Probing the target-specific inhibition of sensitized protein tyrosine phosphatases with biarsenical probes

Article abstract of DOI:10.1039/c4ob02256d

Selective control of enzyme activity is critical for elucidating the roles of specific proteins in signaling pathways. One potential means for developing truly target-specific inhibitors involves the use of protein engineering to sensitize a target enzyme to inhibition by a small molecule that does not inhibit homologous wild-type enzymes. Previously, it has been shown that protein tyrosine phosphatases (PTPs) can be sensitized to inhibition by a biarsenical probe, FlAsH-EDT₂, which inhibits PTP activity by specifically binding to cysteine residues that have been introduced into catalytically important regions. In the present study, we developed an array of biarsenical probes, some newly synthesized and some previously reported, to investigate for the first time the structure-activity relationships for PTP inhibition by biarsenicals. Our data show that biarsenical probes which contain substitutions at the 2′ and 7′ positions are more effective than FlAsH-EDT₂ at inhibiting sensitized PTPs. The increased potency of 2′,7′-substituted probes was observed when PTPs were assayed with both para-nitrophenylphosphate and phosphopeptide PTP substrates and at multiple probe concentrations. The data further indicate that the enhanced inhibitory properties are the result of increased binding affinity between the 2′,7′-substituted biarsenical probes and sensitized PTPs. In addition we provide previously unknown physicochemical and stability data for various biarsenical probes.

Full text of DOI:10.1039/c4ob02256d

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: A. Pomorski, J.
Adamczyk, A. Bishop and A. Krezel, Org. Biomol. Chem., 2014, DOI: 10.1039/C4OB02256D.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
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
Information for Authors.
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 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.
www.rsc.org/obc

Page 1 of 33
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C4OB02256D
Probing the targetꢀspecific inhibition of sensitized
protein tyrosine phosphatases with biarsenical
probes
Adam Pomorski,a Justyna Adamczyk,a Anthony C. Bishop,b and Artur Krężel*,a
a Laboratory of Chemical Biology, Faculty of Biotechnology, University of Wrocław, Joliotꢀ
Curie 14a, 50ꢀ383 Wrocław, Poland.
b Department of Chemistry, Amherst College, Amherst Massachusetts 01002, U.S.A.
* Corresponding Author:
Eꢀmail: krezel@biotech.uni.wroc.pl
Tel: +48 71 375 2765
† Electronic supplementary information (ESI) available.
1

Organic & Biomolecular Chemistry
Page 2 of 33
View Article Online
DOI: 10.1039/C4OB02256D
Abstract
Selective control of enzyme activity is critical for elucidating the roles of specific proteins in
signaling pathways. One potential means for developing truly targetꢀspecific inhibitors
involves the use of protein engineering to sensitize a target enzyme to inhibition by a small
molecule that does not inhibit homologous wildꢀtype enzymes. Previously, it has been shown
that protein tyrosine phosphatases (PTPs) can be sensitized to inhibition by a biarsenical
probe, FlAsHꢀEDT2, which inhibits PTP activity by specifically binding to cysteine residues
that have been introduced into catalytically important regions. In the present study, we
developed an array of biarsenical probes, some newly synthesized and some previously
reported, to investigate for the first time the structureꢀactivity relationships for PTP inhibition
by biarsenicals. Our data show that biarsenical probes which contain substitutions at the 2′
and 7′ positions are more effective than FlAsHꢀEDT2 at inhibiting sensitized PTPs. The
increased potency of 2′,7′ꢀsubstituted probes was observed when PTPs were assayed with
both paraꢀnitrophenylphosphate and phosphopeptide PTP substrates and at multiple probe
concentrations. The data further indicate that the enhanced inhibitory properties are the result
of increased binding affinity between the 2′,7′ꢀsubstituted biarsenical probes and sensitized
PTPs. In addition we provide previously unknown physicochemical and stability data for
various biarsenical probes.
2

Page 3 of 33
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C4OB02256D
Introduction
Selective control of the activity of a single enzyme is a key criterion for chemicalꢀbiology
strategies that aim to unravel a target enzyme’s physiological and/or pathological functions in
the cell. Chemical control of enzyme activity is most frequently achieved with smallꢀ
molecule inhibitors, which can either interact with a target enzyme’s active site or block its
activity by allostery. However, it is often very difficult to design a cellꢀpermeable small
molecule that is specific only toward a target enzyme, particularly if the enzyme is a member
of a large protein family. A complementary approach for achieving targetꢀspecific inhibition
of an enzyme involves sensitizing the enzyme of interest toward a selected molecule that does
not inhibit the wildꢀtype enzyme or other homologous proteins.1,2,3 In a successful
sensitization strategy, which requires engineering of the enzyme to introduce a binding site
for the inhibitor, suppression of enzyme activity is targetꢀspecific, a feature that is especially
desirable in cellꢀsignaling studies.
The classical protein tyrosine phosphatases (PTPs) are a large family of enzymes that
catalyze the dephosphorylation of phosphotyrosine in metazoan signaling networks. Because
the PTPs share a highly conserved active site and because many known inhibitors of wildꢀ
type PTPs are not cellꢀpermeable,4 this protein family potentially presents an ideal case for
sensitization strategies. Previous studies on PTP sensitization have shown that introduction of
spatially arranged cysteine residues into a loop of a PTP domain is sufficient to confer strong
affinity for the membraneꢀpermeable biarsenical probe, FlAsHꢀEDT2, initially designed to
specifically visualize proteins in the cell.5,6,7,8 Binding of FlAsHꢀEDT2 to the sensitized PTP
can cause strong inhibition of PTP activity, which relies on the specific conjugation of the
compound’s two arsenic atoms with the cysteineꢀrich motif that is incorporated into the target
PTP (Fig. 1A). The activities of wildꢀtype PTPs, therefore, are generally unaffected by
FlAsHꢀEDT2.
3

Organic & Biomolecular Chemistry
Page 4 of 33
View Article Online
DOI: 10.1039/C4OB02256D
An array of sensitized Tꢀcell PTP (TCPTP) mutants, which contain tetracysteine (TC)
motifs inserted at various loops within the enzyme’s PTP domain, has been previously
developed.5 These TCPTP mutants exhibited varying degrees of inhibition by FlAsHꢀEDT2,
the most highly sensitive being a construct in which the tetracysteine insertion was placed
directly adjacent to the PTP domain’s WPD loop (4C@187 TCPTP).6 The opening and
closing of the conserved WPD loop is critical for the activity of all classical PTPs,9,10 and it
was hypothesized that FlAsHꢀEDT2 inhibits 4C@187 TCPTP by rigidifying its WPDꢀloop’s
local structure and disrupting its normal dynamics. Subsequent studies showed that TCPTP
mutants with two or three cysteines placed within the WPD loop could also be potently
inhibited by FlAsHꢀEDT2.6
Work to date on PTP sensitization has focused solely on the optimization of protein
engineering, using only FlAsHꢀEDT2 as the inhibitory biarsenical compound. Since multiple
types of biarsenical probes with differing structures and physicochemical properties have
been reported11 we aimed to obtain a library of biarsenical probes, composed of both known
and novel compounds, to test the probes’ abilities to act as inhibitors of sensitized PTPs. We
hypothesized that screening an array of probes with previously engineered PTP mutants
would allow for the identification of optimized biarsenical/enzyme pairs, in which the degree
of engineered PTP inhibition is maximized. In the course of these studies we also made
unexpected observations concerning the differing stabilities of biarsenical probes in solution;
these findings, also reported here, will be of utility to the entire biarsenicalꢀresearch
community.
4

