A study of the reactivity of S(VI)-F containing warheads with nucleophilic amino-acid side chains under physiological conditions

Article abstract of DOI:10.1039/c7ob02028g

Sulfonyl fluorides (SFs) have recently emerged as a promising warhead for the targeted covalent modification of proteins. Despite numerous examples of the successful deployment of SFs as covalent probe compounds, a detailed exploration of the factors influencing the stability and reactivity of SFs has not yet appeared. In this work we present an extensive study on the influence of steric and electronic factors on the reactivity and stability of the SF and related S^VI-F groups. While SFs react rapidly with N-acetylcysteine, the resulting adducts were found to be unstable, rendering SFs inappropriate for the durable covalent inhibition of cysteine residues. In contrast, SFs afforded stable adducts with both N-acetyltyrosine and N-acetyllysine; furthermore, we show that the reactivity of arylsulfonyl fluorides towards these nucleophilic amino acids can be predictably modulated by adjusting the electronic properties of the warhead. These trends were largely conserved when the covalent reaction occurred within a protein binding pocket. We have also obtained a crystal structure depicting covalent modification of the catalytic lysine of a tyrosine kinase (FGFR1) by the ATP analog 5′-O-3-((fluorosulfonyl)benzoyl)adenosine (m-FSBA). Highly reactive warheads were demonstrated to be unstable with respect to hydrolysis in buffered aqueous solutions, indicating that warhead reactivity must be carefully tuned to provide optimal rates of protein modification. Our results demonstrate that the reactivity of SFs complements that of more commonly studied acrylamides, and we hope that this work spurs the rational design of novel SF-containing covalent probe compounds and inhibitors, particularly in cases where a suitably positioned cysteine residue is not present.

Full text of DOI:10.1039/c7ob02028g

View Article Online
View Journal
Organic &
Biomolecular
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: H. Mukherjee, J.
Debreczeni, J. Breed, S. Tentarelli, B. Aquila, J. Dowling, A. Whitty and N. P. Grimster, Org. Biomol.
Chem., 2017, DOI: 10.1039/C7OB02028G.
This is an Accepted Manuscript, which has been through the
Volume 14 Number
1 7 January 2016 Pages 1–372
Royal Society of Chemistry peer review process and has been
accepted for publication.
Organic &
Biomolecular
Chemistry
www.rsc.org/obc
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1477-0520
COMMUNICATION
Takeharu Haino et al.
Solvent-induced emission of organogels based on tris(phenylisoxazolyl)
benzene
rsc.li/obc

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 1 of 12
View Article Online
DOI: 10.1039/C7OB02028G
Journal Name
ARTICLE
A Study of the Reactivity of S(VI)-F Containing Warheads with
Nucleophilic Amino-Acid Side Chains Under Physiological
Conditions†
Received 00th January 20xx,
Accepted 00th January 20xx
DOI: 10.1039/x0xx00000x
www.rsc.org/
H. Mukherjee,a* J. Debreczeni,b J. Breed,b S. Tentarelli,a B. Aquila,a J. E. Dowling,a A. Whittyc and N.
P. Grimstera*
Sulfonyl fluorides (SFs) have recently emerged as a promising warhead for the targeted covalent modification of proteins.
Despite numerous examples of the successful deployment of SFs as covalent probe compounds, a detailed exploration of
the factors influencing the stability and reactivity of SFs has not yet appeared. In this work we present an extensive study
on the influence of steric and electronic factors on the reactivity and stability of the SF and related SVI-F groups. While SFs
react rapidly with N-acetylcysteine, the resulting adducts were found to be unstable, rendering SFs inappropriate for the
durable covalent inhibition of cysteine residues. In contrast, SFs afforded stable adducts with both N-acetyltyrosine and N-
acetyllysine; furthermore, we show that the reactivity of arylsulfonyl fluorides towards these nucleophilic amino acids can
be predictably modulated by adjusting the electronic properties of the warhead. These trends were largely conserved
when the covalent reaction occurred within a protein binding pocket. We have also obtained a crystal structure depicting
covalent modification of the catalytic lysine of
a tyrosine kinase (FGFR1) by the ATP analog 5’-O-3-
((fluorosulfonyl)benzoyl)adenosine (m-FSBA). Highly reactive warheads were demonstrated to be unstable with respect to
hydrolysis in buffered aqueous solutions, indicating that warhead reactivity must be carefully tuned to provide optimal
rates of protein modification. Our results demonstrate that the reactivity of SFs complements that of more commonly
studied acrylamides; and hope that this work spurs the rational design of novel SF-containing covalent probe compounds
and inhibitors, particularly in cases where a suitably positioned cysteine residue is not present
.
belief that reactive functional groups would lead to
promiscuous modification, and result in idiosyncratic
toxicity. Recent advances, however, have led to the
understanding that careful tuning of warhead reactivity can
allow for the generation of highly selective TCIs. TCIs are now
emerging as valuable therapeutic agents,2 spurring the
development of novel approaches for the covalent
modification of target proteins.3,4
Introduction
The covalent modification of proteins by small molecules is an
important and active area of research in both medicinal
chemistry and chemical biology. Indeed, within drug discovery
targeted covalent inhibitors (TCIs) have undergone
a
resurgence in recent years.1 Historically, the development of
covalent inhibitors as drugs was discouraged due to the
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 2 of 12
View Article Online
DOI: 10.1039/C7OB02028G
ARTICLE
Journal Name
Despite renewed interested in TCIs, the ability to modify the
target protein is often limited by the serendipitous placement
of a nucleophilic cysteine residue in, or near, the binding site
of the small molecule. This limitation vastly reduces the
number of targets that can utilize covalent modification;
warheads which expand the range of protein residues that can
be modified are thus highly desirable. In this regard, the
sulfonyl fluoride (-SO2F) group, originally described by Baker,5
has recently reemerged as a promising warhead for the
covalent modification of proteins. Covalent inhibitors and
probe compounds containing sulfonyl fluorides (SFs) and
related SVI-F groups that address a wide range of protein
targets have been described in the literature,4 and several
have been utilized in activity-based protein profiling (ABPP)
experiments.6,7,8 The reason for the rapid adoption of these SF-
based warheads is due to several attractive features, including:
1) their ability to covalently modify a range of nucleophilic
amino acids under physiological conditions, including
tyrosine,9,10,11 lysine,12 proline,13 cysteine,14,15 histidine,16,17
catalytic serine6,18 and threonine;19 2) their ease of
incorporation into complex molecules, owing to the stability of
this group towards
a
wide range of synthetic
transformations;20 3) the biologically benign nature of the
fluoride leaving group; and 4) the stability of the resultant
sulfonate and sulfonamide linkages.21 Recently, Sharpless and
coworkers have expanded such SVI-F-derived scaffolds into the
“third dimension” through the use of SOF4 as a versatile
building block.22 Despite the wealth of literature on the
synthesis and applications of the SF group, a comprehensive
study of the factors contributing to the reactivity of this
functionality has not yet been reported. Similarly, descriptions
of the stability of sulfonyl fluorides in aqueous solutions are
limited to a single report of a small panel of 18F-labelled
sulfonyl fluorides.23 Herein, we describe our efforts to profile
the reactivity of the SF group and other closely related
warheads in order to generate an understanding of the factors
that influence its ability to modify nucleophilic amino acid
residues. In these experiments we have examined the reaction
between the warheads and free amino acids, as well as the
factors affecting the hydrolytic, plasma, and metabolic stability
of compounds incorporating the SF group. Finally, we have
studied the interplay between these factors when the SO2F
group is bound within a kinase active site. Overall, our data
suggest that sulfonyl fluorides and related SVI-F groups
represent versatile warheads for the development of novel
covalent inhibitors and biochemical probe compounds.
