kinase (PSTK) and
then into Sec by SepCysS synthase. Second, the resulting Sec-
INTRODUCTION
■
Selenoproteins are a special class of proteins containing the
21st amino acid, selenocysteine (Sec), which closely resembles
Cys with sulfur replaced by selenium. Sec has unique
properties including a low pKa, high reactivity, and resistance
to excessive oxidation, rendering selenoproteins distinct
biological functions. Selenoproteins are found in all three
domains of life, and there are 25 selenoprotein-encoding genes
in human genome. Implicated in antioxidant defense and
various diseases such as cancer and aging, selenoproteins are
important for human health and development.1−3 Nonetheless,
the function of at least half of the human selenoproteins
remains unclear. Meanwhile, the unique properties of Sec also
provide attractive avenues for various applications, such as
selective protein modification and bioconjugation, chemical
mutagenesis (to Ala, Ser, and dehydroalanine), formation of
novel bonds (selenylsulfide Se−S, diselenide Se−Se, or
selenoether Se−C bonds), facilitation of challenging native
Received: May 2, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.8b04603
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 1. Site-specific incorporation of Sec into proteins. (a) Natural translation machinery for Sec incorporation uses Sec-dedicated genetic factors
and is species dependent. (b) Our method for Sec incorporation obviates all Sec-dedicated factors and is generally compatible with selenoproteins.
tRNASec requires a Sec-specific elongation factor, SelB in
bacteria, and eEFSec in archaea and eukaryotes, to enter the
ribosome A site, where it decodes the opal stop codon UGA as
Sec translationally. Third, the reassignment of the target UGA
codon from a stop signal to Sec is specified by the Sec-
insertion sequence (SECIS), which forms a stem−loop
structure and is recognized by SelB/eEFSec. SECIS is located
immediately next to the UGA codon within the selenoprotein
coding region in bacteria, but is situated in the 3′-untranslated
region many nucleotides downstream of the UGA codon in
mRNAs of archaea and eukaryotes. This complex and species-
dependent mechanism for Sec incorporation makes it
challenging to express selenoproteins recombinantly.
Chemical and biosynthetic methods have been developed to
prepare Sec-containing proteins yet with various limitations.
Solid phase peptide synthesis and native chemical ligation
provide complete control over the number and placement of
Sec in a protein sequence,4,7,25−27 but it becomes challenging
when synthesizing and refolding large proteins. The natural Sec
incorporation machinery of E. coli has been exploited to
produce selenoproteins,28−31 but the species differences in
SECIS sequence and location often introduce unwanted amino
acid changes when an archaeal or eukaryotic selenoprotein is
heterologously expressed in E. coli. To circumvent the need for
SelB and SECIS element in bacteria, efforts have been focused
on engineering the E. coli tRNASec or tRNASec and EF-Tu
together,32,33 allowing the resultant tRNASec to be recognized
by EF-Tu directly. These efforts successfully enabled the
expression of both bacterial and mammalian selenoproteins in
E. coli without additional mutations.32−37 Nonetheless, this
approach still relies on SelA to convert serine to Sec on the
mutant tRNASec, and mutation of tRNASec often decreases the
conversion efficiency thus resulting in misincorporation of
serine. In addition, the engineered tRNASec can be used in
bacteria only.
Expansion of the genetic code using an orthogonal tRNA/
synthetase pair38,39 has the potential to site-specifically
incorporate Sec while completely obviating Sec-dedicated
factors. Recently, an orthogonal tRNALeu/CmnRS pair40 was
employed to incorporate a photocaged Sec analog into a model
protein EGFP in yeast, and the Sec analog was converted into
Sec through photolysis.41 Unfortunately, photolysis to generate
Sec results in significant contamination of dehydroalanine
(>15%).41 The photocaged Sec analog is markedly bulkier than
Sec in size, which may not fit well in buried Sec sites in native
selenoproteins due to potential interference with protein
folding. In fact, no native selenoprotein has been prepared with
this approach.41 The tRNALeu/CmnRS pair cannot be used in
the workhorse E. coli or other bacteria either. In addition,
photolysis using UV light will prevent its effective application
in animal tissues and intact animals.
Exploiting genetic code expansion and chemical depro-
tection, we report here a facile and general method for site-
specific incorporation of Sec into proteins in high specificity.
We evolved an orthogonal tRNAPyl/ASecRS pair to genetically
incorporate unnatural amino acid ASec, a Sec analog with a
B
DOI: 10.1021/jacs.8b04603
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
small allyl protecting group, without needing any Sec-
dedicated factors. Following incorporation, ASec was con-
verted into the native Sec in high efficiency by a facile
palladium-mediated cleavage under mild biocompatible
conditions (Figure 1b). Using this approach, we generated
both artificial and human native selenoproteins from E. coli,
obtaining high enzyme activity owing to the exceptional
specificity of Sec incorporation. We also labeled Sec-
incorporated proteins selectively in the presence of Cys on
live E. coli cell surface, and demonstrated the incorporation of
ASec and conversion into Sec in proteins in mammalian cells.
In addition, the genetically encoded ASec affords a short
terminal alkene for protein cross-linking via thiol-ene reaction
and protein labeling via inverse electron demand Diels−Alder
reaction. This method will be compatible with various cell
types and facilitate the production, engineering, and research
of artificial and native selenoproteins.
with allyl bromide afforded the N-Boc-Se-allyl-L-selenocysteine,
which was deprotected via 4 M HCl/dioxane, affording the
targeted ASec HCl salt as a white solid with high purity (74%
yield, over 3 steps).