Page 5 of 33
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C4OB02256D
Results and discussion
Synthesis of biarsenical probes
The biarsenical probes used in this study were prepared based on resorufin, fluorescein, and
seven fluorescein derivatives. This library includes both new (Et2FlAsHꢀEDT2, Cl4FlAsHꢀ
EDT2, and 5ꢀBuꢀCrAsHꢀEDT2) and previously described compounds (Fig. 2).12,13,14,15 Parent
fluorophores that are not commercially available (2’,7’ꢀdifluorofluorescein16, 2’7’ꢀ
dichlorofluorescein17 and 2’7’ꢀdiethylfluorescein18) were synthesized using the appropriate
resorcinol derivative (4ꢀchloroꢀ, 4ꢀfluoroꢀ, and 4ꢀethyl, respectively) and phthalic anhydride.
Similarly, 4,5,6,7ꢀtetrafluorofluorescein16 and 4,5,6,7ꢀtetrachlorofluorescein17 were obtained
using appropriately substituted phthalic anhydrides in reaction with resorcinol (Fig. 3). 5ꢀ(Nꢀ
butylamide)ꢀfluorescein was obtained in the reaction of 5ꢀcarboxyfluorescein and nꢀ
buthylamide using Oꢀ(benzotriazolꢀ1ꢀyl)ꢀN,N,N′,N′ꢀtetramethyluronium hexafluorophosphate
(HBTU) and N,N,Nꢀdiisopropylethylamine (DIEA) as coupling agents (Fig.4).
Using a standard synthetic route involving mercurization of a parent fluorophore and
subsequent transmetalation with AsCl3,19 all compounds were successfully obtained with
high purity. Details regarding each synthetic step and purification are included in
Supplementary information (Fig. S1 ESI†). Initially, we also attempted to synthesize
biarsenical
probes
based
on
2′,4,5,6,7,7′ꢀhexafluoroꢀ
and
2′,4,5,6,7,7′ꢀ
hexafluorofluorescein16,17. Although the dimercury synthetic intermediates were obtained for
both hexaꢀhalogen fluorophores, the subsequent transmetalation was not successful, despite
several attempts (data not shown).
Physicochemical characterization of biarsenical probes
Since a full complement of physicochemical data for our library of biarsenical probes has not
been reported, we performed a systematic analysis of the probes in complex with an
5

Organic & Biomolecular Chemistry
Page 6 of 33
View Article Online
DOI: 10.1039/C4OB02256D
optimized tetracysteine peptide (TC12),20 the results of which are summarized in Table 1. The
full set of spectra can be found in Figure S2 (ESI†). The characterization data are in good
agreement with previously published results (where available), with one exception: our
absorbance (510 nm) and emission (530 nm) maxima wavelengths for the F2FlAsH−TC12
complex are higher than previously reported values by 10 and 8 nm, respectively.14
We suspect that, since experimental conditions (pH and tetracysteine peptide) are
essentially identical for the differing measurements, the discrepancy may be due to the
significant degradation of the F2FlAsHꢀEDT2 used in the previous study. F2FlAsH’s
corresponding monoarsenical species, has an absorbance maximum of 500 nm (Fig. S3,
ESI†),20 and we observed that DMSO solutions of 2′,7′ꢀsubstituted probes (Cl2FlAsHꢀEDT2,
Et2FlAsHꢀEDT2 and F2FlAsHꢀEDT2) degrade significantly when stored as DMSO solution
overnight at ꢀ20 °C (see below for further discussion of biarsenical stability).
Next, we determined the pKa values for the 3′ꢀphenolic groups of the biarsenical
probes in the library. Although proton dissociation from 3′ꢀphenol is required for the
fluorescent properties of fluorescein and resorufin derivatives, and pKa values are highly
sensitive reporters of the electron rich/poor character of a particular functional group, the pKa
values for almost all of the library’s probe−peptide complexes have not been previously
reported. Our systematic acidꢀbase study shows that in every case the biarsenical
probe−peptide complex has a lower phenolic pKa value than the parent fluorophore (Table 1).
The highest acidity of the phenolic group is observed for Cl2FlAsHꢀTC12 and F2FlAsHꢀ
TC12 complexes due to the high electronegativity of the halogen substituents. These probes
demonstrate stable fluorescence properties over a broad pH range, potentially enabling their
use in staining tetracysteineꢀtagged proteins in a variety of cellular compartments, including
lysosomes. Interestingly, in the biarsenical complexes containing TC12 peptide, a second
dissociation event with a pKa value of ~7 is also observed (Fig. S4, ESI†). We did not
6

Page 7 of 33
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C4OB02256D
observe the second dissociation event, however, when a minimal tetracysteine peptide (Acꢀ
CCPGCCꢀamide) was used (Fig. S5, ESI†). Based on the second dissociation constant values
(Table 1), we believe that the dissociation can be linked to the binding of the fourth cysteine
residue in a tetracysteine motif to an arsenic atom. Our previous results on the Zn(II)
coordination properties of tetracysteine peptides showed that the fourth cysteine residue has a
significantly increased pKa value, resulting in the formation of two complexes, one with three
bound sulfur donors (ZnS3) at neutral pH and one with four donors (ZnS4) at slightly alkaline
conditions.12 A reasonable explanation of this phenomenon is the steric hindrance of four
sulfurs bound to Zn(II) at weakly acidic pH. However, further experiments must be
performed to confirm this hypothesis.
Stability of biarsenical probes
The main route of biarsenical probe degradation in aqueous solution is the breaking of AsꢀC
bond, forming monoarsenical species and subsequently the parent fluorophore, exemplified
by mass spectra of decomposed Et2FlAsH (Fig. S6, ESI†). However, the extent or timeline of
this process has not been precisely measured in previous studies. Experiments with inhibition
of sensitized PTP with biarsenical probe include a 2.5 h preꢀincubation of enzyme with
biarsenical probe (vide infra). We therefore tested whether significant amounts of
monoarsenical species would be formed under these conditions, as significant degradation of
the biarsenicals would, of course, alter the outcome of the inhibition experiments. Using
analytical HPLC we determined that most of the probes do not degrade in significant amounts
(Fig. 5 and Table S1, ESI†). For example, a 6 h incubation gave rise to only a 0.92% increase
of monoarsenical FlAsHꢀEDT and a 0.42% increase in monoarsenical F4FlAsHꢀEDT. By
contrast, the extent of hydrolysis was significant for Et2FlAsHꢀEDT2, which showed a 14.2%
increase in monoarsenical species after 6 h. We also observed that, although all of the
7

