Design, synthesis, and biological evaluation of chicoric acid analogs as inhibitors of HIV-1 integrase

Article abstract of DOI:10.1016/j.bmc.2006.02.030

A series of analogs of the potent HIV-1 integrase (HIV IN) inhibitor chicoric acid (CA) was designed with the intention of ameliorating some of the parent natural product's undesirable properties, in particular its toxicity, instability, and poor membrane

Full text of DOI:10.1016/j.bmc.2006.02.030

Bioorganic & Medicinal Chemistry 14 (2006) 4552–4567
Design, synthesis, and biological evaluation of chicoric acid
analogs as inhibitors of HIV-1 integrase
Trevor T. Charvat,a, Deborah J. Lee,b, W. Edward Robinsonb,c and
A. Richard Chamberlina,c,
*
aDepartment of Chemistry, University of California, Irvine, CA 92697, USA
bDepartment of Pathology and Laboratory Medicine, University of California, Irvine, CA 92697, USA
cChao Family Comprehensive Cancer Center, UC Irvine Medical Center, Orange, CA 92868, USA
Received 14 December 2005; revised 9 February 2006; accepted 10 February 2006
Available online 9 March 2006
Abstract—A series of analogs of the potent HIV-1 integrase (HIV IN) inhibitor chicoric acid (CA) was designed with the intention
of ameliorating some of the parent natural product’s undesirable properties, in particular its toxicity, instability, and poor mem-
brane permeability. More than 70 analogs were synthesized and assayed for three types of activity: (1) the ability to inhibit 30-
end processing and strand transfer reactions using recombinant HIV IN in vitro, (2) toxicity against the CD4+ lymphoblastoid cell
line, MT2, and (3) anti-HIV activity against HIVLAI. CA analogs lacking one of the carboxyl groups of CA and with 3,4,5-trihydr-
oxycinnamoyl sidechains in place of the caffeoyl group of CA exhibited the most potent inhibition of HIV replication and end-pro-
cessing activity. Galloyl-substituted derivatives also displayed very potent in vitro and in vivo activities, in most cases exceeding the
inhibitory effects of CA itself. Conversely, analogous monocarboxy caffeoyl analogs exhibited only modest inhibition, while the cor-
responding 3,4-dihydroxybenzoyl-substituted compounds were devoid of activity.
ꢀ 2006 Elsevier Ltd. All rights reserved.
1. Introduction
some to generate the fully integrated provirus. Integrase
first catalyzes the removal of two nucleotides from the
30-end of each viral long terminal repeat, a reaction
termed ‘30-end processing.’ Next, the free 30-hydroxyls
undergo nucleophilic attack on the host DNA, a reac-
tion termed ‘strand transfer.’ Both of these reactions
can be quantified in vitro using recombinant IN and oli-
gonucleotides homologous to the viral long terminal re-
peats. Additionally, IN can catalyze the reversal of the
Human immunodeficiency virus (HIV) integrase (IN) is
one of three enzymes required by HIV to infect a host
cell. The other two, reverse transcriptase and protease,
are presently exploited as antiviral targets. Although
highly active antiretroviral therapy (HAART) regimens
combining three or more RT or protease inhibitors have
proven to effectively suppress uncontrolled viral replica-
tion, drug resistance and drug-induced side-effects can
hinder complete viral suppression. It thus follows that
the identification of an inhibitor of HIV IN would add
a valuable third component to the antiviral armamen-
tarium. HIV IN is an attractive target because there is
no known homolog in human cells, potentially minimiz-
ing the side-effects associated with other antiviral agents.
integration reaction in vitro,
‘disintegration.’
a reaction termed
L-Chicoric acid (CA) (1) is a potent and selective inhib-
itor of HIV IN. It exhibits IC50 values of 100–500 nM
for the end-processing/strand transfer reactions and
100–200 nM for the disintegration reactions, an ED50
concentration of 1–2 lM, and
a CT5 value of
Integrase is an essential viral enzyme which catalyzes the
covalent joining of the viral genome to the host chromo-
>200 lM, respectively (Fig. 1).1 Furthermore, the CA
scaffold is easily amenable to analog synthesis in order
to determine the structure–activity relationships with
HIV IN. Indeed, numerous SAR studies have been con-
ducted and these investigations have revealed several
structural characteristics requisite for inhibitory activity
and the types of modifications that are tolerated.1–6
Keywords: HIV integrase; Chicoric acid; SAR.
*
Corresponding author. Tel./fax: +1 949 824 7089; e-mail:
archambe@uci.edu
These authors contributed equally to this work.
0968-0896/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2006.02.030

T. T. Charvat et al. / Bioorg. Med. Chem. 14 (2006) 4552–4567
4553
activity exhibited has been attributed to inhibition of
viral entry, rather than integrase inhibition, although
resistance testing and real-time polymerase chain reac-
tion suggest IN is being inhibited in vivo.1,7–11 Converse-
ly, replacement of the caffeoyl groups with galloyl
moieties led to analogs with enhanced potency, but also
increased toxicity.3,4,12 Finally, activity is insensitive to
the stereochemistry of the tartaric acid subunit: (ꢀ)-L,
(+)-D-, and meso-CA all exhibit comparable HIV-IN
and antiviral activity.3,4 Based on these SAR data, there
should be considerable latitude in designing CA analogs
with improved pharmacological properties, particularly
those with modifications of the tartaric acid core region.
Identifying surrogates for the catechol groups will be a
greater challenge, but the fact that the corresponding
galloyl derivatives are active, despite a substantial differ-
ence in structure between gallic acid and caffeic acid, is
encouraging from that perspective as well.
L
-Chicoric Acid (CA)
Bioactivity
HO2C
CO2H
O
O
ED50 = 4.2 μM
CT5 = 264 μM
IC50 (End Processing) = 100-500 nM
IC50 (Disintegration) = 100-200 nM
O
O
HO
OH
HO
OH
1
Figure 1. Structure of L-CA and its bioactivity.
These structure–activity relationship analyses indicate
that at least one carboxylic acid moiety and a bisphenol
{as caffeoyl (3,4-dihydroxycinnamoyl) or galloyl (3,4,5-
trihydroxybenzoyl)} are necessary for anti-HIV activity
(Fig. 2). Removal of the carboxylic acids or replacement
with either cyclic or acyclic alkyl groups resulted in com-
parable in vitro HIV IN inhibition, but no antiviral
activity.4 Derivatives of glyceric acid, serine, isoserine,
and b-aminoalanine (which contain only one carboxylic
acid group) were shown to exhibit anti-HIV IN and
antiviral activity comparable to CA. These analogs also
revealed that conversion of the diester to the mono- and
diamide did not compromise HIV-IN or antiviral inhibi-
tion.1,4 Furthermore, replacing the acid with the corre-
sponding methyl ester resulted in comparable anti-HIV
IN activity; however, the antiviral inhibition by these
analogs was not examined.4 The length of the linker unit
between the heteroatoms has also been briefly examined.
Utilization of lysine in lieu of b-aminoalanine, which in-
creased the linker length from two to four carbons,
resulted in a 10-fold less potent compound with compa-
rable toxicity.1
Although CA and its analogs exhibit potent and selec-
tive activity toward HIV IN, the compounds possess
several structural characteristics that render them quite
poor drug candidates. An analysis of basic medicinal
chemistry principles (i.e., Lipinski’s ‘Rule of 5,’13 Ve-
ber’s bioavailability criteria,14 etc.) reveals that CA is a
rather poor drug candidate for several reasons: (1) its
limited cell permeability (due to the diacid moiety), (2)
the anticipated lability of its two ester linkages toward
hydrolytic enzymes, (3) the potential toxicity that is gen-
erally associated with catechol groups, of which there
are two, and (4) the relatively large number of rotatable
bonds that might limit oral bioavailability (according to
the rules recently published by Veber) (Fig. 3).14 On the
other hand, the very potent CA is an excellent lead com-
pound for optimization because there are readily appar-
ent potential solutions to each of these problems,
respectively: (1) replace the carboxyl groups with prece-
dented surrogates, for example, esters, (2) substitute
more robust linkages, such as amides or ethers, for the
ester groups, (3) identify catechol surrogates by generat-
ing and screening libraries of CA analogs with a variety
of heterocyclic rings in place of the catechol groups, and
(4) synthesize conformationally constrained analogs (or
libraries of analogs) that restrict or prevent rotation
around otherwise freely rotating single bonds in the par-
ent compound. Finally, the relatively simple structure of
CA, and the resultant synthetic accessibility of analogs,
is a further plus as a lead compound. In this paper, we
present our progress toward expanding the SAR around
the CA scaffold while concurrently increasing the mem-
brane permeability and linker stability of CA.
Substitution of the biscatechol moiety has also been ex-
plored. Converting the biscatechol into the correspond-
ing tetramethyl ether4 or replacement with thiophene1
moieties gave completely inactive compounds. Conver-
sion into the tetraacetate, on the other hand, has led
to varied results. Pommier and co-workers reported that
CA tetraacetate analogs exhibit potent inhibition
against HIV IN, but low to moderate antiviral activity.4
Conversely, Reinke et al. reported that CA tetraacetate
displayed significantly decreased HIV IN inhibition rel-
ative to CA and little antiviral activity.1 The antiviral
Required
Tolerated Modifications
Tartaric Acid Stereochemistry
Ester linker (Amide OK)
Aryl moiety (Galloyl OK)
1 Acid
Catechol
2. Chemistry
HO2C
O
CO2H
1 Acid
O
2.1. Synthesis of chicoric acid analogs with increased
stability
O
O
Ester Linker
Aryl Moiety
Catechol
Synthesis of the diamido CA analogs began with a thio-
nyl chloride-mediated Fisher esterification of the four
diamino acids: (DL)-2,3-diaminopropanoic acid (2),
(DL)-2,4-diaminobutanoic acid (3), (DL)-ornithine (4),
and (L)-lysine (5); followed by EDCI mediated biscou-
HO
OH
HO
OH
Tartaric Acid Stereochemistry
Figure 2. Chicoric acid SAR.