Genetic Incorporation of ASec into Proteins. To
genetically encode ASec in E. coli, we evolved a pyrrolysyl-
tRNA synthetase (PylRS) mutant specific for it. A
Methanosarcina mazei PylRS (MmPylRS) mutant, C348W/
W417S, was reported to incorporate S-allyl cysteine (ACys).44
However, when testing this mutant using the EGFP reporter
with a TAG stop codon at the permissive site 182,49 we did not
detect green fluorescence, suggesting that the MmPylRS
C348W/W417S mutant was inefficient in charging ACys. We
then transplanted the mutations to the Methanosarcina barkeri
PylRS (MbPylRS), and the corresponding MbPylRS(C313W/
W382S) showed faint yet ACys-dependent green fluorescence
(Figure S1). To improve the incorporation efficiency, we
generated a focused MbPylRS library by randomizing residues
C313 and W382, and selected the library for ACys specificity
using methods previously described.50 All identified hits
converged into a single mutant (C313W/W382T), which
was 3-fold more efficient than MbPylRS(C313W/W382S) in
incorporating ACys, as demonstrated by fluorescence, SDS-
PAGE, and Western analysis of the expressed EGFP proteins
(Figures S1 and S2). On the basis of this MbPylRS(C313W/
W382T) mutant, we further randomized active site residues
L270, Y271, and L274, but all selected mutants, despite
changes on the DNA level, had the same amino acid sequence
as MbPylRS(C313W/W382T) (SI).
Considering the structural similarity between ACys and
ASec, we attempted using the evolved ACys-specific synthetase
to incorporate ASec into EGFP in E. coli. However, the initial
trial of the MbPylRS(C313W/W382S) or MbPylRS(C313W/
W382T) synthetase did not produce any ASec-containing
proteins, and E. coli cells grew considerably slower when ASec
was added in the media. We reasoned that the observed
toxicity could be due to allyl selenide resulting from the
cleavage of ASec by cysteine-S-conjugate β-lyase in vivo.51 To
alleviate the toxicity, we explored ASec incorporation into
EGFP (182TAG) in the presence of various additives,
including L-Cys, DL-propargylglycine, and 2-(aminooxy)acetic
acid, which inhibit β-lyase or selenium toxicity (Figure S3). We
found that L-Cys could rescue the toxicity of ASec: the yield of
cells grown in the presence of Cys and ASec was significantly
higher than ASec alone (Figure S4). Western analysis showed
that full-length EGFP could now be expressed in the presence
of ASec accompanied by L-Cys only (Figure S3). This L-Cys
enabling effect of ASec incorporation was corroborated by
expressing another protein thioredoxin (Trx) (see below and
Figure S5). The MbPylRS(C313W/W382S) was also able to
incorporate ASec after adding L-Cys, but the MbPylRS-
(C313W/W382T) was 2−3 times more efficient than
MbPylRS(C313W/W382S) for incorporating ASec (Figure
2a). We thus named this MbPylRS bearing mutations of
C313W/W382T as ASecRS for clarity and used it in all
subsequent experiments.
RESULTS
■
Chemical Protection−Deprotection Strategy for Site-
Specific Incorporation of Sec into Proteins. Sec is not
stable in its free form due to its high reactivity, and directly
feeding cells with Sec may result in misincorporation with
endogenous Cys at respective sites because of the structural
similarity between Cys and Sec.42 Inspired by recent progress
on chemically decaging cysteine from the allyl, acetamido-
methyl, or thiazolidine protecting groups using palladium
catalysts under mild conditions,43−47 we envisioned that an
orthorgonal tRNA/synthetase pair could be evolved to
genetically incorporate Se-allyl selenocysteine (ASec) in
response to the amber stop codon and that the allyl group
could be chemically removed by palladium to generate Sec
(Figure 1b). We decided to incorporate ASec because the allyl
group will distinguish ASec from Cys structurally, potentially
allowing a synthetase to be evolved to recognize ASec
specifically. Meanwhile, the allyl group is relatively small in
size, which will mitigate potential interference to protein
folding when ASec is incorporated at native Sec sites. In
addition, since the allyl group can be readily removed from the
thiol of Cys under mild conditions compatible with proteins,
we reasoned that the C−Se bond in ASec would be similarly
cleaved by palladium to generate Sec in proteins.
Chemical Synthesis of ASec. We initially tried to
synthesize ASec starting from the unprotected or fully
protected L-selenocysteine following the published procedures,
yet the yield was less than 10% or the final product had low
quality, respectively (Supporting Information, SI). We then
developed a semiprotected strategy based on a method for the
synthesis of N-Boc-p-methoxy benzyl-L-cysteine (Scheme 1).48
Specifically, the N-Boc-L-selenocystine was reduced by NaBH4
in ethanol. Treatment of the resulting N-Boc-L-selenocysteine
Scheme 1. Synthesis of ASec Starting from Semi-Protected
L-Selenocystine
To determine ASec incorporation fidelity and site flexibility,
we expressed human Trx gene containing an amber stop codon
TAG at either Cys32 or Cys73 site and a C-terminal Hisx6 tag,
together with the tRNAPyl/ASecRS pair. Full-length Trx was
obtained in good yield (3 mg/L) in the presence of 0.5 mM
ASec and 0.8 mM L-Cys, but was undetectable on SDS-PAGE
and Western blot in their absence (Figures 2b and S5).
C
DOI: 10.1021/jacs.8b04603
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 2. Site-specific incorporation of ASec into proteins and conversion to Sec through Pd-mediated cleavage. (a) Western analysis of
incorporation of ASec into EGFP in E. coli. MbPylRS(C313W/W382T) was 3 times more efficient than MbPylRS(C313W/W382S). (b) SDS-
PAGE analysis of affinity purified Trx with ASec incorporated. (c) ESI-MS spectrum of ASec-incorporated Trx, confirming the incorporation of
ASec at site 32. (d) ESI-MS spectrum of Trx after Pd-mediated cleavage confirming the conversion of ASec into Sec.
Analysis of the expressed Trx by ESI-MS confirmed
incorporation of ASec into Trx. An observed peak at 12 786
Da corresponds to Trx lacking the initiator Met and containing
ASec at site 32 (expected [M-Met+H]+ = 12 787 Da). The
observed peak at 12 746 Da corresponds to this protein with
the allyl group cleaved from ASec (expected [M-Met-allyl+H]+
= 12 747 Da) (Figure 2c). Similar ESI-MS results were also
obtained from Trx with ASec incorporated at site 73 (Figure
S6a). No peaks were observed corresponding to Trx
containing a natural amino acid at the TAG site, indicating
the incorporation of ASec into Trx proteins in high specificity.
Conversion of ASec into Sec in Proteins. To fully
convert the incorporated ASec into Sec, we explored using
palladium to cleave the allyl protection group on ASec in
proteins. To facilitate the detection of allyl group cleavage, we
generated a Ub-GyrA fusion construct that bears a TAG codon
at the first Cys position of Mycobacterium xenopi GyrA intein.52
The regeneration of a Cys residue at this site would result in an
active GyrA intein, which would cleave Ub from Ub-GyrA for
ready detection by SDS-PAGE or Western blot (Figure S7a).
We first incorporated ACys at the TAG site, and the intein
activity was blocked by ACys, as shown by no cleavage of Ub
from Ub-GyrA (Figure S7b). For Pd catalyst selection,
[PdCl(allyl)]2 is stable and compatible for use in vivo,53 and
this Pd2+ compound can be readily reduced to Pd(0) upon
nucleophilic attack.53,54 Pd(TPPTS)4 is a water-soluble Pd(0)
species, suitable for our needs for deprotection under native or
physiological conditions.55 We also include other Pd2+ species
for comparison. The results showed that only Pd(0) could
cleave the allyl group from ACys incorporated in protein Ub-
GyrA, with Pd(TPPTS)4 providing the highest efficiency
(>80%) (Figures S7c and S8). The preference of Pd(0) for
allyl removal was also observed for ASec incorporated into
proteins: both Pd(TPPTS)4 and [PdCl(allyl)]2 were able to
cleave ASec incorporated in the MBP-Z protein (Figure S9).
We next treated the ASec incorporated Trx with Pd-
(TPPTS)4, and analyzed the treated protein sample with ESI-
MS (Figure 2d). A peak at 12 746 Da was observed,
corresponding to the conversion of ASec into Sec in >93%
efficiency (expected [M-Met-allyl+H]+ = 12 747 Da). In some
cases, we observed that a small amount of ASec incorporated
Trx already had the ASec converted into Sec after purification
and without Pd treatment, while ACys incorporated Trx
retained its ACys intact (Figures 2C, S10, and S11). This
difference could be due to the fact that the C−Se bond is
weaker than the C−S bond and thus may be more prone to
cleavage when incubation with nickel or cobalt during affinity
purification.56 These results indicate that the incorporated
ASec could be converted into Sec in high efficiency under mild
conditions compatible with proteins, and our method is flexible
with Sec incorporation site.
Preparation of Native Human Selenoprotein GPx1.
Our method uses the TAG amber codon to specify the Sec
incorporation site and obviates the aid of Sec-dedicated SelB,
SelA, and SECIS elements. This straightforward incorporation
mechanism similar to that for canonical amino acids should
allow us to incorporate Sec at any desired site of proteins,
which will be especially useful for recombinantly expressing
mammalian selenoproteins that have internal Sec residues in E.
coli. To further demonstrate the application of our method in
terms of preparation and characterization of native selenopro-
teins, we attempted the production of the human selenoprotein
glutathione peroxidase 1 (GPx1). GPx1 is a major antioxidant
enzyme to combat reactive oxygen species and has
cardiovascular protective function.57,58 It has an essential
Sec49 residue at the active site, the substitution of which with
Cys results in 1000-fold activity loss and with Ser leads to
inactivation. Currently there is no human GPx1 with Sec
incorporated available commercially.
D
DOI: 10.1021/jacs.8b04603
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 3. Preparation of human native selenoprotein GPx1 with high Sec specificity and enzyme activity. (a) Structure of human GPx1 (PDB
code:2F8A) with an orange sphere showing the position of Sec49. (b) SDS-PAGE analysis of purified GPx1 with ASec incorporated. (c) ESI-MS
spectrum of intact GPx1 confirming ASec incorporation in high fidelity. (d) Tandem MS spectrum of trypsin digested GPx1 confirming ASec
incorporation at site 49 specified by the amber codon. U represents ASec in the peptide sequence. The fragment ions that contain ASec are denoted
in red color. (e) ESI-MS spectrum of GPx1 after Pd-mediated cleavage of ASec, confirming conversion of ASec to Sec. Peaks denoted by a star have
mass corresponding to elimination of H2O from the two main peaks. (f) GPx1 glutathione peroxidase activity measurements showing high activity
for the GPx1(Sec49) prepared using our method. The activity was monitored by the absorbance decrease at 340 nm that reflects the consumption
of NADPH, the cofactor of glutathione reductase.
