Synthesis of a phosphoserine mimetic prodrug with potent 14-3-3 protein inhibitory activity

Article abstract of DOI:10.1016/j.chembiol.2012.05.011

Many protein-protein interactions in cells are mediated by functional domains that recognize and bind to motifs containing phosphorylated serine and threonine residues. To create small molecules that inhibit such interactions, we developed methodology for the synthesis of a prodrug that generates a phosphoserine peptidomimetic in cells. For this study, we synthesized a small molecule inhibitor of 14-3-3 proteins that incorporates a nonhydrolyzable difluoromethylenephosphoserine prodrug moiety. The prodrug is cytotoxic at low micromolar concentrations when applied to cancer cells and induces caspase activation resulting in apoptosis. The prodrug reverses the 14-3-3-mediated inhibition of FOXO3a resulting from its phosphorylation by Akt1 in a concentration-dependent manner that correlates well with its ability to inhibit cell growth. This methodology can be applied to target a variety of proteins containing phosphoserine and other phosphoamino acid binding domains.

Full text of DOI:10.1016/j.chembiol.2012.05.011

Chemistry & Biology
Article
Synthesis of a Phosphoserine Mimetic Prodrug
with Potent 14-3-3 Protein Inhibitory Activity
Allison Arrendale,1,2 Keunho Kim,1,2 Jun Young Choi,1 Wei Li,1 Robert L. Geahlen,1 and Richard F. Borch1,
1Department of Medicinal Chemistry and Molecular Pharmacology and the Center for Cancer Research, Purdue University,
West Lafayette, IN 47907, USA
*
2These authors contributed equally to this work
*Correspondence: borch@purdue.edu
DOI 10.1016/j.chembiol.2012.05.011
SUMMARY
phosphate group is difficult because of the lipophilic cell
membrane (Xie et al., 2003). Once the drug enters the cell, it is
Many protein-protein interactions in cells are medi- subject to dephosphorylation by phosphatases, rendering it
inactive.
ated by functional domains that recognize and bind
Numerous strategies have been reported for the design of
phosphate and phosphonate prodrugs, most of them intended
to increase oral bioavailability of the highly charged group.
to motifs containing phosphorylated serine and
threonine residues. To create small molecules that
inhibit such interactions, we developed methodology
for the synthesis of a prodrug that generates a phos-
phoserine peptidomimetic in cells. For this study,
we synthesized a small molecule inhibitor of 14-3-3
proteins that incorporates a nonhydrolyzable difluor-
Pivaloxymethyl (POM) esters have been utilized extensively in
the design of nucleotide prodrugs, and this strategy was used
recently to enhance the cellular permeability of a phosphonate-
based inhibitor of the mitotic regulator Pin1 (Zhao and Etzkorn,
2007). Cyclic phosphotriesters incorporating serine-containing
omethylenephosphoserine prodrug moiety. The pro-
drug is cytotoxic at low micromolar concentrations
peptides have also been used to enhance oral bioavailability
and cell permeability for antiviral acyclonucleoside phospho-
when applied to cancer cells and induces caspase nates such as Cidofovir (Eriksson et al., 2008). We have devel-
oped a strategy for the preparation of phosphate and phospho-
nate prodrugs that are specifically activated intracellularly. The
activation resulting in apoptosis. The prodrug
reverses the 14-3-3-mediated inhibition of FOXO3a
phosphate bridging oxygen was replaced by the difluoromethy-
lene isostere to prevent phosphatase cleavage of the phosphate
group under physiological conditions. This moiety has the
resulting from its phosphorylation by Akt1 in a
concentration-dependent manner that correlates
well with its ability to inhibit cell growth. This method-
ology can be applied to target a variety of proteins
containing phosphoserine and other phosphoamino
acid binding domains.
appropriate pKa to deliver a doubly charged phosphate mimetic
to the binding site (Zheng et al., 2005). This methodology has
been shown to be successful in the development of nucleotide
prodrugs (Freel Meyers et al., 2000; Tobias and Borch, 2001),
inhibitors of phosphotyrosine binding proteins containing Src
homology 2 (SH2) domains (Garrido-Hernandez et al., 2006),
INTRODUCTION
and inhibitors of protein tyrosine phosphatase 1B (PTP1B)
(Boutselis et al., 2007).
We were interested in assessing whether this prodrug tech-
The effort to target protein-protein interactions (PPIs) has
increased in recent years because of reports revealing the nology was applicable to the development of phosphoserine
importance of many such interactions in the development and mimetics. Historically, phosphoserine/threonine sites were
progression of cancer. The arguments against targeting PPIs thought to function allosterically in contrast to phosphotyrosine
with small molecules include the often large protein surface sites that often mediate PPIs (Yaffe and Smerdon, 2001). In
areas involved and the large number of residues that interact recent years, several important PPIs have been discovered
and contribute to the free energy of binding (Lo Conte et al., that are mediated through phosphorylation of serines or threo-
1999). The dogma has been that small molecules cannot repli- nines. These include PPIs involving 14-3-3 proteins, WW
cate these various interactions and still maintain the advantages domains, Forkhead-associated (FHA) domains, and the WD40
of a small molecule. Many PPIs are mediated through phosphor- and LRR domains of F-box proteins (Yaffe and Smerdon,
ylation sites that interact with phosphotyrosine or phosphoser- 2001). For our initial efforts, we focused on 14-3-3 proteins.
