Hypoxia-activated pro-drugs of the KDAC inhibitor vorinostat (SAHA)

Article abstract of DOI:10.1016/j.tet.2020.131170

Hypoxia (lower than normal oxygen) is a characteristic of most solid tumours that results in poor cancer patient prognosis. The difference in cellular environment between normoxia (21percent oxygen) or physoxia (4–7.5percent oxygen) and hypoxia (<2.0percent oxygen) causes increased resistance to radio- and chemotherapy, but also provides the opportunity to selectively release hypoxia-activated pro-drugs. This approach potentially allows targeting of chemotherapies, including lysine deacetylase (KDAC) inhibitors, to the hypoxic fraction of cells. Here, we report initial work on the development of KDAC inhibitors that are selectively released in hypoxic conditions. We have shown that the addition of a 4-nitrobenzyl (NB) or 1-methyl-2-nitroimidazole (NI) bioreductive group onto the hydroxamic acid moiety of SAHA, giving NB-SAHA and NI-SAHA, abolishes KDAC inhibition activity. Both NB-SAHA and NI-SAHA undergo enzyme-mediated bioreduction, in a hypoxia-dependent manner, to release SAHA selectively in <0.1percent oxygen. This work provides an important foundation for further investigations into the targeted release of KDAC inhibitors in hypoxic tumours.

Full text of DOI:10.1016/j.tet.2020.131170

Tetrahedron xxx (xxxx) xxx
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Hypoxia-activated pro-drugs of the KDAC inhibitor vorinostat
(SAHA)*
Ewen D.D. Calder a, Anna Skwarska b, Deborah Sneddon a, Lisa K. Folkes b,
,
, **
Ishna N. Mistry b, Stuart J. Conway a *, Ester M. Hammond b
a Department of Chemistry, Chemistry Research Laboratory, University of Oxford, Mansfield Road, Oxford, OX1 3TA, UK
b Oxford Institute for Radiation Oncology, Department of Oncology, University of Oxford, Old Road Campus Research Building, Oxford, OX3 7DQ, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Hypoxia (lower than normal oxygen) is a characteristic of most solid tumours that results in poor cancer
patient prognosis. The difference in cellular environment between normoxia (21% oxygen) or physoxia (4
e7.5% oxygen) and hypoxia (<2.0% oxygen) causes increased resistance to radio- and chemotherapy, but
also provides the opportunity to selectively release hypoxia-activated pro-drugs. This approach poten-
tially allows targeting of chemotherapies, including lysine deacetylase (KDAC) inhibitors, to the hypoxic
fraction of cells. Here, we report initial work on the development of KDAC inhibitors that are selectively
released in hypoxic conditions. We have shown that the addition of a 4-nitrobenzyl (NB) or 1-methyl-2-
nitroimidazole (NI) bioreductive group onto the hydroxamic acid moiety of SAHA, giving NB-SAHA and
NI-SAHA, abolishes KDAC inhibition activity. Both NB-SAHA and NI-SAHA undergo enzyme-mediated
bioreduction, in a hypoxia-dependent manner, to release SAHA selectively in <0.1% oxygen. This work
provides an important foundation for further investigations into the targeted release of KDAC inhibitors
in hypoxic tumours.
Received 16 October 2019
Received in revised form
17 March 2020
Accepted 31 March 2020
Available online xxx
© 2020 Published by Elsevier Ltd.
1. Introduction
identified as a class of protein modules that bind to acetylated
lysine residues (KAc), making them readers of the epigenetic code
Acetylation of the ε-nitrogen atom of lysine residues was iden-
tified in 1964 [1] and has emerged as a functionally-important
post-translational modification (PTM), with comparable
complexity and significance to phosphorylation [2]. Lysine acety-
lation is found on proteins throughout the cell [3], with studies in
A549, Jurkat, and MV4; 11, cells identifying 3600 occurrences of
lysine acetylation in 1750 proteins [4]. A particular focus on the role
of lysine acetylation in histone proteins, which are involved in gene
transcription, has led to this modification being considered a
component of the epigenetic code [5]. Lysine acetylation is a dy-
namic PTM, with histone/lysine acetyl transferase (H/KAT) en-
zymes identified that transfer an acetyl group from acetyl co-
enzyme-A (AcCoA), and histone/lysine deacetylases (H/KDACs)
that remove the acetyl group [6]. Bromodomains have been
[7e9]. Bromodomain-containing proteins function by promoting
the formation of protein-protein and protein-chromatin in-
teractions, which often affect gene transcription [10,11].
Lysine deacetylating and deacylating enzymes are subdivided
into two mechanistically-distinct classes. The sirtuin enzymes
(Sirt1-7) are NADþ dependent, while the KDACs (KDAC1-11) have a
Zn2þ-dependent mechanism. Five pan-KDAC inhibitors (belinostat,
chidamide, vorinostat, panobinostat, and romidepsin) have been
approved to date for clinical use, mainly for the treatment of T-cell
lymphoma [6]. All of these inhibitors possess a group that bind to
the active site Zn2þ ion. For belinostat, vorinostat (SAHA, 1, Fig. 1A),
and panobinostat this group is a hydroxamic acid. While effective at
binding to the Zn2þ ion in the KDAC enzymes, this moiety can
potentially bind other off-target metals or inhibit other off-target
enzymes. This group is also very polar, hindering diffusion of
hydroxamic acid-based compounds into cells and certain areas of
the body, for example, the brain. In addition, this class of inhibitors
prevents the function of all eleven Zn2þ-dependent KDACs, mean-
ing that the transcription of many genes is affected by these com-
pounds. Perhaps as a consequence of these factors, KDAC inhibitors
*
This paper is dedicated to Professor Steve Davies.
