Synthesis and evaluation of 3-hydroxy-3-phenylpropanoate ester-AZT conjugates as potential dual-action HIV-1 Integrase and Reverse Transcriptase inhibitors

Article abstract of DOI:10.1016/j.bmc.2015.10.039

Novel 3-hydroxy-3-phenylpropanoate ester-azidothymidine (AZT) conjugates have been prepared using Baylis-Hillman methodology, and their potential as dual-action HIV-1 Integrase and Reverse Transcriptase inhibitors has been explored using enzyme inhibition

Full text of DOI:10.1016/j.bmc.2015.10.039

Accepted Manuscript
Synthesis and Evaluation of 3-Hydroxy-3-phenylpropanoate Ester-AZT Con-
jugates as Potential Dual-action HIV-1 Integrase and Reverse Transcriptase
Inhibitors
Meloddy H. Manyeruke, Temitope O. Olomola, Swarup Majumder, Shaakira
Abrahams, Michelle Isaacs, Nicodemus Mautsa, Salerwe Mosebi, Dumisani
Mnkandhla, Raymond Hewer, Heinrich C. Hoppe, Rosalyn Klein, Perry T. Kaye
PII:
S0968-0896(15)30117-6
DOI:
http://dx.doi.org/10.1016/j.bmc.2015.10.039
BMC 12639
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
18 August 2015
17 October 2015
27 October 2015
Please cite this article as: Manyeruke, M.H., Olomola, T.O., Majumder, S., Abrahams, S., Isaacs, M., Mautsa, N.,
Mosebi, S., Mnkandhla, D., Hewer, R., Hoppe, H.C., Klein, R., Kaye, P.T., Synthesis and Evaluation of 3-
Hydroxy-3-phenylpropanoate Ester-AZT Conjugates as Potential Dual-action HIV-1 Integrase and Reverse
Transcriptase Inhibitors, Bioorganic & Medicinal Chemistry (2015), doi: http://dx.doi.org/10.1016/j.bmc.
2015.10.039
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Synthesis and Evaluation of 3-Hydroxy-3-phenylpropanoate Ester-AZT
Conjugates as Potential Dual-action HIV-1 Integrase and Reverse
Transcriptase Inhibitors
Meloddy H. Manyeruke,a Temitope O. Olomola,a S. Majumder,c Shaakira Abrahams,d
Michelle Isaacs,c Nicodemus Mautsa,c Salerwe Mosebi,d Dumisani Mnkandhla,c
Raymond Hewer,d Heinrich C. Hoppe,b,c Rosalyn Kleina,c and Perry T. Kayea,c *
aDepartment of Chemistry, bDepartment of Biochemistry and Microbiology, and cCentre for Chemico- and
Biomedicinal Research, Rhodes University, Grahamstown, 6140, South Africa. E-mail: P.Kaye@ru.ac.za
dBiomed, Advanced Materials Division, Mintek, Randburg, 2125, South Africa.
Abstract. Novel 3-hydroxy-3-phenylpropanoate ester-azidothymidine (AZT) conjugates have been
prepared using Baylis-Hillman methodology, and their potential as dual-action HIV-1 Integrase and
Reverse Transcriptase inhibitors has been explored using enzyme inhibition and computer modelling
techniques; their activity and HeLa cell toxicity have been compared with those of their cinnamate
ester analogues.

Key Words: Dual-action HIV-1 Integrase/Reverse Transcriptase inhibitors, Baylis-Hillman reaction, 3-hydroxy-
3-phenylpropanoate esters, cinnamate esters, azidothymidine (AZT).
1. Introduction
Different stages in the HIV replicative cycle have been identified as targets for
chemotherapeutic intervention and, while an increasing number of target-specific drugs have
been developed for clinical use, their toxicity, the emergence of drug resistance and problems
of patient non-compliance present medicinal chemists with serious challenges.1 HAART
(Highly Active Anti-Retroviral Therapy), a combination therapy designed to minimise the
development of drug resistance, involves the co-administration of an HIV-1 protease (PR)
inhibitor together with a nucleoside and a non-nucleoside HIV-1 reverse transcriptase (RT)
inhibitor.2 Designed Multiple Ligands (DMLs)3-5 or “portmanteau” drugs, as they are also
known,6 offer an alternative strategy in which different, target-specific, drug-like moieties are
combined in the same molecule, and we have previously reported research on Baylis-
Hillman-derived coumarin-AZT conjugates (e.g. 1) as potential, dual-action HIV-1 PR/RT
inhibitors,7-9 Wang et al.6 have reported compound 2 as a dual-action HIV-1 RT/IN inhibitor
with nanomolar IC50 values for both enzymes.
__________
Corresponding author: Prof Perry Kaye. E-mail: P.Kaye@ru.ac.za
1

The apparent HIV-1 IN inhibition properties of caffeic acid phenethyl ester 3 and L-chicoric
acid 4 have been attributed to the presence of cinnamate ester moieties,10,11 while
azidothymidine (AZT) is a known HIV-1 RT inhibitor.12 We now report the preparation of
series of novel 3-hydroxy-3-phenylpropanoate ester-AZT conjugates and the evaluation of their
potential, and that of their cinnamate ester analogues, as dual-action HIV-1 IN/RT inhibitors.
2. Results and Discussion
2.1. Chemistry
Access to the 3-hydroxy-3-phenylpropanoate ester-AZT conjugates 11a-e and 15a-c involved
the application of Baylis-Hillman methodology, the general approach being summarised by
the formal retrosynthetic analysis summarised in Scheme 1. Baylis-Hillman reactions
between salicylaldehydes and methyl or ethyl acrylate are characterised by uncontrolled
cyclisation to complex mixtures of 2H-chromene and coumarin derivatives,13,14 and isolation
of Baylis-Hillman adducts per se requires protection of the phenolic group through
benzylation. The tert-butyl acrylate-derived Baylis-Hillman adducts, however, are stable and
isolable and, in this paper, we report the use of both approaches.15 The O-benzylated
salicylaldehydes 12a-c were obtained in yields of 65-87% by boiling mixtures of the
salicylaldehydes 5a-c and benzyl bromide in aqueous NaOH under reflux for 30 minutes. [An
alternative approach involving the use of K2CO3 and KI for a period of 12 hours gave lower
yields (≤ 54%).] Reaction of the non-benzylated salicylaldehyde derivatives 5b-e with tert-
butyl acrylate 6a, and the O-benzylated salicylaldehyde derivatives 12a-c with methyl
acrylate 6b, led to the isolable Baylis-Hillman adducts 7a-e and 13a-c, respectively. As α,β-
unsaturated esters, the adducts 7b-e and 13a-c underwent conjugate addition by
2