Organic & Biomolecular Chemistry
Page 8 of 33
View Article Online
DOI: 10.1039/C4OB02256D
biarsenical probes described here can be stored in solid form for prolonged times without
degradation, DMSO stock solutions may degrade overnight, even when kept at ꢀ20°C, as we
observed for Cl2FlAsHꢀEDT2, Et2FlAsHꢀEDT2 and F2FlAsHꢀEDT2 (data not shown).
Inhibition of PTP activities by biarsenical probes
In order to assess the PTPꢀinhibitory activity for the library of biarsenical probes we
expressed and purified both wildꢀtype and engineered variants of two PTPs: Tꢀcell PTP
(TCPTP) and hematopoietic PTP (HePTP). The PTP set included four mutants of TCPTP and
one of HePTP (Figure 1), all of which have been previously described.5,6,7 We chose TCPTP
because the previously published optimization of localization of biarsenicalꢀbinding cysteines
was performed on this enzyme; HePTP was also used to confirm that the results can be
applied in other sensitized enzymes as well. The selected mutants have cysteineꢀenriched
biarsenical binding motifs with either two (P181C/E187C TCPTP), three (P181C+2C
TCPTP), or four cysteines (4C TCPTP, 4C@211 HePTP) within or adjacent to their
respective WPD loops. Also included in the study to serve as a negative control was one
mutant that contains a complete tetracysteine binding motif CCPGCC, but not within or near
the protein’s WPD loop (4C@241 TCPTP, Fig. 1B). All proteins were expressed from E. coli
and obtained in high yields, in most cases greater than 20 mg per liter of culture. Purification
of the sixꢀhistidine tagged proteins, carried out using standard protocols, provided pure
protein (Figure S7, ESI†).
An initial screen of the seven PTPs (2 wildꢀtype, 5 engineered) and nine biarsenical
probes (63 PTPꢀprobe combinations) was carried out with the smallꢀmolecule PTP substrate
paraꢀnitrophenylphosphate (pNPP) in the presence of 20 µM TCEP.24 We used TCEP instead
of DTT, since the latter can compete with tetracysteine motifs for binding of biarsenical
probes.15 TCEP has also been previously used in PTP assays,25 and a lack of this reducing
8

Page 9 of 33
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C4OB02256D
agent causes a decrease of enzyme activity during 2.5 h incubation at room temperature, see
Fig. S8 (ESI†). We started by establishing the key kinetic parameters—catalytic rate constant
(kcat) and the Michaelis constant (Km) for pNPP—in the absence or presence (fourꢀfold
excess) of biarsenical probe. A sample MichaelisꢀMenten plot is shown in Figure 6, and
Figure 7 contains a graphical representation of the catalytic efficiencies of different pairs of
enzymes and biarsenical probes. Detailed results can be found in Tables S2ꢀS4 (ESI†). In the
case of wildꢀtype TCPTP no significant inhibition is induced by most of the probes and only
marginal inhibition with FlAsHꢀEDT2, Cl2FlAsHꢀEDT2 and F2FlAsHꢀEDT2. Likewise, wildꢀ
type HePTP was not significantly inhibited by any of the biarsenical probes. These data are in
agreement with previous results using FlAsHꢀEDT2 with wildꢀtype TCPTP and HePTP.5,6,7
Along similar lines, 4C@241 TCPTP, an enzyme that contains a nonꢀsensitizing tetracysteine
motif that is not near its WPD loop, is not substantially inhibited by any of the biarsenicals.
For mutants that contain biarsenicalꢀbinding motifs in their respective WPD loops, the
degree of inhibition varied substantially between proteinꢀprobe combinations. Out of all
combinations tested, 4C TCPTP conjugated to Et2FlAsH resulted in the highest degree of
inhibition (72% reduction in catalytic efficiency, Fig. 6), although other biarsenical
compounds also inhibited the same sensitized PTP very effectively (Fig. 7, Table S4 ESI†).
Highꢀpotency inhibitors were also identified for P181C/E187C TCPTP: both Cl2FlAsHꢀ
EDT2 and Et2FlAsHꢀEDT2 inhibited this enzyme more effectively than FlAsHꢀEDT2. The
2’,7’ꢀsubstituted probes reduced the activity by 43% and 52%, respectively, whereas FlAsHꢀ
EDT2 reduced the activity by only 28%. Likewise, for P181C+2C TCPTP Et2FlAsHꢀEDT2
and F2FlAsHꢀEDT2 offer additional 35% and 23% reduction of activity, respectively
compared to FlAsHꢀEDT2. Similarly, in the case of 4C@211 HePTP the F2FlAsHꢀ and
Et2FlAsHꢀbound enzymes were more strongly inhibited than the FlAsHꢀbound protein.
Taken together, these data demonstrate that substitutions at positions 2′ and 7′ of the
9

Organic & Biomolecular Chemistry
Page 10 of 33
View Article Online
DOI: 10.1039/C4OB02256D
fluorescein scaffold with fluorine, chlorine atoms or ethyl chain serve to improve the
inhibition of sensitized WPDꢀloop PTP mutants.
Based on the results described above we selected the most promising pairs of PTP
mutants and biarsenical probes to test them using a phosphopeptide substrate. Peptide
substrates mimic the natural PTP substrates (phosphoproteins) more closely than the smallꢀ
molecule substrate pNPP.26 Moreover, an assay for inhibition of phosphopeptide
dephosphorylation provides a test for the substrateꢀindependence of biarsenicalꢀinduced PTP
inhibition. To measure PTP activity we used a continuous PTPꢀchymotrypsin coupled assay
in which increased fluorescence is caused by a peptideꢀcleavage event that can occur on nonꢀ
phosphorylated peptide (PTP product), but not on phosphopeptide (PTP substrate). Although
this assay has previously been used in a stopped format, in which the PTP inhibitor sodium
orthovanadate and chymotrypsin are added at the end of the PTP reaction, we found that the
assay worked well in continuous format. Under our continuousꢀassay conditions, PTP activity
is rateꢀlimiting; therefore the observed rate of fluorescence increase is proportional to PTP
concentration. By measuring the activities of enzymes incubated in various molar ratios of
biarsenical probe to enzyme (Fig. 8), we were able to determine the values of dissociation
constant for the 2’,7’ꢀsubstituted biarsenical probes (Table 2). Similarly to results obtained
with pNPP as a substrate, wildꢀtype TCPTP was not inhibited in this assay by any of the
biarsenical probes. In the case of the double cysteine mutant (P181C/E187C TCPTP) there is
no significant difference between FlAsHꢀEDT2 and 2’,7’ꢀsubstituted probes in dissociation
constant. However, when more cysteines are present (P181C+2C and 4C TCPTP), F2FlAsHꢀ
EDT2 demonstrates substantially improved inhibitory capabilities in comparison with FlAsHꢀ
EDT2. F2FlAsHꢀEDT2 not only inhibits the enzyme to higher extent at a 4:1 ratio of
biarsenical probe to protein but also has 14ꢀ (P181C+2C) and 9ꢀfold (4C) higher affinities
toward the target PTPs (Table 2). These values corroborate results obtained using pNPP as a
10