Fig. 1 Structures of sulfonyl fluorides and related SVI-F
warheads examined in this study. For comparison, a small
panel of N-arylacrylamides (10a – e) was also synthesized and
evaluated.
Results and Discussion
and purity from the corresponding commercially available
sulfonyl chlorides, utilizing potassium bifluoride under the
biphasic conditions previously described by Sharpless and co-
Synthesis of Sulfonyl Fluorides and Closely Related Warheads.
workers.24
(dimethylamino)benzenesulfonyl fluoride 1o
An
exception
was
4-
To probe the influence of various structural and electronic
factors on the reactivity and stability of the SF moiety, we first
,
which was
synthesized a panel of monosubstituted aryl SFs (Fig. 1,
1 – 3).
synthesized from 4-fluorobenzenesulfonyl fluoride 1j and
Generally, these compounds were prepared in high yields
dimethylamine in THF. The reaction proved selective for the
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 3 of 12
Journal Name
View Article Online
DOI: 10.1039/C7OB02028G
ARTICLE
desired SNAr product over the isomeric sulfonamide (>9:1). reactivity of our SFs with that of the N-arylacrylamides. We
However, when water was the reaction solvent, the first noted, however, that when the SF library was screened
sulfonamide was the exclusive product, highlighting the under identical (pseudo-first order) conditions we observed
previously reported sensitivity of the sulfonyl fluoride group the formation of a small amount of the corresponding sulfonic
towards activation by hydrogen bond donors.24,25 In addition acid. Due to the magnitude difference between the second-
to this aryl SF library, several other closely related warheads (
– 9) and a range of electronically diverse N-arylacrylamides Information – Appendix 1: Table S2-6), we did not correct the
10a – e) were synthesized (see Supporting Information).
4
order rate constants for the two processes (see Supporting
(
subsequent experiments for the hydrolysis. In addition, in the
case of N-acetylcysteine, the initial adduct expected,
sulfonothioate ester 11, was not observed (Scheme 1). Instead,
11 was subsequently reduced to the corresponding sulfinic
acid 12 via attack of another cysteine, with concomitant
formation of N-acetylcysteine disulfide.29 While this second
step is clearly favored under our assay conditions due to the
presence of excess N-acetylcysteine, it should be noted that
intermediate 11 was never observed by MS analysis, even
when less reactive warheads (vide infra) were assayed or when
the reaction mixtures were analyzed at early time points.
Similarly, attempts to observe 11 by increasing the ratio of SF
to N-acetylcysteine were also unsuccessful; reaction mixtures
containing either equimolar amounts of the two reactants or a
fivefold excess of SF relative to N-acetylcysteine both afforded
the corresponding sulfinic acid and N-acetylcysteine disulfide
as the only observable products.
Comparison between the Reactivity of Sulfonyl Fluorides and N-
Arylacrylamides with Cysteine.
Fig. 2 Plot depicting the second-order rate constants for the
reaction of N-acetylcysteine with N-arylacrylamides (filled
-
circles) or aryl SFs (open circles) vs. the Hammett σp
parameter of the substituent. The dashed and dotted lines
Scheme 1 Reduction of the SF group by N-acetylcysteine.
Proposed intermediate 11 was not observed under our assay
conditions.
indicate best fit lines for the acrylamide (
and SF (
= 1.72, R2 = 0.989) rate data, respectively. Error bars
represent the mean ± standard deviation of three independent
ρ
= 0.79, R2 = 0.947)
ρ
experiments.
Taken together, our results suggest that utilizing the SF group
to stably modify a target protein via reaction with a cysteine
residue should be undertaken with caution due to the inherent
instability of intermediates of type 11. Thus, while sulfonyl
fluorides have been demonstrated to react with cysteine
residues by us (vide supra) and others,14,15 it is instructive to
note that crystal structures exhibiting SF-modified cysteine
residue(s) remain elusive. Similarly, when SFs are used as
probe molecules in biochemical contexts, we anticipate that
SF-cysteine adducts will almost certainly be cleaved by
standard reducing agents used in sample preparations,
possibly leading to misleading selectivity data. We do note,
however, that targeting cysteine residues with SF-based
compounds can in certain cases lead to durable inhibition, if
the intermediate thiosulfonate ester is subsequently captured
intramolecularly by a nucleophilic side chain.14,15
The reaction of a cysteine residue with an N-arylacrylamide is
by far the most commonly used methodology for the site-
specific covalent inhibition of proteins.26 As such, this reaction
has been extensively studied under physiological conditions,
and its rate has been demonstrated to follow a log-linear
relationship with respect to the Hammett value of the
attached aromatic ring.27,28 Therefore, we began our enquiries
by determining the relative rates of reaction of
monosubstituted N-arylacrylamides 10a – e (10 µM) with N-
acetylcysteine (100 µM) in phosphate buffer (100 mM)
containing NaCl (1.0 M) and 5% CH3CN at pH 7.5 at 37 °C (All
subsequent studies described below utilized identical
conditions, but substituted
a different electrophile or
nucleophile as needed – for detailed experimental procedures
and example plots see Supporting Information). Under these
assay conditions we recapitulated the previously observed log-
linear relationship between the Hammett values of the aryl
ring substituents of the N-arylacrylamides and their second
order rate constants (Fig. 2, black closed circles). With these
benchmarking experiments in hand, we sought to compare the
The rapid cleavage of intermediate 11 and the fact that it does
not detectably accumulate indicates that its formation is rate-
limiting, allowing us to measure the rate of the initial reaction
of the thiol with SF by monitoring formation of the
corresponding sulfinic acid 12. A subset of our SF library (1a –
e
) containing identical substituents to the N-arylacrylamides
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 4 of 12
View Article Online
DOI: 10.1039/C7OB02028G
ARTICLE
Journal Name
was exposed to the N-acetylcysteine reactivity assay described capability is valuable in the context of targeted covalent
above. The results showed that the SFs (open circles), too, modification, because it may expand the number of residues in
-
displayed
a
log-linear relationship to the Hammett σp proteins amenable to covalent modification. In particular,
parameter, and that the SFs generally reacted faster than the tyrosine residues are often enriched at protein-protein
corresponding N-arylacrylamides (filled circles) under these interfaces,31 thus SFs may represent promising warheads for
conditions. Indeed, the rate of reaction of nitro derivative 1b these challenging targets. We therefore decided to test our
was ≥16-fold greater than that of the corresponding broad panel of substituted aryl SFs using the LC-MS assay
acrylamide, and was too rapid to be accurately determined. described above, but using N-acetyltyrosine as the
The reactivity of the SFs (ρ
substituents than is the case for the arylacrylamides (
~ 1.7) is more sensitive to aryl nucleophile. As a first point of reference, we observed that the
~ 0.8), initial reaction rate of parent SF 1a was 3.4-fold smaller when
ρ
presumably at least in part because in the SFs the electrophilic N-acetyltyrosine was the nucleophile relative to N-
sulfonyl sulfur atom is directly connected to the aromatic ring. acetylcysteine. A Hammett analysis of the resulting initial rate
Thus, while SFs do not afford stable adducts with cysteine data revealed a strong correlation between the electron-
residues, their reactivity can be tuned over a wide range by withdrawing properties (as measured by the σp- parameter) of
substitution on the aromatic ring; furthermore, this variability the substituent32 and the rate of reaction (Fig. 3, filled circles).