propargylamine 8 to afford the corresponding diastereomeric β-hydroxy products (9b-e and
14a-c). Generally, chromatography afforded a single diastereomer but, in two cases, both
diastereomers (14a1,14a2 and 14b1,14b2 ) were isolated.
In the final steps, the target molecules 11b-e and 15a,b1,b2,c were obtained via “Click
Chemistry” (CuAAC) reactions16 involving AZT 10 and the corresponding alkynylated
intermediates. Flash chromatography, following work-up, afforded the products 15a-c in
yields of up to 97%. Further purification using preparative layer chromatography (PLC)
permitted the isolation of the 5-bromophenyl diastereomers (15b1 and 15b2) in a
diastereomeric ratio (based on the isolated yields) of 72:36, respectively. In all other cases,
NMR analysis of the chromatographed ester-AZT conjugates indicated the presence of a
single diastereomer. Access to the cinnamate ester analogues 17a-e, reported elsewhere,17
involved addition-elimination (effectively, allylic substitution) steps which led to the
unsaturated cinnamate ester intermediates 16a-e (Scheme 3).
Scheme 1.
3

Scheme 2
Scheme 3.
4

2.2. Enzyme Inhibition Assays
Having been designed as potential HIV-1 RT/IN dual-action inhibitors, the target compounds
11a-e, 15a-c and 17a-e were evaluated for their inhibition of each enzyme using the
established inhibitors, nevirapine and raltegravir, as standards for the RT and IN assays,
respectively The propargylated precursors (9a-e, 14a-c and 16a-e) of all three series were
also assayed, in their own right, for HIV-1 IN inhibition potential. The 3-hydroxypropanoate
esters (9) and (11) contain a free phenolic hydroxyl group and the bulky, hydrophobic tert-
butyl moiety, whereas their ester analogues (14) and (16) are methyl esters and, due to O-
benzylation, lack a free, phenolic hydroxyl group. The results of these bioassays are
summarised in Tables 1 and 2.
Table 1.
HIV-1 IN enzyme activity data at inhibitor concentrations
of 10 µM for compounds 9a-e, 14a-c and 16a-e .
IN activitya
Compd.
R1
R2
(%)b
Br
Cl
H
H
83.9 (1.8)
80.6 (2.2)
48.3 (0.8)
63.2 (9.8)
9b
9c
9d
9e
H
OMe
OEt
H
H
H
H
H
H
H
84.5 (2.6)
81.8 (5.1)
56.9 (4.0)
52.0 (2.1)
48.3 (11.5)
14a1
14a2
14b1
14b2
14c
H
Br
Br
Cl
H
H
H
H
31.01 ± 0.14
69.0 (0.1)
16a
16a
Br
Cl
H
H
27.1 (10.0)
90.9 (3.2)
84.8 (6.2)
48.7 (4.9)
16b
16c
16d
16e
H
OMe
OEt
H
a Raltegravir decreases IN activity to 0.76 ± 2.53% under
comparable conditions.
bPercentage enzyme activity relative to untreated controls,
followed, in parentheses, by the standard deviation.
5

Table 2.
HIV-1 RT and IN enzyme activity and toxicity (HeLa cell activity) data
for compounds 11a-e, 15a-c and 17a-e.
Compd.
R1
R2
RT activitya
(%)c
IN activityb
(%)c
HeLa cell
activity (%)c
Br
Cl
H
H
73.7 (7.3)
109.4 (5.1)
68.6 (4.6)
92.2 (13.2)
42.2 (8.1)
54.4 (4.9)
78.4 (5.7)
63.3 (1.9)
85.0
88.8
79.9
95.9
11b
11c
11d
11e
H
OMe
OEt
H
H
H
H
H
H
80.6 (1.5)
73.7 (7.3)
94.2 (11.1)
89.7 (6.7)
96.7 (1.8)
72.1 (8.2)
73.7 (5.9)
67.4 (9.9)
90.9
92.3
84.7
96.3
15a
Br
Br
Cl
15b1
15b2
15c
H
H
118.4 (1.2)
108.3 (2.9)
97.8 (12.8)
95.7 (13.8)
77.5 (16.1)
64.0 (1.0)
70.2 (12.8)
91.0 (4.8)
56.6 (9.9)
81.3 (8.0)
104
17a
17b
17c
17d
17e
Br
Cl
H
H
56.4
104.8
57.7
94.8
H
OMe
OEt
H
a
At 20 µM. (Nevirapine: 9.45% RT activity under comparable conditions.)
b At 10 µM. (Raltegravir reduces IN activity to 0.76 ± 2.53% under comparable conditions.)
c Followed, in parentheses, by the standard deviation.
6

It is evident that all of the propargylated intermediates (Table 1), which contain either 3-
hydroxy-3-phenylpropanoate or cinnamate ester moieties, exhibit some measure of HIV-1 IN
inhibition activity at 10 µM. The % IN activity data for four of them, however, are less than
50% [9d (48.3%); 14c (48.3%); 16b (27.1%); and 16e (48.7%)]. The results for these four
compounds thus correspond to IC50 values of less than 10 µM, indicating significant HIV-1
IN inhibitory promise.
The 3-hydroxy-3-phenylpropanoate ester-AZT conjugates 11b-e and 15a-c (effectively hydrated
cinnamate ester analogues), together with the cinnamate ester-AZT conjugates 17a-e, all
exhibit a measure of HIV-1 IN inhibitory activity (Table 2); however, only one of these
ligands (11b) would have an IC50 value of less than 10 µM. In some cases, the % IN activity
observed for the propargylated precursors remains essentially unchanged on conversion to
the corresponding ester-AZT conjugates [9e and 11e ≈ 63%; 16c and 17c ≈ 91%]. In other
cases, the % IN activity increases [9d (48%) → 11d (78%); 14c (48%) → 15c (67%); and
16b (27%) → 17b (70%)] yet, in others the % IN viability decreases [9b (84%) → 11b
(42%); 9c (81%) → 11c (54%); and 16a (69%) → 17a (64%)].
The HIV-1 RT bioassay data, summarised in Table 2, reflect % RT activity in the presence of
the ligands 11b-e, 15a-c and 17a-e  each at a concentration of 20 µM. Although most of the
ligands show some RT inhibition activity, the levels are disappointing albeit better, in most
cases, than AZT itself (ca. 100.% RT activity at a concentration of 20 µM). Of course, AZT
serves as a nucleoside analogue (NRTI) pro-drug which requires in vivo triphosphorylation.18
This will not occur during in vitro enzyme inhibition assays, and the HIV-1 RT inhibition
data reflected in Table 2 is attributed to binding of the ligands to the allosteric non-nucleoside
(NNRTI) binding pocket. Triphosphorylated derivatives of two of the ligands were
subsequently obtained in low yield; while their 1H NMR spectra were dominated by the ethyl
31
signals corresponding to the triethylammonium internal salts, their P NMR spectra revealed
the expected three 31P signals, as reported by Van Calenbergh and co-workers19 for analogous
systems. Bioassay of these derivatives, however, failed to reveal any significant inhibition of
the HIV-1 enzyme.
7