* Corresponding author. Stuart Conway
** Corresponding author. Ester Hammond
E-mail addresses: Stuart.conway@chem.ox.ac.uk (S.J. Conway), Ester.hammond@
oncology.ox.ac.uk (E.M. Hammond).
https://doi.org/10.1016/j.tet.2020.131170
0040-4020/© 2020 Published by Elsevier Ltd.
Please cite this article as: E.D.D. Calder et al., Hypoxia-activated pro-drugs of the KDAC inhibitor vorinostat (SAHA), Tetrahedron, https://doi.org/
10.1016/j.tet.2020.131170

2
E.D.D. Calder et al. / Tetrahedron xxx (xxxx) xxx
Fig. 1. A. The structure of SAHA (Vorinostat, 1) and an X-ray crystal structure of SAHA (1, carbon ¼ yellow) bound to the active site of KDAC2 (PDB code 4LXZ) [17e19]. B. The
structures of 4-nitrobenzyl SAHA (NB-SAHA, 2) and 1-methyl-2-nitroimidazyl-SAHA (NI-SAHA, 3).
are linked to a number of serious side effects including fatigue,
gastrointestinal issues (diarrhoea, nausea, vomiting), and hemato-
logic complications (anaemia, neutropenia, thrombocytopenia)
[12,13]. In addition, SAHA (1) has a half-life of approximately 0.5 h
in human plasma [14], and 1.5e2.0 h when given orally to patients
with advanced cancer [15]. The main ways in which SAHA (1) is
primarily metabolised are either through glucuronidation of the
hydroxamic acid or hydrolysis of this group to give the corre-
Hypoxia is a characteristic of most solid tumours, which results
from high metabolic demand and abnormal vasculature [25].
Hypoxic tumours are associated with aggressive tumour pheno-
types and increased resistance to radio- and chemotherapy, which
results in poor cancer patient prognosis [26]. Consequently, while
the hypoxic fraction of cells are the most difficult to treat, this is the
fraction of cells that we want to target with chemotherapies,
including KDAC inhibitors. The difference in cellular REDOX envi-
ronment between physoxic (4e7.5% oxygen) and hypoxic (<2.0%
oxygen) [27] cells provides the opportunity to develop pro-drugs
that selectively release their cargo in hypoxia [28]. This strategy
has been employed by us and others with some success [28]. We
have previously reported an hypoxia-activated kinase inhibitor
[29], strategies and protocols for developing and evaluating
hypoxia-activated pro-drugs [19], a new mechanism for imaging of
hypoxia [30], and an hypoxia-activated pro-drug of isopropyl 1-
sponding carboxylic acid, which can undergo subsequent
b-
oxidation. All of these metabolic products are inactive in terms of
KDAC inhibition [13,16]. For these reasons, we are interested in
developing pro-drugs of KDAC inhibitors. These compounds will
potentially circumvent these metabolic liabilities and improve
compound permeation. However, our primary motivation is to
develop pro-drugs that are activated in a given location or context,
with the aim of minimising the side effects of SAHA (1).
A small number of pro-drugs of SAHA have been reported pre-
viously. Schlimme et al. reported a carbamate pro-drug that is
release in cells by hydrolysis [20], and Liu et al. recently reported a
co-pro-drug with paclitaxel [21]. In both cases these compounds
have no mechanism by which they are selectively released in a
given location or context (e.g. cancer cells). In an attempt to target
SAHA to cancerous cells, pro-drugs that are activated by thiols [13],
reactive oxygen species [22], or hydrogen peroxide or peroxynitrite
have been developed [23]. Rubio-Ruiz have reported pro-drug of
SAHA that can be activated using a Pd-functionalised resin in
glioblastoma or lung cancer cell lines [24]. In all of these strategies
the pro-moiety was added to the oxygen atom of the hydroxamic
acid group so as to reduce compound polarity and minimise
metabolism at this position. To the best of our knowledge, however,
there are no reports of SAHA (1) pro-drugs that are activated by
hypoxia (lower than normal oxygen).
thio-b-D-galactopyranoside (IPTG), allowing us to place gene
expression under the control of hypoxia [31]. Here we report initial
work towards the development of KDAC inhibitors that are selec-
tively released in hypoxic conditions (Fig. 1B).
2. Results and discussion
Hypoxia-activated pro-drugs require addition of a bioreductive
group to a functionality that is necessary for the compound to be
active, producing a compound that is inactive until deprotected. To
be released effectively, the functional group needs a relatively low
pKa, so that the fragmentation step, which happens following bio-
reduction, occurs in a facile manner. Based on these criteria, we
selected the hydroxamic acid as a good position to attach the bio-
reductive group. Others have used this position as a suitable point
of attachment when developing pro-drugs of SAHA (vide supra).
Please cite this article as: E.D.D. Calder et al., Hypoxia-activated pro-drugs of the KDAC inhibitor vorinostat (SAHA), Tetrahedron, https://doi.org/
10.1016/j.tet.2020.131170

E.D.D. Calder et al. / Tetrahedron xxx (xxxx) xxx
3
Based on the above considerations, compounds NB-SAHA (2,
Fig. 1B) and NI-SAHA (3, Fig. 2) were designed, in which the
hydroxamic acid is protected with the 4-nitrobenzyl or 1-methyl-2-
nitroimidazole group, respectively.
The synthesis of suberanilic acid (6) and SAHA (1) followed the
route reported by Mai et al. (Scheme 1) [32]. Briefly, suberic acid
was heated with acetic anhydride to give suberoyl anhydride (5),
which was then treated with aniline to give suberanilic acid (6).
Reaction of 6 with ethyl chloroformate, to form the presumed
mixed anhydride, followed by addition of hydroxylamine gave the
hydroxamic acid SAHA (1).