ine/threonine recognition motifs. The free energy of binding in This highly conserved and ubiquitously expressed family of
these cases is contributed primarily by the phosphate-contain- proteins is involved in many vital cellular processes including
ing residue (Bradshaw et al., 1999). This important aspect makes signal transduction, apoptosis, cell-cycle regulation, growth,
these targets much more amenable to inhibition with small and development (Yaffe and Smerdon, 2001). There are seven
molecules. However, developing small molecule inhibitors of different isoforms in humans that interact to form homo or
these sites presents a number of serious obstacles. The intracel- heterodimers. Many of the functions of the 14-3-3 proteins are
lular delivery of drugs that contain the necessary doubly charged regulated by the phosphorylation of their binding partners
764 Chemistry & Biology 19, 764–771, June 22, 2012 ª2012 Elsevier Ltd All rights reserved

Chemistry & Biology
Phosphoserine Mimetic Inhibits 14-3-3 Protein
Figure 1. Synthesis of the a,a-Difluorome-
thyenephosphoserine Mimetic
Reagents and conditions: (a) MeSCF2P(O)(Oi-Pr)2,
t-BuLi, THF/HMPA, ꢀ78ꢁC, 55%; or (b)
HCF2P(O)(OEt)2, LDA, THF/HMPA, ꢀ78ꢁC, 75%;
(c) Boc2O, DMAP, THF; (d) Mg, MeOH, 91%; (e)
conc HCl, MeOH; (f) FmocCl, DIPEA, 92%; (g)
Jones reagent, acetone, ꢀ25ꢁC, 75%.
See Figure S1.
(Muslin et al., 1996). Two high-affinity binding motifs have been be applied to create a variety of inhibitors of important protein
identified: RSXpSXP (mode 1) and RXXXpSXP (mode 2) (Yaffe targets involved in phosphorylation-dependent interactions.
et al., 1997). 14-3-3 proteins bind to many targets including the
tumor suppressor p53, Raf family kinases, FOXO transcription RESULTS AND DISCUSSION
factors, the proapoptotic protein BAD, and Cdc25 phosphatases
through these recognition sequences. Upon binding, 14-3-3 acts Synthesis of Phosphoserine Mimetic Prodrug
as an adaptor, leading to changes in enzymatic activity or When we began to design our synthesis, we considered the
subcellular localization (Aitken, 2006). These characteristics construction of a,a-difluoromethylenephosphoserine analogs
make the 14-3-3 family an interesting target for cancer therapeu- that had been pioneered by Berkowitz and studied by other
tics (Hermeking, 2003).
groups (Berkowitz et al., 1994, 1996; Otaka et al., 1995; Yoko-
Although it has been established that 14-3-3 proteins are matsu et al., 1996; Kawamoto and Campbell, 1997). There are
involved in several serious disorders, much about the details of currently three general strategies known for accessing this
these mechanisms is unknown. The difluoromethylenephospho- moiety: displacement of a primary triflate with phosphonodifluor-
serine moiety may be used as a biological tool to construct omethyl carbanion (Berkowitz et al., 1994, 1996; Yokomatsu
a constitutive pCF2-Ser modification in a native protein or in et al., 1996), reaction of phosphonodifluoromethyl carbanion
the synthesis of small peptides/peptidomimetics to mimic with Garner’s aldehyde (Otaka et al., 1995), and transition
protein binding partners. Phosphatase-resistant phosphoserine metal-mediated alkene addition with a Cu reagent (Kawamoto
mimetics are now achievable, allowing for site-specific phos- and Campbell, 1997). Alternatively, an optically pure aziridine is
phorylation studies and examination of downstream conse- a useful synthon for the preparation of chiral amines, amino
quences of this event (Panigrahi et al., 2009). The utility of the alcohols, and amino acids. Synthesis of an enantiomerically
phosphoserine mimetic has been demonstrated by Cole and pure amino alcohol moiety could be achieved by selective
coworkers (Zheng et al., 2005; Szewczuk et al., 2008), who ring opening of chiral aziridines with various nucleophiles (Hu,
were interested in examining the regulation of arylalkylamine 2004). Oxirane ring opening by a,a-difluoromethylenephospho-
N-acetyltransferase (AANAT), which is responsible for cycle nate anion to furnish g-hydroxy-a,a-difluoromethylenephospho-
regulation of melatonin (Klein, 2007). Phosphorylation at Ser- nates has been reported (Ozouf et al., 2004; Pajkert et al., 2008);
205 was proposed to modulate the acyltransferase activity however, aziridine ring opening with the same anion has not
of AANAT (Zheng et al., 2005). The protein was hypothesized been reported to date. Thus, we speculated that the reaction
to be shielded from proteosomal proteolysis by association to of chiral aziridine with a,a-difluoromethylene diethylphospho-
14-3-3 z through pSer-205. The Cole group synthesized the nate anion might provide facile access to the desired nonhydro-
octapeptide containing a difluoromethyl phosphoserine mimetic lyzable phosphoserine core structure.
