Aryloxy Triester Phosphoramidates as Phosphoserine Prodrugs: A Proof of Concept Study

Article abstract of DOI:10.1002/cmdc.202000034

The specific targeting of protein-protein interactions by phosphoserine-containing small molecules has been scarce due to the dephosphorylation of phosphoserine and its charged nature at physiological pH, which hinder its uptake into cells. To address these issues, we herein report the synthesis of phosphoserine aryloxy triester phosphoramidates as phosphoserine prodrugs that are enzymatically metabolized to release phosphoserine. This phosphoserine-masking approach was applied to a phosphoserine-containing inhibitor of 14-3-3 dimerization, and the generated prodrugs exhibited improved pharmacological activity. Collectively, this provided a proof of concept that the masking of phosphoserine with biocleavable aryloxy triester phosphoramidate masking groups is a viable intracellular delivery system for phosphoserine-containing molecules. Ultimately, this will facilitate the discovery of phosphoserine-containing small-molecule therapeutics.

Full text of DOI:10.1002/cmdc.202000034

Accepted Article
Title: Aryloxy Triester Phosphoramidates as Phosphoserine Prodrugs-
A Proof of Concept Study
Authors: Ageo Miccoli, Binar Dhiani, Peter Thornton, Olivia
Lambourne, Edward James, Hachemi Kadri, and Youcef
Mehellou
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemMedChem 10.1002/cmdc.202000034
Link to VoR: http://dx.doi.org/10.1002/cmdc.202000034
A Journal of
www.chemmedchem.org

10.1002/cmdc.202000034
ChemMedChem
COMMUNICATION
Aryloxy Triester Phosphoramidates as Phosphoserine Prodrugs-
A Proof of Concept Study
Ageo Miccoli,[a] Binar A. Dhiani,[a] Peter J. Thornton,[b] Olivia A. Lambourne,[a] Edward James,[a]
Hachemi Kadri,[c] and Youcef Mehellou*[a]
[a]
Dr. Ageo Miccoli, Dr. Binar A. Dhiani, Ms. Olivia A. Lambourne, Dr. Edward James, and Dr. Youcef Mehellou
School of Pharmacy and Pharmaceutical Sciences
Cardiff University
King Edward VII Avenue, Cardiff University, CF10 3NB, U.K.
E-mail: MehellouY1@cardiff.ac.uk
Dr. Peter J. Thornton
Technology Solutions GBU
Solvay Solutions, Oldbury B69 4LN, U.K
Dr. Hachemi Kadri
[b]
[c]
Department of Chemistry
Durham University
South Road, Durham, DH1 3LE, UK.
Supporting information for this article is given via a link at the end of the document.
Abstract: The specific targeting of protein-protein interactions by
phosphoserine-containing small molecules has been scarce due to
the dephosphorylation of phosphoserine and its charged nature at
physiological pH, which hinders its uptake into cells. To address these
issues, we herein report the synthesis of phosphoserine aryloxy
triester phosphoramidates as phosphoserine prodrugs, which are
enzymatically metabolized to release phosphoserine. This
phosphoserine masking approach was applied to a phosphoserine-
containing inhibitor of 14-3-3 dimerization, and the generated
prodrugs exhibited improved pharmacological activity. Collectively,
this provided a proof of concept of the masking of phosphoserine with
biocleavable aryloxy triester phosphoramidate masking groups as a
viable intracellular delivery system for phosphoserine-containing
molecules. Ultimately, this will facilitate the discovery of
phosphoserine-containing small molecule therapeutics.
masking of the phosphate group of phosphoserine was
previously investigated, the study was limited to
a
phosphoserine mimetic (difluoromethylenephosphoserine)
and not the natural phosphoserine moiety.[4] With this in mind
and in order to improve the drug-like properties of
phosphoserine-containing small molecules, we explored the
application of the aryloxy triester phosphoramidate
technology to phosphoserine. This technology has been
widely used to mask the 5’-O-monophosphate groups of
nucleotides, and has so far led to two FDA-approved drugs;
sofosbuvir and tenefovir alafenamide.[5]
As a proof of concept, we initially synthesized phosphoserine with
the aryloxy triester phosphoramidate masking groups (Figure 1a).