Page 11 of 33
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C4OB02256D
substrate, showing that improved inhibition is substrateꢀindependent, as would be expected if
the biarsenical dye blocks the conformational changes of the WPD loop that are required for
catalysis.9,10 Since Et2FlAsHꢀEDT2 also provided improved reduction of catalytic efficiency,
the molecular basis can be explained with two factors. First, the 2’,7’ꢀsubstituted biarsenical
probes can potentially induce steric hindrance during closing of the WPD loop, due to the
sheer size of the compounds. Second, the presence of halogen atoms may result in the
increased Lewis acidity of the compounds’ arsenic atoms, thereby increasing the stability of
binding with the engineered cysteine residues of the target proteins.
Collectively, our results based on both pNPP and phosphopeptide assays confirm that
cysteineꢀenriched WPD loops can be effectively targeted and that substantial improvements
in inhibition can be realized by using optimized biarsenical probes.
Conclusions
In conclusion, we have developed a library of known and new biarsenical probes and
systematically characterized their physicochemical properties, many of which were
previously unknown. In the course of these studies we found that some biarsenical probes are
unstable in aqueous buffered or DMSO solution, which can affect their longꢀterm application
and storage. Our biarsenical library was screened against a panel of wildꢀtype and sensitized
protein tyrosine phosphatases (PTPs) in order to identify optimal pairs of engineered PTPs
and biarsenicals for selective and potent inhibition of inhibitorꢀsensitized PTPs. Based on
inhibition assays with both smallꢀmolecule and phosphopeptide substrates, we found that the
2′,7′ꢀsubstituted biarsenical probes Cl2FlAsHꢀEDT2, F2FlAsHꢀEDT2 and Et2FlAsHꢀEDT2
are more effective inhibitors of sensitized PTPs than the previously investigated compound
FlAsHꢀEDT2, with F2FlAsHꢀEDT2 being the most effective inhibitor identified to date. Our
results show that substantial improvements in inhibitor effectiveness can be realized through
11

Organic & Biomolecular Chemistry
Page 12 of 33
View Article Online
DOI: 10.1039/C4OB02256D
the systematic investigation of biarsenical/enzyme pairs and suggest that future strategies for
controlling protein activity with biarsenicals should give considerable attention to
optimization of both components of the conjugate, protein and inhibitor.
Experimental procedures
Materials
Unless otherwise noted, all chemicals and reagents were obtained from commercial suppliers
(SigmaꢀAldrich, Merck, Roth, Avantor Performance Materials, or Iris Biotech) at the highest
quality available and were used without further purification. Buffers were prepared using
MilliꢀQ water (Merck Millipore system). Fluorescein, 5ꢀcarboxyfluorescein, resorcinol, 4ꢀ
chlororesorcinol, 4ꢀethylresorcinol, phthalic anhydride, 4,5,6,7ꢀtetrachlorophthalic anhydride,
resorufin were from SigmaꢀAldrich, 4ꢀfluororesorcinol, 4,5,6,7ꢀtetrafluorophthalic anhydride
were from Fluorochem and were used directly as received. NMP, DMSO and DIEA from Iris
Biotech were dried and stored over 3Å molecular sieve.
Synthesis of biarsenical probes
The biarsenical probes used in this study were prepared based on a general synthesis protocol
published previously.12,19 The first step of fluoresceinꢀbased probe synthesis involved the
reaction of the appropriate derivative of phthalic anhydride and resorcinol in methanesulfonic
o
acid (for details see ESI†). After stirring at 70 C for 12 h, reactions were terminated by
o
addition of cold water, and the precipitates were collected by centrifugation at 4 C. All
synthesized or commercially purchased fluorescein or resorufin platforms were subsequently
converted to 4′,5′ꢀdimercury derivatives. Probe synthesis was completed by palladiumꢀ
catalyzed transmetalation of the mercuric derivatives with AsCl3 and capping of the arsenic
12

Page 13 of 33
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C4OB02256D
atoms with 1,2ꢀethanedithiol (EDT). Biarsenical probes were purified by FLASH
chromatography using a gradient of ethyl acetate or methanol in toluene (Gilson PLC 2020
using Interchim PuriFlash SIHP 30 µm, 20 g column, see ESI† for details). Fractions with at
least 95% purity, as estimated by HPLC described below, were collected. The pooled
fractions were concentrated on a rotary evaporator, aliquoted, and dried by Speed Vac. Once
solvent was evaporated, tubes were immediately removed to avoid degradation by overꢀ
drying.27 Because of that solid contained small amounts of solvents used during purification,
which prohibited the estimation of the synthesis yield of biarsenical derivatives. Identification
and purity assessment of the fractions and final product were carried out on by ESIꢀMS
(Applied Biosystems API 2000) and HPLC (Dionex Ultimate 3000 using a Phenomenex
Aeris Peptide C18 column, see ESI† for details). All biarsenical probes were stored as solid
in ꢀ20ºC. Prior to use probes were dissolved in anhydrous DMSO.
4’,5’ꢀbis(1,3,2ꢀdithiarsolanꢀ2ꢀyl)ꢀfluorescein (FlAsHꢀEDT2).
An orange solid was obtained. Analysis on HPLC (Phenomenex Aeris Peptide 3.6 µm C18):
tR = 14.9 min. with gradient from 100 : 0 to 10 : 90 (5 mM NH4CO3 : ACN) in 20 min.
showed purity of 99 %. 1H NMR (500 Hz, CDCl3, ppm): 3.60 (m, 8H, SꢀCH2), 6.52 (d, 2H, J
= 10 Hz, Hꢀ2’, 7’), 6.63 (d, 2H, J = 10 Hz, Hꢀ1’, 8’), 7.20 (d, 1H, J = 5 Hz, Hꢀ7), 7.63 (t, J =
5 Hz, Hꢀ6), 7.69 (t, J = 5 Hz, Hꢀ5), 8.02 (d, 1H, J = 5 Hz, Hꢀ4). 13C NMR (CDCl3): 43.45,
110.68, 112.39, 114.87, 125.29, 128.22, 129.03, 130.78, 135.06, 137.87, 152.50, 162.82,
169.07. MS (m/z) calculated for C24H18As2O5S4 [MꢀH]ꢀ 662.8, found 663.3.
2’,7’ꢀdifluoroꢀ4’,5’ꢀbis(1,3,2ꢀdithiarsolanꢀ2ꢀyl)ꢀfluorescein (F2FlAsHꢀEDT2).
An orange solid was obtained. Analysis on HPLC (Phenomenex Aeris Peptide 3.6 µm C18):
tR = 10.7 min. with gradient from 100 : 0 to 10 : 90 (5 mM NH4CO3 : ACN) in 20 min.
13