and used their EPL technology to look for an interaction between
We recently reported the preparation of enantiomerically en-
pSer-205 AANAT and its 14-3-3 z binding partner (Szewczuk riched aziridines via enzymatic desymmetrization of prochiral
et al., 2008). Through this construct they were able to demon- N-protected serinol (Choi and Borch, 2007). With N-Ts protected
strate that phosphorylation at this site results in binding to 14- chiral aziridine 1 in hand (Figure 1), ring opening of the aziridine
3-3 z and enhances cellular stability of AANAT protein. This was investigated (Fujii et al., 1994; Bergmeier and Seth, 1997).
example demonstrates that this unnatural amino acid is an Addition of the magnesium (Waschbu¨ sch et al., 1997) or
invaluable tool for the dissection of biological pathways.
cadmium (Qiu and Burton, 1996) reagent prepared from diethyl
Until recently, only small molecule stabilizers of 14-3-3 PPIs bromodifluoromethyl phosphonate was investigated without
were known. Fusicoccin and cotylenin have been used success- success. The anion prepared by reaction of diisopropyl methyl-
fully to study the role that 14-3-3 plays in plant physiology and thiodifluoromethylenephosphonate and t-butyllithium (Henry-
cancer (Ottmann et al., 2007). Another recent screen has dit-Quesnel et al., 2003) reacted with 1 to give the desired
identified more PPI stabilizers, but inhibitors of 14-3-3 PPIs are product 2a in 55% yield (Figure 1). However, because of the
still scarce (Rose et al., 2010). In this study, we describe the difficult preparation of the methylthio reagent and subsequent
synthesis and characterization of a nonhydrolyzable phospho- problems with selective cleavage of the isopropyl ester, we
serine peptidomimetic prodrug. This molecule, targeted against turned our attention to the generation and reaction of the lithium
14-3-3 family proteins, induces apoptosis in leukemia cells and anion via direct deprotonation of the difluoromethylenephosph-
reverses the 14-3-3-dependent inhibition of the proapoptotic onate. (Berkowitz et al., 1996) Extensive investigation of bases,
transcription factor FOXO3a. The strategy described here can solvents and reaction conditions gave only trace amounts of
Chemistry & Biology 19, 764–771, June 22, 2012 ª2012 Elsevier Ltd All rights reserved 765

Chemistry & Biology
Phosphoserine Mimetic Inhibits 14-3-3 Protein
Figure 2. Synthesis of the Phosphoserine Phosphonamidate Prodrug
Reagents and conditions: (a) Ac2O, DIPEA, DMAP (cat.), 97%; (b)TMSI; (c) oxalyl chloride; (d) 4-chloro-N-methylbutan-1-amine hydrochloride, DIPEA, 54%; (e)
(5-nitrofuran-2-yl) methanol, LHMDS, THF, ꢀ78ꢁC, 55%; (f) K2CO3, MeOH, ꢀ4ꢁC, 70%; (g) Jones reagent, ꢀ25ꢁC, acetone, 70%.
product 2a. Success was ultimately achieved by reaction of (SMM) screen. It showed promising results in cell-free assays
diethyl difluoromethylenephosphonate with LDA in THF/HMPA and appeared to be an appropriate candidate to demonstrate
followed by addition of aziridine 1. Because of the instability of proof of concept for the phosphoserine prodrug. The prodrug
the lithium anion at higher temperature, the reaction was main- 15 (Figure 3) was synthesized using phosphoserine prodrug
tained at ꢀ78ꢁC overnight, and ring-opened product 2b was precursor 10. PYBOP coupling of 10 with tyramine furnished
obtained in 75% yield along with a small amount of recovered 13; removal of the Fmoc protecting group afforded free amine
starting material. HMPA was the only cosolvent that provided 14, which was coupled with 11 (Figure 4) using PYBOP to furnish
reasonable yields of product, presumably because of its unique the desired inhibitor 15. In order to have the corresponding free
features as a dipolar aprotic solvent. Direct conversion of the ligand available for comparative cell-based studies, compound
tosyl to the Boc protecting group of 2 was achieved via the 19a was prepared (Figure 5). Tyramine was coupled with 5 using
doubly protected N-Boc N-Ts amine; brief sonication in the pres- PyBOP, followed by deprotection of Fmoc with DBU to give
ence of magnesium powder in anhydrous methanol (Nyasse coupling product 16. Because the azide in coupling partner 12
et al., 1997) smoothly furnished N-Boc protected amine 3. was unstable in the presence of TMSI, the diethyl phosphonate
Removal of both TBS and Boc protecting groups was accom- was converted to phosphonic acid 18 prior to coupling. Reaction
plished, and the resulting amine was converted to its Fmoc of 11 and 18 using standard peptide coupling conditions was
analog 4. The Fmoc protected phosphonoamino acid 5 was problematic, so acid 11 was transformed to the NHS-ester 12
accessible via Jones oxidation of alcohol 4. To continue the pro- (Figure 4). Reaction of amine 18 and activated ester 12 in the
drug synthesis (Figure 2), the hydroxyl group in 4 was converted presence of triethylamine gave the free ligand 19a.