Several patterns emerging from the foregoing analysis warrant comment.
i) In general, the inhibitory activities of the ligands against both HIV-1 RT and IN
appear to follow the order: 3-hydroxy-3-phenylpropanoate ester-AZT conjugates 11b-e
> O-benzylated 3-hydroxy-3-phenylpropanoate ester-AZT conjugates 15a-c >
cinnamate ester-AZT conjugates 17a-e.
ii) With few exceptions, incorporation of the AZT moiety appears to increase % IN
activity, suggesting that lengthening the linker between the IN- and RT-active
moieties might be advantageous.
iii) The 5’-bromo-3-hydroxy derivatives 11b and 15b1 show encouraging, simultaneous
promise against both enzymes.
iv) Most of the ester-AZT conjugates exhibit little or no cell toxicity (as illustrated in
Figure 1).
v) Four ester-AZT conjugates (11b, 11c, 11e and 17a), which exhibit greater or
equivalent IN inhibition than their propargylated precursors and low cell toxicity
levels (≥ 85% cell viability at 20 µM), will be considered as candidates for
bioassay against the live virus.
Figure 1. HeLa cell toxicity assay data for
ligands 11b-e, 15a-c and 17a-e at 20 µM.
2.3. Computer Modelling Studies
Docking of the parent cinnamate ester-AZT conjugate 17a into both the HIV-1 IN and the
HIV-1 NNRT receptor cavities was explored. Structures of the cinnamate ester-AZT
conjugate 17a and the respective proteins were prepared using Discovery Studio Visualizer.20
The ligand occupying the receptor cavity and all water molecules were removed from the
original PDB files; hydrogen atoms were added and each atom was assigned an Autodock
8

Type using AutoDock Tools (ADT). The Autodock 4.2 programme21 was used to explore the
binding mode of compound 17a using a flexible dock approach. The protein active-site and
surrounding residues were mapped using the AutoGrid 4.2 algorithm, and a generic algorithm
was used to perform the conformation search. Atom maps were generated for all potential
interactions between the ligand and the active site residues, while selected catalytic residues
were kept flexible. For docking calculations, Gasteiger partial charges22 were assigned to
compound 17a and non-polar hydrogen atoms were merged. All torsions were allowed to
rotate during docking. Docked conformations were then viewed and explored using
Discovery Studio Visualizer.20
In order to explore its binding mode, compound 17a was docked into the HIV-1 NNRT (PDB
1IKW)23 active-site with the Arg8 catalytic residues being kept flexible. As shown in Figure
2, the cinnamate ester-AZT conjugate 17a appears to fit well in the active-site exhibiting
potential hydrogen-bonding interactions between amino acid residues Lys101, Lys103,
Lys172, Ile180, Tyr318, Thr1139 and Glu1138, and both the cinnamate ester and the AZT
moieties, while the O-benzyl group occupies a hydrophobic pocket next to Leu234 and
Phe227 in the HIV-1 NNRT active site. However, based on the disappointing bioassay data,
it would seem that these binding features alone are not sufficient to facilitate RT inhibition
via binding at the HIV-1 NNRT allosteric site.
Since the full crystal structure of HIV-1 IN is not yet known, the cinnamate ester-AZT
conjugate 17a was docked in the core domain structure of the HIV-1 IN enzyme (PDB
1QS4)24 which contains the active-site receptor cavity. A charge of +2 was manually assigned
to the Mg cation in the active site and residues Asp64 and Asp116 were defined as flexible
residues for the simulated docking. The docked conformation of compound 17a in the HIV-1
IN receptor cavity exhibits potential hydrogen-bonding interactions with amino acid residues
Ser119, Lys156, Glu152 and Lys159 (Figure 3). The ligand adopts a conformation in which
the OH group is close to the Mg co-factor (a distance of 3.09 Å) but does not displace the
metal ion which remains complexed to the two aspartate residues. An overlay of the docked
conformation of compound 17a and the core domain crystal structure of the known inhibitor,
5CITEP,25 in the active-site of HIV-1 IN (Figure 4) shows reasonable overlap between the
two ligands; however, the fact that compound 17a extends beyond the binding limits of
5CITEP, might decrease the binding affinity and account for comparatively low, observed
levels of HIV-1 IN inhibition (64% enzyme viability at 10 µM) .
9

Figure 2. Docked conformation of cinnamate ester-AZT conjugate 17a in the HIV-1 NNRT active-
site (PDB 1IKW).23 Protein active-site residues are shown in wire-frame and coloured by atom type,
the ligand is shown as sticks coloured by atom type. Non-polar hydrogen atoms have been omitted
for clarity and H-bonding interactions are shown as green dashed lines with distances in Å.
Figure 3. Overlay of the docked conformations of cinnamate ester-AZT conjugate 17a (in yellow)
and 5CITEP as found in the crystal structure PDB 1QS424 (in atom type colours) in the HIV-1 IN
active-site (surface zone coloured by atom types). Mg2+ is shown as a green sphere and protein
residues are shown in wire-frame, coloured by atom type.
10