NB-SAHA (2) was synthesised as shown in Scheme 2. N-
Hydroxyphthalimide (7) was alkylated with 4-nitrobenzyl chloride
to give 8. The phthalimide group was removed by treatment with
concentrated HCl to give the benzylhydroxylamine 9. Coupling of 9
and suberanilic acid (6) using EDCI gave NB-SAHA (2).
the amine was observed within 15 min and, in addition, fragmen-
tation to SAHA (1) was observed (Fig. S1). After this time, the
reduction of NI-SAHA (3) plateaued, but the formation of SAHA (1)
increased steadily (Fig. S1). NB-SAHA (2) showed no fragmentation
to SAHA (1) in DMF, however, both the corresponding nitroso and
amine derivatives were observed within 15 min (data not shown).
Aliquots of reduced NB-SAHA (2) and NI-SAHA (3) at t ¼ 16 h were
incubated in phosphate buffer for 24 h at 37 ꢀC. While no
discernible increase in SAHA (1) was observed for the reduced NI-
SAHA (3) sample, the reduced NB-SAHA (2) sample showed a
substantial increase in SAHA (1). These data indicate that both NB-
SAHA (2) and NI-SAHA (3) undergo reduction upon treatment with
zinc and ammonium chloride. However, NI-SAHA (3) can undergo
immediate fragmentation to SAHA (1), while NB-SAHA (2) requires
the addition of aqueous phosphate buffer to promote fragmenta-
tion and production of SAHA (1).
NI-SAHA (3) was obtained by alkylation of SAHA (1) with 5-
(chloromethyl)-1-methyl-2-nitro-1H-imidazole (10, Scheme 3),
which was synthesised in a manner similar to that previously re-
ported (Scheme S1) [18,19].
As we have protected the polar hydroxamic acid group, the
solubility of NB-SAHA (2) and NI-SAHA (3) were determined using a
Turbidimetric Aqueous Solubility assay (Cyprotex, see SI for raw
data). These data show that NB-SAHA (2) is estimated to be soluble
We next evaluated the bioreduction and fragmentation of these
compounds to determine whether SAHA (1) would be selectively
released in hypoxic conditions. To achieve this, we used bactosomal
human NADPH-CYP reductase (CYP004, Cypex) at concentrations
of either 9.2 pmol/mL (Fig. 3) or 92 pmol/mL (Fig. S2). A concen-
tration of 9.2 pmol/mL enzyme correlates most accurately with the
degree of hypoxia activation seen in cells [31]. However, in previous
work [19,31], we have not always seen significant bioreduction and
fragmentation of the nitrobenzyl group at the lower enzyme con-
centration. Therefore, to determine whether NB-SAHA would
release SAHA at all we used both concentrations. When 9.2 pmol/
mL enzyme was used, NI-SAHA (3) was completely stable at 21%
oxygen and no release of SAHA (1) was detected (Fig. 3A). At <0.1%
oxygen 3 underwent bioreduction and showed significant release
of SAHA (1) (Fig. 3A). Under the same conditions, NB-SAHA (2) was
stable at 21% oxygen, but showed only a modest release of SAHA (1)
at <0.1% oxygen (Fig. 3B). At the higher enzyme concentration both
compounds showed appreciable bioreduction at both 21% and
<0.1% oxygen. Analysis of the enzymatic reduction of NI-SAHA (3),
using LCMS, showed the presence of the expected mass for SAHA (1,
Fig. S3).
at a concentration of at least 10
mated to be soluble at a concentration of at least 100
be expected, these values are slightly lower than the reported
mM, and that NI-SAHA (3) is esti-
mM. As would
solubility of SAHA (1) (soluble at concentrations of at least 724 mM)
[33].
NB-SAHA (2) and NI-SAHA (3) were then evaluated in an KDAC
inhibition assay (Reaction Biology, see SI for raw data) against
KDAC1-9 and KDAC11 (Table 1). While SAHA (1) shows effective
inhibition of these enzymes, neither NB-SAHA (2) nor NI-SAHA (3)
showed any detectable inhibition at concentrations up to 10 mM.
These data are consistent with other reports of pro-drug strategies
in which the hydroxamic acid of SAHA has been protected (vide
supra).
To determine whether NB-SAHA (2) and NI-SAHA (3) undergo
reduction and fragmentation to give SAHA (1), NB-SAHA (2) and NI-
SAHA (3) were reduced using zinc and ammonium chloride, and
the products analysed using LCMS.38 For NI-SAHA (3), reduction to
These data indicate that NI-SAHA (3), at least, will be capable of
undergoing bioreduction and fragmentation to release SAHA (1) in
cells [19,29]. They indicate that the use of bioreductive groups to
mask the hydroxamic acid moiety of KDAC inhibitors is a viable
Fig. 2. Attachment of the 1-methyl-2-nitroimidazole group to SAHA (1) gives the pro-drug NI-SAHA (3), which does not inhibit KDAC1-9 or 11. Under hypoxic conditions the nitro
group is reduced to form an electron-donating substituent. This reduction results in fragmentation and release of the active KDAC inhibitor SAHA (1).
Please cite this article as: E.D.D. Calder et al., Hypoxia-activated pro-drugs of the KDAC inhibitor vorinostat (SAHA), Tetrahedron, https://doi.org/
10.1016/j.tet.2020.131170

4
E.D.D. Calder et al. / Tetrahedron xxx (xxxx) xxx
Scheme 1. The synthesis of suberanilic acid (6) and SAHA (1). Reagents and conditions: (i) Ac2O, 140 ꢀC; (ii) PhNH2, THF, rt, 27% over two steps; (iii) (a) Ethyl chloroformate, Et3N, THF,
rt; (b) Hydroxylamine hydrochloride, KOH, MeOH, rt, 66% over two steps.