to the acetate 6 (Otaka et al., 1995). The diethyl phosphonate 6
This class of prodrug readily enters cells and is metabolically
was converted to the phosphonamidate 8 using the same activated intracellularly to generate the active phosphate/phos-
method previously reported for the phosphotyrosine mimetics phonate ligand (Garrido-Hernandez et al., 2006; Boutselis
(Boutselis et al., 2007). Briefly, the phosphonate was converted et al., 2007). The mechanism of prodrug activation for compound
to the phosphonyl dichloride by successive treatment with 15 is shown in Figure 6. Upon entry into the cell, the nitrofurfuryl
TMSI and oxalyl chloride, reaction with methyl chlorobutylamine group is removed under physiological conditions via enzymatic
gave phosphonamidic chloride 7, and reaction of 7 with nitrofur- reduction and spontaneous expulsion to furnish phosphonami-
furyl alcohol and base afforded phosphoramidate 8. Careful de- date anion A. Intramolecular cyclization of A results in interme-
protection of acetate 8 with cold potassium carbonate followed diate B, which hydrolyzes spontaneously to furnish phosphonate
by oxidation of alcohol 9 using Jones reagent gave the desired product 19a.
phosphoserine prodrug 10. This route obviates the need for
chiral starting materials and is highly efficient for preparation of Biological Evaluation of 14-3-3 Inhibitor
the prodrug.
We measured the growth inhibitory activity of the prodrug and
the active ligand in human DG75 leukemia cells. After treatment
with compound 15 or compound 19a for 48 hr, the number of
Synthesis of 14-3-3 Inhibitor
While this work was in progress, the identification of a small live cells was counted and compared to treatment with DMSO
molecule 14-3-3 inhibitor was reported (Wu et al., 2010). The alone. The results are shown in Figure 7A. Compound 15 was
inhibitor, compound 19b (Figure 5), is a phosphoserine peptido- a potent inhibitor with an IC50 value of 5 mM; however, the free
mimetic that was identified using a small molecule microarray phosphonate 19a showed no meaningful growth inhibition at
766 Chemistry & Biology 19, 764–771, June 22, 2012 ª2012 Elsevier Ltd All rights reserved

Chemistry & Biology
Phosphoserine Mimetic Inhibits 14-3-3 Protein
Figure 3. Synthesis of the 14-3-3 Inhibitor Prodrug
Reagents and conditions: (a) 4-(2-aminoethyl)phenol, PYBOP, DIPEA, 80%; (b) DBU, DCM, 58%; (c) 11, PYBOP, DIPEA, DMF, 50%.
concentrations up to 100 mM (data not shown). Cell viability was assay, we transfected DG75 cells with plasmids coding for
>90% at concentrations of 15 up to 5 mM but decreased rapidly FOXO3a and/or Akt1 along with two luciferase reporter
at higher concentrations, suggesting that 15 was highly cytotoxic constructs: FOXO3a-luciferase and, as a control for transfection
at concentrations >10 mM. These results are consistent with the efficiency, TK Renilla-luciferase construct. After recovery, cells
facile uptake of the prodrug as compared to the highly charged were treated with increasing concentrations of compound 15
phosphonate ligand.
for 12 hr. Luminescence measurements were carried out using
The potential cytotoxic effects of compound 15 were investi- the Dual-Glo Luciferase assay system (Promega). The results
gated by monitoring the cleavage of poly (ADP-ribose) poly- are shown in Figure 7C. The ectopic expression of FOXO3a led
merase (PARP), a well-known caspase substrate. DG75 cells to a substantial increase in FOXO3a-mediated gene transcrip-
were treated with increasing concentrations of 15 for 24 hr. tion, which was inhibited upon cotransfection of Akt1 as ex-
The cells were lysed and the amounts of full-length and cleaved pected. FOXO3a stimulated transcription was recovered in
PARP were detected by western blot. Staurosporine (STS) was a dose-dependent manner by treatment of cells with increasing
chosen as the positive control; the results are shown in Figure 7B. concentrations of compound 15, consistent with inhibition of the
Significant PARP cleavage was observed at a prodrug concen- 14-3-3-mediated retention of FOXO3a in the cytoplasm. As
tration of 12 mM, suggesting that the cytotoxicity of prodrug 15 a control, cell lysates from this experiment were analyzed by
results from induction of apoptosis.