encompasses the absolute range of reactivities measured for Ortho-substituted sulfonyl fluorides (3a – f, open circles)
the N-arylacrylamides. The cysteine-acrylamide interaction is exhibited roughly the same correlation between reaction rates
-
known to have a kinetic profile that is suitable for use in TCI and the σp parameter, suggesting that electronic rather than
drugs. The above results suggest that a similarly useful kinetic steric factors largely determine the reactivity of this family of
profile is, in principle, achievable with a suitably functionalized compounds (ρortho = 1.52, ρpara = 1.63). We also note the
SF warhead.
possibility of stereoelectronic interactions between the SF
group and an ortho substituent, and observe some evidence
33
for this in ortho-methoxy containing 3f
.
Evaluation of Sulfonyl Fluoride Reactivity against Tyrosine, Lysine
and Serine
We also note that N-arylacrylamide 10b, which exhibited a
rate of reaction with N-acetylcysteine comparable to that of SF
1a, did not afford any 1,4-addition product when screened
against N-acetyltyrosine. Lastly, control experiments revealed
that the reaction was susceptible to general acid-base
catalysis; thus, the concentration and identity of the reaction
buffer did have an effect on the measured rates of reaction
with N-acetyltyrosine (see Supporting Information for details).
We next turned our attention to the reaction between the SF
warhead and lysine. At physiological pH, lysine side chains
exist predominately in the protonated form, and therefore
usually only pKa-perturbed lysines can be successfully targeted
with covalent warheads.12 Nonetheless, solvent-exposed lysine
side chains can be covalently modified by highly reactive
warheads, which can react rapidly with the small fraction of
unionized amine present at equilibrium to form the covalent
linkage. These reactions typically require that the warhead is
appended to a high affinity small molecule, ensuring that the
warhead experiences a long residence time in close proximity
to the nucleophilic lysine, and/or long reaction times.34 We
were therefore intrigued to study the reaction of aryl SFs with
non-pKa perturbed lysines. Using a subset of our panel of aryl
SFs (1a – e), we again repeated our reactivity assay, employing
Nα-acetyllysine as the nucleophile (Supporting Information,
Table S2). Under our assay conditions, each of the five SFs
afforded the corresponding sulfonamide, though in all cases
this reaction was slower than the corresponding reactions of
N-acetyltyrosine (2.9-fold rate reduction for 1a) and N-
acetylcysteine (10-fold rate reduction for 1a). This result
suggests that it may be possible to design small molecules with
highly reactive SF warheads to modify lysines when more
suitable residues are not available, but slow reaction rates and
Fig. 3 Hammett plot showing the dependence of the rate of
modification of N-acetyltyrosine by monosubstituted aryl SFs
-
-
on the σp (ortho and para substituents) or σm (meta
substituents) parameter. The dashed line indicates the best
linear fit for the data (ρ
= 1.53, R2 = 0.913). For rate data in
tabular format, refer to Table S1 (Supporting Information).
Error bars represent the mean ± standard deviation of three
independent experiments.
Having established that the rate of modification of cysteine by
SFs is more susceptible to electronic effects than that of the N-
arylacrylamides, we next chose to examine the ability of our
electronically diverse SFs to modify other amino-acid side
chains. It is well documented that SFs can be utilized to modify
tyrosine residues in proteins in a site-specific manner.11,30 This
4 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 5 of 12
Journal Name
View Article Online
DOI: 10.1039/C7OB02028G
ARTICLE
competing warhead hydrolysis (vide infra), may represent strategies for the rational design of highly selective or highly
significant challenges to this approach. We additionally sought efficient covalent probe molecules.
to explore the potential utility of SF warheads to covalently
modify unactivated serine residues; however, only sulfonic Hydrolytic, Plasma and Metabolic Stability of Aryl Sulfonyl
acid formation (hydrolysis) was observed when N-acetylserine Fluorides
was utilized as the nucleophile under our assay conditions,
even with the most electron deficient SFs.
Effects of pH and Buffer on the Rate of Reaction
Having demonstrated that aryl SFs react with cysteine,
tyrosine, and lysine at rates dependent on the reactivity of the
amino acid side chain as well as the electronic substitution of
the aryl ring, we next tested the effect of pH on the reaction
rates. Using a subset of SFs (1a – e), we performed the same
MS based assay at two additional pH values for all three amino
acid nucleophiles. As expected, increasing the reaction pH to
10.0 greatly increased the reaction rate in all three cases
(Supporting Information, Table S2). This rate acceleration was
so large that no reliable rate data could be obtained for
reactions involving N-acetylcysteine. For the other two amino
acids, only strongly deactivated warheads 1d and 1e afforded
reliable data, showing rate accelerations of 95- and 59-fold,
Fig. 4 Hammett plot showing the dependence of the rate of
-
respectively, for reactions with N-acetyltyrosine, and 161- and
201-fold, respectively, for reactions with Nα-acetyllysine.
Predictably, lowering the reaction pH to 5.0 had the opposite
hydrolysis of monosubstituted aryl SFs on the σm (meta
-
substituent) or σp (ortho and para substituents) parameter.
The dashed line indicates the best linear fit for the data (ρ =
effect. In this case, neutral or deactivated SFs 1a, 1d, and 1e,
1.45, R2 = 0.905). Complete tabulated data is provided in the
Supporting Information (Table S1). Error bars represent the
mean ± standard deviation of three independent experiments.
failed to react with N-acetylcysteine or N-acetyltyrosine, and
1a and 1c – e were all inert towards Nα-acetyllysine at this pH.
For strongly activated SF 1b, we observed 26-, 44-, and 89-fold
decreases in reaction rate for N-acetylcysteine, N-
acetyltyrosine, and Nα-acetyllysine, respectively. Similar trends
were observed when the hydrolytic stabilities of 1a – e were
examined at varying pH (Table S3, Supporting Information).