Scheme 2. The synthesis of NB-SAHA (2). Reagents and conditions: (i) DIPEA, 4-nitrobenzyl chloride, DMF, 70 ꢀC, 84%; (ii) HCl (18 M, aq.), EtOH, 80 ꢀC, 90%; (iii) EDCI$HCl, HOBt, Et3N,
6 then 9, THF:CH2Cl2 (1:1), rt, 8%.
Scheme 3. The synthesis of NI-SAHA (3). Reagents and conditions: (i) NaOHaq (10 M), MeOH, rt, 19%.
Table 1
IC50 values (nM) for SAHA (1), NB-SAHA (2) and NI-SAHA (3) against KDAC1-9 and KDAC11. The colour scale represents a heat map, with ‘hot’ colours showing high enzyme
inhibition. 2 [34].
Please cite this article as: E.D.D. Calder et al., Hypoxia-activated pro-drugs of the KDAC inhibitor vorinostat (SAHA), Tetrahedron, https://doi.org/
10.1016/j.tet.2020.131170

E.D.D. Calder et al. / Tetrahedron xxx (xxxx) xxx
5
Sigma Aldrich and were used without further purification. Brine
refers to a saturated, aqueous solution of sodium chloride. Hexane
refers to a mixture of hexane isomers. Petroleum ether refers to the
fraction boiling between 40 and 60 ꢀC, unless otherwise stated.
Molecular sieves were activated in an oven at 400 ꢀC. Celite® refers
to Celite®545 filter aid, treated with sodium carbonate, flux-
calcined, which was purchased from Sigma Aldrich. Where
appropriate and if not otherwise stated, all non-aqueous reactions
were carried out under an inert atmosphere of argon, using flame-
dried glassware.
Anhydrous solvents were obtained under the following con-
ditions: THF, acetonitrile, dichloromethane, diethyl ether and DMF
were dried by passing them through a column of active basic
alumina according to Grubbs’ procedure and stored over activated
3 Å molecular sieves under argon [36]. Anhydrous methanol and
ethanol was purchased from Sigma Aldrich UK in SureSeal™ bottles
and used without further purification.
Analytical thin layer chromatography (TLC) was performed on
normal phase Merck silica gel 60 F254 aluminium-supported thin
layer chromatography sheets. Spots were visualised by either ab-
sorption under UV light (254 nm), exposure to iodine vapour or
thermal development after dipping into a solution of ammonium
molybdate in sulfuric acid, an aqueous solution of potassium per-
manganate or an ethanolic solution of ninhydrin. Reaction progress
was monitored at appropriate times by TLC.
Normal phase silica gel flash column chromatography was
performed manually using Geduran Silicagel 60 (40e63 mm) under
a positive pressure of compressed nitrogen. In vacuo refers to the
removal of solvents under reduced pressure using a Büchi™ rotary
evaporator in a water bath at 40 ꢀC, unless otherwise specified.
Vacuum transfer refers to the removal of solvents on a manifold
linked to a high vacuum pump at rt.
Fig. 3. The enzymatic (9.2 pmol/mL Bactosomal human NADPH-CYP reductase
[CYP004, Cypex]) reduction of NI-SAHA (3) and NB-SAHA (2) at 21% oxygen (red lines)
and <0.1% oxygen (blue lines). A. NI-SAHA (3) was completely stable at 21% oxygen and
no release of SAHA was detected (red lines). At <0.1% oxygen 3 underwent bio-
reduction and showed significant release of SAHA (1). B. NB-SAHA (2) was stable at
21% oxygen, but showed only a modest release of SAHA (1) at <0.1% oxygen.
1H NMR spectra were recorded on a Bruker AVIIIHD 400
(400 MHz) a Bruker AVII 500 with dual 13C(1H) cryoprobe
(500 MHz) or a Bruker AVIIIHD 500 (500 MHz) spectrometer with
the stated solvents as a reference for the internal deuterium lock.
Chemical shifts are reported as dH in parts per million (ppm) rela-
tive to tetramethylsilane (TMS) where dH (TMS) ¼ 0.00 ppm. The
spectra are calibrated using the solvent peak with the data pro-
vided by Fulmer et al. [37] The multiplicity of each signal is indi-
cated by: s (singlet), br s (broad singlet), d (doublet), dd (doublet of
doublets), ddd (doublet of doublet of doublets), t (triplet), q
(quartet), sept (septet), or m (multiplet). The number of protons (n)
for a given resonance signal is indicated by nH. The shift values of
resonances are quoted to 2 decimal places unless peaks have
similar chemical shifts, in which case 3 decimal places are used.
Where appropriate, coupling constants (J) are quoted in Hz. Iden-
tical proton coupling constants are averaged in each spectrum and
reported to the nearest 0.1 Hz. The coupling constants were
determined by analysis using Bruker TopSpin software (versions 3.2
and 4.0) or Mestrenova software (version 11). 1H spectra were
assigned using 2D NMR experiments including COSY, HSQC and
13Ce1H HMBC.
approach to developing compounds that selectively released in
hypoxic tumours. While SAHA (1) is an excellent model compound
on which to conduct these studies, it has a half-life of approxi-
mately 0.5 h in human plasma [14], and 1.5e2.0 h when given orally
to patients with advanced cancer [15]. Future work will focus on
using more metabolically stable KDAC inhibitors, for example
panobinostat, which has a half-life of approximately 30 h when
given orally to patients with advanced solid tumours [35].
3. Conclusion
In conclusion, we have shown that the addition of a 4-
nitrobenzyl or 1-methyl-2-nitroimidazole bioreductive group
onto the hydroxamic acid moiety of SAHA (1) abolishes KDAC in-
hibition activity, giving NB-SAHA (2) and NI-SAHA (3). NI-SAHA (3)
underwent efficient enzyme-mediated bioreduction in a hypoxia-
dependent manner, releasing SAHA (1) selectively in <0.1% oxy-
gen. This work provides an important foundation for further in-
vestigations into the targeted release of KDAC inhibitors in hypoxic
tumours. Future work will focus on the development of pro-drugs
using more metabolically-stable parent KDAC inhibitors, with the
aim of yielding pro-drugs that are more suitable for use in vivo.