western blot, revealing that the total amount of FOXO3a and
To confirm a direct involvement of 14-3-3 proteins in the action Akt1 is independent of treatment with 15 (Figure S1 available
of 15, we developed an assay to monitor the 14-3-3-dependent online). Concentrations of 15 in excess of 8 mM resulted in
inhibition of FOXO transcription factors. FOXO3a is one of the substantial cytotoxicity, so the effects of higher doses could
Forkhead transcription factors that is phosphorylated by Akt1, not be assessed. However, the concentration of 15 required to
and the phosphorylated protein binds to 14-3-3. This binding recover 50% of the FOXO3a-induced gene expression was
facilitates the nuclear export of FOXO3a by exposing nuclear consistent with the IC50 value for growth inhibition, further sug-
export sequences and its retention in the cytoplasm by masking gesting that the cytotoxicity of 15 is a function of its ability to
nuclear localization sequences. This regulation by the 14-3-3 bind 14-3-3.
proteins allowed us to monitor the inhibition of 14-3-3 directly
To demonstrate that the effect of 15 on FOXO3a reporter
through a FOXO3a-driven luciferase reporter assay. In this activity results from the direct interaction of 19a with the
Figure 4. Synthesis of the Activated NHS-Ester Moiety
Reagents and conditions: (a) EtOCOCH2NH2$HCl, PYBOP, DIPEA, 87%; (b) LiOH, THF/H2O (3:1), 83%; (c) N-hydroxysuccinimide, THF, DIC, 68%.
Chemistry & Biology 19, 764–771, June 22, 2012 ª2012 Elsevier Ltd All rights reserved 767

Chemistry & Biology
Phosphoserine Mimetic Inhibits 14-3-3 Protein
Figure 5. Synthesis of the 14-3-3 Inhibitor Ligand
Reagents and conditions: (a) 4-(2-aminoethyl)phenol, PYBOP, DIPEA, 86%; (b) DBU,DCM, 88%; (c) TMSI, then water, 84%; (d) 12, water/acetone, Et, 69%.
14-3-3 protein, a competitive binding assay was carried out. concentrations (100 mM) were required to see effects in cells
Immobilized 14-3-3 t was incubated with DG75 cell lysate that than are needed in cell-free assays (ꢂ5 mM). In contrast, the pro-
had been transfected with FOXO3a and Akt1; this incubation drug 15 described here exhibits cellular effects at concentrations
produced an immobilized 14-3-3/FOXO3a complex. The that are comparable with those used in cell-free assays
complex was then incubated with different concentrations of measuring PARP cleavage and FOXO3a-induced gene expres-
19a for 1 hr. The bound proteins were then subjected to SDS- sion, suggesting that the prodrug increases significantly the
PAGE, and the amount of FOXO3a remaining on the beads intracellular delivery of the ligand 19a.
following treatment with 19a was analyzed by western blot.
In summary, we have developed and executed a strategy for
The results are shown in Figure 7D; a concentration-dependent the design, synthesis, and characterization of a class of small
disruption of the 14-3-3/FOXO3a complex when the beads molecule prodrug inhibitors of protein-protein interactions. The
were treated with 19a was observed. When the complex was successful design of deliverable, stable phosphoserine mimetics
treated with prodrug 15 (24 mM), there was no change in FOXO3a has implications not only for the development of therapeutic
levels relative to the DMSO control (data not shown).
agents targeting 14-3-3 proteins, but also for the development
Wu et. al. reported that phosphate 19b binds to 14-3-3 with of small molecules that interdict other phosphoserine and phos-
high affinity and is cell permeable. However, much higher phothreonine-binding domains.
Figure 6. Prodrug Activation Mechanism
768 Chemistry & Biology 19, 764–771, June 22, 2012 ª2012 Elsevier Ltd All rights reserved

Chemistry & Biology
Phosphoserine Mimetic Inhibits 14-3-3 Protein
Figure 7. Biological Analysis of a 14-3-3
Inhibitor
(A) Growth inhibition of DG75 cells. DG75 cells
were treated with 1.5, 3, 6, 12, and 24 mM 15 for
48 hr. DMSO was used as a negative control.
Live cells were counted by exclusion of trypan
blue using the BioRad automated cell counter.
Error bars report SEM for three separate experi-
ments.
(B) Activation of apoptotic marker upon treatment
with 15. DG75 cells were treated with 15 at 3, 6,
and 12 mM for 24 hr. Treatment with DMSO and
Staurosporin (STS, 100 nM) provided negative and
positive controls, respectively. The levels of PARP
in full-length and cleaved forms were analyzed by
western blot.
(C) Effect of 15 on the cellular 14-3-3/FOXO3a
interaction. DG75 cells were transfected with
plasmids for the transcription factor FOXO3a,
Akt1, FOXO3a-luciferase, and TK-Renilla lucif-
erase. After a 12 hr recovery, cells were treated
with 1, 3, 6, and 8 mM 15 for 12 hr. Luminescence
measurements were made and presented as
a percentage of luminescence measured in cells
expressing FOXO3a. The bars report mean SE
for three separate experiments.