These data indicate that, at physiological pH, the reaction of SF
with each of the three nucleophilic residues is subject to
specific base catalysis, presumably due to the need to
With an understanding of the relative reactivities of the aryl
SFs, we next focused our attention on their hydrolytic stability.
Modifying our previously described assay, by leaving out any
nucleophilic amino acid and instead including N-
acetylphenylalanine as an inert internal standard, we
monitored the formation of the corresponding sulfonic acid by
MS. As expected, the rate of hydrolysis also displayed a strong
correlation with the electronics of the aryl ring (Figure 4).
+
deprotonate the nucleophile (-SH, -OH or –NH3 ) for attack on
the SF electrophile. The rate would be expected to become pH
independent at values above the pKa of the nucleophile.
However, the highest pH tested here, pH 10, is below the pKa
of the nucleophilic amine of lysine and of the phenolic oxygen
of tyrosine, and for N-acetylcysteine (pKa ~9.5) the reaction at
pH 10 was too fast for the rate to be accurately measured
using our MS-based assay method. The observation that
reaction at pH 7.5 involves specific base catalysis is consistent
with our observation that the reaction is also subject to
general acid-base catalysis by buffer, as described above (Fig.
S1, Supporting Information). We speculate that this buffer
catalysis similarly involves deprotonation of the nucleophile by
basic buffer components. In the context of TCIs, these findings
suggest that rate of covalent modification of proteins by SF-
containing inhibitors could be enhanced if the target residue is
located near an appropriately-positioned basic side chain. The
presence of such a nucleophilic site diad could provide new
Strongly electron-deficient SFs 1b, 1c, 1f, and 1g possessed
half-lives on the order of 5 – 15 minutes under these
conditions, while strongly electron donating substituents
greatly improved the stability of the SF; compounds 1e and 1o
were stable for several days under these conditions.
Interestingly, the rates of hydrolysis of 3b and 3f deviated from
the generally observed trends; this observation mirrors their
behavior in reactivity assays employing N-acetyltyrosine and
supports the hypothesis that a stereoelectronic interaction
between their ortho-substituents and the –SO2F group
influences the reactivity of these two compounds. Lastly, we
note that the acrylamides did not undergo appreciable
hydrolysis under our assay conditions, even when the reaction
was monitored for several days.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 6 of 12
View Article Online
DOI: 10.1039/C7OB02028G
ARTICLE
Journal Name
When a subset of our SF library was subjected to a rat plasma Investigation of Closely Related Warheads
We next studied the reactivity and stability of related SVI-F
groups, including vinyl sulfonyl fluorides, aliphatic sulfonyl
fluorides, aryl fluorosulfates, and aryl sulfonimidoyl fluorides.
To this end, compounds 4 – 9 were synthesized and their rates
of reaction with tyrosine and rates of hydrolysis were
determined (Supporting Information, Table S4). 2-
stability assay in which the loss of SF was recorded over time
(it was not possible to determine if the compounds had
hydrolyzed or reacted with the constituent parts of the rat
plasma), the same correlation was observed (Fig. 5). Although
compound 1o showed remarkable stability (t1/2 = 25.1 h), we
note that it remains relatively unstable when compared to the
N-arylacrylamides, which do not undergo appreciable reaction
under these conditions (data not shown). Taken together, our
results indicate that the electronic parameters of a given SF
warhead must be carefully tuned based on its intended
application. In particular, highly reactive SF probes may exhibit
reduced rates of covalent modification due to competing
hydrolysis under physiological conditions (vide infra).
Phenylethenesulfonyl fluoride
approximately 9-fold less reactive towards N-acetyltyrosine
than 1a and, as expected, this decrease in reactivity
4
was found to be
,
correlated with an increase in hydrolytic stability. Despite the
presence of a strongly electron-deficient olefin, 1,4-addition of
N-acetyltyrosine to the vinylsulfonyl fluoride moiety of
not observed in any assay. Phenylmethanesulfonyl fluoride
(PMSF, ), a protease/esterase inhibitor widely used in the
4 was
5
biological sciences,36 was observed to rapidly hydrolyze at pH
7.5, in agreement with previous reports.37 However, in
contrast to our previous experiments with aryl SFs, this
decrease in hydrolytic stability did not correlate with an
increase in reactivity towards N-acetyltyrosine. This
observation is consistent with hydrolysis occurring through the
intermediacy of a sulfene-like intermediate, rather than via
direct reaction at sulfur. The hypothesis that aliphatic SFs
primarily hydrolyze via sulfene-like intermediates was further
supported when we observed that the rate of hydrolysis of 2-
phenylethanesulfonyl fluoride (PESF,
6
) was roughly 13-fold
, presumably due to the decreased acidity
. Despite this increased stability, we still
under our
slower than that of
5
of the α-protons of
6
did not observe sulfonylation of tyrosine by
6
standard assay conditions, suggesting that sulfene formation
and subsequent hydrolysis were still much faster than
nucleophilic addition to the sulfonyl group. On the other hand,
Fig.
5 Plot showing the dependence of the rate of
decomposition of monosubstituted aryl SFs in rat plasma on
the Hammett σp- parameter of the aryl substituent. The dotted
fluorosulfates
7 and 8 (in which the carboxylic acid groups
were required for solubility reasons) were found to be
extremely stable under these conditions; no hydrolysis was
observed at pH 7.5 over 24 hours at 37 °C, nor did either
compound afford substitution products with either N-
acetyltyrosine or N-acetylcysteine. On extended incubation
with N-acetylcysteine (4 days), however, we did observe the
formation of the corresponding phenol. These phenols are
line indicates the best linear fit for the data (
0.927).
ρ =
= 1.75, R2
Finally, we were interested in investigating the metabolic
stability of our library. Unfortunately, the low molecular
weight and poor ionization in MS signal of a number of the
library members prevented accurate measurements in
standard pharmacokinetic assays. Nevertheless, three
thought to arise from initial formation of a sulfothioate 13
,
.
which is rapidly reduced by N-acetylcysteine to sulfite ester 14
compounds, 1k, 1n, and 1o afforded robust clearance data
Expulsion of SO2 from the sulfite ester then affords the
observed phenols (Scheme 2).
from rat hepatocytes (165, 24, and 170 µL / 106 cells min,
respectively). On the other hand, our library of FSBA analogues
19 (vide infra) afforded robust microsomal stability data.
Within this series of compounds we did not observe a strong
correlation between the Hammett value of the substituent and
the rate of metabolism, suggesting that the metabolism of SF-
containing molecules does not primarily depend on the
intrinsic reactivity of the SF group (Supporting Information, Fig.
S3A). In contrast, a stronger correlation existed between the
measured logD of the compounds and their rate of
metabolism, in agreement with well-known relationships in
medicinal chemistry (Supporting Information, Fig. S3B).35
Scheme 2 Reaction between a generic aryl flurosulfonate and
N-acetylcysteine. As with the sulfonyl fluorides, intermediates
13 and 14 were not observed.