13C NMR spectra were recorded on a Bruker AVIIIHD 400
(101 MHz) or a Bruker AVII 500 with dual 13C(1H) cryoprobe
(126 MHz) spectrometer in the stated solvents with broadband
proton decoupling and an internal deuterium lock. Chemical shifts
are reported as dC in parts per million (ppm) relative to tetrame-
thylsilane (TMS) where dC (TMS) ¼ 0.00 ppm. The spectra are
calibrated using the solvent peak with the data provided by Fulmer
et al. [37] The shift values of resonances are quoted to 1 decimal
place unless peaks have similar chemical shifts, in which case 2
decimal places are used. 13C spectra were assigned using 2D NMR
experiments including HSQC and 13Ce1H HMBC.
4. Experimental section
4.1. General chemistry procedures
Chemicals were purchased from Acros Organics, Alfa Aesar,
Apollo Scientific, Fisher Scientific, Fluka, Fluorochem, Merck or
Electrospray ionisation (ESI) mass spectra were acquired
Please cite this article as: E.D.D. Calder et al., Hypoxia-activated pro-drugs of the KDAC inhibitor vorinostat (SAHA), Tetrahedron, https://doi.org/
10.1016/j.tet.2020.131170

6
E.D.D. Calder et al. / Tetrahedron xxx (xxxx) xxx
using an Agilent 6120 Quadrupole spectrometer or Waters LCT
Premier spectrometer, operating in positive or negative mode, as
indicated, from solutions of MeOH or MeCN. Chemical ionisation
(CI) mass spectra were acquired using a Waters GCT spectrometer.
MS data was processed using MestReNova software (version 11). m/
z values are reported in Daltons and followed by their percentage
abundance in parentheses.
nitrobenzyl)-hydroxylamine (9)38, ethyl (2-amino-1-methyl-imid-
azole-5-yl)carboxylate (S3) [17e19], ethyl (1-methyl-2-nitro-
imidazole-5-yl)carboxylate (S4) [17e19], (1-methyl-2-nitro-imid-
azole-5-yl)methanol (S5) [17e19] and (1-methyl-2-nitro-imid-
azole-5-yl)methyl chloride (10) [17e19] can be found in the
supplementary information.
Accurate mass spectra were obtained using Bruker
mTOF
4.3. N-(40-nitrobenzyloxy)-n0-phenyloctanediamide (2)
spectrometer. m/z values are reported in Daltons. When a com-
pound was not observed by LRMS, only HRMS is quoted.
Melting points were determined using a Kofler hot stage and
are uncorrected. The solvent of crystallisation is shown in
parentheses.
Infrared (IR) spectra were obtained either from neat samples,
either as liquids or solids, or as a thin film using a diamond ATR
module. The spectra were recorded on a Bruker Tensor 27 spec-
trometer. Absorption maxima are reported in wavenumbers
(cmꢁ1). Only the main, relevant peaks have been assigned.
Compound purity was determined by analytical high-
performance liquid chromatography (HPLC) on a PerkinElmer
Flexar system with a Binary LC Pump and UV/Vis LC Detector. All
biologically tested compounds were of >95% purity by HPLC. For
determination of compound purity on reversed phase (RP) a Dionex
Suberanilic acid (6, 0.20 g, 0.80 mmol, 1.0 eq) was dissolved in a
mixture of tetrahydrofuran (10 mL) and dichloromethane (10 mL).
Triethylamine (0.56 mL, 4.0 mmol, 5.0 eq) was added to the solu-
tion followed by 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide
hydrochloride (0.23 g, 1.2 mmol, 1.5 eq) and the reaction was stirred
for 0.75 h at rt. O-(4-Nitrobenzyl)-hydroxylamine (9, 0.27 g,
1.6 mmol, 2.0 eq) was added and the reaction was stirred at rt for a
further 18 h. The reaction was quenched with hydrochloric acid
(1 M, 20 mL) and extracted with dichloromethane (20 mL). The
organic components were washed with hydrochloric acid (1 M,
20 mL), water (20 mL), and brine (20 mL), dried (Na2SO4), filtered,
and concentrated in vacuo. Purification by column chromatography
(ethyl acetate: petroleum ether 30e40, 20e100%) to yield com-
pound 2 (25 mg, 8%) as a colourless solid. Rf 0.52 (ethyl acetate); mp
150e152 ꢀC (from EtOAc); nmax (thin film)/cmꢁ1 3233, 2981, 2885,
1658, 1602, 1524, 1382, 1346, 1253, 1153, 1072; 1H NMR (500 MHz,
DMSO-D6) dH 11.06 (1H, s), 9.83 (1H, s), 8.24 (2H, d, J 8.4), 7.67 (2H,
d, J 8.4), 7.57 (2H, d, J 7.9), 7.27 (2H, dd, J 7.9, 7.6), 7.01 (1H, t, J 7.6),
4.93 (2H, s), 2.27 (2H, t, J 7.4), 1.94 (2H, t, J 7.3), 1.55 (2H, tt, J 7.4, 7.4),
1.48 (2H, tt, J 7.4, 7.3), 1.31e1.18 (4H, m); 13C NMR (126 MHz, DMSO-
D6) dC 180.7, 179.1, 156.7, 153.6, 148.8, 138.8, 138.1, 132.9, 132.4,
128.5, 85.0, 45.8, 41.6, 37.8, 37.7, 34.4, 34.2; HRMS m/z (ESIþ) Found:
400.18600, C21H26N3O5 requires [MþH]þ 400.18670; LRMS m/z
(ESIþ) 422 (100%, [MþNa]þ); HPLC Method A, Retention time e
8.4 min, 95%.