3. Conclusions
Baylis-Hillman methodology has been successfully used to access two series of 3-hydroxy-3-
phenylpropanoate ester-AZT conjugates 11a-e and 15a-c. Enzyme inhibition and computer-
modelling studies of these compounds and their cinnamate ester-AZT analogues 17a-e, as
potential dual-action HIV-1 IN and RT inhibitors, have been undertaken. Some of the HIV-1
IN inhibition levels are encouraging, but structural modification is clearly required to
improve the inhibitions levels.
4. Experimental
4.1. Synthesis
The synthesis of compounds 7b-e, 9b-e, 13a-c, 14b1, 16a-e and 17a-e have been reported
previously.17 The intermediates 14a1,a2,b1,c and the 3-hydroxypropanoate ester-AZT
conjugates 11b-e and 15a-c are all new and were obtained as follows.
4.1.1. The general procedure for the aza-Michael reaction of the Baylis-Hillman
adducts 7a-e and 13a-c with propargylamine is illustrated by the following example.
Propargylamine 8 (0.192 mL, 3 mmol) was added to a solution of methyl 3-(2-
benzyloxyphenyl)-3-hydroxy-2-methylenepropanoate 13a (0.33, 1.1 mmol) in dry THF (1.5
mL) and the mixture stirred at r.t. for 48 h. The reaction mixture was then concentrated in
vacuo to afford the crude product (0.30 g, 76%), which was flash chromatographed [on silica
gel; elution with hexane–EtOAc (1:1)] to afford two fractions comprising the diastereomeric
methyl 3-[2-(benzyloxyphenyl)-3-hydroxy-2-[(2-propynylamino)methyl]propanoate esters
14a1 and 14a2.
Diastereomer 14a1 (60%) as a yellow oil [HMRS: m/z calculated for C21H24NO4 (MH+)
354.1705. Found 354.1695]; δH (400 MHz; CDCl3) 2.16 (1H, s, C≡CH), 2.90 (1H, dd, J = 6.6
and 12.0 Hz, CHCHa), 3.01 (1H, dd, J = 4.7 and 12.1 Hz, CHCHb), 3.15 (1H, 2 x d, J = 4.0
Hz, CHCO), 3.32 (1H, dd, J = 2.6 and 6.2 Hz, CHaC≡CH), 3.39 (1H, m, CHbC≡CH) 3.55
(3H, s, OCH3), 5.10 (2H, 2 x overlapping d, OCH2), 5.32 (1H, d, J = 5.3 Hz, CHOH), 6.93
(1H, d, J = 8.2 Hz, ArH) 6.99 (1H, t, J = 7.5 Hz, ArH) and 7.20-7.50 (7H, overlapping m,
ArH); δC (100 MHz; CDCl3) 38.0 and 48.4 (2 × NCH2), 50.4 (CHCO), 51.5 (CH3), 70.0
(OCH2), 70.2 (CHOH), 71.6 and 81.4 (C≡CH), 111.4, 120.9, 127.16, 127.22, 128.0, 128.5,
128.6, 130.0 and 136.7 (Ar-C), 155.0 [ArC(OBn)] and 174.1 (C=O).
11

Diastereomer 14a2 (40%) as a yellow oil [HMRS: m/z calculated for C21H24NO4 (MH+)
354.1705. Found 354.1699]; max/cm-1 (C=O); δH (400 MHz; CDCl3) 2.19 (1H, s, C≡CH),
2.53 (1H, dd, J = 3.9 and 12.3 Hz, CHCHa), 3.15 (1H, q, J = 3.5 Hz. CHCO), 3.19-3.29 (2H,
m, CHCHb and CHaC≡CH), 3.41 (1H, m, CHbC≡CH), 3.76 (3H, s, OCH3), 5.12 (2H, d, J =
5.6 Hz, OCH2), 5.73 (1H, d, J = 3.5 Hz, CHOH), 6.94 (1H, d, J = 8.2 Hz, ArH), 7.04 (1H, t, J
= 7.5 Hz, ArH), 7.23-7.49 (6H, overlapping m, ArH) and 7.56 (1H, d, J = 8.0 Hz, ArH); δC
(100 MHz; CDCl3) 38.3 and 45.9 (NCH2), 47.5 (CHCO), 52.0 (CH3), 69.6 (OCH2), 71.7
(CHOH), 72.1 and 80.8 (C≡CH), 111.4, 120.8, 126.7, 127.1, 127.7, 128.3, 128.4, 131.2 and
136.9 (ArC), 154.5 [ArC(OBn)] and 174.1 (C=O).
4.1.1.1. The diastereomeric methyl 3-(2-benzyloxy-5-bromophenyl)-3-hydroxy-2-[(2-
propynylamino)methyl]propanoate esters 14b1 and 14b2 were obtained following flash
chromatography of the crude product (0.39 g, 84%).
Diastereomer 14b1 (56%) as a cream solid, m.p. 78-82 oC (Lit.17 97-99 °C).
o
Diastereomer 14b2 (44%) as cream solid, m.p. 68-70 C; [HMRS: m/z calculated for
C21H23BrNO4 (MH+) 432.0810. Found 432.0816]; max/cm-1 3336 (OH) and 1719 (C=O); δH
(400 MHz; CDCl3) 2.29 (1H, m, C≡CH), 2.58 (1H, dd, J = 3.7 and 12.4 Hz, CHCHa), 3.09
(1H, q, J = 3.4 Hz, CHCO), 3.20 (1H, dd, J = 3.2 and 12.5 Hz, CHCHb), 3.37 (2H, qd, J =
17.0 and 2.6 Hz, CH2C≡CH), 3.76 (3H, s, OCH3), 5.08 (2H, 2 x overlapping d, J = 12.0 Hz,
OCH2), 5.71 (1H, d, J = 3.1 Hz, CHOH), 6.80 (1H, d, J = 8.6 Hz, ArH), 7.27-7.48 (6H,
overlapping m, ArH) and 7.72 (1H, d, J = 2.7 Hz, ArH); δC (100 MHz; CDCl3) 38.3 and 45.6
(NCH2), 46.9 (CHCO), 52.1 (CH3), 70.0 (OCH2) 71.4 (CHOH), 72.4 and 80.6 (C≡CH),
113.1, 113.6, 126.7, 127.9, 128.5, 130.3, 130.9, 133.8, and 136.4 (ArC), 153.5 [ArC(OBn)]
and 173.8 (C=O).
4.1.1.2. Methyl 3-(2-benzyloxy-5-chlorophenyl)-3-hydroxy-2-[(2-propynylamino)methyl]-
o
propanoate 14c as a cream solid (0.41 g, 97%), m.p. 88-92 C; [HMRS: m/z calculated for
C21H23ClNO4 (MH+) 388.1316. Found 388.1320]; max/cm-1 3460 (OH) and 1717 (C=O); δH
(400 MHz; MeOD-d4) 2.48 (1H, t, J = 2.6 Hz, C≡CH), 2.66 (1H, m, CHCO), 2.90-3.00 (2H,
m, CHCH2), 3.19 (2H, d, J = 2.6 Hz, CH2C≡CH), 3.56 (3H, s, OCH3), 5.09 (2H, d, J = 2.7
Hz, OCH2), 5.27 (1H, d, J = 6.1 Hz, CHOH), 6.99 (1H, d, J = 8.8 Hz, ArH), 7.19 (1H, dd, J =
2.8 and 8.7 Hz, ArH) and 7.29-7.50 (6H, overlapping m, ArH); δC (100 MHz; MeOD-d4) 38.2
and 48.4 (NCH2), 52.1 (CHCO), 53.6 (CH3), 68.2 (CHOH), 71.6 (OCH2), 73.4 and 81.8
(C≡CH), 114.4, 127.0, 128.2, 128.7, 129.1, 129.2, 129.7, 134.5 and 138.1 (ArC), 155.3
[Ar(OBn)] and 175.1 (C=O).
12