6 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 7 of 12
Journal Name
Due to
View Article Online
DOI: 10.1039/C7OB02028G
ARTICLE
the
possibility
of
substitution
on ultimately afforded a mixture of products, with phenyl N-
nitrogen, the sulfonimidoyl fluoride group [R-SO(=NR’)F], may acetylcysteine disulfide 17 as the major component. We
represent a more versatile covalent warhead than the -SO2F propose that these reactions initially afford S-sulfonylated
group. Indeed, we recognized that not only does the N- substitution products 15, which are rapidly reduced to the
substituent represent a novel way to modulate the reactivity corresponding sulfinamides 16 and N-acetylcysteine disulfide,
of the sulfonimidoyl fluoride warhead, but it also offers a new in an analogous reaction to that of the sulfonyl fluorides and
vector with which to attach the covalent warhead to a larger fluorosulfates. The sulfinamides, unlike the deprotonated
molecule. Despite the potential advantages afforded by sulfinic acids produced in the sulfonyl fluoride reaction, are
sulfonimidoyl fluorides, which first appeared in the literature then further reduced by additional cysteine, ultimately
in 1983,38 reports of their applications remain scarce. We thus affording a mixture of disulfides (Scheme 3). We note that
synthesized
a
small
panel
of
N-substituted similar reactions of sulfinamides and thiols have previously
benzenesulfonimidoyl fluorides to determine the effects of N- been described in the literature. 40
substitution on the stability and reactivity of this class of
compounds.
Of the eight sulfonimidoyl fluorides, 9a – h, tested, only the
four most electron deficient, 9a – d, afforded adducts with N-
acetyltyrosine at pH 7.5 at 37 °C. In particular, sulfonimidoyl
fluorides bearing N-acyl (9a, 9b) substituents were found to
label N-acetyltyrosine more rapidly than PhSO2F (relative rates
= 2.5 and 1.2, respectively), whereas the rate of reaction of N-
Boc substrate (relative rate = 0.98) was similar to that of
PhSO2F, and that of N-(4-(trifluoromethyl))phenyl warhead 9d
(relative rate = 0.05) was substantially reduced (Supporting
Information, Table S5). In contrast to our observations of
sulfonyl fluorides, however, we found that variation of the N-
substituent could improve hydrolytic stability without
compromising reactivity – thus 9b (R = Piv) and 9c (R = Boc)
were both found to be more stable towards hydrolysis than
PhSO2F. Possibly, increasing the steric bulk of the N-
substituent increases the hydrolytic stability of these
warheads, while having a less pronounced effect on the
reactivity of this group towards weaker nucleophiles, though
how this might occur is unclear. Alternatively, it is possible that
sulfonimidoyl fluorides react through different mechanisms
depending on the nature of the N-substituent.
Notwithstanding these open questions, we regard
sulfonimidoyl fluorides bearing N-carboxyl and N-acyl groups
as promising, yet underexplored scaffolds for the development
of novel tyrosine-targeting covalent probe compounds, though
further experiments are undoubtedly required to fully
elucidate the reactivity and utility of these groups.
Scheme 3 Formation of N-acetylcysteine phenyl disulfide from
the reaction of N-acetylcysteine and benzenesulfonimidoyl
fluorides.
“In-Pocket” Reactivity of Sulfonyl Fluorides.
The reactivity profiling experiments described above
demonstrated that the reactivity and stability of substituted
aryl SFs was determined almost entirely by their electronic
properties. We recognized, however, that such a simplified
system may not accurately predict the reactivity of the sulfonyl
fluoride group and the target residue when covalent bond
formation is occurring in the context of a protein-inhibitor
complex. In order to assess the reactivity and stability profiles
of the SF group in such a system, we turned to two sulfonyl
fluoride-containing ATP analogs, 5’-O-(4-fluorosulfonyl)benzoyl
41
adenosine (p-FSBA, 18
)
and its more recently disclosed
isomer m-FSBA (19a, Figure 6).42 18 is a non-selective kinase
inhibitor, which has been demonstrated to covalently modify
the catalytic lysine residue43 (and potentially, in some cases,
other proximal residues)44 within the ATP binding pocket; m-
FSBA has been hypothesized to react in an analogous
manner.42 In order to probe the generality of the FSBA-kinase
interaction, we selected two kinases, FGFR1 and SYK, as
targets for covalent modification. In particular, an analysis of
previously published crystal structures of tyrosine kinase
FGFR1,45 suggested that it may be possible to incorporate
various electronically-modulating substituents ortho to the SF
on the aryl ring of m-FSBA without interfering with ligand
binding. A library of electronically diverse m-FBSA analogues
would subsequently allow us to test whether our previously
determined structure-reactivity relationships were conserved
in more complex systems.
In contrast, other N-aryl warheads 9e and 9f were observed to
rapidly decompose under assay conditions (rates of
decomposition relative to 1a = 6.6 and 5.3, respectively); we
hypothesized that this reduction in stability may be due to
an alternate pathway for decomposition,39 possibly involving
oxidation of their electron-rich arene rings. N-alkyl compounds
9g and 9h were inert under these conditions. Remarkably,
these two compounds were found to be exceptionally
stable even at pH 10.0, with no decomposition observed after
8 h at 37 °C. Attempts to access the substitution chemistry
of 9e – h using N-acetylcysteine as the nucleophile were
similarly unsuccessful; 9g and 9h remained inert, while
decomposition of 9e and 9f still outcompeted substitution
pathways. Sulfonimidoyl fluorides 9a – d did react with N-
acetylcysteine, though as before, the direct substitution
products were never observed. Rather, all four compounds
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 8 of 12
View Article Online
DOI: 10.1039/C7OB02028G
ARTICLE
Journal Name
Fig. 6 Structures of p- and m-FSBA, novel m-FSBA analogs, and observed within the crystal structure containing 19h. In one of
FSBA-derived fluorosulfates.
the two binding sites within the asymmetric unit of the FGFR1-
19h complex, the covalent bond is clearly visible and the ligand
binding mode is essentially identical to that of 19a (Fig. 7).
However, the electron density around the ethyl group was
weak, and thus the precise orientation of this group could not
be determined unambiguously. In the other binding site within
the asymmetric unit cell, the electron density indicated a
mixture of two species in a roughly 1 : 1 ratio. In the first of
these, the covalent bond had formed between the sulfur atom
of 19h and Lys514 of FGFR1. The second species depicted the
apparent noncovalent interaction between 19h and FGFR1,
thus the sulfonyl fluoride group was intact and was located
further away from the key lysine residue. Nevertheless, given
the overall degree of similarity of both the ligand binding
mode and the FGFR1 binding site between the co-crystal
structures of 19a and 19h, we were confident that substitution
at the 4’-position of the m-FSBA analog would largely be
tolerated.