(D) Effect of 19a on the molecular 14-3-3/FOXO3a
interaction. Immobilized 14-3-3 t was incubated
with DG75 cell lysates which were transiently
transfected with FOXO3a and Akt1. The immobilized 14-3-3/FOXO3a complex was incubated with 3, 12, or 24 mM 19a for one hour. The amount of FOXO3a
remaining on the beads was eluted and analyzed by western blot. The bars report mean SE for three separate experiments.
at 0ꢁC for 15 min in buffer containing 25 mM HEPES (pH 7.2), 150 mM NaCl,
SIGNIFICANCE
1% Nonidet P-40, 10 mg/ml aprotinin, and 10 mg/ml leupeptin. Supernatants
from a 5 min centrifugation at 4000 3 g were separated by SDS-polyacryl-
A facile and high-yield synthesis of a nonhydrolyzable
amide gel electrophoresis and analyzed by western blot to determine the
phosphoserine mimetic prodrug has been developed that
amount of full-length and cleaved PARP. Rabbit antibody against PARP was
provides efficient intracellular delivery of the highly charged
purchased from Cell Signaling Technology.
phosphonoserine moiety. The difluoromethylenephospho-
serine prodrug can be readily incorporated into other
phosphoserine mimetics, providing a unique avenue to the
investigation of proteins that are regulated by phosphoser-
ine recognition sites. The utility of this strategy was demon-
strated in the synthesis of a low micromolar inhibitor of
14-3-3 family proteins that exhibited exciting biological
results in cell-based assays.
FOXO3a-Luciferase Assay
DG75 cells (8 3 107; 1 3 107 cells per treatment) were transfected with 10 mg
FOXO3a-Luciferase, 20 mg FOXO3a, 20 mg Akt1, and/or 2 mg TK-Renilla
Luciferase plasmids by elecotroporation (240 V, 960 mF). The cells were
allowed to recover on ice for 15 min. The cells were then diluted in 10 ml
FBS-containing RPMI media and allowed to incubate at 37ꢁC in 5% CO2. After
12 hr, prodrug 15 in DMSO was added at a final concentration of 1, 3, 6, and
8 mM. After incubating for another 12 hr, the cells were counted, resuspended,
and plated at 1 3 106 cells/ml. The cells were harvested 6 hr later and analyzed
for both firefly and Renilla luciferase activities using the Dual-Glo luciferase
assay system (Promega). The FOXO3a reporter activities were normalized
against those of the TK-Renilla luciferase internal controls.
EXPERIMENTAL PROCEDURES
Cell Culture
DG75 cells were cultured in RPMI-1640 supplemented with 10% fetal bovine
serum, 50 mM 2-mercaptoethanol, 1 mM sodium pyruvate, 100 IU/ml penicillin
G, and 100 mg/ml streptomycin.
Immobilized 14-3-3 t Binding Assay
GST-14-3-3 t was expressed in bacteria and bound to glutathione-Sepharose
beads. DG75 cells (1 3 107) were transfected with 20 mg FOXO3a and 20 mg
Akt1 expression plasmids by electroporation. The cells were allowed to
recover on ice for 15 min. The cells were then diluted in 10 ml FBS-containing
RPMI media and allowed to incubate at 37ꢁC in 5% CO2. After incubating for
24 hr, the cells were harvested and lysed in NP40 lysis buffer for 15 min on ice.
Aliquots of the lysate were incubated with immobilized 14-3-3 t for one hour at
4ꢁC. After incubation, the lysate was removed and the beads were washed
with NP40 lysis buffer two times. NP40 lysis buffer (200 ml) was added to
the beads followed by 2 ml DMSO or 19a in DMSO. This mixture was allowed
to incubate for one hour at 4ꢁC. Lysate was removed and the beads were
washed two times with lysis buffer. Bound proteins were separated by SDS-
PAGE and detected by western blotting using the FOXO3a antibody (Cell
Signaling).
Growth Inhibition Assay
DG75 cells at 1 3 105 cells/ml were aliquotted in a 96-well plate with 100 ml per
well. Serial dilution (1 ml) of stock solutions of 15 in DMSO, 19a in DMSO, or
DMSO alone was added to each well to reach the desired final concentration.
The plate was then incubated at 37ꢁC at 5% CO2 for 48 hr. Live cells, as
determined by trypan blue exclusion, were counted using a Bio-Rad TC10
automated cell counter. The number of live cells in the drug-treated wells
was compared to those in wells that were treated with DMSO.
PARP Cleavage Assay
DG75 cells (1 3 106 cells/ml) were treated with DMSO, prodrug 15, or STS for
24 hr. The cells were harvested and washed twice with PBS. Cells were lysed
Chemistry & Biology 19, 764–771, June 22, 2012 ª2012 Elsevier Ltd All rights reserved 769

Chemistry & Biology
Phosphoserine Mimetic Inhibits 14-3-3 Protein
SUPPLEMENTAL INFORMATION
Hu, X.E. (2004). Nucleophilic ring opening of aziridines. Tetrahedron 60, 2701–
2743.
Supplemental Information includes one figure and Supplemental Experimental
Procedures and can be found with this article online at doi:10.1016/j.chembiol.
2012.05.011.
Kawamoto, A., and Campbell, M. (1997). A new method for the synthesis of
a phosphonic acid analogue of phosphoserine via a novel 1,1-difluorophosph-
onate intermediate. J. Fluor. Chem. 81, 181–186.