Organic & Biomolecular Chemistry
Page 14 of 33
View Article Online
DOI: 10.1039/C4OB02256D
showed purity of 97 %. 1H NMR (500Hz, CDCl3, ppm): 3.60 (m, 8H, SꢀCH2), 6.41 (d, 2H, J
= 10 Hz, Hꢀ1’, 8’), 7.21 (d, 1H, J = 5 Hz, Hꢀ4), 7.66 (t, J = 5 Hz, Hꢀ6), 7.72 (t, J = 5 Hz, Hꢀ
5), 8.03 (d, 1H, J = 5 Hz, Hꢀ7). 13C NMR (500 Hz, CDCl3, ppm): 43.39, 108.76, 115.28,
115.44, 123.66, 125.27, 126.32, 127.97, 128.78, 130.19, 132.34, 134.19, 135.17, 145.29 (d,
JCF = 240 Hz), 154.61 (d, JCF = 246 Hz), 151.15, 168.20.. MS (m/z) calculated for
C24H18As2F2O5S4 [MꢀH]ꢀ 699.8, found 699.3.
2’,7’ꢀdichloroꢀ4’,5’ꢀbis(1,3,2ꢀdithiarsolanꢀ2ꢀyl)ꢀfluorescein (Cl2FlAsHꢀEDT2).
A red solid was obtained. Analysis on HPLC (Phenomenex Aeris Peptide 3.6 µm C18): tR =
9.7 min. with gradient from 100 : 0 to 10 : 90 (5 mM NH4CO3 : ACN) in 20 min. showed
purity of 90 %. 1H NMR (500 Hz, CDCl3, ppm): 3.62 (m, 8H, SꢀCH2), 7.21 (d, 1H, J = 5Hz,
13
Hꢀ4), 7.70 (t, J = 5 Hz, Hꢀ6), 7.75 (t, J = 5 Hz, Hꢀ5), 8.06 (d, 1H, J = 5 Hz, Hꢀ7). C NMR
(500 Hz, CDCl3, ppm): 43.42, 110.91, 113.80, 118.42, 123.65, 125.04, 127.97128.78,
130.07, 135.31, 157.54. MS (m/z) calculated for C24H18As2Cl2O5S4 (monoisotopic peak) [Mꢀ
H]ꢀ 730.8, found 731.1.
2’,7’ꢀdiethylꢀ4’,5’ꢀbis(1,3,2ꢀdithiarsolanꢀ2ꢀyl)ꢀfluorescein (Et2FlAsHꢀEDT2).
A red solid was obtained. Analysis on HPLC (Phenomenex Aeris Peptide 3.6 µm C18): tR =
16.1 min. with gradient from 100 : 0 to 10 : 90 (5 mM NH4CO3 : ACN) in 20 min. showed
purity of 96 %. 1H NMR (500 Hz, CDCl3, ppm): 1.0 (t, 3H, J = 5 Hz), 2.40 (m, 2H), 3.58 (m,
8H, SꢀCH2), 7.21 (d, 1H, J = 5 Hz, Hꢀ4), 7.64 (t, J = 5 Hz, Hꢀ6), 7.69 (t, J = 5 Hz, Hꢀ5), 8.04
(d, 1H, J = 5 Hz, Hꢀ7). 13C NMR (500 Hz, CDCl3, ppm): 14.06, 23.56, 43.66, 110.13, 112.14,
128.47, 129.28, 129.74, 135.22, 160.6. MS (m/z) calculated for C28H26As2O5S4 [MꢀH]ꢀ 718.9,
found 719.3.
14

Page 15 of 33
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C4OB02256D
4,5,6,7ꢀtetrachloroꢀ4’,5’ꢀbis(1,3,2ꢀdithiarsolanꢀ2ꢀyl)ꢀfluorescein (F4FlAsHꢀEDT2).
An orange solid was obtained. Analysis on HPLC (Phenomenex Aeris Peptide 3.6 µm C18):
tR = 11.6 min. with gradient from 100 : 0 to 10 : 90 (5 mM NH4CO3 : ACN) in 20 min.
showed purity of 97 %.1H NMR (500 Hz, CDCl3, ppm): 3.57 (m, 8H, SꢀCH2), 6.60 (d, 2H, J
= 10 Hz, Hꢀ2’, 7’), 6.78 (d, 2H, J = 10 Hz, Hꢀ1’, 8’). 13C NMR (500 Hz, CDCl3, ppm): 43.29,
107.59, 112.81, 115.00, 125.04, 127.97, 128.78, 134.30, 135.27 (m), 141.04 (m), 143.35 (dd,
JCF =250, 4 Hz), 146.92 (dd, JCF = 255 Hz, 4 Hz), 151.79, 163.17. MS (m/z) calculated for
C24H14As2F4O5S4 [MꢀH]ꢀ 734.8, found 735.1.
4,5,6,7ꢀtetrachloroꢀ4’,5’ꢀbis(1,3,2ꢀdithiarsolanꢀ2ꢀyl)ꢀfluorescein (Cl4FlAsHꢀEDT2).
A red solid was obtained. Analysis on HPLC (Phenomenex Aeris Peptide 3.6 µm C18): tR =
8.7 min. with gradient from 100 : 0 to 10 : 90 (5 mM NH4CO3 : ACN) in 20 min. showed
1
purity of 97 %. H NMR (500 Hz, CDCl3, ppm): 3.56 (m, 8H, SꢀCH2), 6.58 (d, 2H, J = 10
Hz, Hꢀ2’, 7’), 6.75 (d, 2H, J = 10 Hz, Hꢀ1’, 8’). 13C NMR (500 Hz, CDCl3, ppm): 43.30,
107.09, 112.63, 115.08, 122.46, 125.29, 128.21, 129.02, 129.38, 149.35, 152.31, 163.11. MS
(m/z) calculated for C24H14As2Cl4O5S4 (monoisotopic peak) [MꢀH]ꢀ 798.7, found 799.1.
5ꢀCarboxyꢀ4’,5’ꢀbis(1,3,2ꢀdithiarsolanꢀ2ꢀyl)ꢀfluorescein (5ꢀCrAsHꢀEDT2).
An orange solid was obtained. Analysis on HPLC (Phenomenex Aeris Peptide 3.6 µm C18):
tR = 9.8 min. with gradient from 100 : 0 to 10 : 90 (5 mM NH4CO3 : ACN) in 20 min. showed
purity of 97 %.1H NMR (500 Hz, CDCl3, ppm): 3.60 (m, 8H, SꢀCH2), 6.53 (d, 2H, J = 10 Hz,
Hꢀ2’, 7’), 6.62 (d, 2H, J = 10 Hz, Hꢀ1’, 8’), 8.00 (d, 1H, J = 10Hz, Hꢀ7), 8.40 (d, 1H, J = 10
Hz, Hꢀ6), 8.72 (s, 1H, Hꢀ4). 13C NMR (500 Hz, CDCl3, ppm): 15.50, 17.91, 29.90, 30.91,
15