This structure conclusively established that m-FSBA specifically
modifies the catalytic lysine of FGFR1; only
a single
modification of FGFR1 was ever observed by MS. It
also suggested that substitution at the 4-position of the aryl
ring can be accommodated, as the nearest protein residue
appears to be greater than 5 Å away. Confident that ortho
substitution on m-FSBA would not significantly disrupt binding
of the small molecule to the kinase domain of FGFR1, we
synthesized a panel of m-FSBA analogs (19a – j, Scheme 4).
Briefly, 4-substituted benzoic acids 22 were treated with hot
chlorosulfonic acid, then the resulting crude sulfonyl chlorides
were directly converted to the corresponding sulfonyl fluorides
23 with potassium bifluoride under biphasic conditions.
Utilizing a modified literature procedure, the crude benzoic
acids were then converted to the corresponding acid chlorides,
which were treated with adenosine and DMPU in the absence
of base to afford the substituted FSBA analogs 19 in low to
moderate yields.
To accomplish this aim, we first sought to obtain evidence that
substituents on the aryl ring of the FSBA analogs could be
accommodated within the ATP binding pocket of FGFR1. To
this end, we obtained two X-ray crystal structures of m-FSBA
analogs (19a PDB: 5O49 and 19h PDB: 5O4A) covalently bound
to the catalytic lysine (Lys514) of FGFR1 (Fig. 7). Both structures
contain two molecules in the asymmetric unit of the crystal
structure, with minimal differences in ligand binding. As
expected, the adenosine moiety of forms H-bond interactions
with the hinge of the kinase in both structures. The phenyl
ring occupies the space underneath the glycine rich loop and
positions the –SO2F group in close proximity to Lys514. The
glycine-rich loop appears to be flexible and partially disordered
in both structures, as indicated by weaker electron density
within this region. In spite of the inherent flexibility of this
loop, a covalent bond is clearly visible between Lys514 and the
sulfur atom of 19a. In contrast, multiple binding modes were
Fig. 7 X-Ray structures of m-FSBA analogs 19a (purple - PDB: 5O49) and 19h (blue - PDB: 5O4A) covalently bound to the catalytic
lysine residue (Lys514) of FGFR1.
8 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 9 of 12
View Article Online
DOI: 10.1039/C7OB02028G
Journal Name
ARTICLE
observed between the rate of modification of the kinases and
-
the σp parameter of the ortho-group, with the exception of
Scheme 4 Synthesis of FSBA analogs.
analogs 19b (R = NO2) and 19j (R = OCH3). The Hammett
ρ
values for these reactions were determined to be 0.77 and
1.12 for FGFR1 and SYK, respectively, in reasonable agreement
with the results obtained using isolated amino acids. FSBA
analog 19b, however, did not afford a covalent adduct with
either kinase domain. Indeed, 19b was fully converted to the
corresponding sulfonic acid within 10 min under our assay
conditions (LC-MS analysis). On the other hand, FSBA analog
19j was observed to react with the kinases far slower than
would be expected based on electronic properties alone,
suggesting that the electronic differences associated with the
methoxy group was not well tolerated. We also monitored the
rate of modification of FGFR1 by p-FSBA, and determined it to
be slower than that of m-FBSA (relative rate = 0.65) despite
electronic activation from the para-ester group. Here, we
attributed this difference to poor overlap between the
incoming amine and the sulfur center. For completeness, we
We subsequently determined the rates of covalent
modification of the kinase domains FGFR1 and SYK for our
panel of FSBA analogues using an LC-MS protocol (see
Supporting Information). We employed a large excess of the
FSBA analogs relative to the kinase to fully saturate the kinase
ATP binding site, thus enabling us to decouple the rate of
(irreversible) covalent modification from the non-covalent
equilibrium reaction. These experiments revealed an
interesting correlation between the rate of covalent
modification and the electronic properties of the FSBA
analogues (Figure 8). In line with our hypothesis that ortho
substituents would have minimal effects on the binding of the
m-FSBA analogues to FGFR1, analogs 19g – i (which all possess
similar Hammett parameters) had essentially identical reaction
rates, despite a large difference in steric bulk in close proximity
to the reactive warhead. A fairly robust correlation was
synthesized the two fluorosulfonate analogues 20 and 21
.
These two compounds showed only traces of modification of
FGFR1, even after extended reaction times, in line with our
expectations based on profiling experiments involving the
fluorosulfonate group.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 10 of 12
View Article Online
DOI: 10.1039/C7OB02028G
ARTICLE
Journal Name
modify nucleophilic residues other than cysteine, especially
tyrosines. Unfortunately, this aspect of SF chemistry comes at
the expense of reduced stability in aqueous media. Our results
provide a framework for the rational design of SF-based
warheads with specific kinetic profiles. These SF-based
covalent probes should be especially useful in cases where a
proximal cysteine residue is not present, thereby eliminating
the potential use of acrylamides and related groups. In
particular, we note that the rates of reaction measured for SF
warheads span the same range as those obtained in assays
employing N-arylacrylamides. Overall, the data reported
herein demonstrate that SFs possess predictable reactivity and
stability profiles and are able to engage numerous amino acids
in a covalent manner; we thus consider SFs and related groups
to be valuable warheads for the development of novel TCIs
and probe compounds.
Experimental
Fig. 8 Hammett plot showing the dependence of the rate of
modification of FGFR1 and SYK kinases domain by substituted
FSBA analogs on the σp- parameter of the substituent. Best
General Procedure for LC-MS Assays for S (IV)-
Reactivity Profiling
F
linear fits for FGFR1 (dashes,
(dots,
1.12, R2 = 0.771) are shown, excluding the point
where R = OCH3 19j, σ = -0.27, see text). Note that FSBA
ρ =
0.77, R2 = 0.801) and SYK
ρ
=
Reactions were conducted in 2-mL glass vials. Each vial was
charged with 250 µL water, 100 µL sodium hydrogen
phosphate / sodium dihydrogen phosphate buffer (1.0 M, pH
7.5), 400 µL N-acetyltyrosine solution (25 mM), 200 µL NaCl
solution (5.0 M), and 40 µL CH3CN. The reaction buffer was
prewarmed to 37 °C, then the reactions were initiated by the
addition of sulfonyl fluoride (10 µL, 100 mM in CH3CN). The
reaction vials were immediately placed in the preheated (37
°C) autosampler compartment of the mass spectrometer and
spectra were obtained at regular intervals (minimum sampling
interval ~2.5 minutes) using the same LC-MS conditions as
listed in General Experimental Considerations. For reactions
conducted with other amino acids, the corresponding N-acetyl
amino acid was used in place of N-acetyltyrosine. For stability
(
analog 19b (R = NO2) did not afford and adduct with either
kinase under assay conditions.