Acclaim® 120 column (C18, 5
m
m, 12 Å, 4.6 ꢂ 150 mm) was
employed with water as eluent A and acetonitrile as eluent B.
Samples were injected in methanol, water or acetonitrile. For each
compound, one of two methods were used as described below:
Method A refers to no modifier added.
Method B refers to the addition of 0.1% v/v TFA.
Time (min)
Flow (mL/min)
%A
%B
Slope
5
10
5
1.5
1.5
1.5
95%
5%
5%
5%
95%
95%
0
1
0
4.2. Semi-preparative HPLC details
HPLC purification of NI-SAHA (3) was carried out on Waters
Autopurification system,* equipped with a Waters X-Bridge OBD
semi-prep column (19 mm ꢂ 50 mm, 5
of 1 mL, eluting with H2Oþ0.1% Formic acid/MeOH þ0.1% formic
acid. The crude samples (in MeOH) were filtered (nylon, 0.2 m)
and injected in 750 L aliquots, with mass-directed purification
with an ACQUITY QDa performance mass spectrometer.
mm), with an injection loop
4.4. N-((10-methyl-20-nitroimidazol-50-yl)methoxy)-n0-
phenyloctanediamide (3)
m
m
*Full details: Waters 2767 Sample manager; Waters 2998
Photodiode array detector; Waters ACQUITY QDa; Waters SFO
(systems fluidics organiser) and Waters 2545 Binary Gradient
Module.
Method:
A ¼ H2O þ 0.1% formic acid.
B ¼ MeOH þ0.1% formic acid.
N-Hydroxy-N0-phenyloctanediamide (1, 20 mg, 0.076 mmol, 1.0
eq) was dissolved in a mixture of methanol (0.8 mL) and aqueous,
10 M sodium hydroxide (0.016 mL, 0.16 mmol, 2.1 eq). (1-Methyl-2-
nitro-imidazole-5-yl)methyl chloride (10, 20 mg, 0.11 mmol, 1.5 eq)
was added, the reaction was stirred at rt for 11 h then concentrated
in vacuo. The residue was dissolved in chloroform and filtered then
concentrated in vacuo. The residue was triturated first with chlo-
roform and then with ethyl acetate before the residue was dis-
solved in a minimal amount of methanol and purified by semi-
preparative HPLC, using general conditions (product elution at
10.5 min). Solvent was removed with a stream of nitrogen, over
18 h at rt, then dried by vacuum transfer to yield compound 3 as a
colourless solid (7.5 mg, 19%). Rf 0.28 (5% ethanol in chloroform);
Time (min)
Flow (mL/min)
%A
%B
Slope
0
1
25
25
25
25
25
95
95
5
5
95
5
5
95
95
5
0
0
1
0
1
17
18
20
Experimental details and characterisation for suberanilic acid
(6) [31], N-hydroxy-N0-phenyloctanediamide (SAHA, 1) [37], N-
phthalimido-O-(40-nitrobenzyl)-hydroxylamine
(8)38
,
O-(4-
Please cite this article as: E.D.D. Calder et al., Hypoxia-activated pro-drugs of the KDAC inhibitor vorinostat (SAHA), Tetrahedron, https://doi.org/
10.1016/j.tet.2020.131170

E.D.D. Calder et al. / Tetrahedron xxx (xxxx) xxx
7
mp 137e139 ꢀC (from water); nmax (thin film)/cmꢁ1 3200, 2934,
2860, 1663, 1559, 1544, 1495, 1443, 1333, 1278, 1189; 1H NMR
(500 MHz, DMSO-D6) dH 11.05 (1H, s), 9.84 (1H, s), 7.57 (2H, d, J 8.0),
7.27 (2H, dd, J 8.0, 7.6), 7.25 (1H, s), 7.01 (1H, t, J 7.6), 4.89 (2H, s),
4.03 (2H, s), 2.28 (2H, t, J 7.4), 1.94 (2H, t, J 7.3), 1.57 (2H, tt, J 7.4, 7.3),
1.47 (2H, tt, J 7.3, 7.3),1.32e1.20 (4H, m); 13C NMR (126 MHz, DMSO-
D6) dC 171.2, 169.8, 139.3, 132.8, 129.5, 128.6, 122.9, 119.0, 65.4, 36.3,
34.4, 32.1, 28.4, 28.3, 25.0, 24.7; HRMS m/z (ESIþ) Found: 404.19382,
combined with K2HPO4 were injected at 10
m
L, and AUC values
normalised for dilution. NI-SAHA was detected with a cone voltage
of 13 V at selected ion recording (SIR) of m/z 404.2 (M þ H),
retention time 3.7 min, NB-SAHA with a cone voltage of 13 V and
SIR of m/z 400.2 (M þ H), retention time 5.5 min and SAHA with a
cone voltage of 11 V and SIR of m/z 265.15 (M þ H), retention time
2.5 min. In addition, analytes were monitored in the same system
using absorbance spectroscopy with a Waters 2996 photodiode
array detector.
C
19H26N5O5 requires [MþH]þ 404.19285; LRMS (ESIþ) 830
([2 M þ Na]þ, 100%), 808 ([2 M þ H]þ, 35%), 608 (21%), 426
([MþNa]þ, 64%), 404 ([MþH]þ, 23%), 258 (13%), 102 (24%); HPLC
Method B, Retention time e 7.0 min, >99%.