Organic & Biomolecular Chemistry
Page 16 of 33
View Article Online
DOI: 10.1039/C4OB02256D
43.73, 49.76, 66.10, 110.21, 112.83, 115.29, 124.51, 127.54, 130.89, 136.74, 152.63, 168.34.
MS (m/z) calculated for C25H18As2O7S4 [MꢀH]ꢀ 706.8, found 707.3.
5ꢀ(Nꢀbutylamide)ꢀ4’,5’ꢀbis(1,3,2ꢀdithiarsolanꢀ2ꢀyl)ꢀfluorescein (5ꢀBuCrAsHꢀEDT2).
An orange solid was obtained. Analysis on HPLC (Phenomenex Aeris Peptide 3.6 µm C18):
tR = 9.7 min. with gradient from 100 : 0 to 10 : 90 (5 mM NH4CO3 : ACN) in 20 min. showed
1
purity of 95 %. H NMR (500 Hz, DMSO, ppm): 0.92 (t, 3H, J = 5 Hz), 1.35 (m, 2H), 1.54
(m, 2H), 2.18 (t, 2H, J = 5 Hz), 3.30 (m, 8H, SꢀCH2), 6.58 (d, 2H, J = 10 Hz, Hꢀ2’, 7’), 6.72
(d, 2H, J = 10 Hz, Hꢀ 1’, 8’), 7.2 (m, 2H, Hꢀ5, 6), 8.25 (d, 1H, J = 10 Hz, Hꢀ7), 8.39 (d, 1H, J
= 10 Hz, Hꢀ6), 8.45 (s, 1H, Hꢀ4). 13C NMR (500 Hz, DMSOꢀd6, ppm): 13.68, 17.20, 30.08,
39.50, 42.13, 113.29 118.42, 121.42, 128.19, 128.88, 129.61, 132.06, 142.62, 147.32, 156.58,
164.13, 172.02. MS (m/z) calculated for C29H27As2NO6S4 [MꢀH]ꢀ 763.9, found 763.9.
4,5ꢀBis(1,3,2ꢀdithiarosolanꢀ2ꢀyl)ꢀresorufin (ReAsHꢀEDT2).
An orange solid was obtained. Analysis on HPLC (Phenomenex Aeris Peptide 3.6 µm C18):
tR = 13.3 min. with gradient from 100 : 0 to 10 : 90 (5 mM NH4CO3 : ACN) in 20 min.
showed purity of 97 %.1H NMR (500 Hz, CDCl3, ppm): 3.54 (m, 8H, SꢀCH2), 6.22 (d, 2H, J
= 10 Hz, Hꢀ2, 7), 6.30 (d, 2H, J = 10 Hz, Hꢀ1, 8). 13C NMR (500 Hz, CDCl3, ppm): 14.11,
29.69, 31.92, 33.46, 125.29. MS (m/z) calculated for C29H27As2NO6S4 [MꢀH]ꢀ 543.8, found
544.2.
Peptide synthesis
Tetracysteine peptides AcꢀFLNCCPGCCMEPꢀamide (TC12) and AcꢀCCPGCCꢀamide (TC6)
were synthesized by solid phase synthesis using the Fmoc strategy. Tenta Gel R Ram resin
was utilized for amide peptides and 2ꢀchlorotrityl for carboxyl peptides. Peptides were
16

Page 17 of 33
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C4OB02256D
synthesized in Liberty 1 microwaveꢀassisted synthesizer (CEM). Couplings of amino acids
were performed with 3 eq. of NꢀαꢀFmocꢀprotected amino acid, HBTU (3 eq.) and DIEA (5
eq.) in DMF. Peptides were terminated by acetylation with Ac2O. For that purpose resin was
mixed with 4 eq. of Ac2O, 4 eq. of DIEA in DMF for 4 h. Peptide cleavage was achieved
with mixture of 90% of TFA, 5% thioanisole, 3% anisole and 2% 1,2ꢀethanedithiol over 1.5
h, followed by precipitation in cold (ꢀ80°C) diethyl ether. Crude peptide pellets were
collected by centrifugation. Peptides were purified on HPLC (Dionex Ultimate 3000) using
semiꢀpreparative Phenomenex GeminiꢀNX C18 column and gradient of 0.1% TFA in
acetonitrile with 0.1% TFA. The purified peptide was identified by ESI mass spectrometry
using API 2000 (Applied Biosystems) instrument. MS (m/z) calculated for TC12 and TC6
[M+H]+ were 1358.7, 626.8 and found 1358.4, 626.6, respectively.
Physicochemical properties of biarsenical probes
Electronic spectra were obtained on a Jasco Vꢀ650 spectrophotometer. Fluorescence was
recorded on a Horiba Scientific FluoroMaxꢀ4 spectrofluorimeter. All measurements were
recorded in 50 mM Na+ꢀHEPES buffer with 100 mM NaCl and 200 µM TCEP at pH 7.4,
25oC. Excitation wavelengths were chosen based on the biarsenicalꢀprobe complex’s
maximal absorption. All probe−peptide conjugates were prepared with the optimized
tetracysteine peptide, TC12.15 Probe−peptide conjugates were obtained by initial incubation
of 10 µM biarsenical probe with 15 µM TC peptide. All spectra were recorded after 2 h.
Determination of pKa values.
Solutions containing 1 µM biarsenical probe with 6 µM TC12 peptide in 50 mM Na+ꢀborate
buffer with 50 µM TCEP at pH 10.0 were incubated at room temperature until no increase of
fluorescence was observed. Samples were then titrated with HCl, and the resulting and pH
17