Conclusions
We have completed an extensive study of the effects of
electronic and steric factors on the reactivity and stability of
the sulfonyl fluoride and closely related SVI-F groups. While
alkylsulfonyl fluorides are typically too unstable to effect the
modification of amino acids, we found that the reactivity of
arylsulfonyl fluorides can be predictably adjusted by profiling, N-acetylphenylalanine was used in place of N-
acetyltyrosine. For reactions conducted at other pH values, the
sodium phosphate buffer was replaced with either sodium
acetate/acetic acid buffer (1.0 M, pH 5.0) or sodium
bicarbonate/sodium carbonate buffer (1.0 M, pH 10.0).
Reaction products were identified by mass spectrometry (ES-
ionization); sulfonyl fluorides which did not yield ions under
these conditions were identified by comparison of their LC UV
chromatogram with that of a purified sample. Samples were
quantified by integration of the UV traces. For stability studies,
extent of hydrolysis was determined, where possible, by UV
quantification of the corresponding sulfonic acid. In certain
modulating the electronic properties of the warhead, enabling
the rational design of covalent compounds targeting a given
residue. Furthermore, sulfonimidoyl fluorides bearing
electron-withdrawing groups on nitrogen were found to
exhibit comparable rates of reaction with tyrosine, relative to
SFs, while in certain cases also displaying enhanced stability,
and may thus represent an even more promising starting point
for the development of novel TCIs or other probe compounds.
Although the hydrolytic stability of the sulfonyl fluoride moiety
proved to be a liability when highly reactive SF warheads were
incorporated into probe molecules, it is the less reactive SFs cases, however, the sulfonic acid eluted within the first 0.30
minutes of the LC run; this interval could not reliably be
quantified due to the presence of inorganic salts in the
reaction mixture. In these cases, extent of hydrolysis was
determined by the loss of parent sulfonyl fluoride signal in
comparison to N-acetylphenylalanine. In all cases, the reaction
rates reported in the text were obtained by measuring the
formation of product over time in the interval where this rate
was linear (typically <10% product formation). For example
whose reactivity matches that of the well-validated N-
arylacrylamide class of TCI warheads. Moreover, this trend
towards reduced stability for the more reactive SFs was not
observed when the metabolic stability of these compounds
was examined in rat hepatocytes. This indicates that the
intrinsic reactivity of the sulfonyl fluoride group may not
necessarily be a useful predictive marker when considering the
metabolic clearance of SF-containing probe molecules.
plots see Figure S1
.
In many ways, the reactivity of the SF moiety complements
that of the N-arylacryamide group due to its ability to stably
10 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 11 of 12
Journal Name
View Article Online
DOI: 10.1039/C7OB02028G
ARTICLE
1 J. Singh, R. C. Petter, T. A. Baillie and A. Whitty, Nat. Rev.
Drug. Disc., 2011, 10, 307.
Rat Plasma Stability Assay
Each well of a deep well 96-well plate was charged with 200 µL
acetonitrile. Blank samples were prepared by adding 50 µL rat
plasma (purchased from BioreclamationIVT (Westbury
NY). Part number was RATPLEDTA3-WH. (Wistar Hanover
rats, plasma in K3 EDTA)) to the acetonitrile. The plate was
then incubated on ice, protected from light. Assay samples
were prepared by adding 5 µL of each of the indicated sulfonyl
fluorides (4 mM in DMSO) to 995 µL rat plasma at room
temperature. The samples were mixed thoroughly, then were
allowed to incubate at room temperature. At intervals of 0, 15,
30, 60, 120, 240, and 360 minutes, a 50 µL aliquot of the
reaction mixture was removed and added to the deep well
plate containing cold acetonitrile. The plate was mixed using a
2 D. A. Cross, S. E. Ashton, S. Ghiorghiu, C. Eberlein, C. A.
Nebhan, P. J. Spitzler, J. P. Orme, M. R. V. Finlay, R. A.
Ward, M. J. Mellor, G. Hughes, A. Rahi, V. N. Jacobs,
M. Red Brewer, E. Ichihara, J. Sun, H. Jin, P. Ballard, K. Al-
Kadhimi, R. Rowlinson, T. Klinowska, G. H. Richmond,
M. Cantarini, D. W. Kim, M. R. Ranson and W. Pao, Cancer
Discovery, 2014, 4, 1046.
3 G. Akçay, M. A. Belmonte, B. Aquila, C. Chuaqui, A. W.
Hird, M. L. Lamb, P. B. Rawlins, N. Su, S. Tentarelli, N. P.
Grimster and Q. Su. Nat. Chem. Bio., 2016, 12, 931.
4 I. M. Serafimova, M. A. Pufall, S. Krishnan, K. Duda, M. S.
Cohen, R. L. Maglathlin, J. M. McFarland, R. M. Miller, M.
plate shaker (ca. 5 s) to ensure complete mixing after each Frödin and J. Taunton, Nat. Chem. Biol., 2012, 8, 471.
addition of plasma solution. After the final aliquot of plasma
was added to the plate, the plate was centrifuged (4 000 x g, 5
min). The supernatant was withdrawn from each well and
subjected to LC-MS/MS analysis using an Agilent 1290 UHPLC
coupled to an Agilent 6490 triple quad mass spectrometer. The
MS data was integrated using the MRM peak area for each
sulfonyl fluoride. To obtain the reported t1/2 values, ln (St –
Bt), where St = sulfonyl fluoride peak area at time t and Bt =
blank area at time t, was plotted against time, affording slope
m. The t1/2 values were then calculated (assuming first-order
kinetics) as t1/2 = -ln (2)/m.
5 B. R. Baker, Acc. Chem. Res., 1969, 2, 129.
6 D. A. Shannon, C. Gu, C. J. McLaughlin, M. Kaiser, R. A.
L. van der Hoorn and E. Weerapana, ChemBioChem, 2012,
13, 2327.
7 C. Gu, D. A. Shannon, T. Colby, Z. Wang, M. Shabab, S.
Kumari, J. G. Villamor, C. J. McLaughlin, E. Weerapana, M.
Kaiser, B. F. Cravatt and R. A. L. van der Hoorn, Chem. &
Bio., 2013, 20, 457.
8 S. S. Khandekar, B. Feng, T. Yi, S. Chen, N. Laping and N.
J. Bramson, Biomol. Screening, 2005, 10, 447.
9 F. S. Esch and W. S. Allison, J. Biol. Chem., 1980, 255,
9388.
10 D. Manvar, K. Singh and V. N. Pandey, Biochem., 2013,
52, 432.
11 E. C. Hett, H. Xu, K. F. Geoghegan, A. Gopalsamy, R. E.
Kyne Jr., C. A. Menard, A. Narayanan, M. D. Parikh, S. Liu,
L. Roberts, R. P. Robinson, M. A. Tones and L. H. Jones,
ACS Chem. Biol., 2015, 10, 1094.
12 N. P. Grimster, S. Connelly, A. Baranczak, J. Dong, L. B.
Krasnova, K. B. Sharpless, E. T. Powers, I. A. Wilson, J. W.
Kelly, J. Am. Chem. Soc., 2013, 135, 5656.