Declaration of interests
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
4.5. Zinc reduction assay
The chemical reduction was carried out as previously described
[19]. Briefly, the compound of interest, NB-SAHA or NI-SAHA
Acknowledgements
(2.5
mmol), was combined in DMF (2 mL) with aqueous ammo-
nium chloride (20
m
L, 10% w/v). A sample (100 L) of this was
m
E.D.D.C., I.N.M., A.S., S.J.C., and E.M.H. thank the Medical
Research Council for funding (MR/N009460/1). S.J.C., D.S. and E.M.H
thank the Engineering and Physical Sciences Research Council for
Programme Grant support (EP/S019901/1). E.M.H. received
research funding from Cancer Research UK. S.J.C. thanks St Hugh’s
College, Oxford, for research support. The authors thank Dr Nicola
Farrer for use of her HPLC equipment.
removed (t ¼ 0 h) before zinc powder (5 mg, 0.077 mmol) was
added. At predetermined timepoints (t ¼ 0.25, 0.5, 1, 4, 16, 24 h), a
sample (100
filter (0.2
reduction. To determine the percentage fragmentation, a sample of
the reaction mixture (5 L) at selected timepoints were collected
mL) was removed and filtered through a nylon syringe
mM, 4 mm). Analysis by LCMS determined the percentage
m
and combined with K2HPO4 buffer (95
mL, 0.1 M, pH 7.4) and
incubated at 37 ꢀC for 24 h. LCMS analysis was carried out on a
Waters 2695 separations module. Separation was achieved on an
Appendix A. Supplementary data
ACE C18, 3
m
m, 150 ꢂ 3 mm column (Hichrom, UK) maintained at
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.tet.2020.131170.
40 ꢀC. Samples were eluted in A: 0.1% formic acid, B. Acetonitrile
containing 0.1% formic acid with a flow rate of 0.4 mL/min, and a
linear gradient of 40e90% B in 6 min, total run time 10 min. Ana-
lytes were measured using a mass spectrometer with a Waters
Acquity QDa detector with electrospray ionisation in a positive
ionisation mode (0.8 kV), source temperature 600 ꢀC.
References
[1] V. Allfrey, R. Faulkner, A. Mirsky, Proc. Natl. Acad. Sci. U.S.A. 51 (1964) 786.
[2] T. Kouzarides, EMBO J. 19 (2000) 1176.
[3] C. Choudhary, B.T. Weinert, Y. Nishida, E. Verdin, M. Mann, Nat. Rev. Mol. Cell
Biol. 15 (2014) 536.
[4] C. Choudhary, C. Kumar, F. Gnad, M.L. Nielsen, M. Rehman, T.C. Walther,
J.V. Olsen, M. Mann, Science 325 (2009) 834.
4.6. Enzyme reduction assay
[5] B.M. Turner, Nat. Cell Biol. 9 (2007) 2.
Bactosomal human NADPH-CYP reductase (9.2 or 92 pmol/mL,
CYP004, Cypex) was used in combination with NADPH-
regenerating system (Cat. No 451220 and Cat No 451200, Corn-
ing) as described previously [19]. NI-SAHA and NB-SAHA were
prepared as 10 mM stock solutions in DMSO, and their final con-
[6] M. Schiedel, S.J. Conway, Curr. Opin. Chem. Biol. 45 (2018) 166.
[7] D.S. Hewings, T.P.C. Rooney, L.E. Jennings, D.A. Hay, C.J. Schofield, P.E. Brennan,
S. Knapp, S.J. Conway, J. Med. Chem. 55 (2012) 9393.
[8] S.J. Conway, ACS Med. Chem. Lett. 3 (2012) 691.
[9] M. Brand, A.M. Measures, B.G. Wilson, W.A. Cortopassi, R. Alexander, M. Hoss,
€
D.S. Hewings, T.P.C. Rooney, R.S. Paton, S.J. Conway, ACS Chem. Biol. 10 (2015)
22.
centration in the reaction mixture was 10
were carried out under normoxic (21% O2) or hypoxic (<0.1% O2)
conditions. Samples (50 L) were taken at regular intervals,
quenched with 50 L MeCN/MeOH (1:1) and analysed by HPLC or
mM. Enzymatic reactions
[10] M. Schiedel, M. Moroglu, D.M.H. Ascough, A.E.R. Chamberlain, J.J.A.G. Kamps,
A.R. Sekirnik, S.J. Conway, Angew. Chem. Int. Ed. 58 (2019) 17930.
[11] A.G. Cochran, A.R. Conery, R.J. Sims, Nat. Rev. Drug Discov. 18 (2019) 609.
[12] B.E. Gryder, Q.H. Sodji, A.K. Oyelere, Future Med. Chem. 4 (2012) 505.
[13] K.B. Daniel, E.D. Sullivan, Y. Chen, J.C. Chan, P.A. Jennings, C.A. Fierke,
S.M. Cohen, J. Med. Chem. 58 (2015) 4812.
m
m
LCMS. HPLC analysis was carried out on a Waters 2695 separations
module with a Waters 2996 photodiode array detector. Separation
[14] R. Konsoula, M. Jung, Int. J. Pharm. 361 (2008) 19.
[15] W.K. Kelly, O.A. O’Connor, L.M. Krug, J.H. Chiao, M. Heaney, T. Curley,
B. MacGregore-Cortelli, W. Tong, J.P. Secrist, L. Schwartz, S. Richardson, E. Chu,
S. Olgac, P.A. Marks, H. Scher, V.M. Richon, J. Clin. Oncol. 23 (2005) 3923.
[16] E.H. Rubin, N.G.B. Agrawal, E.J. Friedman, P. Scott, K.E. Mazina, L. Sun, L. Du,
J.L. Ricker, S.R. Frankel, K.M. Gottesdiener, J.A. Wagner, M. Iwamoto, Clin. Canc.
Res. 12 (2006) 7039.
was achieved on a RPB C18 reversed phase column, 5
m
m, 100 ꢂ 3.2
column (Hichrom, UK) maintained at 35 ꢀC. Samples were eluted in
A: 0.1% formic acid, B: acetonitrile with a flow rate of 0.5 mL/min,
and a linear gradient of 20e100% B in 6 min, total run time 10 min.
Analytes were detected at 254 nm; NI-SAHA had a retention time of
4.8 min, NB-SAHA had a retention time of 5.6 min and SAHA had a
retention time of 3.8 min.
[17] I. Parveen, D.P. Naughton, W. Whish, M.D. Threadgill, Bioorg. Med. Chem. Lett
9 (1999) 2031.
€
[18] L.J. O’Connor, C. Cazares-Korner, J. Saha, C.N.G. Evans, M.R.L. Stratford,
E.M. Hammond, S.J. Conway, Org. Chem. Front. 2 (2015) 1026.
[19] L.J. O’Connor, C. Cazares-Korner, J. Saha, C.N.G. Evans, M.R.L. Stratford,
E.M. Hammond, S.J. Conway, Nat. Protoc. 11 (2016) 781.
[20] S. Schlimme, A.-T. Hauser, V. Carafa, R. Heinke, S. Kannan, D.A. Stolfa,
S. Cellamare, A. Carotti, L. Altucci, M. Jung, W. Sippl, ChemMedChem 6 (2011)
1193.
[21] S. Liu, K. Zhang, Q. Zhu, Q. Shen, Q. Zhang, J. Yu, Y. Chen, W. Lu, Bioorg. Med.
Chem. 27 (2019) 1405.
[22] S.D. Bhagat, U. Singh, R.K. Mishra, A. Srivastava, ChemMedChem 13 (2018)
2073.
LCMS analysis was carried out on a Waters 2695 separations
module. Separation was achieved on an ACE C18, 3
m
m, 150 ꢂ 3 mm
column (Hichrom, UK) maintained at 40 ꢀC. Samples were eluted in
A: 0.1% formic acid, B. Acetonitrile containing 0.1% formic acid with
a flow rate of 0.4 mL/min, and a linear gradient of 40e90% B in
6 min, total run time 10 min. Analytes were measured using a mass
spectrometer with a Waters Acquity QDa detector with electro-
spray ionisation in a positive ionisation mode (0.8 kV), source
[23] Y. Liao, L. Xu, S. Ou, H. Edwards, D. Luedtke, Y. Ge, Z. Qin, ACS Med. Chem. Lett.
9 (2018) 635.
temperature 600 ꢀC. Samples in DMF were injected at 1
mL, samples
Please cite this article as: E.D.D. Calder et al., Hypoxia-activated pro-drugs of the KDAC inhibitor vorinostat (SAHA), Tetrahedron, https://doi.org/
10.1016/j.tet.2020.131170

8
E.D.D. Calder et al. / Tetrahedron xxx (xxxx) xxx
[24] B. Rubio-Ruiz, J.T. Weiss, A. Unciti-Broceta, J. Med. Chem. 59 (21) (2016) 9974.
[25] M. Hockel, P. Vaupel, J. Natl. Cancer Inst. 93 (2001) 266.
391.
€
[33] Y.Y. Cai, C.W. Yap, Z. Wang, P.C. Ho, S.Y. Chan, K.Y. Ng, Z.G. Ge, H.S. Lin, J. Clin.
Pharm. Therapeut. 35 (2010) 521.
[34] J. Arts, P. King, A. Marien, W. Floren, A. Belien, L. Janssen, I. Pilatte, B. Roux,
L. Decrane, R. Gilissen, I. Hickson, V. Vreys, E. Cox, K. Bol, W. Talloen, I. Goris,
L. Andries, Jardin, M. Du, M. Janicot, M. Page, K. van Emelen, P. Angibaud, Clin.
Canc. Res. 15 (2009) 6841.
[35] S. Sharma, P.O. Witteveen, M.P. Lolkema, D. Hess, H. Gelderblom, S.A. Hussain,
M.G. Porro, E. Waldron, S.-Z. Valera, S. Mu, Canc. Chemother. Pharmacol. 75
(2015) 87.
[36] A.B. Pangborn, M.A. Giardello, R.H. Grubbs, R.K. Rosen, F. Timmers,
J. Organometallics 15 (1996) 1518.
[26] J.M. Brown, W.R. Wilson, Nat. Rev. Canc. 4 (2004) 437.
[27] S.R. McKeown, Br. J. Radiol. 87 (2014) 20130676.
[28] I.N. Mistry, M. Thomas, E.D.D. Calder, S.J. Conway, E.M. Hammond, Int. J.
Radiat. Oncol. Biol. Phys. 98 (2017) 1183.
[29] C. Cazares-Korner, I.M. Pires, I.D. Swallow, S.C. Grayer, L.J. O’Connor,
M.M. Olcina, M. Christlieb, S.J. Conway, E.M. Hammond, ACS Chem. Biol. 8
(2013) 1451.
[30] L.J. O’Connor, I.N. Mistry, S.L. Collins, L.K. Folkes, G. Brown, S.J. Conway,
E.M. Hammond, ACS Cent. Sci. 3 (2017) 20.
[31] S.L. Collins, J. Saha, L.C. Bouchez, E.M. Hammond, S.J. Conway, ACS Chem. Biol.
13 (2018) 3354.
€
€
[37] G.R. Fulmer, A.J.M. Miller, N.H. Sherden, H.E. Gottlieb, A. Nudelman,
B.M. Stoltz, J.E. Bercaw, K.I. Goldberg, Organometallics 29 (2010) 2176.
[32] A. Mai, M. Esposito, G. Sbardella, S. Massa, Org. Prep. Proced. Int. 33 (4) (2001)
Please cite this article as: E.D.D. Calder et al., Hypoxia-activated pro-drugs of the KDAC inhibitor vorinostat (SAHA), Tetrahedron, https://doi.org/
10.1016/j.tet.2020.131170