For this, the N- and C-terminals of phosphoserine had to be
protected[6] first to allow for the selective addition of the aryloxy
triester phosphoramidate moiety to the side chain hydroxyl group.
The synthesis was initiated by the chlorination of tert-butanol (1)
by copper(I) chloride with the peptide coupling reagent N,N'-
dicyclohexylcarbodiimide (DCC).[6] The generated compound, 2,
was then reacted with the commercially available N-Boc L-serine,
3, in DCM to yield the N- and C-protected L-serine 4.[6] This latter
compound was subsequently reacted with phenyl L-alanine
methyl ester phosphorochloridate (7), which had been
synthesized according to reported procedures[7] by reacting L-
alanine methyl ester hydrochloride (6) with the commercially
available phenyl dichlorophosphate (5) in DCM and in the
presence of triethylamine (NEt3). This reaction yielded the desired
phosphoserine aryloxy triester phosphoramidate, compound 8, as
a white solid in a good yield (62%).
Protein phosphorylation at serine residues is a fundamental
phenomenon that is used by cells to affect the function,
localization and degradation of proteins.[1] In some cases,
phosphorylated serine residues mediate protein-protein
interactions via the docking of the phosphoserine residue
into a positively charged pocket within the partner protein. An
example of this is the serine phosphorylation of the adaptor
protein 14-3-3, which facilitates its homodimerization.[2]
Attempts at inhibiting these phosphoserine-mediated
protein-protein interactions with small molecules have mostly
led, by design, to small molecules that lack phosphoserine.
The move to discard phosphoserine groups from these
molecules was driven by the fact that phosphoserine carries
two negative charges at physiological pH, which limit its
(passive) cellular uptake.[3] Additionally, phosphoserine is
also subject to dephosphorylation by alkaline phosphatases,
a process that yields serine-containing derivatives that do not
often retain potent pharmacological activity compared to their
parent phosphoserine-containing molecules. Although the
With the phosphoserine aryloxy triester phosphoramidate in hand,
we first studied whether the aryl and amino acid ester groups that
mask the phosphate group of phosphoserine could be
metabolized in vitro to release the unmasked phosphoserine
species. It is now well established that the metabolism of the
aryloxy triester phosphoramidate moieties is initiated by the
cleavage of the ester motif by carboxypeptidase Y to yield
1
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202000034
ChemMedChem
COMMUNICATION
O
A.
C.
Compound 8
HO
3
OH
O
OH
(δp = 3.2 ppm)
Metabolite C
NHBoc
Cl
(δp = 6.4 ppm)
OtBu
HO
DCC, Cu(I)Cl cat.
5 days, dark
DCM, 50%
NHBoc
t = 18 hr
1
2
4
O
O
O
P
Compound 8
OtBu
NHBoc
O
NH
(δp = 3.2 ppm)
2.2. eq. NEt3
DCM, 62%
Metabolite C
O
O
(δp = 6.4 ppm)
O
P
O
H2N
HCl
DCM, NEt3, 78%
Cl
HN
O
P
6
O
8
O
t = 3 hr
Cl
O
Cl
O
Compound 8
(δp = 3.1 and
3.2 ppm)
7
5
O
Metabolite C
B.
(δp = 6.4 ppm)
O
O
O
t = 30 min
O
OtBu
O
OtBu
O
P
OtBu
P
P
O
NH
O
NH
O
Carboxypeptidase Y
NHBoc
Cyclisation
O
O
NHBoc
Metabolite A
O
O
O
Compound 8
(δp = 3.1 and
3.2 ppm)
NHBoc
O
(δp = 4.1 ppm)
NH
O
O
Metabolite C
(δp = 6.4 ppm)
Metabolite A
Metabolite B
t = 5 min
Phosphoserine
phosphoramidate 8
Compound 8
(δp = 3.1 and
3.2 ppm)
O
O
O
OtBu
Phosphoramidase-type
enzyme
O
P
OtBu
P
O
NHBoc
O
O
H2O
NH
O
NHBoc
O
t = 0
O
O
Metabolite C
Phosphoserine
ppm
20
15
10
5
0
-5
Fig. 1. Synthesis and in vitro esterase-mediated metabolism of phosphoserine aryloxy triester phosphoramidates. A. Synthesis of a phosphoserine aryloxy triester
phosphoramidates (8). B. Mechanism of aryloxy triester phosphoramidates metabolism. C. 31P-NMR in vitro enzymatic assay of the breakdown of the
phosphoserine phosphoramidate by carboxypeptidase Y.
metabolite A (Figure 1b).[5a, 8] The formed carboxylate group then
performs a nucleophilic attack onto the phosphate group leading
to the release of the aryl group and the formation of a highly
substrate for the enzyme and thus was processed quicker.
From 30 minutes onward, only two 31P-NMR peaks
persisted; dp = 3.2 and 6.4 ppm. After 18 h, the peak at dp
=
unstable five-membered anhydride ring (metabolite B, Figure 1b). 6.4 ppm became the major peak with only one of the peaks
This is subsequently opened up by a water molecule to generate
the phosphoramidate metabolite C (Figure 1b). Finally, the
phosphoramidase-type enzyme Hint-1[9] cleaves the P-N bond of
metabolite C to release the unmasked monophosphate group.
With this in mind, we incubated compound 8 at 37 °C with
commercially available recombinant carboxypeptidase Y and
monitored the sample by 31P-NMR (Figure 1c).
from the original phosphoserine phosphoramidate two
singlets remained. Mass spectrometry analysis of the
products generated from this in vitro enzymatic assay after
18 h confirmed the breakdown of the phosphate masking
groups and the release of metabolite
S1). As in this in vitro assay, there was no phosphoramidase-
type enzyme, metabolite was not further processed to
release the fully unmasked phosphoserine species. To
predict whether metabolite would be a good substrate for
C (Supporting Figure
C
As shown in Figure 1c, at t = 0 min, the 31P-NMR of
compound
8
shows a sharp two singlets (dp = 3.1 and 3.2
C
ppm), which correspond to the two diastereoisomers (R and
S) of the phosphoserine aryloxy triester phosphoramidate.
This is typical of the aryloxy triester phosphoramidates as
they have a chiral phosphorous and most current synthetic
Hint-1, we performed in silico docking of the metabolite into
the crystal structure of the human Hint-1 co-crystallized with
AMP (PDB 1KPF) as previously reported.[10] The results
showed that metabolite
C sits in the Hint-1 active site and the
routes yield these compounds as
a
mixture of two
phosphate group of phosphoserine forms key interactions
with catalytic residues of Hint-1 (serine107, histidine112 and
histidine114) suggesting that it could be a good substrate for
this enzyme (Supporting Figure S2).[9b] These docking
results are in line with previous docking studies for the same
enzyme that predicted and verified whether the P-N bond of
the docked substrates could be cleaved off by Hint-1.[9b]
Together, these results indicate that phosphoserine aryloxy
triester phosphoramidates are metabolized in a similar
fashion to the nucleoside monophosphate aryloxy triester
phosphoramidates.
7b, 10]
diastereoisomers.[5a,
After 5 minutes incubation with
recombinant carboxypeptidase Y, two new peaks appeared;
one appeared briefly at dp = 4.1 ppm while another peak at
dp = 6.4 ppm also appeared and remained throughout the
rest of the experiment. This is typical for aryloxy triester
phosphoramidates
carboxypeptidase Y showed a similar pattern with the peak
at p = 4.1 ppm corresponding to metabolite whilst that at
dp = 6.4 ppm corresponding to metabolite
Figure 1b).[10-
Notably, after 30 minutes incubation, only the 31P-NMR
peak at
p = 3.2 ppm remained from the two 31P-NMR peaks
of the parent compound . This indicated that one of the
as
previous
studies
using
d
A
C
(
11]
d
In order to establish the applicability of the aryloxy triester
8
phosphoramidate
pharmacological activity of phosphoserine-containing
compounds, we applied this approach to 14-3-3
approach
in
improving
the
diastereoisomers of phosphoserine phosphoramidate, which
has a 31P-NMR that corresponds to dp = 3.1 ppm, is a better
a
2
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202000034
ChemMedChem
COMMUNICATION
OH
OH
O
O
O
O
BocHN
BocHN
BocHN
H2N
OH
N
OH
N
Tyramine, PyBOP,
NEt3, THF, 12 h, 95%.
18, PyBOP, NEt3,
DMF, 12 h, 80%.
BnBr, NaH, DMF,
10 h, 70%.
H
1) TFA, DCM, 48 h, 97%.
2) H2, 10% Pd/C, EtOAc,
EtOH, 2 h, 99%.
H
OBn
OBn
OH
OH
9
10
11
3
OH
O
O
H
N
N
H
N
H
1) For compound 13:
a. N,N,N',N'-Tetramethylphosphorodiamidic
chloride, tBuMgCl, MeCN, 24 h, 25%.
b. HCl, dioxane, 24 h, 99%.
O
N3
O
P
13: R1 = R2 = OH.
14: R1 = R2 = Et.
O
OR1
2) For compound 14:
Diethyl chlorophosphate, tBuMgCl,
MeCN, 24 h, 30%.
OH
O
O
OR2
H
N
N
N
H
O
H
OH
N3
OH
O
O
H
N
12
N
N
H
H
O
7, 20a-c, tBuMgCl, MeCN, 24 h.
N3
O
15a: R3 = Me (29%).
15b: R3 = iPr (27%).
15c: R3 = Bn (29%).
O
P
O
NH
R3O
O
B.
O
H
O
C.
O
O
N
O
O
Cl
N
R4
ethyl glycine
hydrochloride
P
OEt
OH
L-alanine amino acid
ester hydrochloride
P
O
OH
O
N
O
Cl
Cl
H
H
NaOH, THF,
1 h, 98%
PyBOP, DIPEA,
DCM, 3 h, 92%
O
O
N3
NEt3, (Et)2O,
-78 C, 3 h.
7: R4 = Me (84%).
20a: R4 = iPr (85%).
20b: R4 = Bn (83%).
N3
N3
17
18
19
16
Fig. 2. A. Synthesis of the phosphoserine-containing 14-3-3 dimerization inhibitor (13), diethyl phosphate (14) and its aryloxy triester phosphoramidates (15a-c).
The serine motif is shown in red while its phosphate and masked phosphate groups are shown in blue. B. Synthesis of the intermediate 2-((4-
azidobenzoyl)oxy)acetic acid (18). C. Synthesis of phenyl L-alanine ester phosphorochloridates (7 and 20a-b).
dimerization inhibitor 13
(
Figure
2
), which showed
This compound
phosphoramidates of the 14-3-3 dimerization inhibitor, 13
the synthesis was accomplished by coupling compound 12
,
promising, but not potent activity in cells [12]
.
is an ideal candidate for the application of the aryloxy triester
phosphoramidate technology as it contains a phosphoserine
motif that is essential for its pharmacological activity.[12]
The synthesis of the aryloxy triester phosphoramidates of
compound 13 commenced by the regioselective benzylation
to the appropriate phosphorochloridate (
7
,
20a or 20b
,
t
Figure 2c) in the presence of BuMgCl in MeCN to give the
desired aryloxy triester phosphoramidates (15a-c), as a
mixture of two diastereoisomers, in low yields (~ 30%).
Notably, the phosphorochloridates
7
,
20a-b were
of the side chain hydroxyl group of N-Boc-L-serine (
proceeded with 70% yield (Figure 2a). The carboxylic acid
of the product, , was then conjugated to the amine group of
3
), which
synthesised as shown in Figure 2c and reported
previously.[7]
9
In order to determine the efficacy of the synthesized aryloxy
triester phosphoramidate prodrugs to inhibit 14-3-3
dimerization and cell proliferation, we treated the lung cancer
cell line, A549, with the 14-3-3 dimerization inhibitor, 13, its
diethyl phosphate derivative, 14, and the aryloxy triester
tyramine to form a peptide bond using the peptide coupling
reagent PyBOP in THF in the presence of NEt3. This was
then followed by sequential deprotection of the amine group
and the side chain hydroxyl moiety using trifluoroacetic acid
in DCM and hydrogenation, respectively. This resultant
compound 11 was reacted with compound 18, which had
been generated[13] in two steps from the commercially
available 4-azidobenzoic acid (Figure 2b), to generate
compound 12.
For the synthesis of the reported 14-3-3 dimerisation
inhibitor, 13, compound 12 was reacted with N,N,N’,N’-
tetramethylphosphorodiamidic chloride in MeCN in the
presence of the Grignard reagent tertbutyl magnesium
chloride (tBuMgCl). The generated phosphorodiamidate
derivative was then converted to the desired monophosphate
species, 13, by treatment with HCl in dioxane. We also
synthesized the derivative of compound 13 where the
phosphate group is fully masked by diethyl groups,
compound 14, to be used as a negative control in the
biological assays. The synthesis was achieved by reacting
compound 12 with diethyl chlorophosphate in the presence
of tBuMgCl in MeCN to flourish the desired compound 14 in
30% yield. For the synthesis of the aryloxy triester
phosphoramidates 15a
indicated concentrations (12.5, 25, 50 and 100
-
c
as well as compound 12 at the
M) for 48
µ
and 72 h (Figure 3). Cell viability was then determined using
standard MTT assays. The results showed that, as expected,
the unphosphorylated compound 12, its phosphorylated
derivative 13 and the diethyl phosphate compound 14, did
not show any significant effect on cell viability.
Encouragingly, the aryloxy triester phosphoramidates 15a-c,
apart from 15a, did show significant dose-dependent
reduction of A549 cell viability (Figure 3). This is because
these aryloxy triester phosphoramidates are neutral at
physiological pH, unlike the parent compound 13, and thus
they are more readily taken up by cells via passive diffusion.
The fact that among this series of aryloxy triester
phosphoramidates, the isopropyl and benzyl ester
phosphoramidates (15b and 15c, respectively) showed
significant biological activity and the methyl ester
phosphoramidate, 15a
, did not demonstrate profound
biological activity, is in line with the established structure-
3
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202000034
ChemMedChem
COMMUNICATION
Compound 12
Compound 14
Compound 13
[concentration] µM
[concentration] µM
[concentration] µM
Prodrug 15a
Prodrug 15b
Prodrug 15c
[concentration] µM
[concentration] µM
[concentration] µM
Fig. 3. Cell viability of the unmasked phosphoserine compound (12), phosphoserine-containing 14-3-3 dimerization inhibitor (13), its diethyl phosphate derivatives
(14) and the aryloxy triester phosphoramidates (15a-c). Cell viability was determined by standard MTT assay. The compounds were incubated with A549 lung
cancer cell line for 48 h (white bar) and 72 h (grey bar) at the indicated concentrations. The percentage of cell viability was calculated and presented as normalized
value to control DMSO. Error bars show standard error from triplicate experiments.
[7] a) M. S. Davey, R. Malde, R. C. Mykura, A. T. Baker, T. E. Taher, C. S. Le
activity relationship of aryloxytriester phosphoramidates of
5’-O-nucleoside monophosphates.[5a, 7b]
Duff, B. E. Willcox, Y. Mehellou, J. Med. Chem. 2017; b) L. Osgerby, Y. C.
Lai, P. J. Thornton, J. Amalfitano, C. S. Le Duff, I. Jabeen, H. Kadri, A.
In summary, masking the phosphate group of phosphoserine
with an aryl motif and an amino acid ester serves as a useful
approach for the intracellular delivery of phosphoserine.
Indeed, this work showed using a combination of in vitro
Miccoli, J. H. R. Tucker, M. M. K. Muqit, Y. Mehellou, J. Med. Chem. 2017,
60(8), 3518-3524.
[8] a) G. Birkus, R. Wang, X. Liu, N. Kutty, H. MacArthur, T. Cihlar, C. Gibbs,
S. Swaminathan, W. Lee, M. McDermott, Antimicrob. Agents Chemother.
enzymatic and in silico docking studies that these
phosphoserine aryloxy triester phosphoramidates are
metabolized to release the unmasked phosphoserine moiety.
The application of this approach to a 14-3-3 phosphoserine-
2007, 51(2), 543-550; b) J. Balzarini, A. Karlsson, S. Aquaro, C. F. Perno,
D. Cahard, L. Naesens, E. De Clercq, C. McGuigan, Proc. Nat. Acad. Sci.
USA 1996, 93(14), 7295-7299.
[9] a) J. Cheng, X. Zhou, T. F. Chou, B. Ghosh, B. Liu, C. R. Wagner, Bioorg.
Med. Chem. Lett. 2009, 19(22), 6379-6381; b) E. Murakami, T. Tolstykh, H.
Bao, C. Niu, H. M. Steuer, D. Bao, W. Chang, C. Espiritu, S. Bansal, A. M.
Lam, M. J. Otto, M. J. Sofia, P. A. Furman, J. Biol. Chem. 2010, 285(45),
34337-34347.
containing
molecule
significantly
improved
its
pharmacological efficacy in cancer cells. Together, this
validates this approach of phosphoserine masking with
biocleavable motifs, and this will in the future facilitate the
discovery of phosphoserine-containing molecules as
powerful tool compounds and potential therapeutics.
[10] Y. Mehellou, R. Valente, H. Mottram, E. Walsby, K. I. Mills, J. Balzarini, C.
McGuigan, Bioorg. Med. Chem. 2010, 18(7), 2439-2446.
[11] M. Slusarczyk, M. H. Lopez, J. Balzarini, M. Mason, W. G. Jiang, S.
Blagden, E. Thompson, E. Ghazaly, C. McGuigan, J. Med. Chem. 2014,
57(4), 1531-1542.
[12] H. Wu, J. Ge, S. Q. Yao, Angew. Chem. 2010, 49(37), 6528-6532.
[13] H. Garrido-Hernandez, K. D. Moon, R. L. Geahlen, R. F. Borch, J. Med.
Chem. 2006, 49(11), 3368-3376.
Experimental Section
Experimental details are found in the Supporting Information.
Keywords: Phosphoserine • Phosphoramidate • Prodrug • 14-3-
3 • Cancer
[1] a) T. Pawson, J. D. Scott, Trends Biochem. Sci. 2005, 30(6), 286-290; b)
T. Hunter, Cell 1995, 80(2), 225-236; c) P. Cohen, Trends Biochem. Sci.
1992, 17(10), 408-413.
[2] M. B. Yaffe, K. Rittinger, S. Volinia, P. R. Caron, A. Aitken, H. Leffers, S. J.
Gamblin, S. J. Smerdon, L. C. Cantley, Cell 1997, 91(7), 961-971.
[3] L. Xie, S. Y. Lee, J. N. Andersen, S. Waters, K. Shen, X. L. Guo, N. P.
Moller, J. M. Olefsky, D. S. Lawrence, Z. Y. Zhang, Biochemistry 2003,
42(44), 12792-12804.
[4] A. Arrendale, K. Kim, J. Y. Choi, W. Li, R. L. Geahlen, R. F. Borch, Chem.
Biol. 2012, 19(6), 764-771.
[5] a) Y. Mehellou, H. S. Rattan, J. Balzarini, J. Med. Chem. 2018, 61(6), 2211-
2226; b) Y. Mehellou, ChemMedChem 2016, 11(11), 1114-1116.
[6] S. Koch, D. Schollmeyer, H. Lowe, H. Kunz, Chemistry 2013, 19(22), 7020-
7041.
4
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202000034
ChemMedChem
COMMUNICATION
Entry for the Table of Contents
Aryl
masking group
Phosphoserine
O
O
O
O
O
Intracellular
metabolism
O
O
O
HO
P
P
P
O
O
O
O
HN
O
O
HN
HO
HN
HN
HN
O
lipid
bilayer
O
O
Amino acid
ester masking group
O
Phosphoserine
The development of phosphoserine-containing therapeutics has been limited by their poor drug-like properties. To address this, we
applied the aryloxy triester phosphoramidate prodrug technology to phosphoserine and a phosphoserine-containing 14-3-3-
dimerization inhibitor. These prodrugs exhibited improved biological activity highlighting the promise of this prodrug technology in the
discovery of phosphoserine-containing therapeutic.
Institute and/or researcher Twitter usernames: @Youcef_Mehellou, @PharmacyCU
5
This article is protected by copyright. All rights reserved.