4554
T. T. Charvat et al. / Bioorg. Med. Chem. 14 (2006) 4552–4567
Problematic Functionality Requiring Modification
Chicoric Acid
Limits translocation across cell membrane
Susceptible to esterase hydrolysis
HO2C
O
CO2H
O
O
O
Metal Chelator
Susceptible to aerobic oxidation to o-quinone
HO
OH
HO
OH
15 Rotatable Bonds
> 1700 possible conformations
Figure 3. CA’s undesirable structural characteristics.
pling to three of the protected aromatic acids: 3,4-
dimethoxycarbonyloxybenzoic acid (10), 3,4,5-trimeth-
oxycarbonyloxybenzoic acid (11), and 3,4-dimethoxy-
carbonyloxycinnamic acid (12)3 (Scheme 1). The
resultant diamido esters were then saponified with lithi-
um hydroxide in a 1:1:0.5 solution of tetrahydrofu-
ran:water:methanol to afford globally deprotected
diamido CA analogs 25–36.
The second series of diamido CA analogs synthesized
contained the fourth aryl sidechain, 3,4,5-trihydroxycin-
namic acid (37) (Scheme 2). Synthesis of the CA deriva-
tives was accomplished in a similar manner to the
previously assembled analogs, except a permethyl ether
protected aromatic acid was employed (the only com-
mercially available trihydroxycinnamic acid derivative
is trimethoxycinnamic acid). The four previously em-
EDCI
NEt3
SOCl2
MeOH
CH2Cl2
reflux
CO2H
CO2Me
H2N
H2N
NH2
O
OH
NH2
m
0 ˚C−reflux
quantitative
m
HCl
2 HCl
n
2, m = 1
3, m = 2
4, m = 3
5, m = 4
6, m = 1
7, m = 2
8, m = 3
9, m = 4
MeO2CO
R
OCO2Me
10, R = H, n = 0
11, R = OCO2Me, n = 0
12, R = H, n = 1
22−83%
HO
m
MeO
O
O
O
O
O
O
m
NH HN
NH HN
LiOH
THF/Water/MeOH
(1:1:0.5)
n
n
10−65%
HO
R
R
OH
MeO2CO
R
R
OCO2Me
OCO2Me
OH
OH
OCO2Me
13, R = H, m = 1, n = 0
14, R = OCO2Me, m = 1, n = 0
15, R = H, m = 1, n = 1
16, R = H, m = 2, n = 0
17, R = OCO2Me, m = 2, n = 0
18, R = H, m = 2, n = 1
19, R = H, m = 3, n = 0
20, R = OCO2Me, m = 3, n = 0
21, R = H, m = 3, n = 1
22, R = H, m = 4, n = 0
23, R = OCO2Me, m = 4, n = 0
24, R = H, m = 4, n = 1
25, R = H, m = 1, n = 0
26, R = OH, m = 1, n = 0
27, R = H, m = 1, n = 1
28, R = H, m = 2, n = 0
29, R = OH, m = 2, n = 0
30, R = H, m = 2, n = 1
31, R = H, m = 3, n = 0
32, R = OH, m = 3, n = 0
33, R = H, m = 3, n = 1
34, R = H, m = 4, n = 0
35, R = OH, m = 4, n = 0
36, R = H, m = 4, n = 1
Scheme 1. Synthesis of diamido CA analogs.

T. T. Charvat et al. / Bioorg. Med. Chem. 14 (2006) 4552–4567
4555
1. NEt3
DCM or DMF
0 ˚C
SOCl2
MeOH
CO2H
NH2
CO2Me
H2N
H2N
NH2
m
0˚C−reflux
quantitative
CO2H
OMe
2.
m
HCl
2 HCl
2, m = 1
3, m = 2
4, m = 3
5, m = 4
6, m = 1
7, m = 2
8, m = 3
9, m = 4
MeO
MeO
37
EDCI
0˚C−60˚C
33−76%
CO2H
CO2Me
O
NH HN
O
m
O
O
m
BBr3
NH HN
CH2Cl2
0˚C
24−54%
HO
OH
OH
MeO
MeO
OMe
OMe
HO
OH
HO
OMe
MeO
38, m = 1
39, m = 2
40, m = 3
41, m = 4
42, m = 1
43, m = 2
44, m = 3
45, m = 4
Scheme 2. Synthesis of trihydroxycinnamoyl CA analogs.
ployed amino acids were esterified, biscoupled to the
aromatic acid, and finally globally deprotected with bor-
on tribromide to afford diamido CA analogs 42–45.
ond was the toxicity of each analog against the CD4+
lymphoblastoid cell line, MT2 (Fig. 5). The third and fi-
nal biological activity was the anti-HIV activity of each
compound against HIVLAI (Fig. 5). Each assay was per-
formed as described previously.1,3
2.2. Synthesis of chicoric acid analogs with improved
stability and membrane permeability
The first series of potentially cell permeable CA analogs
contain a pivaloyloxymethyl ester promoiety attached to
the pendant free acid of the amino acid linker. Synthesis
of the acyloxymethyl esters began with EDCI biscou-
pling of diamino ester 6 with the commercially available
permethoxyaromatic acids 37 and 46–47, followed by
LiOH-mediated hydrolysis of the methyl esters. The
resultant carboxylic acids 50–52 were deprotonated with
triethylamine, followed by addition of excess chloro-
methyl pivalate to afford the desired acyloxymethyl es-
ters (53–55) in good yield (Scheme 3).
4. Results and discussion
The first of CAs undesirable structural characteristics
addressed are the labile ester bonds. The undesired es-
ters were replaced by a significantly more robust dia-
mide-linkage. Based on the biscaffeoyl monoacid CA
analogs that exhibit potent in vitro HIV IN inhibition,1
a small library of amide-linked CA analogs were de-
signed in which both the aromatic ring and linker
length between the diamines are varied. Four different
aromatic acids (3,4-dihydroxybenzoic acid, gallic acid,
caffeic acid, and 3,4,5-trihydroxycinnamic acid) and
four different diamines {(DL)-2,3 diaminopropanoic
acid, (DL)-2,4-diaminobutyric acid, (DL)-ornithine, and
L-lysine} were utilized in which the linker length will
be two-to-five carbons. The above four acids were stra-
tegically chosen because they contain all the possible
permutations of aromatic hydroxyl groups and unsatu-
ration present in the two active aryl substituents (gallic
acid and caffeic acid): two or three hydroxy moieties at-
tached to either benzoic or cinnamic acid. The diamines
were also carefully chosen based on the HIV inhibition
(as mentioned in the CA SAR section) of biscaffeoyl
(DL)-2,3 diaminopropanoic acid (linker length equals
two) and biscaffeoyl L-lysine (linker length equals five).
However, linker lengths of three {(DL)-2,4-diaminobu-
tyric acid} and four {(DL)-ornithine} carbons have
not been examined.
A second series of CA analogs with potential enhanced
cell permeation possess a methyl ester promoiety again
attached at the carboxylic acid terminus. Employing
the previously synthesized permethylcarbonyloxy CA
analogs 13–24, the desired methyl ester derivatives were
obtained in 15–87% yield following sodium methoxide
mediated deprotection of the phenolic methylcarbonates
(Scheme 4).
3. Biology
Three biological activities were determined for each CA
analog. The first was the ability of each compound to
inhibit recombinant HIV IN in vitro in the 30-end pro-
cessing and strand transfer reactions (Fig. 4). The sec-

4556
T. T. Charvat et al. / Bioorg. Med. Chem. 14 (2006) 4552–4567
CO2Me
1. NEt3
O
O
DCM or DMF
0˚C
LiOH
NH HN
CO2Me
NH2
MeOH/H2O
H2N
n
n
93−99%
CO2H
2.
2 HCl
R
R
n
6
MeO
OMe
MeO
OMe
R
48, R = OMe, n = 0
49, R = H, n = 1
38, R = OMe, n = 1
MeO
EDCI
OMe
0˚C−60˚C
46, R = OMe, n = 0
47, R = H, n = 1
37, R =OMe, n = 1
72− 76%
O
O
O
O
CO2H
O
O
O
NH HN
O
Cl
O
O
n
n
NH HN
NEt3
R
R
DMF
70˚C
n
n
MeO
OMe
MeO
OMe
3h
R
R
40−60%
50, R = OMe, n = 0
51, R = H, n = 1
MeO
OMe
MeO
OMe
52, R = OMe, n = 1
53, R = OMe, n = 0
54, R = H, n = 1
55, R = OMe, n = 1
Scheme 3. Synthesis of pivaloyloxymethyl ester CA analogs.
MeO
O
MeO
O
O
O
O
m
O
m
NH HN
NH HN
NaOMe
n
n
MeOH
n
n
15−87%
MeO2CO
R
R
OCO2Me
OCO2Me
HO
R
R
OH
OCO2Me
OH
OH
13, R = H, m = 1, n = 0
14, R = OCO2Me, m = 1, n = 0
15, R = H, m = 1, n = 1
16, R = H, m = 2, n = 0
17, R = OCO2Me, m = 2, n = 0
18, R = H, m = 2, n = 1
19, R = H, m = 3, n = 0
20, R = OCO2Me, m = 3, n = 0
21, R = H, m = 3, n = 1
22, R = H, m = 4, n = 0
23, R = OCO2Me, m = 4, n = 0
24, R = H, m = 4, n = 1
56, R = H, m = 1, n = 0
57, R = OH, m = 1, n = 0
58, R = H, m = 1, n = 1
59, R = H, m = 2, n = 0
60, R = OH, m = 2, n = 0
61, R = H, m = 2, n = 1
62, R = H, m = 3, n = 0
63, R = OH, m = 3, n = 0
64, R = H, m = 3, n = 1
65, R = H, m = 4, n = 0
66, R = OH, m = 4, n = 0
67, R = H, m = 4, n = 1
Scheme 4. Synthesis of diamido CA methyl ester analogs.
The second of CA’s undesirable structural characteris-
tics addressed is the diacid moiety, which leads to its
poor cell permeation profile. The obvious method to
improve CA’s membrane permeability is to remove
one carboxylic acid to afford the unsymmetrical mono-
acid. Numerous monocarboxylic acids have been syn-