Organic & Biomolecular Chemistry
Page 18 of 33
View Article Online
DOI: 10.1039/C4OB02256D
and fluorescence intensities were measured and fitted using Hill’s equation (Eq. S1, ESI†) in
order to obtain pKa values and Hill coefficients for the phenolic group at position 3′. When
TC6 peptide was used the same methodology was applied and data was fitted to Hill’s
equation (Eq. S2, ESI†)
Stability of biarsenical probes.
Solutions of 25 µM biarsenical probes in 50 mM Na+ꢀHEPES buffer and 100 mM NaCl at pH
7.4 were incubated in darkness over periods of 1, 2, 3 or 6 h and analyzed by analytical
HPLC (Dionex Ultimate 3000 using column C18 Phenomenex Aeris Peptide) using a 5 mM
ammonium carbonate/acetonitrile gradient. Since monoarsenical and biarsenical species have
absorption maxima that differ by ~10 nm, the percent composition of the two species was
estimated by comparing the peak areas from chromatograms collected at 280 nm.
Protein expression and purification
Plasmids encoding the wildꢀtype and mutant forms of sixꢀhistidine tagged TCPTP and
HePTP were described previously (see Table 2 for mutant aminoꢀacid sequences).5ꢀ7 The
appropriate plasmids were used to transform BL21(DE3)ꢀcodonPLUSꢀRIL E. coli. Single
colonies were picked and used to inoculate 1000 ml LB cultures, which were grown to midꢀ
log phase at 37°C, and induced with either 0.2 mM IPTG (TCPTP) or 0.04% arabinose
o
(HePTP) for 16 h at 26°C. Bacterial cells were centrifuged and frozen at ꢀ20 C. Cell lysis
was achieved by pellet sonication on ice at 80 W for 5 seconds with 10 second intervals.
Enzyme purification was carried out using NiꢀNTA Agarose (Qiagen) according to
manufacturer’s instructions. Purified proteins were concentrated on 7 kDa cutꢀoff CentriPrep
Centrifugal Filter Devices (Milipore) and exchanged into pH 7.0 buffer containing 50 mM
3,3ꢀdimethylglutarate, 1 mM EDTA, 150 mM NaCl, and 1 mM dithiothreitol (DTT), flashꢀ
18

Page 19 of 33
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C4OB02256D
frozen in liquid nitrogen, and stored at ꢀ80°C. Protein purity was assessed by SDSꢀPAGE.
Protein concentrations were determined spectrophotometrically utilizing Bradford method
using BSA as a standard.28
PTP activity assay with pNPP substrate
Solutions of TCPTP or HePTP (0.24 µM) in PTP activity buffer (50 mM 3,3ꢀdimethylglutaric
acid pH 7.0, 50 mM NaCl, 1 mM EDTA with 20 µM TCEP) were incubated in darkness with
or without biarsenical probe (0.96 µM) in protein lowꢀbinding tubes (Eppendorf). The final
concentration of DTT from storage buffer was less than 1 µM. After 2.5 h incubation at room
temperature, the enzymatic activity was assayed using paraꢀnitrophenylphosphate (pNPP).
Reactions were carried out in a total volume of 200 µL at 22°C. To a solution of 0.25ꢀ10 mM
pNPP in PTP activity buffer, 100 nM of appropriate PTP enzyme was added. After 3ꢀ8 min,
depending on the enzyme activity, reactions were quenched by the addition of 40 µL of 5 M
NaOH. The absorbance of the hydrolyzed product was measured at 405 nm in a microplate
reader (Asys UVM 340, Biochrom). All experiments were performed in triplicates.
Enzymatic kinetic parameters were obtained by nonlinear regression fitted to the Michaelisꢀ
Menten equation using Origin 8.6 software. The calculation of catalytic efficiency error is
described in Supplementary information (ESI†).
PTP activity assays with phosphorylated peptide substrate
PTP activities were assayed using the fluorogenic (FRET) phosphorylated peptide MCAꢀGlyꢀ
AspꢀAlaꢀGluꢀTyr(PO3H2)ꢀAlaꢀAlaꢀLys(DNP)ꢀArgꢀamide (SigmaꢀAldrich, SCP0229) in a
modified PTPꢀchymotrypsin coupled assay.26 MCA and DNP correspond to 7ꢀ
methoxycumarinꢀ4ꢀacetyl (donor) and 2,4ꢀdinitrophenyl (quencher) moiety, respectively. The
assay, which monitors the fluorescence increase induced by peptide cleavage once the
19

Organic & Biomolecular Chemistry
Page 20 of 33
View Article Online
DOI: 10.1039/C4OB02256D
peptide is dephosphorylated, was performed using enzymes with and without biarsenical
probes. To a cuvette containing PTP activity buffer with 50 µM TCEP, 500 nM
phosphorylated peptide (dissolved in DMSO) was added. Next, chymotrypsin in PTP activity
buffer (final concentration 0.05% (w/v) – 20 µM) was added and the reaction was monitored
on a Fluoromaxꢀ4 spectrofluorimeter (excitation: 328 nm, emission: 395 nm). Once
chymotrypsin had cleaved all of the dephosphorylated peptide present in the phosphopeptide
solution, TCPTP was added (wildꢀtype or mutant, with or without a biarsenical probe) to a
final concentration of 1 nM. Prior to the measurement the enzyme was incubated for 2.5 h
with in molar ratios of TCPTP to biarsenical probes of 1 to 0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 1, 2, 4
or 8, prepared as described above. All experiments were performed in triplicates.
Acknowledgements
This research was supported by the National Science Centre (NCN) under Opus grant, No.
2011/01/B/ST5/00830 (A.K.) and partially by 2012/07/N/NZ1/03079 (A.P.). A.P., J.A., A.K,
and instrumentation were supported by the Polish Foundation for Science under Focus
projects F1/2010, F1/2010/P/2013, FG1/2010. A.C.B. acknowledges the National Institute of
General Medical Sciences of the National Institutes of Health (R15GM071388) and the
Henry Dreyfus TeacherꢀScholar Awards Program for funding. We thank Tomasz
Kochańczyk for help in figure preparation.
20