4.1.2. General procedure for the preparation of the 3-hydroxypropanoate ester-AZT
conjugates is illustrated by the following example.
tert-Butyl 3-(5-bromo-2-hydroxyphenyl)-3-hydroxy-2-[(2-propynylamino)methyl]propanoate
9b (0.15 g, 0.4 mmol) was dissolved in CH3CN (2 mL) and AZT (0.12 g, 0.44 mmol), Et3N
(67 µL) and CuI (0.008 g) were added to the solution. The mixture was stirred for 48 hours
then diluted with DCM (10 mL), washed with saturated aq. NH4Cl solution (5 mL), followed
by brine (5 mL), and dried over anhydr. MgSO4. The organic solution was concentrated in
vacuo and the crude product purified by column chromatography [on silica gel; elution with
hexane-EtOAc (2:1)] to afford the tert-butyl 3-hydroxypropanoate ester-AZT conjugate 11b
as a pale yellow oil (0.32 g, 13%); [HMRS: m/z calculated for C27H36BrN6O8 (MH+)
651.1778. Found 651.1790]; max/cm-1 1683 (C=O); δH (400 MHz; CD3OD) 1.30 [9H, s,
C(CH3)3], 1.86 (3H, s, CH3), 2.6 – 3.1 (5H, overlapping m, CH2CHN, CHCH2N and CHCO),
3.7 - 3.8 (4H, overlapping m, CH2CN and CH2OH), 4.30 (1H, s, OCHCHN), 5.15 (1H, s,
CHOH), 5.36 (1H, s, OCHCH2OH), 6.44 (1H, q, J = 5.6 Hz, OCHN), 6.64 (1H, m, ArH),
7.15 (1H, m, ArH) 7.34 (1H, d, J = 10.0 Hz, ArH) 7.89 (2H, m, ArH); δC (100 MHz; CD3OD)
12.5 (CH3Ar), 28.2 [C(CH3)3], 39.1 (CH2CHN), 44.8 and 48.0 (CH2N), 52.1 (CHCO), 60.9
(CHN), 62.1 (CH2OH), 71.1 (CHOH), 82.2 [C(CH3)3], 86.4 (HOCH2CHO), 86.7 (NCHO),
111.6, 112.0, 118.2, 123.9, 131.7, 131.8, 132.0, 132.1, 138.3, 152.3, 155.1, 166.4 and 173.5
(ArC, C=C and C=O).
4.1.2.1. The tert-butyl 3-hydroxypropanoate ester-AZT conjugate 11c as a pale yellow solid
(0.036 g, 14%), m.p. 100-102 oC; [HMRS: m/z calculated for C27H36ClN6O8 (MH+) 607.0610.
Found 607.1104]; max/cm-1 1683 (C=O); δH (400 MHz; CD3OD) 1.29 [9H, s, C(CH3)3], 1.84
(3H, s, CH3), 2.58-2.97 (4H, m, CH2CHN, and CHCH2N), 3.03 (1H, s, CHCO), 3.60-3.94
(4H, m, CH2CN and CH2OH), 4.29 (1H, s, OCHCHN), 5.15 (1H, d, J = 6.2 Hz, CHOH), 5.35
(1H, s, OCHCH2OH), 6.42 (1H, s, OCHN), 6.63 (1H, d, J = 8.3 Hz, ArH), 7.17 (1H, m,
ArH), 7.33 (1H, s, ArH), 7.67-8.05 (2H, m, ArH); δC (100 MHz; CD3OD) 12.5 (CH3Ar), 28.2
[C(CH3)3], 39.1 (CH2CHN), 44.8 and 48.2 (CH2N), 52.0 (CHCO), 60.9 (CHN), 62.1
(CH2OH), 71.1 (CHOH), 82.2 [C(CH3)3], 86.4 (HOCH2CHO), 86.6 (NCHO), 111.6, 118.3,
119.3, 124.0, 131.7, 132.1, 135.1, 138.2, 140.1, 152.2, 155.1, 166.4 and 173.5 (ArC, C=C and
C=O).
4.1.2.2. The tert-butyl 3-hydroxypropanoate ester-AZT conjugate 11d as a pale yellow oil
(0.16 g, 66%); [HMRS: m/z calculated for C28H39N6O9 (MH+) 603.2779. Found 603.2783];

max/cm-1 1683 (C=O); δH (400 MHz; CD3OD) 1.30 [9H, s, C(CH3)3], 1.87 (3H, s, CH3), 2.6
13

– 3.1 (5H, series of multiplets, CH2CHN, CHCH2N and CHCO), 3.31 (2H, s, CH2CN), 3.76-
3.84 (5H, overlapping s and m, OCH3 and CH2OH), 4.30 (1H, m, OCHCHN), 5.27 (1H, d, J
= 6.1 Hz, CHOH), 5.36 (1H, m, OCHCH2OH), 6.44 (1H, t, J = 6.4 Hz, OCHN), 6.7-6.9 (2H,
m, ArH), 7.89 (2H, m, ArH); δC (100 MHz; CD3OD) 12.5 (CH3Ar), 28.2 [C(CH3)3], 39.1
(CH2CHN), 44.7 and 48.1 (CH2N), 51.9 (CHCO), 56.6 (OCH3), 60.9 (CHN), 62.1 (CH2OH),
71.4 (CHOH), 82.1 [C(CH3)3], 86.4 (HOCH2CHO), 86.7 (NCHO), 111.5, 111.7, 119.9,
120.6, 124.1, 129.5, 138.3, 144.7, 146.7, 148.7, 152.3, 166.4 and 173.8 (ArC, C=C and
C=O).
4.1.2.3. The tert-butyl 3-hydroxypropanoate ester-AZT conjugate 11e as a pale yellow oil
(0.30 g, 10%); [HMRS: m/z calculated for C29H41N6O9 (MH+) 617.2935. Found 617.2956];

max/cm-1 1682 (C=O); δH (400 MHz; CD3OD) 1.32 [9H, s, C(CH3)3], 1.35 (3H, t, J = 6.9
Hz, OCH2CH3), 1.86 (3H, s, CH3), 2.68 (1H, ddd, J = 6.1, 8.6 and 14.3 Hz, CHaCHN), 2.7-
2.9 (3H, overlapping m, CHCH2N and CHbCHN), 3.07 (3H, td, J = 4.8 and 7.0 Hz, CHCO),
3.65-3.79 (2H, m, CH2CN), 3.81-3.90 (2H, m, CH2OH), 3.99-4.07 (2H, m, OCH2CH3), 4.30
(1H, dt, J = 3.0 and 5.7 Hz, OCHCHN), 5.25 (1H, d, J = 6.2 Hz, CHOH), 5.35 (1H, m,
OCHCH2OH), 6.43 (1H, t, J = 6.4 Hz, OCHN), 6.71 (1H, t, J = 7.9 Hz, ArH), 6.79 (1H, dd,
J = 1.6 and 8.1 Hz, ArH), 6.85 (1H, dd, J = 1.6 and 7.7 Hz, ArH), 7.88 (2H, m, ArH); δC (100
MHz; CD3OD) 12.5 (CH3Ar), 15.1 (CH2CH3), 28.2 [C(CH3)3], 39.0 (CH2CHN), 44.8 and
48.2 (CH2N), 52.0 (CHCO), 60.9 (CHN), 62.1 (CH2OH), 65.6 (OCH2CH3), 71.5 (CHOH),
82.0 [C(CH3)3], 86.4 (HOCH2CHO), 86.7 (NCHO), 111.6, 112.7, 119.9, 120.6, 124.0, 129.6,
138.3, 144.9, 146.9, 147.8, 152.3, 166.4 and 173.8 (ArC, C=C and C=O).
4.1.2.4. The methyl O-benzylated 3-hydroxypropanoate ester-AZT conjugate 15a as a pale
o
yellow solid (0.069 g, 55%), m.p. 64-66 C; [HMRS: m/z calculated for C31H37N6O8 (MH+)
621.2673. Found 621.2683]; max/cm-1 1683 (C=O); δH (400 MHz; CD3OD) 1.87 (3H, s,
CH3), 2.48 (1H, m, CHaCHN), 2.63 (1H, m, CHCHaN), 2.70-2.87 (2H, overlapping m,
CHCHbN and CHbCHN), 3.04 (1H, m, CHCO), 3.58 (3H, s, OCH3), 3.63-3.75 (3H,
overlapping signals, CH2OH and CHaCN), 3.84 (1H, d, J = 6.0 Hz, CHbCN), 4.22 (1H, br s,
OCHCHN), 5.07 (2H, s, OCH2Ph), 5.29 (2H, d, J = 7.6 Hz, CHOH and OCHCH2OH), 6.42
(1H, t, J = 6.4 Hz, OCHN), 6.96 (2H, m, ArH), 7.18-7.44 (7H, overlapping m, ArH), 7.72
(1H, d, J = 7.6 Hz, ArH) and 7.87 (1H, s, ArH); δC (100 MHz; CD3OD) 12.5 (CH3Ar), 39.0
(CH2CHN), 44.6 and 48.8 (CH2N), 52.2 (OCH3), 54.1 (CHCO), 60.8 (CHN), 62.1 (CH2OH),
68.7 (CHOH), 71.1 (OCH2Ph), 86.3 (HOCH2CHO), 86.6 (NCHO), 111.7, 113.0, 122.0,
14