13 G. V. Crichlow, J. B. Lubetsky, L. Leng, R. Bucala and E.
J. Lolis, Biochem., 2009, 48, 132.
14 A. E. Annamalai and R. F. Colman, J. Biol. Chem., 1981,
256, 10276.
15 C. Oudot, J.-M. Jault, M. Jaquinod, D. Negre, J.-F. Prost,
A. J. Cozzone and J.-C. Cortay, Eur. J. Biochem., 1998, 258,
579.
16 D. A. Bullough and W. S. Allison, J. Biol. Chem., 1986,
261, 5722.
17 K. W. Harlow and R. L. Switzer, J. Biol. Chem., 1990,
265, 5487.
Profiling the Reactivity of FBSA Analogues against
SYK or FGFR1
Reactions were performed in 2-mL glass vials fitted with 0.2-
mL polypropylene inserts. Each insert was charged with 100 µL
reaction buffer (described above) containing FGFR1 or SYK
kinase domain (10 µM). The reaction vials were preheated to
37 °C, then the reactions were initiated by the addition of a
DMSO solution of FSBA analogue (10 mM, 1.0 µL, 10 eq.
relative to kinase). The reaction vials were immediately placed
in the preheated (37 °C) autosampler compartment of the
mass spectrometer. Mass spectra were obtained at regular
intervals (minimum sampling interval ~6 minutes) using a
Waters UPLC linked to a Waters SQD mass spectrometer
(Column temperature = 27 °C, UV = 210-400 nm; positive
electrospray ionization) equipped with a Phenomenex Jupiter
5u C4 300A column (5 µm, 2.0 x 50 mm), eluting with 2 – 98%
solvent B in solvent A over 4.5 minutes followed by a 1 minute
hold at 98% solvent B, where solvent A = water containing
0.1% formic acid and solvent B = CH3CN containing 0.1% formic
acid. The mass spectra corresponding to the proteins were
averaged and deconvoluted; the deconvoluted spectra were
then integrated to afford the initial rate data.
18 D. E. Fahrney and A. M. Gold, J. Am. Chem. Soc., 1963,
85, 997.
19 C. Dubiella, H. Cui, M. Gersch, A. J. Brouwer, S. A.
Sieber, A. Krüger, R. M. J. Liskamp and M. Groll, Angew.
Chem. Int. Ed., 2014, 53, 11969.
20 For a review: J. Dong; L. Krasnova, M. G. Finn and K. B.
Sharpless, Angew. Chem. Int. Ed., 2014, 53, 9430 and
references therein.
21 It should be noted, however, that the adducts formed
between sulfonyl fluorides and cysteine and histidine residues
are generally unstable towards reduction (cysteine) or
hydrolysis (histidine).
Acknowledgements
We thank A. Hird and S. Edmonson for proofreading the
manuscript and helpful discussions. We thank the AstraZeneca
PostDoc Program for funding this project.
Notes and references
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 11
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 12 of 12
View Article Online
DOI: 10.1039/C7OB02028G
ARTICLE
Journal Name
22 S. Li, P. Wu, L. E. Moses, K. B. Sharpless, Ang. Chem.
Int. Ed., 2017, 56, 2903.
23 L. Matesic, N. A. Wyatt, B. H. Fraser, M. Roberts, T. Q.’
Pham and I. Greguric, J. Org. Chem., 2013, 78, 11262.
24 S. Narayan, J. Muldoon, M. G. Finn, V. V. Fokin, H. C.
Kolb and K. B. Sharpless, Angew. Chem. Int. Ed., 2005, 44,
3275.
25 D. N. Kevill and M. J. D’Souza, J. Org. Chem., 2004, 69,
7044. and references therein.
26 For a review: T. Barf and A. Kaptein, J. Med. Chem.,
2012, 55, 6243.
27 V. J. Cee, L. P. Volak, Y. Chen, M. D. Bartberger, C.
Tegley, T. Arvedson, J. McCarter, A. S. Tasker, C. Fotsch, J.
Med. Chem., 2015, 58, 9171.
28 U. P. Dahal, A. M. Gilbert, R. S. Obach, M. E. Flanagan,
J. M. Chen, C. Garcia-Irizarry, J. T. Starr, B. Schuff, D. P.
Uccello and J. A. Young, Med. Chem. Commun., 2016, 7,
864.
29 J.-P. Mahieu, M. Gosselet, B. Sebille and Y. Beuzard,
Synth. Comm., 1986, 16, 1709.
30 A. Narayanan and L. H. Jones, Chemical Science, 2015, 6,
2650, and references therein.
31 A. A. Bogan and K. S. Thorn, J. Mol. Biol., 1998, 280, 1.
32 Hammett parameters were obtained from: C. Hansch, A.
Leo and R. W. Taft, Chem. Rev., 1991, 91, 165.
33 We note that the 19F NMR shift of 3f (δ 58.19) compared
to that of 1d (δ 65.96). Of note, the 19F chemical shift of the -
SO2F group of 3f is farther upfield than any of the other
benzenesulfonyl fluorides 1a – 3e. This is also observed when
comparing the 19F chemical shift of the –SO2F group of
FSBA analog 19j (δ 59.34) to that of analogs 19a – i.
34 For a review, see: E. Baslé, N. Joubert and M. Pucheault,
Chemistry & Biology, 2010, 17, 213, and references therein.
35 B. Testa, P. Crivori, M. Reist and P.-A. Carrupt,
Perspectives Drug Disc. Des., 2000, 19, 179.
36 D. H. Lee and A. L. Goldberg, Trends Cell Biol., 1998, 8,
397.
37 G. T. James, Analytical Biochem., 1978, 86, 574.
38 C. R. Johnson, G. B. Kostaki, J. H. Cantillo, N. A.
Meanwell, M. F. D. Reinhard, J. R. Zeller and G. P. Vonk, J.
Org. Chem., 1983, 48, 1.
39 The expected hydrolysis products of 9e and 9f were not
observed in our hydrolysis assay, whereas 9a – c cleanly
afforded the corresponding sulfonamides under these
conditions. 9d afforded only traces of the corresponding
sulfonamide.
40 V. Clarke and E. R. Cole, Phosphorus, Sulfur, Silicon, and
the Related Elements, 1994, 91, 45.
41 J. L. Wyatt and R. F. Colman, R. F. Biochemistry, 1977,
17, 1333.
42 Y.-L. Hsu, C.-C. Yang, C.-Y. Lan, J.-J. Lin, and L.-C. Lo,
J. Chin. Chem. Soc., 2013, 60, 846.
43 M. J. Zoller, N. C. Nelson and S. S. J. Taylor, Biol. Chem.,
1981, 256, 10837.
44 J. A. Schmidt and R. F. Colman, J. Biol. Chem., 1984,
259, 14515.
45 Y. Yosaatmadja, A. V. Patterson, J. B. Smaill and C. J.
Squire, Acta Crystallogr. D, 2015, 71, 525.
12 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins