Aldehydes can switch the chemoselectivity of electrophiles in protein labeling

Article abstract of DOI:10.1039/c8ob02897d

We show that the chemoselectivity of an electrophile in protein labeling can be promiscuous. An aldehyde enables switching of chemoselectivity of an epoxide and a sulfonate ester along with an enhanced rate of reaction. The chemical technology renders single-site installation of diverse probes on a protein and delivers analytically pure tagged proteins.

Full text of DOI:10.1039/c8ob02897d

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
COMMUNICATION
Aldehydes can switch the chemoselectivity of
electrophiles in protein labeling†
Cite this: Org. Biomol. Chem., 2018,
16, 9377
Srinivasa Rao Adusumalli,
Dattatraya Gautam Rawale
and Vishal Rai
*
Received 20th November 2018,
Accepted 28th November 2018
DOI: 10.1039/c8ob02897d
rsc.li/obc
We show that the chemoselectivity of an electrophile in protein resemble the chemistry of small molecules. However, the low
labeling can be promiscuous. An aldehyde enables switching of residence time of the ligand and the heterogeneity in the dis-
chemoselectivity of an epoxide and a sulfonate ester along with an tribution of the ligand–electrophile conjugate on the surface
enhanced rate of reaction. The chemical technology renders of proteins are inevitable. It might explain the lack of site-
single-site installation of diverse probes on a protein and delivers selectivity exhibited by an electrophile in bioconjugation. At
analytically pure tagged proteins.
the same time, these ligands can alter the chemoselectivity of
an electrophile.6
The application of well-established organic reactions to pro-
teins faces severe challenges associated with the reactivity and
selectivity. Frequently, the results of protein bioconjugation
with electrophiles are challenging to reproduce. It indicates
that our understanding of the reaction parameters in protein
bioconjugation needs further efforts. The preferential reactivity
of an electrophile with one nucleophile over the other defines
its chemoselectivity.1 In general, this attribute of an electro-
phile remains conserved for its structural derivatives within
the chemical reaction of small molecules. It inspires the use of
an electrophilic reagent as warheads for labeling proteins
through covalent bond formation. Also, it determines the
reagent design for irreversible inhibition of proteases.2 These
inhibitors include a broad range of ligands conjugated to a
specific electrophile. The same set of principles is also appli-
cable to activity-based protein profiling.3 It is noteworthy that
many examples of ligand–electrophile conjugates exhibit poor
reactivity and selectivity.
We wondered what would happen if a functional group
with defined reactivity substitutes the ligand exhibiting non-
covalent interactions. We envisaged that the hypothesis could
be examined with an electrophile–electrophile conjugate (E1–
E2, Fig. 1a). We can ask two important questions with this
model. (a) Can we enhance the residence time of E1 through a
kinetically preferred reversible transformation of E2? (b) Can
the subsequent latent nucleophile (Nu1, Fig. 1a) alter the
chemoselectivity of E1? In case the switching is exclusive, the
method can provide a reliable tool for regulating the chemo-
selectivity and enabling single-site labeling of the proteins.
We selected two established examples of chemoselective
electrophiles from the literature, epoxides and sulfonate
esters. In recent years, epoxides provided the electrophilic
warhead for enzyme inhibition, enzymatic activity profiling,
and developing functional probes.7 The ligand part of such
ligand–epoxide conjugates bears a variety of functional groups.
If considered carefully, it indicates that the choice of ligand
would have an impact on the bioconjugation. It might emerge
from the distribution of the electrophile on the surface of a
protein. In the best case scenario, a highly selective ligand–
protein interaction4 can lead to the high residence time5 (or
local concentration) of the ligand–electrophile conjugate at
one site of a protein. Here, the relative stoichiometry would
Organic and Bioconjugate Chemistry Laboratory, Department of Chemistry, Indian
Institute of Science Education and Research Bhopal, Bhauri, Bhopal Bypass Road,
Bhopal 462066, India. E-mail: vrai@iiserb.ac.in
†Electronic supplementary information (ESI) available: Additional results and
discussion, experimental procedures, characterization data and spectra. See
DOI: 10.1039/c8ob02897d
Fig. 1 (a) Can an electrophile exhibit promiscuous chemoselectivity?
(b) The reactivity preferences of the epoxide and sulfonate ester.
This journal is © The Royal Society of Chemistry 2018
Org. Biomol. Chem., 2018, 16, 9377–9381 | 9377

View Article Online
Communication
Organic & Biomolecular Chemistry
For example, a natural product (E64) primed with an epoxide tion 1). The sluggish reaction results in 10% labeling of His
delivered a potent, non-selective, calpain inhibitor.8
(H105) residue over a period of 24 h (Fig. S12, ESI†). As per
The ligand conjugated epoxides can also be customized to our hypothesis, we selected an aldehyde as the electrophile
track the activity of cysteine proteases.9 The appropriate design and conjugated it with an epoxide for the synthesis of reagent
of the ligand is essential to harness the desired selectivity. In a 2b. Unfortunately, 2b led transformation was still too slow for
mechanistically related methodology, a ligand was found to any practical utility (Fig. 2a, reaction 2 and Fig. S13, ESI†). The
increase the residence time of a conjugated electrophile in poor conversions rendered an insufficient amount of product
proximity-driven protein labeling.10 Using this approach, for the identification of the site of labeling. To solve the
Sames10 and Hamachi11 employed a ligand–epoxide conjugate problem, we examined a series of aldehydes (2c–2e, 2g) and a
for chemoselective and site-selective His modification. In such ketone (2f) (Fig. 2a, reactions 3a–3e and Fig. S14, ESI†). We
examples, Cys is inaccessible or blocked by another reagent identified aldehyde 2g for conjugation with the epoxide and
before the bioconjugation step as it shows preferential reactiv- further screening. As per our investigations,14 the vicinal
ity in comparison to His (Fig. 1b). The activity-based protein hydroxyl group plays a critical role in the stabilization of the
profiling (ABPP) also uses the reactivity bias of an electro- imine. Interestingly, the reagent 2h led to the formation of
phile.12 Cravatt established that the reactivity preference of a 62% mono-labeled RNase A 7a in overall 12 h (Fig. 2a, reaction
sulfonate ester based ABPP probe is Glu, Asp, Cys, Arg, and Tyr 4). The extension of the reaction time was avoided to prevent
respectively (Fig. 1b).13
bis-labeling. MS, α-chymotrypsin digestion, peptide mapping,
In this work, we demonstrate that an electrophile can and MS–MS determined the homogeneity of the labeled
exhibit promiscuous chemoselectivity when employed with protein 7a (Fig. S15, ESI†). The sequencing confirmed the
proteins. The chemoselectivity of the epoxide conjugated to an chemoselectivity switching with N-terminus Nα-NH2 as the
aldehyde is switched from His to result in labeling of the exclusive site of conjugation. The structure and the enzymatic
N-terminus α-amine (Nα-NH2). The absolute switching of activity of labeled protein 7a remain unperturbed (Fig. 2b and
chemoselectivity is remarkable and enables single-site labeling c and Fig. S28 and S29, ESI†).
of proteins. Later, we extended the concept to sulfonate esters
A few reports15–17 and our earlier investigations18 suggest
that prefer to react with carboxylate (Glu, Asp), thiolate (Cys), that Nα-NH2 could exhibit higher reactivity in preference to Nε-
guanidine (Arg), and phenol (Tyr). Again, the N-terminus Nα- NH2. This feature can allow the selective formation of carbino-
NH2 is the exclusive site of modification. The excellent selecti- lamine 8 with Nα-NH2 of RNase A 1a and aldehyde E2
vity for the single-site of the protein provides a platform for (Fig. 3).19 In turn, it can result in enhanced residence time of
installation of an affinity tag, NMR tag, and a fluorophore. The the electrophile E1 near that latent nucleophile Nu1 generated
overall protocol includes purification and isolation of analyti- preferentially at the N-terminus. The high relative concen-
cally pure tagged proteins.
tration of mono-labeled imine 6e in the reaction of aldehyde
Initially, we identified the epoxide as the target functional 2g and RNase A 1a supports the hypothesis (reaction 3e,
group that reacts preferably with His residues in the absence Fig. 2a). Next, the carbinolamine intermediate 8 can undergo
of accessible Cys.12 In a control experiment, we vortexed the dehydration to equilibrate with a Schiff base 11 and imidazoli-
model protein RNase A 1a with the epoxide 2a (Fig. 2a, reac- dinone 12 in a reversible reaction (Fig. 3, path a). The intra-
Fig. 2 (a). Chemical modification of RNase A with reagents (reaction 1) 2a, (reaction 2) 2b, (reaction 3) 2c–2g, and (reaction 4) 2h. Effect of labeling
on the structure of proteins, (b) circular dichroism (CD) spectra of RNase A (1a) and labeled RNase A (7a) and (c) comparison of enzymatic activity
between native RNase A (1a) and labeled RNase A (7a) at 300 nm. PB: phosphate buffer.
9378 | Org. Biomol. Chem., 2018, 16, 9377–9381
This journal is © The Royal Society of Chemistry 2018

View Article Online
Organic & Biomolecular Chemistry
Communication
Fig. 4 N-Terminus modification of RNase A. (a) Control experiment. (b)
Reactivity preference for the N-terminus α-amine.
Fig. 3 Potential regulators for chemoselectivity switching.
molecular attack of the lone pair of imine nitrogen on epoxide
(11) can also render amino alcohol 10 (Fig. 3, path a).20
However, the secondary amine of transient carbinolamine can
also trigger an irreversible intramolecular epoxide ring
opening reaction to generate intermediate 9 (Fig. 3, path b).
Subsequently, the lone pair migration from oxygen can
drive the facile C–N bond cleavage and equilibrate with the
amino aldehyde 10. In a control experiment, we vortexed
RNase A 1a with 2,4-dihydroxybenzaldehyde S8 and the
epoxide 2a (Fig. S30, ESI†). A considerable drop in the conver-
sion (<5%) strongly indicates that labeled RNase A 7a is
formed through an intramolecular pathway. The formation of
hemiaminals and imines and their reactions are sensitive to
the pH of the buffer.21 We designed a control experiment to
investigate the participation of intermediate 8. The carbinola-
mine is activated under acidic conditions to accelerate its de-
hydration to an imine (Fig. S31, ESI†). Interestingly, the extent
of protein labeling decreases drastically at pH 6.0 and
becomes negligible at pH 5.0. It supports the role of intermedi-
ate 8 through path b in the switching of chemoselectivity. We
argued that if the reactivity of the carbinolamine is enabling
rapid intramolecular epoxide opening, the length and geome-
try of the spacer between E1 and E2 could be critical.
Interestingly, the variation in the length of the linear spacer
(Fig. S15 and S16†) led to reduced conversion (Fig. S16, S17,
and S32, ESI†). These observations support the proposed intra-
molecular nucleophilic addition of the carbinolamine to the
epoxide (Fig. 3, path b).
mation was observed upon its treatment with the reagent 2i,
whereas the reaction of sulfonate ester 2j (25 equiv.) with apro-
tinin 1b results in mono-labeled product 14b (40% conv.,
Fig. 4b). The enzymatic digestion of 14b followed by peptide
mapping and MS–MS confirmed that the N-terminus was
modified site-selectively (Fig. S19, ESI†).
The exclusive labeling of Nα-NH2 by the epoxide and sulfo-
nate ester highlights the promiscuous electrophilicity of a bio-
conjugation reagent. The methodology installs a functionally
orthogonal group (E2, o-hydroxybenzaldehyde) site-selectively
at the N-terminus of a protein (15). It is primed for the late-
stage modification with O-hydroxylamine derivatives of the
desired tag. At first, we attached an affinity probe using 16 to
result in tagged RNase A 17 (Fig. 5, reaction 1). Next, we deriva-
tized the labeled protein 15 with a biophysical 19F-NMR probe
18 and fluorophore 20 through oxime formation (Fig. 5, reac-
tions 2 and 3). All the tagged proteins (17, 19 and 21) were
formed in good conversion (ESI-MS) (Fig. S20–S22, ESI†).
Next, we demonstrated the versatility of the o-hydroxyben-
zaldehyde label (E2) in the purification of labeled proteins
(label–purify–tag–purify, Fig. 6a–d). We initiated a sequence of
orthogonal steps by immobilizing the labeled protein 15 on a
hydrazide activated resin through hydrazone formation (23,
To investigate further, we selected the sulfonate ester as the
next electrophile. The vortexing of RNase A 1a with sulfonate
ester 2i results in no transformation in 48 h (Fig. 4a).
Interestingly, the o-hydroxybenzaldehyde conjugated sulfonate
ester 2j resulted in single-site modification, and the late-stage
oxime formation led to 14a (48%, Fig. 4b). We substantiated
the characterization of the product by MS, protein digestion,
peptide mapping, and sequencing of K1-F8 by MS–MS
(Fig. S18, ESI†). The late-stage modification is essential to
eliminate the analytical complications created by the presence
of an imine in the reaction mixture. Next, we applied this
method to modify aprotinin 1b. However, no product for- Fig. 5 Installation of tags through oxime formation.
This journal is © The Royal Society of Chemistry 2018
Org. Biomol. Chem., 2018, 16, 9377–9381 | 9379

View Article Online
Communication
Organic & Biomolecular Chemistry
Fig. 6 (a) Chemoselectivity switching: label–purify–tag–purify protocol delivers analytically pure single-site tagged proteins. (b) Deconvoluted
ESI-MS spectrum of biotin-tagged RNase A (17), (c) 19F NMR spectrum of fluorine-tagged RNase A (19). (d) Steady-state fluorescence spectrum of
fluorophore-tagged RNase A (21).
Fig. 6a). At first, this step allows the recovery of the unreacted ation delivers the single-site installation of the aldehyde func-
native protein 1a. Subsequently, the immobilized labeled tional group. We used this functionality in the installation of
protein is released with O-hydroxylamine and gives the oxime various tags through a bio-orthogonal transformation. The
product 24a (Fig. S23, ESI†).
methodology is integrated with the purification protocol to
The immobilized labeled protein 23 was released with the enable the recovery of unreacted native proteins along with the
derivatives of O-hydroxylamine (16, 18, 20) to render tagged isolation of analytically pure tagged proteins. The work also
proteins (17, 19, 21). The excess tagging reagent is removed by provides a note of caution for the extrapolation of chemo-
centrifugal spin concentration. The overall label–purify–tag– selectivity within the derivatives of an electrophile.
purify protocol results in analytically pure tagged proteins
(Fig. S24–S26, ESI†). The purified 19F NMR probe labeled
RNase A 19 shows a sharp signal at −62.65 ppm by 19F NMR
spectroscopy (Fig. 6c, Fig. S33, ESI†). The coumarin attached
Conflicts of interest
There are no conflicts to declare.
protein (21) shows the emission band at 430 nm (excitation at
333 nm, Fig. 6d, Fig. S34, ESI†).
Acknowledgements
Conclusions
This work is supported by the SERB, the DBT, the DAE, and
the IISER Bhopal. S. R. A. and D. G. R. are the recipients of a
research fellowship from the CSIR and the UGC.
N-Terminus protein labeling has always been dependent on
the electrophiles that exhibit chemoselectivity towards primary
amines. This work highlights the promiscuous electrophilicity
and opens up the avenues for single-site Nα-NH2 labeling by
electrophiles that are intrinsically biased for other nucleophilic
residues. The result establishes that a fast reversible reaction
can guide and enhance the rate of slow irreversible transform-
ation by multiple folds. The reactivity of the N-terminus
α-amine reconstitutes and alters the chemoselectivity of an
electrophile from a protein residue to Nα-NH2. We demonstrate
the principle through an epoxide and sulfonate ester tethered
to an aromatic o-hydroxyaldehyde. The chemical transform-
Notes and references
1 N. A. Afagh and A. K. Yudin, Angew. Chem., Int. Ed., 2010,
49, 262.
2 J. C. Powers, J. L. Asgian, O. D. Ekici and K. E. James,
Chem. Rev., 2002, 102, 4639.
3 L. I. Willems, H. S. Overkleeft and S. I. van Kasteren,
Bioconjugate Chem., 2014, 25, 1181.
9380 | Org. Biomol. Chem., 2018, 16, 9377–9381
This journal is © The Royal Society of Chemistry 2018

View Article Online
Organic & Biomolecular Chemistry
Communication
4 A. Dickson and S. D. Lotz, J. Phys. Chem. B, 2016, 120, 5377. 13 E. Weerapana, G. M. Simon and B. F. Cravatt, Nat. Chem.
5 H. Lu and P. J. Tonge, Curr. Opin. Chem. Biol., 2010, 14, 467. Biol., 2008, 4, 405.
6 H. Wang, Y. Koshi, D. Minato, H. Nonaka, S. Kiyonaka, 14 S. R. Adusumalli, D. G. Rawale, U. Singh, P. Tripathi,
Y. Mori, S. Tsukiji and I. Hamachi, J. Am. Chem. Soc., 2011,
133, 12220.
R. Paul, N. Kalra, R. K. Mishra, S. Shukla and V. Rai, J. Am.
Chem. Soc., 2018, 140, 15114.
7 D. Greenbaum, A. Baruch, L. Hayrapetian, Z. Darula, 15 D. Chen, M. M. Disotuar, X. Xiong, Y. Wang and D.
A. Burlingame, K. F. Medzihradszky and M. Bogyo, Mol.
Cell. Proteomics, 2002, 1, 60.
8 I. T. Schiefer, S. Tapadar, V. Litosh, M. Siklos, R. Scism,
G. T. Wijewickrama, E. P. Chandrasena, V. Sinha,
H.-C. Chou, Chem. Sci., 2017, 8, 2717.
16 A. O.-Y. Chan, C.-M. Ho, H.-C. Chong, Y.-C. Leung,
J.-S. Huang, M.-K. Wong and C.-M. Che, J. Am. Chem. Soc.,
2012, 134, 2589.
E. Tavassoli, M. Brunsteiner, M. Fa’, O. Arancio, P. Petukhov 17 X. Chen, K. Muthoosamy, A. Pfisterer, B. Neumann and
and G. R. J. Thatcher, J. Med. Chem., 2013, 56, 6054. T. Weil, Bioconjugate Chem., 2012, 23, 500.
9 D. Greenbaum, K. F. Medzihradszky, A. Burlingame and 18 R. Singudas, S. R. Adusumalli, P. N. Joshi and V. Rai, Chem.
M. Bogyo, Chem. Biol., 2000, 7, 569. Commun., 2015, 51, 473.
10 G. Chen, A. Heim, D. Riether, D. Yee, Y. Milgrom, 19 J. M. Gilmore, R. A. Scheck, A. P. Esser-Kahn,
M. A. Gawinowicz and D. Sames, J. Am. Chem. Soc., 2003,
125, 8130.
N. S. Joshi and M. B. Francis, Angew. Chem., Int. Ed., 2006,
45, 5307.
11 Y. Takaoka, H. Tsutsumi, N. Kasagi, E. Nakata and 20 M. Raj, H. Wu, S. L. Blosser, M. A. Vittoria and P. S. Arora,
I. Hamachi, J. Am. Chem. Soc., 2006, 128, 3273. J. Am. Chem. Soc., 2015, 137, 6932.
12 D. A. Shannon and E. Weerapana, Curr. Opin. Chem. Biol., 21 W. P. Jencks, Catalysis in Chemistry and Enzymology,
2015, 24, 18.
McGraw- Hill, New York, 1969.
This journal is © The Royal Society of Chemistry 2018
Org. Biomol. Chem., 2018, 16, 9377–9381 | 9381