Using the orthogonal tRNAPyl/ASecRS pair, we expressed
GPx1 with ASec incorporated at site 49 in E. coli in the yield of
3 mg/L (Figure 3a, b). The purified GPx1 intact protein was
analyzed by ESI-MS (Figure 3c). The observed peak at 21 671
Da corresponds to GPx1 containing ASec49 (expected [M +
H]+ = 21 671 Da), and the peak at 21540 Da corresponds to
GPx1 lacking the initiator Met and containing ASec49
(expected [M-Met+H]+ = 21539 Da). No peaks corresponding
to GPx1 containing any natural amino acids at site 49 were
observed. In addition, we also analyzed GPx1 using trypsin
digestion followed by tandem MS (Figure 3d). The precursor
ion for the ASec-containing peptide was identified in high
accuracy (expected M/Z = 888.46, observed 888.47), and a
series of b and y fragment ions of this peptide clearly indicate
that ASec was unambiguously incorporated at site 49 specified
by the amber codon. Moreover, the isotopic pattern of the
ASec-containing peptide was also analyzed, which was
consistent with the expected isotopic distribution of selenium
(Figure S12), further confirming ASec incorporation. Alto-
gether these data indicated that ASec was incorporated into
GPx1 at the amber codon specified position in high fidelity
without contamination of other amino acids.
E
DOI: 10.1021/jacs.8b04603
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 4. Selective labeling of incorporated Sec on live E. coli cell surface in the presence of endogenous Cys. (a) Scheme showing the procedures
for Sec incorporation and labeling on E. coli cell surface. (b) E. coli cells expressing Sec-incorporated outer membrane protein was selectively labeled
by the Alexa Fluor 488 dye at pH 5.0.
Figure 5. Genetically incorporated ASec enables photoinduced GST dimerization in living cells and site-specific protein labeling through IEDDA.
(a) Western blot analysis of GST dimerization through thiol−ene reaction upon UV illumination of living cells. (b) Purified MBP-Z with ASec
incorporated was fluorescently labeled with Cy3 tetrazine. Top: Coomassie stained SDS-PAGE gel; bottom: fluorescence image of the same gel
before Coomassie blue staining.
After palladium treatment of GPx1, ASec was almost fully
converted into Sec as shown by ESI-MS analysis (Figure 3e).
The observed peak at 21630 Da corresponds to GPx1
containing Sec49 (expected [M + H]+ = 21630 Da), and the
peak at 21498 Da corresponds to GPx1 lacking the initiator
Met and containing Sec49 (expected [M-Met+H]+ = 21498
Da). No peak corresponding to ASec-GPx1 was detected. The
peroxidase activity of resultant GPx1 was measured using a
glutathione/glutathione reductase coupled assay, and a high
activity of 25 Units (Figure 3f) was obtained (1 unit was
defined as per μmol of NADPH oxidized per min per mg of
enzyme). The enzymatic activity measured compares favorably
with what have been reported in the literature (Table
S2).30,32,36 These results indicate that our method could
produce native selenoprotein with high Sec incorporation
specificity and enzyme activity.
used because of the high reaction rate and readily available
reagents, but it lacks site specificity when multiple cysteine
exists.59 For Sec, one of the most appealing applications is
selective labeling in the presence of all natural amino acids
including Cys, which obviates the mutation of endogenous Cys
residues and affords the shortest linkage placing the resultant
modification close to the protein backbone. Due to the large
pKa difference between Cys (pKa 8.3) and Sec (pKa 5.2), it is
possible to poise the pH to selectively label Sec only with Cys
residues unaffected on a single selenoprotein in vitro.5,60 Since
our method can genetically incorporate ASec into proteins in
live cells, we explored if Sec could be converted from ASec in
live cells and if the resultant Sec could be selectively labeled on
the surface of live cells.
We incorporated ASec into an outer membrane protein
eCPX (enhanced circularly permuted outer membrane protein
OmpX) in E. coli to display the ASec residue on the cell
surface,61 treated the cells with Pd(TPPTS)4 to generate Sec,
Selective Labeling of Sec on Live E. coli Cell Surface.
Protein labeling through cysteine thiol alkylation is frequently
F
DOI: 10.1021/jacs.8b04603
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 6. Genetic incorporation of ASec and its deprotection in mammalian cells. (a) Confocal fluorescence image to detect the incorporation of
ASec into EGFP-182TAG in HeLa cells. (b) FACS analyses of ASec incorporation into EGFP-182TAG in HeLa cells under conditions indicated
above. The usage of ASec above 0.2 mM led to increased toxicity. Adding Cys rescued the toxicity, and thus increased incorporation efficiency. The
data point of 0.2 mM ASec alone after 48 h was not collected due to ASec toxicity. (c) Western blot analysis of HeLa reporter cell lysate to detect
ASec incorporation into EGFP. (d) Fluorescence gel image of various EGFP samples labeled with Cy3 maleimide. Deprotection of ASec into Sec in
EGFP in HeLa cells led to Cy3 labeling of Sec.
and then incubated cells with a membrane impermeable, Cys-
reactive dye Alexa Fluor 488 C5 maleimide (Figure 4a). At pH
7.4, control cells without ASec incorporation were all
fluorescently labeled through the dye reacting with endoge-
nous Cys at the cell surface; as expected, the labeling was
abolished when pH was decreased to 5.0. In contrast, cells with
ASec incorporation into eCPX followed by Pd deprotection
were clearly labeled by the dye at pH 5.0, indicating the
selective labeling of Sec on live E. coli cell surface in the
presence of excess thiols (Figure 4b).
Protein Cross-Linking and Labeling via the Terminal
Alkene of ASec. Besides the seleno group, ASec also affords a
biorthogonal alkene functional group. The side chain length
and size of ASec are comparable to those of canonical amino
acids, which will minimize the potential perturbation to
proteins. Labeling or reaction through this terminal alkene will
also be valuable for maintaining the modification close to the
protein backbone. We therefore explored biorthogonal usage
of this alkene group as well.
Photocatalytic thiol−ene reaction has been widely employed
for protein labeling and bioconjugations.62 To the best of our
knowledge, it has not be applied for protein−protein cross-
linking. Here we genetically encoded ASec into the 103rd
position of the E. coli glutathione transferase (GST), and
mutated the proximal Lys107 at the dimer interface into Cys.
After expressing ASec-containing GST in E. coli cells, the cells
were illuminated with 302 nm UV light. Western blot analysis
of cell lysate showed that GST dimerization was strongly
induced after 30 min (Figure 5a). Through photocontrolled
thiol−ene reaction, genetically encoded ASec thus may provide
a valuable tool for protein cross-linking.
In addition to the thiol−ene reaction, a terminal alkene can
also be utilized through inverse electron demand Diels−Alder
reaction (IEDDA).63−65 Since ASec can selectively introduce a
terminal alkene onto the surface of a protein, we tested
whether ASec would be compatible with IEDDA for protein
labeling. We site specifically incorporated ASec into the MBP-
Z fusion protein, and treated the purified protein with or
G
DOI: 10.1021/jacs.8b04603
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
without Cy3 tetrazine derivative. The MBP-Z protein was
fluorescently labeled by the Cy3 dye, and the labeling was
stable and resistant to reduction by DTT (Figure 5b). These
results demonstrate that the genetically incorporated ASec
could be used as a biorthogonal reactive handle for labeling
through IEDDA.
Sec analog, our method affords high specificity for Sec
incorporation, as shown in the preparation of the native
selenoprotein GPx1, and enables ASec/Sec incorporation into
proteins in mammalian cells.
To achieve protein translation in high fidelity, our approach
features a “chemical protection−deprotection” strategy to solve
the challenge of distinguishing Sec from the structural similar
Cys. Nature solves this challenge by biosynthesizing Sec on the
tRNA without evolving an aminoacyl-tRNA synthetase specific
for Sec. Our chemical protection−deprotection strategy
represents a facile alternative that we expect can be generally
useful for the incorporation of various intractable amino acids
due to their structural, physical, chemical, or biological
properties. For instance, we recently devised a similar strategy
to genetically incorporate phosphotyrosine, which is cell-
impermeable and chemically labile inside cells.69 In addition, in
solid phase peptide synthesis and Sec-mediated multiple steps
of challenging protein ligations, Sec needs to be protected and
regenerated using p-methoxybenzyl, 2,2′-dithiobis(5-nitropyr-
idine), or selenazolidine groups.70−73 The facile conversion of
ASec to Sec shown here may provide an attractive alternative
route to such applications.
Another obstacle we overcame in this work is circumventing
the toxicity of the target Uaa. It has been generally considered
infeasible to genetically encode Uaas that are toxic to cells.74
By devising a strategy to inhibit the toxicity, we showed here
the successful incorporation of the toxic ASec, which may
guide future efforts in incorporating other toxic yet useful
Uaas. As a starting point, we first demonstrated that it is
feasible to incorporate a Sec-derived Uaa into proteins in
mammalian cells followed by in cellulo deprotection to
generate Sec. Therefore, our method has the potential to be
compatible for use in both prokaryotic and eukaryotic cells. We
expect that the use of tRNAPyl/ASecRS in eukaryotic cells will
serve as a valuable tool for studying and charactering the native
selenoproteins with unknown biological functions.
Genetic Incorporation of ASec into Proteins in
Mammalian Cells. Sec and Sec-derived Uaas have not been
incorporated into proteins in mammalian cells using
engineered tRNA/synthetase pairs. As the tRNAPyl/ASecRS
was derived from the tRNAPyl/PylRS, which is orthogonal in
prokaryotic and eukaryotic cells,50,66 we tested the incorpo-
ration of ASec into proteins in mammalian cells using the
tRNAPyl/ASecRS pair. We transfected HeLa-EGFP-182TAG
reporter cells49 with plasmid pMP-3xPyltRNA-ASecRS, which
expresses ASecRS and 3 copies of tRNAPyl genes. Suppression
of the 182TAG codon of the genome-integrated EGFP gene in
the reporter cells would produce full-length EGFP rendering
cells fluorescent. Using fluorescence confocal microscopy, we
observed strong green fluorescence from cells cultured in the
presence of ASec only (Figure 6a). FACS analysis of cells
confirmed ASec-dependent cellular fluorescence, and showed
that more cells became green fluorescent with ASec incubation
time (Figure 6b). ASec could be used in the final concentration
of up to 0.2 mM, and coadding 3.2 mM of Cys further
ameliorated cellular toxicity of ASec. The successful incorpo-
ration of ASec into EGFP in HeLa cells was also demonstrated
by Western blot analysis of reporter cell lysates with an anti-
GFP antibody (Figure 6c): Full-length EGFP was produced
only when ASec was added to the growth media.
We next applied [PdCl(allyl)]2 to these cells for in vivo
deprotection of ASec. [PdCl(allyl)]2 has been shown as an
effective Pd cleavage reagent with little cellular toxicity.53,67
Similarly, we found that the EGFP expression level was not
affected by [PdCl(allyl)]2 treatment of the HeLa-EGFP-
182TAG reporter cells (Figure 6c). To investigate the cleavage
of ASec into Sec in mammalian cells, we utilized Sec’s unique
property of selective labeling at pH 5 due to its lower pKa and
high nucleophilicity.5,41 ASec-EGFP expressing HeLa cells
were treated with [PdCl(allyl)]2, and EGFP was trapped by an
EGFP specific antibody and then labeled with a Cy3 maleimide
through Se-Michael addition at pH 5. The samples were
separated by SDS-PAGE and imaged for Cy3 fluorescence. As
a negative control, Uaa p-azido-L-phenylalanine (AzF) was
incorporated at the same site into EGFP by transfecting the
HeLa-EGFP-182TAG reporter cells with plasmid pIRE-Azi3
and incubating with 1 mM of Azi,68 and the sample was
similarly treated. As shown in Figure 6d, Cy3 fluorescence was
detected only on the EGFP with ASec incorporated and
treated with [PdCl(allyl)]2, suggesting the successful depro-
tection of ASec into Sec in mammalian cells.
We further demonstrated the selective labeling of the
incorporated Sec in the presence of Cys on live E. coli cell
surface. With this applicability, there is no need to mutate
endogenous Cys residues of the protein to be labeled, which is
especially useful for proteins containing essential Cys residues
or disulfide bonds. A unique advantage of labeling through the
seleno group of Sec is that it results in the shortest possible
linkage (similar to Cys) placing the label closer to the protein
backbone, which is valuable in applications such as the
attachment of biophysical probes (e.g., spin labels for electron
̈
paramagnetic resonance and fluorescent labels for Forster
resonance energy transfer) for measurements in high accuracy.
In addition, no catalyst is needed for seleno-mediated labeling,
and many probes compatible for thiol and seleno labeling are
readily available commercially. However, for applications in
which linkage length is more appropriate on the scale of other
canonical amino acids such as Met and Lys, the site-specifically
incorporated ASec also affords a bioorthogonal alkene handle
for selective protein labeling and bioconjugations through
metathesis,75 thiol−ene reaction,76 or oxidative heck cou-
pling.77 We have demonstrated the use of the terminal alkene
of ASec for photoinduced GST dimerization through thiol−
ene reaction and for site-specific protein labeling through
IEDDA reaction.
DISCUSSION
■
We have developed a new general method to prepare
selenoproteins with Sec site-selectively introduced in high
specificity. Through genetic incorporation of the allyl protected
ASec in response to the amber TAG codon followed by facile
palladium-mediated cleavage to generate Sec, our method
obviates Sec-dedicated factors completely and thus generally
enables the expression in E. coli of various selenoproteins
regardless of the Sec position and source species. In
comparison with existing methods using either tRNASec
mutants compatible with the common EF-Tu or a photocaged
In summary, the generality and robustness of this method to
prepare native selenoproteins would enable the study of
diverse selenoproteins currently elusive to procure; and facile
H
DOI: 10.1021/jacs.8b04603
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(17) Metanis, N.; Keinan, E.; Dawson, P. E. Angew. Chem., Int. Ed.
2010, 49, 7049.
(18) Wang, Y. J.; Szantai-Kis, D. M.; Petersson, E. J. Org. Biomol.
Chem. 2016, 14, 6262.
(19) Kulkarni, S. S.; Sayers, J.; Premdjee, B.; Payne, R. J. Rapid and
efficient protein synthesis through expansion of the native chemical
ligation concept. Nat. Rev. Chem. 2018, 2.
introduction of Sec into non-Sec-containing proteins in high
specificity would dramatically expand the scope and versatility
of chemical biology means for general protein research and
exploitation.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.8b04603.
■
S
̈
(20) Bock, A.; Forchhammer, K.; Heider, J.; Leinfelder, W.; Sawers,
G.; Veprek, B.; Zinoni, F. Mol. Microbiol. 1991, 5, 515.
(21) Driscoll, D. M.; Copeland, P. R. Annu. Rev. Nutr. 2003, 23, 17.
̈
(22) Ambrogelly, A.; Palioura, S.; Soll, D. Nat. Chem. Biol. 2007, 3,
29.
(23) Bulteau, A. L.; Chavatte, L. Antioxid. Redox Signaling 2015, 23,
775.
Experimental details for chemical synthesis, library
construction, synthetase screening, ASec incorporation
and deprotection, protein expression, purification and
characterization, Sec labeling, enzyme assay, and mass
spectrometry (Figures S1−S14) (PDF)
(24) Gonzalez-Flores, J. N.; Shetty, S. P.; Dubey, A.; Copeland, P. R.
Biomol. Concepts 2013, 4, 349.
(25) Quaderer, R.; Sewing, A.; Hilvert, D. Helv. Chim. Acta 2001, 84,
1197.
(26) Hondal, R. J.; Nilsson, B. L.; Raines, R. T. J. Am. Chem. Soc.
2001, 123, 5140.
(27) Gieselman, M. D.; Xie, L.; van der Donk, W. A. Org. Lett. 2001,
3, 1331.
(28) Thyer, R.; Filipovska, A.; Rackham, O. J. Am. Chem. Soc. 2013,
135, 2.
(29) Brocker, M. J.; Ho, J. M.; Church, G. M.; Soll, D.;
O’Donoghue, P. Angew. Chem., Int. Ed. 2014, 53, 319.
(30) Cheng, Q.; Arner, E. S. J. Biol. Chem. 2017, 292, 5476.
(31) Arner, E. S. Methods Enzymol. 2002, 347, 226.
(32) Aldag, C.; Brocker, M. J.; Hohn, M. J.; Prat, L.; Hammond, G.;
Plummer, A.; Soll, D. Angew. Chem., Int. Ed. 2013, 52, 1441.
(33) Haruna, K.; Alkazemi, M. H.; Liu, Y.; Soll, D.; Englert, M.
Nucleic Acids Res. 2014, 42, 9976.
(34) Miller, C.; Brocker, M. J.; Prat, L.; Ip, K.; Chirathivat, N.;
Feiock, A.; Veszpremi, M.; Soll, D. FEBS Lett. 2015, 589, 2194.
(35) Thyer, R.; Robotham, S. A.; Brodbelt, J. S.; Ellington, A. D. J.
Am. Chem. Soc. 2015, 137, 46.
(36) Fan, Z.; Song, J.; Guan, T.; Lv, X.; Wei, J. ACS Synth. Biol.
2018, 7, 249.
(37) Mukai, T.; Sevostyanova, A.; Suzuki, T.; Fu, X.; Soll, D. Angew.
Chem., Int. Ed. 2018, 57, 7215.
(38) Wang, L.; Brock, A.; Herberich, B.; Schultz, P. G. Science 2001,
292, 498.
(39) Wang, L. Acc. Chem. Res. 2017, 50, 2767.
(40) Kang, J. Y.; Kawaguchi, D.; Coin, I.; Xiang, Z.; O’Leary, D. D.;
Slesinger, P. A.; Wang, L. Neuron 2013, 80, 358.
(41) Rakauskaite, R.; Urbanaviciute, G.; Ruksenaite, A.;
Liutkeviciute, Z.; Juskenas, R.; Masevicius, V.; Klimasauskas, S.
Chem. Commun. (Cambridge, U. K.) 2015, 51, 8245.
(42) Muller, S.; Senn, H.; Gsell, B.; Vetter, W.; Baron, C.; Bock, A.
Biochemistry 1994, 33, 3404.
(43) Maity, S. K.; Jbara, M.; Laps, S.; Brik, A. Angew. Chem., Int. Ed.
2016, 55, 8108.
(44) Exner, M. P.; Kuenzl, T.; To, T. M.; Ouyang, Z.; Schwagerus,
S.; Hoesl, M. G.; Hackenberger, C. P.; Lensen, M. C.; Panke, S.;
Budisa, N. ChemBioChem 2017, 18, 85.
(45) Jbara, M.; Maity, S. K.; Seenaiah, M.; Brik, A. J. Am. Chem. Soc.
2016, 138, 5069.
(46) Jbara, M.; Maity, S. K.; Brik, A. Angew. Chem., Int. Ed. 2017, 56,
10644.
(47) Maity, S. K.; Mann, G.; Jbara, M.; Laps, S.; Kamnesky, G.; Brik,
A. Org. Lett. 2016, 18, 3026.
AUTHOR INFORMATION
Corresponding Author
*Lei.wang2@ucsf.edu
■
ORCID
Sharon Rozovsky: 0000-0003-4902-0777
Lei Wang: 0000-0002-5859-2526
Author Contributions
⊥These two authors contributed equally to this work.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Dr. Hyunil Jo for the helpful discussion and
purification of ASec compound at early stage, Dr. Hai Huang
and Dr. Qiang Zhang for help with the laser confocal imaging.
S.R. acknowledges the support of NSF (MCB-1616178); and
L.W. acknowledges the support of the NIH (R01GM118384,
RF1MH114079).
REFERENCES
■
(1) Hatfield, D. L.; Tsuji, P. A.; Carlson, B. A.; Gladyshev, V. N.
Trends Biochem. Sci. 2014, 39, 112.
(2) Liu, J.; Rozovsky, S. Antioxid. Redox Signaling 2015, 23, 795.
(3) Reich, H. J.; Hondal, R. J. ACS Chem. Biol. 2016, 11, 821.
(4) Metanis, N.; Hilvert, D. Curr. Opin. Chem. Biol. 2014, 22, 27.
(5) Liu, J.; Chen, Q.; Rozovsky, S. J. Am. Chem. Soc. 2017, 139,
3430.
(6) Mousa, R.; Notis Dardashti, R.; Metanis, N. Angew. Chem., Int.
Ed. 2017, 56, 15818.
(7) Malins, L. R.; Mitchell, N. J.; Payne, R. J. J. Pept. Sci. 2014, 20,
64.
(8) Arner, E. S. Exp. Cell Res. 2010, 316, 1296.
(9) Malins, L. R.; Mitchell, N. J.; McGowan, S.; Payne, R. J. Angew.
Chem., Int. Ed. 2015, 54, 12716.
(10) Mitchell, N. J.; Malins, L. R.; Liu, X.; Thompson, R. E.; Chan,
B.; Radom, L.; Payne, R. J. J. Am. Chem. Soc. 2015, 137, 14011.
́
́
(11) Boutureira, O.; Bernardes, G. J. L.; Fernandez-Gonzalez, M.;
Anthony, D. C.; Davis, B. G. Angew. Chem., Int. Ed. 2012, 51, 1432.
(12) Wang, J.; Schiller, S. M.; Schultz, P. G. Angew. Chem., Int. Ed.
2007, 46, 6849.
(48) Bhuyan, B. J.; Mugesh, G. Org. Biomol. Chem. 2011, 9, 1356.
(49) Wang, W.; Takimoto, J. K.; Louie, G. V.; Baiga, T. J.; Noel, J.
P.; Lee, K. F.; Slesinger, P. A.; Wang, L. Nat. Neurosci. 2007, 10, 1063.
(50) Takimoto, J. K.; Dellas, N.; Noel, J. P.; Wang, L. ACS Chem.
Biol. 2011, 6, 733.
(13) Beld, J.; Woycechowsky, K. J.; Hilvert, D. J. Biotechnol. 2010,
150, 481.
(14) Metanis, N.; Hilvert, D. Chem. Sci. 2015, 6, 322.
(15) Dery, S.; Reddy, P. S.; Dery, L.; Mousa, R.; Dardashti, R. N.;
Metanis, N. Chem. Sci. 2015, 6, 6207.