Klein, D.C. (2007). Arylalkylamine N-acetyltransferase: ‘‘the timezyme.’’ J. Biol.
ACKNOWLEDGMENTS
Chem. 282, 4233–4237.
Lo Conte, L., Chothia, C., and Janin, J. (1999). The atomic structure of protein-
We thank Dr. Kavita Shah for providing the FOXO3a and FOXO3a-Luciferase
plasmids. We appreciate the assistance of Dr. Wen-Horng Wang for
preparing the immobilized 14-3-3 protein. We would also like to thank
Dr. Victoria Martin for her assistance in the FOXO3a luciferase reporter
assay. Support from the Purdue Cancer Center Support Grant P30
CA23168 for services provided by the NMR and Mass Spectrometry Share
Resources is appreciated. Financial support from NCI grant CA037372 is
gratefully acknowledged.
protein recognition sites. J. Mol. Biol. 285, 2177–2198.
Muslin, A.J., Tanner, J.W., Allen, P.M., and Shaw, A.S. (1996). Interaction of
14-3-3 with signaling proteins is mediated by the recognition of phosphoser-
ine. Cell 84, 889–897.
Nyasse, B., Grehn, L., and Ragnarsson, U. (1997). Mild, efficient cleavage of
arenesulfonamides by magnesium reduction. Chem. Commun. (Camb.) 1,
1017–1018.
Otaka, A., Miyoshi, K., Burke, T., Roller, P., Kubota, H., Tamamura, H., and
Fujii, N. (1995). Synthesis and application of N-boc-L-2-amino-4-(diethyl-
phosphono)-4,4-difluorobutanoic acid for solid-phase synthesis of nonhy-
drolyzable phosphoserine peptide analogues. Tetrahedron Lett. 36,
927–930.
Received: September 26, 2011
Revised: May 8, 2012
Accepted: May 11, 2012
Published: June 21, 2012
Ottmann, C., Marco, S., Jaspert, N., Marcon, C., Schauer, N., Weyand, M.,
Vandermeeren, C., Duby, G., Boutry, M., Wittinghofer, A., et al. (2007).
Structure of a 14-3-3 coordinated hexamer of the plant plasma membrane
H+ -ATPase by combining X-ray crystallography and electron cryomicro-
scopy. Mol. Cell 25, 427–440.
REFERENCES
Aitken, A. (2006). 14-3-3 proteins: a historic overview. Semin. Cancer Biol. 16,
162–172.
Bergmeier, S.C., and Seth, P.P. (1997). Formation of scalemic aziridines via the
Ozouf, P., Binot, G., Pommelet, J.-C., and Lequeux, T.P. (2004). Syntheses of
u-hydroxy-a,a-difluoromethylphosphonates by oxacycle ring-opening reac-
tions. Org. Lett. 6, 3747–3750.
nucleophilic opening of aziridines. J. Org. Chem. 62, 2671–2674.
Berkowitz, D., Shen, Q., and Maeng, J. (1994). Synthesis of the (a,a-difluor-
oalkyl)phosphonate analogue of phosphoserine. Tetrahedron Lett. 35, 6445–
6448.
¨
Pajkert, R., Kolomeitsev, A., Milewska, M., Roschenthaler, G., and Koroniak,
H. (2008). TiCl4 and grignard reagent-promoted ring-opening reactions of
various epoxides: synthesis of g-hydroxy-a,a-difluoromethylenephospho-
nates. Tetrahedron Lett. 49, 6046–6049.
Berkowitz, D.B., Eggen, M., Shen, Q., and Shoemaker, R.K. (1996). Ready
access to fluorinated phosphonate mimics of secondary phosphates.
synthesis of the (a,a-difluoroalkyl)phosphonate analogues of L-phosphoser-
ine, L-phosphoallothreonine, and L-phosphothreonine. J. Org. Chem. 61,
4666–4675.
Panigrahi, K., Eggen, M., Maeng, J.-H., Shen, Q., and Berkowitz, D.B. (2009).
The a,a-difluorinated phosphonate L-pSer-analogue: an accessible chemical
tool for studying kinase-dependent signal transduction. Chem. Biol. 16,
928–936.
Boutselis, I.G., Yu, X., Zhang, Z.Y., and Borch, R.F. (2007). Synthesis and cell-
based activity of a potent and selective protein tyrosine phosphatase 1B inhib-
itor prodrug. J. Med. Chem. 50, 856–864.
Qiu, W., and Burton, D.J. (1996).
A facile and general preparation of
a,a-difluoro benzylic phosphonates by the CuCl promoted coupling reaction
of the (diethylphosphonyl) difluoromethylcadmium reagent with aryl lodides.
Tetrahedron Lett. 37, 2745–2748.
Bradshaw, J.M., Mitaxov, V., and Waksman, G. (1999). Investigation of phos-
photyrosine recognition by the SH2 domain of the Src kinase. J. Mol. Biol. 293,
971–985.