Page 21 of 33
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C4OB02256D
Table 1. Spectral and chemical properties of biarsenical probes conjugated to the peptide Acꢀ
FLNCCPGCCMEPꢀamide (TC12). Superscripted numbers indicate references for previously
published values.
λex / λem
(nm)
pKa of parent
fluorophore
Probeꢀpeptide complex
FlAsHꢀTC12
pKa1
pKa2
509/530
(511/527)20
5.07 ± 0.01
(5.4)8
6.71 ± 0.06
6.4321
510/530
3.66 ± 0.02
6.24 ± 0.04
4.816
(500/522)14
519/539
(520/539)12
F2FlAsHꢀTC12
3.32 ± 0.03
4.93 ± 0.01
4.95 ± 0.02
5.02 ± 0.01
5.36 ± 0.01
5.19 ± 0.01
4.180 ± 0.007
6.36 ± 0.08
7.09 ± 0.04
7.2 ± 0.1
4.7217
6.61 ± 0.0318
6.116
Cl2FlAsHꢀTC12
Et2FlAsHꢀTC12
F4FlAsHꢀTC12
Cl4FlAsHꢀTC12
5ꢀCrAsHꢀTC12
5ꢀBuCrAsHꢀTC12
ReAsHꢀTC12
521/542
528/545
(528/544)14
530/546
6.96 ± 0.08
7.6 ± 0.1
5.9717
6.422
511/536
(ꢀ/534)a,13
512/538
6.74 ± 0.05
6.21 ± 0.02
ꢀ
597/606
(597/608)15
5.823
a Theses values were determined using peptides with unoptimized TC sequence ꢀ CCXXCC.
21

Organic & Biomolecular Chemistry
Page 22 of 33
View Article Online
DOI: 10.1039/C4OB02256D
Table 2. Dissociation constants of biarsenicalꢀPTP enzyme complexes. The values were
determined by doseꢀdependent inhibition of the indicated probeꢀresponsive TCPTP mutants.
The enzyme (final concentration 1 nM) was incubated for 2.5h with 0.05, 0.1, 0.2, 0.3, 0.4,
0.5, 1, 2, 4 or 8 equivalents of biarsenical probe. Experimental data, partially shown in Fig. 8
were fitted to Equation S6 (ESI†)
P181C/E187C
(nM)
P181C+2C
(nM)
4C
Biarsenical probe
FlAsHꢀEDT2
(nM)
0.60 ± 0.09
1.1 ± 0.3
1.2 ± 0.3
0.8 ± 0.2
2.1 ± 0.3
0.15 ± 0.04
0.9 ± 0.2
0.8 ± 0.1
0.09 ± 0.03
1.2 ± 0.4
F2FlAsHꢀEDT2
Cl2FlAsHꢀEDT2
Et2FlAsHꢀEDT2
0.6 ± 0.2
1.1 ± 0.3
22

Page 23 of 33
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C4OB02256D
Figure 1. Biarsenical probe acting as an allosteric inhibitor. (A) Spatially arranged cysteines
are introduced (light green) in the WPD loop of a protein tyrosine phosphatase (PTP). The
mutation causes only limited changes in the catalytic efficiency of the enzyme. However,
when the biarsenical probe FlAsH is conjugated, the enzyme is strongly inhibited. (B) Amino
acid sequences of wild type and cysteine mutants of TCPTP and HePTP used in this study.
Red letters highlight amino acid residues substituted or added to the wild type protein
sequence. Protein model was based on structure of TCPTP (PBD: 1L8K).
23

Organic & Biomolecular Chemistry
Page 24 of 33
View Article Online
DOI: 10.1039/C4OB02256D
S
S
S
S
As
As
S
S
S
S
HO
R
O
O
R
As
As
HO
O
N
O
COOH
X
X
X
ReAsH-EDT2
Z
R,X = H Z = COOH
R = H X,Z = F
5-CrAsH-EDT2
F4FlAsH-EDT2
Cl4FlAsH-EDT2
5-BuCrAsH-EDT2
R,X,Z = H
FlAsH-EDT2
R = F X,Z = H
R = Cl X,Z = H
F2FlAsH-EDT2
Cl2FlAsH-EDT2
R = H X,Z = Cl
R = C2H5 X,Z = H
Et2FlAsH-EDT2
R,X = H Z = CO-NH-C4H9
Figure 2. Structures of biarsenical EDTꢀcapped probes based on different parent
fluorophores used in this work.
24

Page 25 of 33
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C4OB02256D
X
O
HO
O
O
X
Z
HO
OH
60 oC, 15 h
O
+
2x
R
R
CH3SO3H
R
X
X
COOH
O
X
X
Z
HgO
25 oC, 4 h
CF3
CF3COOH
O
F3C
O
S
S
S
S
As
As
O
O
Hg
Hg
HO
R
O
O
R
70 oC, 3 h
HO
R
O
O
R
1. AsCl3, DIEA
Pd(OAc)2, NMP
COOH
X
X
X
2. EDT
X
X
COOH
X
Z
Z
Figure 3. Schematic pathway for the synthesis of biarsenical probes built on the fluorescein
platform. Stages include the synthesis of parent the fluorophore, its subsequent mercurization,
transmetalation with AsCl3, and final capping with EDT. It should be noted that first reaction
results in forming of two isomers, where Z is located at position 5 or 6 in the product. Since
the probes used in this study are single isomers the Z was placed at position 5.
25

Organic & Biomolecular Chemistry
Page 26 of 33
View Article Online
DOI: 10.1039/C4OB02256D
HO
O
O
HO
O
O
COOH
COOH
25 oC, 15 h
+
H2N
HBTU, DIEA
EDT
COOH
O
N
H
Figure 4. Synthesis of 5ꢀ(Nꢀbutylamide)ꢀfluorescein, a parent platform used for the synthesis
of 5ꢀBuCrAsHꢀEDT2.
26

Page 27 of 33
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C4OB02256D
15
10
5
1h
3h
6h
0
Figure 5. Decomposition of biarsenical probes in aqueous pH buffered solution. 25 ꢁM
biarsenical probes were incubated for 1, 3 and 6 h in 50 mM Na+ꢀHEPES buffer, 100 mM
NaCl, pH 7.4. The percentage of monoarsenical species was determined using analytical
HPLC. Details can be found in Table S1 (ESI†).
27

Organic & Biomolecular Chemistry
Page 28 of 33
View Article Online
DOI: 10.1039/C4OB02256D
15
10
5
0
0
2
4
6
8
10
pNPP (mM)
Figure 6. MichaelisꢀMenten plot for TCPTP 4C incubated without biarsenical probe (black),
with FlAsHꢀEDT2 (blue) or Et2FlAsHꢀEDT2 (red). Assay was performed at various
concentrations of paraꢀnitrophenylphosphate (pNPP) using 100 nM enzyme or enzymeꢀ
biarsenical complex as described in Materials and Methods. Data was fitted to the Michaelisꢀ
Menten equation.
28

Page 29 of 33
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C4OB02256D
Figure 7. Graphical representation of catalytic efficiency (kcat/Km, sꢀ1 mMꢀ1) values for PTP
mutants in the presence of 4 equivalents of biarsenical probes and assayed with pNPP. Blue
color demonstrate the typical enzyme activity (not inhibited) and red complete inhibition of
activity. Gradient color is directly proportional to the degree of inhibition of the enzyme. kcat,
Km and kcat/Km values are listed with their errors in Table S2ꢀS4 (ESI†).
29