T. T. Charvat et al. / Bioorg. Med. Chem. 14 (2006) 4552–4567
4557
acidic phenols play in the CA analogs poor cell perme-
ation profile.
Analysis of the biological activity of the CA analogs re-
veals positive SAR progress was accomplished (Tables 1
and 2). Seven compounds (the boxed entries) exhibited
more potent anti-HIV activity relative to the lead com-
pound, L-CA. Examination of the catecholic sidechains
reveals that the free acid CA analogs with 3,4,5-trihydr-
oxycinnamoyl aryl substituents (entries 5, 9, 3, and 17)
displayed the strongest antiviral and end-processing
activity; for example, Orn-linked derivative 44 (entry
13) demonstrated an antiviral EC50 concentration of
1.1 lM and an anti-IN IC50 of 380 nM. The gallic
acid-coupled analogs (entries 4, 8, 12, and 16) were also
potent inhibitors of HIV replication, with EC50s of 2.5–
7.6 lM. The caffeoyl-substituted derivatives (with the
identical catecholic sidechains as CA), on the other
hand, exhibited less potent antiviral and end-processing
inhibition relative to CA, leading to anti-HIV EC50 con-
centrations of 4.5 to >62.5 lM (the upper limit of the as-
say). Finally, all the 3,4-dihydroxy-coupled derivatives
were inactive in the anti-HIV assay. The length of the
amino acid linker connecting the catecholic sidechains
also modulated the activity of the CA analogs. The
2,4-diaminobutyric acid (entries 6–9) and ornithine (en-
tries 10–13) linkers led to significantly more potent end-
processing and antiviral activity compared to the corre-
sponding 2,3-diaminopropanoic acid (entries 2–5) and
lysine (entries 14–17) derivatives (Fig. 4).
Figure 4. Representative IC50 determination. IN was incubated with
compound 33 and substrate for 1 h at 37 ꢁC. Substrate was separated
from products by denaturing polyacryl amide gel electrophoresis
(PAGE). Products and substrate were quantified by phosphorimager
analysis and the IC50 calculated through the linear portion of the curve
using CalcuSyn for Windows. Lane 1, substrate alone; lanes 2–4 IN
plus substrate; lane 5, 25 lM L-TA (negative control); lane 6, 25 lM
L-CA (positive control); lanes 7–24, IN plus substrate plus compound
33 in triplicate with concentrations in micromolar above the lanes.
STP, strand transfer products; arrow, V1/V2 substrate; dot, 30-end-
processing products.
Although the potency of numerous CA analogs im-
proved relative to CA, the cytotoxicity of the amide-
linked compounds also increased. The toxicity effects
appear unaffected by the diaryl substituent, as all four
aryl moieties displayed comparable cytotoxicity levels.
Thus, the diamide linker appears responsible for the
lethal effects of the compounds (Fig. 5).
Figure 5. Representative cell toxicity and anti-HIV results. MT-2 cells
were incubated in triplicate with 2-fold serial dilutions of compound 33
for 1 h at 37 ꢁC. For cell toxicity determination (closed circles), cells
were incubated at 37 ꢁC for three days and then harvested for cell
viability determined using Finter’s neutral red dye. For anti-HIV
activity (open circles), HIVLAI was added at a multiplicity of infection
<1 and the cells were incubated for three days at 37 ꢁC. Cells were
harvested for cell viability and quantified using Finter’s neutral red
dye. The percentage of cells protected from HIV-induced death was
calculated at each concentration. Points are means of triplicates.
The increase in cytotoxicity levels associated with the
diamido CA analogs in turn led to reduced therapeu-
tic indices (TIs) relative to CA (TI = 63). The 3,4,5-
trihydroxycinnamoyl- and galloyl-coupled analogs
were the most selective, leading to TIs generally
greater than 10. The most potent and selective inhib-
itor of HIV was analog 44 (entry 13), exhibiting a TI
of 33.6.
thesized and shown to be potent HIV IN inhibitors
in vitro;1,4 however, the ability of all these analogs to
translocate into the cell has not been examined. Our ef-
forts were thus focused on monocarboxylate analogs,
but it is possible that they will also be cell imperme-
able. Thus, we also investigated several pro-drug strat-
egies to promote permeation of the analogs into the
cell. Numerous pro-drug strategies have been devel-
oped that successfully convert cell impermeable com-
pounds into membrane permeable counterparts. These
strategies are generally based on transforming the car-
boxylic acid to an ester or related derivative that con-
tains either a lipophilic moiety or functional group that
is a substrate for a membrane transport system. The
pro-drug strategies we evaluated are based on the en-
hanced lipophilicity method and include methyl and
pivaloyloxymethyl esters. Both ester pro-drugs have
been shown to be enzymatically cleaved in vivo to re-
veal the target acid.19 Furthermore, methylation of
the phenols was performed to examine the role the
Conversion of the pendant free acid of the CA analogs
to the corresponding methyl ester (Table 2) generally
led to a considerable decrease in both antiviral and
end-processing activity. However, the galloyl-coupled
esters (entries 20, 26, 30, and 34) retained substantial
levels of potency (the 3,4,5-trihydroxycinnamoyl deriva-
tives were not synthesized). Similarly, methylating the
catechols (entries 21–23, 27, 31, and 35) abolished all
anti-HIV inhibition (due in part to a significantly de-
creased solubility profile). Thus, simple ester and methyl
ether pro-drug strategies appear ineffective as methods
to improve the bioactivity of the diamido CA analogs.
Finally, the pivaloyloxymethyl ester pro-drug strategy
was unable to be evaluated due to the poor aqueous sol-
ubility of the resultant esterified CA analogs.

4558
T. T. Charvat et al. / Bioorg. Med. Chem. 14 (2006) 4552–4567
Table 1. Biological analysis—amide-linked CA analogs with increased stability
HO
O
O
O
n
NH HN
R1
R1
R1
Entry Compound
n
Cell Toxicity Anti-HIV TI
Anti-IN
μ
μ
( M)
( M)
3'-Processing
μ
μ
CT5
264
39
EC50
4.2
10 M Screen IC50 ( M)
NAa
1
2
1 (CA)
25
-
-
62.9
0.4
0.53
BAb
NAa
2%
1
94
CAc
3
4
27
26
42
28
30
29
43
31
33
32
44
34
36
35
45
1
1
1
2
2
2
2
3
3
3
3
4
4
4
4
30
32
30
75
39
56
29
17
38
29
37
104
51
38
39
4.5
3
6.7
10.7
18.8
-
100%
100%
0.81
0.38
GAd
TAe
BAb
CAc
GAd
TAe
BAb
CAc
GAd
TAe
BAb
CAc
GAd
TAe
NAa
0%
NAa
NAa
2.3
5
1.6
6
>62.5
67.2
2.7
7
0.6
20.7
17.1
-
93%
100%
8
0.54
NAa
0%
NAa
NAa
2.9
9
1.7
10
11
12
13
14
15
16
17
>62.5
7.2
5.3
11.6
33.6
-
90%
95%
2.5
0.39
0.38
NAa
0%
1.1
NAa
NAa
2.25
0.81
>62.5
>62.5
7.6
-
78%
100%
5.0
22.9
NAa
1.7
aExperiment not performed.
b3,4-Dihydroxybenzoyl.
cCaffeoyl.
dGalloyl.
e3,4,5-Trihydroxycinnamoyl.
5. Conclusions
inhibition of HIV replication and end-processing, while
the b-aminoalanine- (two carbons) and lysine- (five car-
bons) coupled derivatives exhibited diminished in vitro
and in vivo activity. Finally, initial attempts to mask
the problematic carboxylic acid and catecholic function-
ality were unsuccessful, as the methyl ester and permeth-
yl ether CA analogs failed to improve the bioactivity of
the parent compounds.
The CA analogs with 3,4,5-trihydroxycinnamoyl aryl
sidechains exhibited the most potent inhibition of HIV
replication and end-processing activity. The galloyl-
substituted derivatives also displayed very potent in vitro
and in vivo activity, in most examples exceeding the
inhibitory effects of CA. Conversely, the caffeoyl-cou-
pled analogs led to modest inhibition, while the 3,4-di-
hydroxybenzoyl-substituted compounds were devoid of
activity.
6. Experimental
The linker length of the amino acid bridge also played a
significant role in the bioactivity of the CA analogs.
Interestingly, the c-aminoalanine and ornithine linkers,
with three and four carbons, respectively, led to potent
6.1. General information
Nuclear magnetic resonance spectra were recorded on a
Bruker DRX 500 (500 MHz), GN 500 (500 MHz), or a

T. T. Charvat et al. / Bioorg. Med. Chem. 14 (2006) 4552–4567
4559
Table 2. Biological analysis—amide-linked CA analogs with improved stability and membrane permeability
R2O
O
O
O
NH HN
R1
R1
Entry
Compound
n
R1
R2
Cell toxicity (lM)
Anti-HlV (lM)
TI
Anti-IN 30-processing
10 lM Screen IC50 (lM)
CT50
EC50
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
66
68
67
55
56
38
69
71
70
39
72
74
73
40
75
77
76
41
63
64
65
1
1
1
1
1
1
2
2
2
2
3
3
3
3
4
4
4
4
1
1
1
BAc
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
AMi
AMi
AMi
29
49
>62.5
58
10
—
0.8
1.5
—
—
—
—
—
3.6
—
—
—
4.2
—
—
—
1.7
—
—
—
—
0%
NAa
1.95
0.89
—
CAd
98%
100%
—
GAe
15
OMe-GAf
OMe-CAg
OMe-TAh
BAc
Insolb
Insolb
Insolb
86
—
—
—
—
—
—
—
>62.5
>62.5
19.4
ꢁ62.5
>62.5
>62.5
10
0%
NAa
NAa
0.44
NAa
NAa
NAa
0.74
NAa
NAa
NAa
2.4
CAd
GAe
66
69
52%
100%
NAa
0%
OMe-TAh
BAc
282
37
CAd
GAe
33
42
99%
96%
NAa
0%
OMe-TAh
BAc
56
70
ꢁ31.3
>62.5
>62.5
35.9
ꢁ31.25
—
CAd
GAe
68
62
55%
100%
NAa
—
OMe-TAh
OMe-GAf
OMe-CAg
OMe-TAh
104
Insolb
Insolb
Insolb
NAa
—
—
—
—
—
—
—
a Experiment not performed.
b Compound was insoluble under assay conditions.
c 3,4-Dihydroxybenzoyl.
d Caffeoyl.
e Galloyl.
f Galloyl 3,4,5-trimethylether.
g Caffeoyl 3,4-dimethylether.
h 3,4,5-Trimethoxycinnamoyl.
i Pivaloyloxymethyl.
Bruker DRX 400 (400 MHz) spectrometer. Chemical
shifts are reported in ppm (d) referenced to CHCl3
(7.24 ppm) or CH3OH (3.31 ppm) for 1H NMR and
CDCl3 (77.0 ppm) or CD3OD (49.15 ppm) for 13C
NMR. Coupling constant, JHH, values are given in hertz.
Data are reported as follows: chemical shift, multiplicity
(app, apparent; par obsc, partially obscured; br, broad; s,
singlet; d, doublet; t, triplet; q, quartet; qn, quintet; and
m, multiplet), coupling constant, and integration. Infra-
red spectra (IR) were taken with a Perkin and Elmer
Model 1600 Series FTIR spectrophotometer. Optical
rotations were obtained with a JASCO DIP-360 digital
polarimeter. Melting points (mp) were obtained from a
Laboratory Devices Mel-Temp melting-point apparatus
and are reported uncorrected. High resolution mass
spectrometry was obtained from the Departmental Mass
Spectrometry facility. Thin-layer chromatography
(TLC) was performed on 0.25 mm Merck pre-coated sil-
ica gel plates (60 F-254), and silica gel chromatography
was performed using ICN 200–400 mesh silica gel. Rota-
ry evaporation was used to remove solvents. Inert atmo-
sphere operations were conducted under nitrogen passed
through a Drierite tube in oven- or flame-dried glass-
ware. Anhydrous tetrahydrofuran (THF), methylene
chloride (CH2Cl2), diethyl ether (Et2O), methanol
(MeOH), ethanol (EtOH), dimethylformamide (DMF),
diisopropylamine, and triethylamine were filtered
through two columns of activated basic alumina under
Ar(g) and transferred under Ar(g) in a solvent purifica-
tion system designed and manufactured by J. C. Meyer.
All reagents were used as purchased from Aldrich, Lan-
caster, or Acros; unless stated otherwise. The term ‘shak-
en’ when used below refers to agitation of the reaction
mixture on a standard orbital shaker.
6.2. Inhibition of integrase
Integrase inhibition assays were performed as de-
scribed previously.1,3,9,10,12,15 Briefly, the V1 oligonu-
cleotide was 50-end labeled using c-[32P]ATP and T4
polynucleotide kinase. The labeled oligonucleotide
was annealed to its complementary V2 oligonucleo-
tide.16 In a final volume of 20 ll, 1 ll of each com-
pound was incubated with 1.5 pmol of recombinant
HIV IN purified from Escherichia coli and 2 pmol of
V1/V2 in a reaction buffer of 20 mM 4-(2-hydroxyeth-

4560
T. T. Charvat et al. / Bioorg. Med. Chem. 14 (2006) 4552–4567
yl)-1-piperazineethanesulfonic acid (HEPES), pH 7.5,
10 mM dithiothreitol, 0.05% Nonidet P-40, and
10 mM MnCl2 at 37 ꢁC for 1 h. Reactions were then
stopped by addition of EDTA to a final concentration
of 18 mM. Products were separated from substrate by
electrophoresis through a denaturing 15% polyacryl-
amide gel and quantified using a Storm phosphorim-
ager (Molecular Dynamics) and ImageQuant
software. Analogs were first screened for activity at
10 lM; any compound that inhibited HIV IN by
50% or greater at that concentration was selected for
calculation of the fifty percent inhibitory concentra-
tion (IC50). In triplicate reactions, with inhibitor con-
centrations ranging from 30 nM to 10 lM, active
inhibitors were studied. Following quantification, the
IC50 was calculated through the linear range of the
log dose–response curve using CalcuSyn for Windows
software (Biosoft).
by low-speed centrifugation followed by filtration
through 0.45 lm cellulose acetate filters.
6.4. Preparatory HPLC purification
The target compound was purified via reverse-phase
preparatory HPLC utilizing a Column Engineering Ma-
trix C18 column (5 lm, 20 · 50 mm, 0.1% TFA/H2O
(eluent A):0.08% TFA/MeCN (eluent B)).
6.5. Analytical HPLC analysis
The purity of the target compound was analyzed via re-
verse-phase HPLC utilizing a Column Engineering Ma-
trix C18 column (5 lm, 4.6 · 50 mm, 95:5–5:95 0.1%
TFA/H2O (eluent A):0.08% TFA/MeCN (eluent B) over
10 min).
6.6. (DL)-2,3-Diaminopropionic acid methyl ester dihy-
drochloride (6)
6.3. Cell toxicity and anti-HIV activities of CA analogs
Cell toxicity and anti-HIV activities of each CA analog
were determined as described previously.1,3,9,10,12,15,17,18
Each analog was dissolved in H2O, 25% EtOH, or 33%
dimethylsulfoxide plus 33% EtOH, depending on its sol-
ubility profile. In triplicate wells of a 96-well plate, each
analog was diluted in 2 fold serial dilutions. Next,
approximately 5 · 105 MT2 cells were added and the
plates were incubated for 72 h at 37 ꢁC. Cells were trans-
ferred to poly-L-lysine coated 96-well plates, stained with
Finter’s Neutral red dye, and the viability was deter-
mined as the A540. The CT5 was calculated through the
linear range of the dose response curve using CalcuSyn
for Windows. The CT5, a dose where 95% of the cells
are viable, was chosen because it is a truly non-toxic
dose. Since HIV is an obligate intracellular parasite,
any anti-HIV activity at the CT5 is due to an effect of
the analog on viral, rather than cellular, proteins.
To a dry 250 mL three-necked round-bottomed flask
fitted with a drying tube was added (DL)-2,3 diamino-
propanoic acid hydrochloride (2.00 g, 14.23 mmol) and
77 mL methanol. The heterogeneous solution was
cooled to 0 ꢁC, and thionyl chloride (7.8 mL,
106.8 mmol) was added dropwise over 15 min. The reac-
tion mixture quickly became homogeneous and was stir-
red for 20 min at 0 ꢁC. The reaction mixture was
subsequently fitted with a condenser and heated at reflux
for 5 h. The resultant solution was concentrated in vac-
uo, the crude product was diluted with methanol, and
the solution was concentrated again to remove any
residual acid. The crude solid was then placed on the
high vacuum pump overnight to afford 2.71 g (100%)
of 6 as a white solid, which was used directly in the sub-
sequent reactions (e.g., 6.7, 6.8, and 6.9).
6.7. (DL)-2,4-Diaminobutyric acid methyl ester dihydro-
chloride (7)
Once the CT5 was determined, each analog was tested
for anti-HIV activity. Each analog was 2 fold serially
diluted in triplicate. Next, 5 · 105 MT2 cells were add-
ed to each well and the cells were incubated for 1 h at
37 ꢁC. Next, approximately 5 · 105 infectious particles
of HIVLAI were added to each well and the cells were
incubated for approximately 72 h at 37 ꢁC. Cells were
harvested, stained for viability using Finter’s neutral
red, and viability calculated as described for cell tox-
icity assays. The fifty percent effective concentration
(EC50) was calculated based on protection from
HIV-induced cytopathic effect. Percent protection
was relative to cell controls with no virus added
(100% viable) and virus control infections in which
no analog was added to cells plus virus (0% viable).
The range between these two controls was between
0.25 and 0.35 A540. The therapeutic index, or TI,
was calculated as the ratio of CT5 to EC50. Both
MT2 cells and H9 cells infected with HIVLAI were cul-
tured in RPMI-1640 media supplemented with 25 mM
HEPES, 12.5% fetal bovine serum, and 2 mM L-gluta-
mine. All cells and virus were obtained from the
National Institutes of Health AIDS Research and Ref-
erence Reagent Program. HIVLAI was clarified of cells
Following the procedure for 6, diamine 7 (384 mg) was
produced in 99% yield from (DL)-2,4-diaminobutyric
acid dihydrochloride (362 mg, 1.90 mmol) and used
directly in the subsequent reactions (6.10, 6.11, and
6.12).
6.8. (DL)-2,5-diaminopentanoic acid methyl ester dihy-
drochloride (8)
Following the procedure for 6, diamine 8 (2.58 g) was gen-
erated in 100% yield from (DL)-2,5-diaminopentanoic
acid hydrochloride (2.00 g, 11.86 mmol) and used directly
in the subsequent reactions (6.13, 6.14, and 6.15).
6.9. L-2,6-Diaminohexanoic acid methyl ester dihydro-
chloride (9)
Following the procedure for 6, diamine 9 (2.50 g) was
produced in 98% yield from (L)-2,5-diaminohexanoic
acid hydrochloride (2.00 g, 10.95 mmol) and used direct-
ly in the subsequent reactions (6.16, 6.17, and 6.18).

T. T. Charvat et al. / Bioorg. Med. Chem. 14 (2006) 4552–4567
4561
6.10. (DL)-2,3-Bis(3,4-dimethoxycarbonyloxybenzoylami-
no)propionic acid (13)
nyl)acrylic acid (600 mg, 2.03 mmol) were converted to
diamide 18 (439 mg, 79% yield), which was used directly
in the subsequent reactions.
A 100 mL three-necked flask was charged with dia-
triethylamine
mine
6
(142 mg,
0.741 mmol),
6.16. (DL)-2,5-Bis(3,4-dimethoxycarbonyloxybenzoylami-
no)pentanoic acid (19)
(0.204 mL, 1.48 mmol), and methylene chloride
(3.7 mL). The resultant homogeneous solution was
cooled to 0 ꢁC, and 3,4-dimethoxycarbonyloxybenzoic
acid (500 mg, 1.85 mmol) was added. The reaction
mixture was stirred for 10 min, followed by the addi-
tion of EDCI (377 mg, 1.93 mmol). After stirring for
10 min at 0 ꢁC, the reaction mixture was warmed to
rt and then heated at reflux for 48 h. The resultant
solution was concentrated in vacuo, partitioned be-
tween EtOAc and 10% HCl, and extracted with
EtOAc (4· 40 mL). The combined organics were
washed with saturated NaHCO3 (2· 40 mL), 10%
HCl (1· 40 mL), and brine (1· 40 mL); followed by
drying with MgSO4 and concentration in vacuo to
generate the crude diamide (409 mg, 66% yield), which
was used directly in the subsequent reactions.
Following the procedure for 13, diamine 8 (195 mg,
0.889 mmol) and 3,4-dimethoxycarbonyloxybenzoic
acid (600 mg, 2.22 mmol) were converted to diamide
19 (480 mg, 83% yield), which was used directly in the
subsequent reactions.
6.17. (DL)-2,5-Bis(3,4,5-trimethoxycarbonyloxybenzoyla-
mino)pentanoic acid (20)
Following the procedure for 13, diamine 8 (153 mg,
0.699 mmol) and 3,4,5-trimethoxycarbonyloxybenzoic
acid (575 mg, 1.67 mmol) were converted to diamide
20 (156 mg, 28% yield), which was used directly in the
subsequent reactions.
6.11. (DL)-2,3-Bis(3,4,5-trimethoxycarbonyloxybenzoyla-
mino)propionic acid (14)
6.18. (DL)-2,5-Bis[3-(3,4-dimethoxycarbonyloxyphe-
nyl)acryloylamino]pentanoic acid (21)
Following the procedure for 13, diamine 6 (151 mg,
0.793 mmol) and 3,4,5-trimethoxycarbonyloxybenzoic
acid (600 mg, 1.74 mmol) were converted to diamide
14 (252 mg, 41% yield), which was used directly in the
subsequent reactions.
Following the procedure for 13, diamine 8 (178 mg,
0.811 mmol) and 3-(3,4-dimethoxycarbonyloxyphe-
nyl)acrylic acid (600 mg, 2.03 mmol) were converted to
diamide 21 (383 mg, 67% yield), which was used directly
in the subsequent reactions.
6.12. (DL)-2,3-Bis[3-(3,4-dimethoxycarbonyloxyphe-
nyl)acryloylamino]propionic acid (15)
6.19. L-2,6-Bis(3,4-dimethoxycarbonyloxybenzoylami-
no)hexanoic acid (22)
Following the procedure for 13, diamine 6 (129 mg,
0.676 mmol) and 3-(3,4-dimethoxycarbonyloxyphe-
nyl)acrylic acid (500 mg, 1.69 mmol) were converted to
diamide 15 (416 mg, 38% yield), which was used directly
in the subsequent reactions.
Following the procedure for 13, diamine 9 (173 mg,
0.742 mmol) and 3,4-dimethoxycarbonyloxybenzoic
acid (500 mg, 1.85 mmol) were converted to diamide
22 (370 mg, 75% yield), which was used directly in the
subsequent reactions.
6.13. (DL)-2,4-Bis(3,4-dimethoxycarbonyloxybenzoylami-
no)butyric acid (16)
6.20. L-2,6-Bis(3,4,5-trimethoxycarbonyloxybenzoylami-
no)hexanoic acid (23)
Following the procedure for 13, diamine 7 (177 mg,
0.863 mmol) and 3,4-dimethoxycarbonyloxybenzoic
acid (583 mg, 2.16 mmol) were converted to diamide
16 (313 mg, 49% yield), which was used directly in the
subsequent reactions.
Following the procedure for 13, diamine 9 (129 mg,
0.552 mmol) and 3,4,5-trimethoxycarbonyloxybenzoic
acid (475 mg, 1.38 mmol) were converted to diamide
23 (250 mg, 56% yield), which was used directly in the
subsequent reactions.
6.14. (DL)-2,4-Bis(3,4,5-trimethoxycarbonyloxybenzoyla-
mino)butyric acid (17)
6.21. L-2,6-Bis[3-(3,4-dimethoxycarbonyloxyphe-
nyl)acryloylamino]hexanoic acid (24)
Following the procedure for 13, diamine 7 (144 mg,
0.699 mmol) and 3,4,5-trimethoxycarbonyloxybenzoic
acid (575 mg, 1.67 mmol) were converted to diamide
17 (308 mg, 56% yield), which was used directly in the
subsequent reactions.
Following the procedure for 13, diamine 9 (157 mg,
0.676 mmol) and 3-(3,4-dimethoxycarbonyloxyphe-
nyl)acrylic acid (500 mg, 1.69 mmol) were converted to
diamide 24 (419 mg, 87% yield), which was used directly
in the subsequent reactions.
6.15. (DL)-2,4-Bis[3-(3,4-dimethoxycarbonyloxyphe-
nyl)acryloylamino]butyric acid (18)
6.22. (DL)-2,3-Bis(3,4-dihydroxybenzoylamino)propionic
acid (25)
Following the procedure for 13, diamine 7 (166 mg,
0.811 mmol) and 3-(3,4-dimethoxycarbonyloxyphe-
To crude diamide 13 (75 mg, 0.121 mmol) in 0.85 mL
THF and 0.43 mL MeOH was added LiOH (0.84 mL,

4562
T. T. Charvat et al. / Bioorg. Med. Chem. 14 (2006) 4552–4567
1.0 M) dropwise. The resultant solution was shaken for
2.5 h, quenched with 2.0 mL of 10% HCl, lyophilized,
and purified via preparatory HPLC (98–45%, eluent
A) to afford 18 mg (40%) of acid 25: 1H NMR
(500 MHz, CD3OD) d 7.31 (d, J = 2.2, 1H), 7.27 (d,
J = 2.2, 1H), 7.25 (dd, J = 8.3, 2.2, 1H), 7.19 (dd,
J = 8.3, 2.2, 1H), 6.81 (d, J = 8.3, 1H), 6.79 (d, J = 8.3,
1H), 4.73 (dd, J = 6.6, 5.3, 1H), 3.85 (app d, J = 5.2,
2H); HRMS (CI/methanol) m/z calcd for C17H16N2O8-
Na (M+Na)+ 399.0804, found 399.0812; Analytical
HPLC tR = 3.65 min.
6.27. (DL)-2,4-Bis[3-(3,4-dihydroxyphenyl)acryloylami-
no]butyric acid (30)
Following the procedure for 25, crude diamide 18
(75 mg, 0.109 mmol) was hydrolyzed to afford 20 mg
1
(42%) of acid 30: H NMR (500 MHz, CD3OD) d 7.44
(d, J = 15.7, 1H), 7.40 (d, J = 15.7, 1H), 7.03 (d,
J = 2.0, 1H), 7.01 (d, J = 2.0, 1H), 6.93 (dd, J = 8.2,
2.0, 1H), 6.90 (dd, J = 8.2, 2.0, 1H), 6.77 (d, J = 8.2,
1H), 6.76 (d, J = 8.2, 1H), 6.48 (d, J = 15.7, 1H), 6.36
(d, J = 15.7, 1H), 4.59 (dd, J = 9.1, 4.7, 1H), 3.46–3.55
(m, 1H), 3.27–3.37 (m, 1H), 2.14–2.24 (m, 1H), 1.93–
2.01 (m, 1H); HRMS (CI/methanol) m/z calcd for
C22H22N2O8Na (M+Na)+465.1274, found 465.1275;
Analytical HPLC tR = 4.38 min.
6.23. (DL)-2,3-Bis(3,4,5-trihydroxybenzoylamino)propi-
onic acid (26)
Following the procedure for 25, crude diamide 14
(40 mg, 0.0519 mmol) was hydrolyzed to afford 4 mg
(10%) of acid 26: H NMR (500 MHz, CD3OD) d 6.89
6.28. (DL)-2,5-Bis(3,4-dihydroxybenzoylamino)pentanoic
acid (31)
1
(s, 2H), 6.84 (s, 2H), 4.70 (dd, J = 6.7, 5.3, 1H), 3.81–
3.85 (m, 2H); HRMS (CI/methanol) m/z calcd for
C17H15N2O10Na2 (M + 2Na–H)+ 453.0522, found
453.0526; Analytical HPLC tR = 3.23 min.
Following the procedure for 25, crude diamide 19
(75 mg, 0.115 mmol) was hydrolyzed to afford 24 mg
(52%) of acid 31: H NMR (500 MHz, CD3OD) d 7.31
1
(d, J = 2.1, 1H), 7.27 (d, J = 2.2, 1H), 7.27 (dd,
J = 8.3, 2.2, 1H), 7.19 (dd, J = 8.3, 2.2, 1H), 6.81 (d,
J = 8.3, 1H), 6.79 (d, J = 8.3, 1H), 4.58 (dd, J = 9.2,
4.9, 1H), 3.39 (t, J = 7.0, 2H), 2.0–2.1 (m, 1H), 1.84–
1.94 (m, 1H), 1.69–1.83 (m, 2H); HRMS (CI/methanol)
m/z calcd for C19H20N2O8Na (M+Na)+ 427.1117, found
427.1120; Analytical HPLC tR = 3.70 min.
6.24. (DL)-2,3-Bis[3-(3,4-dihydroxyphenyl)acryloylami-
no]propionic acid (27)
Following the procedure for 25, crude diamide 15
(75 mg, 0.111 mmol) was hydrolyzed to afford 9 mg
1
(20%) of acid 27: H NMR (500 MHz, CD3OD) d 7.42
(d, J = 15.7, 1H), 7.41 (d, J = 15.7, 1H), 7.02 (d,
J = 2.1, 1H), 7.00 (d, J = 2.1, 1H), 6.88–6.94 (m, 2H),
6.77 (d, J = 8.1, 1H), 6.76 (d, J = 8.1, 1H), 6.45 (d,
J = 15.7, 1H), 6.37 (d, J = 15.7, 1H), 4.70 (dd, J = 7.5,
4.8, 1H), 3.84 (dd, J = 13.8, 4.7, 1H), 3.68 (dd,
J = 13.8, 7.5, 1H); HRMS (CI/methanol) m/z calcd for
C21H20N2O8Na (M+Na)+ 451.1117, found 451.1115;
Analytical HPLC tR = 4.29 min.
6.29. (DL)-2,5-Bis(3,4,5-trihydroxybenzoylamino)penta-
noic acid (32)
Following the procedure for 25, crude diamide 20
(75 mg, 0.0940 mmol) was hydrolyzed to afford 10 mg
1
(24%) of acid 32: H NMR (500 MHz, CD3OD) d 6.89
(s, 2H), 6.84 (s, 2H), 4.56 (dd, J = 9.1, 5.0, 1H), 3.38
(t, J = 6.8, 2H), 1.99–2.08 (m, 1H), 1.81–1.91 (m, 1H),
1.64–1.78 (m, 2H); HRMS (CI/methanol) m/z calcd for
C19H20N2O10Na (M+Na)+ 459.1016, found 459.1001;
Analytical HPLC tR = 3.22 min.
6.25. (DL)-2,4-Bis(3,4-dihydroxybenzoylamino)butyric
acid (28)
Following the procedure for 25, crude diamide 16
(70 mg, 0.110 mmol) was hydrolzyed to afford 9 mg
(21%) of acid 28: 1H NMR (500 MHz, CD3OD) d
7.36 (d, J = 2.2, 1H), 7.32 (dd, J = 8.3, 2.2, 1H), 7.29
(d, J = 2.2, 1H), 7.21 (dd, J = 8.3, 2.2, 1H), 6.83 (d,
J = 8.3, 1H), 6.80 (d, J = 8.3, 1H), 4.68 (dd, J = 8.9,
4.9,1H), 3.60–3.68 (m, 1H), 3.34–3.40 (m, 1H), 2.21–
2.30 (m, 1H), 2.06–2.16 (m, 1H); HRMS (CI/methanol)
m/z calcd for C18H18N2O8Na (M+Na)+ 413.0961,
found 413.0974; Analytical HPLC tR = 3.64 min.
6.30. (DL)-2,5-Bis[3-(3,4-dihydroxyphenyl)acryloylami-
no]pentanoic acid (33)
Following the procedure for 25, crude diamide 21
(75 mg, 0.107 mmol) was hydrolyzed to afford 24 mg
1
(49%) of acid 33: H NMR (500 MHz, CD3OD) d 7.42
(d, J = 15.9, 1H), 7.38 (d, J = 16.0, 1H), 7.02 (d,
J = 2.0, 1H), 7.00 (d, J = 2.0, 1H), 6.91 (dd, J = 8.4,
2.0, 1H), 6.89 (dd, J = 8.6, 2.0, 1H), 6.77 (d, J = 8.2,
1H), 6.76 (d, J = 8.2, 1H), 6.47 (d, J = 15.7, 1H), 6.35
(d, J = 15.7, 1H), 4.54 (dd, J = 8.8, 4.8, 1H), 3.34 (t,
J = 6.3, 2H), 1.94–2.03 (m, 1H), 1.75–1.84 (m, 1H),
1.65–1.74 (m, 2H); HRMS (CI/methanol) m/z calcd for
C23H24N2O8Na (M+Na)+ 479.1530, found 479.1421;
Analytical HPLC tR = 4.39 min.
6.26. (DL)-2,4-Bis(3,4,5-trihydroxybenzoylamino)butyric
acid (29)
Following the procedure for 25, crude diamide 17
(75 mg, 0.0957 mmol) was hydrolyzed to afford 5 mg
1
(12%) of acid 29: H NMR (500 MHz, CD3OD) d 6.95
(s, 2H), 6.87 (s, 2H), 4.67 (dd, J = 8.9, 5.0, 1H), 3.60–
3.70 (m, 2H), 2.20–2.28 (m, 1H), 2.05–2.15 (m, 1H);
HRMS (CI/methanol) m/z calcd for C18H18N2O10Na
(M+Na)+ 445.0859, found 445.0852; Analytical HPLC
tR = 3.19 min.
6.31. L-2,6-Bis(3,4-dihydroxybenzoylamino)hexanoic acid
(34)
Following the procedure for 25, crude diamide 22
(75 mg, 0.13 mmol) was hydrolyzed to afford 30 mg

T. T. Charvat et al. / Bioorg. Med. Chem. 14 (2006) 4552–4567
4563
1
(64%) of acid 34: H NMR (500 MHz, CD3OD) d 7.30
4.80 (dd, J = 12.3, 5.6, 1H), 3.82–3.86 (m, 14H),
3.81 (s, 6H), 3.76 (s, 3H); 13C NMR (125 MHz,
CDCl3) d 171.0, 167.5, 166.6, 153.6, 142.3, 142.1,
140.0, 130.2, 130.2, 119.3, 119.3, 105.3, 105.3, 61.1,
56.3, 56.3, 54.1, 53.1, 42.2; HRMS (CI/methanol) m/
z calcd for C28H34N2O10Na (M+Na)+ 581.2111, found
581.2101.
(d, J = 2.2, 1H), 7.26 (d, J = 2.1, 1H), 7.25 (dd,
J = 8.3, 2.2, 1H), 7.16 (dd, J = 8.3, 2.2, 1H), 6.79 (d,
J = 8.3, 1H), 6.77 (d, J = 8.3, 1H), 4.55 (dd, J = 9.4,
4.9, 1H), 3.35 (app dt, J = 7.2, 2.1, 2H), 1.96–2.05 (m,
1H), 1.85–1.93 (m, 1H), 1.61–1.72 (m, 2H), 1.46–1.59
(m, 2H); HRMS (CI/methanol) m/z calcd for
C20H22N2O8Na (M+Na)+ 441.1274, found 441.1261;
Analytical HPLC tR = 3.87 min.
6.35. (DL)-2,3-Bis[3-(3,4,5-trihydroxyphenyl)acryloylami-
no]propionic acid (42)
6.32. L-2,6-Bis(3,4,5-trihydroxybenzoylamino)hexanoic
acid (35)
To a solution of permethylated ester 38 (50 mg,
0.090 mmol) in 0.55 mL methylene chloride at 0 ꢁC
was added boron tribromide (1.3 mL, 1.0 M) dropwise.
The resultant solution was stirred for 5 h 30 min at
0 ꢁC and then quenched with 1.0 mL of 4 M NaOH
and 0.5 mL MeOH. The methylene chloride was subse-
quently removed in vacuo, followed by preparatory
HPLC (95–60% eluent A) purification to provide
Following the procedure for 25, crude diamide 23
(75 mg, 0.0924 mmol) was hydrolyzed to afford 32 mg
1
(65%) of acid 35: H NMR (500 MHz, CD3OD) d 6.89
(s, 2H), 6.83 (s, 2H), 4.54 (dd, J=9.0, 5.0, 1H), 3.30–
3.36 (m, 2H), 1.95–2.04 (m, 1H), 1.82–1.90 (m, 1H),
1.60–1.70 (m, 2H), 1.46–1.57 (m, 2H); HRMS (CI/meth-
anol) m/z calcd for C20H22N2O10Na (M+Na)+ 473.1172,
found 473.1165; Analytical HPLC tR = 3.42 min.
1
19 mg (46%) acid 42: H NMR (500 MHz, CD3OD) d
7.35 (d, J = 15.7, 1H), 7.32 (d, J = 15.7, 1H), 6.59 (s,
2H), 6.57 (s, 2H), 6.41 (d, J = 15.7, 1H), 6.34 (d,
J = 15.7, 1H), 4.69 (dd, J = 7.5, 4.7, 1H), 3.84 (dd,
J = 13.9, 4.8, 1H), 3.67 (dd, J = 13.8, 7.5, 1H); HRMS
(CI/methanol) m/z calcd for C21H20N2O10Na (M+Na)+
6.33. L-2,6-Bis[3-(3,4-dihydroxyphenyl)acryloylami-
no]hexanoic acid (36)
Following the procedure for 25, crude diamide 24
(75 mg, 0.105 mmol) was hydrolyzed to afford 11 mg
(26%) of acid 36: 1H NMR (500 MHz, CD3OD) d
7.41 (d, J = 15.7, 1H), 7.38 (d, J = 15.7, 1H), 7.02
(d, J = 2.0, 1H), 6.99 (d, J = 2.0, 1H), 6.90 (dd,
J = 8.3, 2.2, 1H), 6.87 (dd, J = 8.2, 2.0, 1H), 6.76 (d,
J = 8.2, 1H), 6.75 (d, J = 8.2, 1H), 6.48 (d, J = 15.7,
1H), 6.35 (d, J = 15.7, 1H), 4.51 (dd, J = 9.0, 4.8,
1H), 3.30–3.35 (m, 2H), 1.91–1.99 (m, 1H), 1.77–1.84
(m, 1H), 1.57–1.67 (m, 2H) 1.46–1.55 (m, 2H); HRMS
(CI/methanol) m/z calcd for C24H26N2O8Na (M+Na)+
483.1016,
tR = 3.75 min.
found
483.1000;
Analytical
HPLC
6.36. (DL)-2,4-Bis[3-(3,4,5-trihydroxyphenyl)acryloylami-
no]butyric acid (43)
A 100 mL three-necked flask was charged with diamine
7
(206 mg, 1.00 mmol), triethylamine (0.280 mL,
2.01 mmol), and methylene chloride (5 mL). The resul-
tant homogeneous solution was cooled to 0 ꢁC and
3,4,5-trimethoxycinnamic acid (598 mg, 2.51 mmol)
was added. The reaction mixture was stirred for
10 min, followed by the addition of EDCI (512 mg,
2.61 mmol). After stirring for 10 min at 0 ꢁC, the reac-
tion mixture was warmed to rt and then heated at reflux
for 48 h. The resultant solution was concentrated in vac-
uo, partitioned between EtOAc and 10% HCl, and
extracted with EtOAc (4· 40 mL). The combined organ-
ics were washed with saturated NaHCO3 (2· 40 mL),
10% HCl (1· 40 mL), and brine (1· 40 mL), followed
by drying with MgSO4 and concentration in vacuo to
generate the crude diamide (478 mg) in 83% yield.
493.1587, found
tR = 4.55 min.
493.1576;
Analytical
HPLC
6.34. (DL)-2,3-Bis[3-(3,4,5-trimethoxyphenyl)acryloyl-
amino]propionic acid methyl ester (38)
A 100 mL three-necked flask was charged with dia-
mine 6 (532 mg, 2.79 mmol), triethylamine (0.777 mL,
5.57 mmol), and methylene chloride (30 mL). The
resultant homogeneous solution was cooled to 0 ꢁC
and 3,4,5-trimethoxycinnamic acid (1.46 g, 6.13 mmol)
was added. The reaction mixture was stirred for
10 min, followed by the addition of EDCI (1.22 g,
6.41 mmol). After stirring for 10 min at 0 ꢁC, the reac-
tion mixture was warmed to rt and then heated at re-
flux for 48 h. The resultant solution was concentrated
in vacuo, partitioned between EtOAc and 10% HCl,
and extracted with EtOAc (4· 40 mL). The combined
organics were washed with saturated NaHCO3 (2·
40 mL), 10% HCl (1· 40 mL), and brine (1· 40 mL);
followed by drying with MgSO4 and concentration
in vacuo. The resultant diamide was subsequently
recrystallized from EtOAc/hexanes to produce 1.18g
of diamide 38 (76%) as a white solid: 1H NMR
(500 MHz, CDCl3) d 7.52 (d, J = 6.4, 1H), 7.49 (d,
J = 6.4, 1H), 7.21 (d, J = 6.9, 1H), 6.64–6.69 (m,
5H), 6.39 (d, J = 15.6, 1H), 6.31 (d, J = 15.6, 1H),
To a solution of the crude diamide (50 mg, 0.087 mmol) in
0.53 mL methylene chloride at 0 ꢁC was added boron
tribromide (1.2 mL, 1.0 M) dropwise. The resultant solu-
tion was stirred for 5 h 30 min at 0 ꢁC and then quenched
with 1.0 mL of 4 M NaOH and 0.5 mL MeOH. The meth-
ylene chloride was subsequently removed in vacuo, fol-
lowed by preparatory HPLC (95–60% eluent A)
1
purification to provide 10 mg (24%) acid 43: H NMR
(500 MHz, CD3OD) d 7.37 (d, J = 15.7, 1H), 7.32 (d,
J = 15.7, 1H), 6.61 (s, 2H), 6.58 (s, 2H), 6.45 (d,
J = 15.6, 1H), 6.33 (d, J = 15.6, 1H), 4.58 (dd, J = 9.2,
4.8, 1H), 3.47–3.56 (m, 1H), 3.26–3.38 (m, 1H), 2.15–
2.25 (m, 1H), 1.93–2.03 (m, 1H); HRMS (CI/methanol)
m/z calcd for C22H22N2O10Na (M+Na)+ 497.1172, found
497.1162; Analytical HPLC tR = 3.75 min.

4564
T. T. Charvat et al. / Bioorg. Med. Chem. 14 (2006) 4552–4567
6.37. (DL)-2,5-Bis[3-(3,4,5-trihydroxyphenyl)acryloylami-
no]pentanoic acid (44)
60% eluent A) purification, to provide 20 mg (48%) acid
45: H NMR (500 MHz, CD3OD) d 7.34 (d, J = 15.6,
1
1H), 7.30 (d, J = 15.6, 1H), 6.59 (s, 2H), 6.57 (s, 2H),
6.44 (d, J = 15.6, 1H), 6.32 (d, J = 15.6, 1H), 4.50 (dd,
J = 8.9, 4.9, 1H), 3.28–3.32 (m, 2H), 1.90–1.99 (m,
1H), 1.74–1.83 (m, 1H), 1.57–1.66 (m, 2H), 1.44–1.55
(m, 2H); HRMS (CI/methanol) m/z calcd for
C24H26N2O10Na (M+Na)+ 525.1485, found 525.1495;
Analytical HPLC tR = 4.00 min.
A 100 mL three-necked flask was charged with diamine
8
(220 mg, 1.00 mmol), triethylamine (0.280 mL,
2.01 mmol), and methylene chloride (5 mL). The resul-
tant homogeneous solution was cooled to 0 ꢁC and
3,4,5-trimethoxycinnamic acid (598 mg, 2.51 mmol)
was added. The reaction mixture was stirred for
10 min, followed by the addition of EDCI (512 mg,
2.61 mmol). After stirring for 10 min at 0 ꢁC, the reac-
tion mixture was warmed to rt and then heated at reflux
for 48 h. The resultant solution was concentrated in vac-
uo, partitioned between EtOAc and 10% HCl, and
extracted with EtOAc (4· 40 mL). The combined organ-
ics were washed with saturated NaHCO3 (2· 40 mL),
10% HCl (1· 40 mL), and brine (1· 40 mL), followed
by drying with MgSO4 and concentration in vacuo to
generate the crude diamide (197 mg) in 33% yield.
6.39. (DL)-2,3-Bis(3,4,5-trimethoxybenzoylamino)propi-
onic acid methyl ester (48)
A 100 mL three-necked flask was charged with diamine
(532 mg, 2.79 mmol), triethylamine (0.777 mL,
6
5.57 mmol), and DMF (30 mL). The resultant homoge-
neous solution was cooled to 0 ꢁC and acid 46 (1.30 g,
6.13 mmol) was added. The reaction mixture was stirred
for 10 min, followed by the addition of EDCI (1.22 g,
6.41 mmol). After stirring for 10 min at 0 ꢁC, the reac-
tion mixture was warmed to rt and then placed in a
60 ꢁC oil bath and stirred for 48 h. The resultant solu-
tion was concentrated in vacuo, partitioned between
EtOAc and 10% HCl, and extracted with EtOAc (4·
40 mL). The combined organics were washed with satu-
rated NaHCO3 (2· 40 mL), 10% HCl (1· 40 mL), and
brine (1· 40 mL), followed by drying with MgSO4 and
concentration in vacuo. The resultant diamide was sub-
sequently recrystallized from EtOAc/hexanes to produce
1.03 g ester 48 (73%) as a white solid: 1H NMR
(500 MHz, CDCl3) d 7.87 (d, J = 6.5, 1H), 7.26 (t,
J = 5.8, 1H), 7.09 (s, 2H), 7.02 (s, 2H), 4.76–4.82 (m,
1H), 3.86–3.93 (m, 2H), 3.87 (s, 6H), 3.85, (s, 3H),
3.84 (s, 3H), 3.83, (s, 6H), 3.75, (s, 3H); 13C NMR
(125 MHz, CDCl3) d 170.8, 169.1, 167.8, 153.4, 153.4,
141.5, 141.5, 129.0, 128.6, 104.8, 104.8, 61.1, 61.1,
56.4, 56.4, 54.9, 53.1, 42.7; HRMS (CI/MeOH) m/z
calcd for C24H30N2O10Na (M+Na)+ 507.1979, found
507.1970.
To a solution of the crude diamide (20 mg, 0.034 mmol)
in 0.20 mL methylene chloride at 0 ꢁC was added boron
tribromide (0.48 mL, 1.0 M) dropwise. The resultant
solution was stirred for 5 h 30 min at 0 ꢁC and then
quenched with 0.5 mL of 4 M NaOH and 0.3 mL
MeOH. The methylene chloride was subsequently re-
moved in vacuo, followed by preparatory HPLC (95–
60% eluent A) purification to provide 9 mg (54%) acid
1
44: H NMR (500 MHz, CD3OD) d 7.34 (d, J = 15.9,
1H), 7.31 (d, J = 15.9, 1H), 6.59 (s, 2H), 6.57 (s, 2H),
6.44 (d, J = 15.6, 1H), 6.32 (d, J = 15.6, 1H), 4.52 (dd,
J = 8.8, 4.9, 1H), 3.30–3.35 (m, 2H), 1.93–2.01 (m, 1H),
1.76–1.84 (m, 1H), 1.64–1.75 (m, 2H); HRMS (CI/meth-
anol) m/z calcd for C23H24N2O10Na (M+Na)+ 511.1329,
found 511.1330; Analytical HPLC tR = 3.86 min.
6.38. L-2,6-Bis[3-(3,4,5-trihydroxyphenyl)acryloylami-
no]hexanoic acid (45)
A 100 mL three-necked flask was charged with diamine
9
(196 mg, 0.840 mmol), triethylamine (0.234 mL,
6.40. (DL)-2,3-Bis[3-(3,4-dimethoxyphenyl)acryloylami-
no]propionic acid methyl ester (49)
1.68 mmol), and methylene chloride (4.2 mL). The resul-
tant homogeneous solution was cooled to 0 ꢁC and
3,4,5-trimethoxycinnamic acid (500 mg, 2.10 mmol)
was added. The reaction mixture was stirred for
10 min, followed by the addition of EDCI (428 mg,
2.19 mmol). After stirring for 10 min at 0 ꢁC, the reac-
tion mixture was warmed to rt and then heated at reflux
for 48 h. The resultant solution was concentrated in vac-
uo, partitioned between EtOAc and 10% HCl, and
extracted with EtOAc (4· 40 mL). The combined organ-
ics were washed with saturated NaHCO3 (2· 40 mL),
10% HCl (1· 40 mL), and brine (1· 40 mL) followed
by drying with MgSO4 and concentration in vacuo to
generate the crude diamide (448 mg) in 89% yield.
Following the procedure for 38, diamide 49 (990 mg)
was generated in 71% yield from acid 47 (1.28 g,
6.13 mmol), and diamine 6 (532 mg, 2.79 mmol), which
was used directly in the subsequent reactions.
6.41. (DL)-2,3-Bis(3,4,5-trimethoxybenzoylamino)propi-
onic acid (50)
Ester 48 (500 mg, 0.988 mmol) was dissolved in methy-
lene chloride, followed by concentration in vacuo until
almost all the solvent was removed (the solid ester is
insoluble in methanol). The resultant solution was dis-
solved in methanol (5.0 mL) and water (2.0 mL), fol-
lowed by the addition of LiOH (2.96 mL, 1.0 M).
After stirring overnight (8 h), the methanol was removed
in vacuo, followed by acidification with 10% HCl and
placement at 4 ꢁC for 5 d. The resultant white solid
was filtered and then washed with cold water to afford
480 mg (99%) of acid 50, which was used directly in
the subsequent reactions.
To a solution of the crude diamide (50 mg, 0.083 mmol)
in 0.50 mL methylene chloride at 0 ꢁC was added boron
tribromide (1.2 mL, 1.0 M) dropwise. The resultant
solution was stirred for 5 h 30 min at 0 ꢁC and then
quenched with 1.0 mL of 4 M NaOH and 0.5 mL
MeOH. The methylene chloride was subsequently re-
moved in vacuo, followed by preparatory HPLC (95–

T. T. Charvat et al. / Bioorg. Med. Chem. 14 (2006) 4552–4567
4565
6.42. (DL)-2,3-Bis[3-(3,4-dimethoxyphenyl)acryloylami-
no]propionic acid (51)
1H), 6.76 (d, J = 8.2, 1H), 6.45 (d, J = 15.7, 1H), 6.36
(d, J = 15.7, 1H), 4.69 (dd, J = 6.7, 5.3, 1H), 3.80 (dd,
J = 13.9, 5.3, 1H), 3.76 (s, 3H), 3.68 (dd, J = 13.8, 6.8,
1H); HRMS (CI/methanol) m/z calcd for C22H22N2O8-
Na (M+Na)+ 465.1274, found 465.1255; Analytical
HPLC tR = 4.69 min.
Ester 49 (500 mg, 1.00 mmol) was dissolved in meth-
ylene chloride, followed by concentration in vacuo
until almost all the solvent was removed (the solid
ester is insoluble in methanol). The resultant solution
was dissolved in methanol (2.0 mL), THF (6.0 mL),
and water (2.0 mL), followed by the addition of
LiOH (3.01 mL, 1.0 M). After stirring overnight
(8 h), the methanol was removed in vacuo, followed
by acidification with 10% HCl and placement at
4 ꢁC for 5 d. The resultant white solid was filtered
and then washed with cold water to afford 482 mg
(99%) of acid 51, which was used directly in the sub-
sequent reactions.
6.47. (DL)-2,4-Bis(3,4-dihydroxybenzoylamino)butyric
acid methyl ester (59)
Following the procedure for 57, crude diamide 16
(75 mg, 0.118 mmol) was hydrolyzed to afford 10 mg
(21%) of ester 59: 1H NMR (500 MHz, CD3OD) d
7.36 (d, J = 2.2, 1H), 7.31 (dd, J = 8.3, 2.2, 1H), 7.29
(d, J = 2.2, 1H), 7.21 (dd, J = 8.3, 2.2, 1H), 6.83 (d,
J = 8.3, 1H), 6.79 (d, J = 8.3, 1H), 4.71 (dd, J = 8.6,
4.9, 1H), 3.58–3.66 (m, 1H), 3.68 (s, 3H), 3.30–3.36
(m, 1H), 2.18–2.26 (m, 1H), 2.07–2.16 (m, 1H); HRMS
(CI/methanol) m/z calcd for C19H20N2O8Na (M+Na)+
427.1117, found 427.1102; Analytical HPLC tR = 4.04 min.
6.43. (DL)-2,3-Bis[3-(3,4,5-trimethoxyphenyl)acryloyla-
mino]propionic acid (52)
Following the procedure for acid 50, ester 38 (500 mg,
0.896 mmol) was converted into acid 52 (454 mg) in
93% yield, which was used directly in the subsequent
reactions.
6.48. (DL)-2,4-Bis(3,4,5-trihydroxybenzoylamino)butyric
acid methyl ester (60)
Following the procedure for 56, crude diamide 17
(75 mg, 0.0957 mmol) was hydrolyzed to afford 29 mg
(70%) of ester 60: 1H NMR (500 MHz, CD3OD) d
6.95 (s, 2H), 6.86 (s, 2H), 4.69–472 (m, 1H), 3.65–3.72
(m, 2H), 2.16–2.24 (m, 1H), 2.06–2.14 (m, 1H); HRMS
(CI/methanol) m/z calcd for C23H24N2O8Na (M+Na)+
6.44. (DL)-2,3-Bis(3,4-dihydroxybenzoylamino)propionic
acid methyl ester (56)
To crude diamide 13 (75 mg, 0.121 mmol) was added
1.0 M NaOMe (0.51 mL). The resultant solution was
shaken for 30 min, neutralized with 10% HCl, and puri-
fied via preparatory HPLC (95–45%, eluent A) to afford
479.1430,
tR = 3.57 min.
found
479.1409;
Analytical
HPLC
1
41 mg (87%) of ester 56: H NMR (500 MHz, CD3OD)
d 7.30 (d, J = 2.2, 1H), 7.27 (d, J = 2.2, 1H), 7.25 (dd,
J = 8.3, 2.2, 1H), 7.19 (dd, J = 8.3, 2.2, 1H), 6.82 (d,
J = 8.3, 1H), 6.79 (d, J = 8.3, 1H), 4.72 (dd, J = 6.9,
4.9, 1H), 3.79–3.88 (m, 2H), 3.76 (s, 3H); HRMS (CI/
methanol) m/z calcd for C18H18N2O8Na (M+Na)+
6.49. (DL)-2,4-Bis[3-(3,4-dihydroxyphenyl)acryloylami-
no]butyric acid methyl ester (61)
Following the procedure for 56, crude diamide 18
(75 mg, 0.109 mmol) was hydrolyzed to afford 14 mg
(28%) of ester 61: 1H NMR (500 MHz, CD3OD) d
7.42 (d, J = 15.7, 1H), 7.40 (d, J = 15.7, 1H), 7.03
(d, J = 2.0, 1H), 7.00 (d, J = 2.0, 1H), 6.93 (dd,
J = 8.2, 2.0, 1H), 6.91 (dd, J = 8.3, 2.0, 1H), 6.77 (d,
J = 8.2, 1H), 6.76 (d, J = 8.1, 1H), 6.46 (d, J = 15.7,
1H), 6.36 (d, J = 15.7, 1H), 4.61 (dd, J = 9.0, 4.8,
1H), 3.74 (s, 3H), 3.70–3.77 (m, 1H), 3.43–3.51 (m,
1H), 2.10–2.20 (m, 1H), 1.93–2.01 (m, 1H); HRMS
413.0961, found
tR = 3.97 min.
413.0961;
Analytical
HPLC
6.45. (DL)-2,3-Bis(3,4,5-trihydroxybenzoylamino)propi-
onic acid methyl ester (57)
To crude diamide 14 (40 mg, 0.0519 mmol) was added
0.50 M NaOMe (0.66 mL). The resultant solution was
shaken for 1 h, neutralized with 10% HCl, and purified
via preparatory HPLC (95–55%, eluent A) to afford
(CI/methanol)
m/z
calcd
for
C23H24N2O8Na
(M+Na)+ 479.1430, found 479.1409; Analytical HPLC
tR = 4.69 min.
1
6 mg (15%) of ester 57: H NMR (500 MHz, CD3OD)
d 6.89 (s, 2H), 6.83 (s, 2H), 4.68 (dd, J = 6.5, 5.0, 1H),
3.79–3.85, (m, 2H), 3.76 (s, 3H); HRMS (CI/methanol)
m/z calcd for C18H18N2O10Na (M+Na)+ 445.0859,
found 445.0841; Analytical HPLC tR = 3.57 min.
6.50. (DL)-2,5-Bis(3,4-dihydroxybenzoylamino)pentanoic
acid methyl ester (62)
Following the procedure for 57, crude diamide 19
(75 mg, 0.115 mmol) was hydrolyzed to afford 32 mg
(66%) of ester 62: H NMR (500 MHz, CD3OD) d 7.31
6.46. (DL)-2,3-Bis[3-(3,4-dihydroxyphenyl)acryloylami-
no]propionic acid methyl ester (58)
1
(d, J = 2.1, 1H), 7.27 (d, J = 2.1, 1H), 7.27 (dd, J = 8.2,
2.2, 1H), 7.20 (dd, J = 8.3, 2.2, 1H), 6.81 (d, J = 8.4,
1H), 6.79 (d, J = 8.5, 1H), 4.59 (dd, J = 9.31, 5.1, 1H),
3.73 (s, 3H), 3.38 (t, J = 7.0, 2H), 1.96–2.0 (m, 1H),
1.83–2.01 (m, 2H), 1.65–1.80 (m, 1H); HRMS (CI/meth-
anol) m/z calcd for C20H22N2O8Na (M+Na)+ 441.1274,
found 441.1265; Analytical HPLC tR = 4.16 min.
Following the procedure for 56, crude diamide 14
(75 mg, 0.111 mmol) was hydrolyzed to afford 26 mg
(53%) of ester 58: 1H NMR (500 MHz, CD3OD) d
7.41 (d, J = 15.7, 1H), 7.40 (d, J = 15.7, 1H), 7.02 (d,
J = 2.0, 1H), 7.00 (d, J = 2.0, 1H), 6.93 (dd, J = 8.5,
2.0, 1H), 6.91 (dd, J = 8.4, 2.0, 1H), 6.77 (d, J = 8.2,

4566
T. T. Charvat et al. / Bioorg. Med. Chem. 14 (2006) 4552–4567
6.51. (DL)-2,5-Bis(3,4,5-trihydroxybenzoylamino)penta-
noic acid methyl ester (63)
J = 2.0, 1H), 6.90 (dd, J = 8.2, 2.0, 1H), 6.87 (dd, J = 8.2,
2.0, 1H), 6.76 (d, J = 8.2, 1H), 6.75 (d, J = 8.2, 1H), 6.45
(d, J = 15.6, 1H), 6.35 (d, J = 15.6, 1H), 4.51 (dd, J = 9.0,
5.1, 1H), 3.73 (s, 3H), 3.29–3.35 (m, 2H), 1.85–1.95 (m,
1H), 1.74–1.82 (m, 1H), 1.56–1.65 (m, 2H) 1.43–1.52
(m, 2H); HRMS (CI/methanol) m/z calcd for
C25H28N2O8Na (M+Na)+ 507.1743, found 507.1722;
Analytical HPLC tR = 5.15 min.
Following the procedure for 57, crude diamide 20
(75 mg, 0.0940 mmol) was hydrolyzed to afford 10 mg
(24%) of ester 63: 1H NMR (500 MHz, CD3OD) d
6.89 (s, 2H), 6.84 (s, 2H), 4.57 (dd, J = 9.2, 5.0, 1H),
3.73 (s, 3H), 3.37 (t, J = 7.0, 2H), 1.94–2.03 (m, 1H),
1.82–1.91 (m, 1H), 1.64–1.78 (m, 2H); HRMS (CI/meth-
anol) m/z calcd for C20H22N2O10Na (M+Na)+ 473.1172,
found 473.1174; Analytical HPLC tR = 3.62 min.
Acknowledgments
6.52. (DL)-2,5-Bis[3-(3,4-dihydroxyphenyl)acryloylami-
no]pentanoic acid methyl ester (64)
This work was supported in part by grants from the
Public Health Service 1R21-AI054305 (WER), 5T32-
GM08620 (DJL), and the Burroughs-Wellcome Fund
99-2609 (WER). We also thank the UC Irvine Chao
Family Comprehensive Cancer Center for partial sup-
port. WER is a Burroughs-Wellcome Fund Clinical
Scientist in Translational Research. The authors
thank Brenda McDougall for excellent technical
assistance.
Following the procedure for 57, crude diamide 21
(75 mg, 0.107 mmol) was hydrolyzed to afford 29 mg
(58%) of ester 64: 1H NMR (500 MHz, CD3OD) d
7.41 (d, J = 15.7, 1H), 7.39 (d, J = 15.6, 1H), 7.02 (d,
J = 2.0, 1H), 7.00 (d, J = 2.0, 1H), 6.91 (dd, J = 8.2,
2.0, 1H), 6.89 (dd, J = 8.3, 2.0, 1H), 6.77 (d, J = 8.2,
1H), 6.75 (d, J = 8.2, 1H), 6.45 (d, J = 15.7, 1H), 6.36
(d, J = 15.8, 1H), 4.54 (dd, J = 8.9, 5.1, 1H), 3.74 (s,
3H), 3.33 (t, J = 7.0, 2H), 1.90–1.97 (m, 1H), 1.74–1.83
(m, 1H), 1.60–1.72 (m, 2H); HRMS (CI/methanol) m/z
calcd for C24H26N2O8Na (M+Na)+ 493.1587, found
493.1583; Analytical HPLC tR = 4.92 min.
References and notes
1. Reinke, R. A.; King, P. J.; Victoria, J. G.; McDougall, B.
R.; Ma, G.; Mao, Y.; Reinecke, M. G.; Robinson, W. E.,
Jr. J. Med. Chem. 2002, 45, 3669–3683.
2. Artico, M.; Di Santo, R.; Costi, R.; Novellino, E.; Greco,
G.; Massa, S.; Tramontano, E.; Marongiu, M. E.; De
Montis, A.; La Colla, P. J. Med. Chem. 1998, 41, 3948–
3960.
6.53. L-2,6-Bis(3,4-dihydroxybenzoylamino)hexanoic acid
methyl ester (65)
Following the procedure for 57, crude diamide 22
(75 mg, 0.113 mmol) was hydrolyzed to afford 26 mg
(53%) of ester 65: 1H NMR (500 MHz, CD3OD) d
7.30 (d, J = 2.1, 1H), 7.26 (d, J = 2.2, 1H), 7.25 (dd,
J = 8.3, 2.2, 1H), 7.17 (dd, J = 8.3, 2.2, 1H), 6.79 (d,
J = 8.2, 1H), 6.77 (d, J = 8.2, 1H), 4.55 (dd, J = 9.4,
5.2, 1H), 3.72 (s, 3H), 3.34 (dt, J = 6.8, 1.9, 2H), 1.92–
2.01 (m, 1H), 1.83–1.91 (m, 1H), 1.59–1.71 (m, 2H),
1.44–1.56 (m, 2H); HRMS (CI/methanol) m/z calcd for
C21H25N2O8 (M+H)+ 433.1611, found 433.1613; Ana-
lytical HPLC tR = 4.32 min.
3. King, P. J.; Ma, G. X.; Miao, W. F.; Jia, Q.; McDougall,
B. R.; Reinecke, M. G.; Cornell, C.; Kuan, J.; Kim, T. R.;
Robinson, W. E. J. Med. Chem. 1999, 42, 497–509.
4. Lin, Z. W.; Neamati, N.; Zhao, H.; Kiryu, Y.; Turpin, J.
A.; Aberham, C.; Strebel, K.; Kohn, K.; Witvrouw, M.;
Pannecouque, C.; Debyser, Z.; De Clercq, E.; Rice, W. G.;
Pommier, Y.; Burke, T. R. J. Med. Chem. 1999, 42, 1401–
1414.
5. Kim, S. N.; Lee, J. Y.; Kim, H. J.; Shin, C.; Park, H.; Lee,
Y. S. Chem. Lett. 2000, 10, 1879–1882.
6. Hwang, D. J.; Kim, S. N.; Choi, J. H.; Lee, Y. S. Bioorg.
Med. Chem. 2001, 9, 1429–1437.
7. Pluymers, W.; Neamati, N.; Pannecouque, C.; Fikkert, V.;
Marchand, C.; Burke, T. R.; Pommier, Y.; Schols, D.; De
Clercq, E.; Debyser, Z.; Witvrouw, M. Mol. Pharmacol.
2001, 59, 403.
6.54. L-2,6-Bis(3,4,5-trihydroxybenzoylamino)hexanoic
acid methyl ester (66)
8. King, P. J.; Robinson, W. E., Jr. J. Virol. 1998, 8420–
8424.
Following the procedure for 57, crude diamide 23
(75 mg, 0.0924 mmol) was hydrolyzed to afford 7 mg
(16%) of ester 66: 1H NMR (500 MHz, CD3OD) d
6.88 (s, 2H), 6.83 (s, 2H), 4.53 (dd, J = 9.2, 5.3, 1H),
3.72 (s, 3H), 3.28–3.36 (m, 2H), 1.92–2.00 (m, 1H),
1.81–1.88 (m, 1H), 1.58–1.70 (m, 2H), 1.43–1.56 (m,
2H); HRMS (CI/methanol) m/z calcd for C21H24N2O10-
Na (M+Na)+ 487.1329, found 487.1328; Analytical
HPLC tR = 3.85 min.
9. King, P. J.; Lee, D. J.; Reinke, R. A.; Victoria, J. G.;
Beale, K.; Robinson, W. E., Jr. Virology 2003, 147–161.
10. Lee, D. J.; Robinson, W. E., Jr. J. Virol. 2004, 5835–5847.
11. Reinke, R. A.; Lee, D. J.; McDougall, B. R.; King, P. J.;
Victoria, J.; Mao, Y.; Lei, X.; Reinecke, M. G.; Robinson,
W. E., Jr. Virology 2004, 203–219.
12. Robinson, W. E.; Cordeiro, M.; AbdelMalek, S.; Jia, Q.;
Chow, S. A.; Reinecke, M. G.; Mitchell, W. M. Mol.
Pharmacol. 1996, 50, 846–855.
13. Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P.
J. Adv. Drug Delivery Rev. 1997, 23, 3–25.
14. Veber, D. F.; Johnson, S. R.; Cheng, H. Y.; Smith, B. R.;
Ward, K. W.; Kopple, K. D. J. Med. Chem. 2002, 45,
2615–2623.
15. Robinson, W. E.; Reinecke, M. G.; AbdelMalek, S.; Jia,
Q.; Chow, S. A. Proc. Natl. Acad. Sci. U.S.A. 1996, 93,
6326–6331.
6.55. L-2,6-Bis[3-(3,4-dihydroxyphenyl)acryloylami-
no]hexanoic acid methyl ester (67)
Following the procedure for 57, crude diamide 24 (72 mg,
0.101 mmol) was hydrolyzed to afford 27 mg (55%) of es-
ter 67: 1H NMR (500 MHz, CD3OD) d 7.41 (d, J = 15.4,
1H), 7.38 (d, J = 15.3, 1H), 7.01 (d, J = 2.0, 1H), 6.99 (d,

T. T. Charvat et al. / Bioorg. Med. Chem. 14 (2006) 4552–4567
4567
16. Chow, S. A. Methods 1997, 306–317.
17. Essey, R. J.; McDougall, B. R.; Robinson, W. E., Jr.
Antiviral Res. 2001, 189–202.
18. Montefiori, D. C.; Robinson, W. E., Jr.; Schuffman, S. S.;
Mitchell, W. M. J. Clin. Microbiol. 1988, 231–235.
19. Bundgaard, H. Adv. Drug Delivery Rev. 1989, 3, 39–65.