123.7, 123.8, 128.3, 128.5, 128.9, 129.6, 129.8, 132.0, 138.2, 138.6, 147.0, 147.1, 152.2,
156.8, 166.4 and 175.6 (ArC, C=C and C=O).
4.1.2.5. The disatereomeric methyl O-benzylated 3-hydroxypropanoate ester-AZT conjugates
15b1 and 15b2.
o
Diastereomer 15b1 as a pale yellow solid (0.1 g, 72%), m.p. 96-100 C; [HMRS: m/z
calculated for C31H36BrN6O8 (MH+) 699.1778. Found 699.1805]; max/cm-1 1683 (C=O); δH
(400 MHz; CD3OD) 1.87 (3H, s, CH3), 2.53 (1H, m, CHaCHN), 2.64 (1H, m, CHCHaN), 2.79
(1H, m, CHCHbN), 2.87 (1H, t, J = 10.4 Hz, CHbCHN), 2.99 (1H, m, CHCO), 3.55 (3H, s,
OCH3), 3.7 (3H, overlapping signals, CH2OH and CHaCN), 3.84 (1H, d, J = 12.0 Hz,
CHbCN), 4.24 (1H, br s, OCHCHN), 5.06 (2H, d, J = 3.9 Hz, OCH2Ph), 5.23 (1H, d, J = 4.0
Hz, CHOH), 5.31 (1H, m, OCHCH2OH), 6.42 (1H, t, J = 6.4 Hz, OCHN), 6.91 (1H, d, J =
8.8 Hz, ArH), 7.2-7.4 (6H, overlapping m, ArH), 7.50 (1H, s, ArH), 7.78 (1H, s, ArH) and
7.88 (1H, s, ArH); δC (100 MHz; CD3OD) 12.5 (CH3Ar), 39.1 (CH2CHN), 44.6 and 48.8
(CH2N), 52.1 (OCH3), 53.7 (CHCO), 60.9 (CHN), 62.1 (CH2OH), 68.2 (CHOH), 71.5
(OCH2Ph), 86.4 (HOCH2CHO), 86.6 (NCHO), 111.7, 114.2, 115.0, 123.76, 123.83, 128.6,
129.1, 129.7, 131.1, 132.3, 134.9, 138.1, 138.2, 147.08, 147.13, 152.3, 155.8, 166.4 and
175.1 (ArC, C=C and C=O).
o
Diastereomer 15b2 as a pale yellow solid (0.05 g, 36%), m.p. 97-100 C; [HMRS: m/z
calculated for C31H36BrN6O8 (MH+) 699.1778. Found 699.1786]; max/cm-1 1683 (C=O); δH
(400 MHz; CD3OD) 1.87 (3H, s, CH3), 2.46 (1H, m, CHaCHN), 2.63 (1H, m, CHCHaN),
2.6.8-2.89 (2H, overlapping m, CHCHbN and CHbCHN), 3.03 (1H, m, CHCO), 3.59 (3H, s,
OCH3), 3.61-3.76 (3H, overlapping m, CH2OH and CHaCN), 3.83 (1H, d, J = 12.0 Hz,
CHbCN), 4.23 (1H, br s, OCHCHN), 5.07 (2H, m, OCH2Ph), 5.27 (2H, d, m, CHOH and
OCHCH2OH), 6.42 (1H, t, J = 6.0 Hz, OCHN), 6.96 (2H, t, J = 10.0 Hz, ArH), 7.14-7.43
(6H, overlapping m, ArH), 7.71 (1H, d, J = 7.8 Hz, ArH) and 7.87 (1H, s, ArH); δC (100
MHz; CD3OD) 12.5 (CH3Ar), 39.0 (CH2CHN), 44.5 and 48.8 (CH2N), 52.2 (OCH3), 54.2
(CHCO), 60.9 (CHN), 62.1 (CH2OH), 68.8 (CHOH), 71.2 (OCH2Ph), 86.4 (HOCH2CHO),
86.7 (NCHO), 111.7, 113.0, 122.0, 123.7, 128.3, 128.5, 129.0, 129.4, 129.6, 129.80, 129.84,
132.0, 138.2, 138.6, 147.1, 152.3, 156.8, 166.4 and 175.7 (ArC, C=C and C=O).
4.1.2.6. The methyl O-benzylated 3-hydroxypropanoate ester-AZT conjugate 15c as a pale
o
yellow solid (0.24 g, 92%), m.p. 86-90 C; [HMRS: m/z calculated for C31H36ClN6O8 (MH+)
655.2283. Found 655.2297]; max/cm-1 (C=O); δH (400 MHz; CD3OD) 1.86 (3H, s, CH3),
2.53 (1H, m, CHaCHN), 2.63 (1H, m, CHCHaN), 2.77 (1H, m, CHCHbN), 2.87 (1H, t, J =
15

10.7 Hz, CHbCHN), 2.99 (1H, m, CHCO), 3.54 (3H, s, OCH3), 3.62-3.75 (3H, overlapping
signals, CH2OH and CHaCN), 3.84 (1H, dd, J = 2.9 and 12.2 Hz, CHbCN), 4.24 (1H, m,
OCHCHN), 5.05 (2H, d, J = 4.0 Hz, OCH2Ph), 5.23 (1H, d, J = 4.0 Hz, CHOH), 5.30 (1H,
m, OCHCH2OH), 6.42 (1H, t, J = 6.4 Hz, OCHN), 6.95 (1H, d, J = 8.7 Hz, ArH), 7.15 (1H,
dd, J = 2.8 and 8.9 Hz, ArH), 7.24-7.42 (6H, overlapping m, ArH), 7.78 (1H, s, ArH) and
7.87 (1H, s, ArH); δC (100 MHz; CD3OD) 12.5 (CH3Ar), 39.0 (CH2CHN), 44.6 and 48.8
(CH2N), 52.1 (OCH3), 53.7 (CHCO), 60.9 (CHN), 62.1 (CH2OH), 68.2 (CHOH), 71.5
(OCH2Ph), 86.4 (HOCH2CHO), 86.6 (NCHO), 111.6, 114.5, 123.8, 123.9, 126.9, 128.1,
128.6, 129.1, 129.3, 129.7, 134.5, 138.1, 138.2, 147.0, 147.1, 152.2, 155.3, 166.4 and 175.1
(ArC, C=C and C=O).
4.2. Bioassay procedures.
4.2.1. HIV-1 RT assay. Quantification of the inhibitory effect of the ligands was performed
in triplicate using a commercially available HIV-RT kit (Roche Applied Science, USA) as
per the manufacturer’s instructions. The inhibitory activity of the reverse transcriptase
inhibitors was calculated as the percentage of the enzyme activity compared to a sample that
does not contain any inhibitor.
4.2.2. HIV-1 IN assay. The HIV-1 IN strand transfer inhibition assay was adapted from
previously described methods.26 Briefly, 0.15 µM double-stranded biotinylated donor DNA
(5’-biotin-GTGTGGAAAATCTCTAGCA-3’ and 5’-ACTGCTAGAGATTTTCCACAC-3’)
was added to the wells of streptavidin-coated 96-well microtiter plates (R&D Systems, USA).
Following incubation at room temperature for 60 minutes and a stringent wash step, 1 µM
purified recombinant HIV-1 subtype B IN (in the presence of MgCl2 and MnCl2) was
assembled onto the pre-processed donor DNA through incubation for 30 minutes at 22 °C.
Following a wash step, the test compounds and raltegravir were titrated into individual wells at
a final concentration of 10 µM. The microtiter plates were incubated for 30 minutes at 37 °C,
washed and the strand transfer reaction was initiated through the addition of 0.25 µM double-
stranded FITC-labelled target DNA (5’-TGACCAAGGGCTAATTCACT-FITC-3’ and 5’-
AGTGAATTAGCCCTTGGTCA-FITC-3’) in Hepes buffer containing MgCl2 and MnCl2.
After an incubation period of 60 minutes at 37 °C, the plates were washed as before and an
alkaline phosphatase (AP) – conjugated anti-FITC secondary antibody (Sigma, USA) was
added. Finally, the plates were washed and substrate (BluePhos, KPL, USA) was added to
16

allow for detection at 620 nm using a microplate reader (xMarkTM, Bio-Rad, USA). All
inhibition values are the average of triplicate experiments.
4.2.3. Cell toxicity assay. HeLa cells were cultured in DMEM supplemented with 10% foetal
bovine serum and penicillin/streptomycin/amphotericin B and plated at density of 2x104 cells
in 150µL medium per well in 96 well plates. The following day, compounds were added to
duplicate wells at a final concentration of 20 µM and incubated with the cells for 24 h. Cell
viability was assessed by adding 20 µL per well resazurin-based toxicology assay reagent
(Sigma-Aldrich), incubating for 2-4 h and measuring fluorescence using 560 nm and 590 nm
excitation and emission wavelengths, respectively, in a Molecular Devices Spectramax M3
plate reader. Fluorescence readings in compound-treated wells were used to calculate cell
viability as a percentage of the average readings obtained from wells without the test
compounds.
ACKNOWLEDGEMENTS
This research project was funded by the South African Medical Research Council (MRC)
with funds from National Treasury under its Economic Competitiveness and Support
Package. The authors also thank the MRC for a bursary (to M.H.M.), Rhodes University for
a bursary (to T.O.O.) and generous financial support, and Aspen Pharmacare for the supply of
azidothymidine (AZT) used for this work,
REFERENCES
[1]
Palella, F.J.; Delaney, K.M.; Moorman, A.C.; Loveless, M.O., Fuhrer, J.; Satten, G.A.
N. Engl. J. Med. 1998, 338, 853-860.
[2]
[3]
[4]
[5]
Chesney, M. AIDS Patient Care and STDs 2003, 17, 169-177.
Morphy, R; Rankovic, Z. J. Med. Chem. 2005, 48, 6523-6543.
Munoz-Torrero, D.; Camps, P. Curr. Med. Chem. 2006, 13, 399-422.
Yamamoto, M; Ikeda, S.; Kondo, H.; Inoue, S. Bioorg. Med. Chem. Lett. 2002, 12,
375-378.
17

[6]
[7]
[8]
[9]
Wang, Z.; Bennett, E.M.; Wilson, D.J.; Salomon, C.; Vince, R. J. Med. Chem. 2007,
50, 3416-3419.
Kaye, P.T.; Musa, M.A.; Nchinda, A.T.; Nocanda, X.W. Synth. Commun. 2004,
34(14), 2575-2589.
Olomola, T.O.; Klein, R.; Lobb, K.A.; Sayed, Y; Kaye, P.T. Tetrahedron Lett. 2010,
51, 6325-6328.
Olomola, T.O.; Klein, R.; Mautsa, N.; Sayed, Y.; Kaye, P.T. Bioorg. Med. Chem.
2013, 21, 1964-1971.
[10] Dayam, R.; Gundla, R.; Al-Mawsawi, L.Q.; Neamati, N. Med. Res. Rev. 2008, 28,
118-154.
[11] Bodiwala, H.S.; Sabde, S.; Gupta, P.; Mukherjee, R.; Kumar, R.; Garg, P.; Bhutani,
K.K.; Mitra, D.; Singh, I.P. Bioorg. Med. Chem. 2011, 19(3), 1256-1263.
[12] Morphy, R; Rankovic, Z. J. Med. Chem. 2005, 48, 6523-6543.
[13] Bacsa, J.; Kaye, P.T.; Robinson, R.S. S. Afr. J. Chem., 1998, 51, 47-54.
[14] Kaye, P.T.; Musa, M.A.; Nocanda, X.W.; Robinson, R.S. Org. Biomol. Chem., 2003,
1, 1133-1138.
[15] Kaye, P.T.; Musa, M.A.; Nocanda, X.W. Synthesis, 2003, (4), 531-534.
[16] Kolb, H.C.; Sharpless, K.B. Drug Discov. Today 2003, 8, 1128-1137.
[17] Olomola, T.O.; Klein, R.; Kaye, P.T. Tetrahedron, 2014, 70, 9449-9445, and
references cited therein.
[18] Furman, P. A.; Fyfe, J. A.; St. Clair, M. H.; Weinhold, K.; Rideout, J. L.; Freeman, G.
A.; Lehrman, S. N.; Bolognesi, D. P.; Broder, S.; Mitsuya, H.; Barry, D. W. Proc.
Natl. Acad. Sci. U.S.A. 1986, 83, 8333-8337.
[19] Toti, K.S.; Derudas, M.; Pertusati, F.; Sinnaeve, D.; Van den Broeck, F.;
Margamuljana, L.; Martins, J.C.; Herdewijn, P.; Balzarini, J.; Christopher McGuigan,
C.; Van Calenbergh, S. J. Org. Chem., 2014, 79, 5097−5112.
[20] Discovery Visual Studio, Release 3.1, San Diego: Accelrys Software Inc., 2011.
[21] G.M. Morris, G.M.; Goodsell, D.S ; Halliday, R.S.; Huey, R.; Hart, W.E.; Belew,
R.K.; Olson, A.J. J. Comput. Chem. 1998, 19, 1639-1662.
[22] Gasteiger, J.; Marsili, M.. Tetrahedron 1980, 36, 3219-3228.
[23] Lindberg, J.; Sigurdsson, S.; Lowgren, S.; Andersson, H.O; Sahlberg, C.; Noreen, R.;
Fridborg, K.; Zhang, H.; Unge, T. Eur. J. Biochem. 2002, 269, 1670-1677.
18

[24] Goldgur, Y.; Craigie, R.; Cohen, G.H.; Fujiwara, T.; Yoshinaga, T.; Fujishita, T.;
Sugimoto, H.; Endo, T.; Murai, H.; Davies, D.R. Proc. Natl. Acad. Sci. U.S.A. 1999,
96 (23), 13040-13043.
[25] Wei, C.; Mei, Y.; Zhang, D. Chem. Phys. Lett. 2010, 495, 121–124.
[26] D.J. Hazuda, J.C. Hastings, A.L. Wolfe, and E.A. Emini. Nucleic Acids Res. 1994, 22,
1121-1122.
19

Graphical Abstract
3-Hydroxy-3-phenylpropanoate ester-AZT conjugates have
been prepared and their potential as dual-action HIV-1
Integrase and Reverse Transcriptase inhibitors has been
explored and compared with their cinnamate ester analogues.