Rose, R., Erdmann, S., Bovens, S., Wolf, A., Rose, M., Hennig, S., Waldmann,
H., and Ottmann, C. (2010). Identification and structure of small-molecule
stabilizers of the 14-3-3 protein-protein interactions. Angew. Chem. Int. Ed.
Engl. 49, 4129–4132.
Choi, J.Y., and Borch, R.F. (2007). Highly efficient synthesis of enantiomeri-
cally enriched 2-hydroxymethylaziridines by enzymatic desymmetrization.
Org. Lett. 9, 215–218.
Eriksson, U., Peterson, L.W., Kashemirov, B.A., Hilfinger, J.M., Drach, J.C.,
Borysko, K.Z., Breitenbach, J.M., Kim, J.S., Mitchell, S., Kijek, P., and
McKenna, C.E. (2008). Serine peptide phosphoester prodrugs of cyclic
cidofovir: synthesis, transport, and antiviral activity. Mol. Pharm. 5,
598–609.
Szewczuk, L., Tarrant, M., Sample, V., Drury, W., Zhang, J., and Cole, P.A.
(2008). Analysis of serotonin N-acetyltransferase regulation in vitro and in
live cells using protein semisynthesis. Biochemistry 47, 10407–10419.
Tobias, S.C., and Borch, R.F. (2001). Synthesis and biological studies of novel
nucleoside phosphoramidate prodrugs. J. Med. Chem. 44, 4475–4480.
Freel Meyers, C.L., Hong, L., Joswig, C., and Borch, R.F. (2000). Synthesis and
biological activity of novel 5-fluoro-20-deoxyuridine phosphoramidate pro-
drugs. J. Med. Chem. 43, 4313–4318.
Waschbu¨ sch, R., Samadi, M., and Savignac, P. (1997). A useful magnesium
reagent for the preparation of 1,1-difluoro-2-hydroxyphosphonates from
diethyl bromodifluoromethylphosphonate via a metal—halogen exchange
reaction. J. Organomet. Chem. 529, 267–278.
Fujii, N., Nakai, K., Habashita, H., Hotta, Y., Tamamura, H., Otaka, A., and
Ibuka, T. (1994). Synthesis of optically pure 2-aziridinemethanols: versatile
synthetic building blocks. Chem. Pharm. Bull. (Tokyo) 42, 2241–2250.
Wu, H., Ge, J., and Yao, S.Q. (2010). Microarray-assisted high-throughput
identification of a cell-permeable small-molecule binder of 14-3-3 proteins.
Angew. Chem. Int. Ed. Engl. 49, 6528–6532.
Garrido-Hernandez, H., Moon, K.D., Geahlen, R.L., and Borch, R.F. (2006).
Design and synthesis of phosphotyrosine peptidomimetic prodrugs. J. Med.
Chem. 49, 3368–3376.
Xie, L., Lee, S.-Y., Andersen, J.N., Waters, S., Shen, K., Guo, X.-L., Moller,
N.P., Olefsky, J.M., Lawrence, D.S., and Zhang, Z.-Y. (2003). Cellular effects
of small molecule PTP1B inhibitors on insulin signaling. Biochemistry 42,
12792–12804.
Henry-dit-Quesnel, A., Toupet, L., Pommelet, J.C., and Lequeux, T. (2003). A
difluorosulfide as a Freon-free source of phosphonodifluoromethyl carbanion.
Org. Biomol. Chem. 1, 2486–2491.
Hermeking, H. (2003). The 14-3-3 cancer connection. Nat. Rev. Cancer 3,
Yaffe, M.B., and Smerdon, S.J. (2001). PhosphoSerine/threonine binding
931–943.
domains: you can’t pSERious? Structure 9, R33–R38.
770 Chemistry & Biology 19, 764–771, June 22, 2012 ª2012 Elsevier Ltd All rights reserved

Chemistry & Biology
Phosphoserine Mimetic Inhibits 14-3-3 Protein
Yaffe, M.B., Rittinger, K., Volinia, S., Caron, P.R., Aitken, A., Leffers, H.,
Gamblin, S.J., Smerdon, S.J., and Cantley, L.C. (1997). The structural basis
for 14-3-3:phosphopeptide binding specificity. Cell 91, 961–971.
Zhao, S., and Etzkorn, F.A. (2007). A phosphorylated prodrug for the inhibition
of Pin1. Bioorg. Med. Chem. Lett. 17, 6615–6618.
Yokomatsu, T., Sato, M., and Shibuya, S. (1996). Lipase-catalyzed enantiose-
lective acylation of prochiral 2-(u-phosphono)alkyl-1,3-propanediols:
Application to the enantioselective synthesis of u-phosphono-a-amino acids.
Tetrahedron Asymmetry 7, 2743–2754.
Zheng, W., Schwarzer, D., Lebeau, A., Weller, J.L., Klein, D.C., and Cole, P.A.
(2005). Cellular stability of serotonin N-acetyltransferase conferred by phos-
phonodifluoromethylene alanine (Pfa) substitution for Ser-205. J. Biol. Chem.
280, 10462–10467.
Chemistry & Biology 19, 764–771, June 22, 2012 ª2012 Elsevier Ltd All rights reserved 771