(51) Commandeur, J. N.; Andreadou, I.; Rooseboom, M.; Out, M.;
de Leur, L. J.; Groot, E.; Vermeulen, N. P. J. Pharmacol Exp Ther
2000, 294, 753.
(16) Hondal, R. J.; Marino, S. M.; Gladyshev, V. N. Selenocysteine
in Thiol/Disulfide-Like Exchange Reactions. Antioxid Redox Signal
2012.
I
DOI: 10.1021/jacs.8b04603
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(52) Evans, T. C., Jr.; Benner, J.; Xu, M. Q. Protein Sci. 1998, 7,
2256.
(53) Li, J.; Yu, J.; Zhao, J.; Wang, J.; Zheng, S.; Lin, S.; Chen, L.;
Yang, M.; Jia, S.; Zhang, X.; Chen, P. R. Nat. Chem. 2014, 6, 352.
(54) Wei, C. S.; Davies, G. H.; Soltani, O.; Albrecht, J.; Gao, Q.;
Pathirana, C.; Hsiao, Y.; Tummala, S.; Eastgate, M. D. Angew. Chem.,
Int. Ed. 2013, 52, 5822.
(55) Amatore, C.; Blart, E.; Genet, J. P.; Jutand, A.; Lemaire-
Audoire, S.; Savignac, M. J. Org. Chem. 1995, 60, 6829.
(56) Cooper, A. J.; Krasnikov, B. F.; Niatsetskaya, Z. V.; Pinto, J. T.;
Callery, P. S.; Villar, M. T.; Artigues, A.; Bruschi, S. A. Amino Acids
2011, 41, 7.
(57) Brigelius-Flohe, R.; Maiorino, M. Biochim. Biophys. Acta, Gen.
Subj. 2013, 1830, 3289.
(58) Blankenberg, S.; Rupprecht, H. J.; Bickel, C.; Torzewski, M.;
Hafner, G.; Tiret, L.; Smieja, M.; Cambien, F.; Meyer, J.; Lackner, K.
J.; AtheroGene, I. N. Engl. J. Med. 2003, 349, 1605.
(59) Gunnoo, S. B.; Madder, A. ChemBioChem 2016, 17, 529.
(60) Hofer, T.; Thomas, J. D.; Burke, T. R., Jr.; Rader, C. Proc. Natl.
Acad. Sci. U. S. A. 2008, 105, 12451.
(61) Rice, J. J.; Daugherty, P. S. Protein Eng., Des. Sel. 2008, 21, 435.
(62) Hoyle, C. E.; Bowman, C. N. Angew. Chem., Int. Ed. 2010, 49,
1540.
(63) Oliveira, B. L.; Guo, Z.; Bernardes, G. J. L. Chem. Soc. Rev.
2017, 46, 4895.
(64) Oliveira, B. L.; Guo, Z.; Boutureira, O.; Guerreiro, A.; Jimenez-
Oses, G.; Bernardes, G. J. Angew. Chem., Int. Ed. 2016, 55, 14683.
(65) Niederwieser, A.; Spate, A. K.; Nguyen, L. D.; Jungst, C.;
Reutter, W.; Wittmann, V. Angew. Chem., Int. Ed. 2013, 52, 4265.
(66) Blight, S. K.; Larue, R. C.; Mahapatra, A.; Longstaff, D. G.;
Chang, E.; Zhao, G.; Kang, P. T.; Green-Church, K. B.; Chan, M. K.;
Krzycki, J. A. Nature 2004, 431, 333.
(67) Wang, J.; Zheng, S.; Liu, Y.; Zhang, Z.; Lin, Z.; Li, J.; Zhang, G.;
Wang, X.; Li, J.; Chen, P. R. J. Am. Chem. Soc. 2016, 138, 15118.
(68) Coin, I.; Katritch, V.; Sun, T.; Xiang, Z.; Siu, F. Y.; Beyermann,
M.; Stevens, R. C.; Wang, L. Cell 2013, 155, 1258.
(69) Hoppmann, C.; Wong, A.; Yang, B.; Li, S.; Hunter, T.; Shokat,
K. M.; Wang, L. Nat. Chem. Biol. 2017, 13, 842.
(70) Harris, K. M.; Flemer, S., Jr.; Hondal, R. J. J. Pept. Sci. 2007, 13,
81.
(71) Reddy, P. S.; Dery, S.; Metanis, N. Angew. Chem., Int. Ed. 2016,
55, 992.
(72) Mitchell, N. J.; Sayers, J.; Kulkarni, S. S.; Clayton, D.; Goldys,
A. M.; Ripoll-Rozada, J.; Barbosa Pereira, P. J.; Chan, B.; Radom, L.;
Payne, R. J. Chem. 2017, 2, 703.
(73) Schroll, A. L.; Hondal, R. J.; Flemer, S., Jr. J. Pept. Sci. 2012, 18,
155.
(74) Wang, L.; Schultz, P. G. Angew. Chem., Int. Ed. 2005, 44, 34.
(75) Lin, Y. Y. A.; Boutureira, O.; Lercher, L.; Bhushan, B.; Paton, R.
S.; Davis, B. G. J. Am. Chem. Soc. 2013, 135, 12156.
(76) Hoogenboom, R. Angew. Chem., Int. Ed. 2010, 49, 3415.
(77) Ourailidou, M. E.; van der Meer, J.-Y.; Baas, B.-J.; Jeronimus-
Stratingh, M.; Gottumukkala, A. L.; Poelarends, G. J.; Minnaard, A. J.;
Dekker, F. J. ChemBioChem 2014, 15, 209.
J
DOI: 10.1021/jacs.8b04603
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX