Unsymmetrical cyclotriazadisulfonamide (CADA) compounds as human CD4 receptor down-modulating agents

Article abstract of DOI:10.1021/jm2002603

Cyclotriazadisulfonamide (CADA) inhibits HIV at submicromolar levels by specifically down-modulating cell-surface and intracellular CD4. The specific biomolecular target of CADA compounds is unknown, but previous studies led to an unsymmetrical binding model. To test this model, methods were developed for effective synthesis of diverse, unsymmetrical CADA compounds. A total of 13 new, unsymmetrical target compounds were synthesized, as well as one symmetrical analogue. The new compounds display a wide range of potency for CD4 down-modulation in CHO-CD4-YFP cells. VGD020 (IC₅₀ = 46 nM) is the most potent CADA compound discovered to date, and VGD029 (IC₅₀ = 730 nM) is the most potent fluorescent analogue. Structure-activity relationships are analyzed from the standpoint of additive or nonadditive energy effects of different substituents. They appear to be consistent with the zipper-type mechanism in which entropy costs are reduced for additional stabilizing interactions between the small molecule and its protein target.

Full text of DOI:10.1021/jm2002603

ARTICLE
pubs.acs.org/jmc
Unsymmetrical Cyclotriazadisulfonamide (CADA) Compounds
as Human CD4 Receptor Down-Modulating Agents
Violeta G. Demillo,† Florian Goulinet-Mateo,† Jessica Kim,† Dominique Schols,‡ Kurt Vermeire,‡
and Thomas W. Bell*,†
†Department of Chemistry, University of Nevada, Reno, Nevada 89557-0216, United States
‡Rega Institute for Medical Research, Katholieke Universiteit Leuven, B-3000 Leuven, Belgium
S Supporting Information
b
ABSTRACT: Cyclotriazadisulfonamide (CADA) inhibits HIV at submicro-
molar levels by specifically down-modulating cell-surface and intracellular
CD4. The specific biomolecular target of CADA compounds is unknown, but
previous studies led to an unsymmetrical binding model. To test this model,
methods were developed for effective synthesis of diverse, unsymmetrical
CADA compounds. A total of 13 new, unsymmetrical target compounds were
synthesized, as well as one symmetrical analogue. The new compounds display
a wide range of potency for CD4 down-modulation in CHO CD4-YFP cells. VGD020 (IC50 = 46 nM) is the most potent CADA
3
compound discovered to date, and VGD029 (IC50 = 730 nM) is the most potent fluorescent analogue. Structureꢀactivity
relationships are analyzed from the standpoint of additive or nonadditive energy effects of different substituents. They appear to be
consistent with the zipper-type mechanism in which entropy costs are reduced for additional stabilizing interactions between the
small molecule and its protein target.
’ INTRODUCTION
The cluster of differentiation 4 (CD4) glycoprotein is mainly
expressed on the surface of thymocytes, T-helper lymphocytes,
and macrophage/monocyte cells.1 This intercellular adhesion
molecule plays a central role in the immune response via its
function in several steps of physiological T-cell activation.2,3
Accordingly, CD4+ T cells are believed to be important in the
pathogenesis of a variety of immunologically based diseases (e.g.,
asthma, rheumatoid arthritis, and diabetes), and anti-CD4 mono-
clonal antibodies have been explored as a therapeutic strategy for
these diseases.4ꢀ7 Anti-CD4 antibodies are also being explored
for treatment of cutaneous T-cell lymphomas, since malignant
cells express CD4 in all stages of disease progression.8 CD4 is also
the main receptor enabling infection by the human immunode-
Figure 1. (a) Structures of previously synthesized CADA compounds,
ficiency virus (HIV), so it has become an important target for
therapeutic intervention.9ꢀ12 Viral entry inhibitors are actively
being pursued as novel anti-HIV agents,13ꢀ16 including anti-
CD4 antibodies17,18 and chemokine receptor antagonists.19,20
Several studies have shown that down-modulation of CD4
effectively protects cells from HIV infection,21ꢀ25 and this has
even been found to be a strategy by which the African green
monkey avoids progression to acquired immune deficiency
syndrome (AIDS) after infection by simian immunodeficiency
virus (SIV).26
The small molecule cyclotriazadisulfonamide (CADA,
Figure 1) potently inhibits in vitro replication of a broad
spectrum of HIV-1 strains, as well as HIV-2 and SIV, by down-
modulating cell surface CD4.27 Consistent with this entry-
inhibitor mechanism of action, the anti-HIV effect of CADA is
including unsymmetrical analogues 3 and 4. (b) Proposed unsymme-
trical two-site binding model. Potency IC50(CADA)/IC50-
(compound). IC50 is for CD4 down-modulation in MT-4 cells (ref 30).
=
synergistic with those of antiviral drugs operating by other
mechanisms, including inhibitors of HIV reverse transcriptase,
protease, and attachment/entry.28 Moreover, in a series of more
than 25 CADA analogues, anti-HIV potency was directly corre-
lated with CD4 down-modulation ability.25 It was recently
discovered that CADA inhibits the synthesis of functional CD4
by blocking cotranslational translocation of nascent CD4 across
the endoplasmic reticulum (ER) membrane.29 This effect is
Received: March 6, 2011
Published: July 29, 2011
r
2011 American Chemical Society
5712
dx.doi.org/10.1021/jm2002603 J. Med. Chem. 2011, 54, 5712–5721
|

Journal of Medicinal Chemistry
ARTICLE
Scheme 1
apparently specific for the CD4 signal peptide sequence, and it
results in degradation of nascent CD4 in the cytosol.
known. The goal of the studies described here was mainly to
explore other unsymmetrical compounds in order to test our
“two-site” binding model and, hopefully, to discover more potent
compounds.
Given that the specific biomolecular target of CADA com-
pounds has not been identified, numerous analogues with
different arenesulfonyl side arms and tail groups (ArSO2 and
R, respectively, as defined in Figure 1) have been synthesized and
used to develop a 3D QSAR model for CD4 down-modulation.30
Some interesting effects of these substitutions appear in the table
in Figure 1. The column labeled “Potency” compares activities of
the analogues with that of CADA (IC50 ≈ 0.8 μM). Two
symmetrical analogues, 1 (QJ028)25 and 2 (KKD023),30 show
that either substitution of the benzyl tail by cyclohexylmethyl or
substitution of the p-tolyl groups of the side arms produces
significant increases in potency. Replacement of one p-toluene-
sulfonyl side arm of either CADA or 1 with 5-dimethylamino-
’ RESULTS AND DISCUSSION
Synthesis. Three retrosynthetic pathways for symmetrical and
unsymmetrical CADA compounds are shown in Scheme 1. All
three use the same disconnection of the macrocyclic ring of the
target molecule to an open-chain disulfonamide intermediate.
This final step of the synthesis can be accomplished by At-
kinsꢀRichman macrocyclization32,33 of the corresponding bis-
(tosylamide) dianion with 2-chloromethyl-3-chloropropene, as
described previously for preparation of CADA compounds for
structural and QSAR studies.30 The AtkinsꢀRichman macro-
cyclization requires large volumes of anhydrous DMF and gives
limited yields because of the formation of byproducts and con-
sequential losses in chromatographic purification. The syntheses
reported here generally employ an improved, recently developed,
palladium-catalyzed cyclization method.34
Previously reported symmetrical CADA compounds were
prepared by pathway I in Scheme 1, as exemplified by the
synthesis of 2 shown in Scheme 230 The requisite open-chain
disulfonamide intermediate (8) was prepared from triamine 7,
which was obtained by reduction of bis(2-cyanoethyl)benzyl-
amine (6), in turn prepared by benzylation of bis(cyanoethyl)-
amine (5). For purpose of comparison with the new, unsymmetrical
compounds reported here, 2 was debenzylated with R-chlor-
oethyl chloroformate (ACE-Cl)35,36 and the resulting secondary
amine, 9 (KKD025),30 underwent reductive alkylation to give
the novel symmetrical analogue 10 (VGD045).37
naphthalene-1-sulfonyl (dansyl) gave fluorescent probes31
3
(KKD015)30 or 4 (KKD016),30 respectively. The potencies of
these unsymmetrical analogues are comparable to that of CADA,
while the symmetrical didansyl analogue corresponding to 3 had
no detectable activity (structure and data not shown).
The finding that one p-toluenesulfonyl (tosyl) side arm of
CADA can be changed to a bulky dansyl group without strongly
affecting potency, while the same replacement of both tosyl side
arms destroys potency, led to the unsymmetrical binding-con-
formation model shown in Figure 1. This model is based on the
unsymmetrical conformation observed in the X-ray crystal
structures of CADA and three analogues, which was also used
for aligning analogue structures for 3D QSAR analysis.30 In these
structures, the 12-membered ring is twisted about the isobuty-
lene headgroup such that one sulfonamide nitrogen is approxi-
mately in the isobutylene plane and the other is out of this plane.
This places the two arenesulfonyl side arms in very different
orientations with respect to the macro ring. If this is the bioactive
conformation of the drug, then the two side arms occupy quite
different sites of the biomolecular target, labeled “Site 1” and
“Site 2” in Figure 1. It stands to reason that these two sites may
have very different size constraints and that only one can
accommodate the bulky dansyl side arm. At the time of their
synthesis, 3 and 4 were the only unsymmetrical CADA analogues
Synthesis of unsymmetrical CADA compounds requires ac-
cess to unsymmetrical disulfonamide intermediates. Pathway II
(Scheme 1) shows the route used for the preparation of unsym-
metrical, dansyl-containing analogues 3 and 4.30 An inexpensive
starting material, bis(3-aminopropyl)amine, was desymmetrized
by condensation with acetaldehyde to form a cyclic aminal
(a hexahydropyrimidine) bearing primary, secondary, and
5713
dx.doi.org/10.1021/jm2002603 |J. Med. Chem. 2011, 54, 5712–5721

Journal of Medicinal Chemistry
ARTICLE
Scheme 2
tertiary nitrogen atoms. Tosylation at the more nucleophilic
secondary nitrogen followed by dansylation of the primary
nitrogen and ring-opening proceeded in low yield (18% overall),
but after installation of benzyl or cyclohexylmethyl tail groups
and AtkinsꢀRichman macrocyclization, the corresponding CADA
analogues 3 and 4 were obtained.
A new approach to unsymmetrical CADA compounds is
shown as retrosynthetic pathway III (Scheme 1). Here, the
unsymmetrical disulfonamide intermediate is obtained by chain
elongation of an N-alkyl-N-sulfonyl-1,3-diaminopropane. This
intermediate, in turn, is prepared by reductive alkylation of a
monotosyl or monodansyl derivative of 1,3-diaminopropane.
Pathway III shares with pathway II the desymmetrization of an
inexpensive starting material, but the new approach is more
effective because yields are higher overall.
The syntheses of 12 new, unsymmetrical CADA compounds
by pathway III are shown in Scheme 3.37 For compounds in
which one of the arenesulfonyl groups is tosyl (Ar1 = p-tolyl), the
synthesis began with reaction of tosyl chloride with an excess of
1,3-diaminopropane,38,39 giving the monotosyl derivative (11) in
48% yield. While this is only a modest yield, the reactants are
inexpensive and a low yield is best tolerated early in a synthesis.
Dansylation of 1,3-diaminopropane gave monodansyl derivative
12 in 99% yield. Reductive alkylation of 11 and 12 by reaction
with cyclohexanecarboxaldehyde and sodium borohydride gave
the corresponding N-cyclohexylmethyl derivatives (13 and 14)
in high yield. In the tosyl case (13), the crude product contained
a minor impurity of the dialkylation side product, but this could
be removed by chromatography at a later stage of the synthesis.
Eight new, unsymmetrical CADA compounds (28ꢀ34 and
37) were synthesized by the longer, five-step route shown in
Scheme 3.37 All of these bear a tosyl side arm (Ar1 = p-tolyl) and
were synthesized from diamine intermediate 13 by alkylation
with 3-bromopropylphthalimide40 to give intermediate 15 in
84% yield. Cleavage of the phthalimide protecting group with
hydrazine gave triamine 16 (96%), which was sulfonylated with
various arenesulfonyl chlorides to afford open-chain intermedi-
ates 17ꢀ23 and 26 in 34ꢀ78% yield, after purification by column
chromatography. In some cases, the shorter, three-step route
shown in Scheme 3 was compared. In one case, reaction of
intermediate 13 with N-(3-bromopropyl)benzenesulfonamide
(Ar2 = phenyl) gave the corresponding open-chain triamine
(24) in higher yield than was obtained overall for three steps via
intermediates 15 and 16. The two dansylated analogues 36 and
38 were also prepared via the shorter route by reaction of inter-
mediate 14 with N-(3-bromopropyl)-4-methoxybenzenesulfo-
namide or N-(3-bromopropyl)-4-nitrobenzenesulfonamide to
give open-chain intermediate 25 or 27, respectively.
All of the 11 new unsymmetrical CADA compounds were
prepared from the open-chain intermediates by a recently
developed cyclization method using catalytic amounts of
palladium(0) and 1,4-bis(diphenylphosphino)butane (dppb).34
This reaction uses smaller amounts of a more volatile solvent
(acetonitrile) and generally produces fewer byproducts than the
AtkinsꢀRichman method. Nevertheless, the yields of the pure
macrocycles ranged widely (19ꢀ53%), apparently because of
losses from repeated chromatographic separations required for
complete purification in a few cases. A one-armed analogue, 39
(VGD040),37 was also prepared by cleavage of the 2-nitroben-
zenesulfonyl side arm of 26 with 2-mercaptoethanol.41
After the high potency of compound 31 (VGD020) was
discovered, benzyl tail analogue 42 (VGD027)37 was synthesized
by the route shown in Scheme 4.37 This parallels the shorter,
three-step route to cyclohexylmethyl tail compounds shown in
Scheme 3. Thus, the monotosyl derivative of 1,3-diaminopro-
pane (11) was reductively alkylated with benzaldehyde to give 40
in 70% yield. Reaction of this diamine with N-(3-brompropyl)-4-
methoxybenenesulfonamide gave open-chain triamine 41
(VGD026)37 in 54% yield, after purification by column chroma-
tography. Palladium catalyzed macrocyclization, followed by
purification as the hydrochloride salt afforded pure 42 HCl
3
3
H2O in 56% yield.
5714
dx.doi.org/10.1021/jm2002603 |J. Med. Chem. 2011, 54, 5712–5721

Journal of Medicinal Chemistry
ARTICLE
Scheme 3
Potency. Samples of the 14 new CADA compounds were
tested for CD4 down-modulation either in their analytically pure
(>95%) free base forms or as hydrochloride salts, as indicated in
the Experimental Section for compositions determined by com-
bustion microanalysis. Most of the compounds tested were
stoichiometric hydrates of monohydrochloride salts (or dihy-
drochoride salts when a basic site was present in a side arm). The
CD4 down-modulation potencies of the new CADA compounds
described in this publication are given in Table 1 as IC50 values
(concentration producing 50% decrease in CD4 in Chinese
hand, we believe that energetic analysis of the results is still useful,
as it may lead to postulation of new compounds for synthesis and
testing.
Replacement of the benzyl tail group of CADA with cyclohex-
ylmethyl (1) or the tosyl side arms of CADA with 4-methox-
ybenzenesulfonyl (2) was previously found to increase potency
for CD4 down-modulation, as shown in Figure 1; the current
results in a different cell system reproduce these trends. Thus, it
was of interest to compare the effects of different combinations of
these substitutions. As seen in Table 1, replacement of both tosyl
side arms of CADA with 4-methoxybenzenesulfonyl in addition
to replacing the tail with cyclohexylmethyl gives a slight further
increase in potency (10). The two analogues with single side arm
substitutions, 42 with a benzyl tail and 31 with a cyclohexyl-
methyl tail, were of greater interest to test the proposed, unsym-
metrical, two-site binding model shown in Figure 1. The potency
of 42 is between those of the symmetrical benzyl tail compounds
CADA and 2, while unsymmetrical cyclohexylmethyl tail ana-
logue 31 is more potent than either symmetrical analogue, 1 or
10. Compound 31 is the most potent CADA analogue synthe-
sized to date.
hamster ovary/CD4/yellow fluorescent protein (CHO CD4-
3
YFP) cells after 24 h of drug treatment). Also compared in
Table 1 are the CD4 down-modulation potencies of the five
structurally related compounds shown in Figure 1, the symme-
trical compounds CADA, 1, and 2, and unsymmetrical analogues
3 and 4. Also given for each compound in Table 1 are pIC50
values (ꢀlog10 IC50), which is, in principle, proportional to the
negative free energy of binding of the small molecule to its
binding site. This is rigorously true if the compounds bind only to
the site of CD4 down-modulation or do not distribute differently
between various binding sites. Since the IC50 values are measured
in cell-based assays, different distributions of different com-
pounds between various binding sites is possible. On the other
Relative to CADA, monodansyl analogues 3 and 4 appear to
be less potent than found previously in different cell systems,
5715
dx.doi.org/10.1021/jm2002603 |J. Med. Chem. 2011, 54, 5712–5721

Journal of Medicinal Chemistry
ARTICLE
Scheme 4
Table 1. CD4 Down-Modulating Activities of CADA Com-
pounds in CHO CD4-YFP Cells
3
IC50 (μM)b
c
compda
mean
(
SDV
pIC50
CADA
1
0.56
0.19
0.23
5.9
(
(
(
(
(
(
(
0.046
0.003
0.003
3.9
6.25
6.73
6.64
5.23
5.97
6.11
6.44
4.64
7.34
6.03
6.75
6.12
6.36
6.53
6.14
6.65
4.73
6.87
2
3
4
1.06
0.78
0.36
>25
0.046
0.94
0.18
0.76
0.43
0.30
0.73
0.22
19
0.063
0.033
0.031
28
29
30
31
32
33
34
35
42
36
37
39
10
(
(
(
(
(
(
(
(
(
(
0.002
0.035
0.015
0.033
0.012
0.027
0.050
0.024
9.3
though cyclohexylmethyl tail analogue 4 is still found to be more
potent than benzyl tail system 3. Replacement of the second tosyl
side arm of 4 with 4-methoxybenzenesulfonyl further improves
potency, making 36 the most potent fluorescent CADA analogue
to date. Compound 32, bearing a naphthalene-1-sulfonyl in place
of the dansyl group of 4, is marginally more potent than 4,
suggesting that the dimethylamino group of the dansyl system
does not contribute to binding. The isomer with a naphthalene-
2-sulfonyl side arm (33) is much more potent than 32 and
slightly more potent than ditosyl side arm analogue 1, suggesting
the importance of side arm size and shape. Replacing one of the
naphthalene-1-sulfonyl carbons with nitrogen, as in isoquinoline-
5-sulfonyl analogue 30, greatly decreases potency, as does
complete removal of one side arm in 39. Replacement of one
tosyl methyl group of 1 with bulky tert-butyl in 28 or electro-
negative nitro in 34 causes nearly 4-fold decreases in activity,
while replacement of a methyl with smaller chloro in 29 or with
hydrogen in 35 causes more modest decreases in potency.
Interestingly, movement of a nitro group from the para position
of one benzene side arm in 34 to the ortho position in 37 restores
potency almost to the level of 1.
Additivity. The mechanism by which CADA compounds
inhibit cotranslational translocation of CD4 across the ER
membrane is not known. It is likely that the biomolecular target
is a protein, such as the Sec61 channel,42 the CD4 signal peptide
sequence, or a combination of both. The third possibility is
precedented by the mechanism of inhibition of cotranslational
translocation of certain proteins in endothelial cells by cotransins43
and related compounds.44 Analyzing structureꢀactivity relation-
ships can help to infer molecular characteristics of the drug-
binding site, to increase potency, and to develop analogues as
tools for elucidating the drug target. A common approach in
medicinal chemistry is to consider the contribution of each
molecular fragment to the total binding energy of the drug.
Proteinꢀligand affinity, as described by the binding or asso-
ciation constant Ka (or the inverse, dissociation constant Kd) is
determined by the equilibrium free energy ΔG. Free energy is
composed of enthalpic (ΔH) and entropic (ΔS) contributions,
according to the Gibbs equation, ΔG = ΔH ꢀ TΔS. The driving
force for association of a small molecule with a protein is thus
determined by changes in both enthalpy and entropy. Enthalpic
contributions include ligandꢀprotein interactions (e.g., van der
0.13
0.006
a CADA, 1, and 4 were used as HCl salts, while 2 and 3 were used in their
free base forms, as described in ref 30. All VGD compounds were used in
their free base or salt forms, as indicated for combustion microanalysis
results in the Experimental Section. b IC50: inhibitory concentration
50%, i.e., concentration at which 50% down-modulation of CD4
expression was measured in CHO CD4-YFP after 24 h of drug
3
treatment. Values are the mean + standard deviation from three
independent experiments (except for 35, 37, and 39, for which two
independent experiments were done). c pIC50 = ꢀlog(IC50).
Waals, hydrogen bonding, ionꢀdipole, and dipoleꢀdipole), as
well as altered interactions within the protein structure induced
by ligand binding. Entropic contributions result from decreased
ligand translational and rotational entropy, residual motion of
both the ligand and protein in the bound state, and solvent
effects, such as increased motion of water molecules released
from both the ligand and protein upon binding. It is well
recognized that these diverse factors confound simplistic at-
tempts to assess the individual contributions of fragments in drug
design.45,46 Enthalpic contributions of each fragment may be
additive, but there is no reason to expect entropic contributions
to be additive, according to first principles. A general observation
termed “enthalpy/entropy compensation” is that enthalpically
favorable interactions often produce opposite entropy effects,
resulting in small changes in free energy.45,46
Despite the fundamental fallacy of adding group energy
increments to obtain binding energies, detection of additive or
nonadditive effects may produce useful insights about drugꢀ
receptor interactions. For example, Williams proposed that
incremental introduction of new binding interactions, or increas-
ing the strength of existing interactions, can be “cooperative”
when residual motion of the ligand within the binding site is not
further reduced.45 Thus, the entropic cost of restricting motion
(of both the ligand and the protein) has already been paid, and
new (or stronger) interactions are less compensated by entropy.
There may also be an enthalpic component to this cooperativity
5716
dx.doi.org/10.1021/jm2002603 |J. Med. Chem. 2011, 54, 5712–5721

Journal of Medicinal Chemistry
ARTICLE
Figure 2. Thermodynamic cycle comparing energy effects of substitutions, starting with CADA (upper left). The pIC50 for CD4 down-modulation is
given at lower right of each compound structure. Numbers above or beside each arrow represent change in pIC50 for each substitution.
in that proteins can become more tightly packed upon ligand
binding, a possible explanation for a number of exceedingly
strong bimolecular interactions.47 In this context, positive co-
operativity occurs when ligand binding is stronger than the sum
of its parts, which is different from the definition of cooperativity
used for allosteric interactions in proteins with more than one
binding site. A recent thermodynamic and structural study of
thrombin inhibitors examining additivity of functional group
contributions found that the greatest cooperative interactions
were associated with decreasing loss of residual entropy upon
introduction of additional interactions (zipper-type mechanism).48
In the present study, it was of interest to examine the energetic
contributions of substituents in the most potent compound, 31.
The relevant molecular structures and corresponding changes in
pIC50 values are displayed in the thermodynamic cycle shown in
Figure 2. As stated previously, pIC50 values are directly propor-
tional to the negative free energy of binding if the compounds
bind only to the site of CD4 down-modulation or do not
distribute differently between various binding sites. Replacement
of one tosyl side arm of CADA by 4-methoxybenzenesulfonyl in
42 increases potency by 0.28 pIC50 units, while replacement of
the benzyl tail with cyclohexylmethyl in 1 increases potency by
0.48 pIC50 units. The same substitutions on 1 and 42 have about
double the energetic effect (0.61 and 0.81, respectively). Thus,
functional group contributions in the transformation of CADA to
31 are nonadditive and show positive cooperativity. It is tempting
to conclude that either substitution on CADA results in most of
the loss of residual motion attending binding of 31 to its target.
Accordingly, the structure of 31 is very well adapted to fit the
binding site and replacement of the second tosyl side arm with
4-methoxybenzenesulfonyl in 10 decreases potency by 0.47 pIC50
units, possibly because the larger methoxy group fits less well
than methyl in the binding site. The same replacement in 42
causes a slight increase in potency for 2, perhaps because the
benzyl tail causes a different placement of the side arms in their
respective binding sites. The high potency of unsymmetrical
compound 31 agrees with the unsymmetrical binding model
shown in Figure 1, though it remains to be determined which one
of the two different side arms fits into which of the two sites.
From the binding data available in Table 1, it is possible to
construct a second thermodynamic cycle, which is shown in
Figure 3. For monodansyl compound 3, replacement of the
benzyl tail by cyclohexylmethyl improves potency by 0.74 pIC50
units (4), whereas changing the dansyl side arm to 4-methox-
ybenzenesulfonyl causes a greater enhancement (1.30 pIC50
units, 42). The corresponding substitutions on 42 or 4 produce
almost the same energetic effects (changes of 0.81 and 1.37,
respectively). This differs from the increased effect of the second
substitution in either pathway converting CADA to 31
(Figure 2). In Figure 3, the starting structure, 3, is about 10-fold
less potent than CADA; hence, its molecular structure is more
poorly adapted to the binding site and a large degree of residual
motion is expected in the complex. Replacement of the benzyl
tail by cyclohexylmethyl would not be expected to greatly reduce
residual motion, but changing dansyl to 4-methoxybenzenesul-
fonyl should. These expected differences in residual motion
changes might account for the apparent lack of cooperativity in
this case. Of course, the presence of the dimethylamino group in
dansyl introduces the possibility of solvent-related entropy
changes not associated with motion in the drugꢀprotein com-
plex. Finally, replacement of the tosyl side arm of 4 with
4-methoxybenzenesulfonyl causes a modest increase in potency
for 36, as seen for the same substitution in 42 (Figure 2).
Symmetry. The molecular structures of many drugs are
symmetrical, as defined by symmetry elements including rota-
tional axes, mirror planes, or centers of inversion.49ꢀ53 In many
cases, drug symmetry may reflect the method of synthesis
because it is often easier to synthesize symmetrical structures,
as is the case with CADA compounds. While amino acids,
nucleotides, and carbohydrates are asymmetric (chiral), in rare
cases the drug-binding site may be dissymmetric and possess
rotational symmetry, for example, in multimeric protein assem-
blies that often occur in viruses.53 The binding site for CADA
compounds, which probably involves the Sec 61 translocon, the
CD4 signal peptide sequence, or both, is unlikely to be sym-
metric. Therefore, it is understandable that unsymmetrical com-
pound 31 is more potent than either of its symmetrical analogues,
1 or 10 (cf. Figure 2). The slightly greater potency of symmetrical
5717
dx.doi.org/10.1021/jm2002603 |J. Med. Chem. 2011, 54, 5712–5721

Journal of Medicinal Chemistry
ARTICLE
Figure 3. Thermodynamic cycle comparing energy effects of substitutions, starting with 3 (upper left). The pIC50 for CD4 down-modulation is given at
lower right of each compound structure. Numbers above or beside each arrow represent change in pIC50 for each substitution.
2 over unsymmetrical 42, however, suggests that symmetry is not
necessarily more important than the physicochemical properties
of substituents. As stated previously, potencies were measured in
whole cell assays, and some observed differences might also arise
from different distribution of the compounds between more than
one potential binding site. Compound 31 is the most potent
CADA compound to date, and its potency is consistent with
unsymmetrical binding models, such as that displayed in Figure 1.
Its discovery paves the way for future, conformationally con-
trolled analogues capable of separately probing the characteris-
tics of the two different side arm binding sites.
with different concentrations of CADA or its analogues at 37 °C. Cells
were then washed, fixed in 1% formaldehyde, and analyzed immediately.
Data were acquired with a FACSCalibur flow cytometer (BD Biosciences)
using the 488 nm laser line and CellQuest software (BD Biosciences).
YFP was measured with the FL-1 detector, and data were analyzed with
FLOWJO software (Tree Star, San Carlos, CA). Down-modulation of
CD4 was evaluated by the decrease in fluorescence intensity on CADA-
treated cells relative to matched, untreated cells. To calculate the
efficiency of CD4 down-modulation, the median fluorescence intensity
(MFI) for YFP for each sample was expressed as a percentage of the MFI
of control cells (after subtracting the background MFI of the nontrans-
fected control cells).
General Methods. All reactions were performed under an atmo-
sphere of dry nitrogen. Reagents and solvents purchased from Aldrich
Chemical Co., Acros Organics, or Fisher Scientific were of ACS reagent
grade or better and were used without purification, unless indicated
otherwise. Anhydrous acetonitrile was distilled from CaH2. HCl (2 N) in
methanol/water was prepared from 42 mL of concentrated aqueous HCl
(12.1 N) and 210 mL of methanol. For macrocyclization reactions, the
disulfonamide intermediate (previously washed with aqueous NaOH, as
described for deprotonation of HCl salts) and 2-methylene 1,3-
propanebis(tert-butylcarbonate) were dried in vacuo for at least 16 h
and then dissolved in anhydrous acetonitrile with 1,4-bis(diphenylpho-
sphino)butane (dppb), and Pd2(dba)3 was added last. Column chro-
matography was performed with Sorbent Technologies neutral alumina
(50ꢀ200 μm) or Sorbent Technologies standard grade silica (32ꢀ63
μm), unless noted otherwise. Melting points were measured on Tho-
’ CONCLUSIONS
Effective methods have been developed for the synthesis of
diverse, unsymmetrical CADA compounds. CD4 down-modula-
tion assays of the 14 new compounds displayed a wide range of
potency and revealed the most potent CADA compound dis-
covered to date, 31, and the most potent fluorescent analogue,
36. Relative to CADA, group contributions to the potency of 31
are nonadditive and show positive cooperativity, consistent with
the zipper-type mechanism in which additional binding interac-
tions exact lower entropy costs due to smaller changes in residual
motion. This conclusion is based on the assumption that the
potencies of these compounds are directly related to their affinity
for the CD4 down-modulation binding site. Compound 31 is
more potent than either of its symmetrical analogues, consistent
with the proposed, unsymmetrical binding model. These results
lay important groundwork for developing effective tools for
elucidating the exact molecular mechanism by which CADA
compounds inhibit cotranslational translocation of nascent CD4.
Discovery of this mechanism could pave the way to designing
drugs for inhibiting expression of other membrane proteins.
1
mas-Hoover or Mel-Temp apparatus and are uncorrected. H NMR
(400 or 500 MHz) and 13C NMR (100 or 125 MHz) spectra were
acquired on a Varian 400 or Varian Unity +500 spectrometer. All
chemical shifts (δ) are reported in ppm units relative to solvent
1
resonances, as follows: H, CDCl3/TMS = 0.00, DMSO-d6 = 2.50,
CD3OD = 3.31; 13C, CDCl3 = 77.23, DMSO-d6 = 39.7, CD3OD =
49.15. Infrared spectra (IR) were recorded on a Nicolet 6700 FTIR
spectrometer. Mass spectra (MS) were acquired on a Waters Micromass
ZQ electrospray ionization quadrupole mass spectrometer with positive
ion detection (capillary voltage of 3.5 kV). High-resolution mass spectra
(HRMS) were acquired on an Agilent 6230 TOF mass spectrometer.
Samples for elemental analysis were dried at 78 °C (0.1 mm) for 2 days,
’ EXPERIMENTAL SECTION
Flow Cytometry. To study the effect of CADA on CD4 expression,
CHO cells, stably expressing CD4-YFP (huCD4 fused at its COOH-
terminus to the yellow fluorescent protein [YFP]), were treated for 24 h
5718
dx.doi.org/10.1021/jm2002603 |J. Med. Chem. 2011, 54, 5712–5721

Journal of Medicinal Chemistry
ARTICLE
unless stated otherwise, and microanalysis was performed by NuMega
Resonance Labs, Inc. Samples tested for CD4 down-modulation were
greater than 95% pure, as shown by combustion microanalysis.
N-(3-p-Toluenesulfonamidopropyl)benzylamine (40). A mix-
ture of 2.01 g (8.80 mmol) of N-(3-aminopropyl)-p-toluenesulfonamide
(11), 1.05 g (9.86 mmol) of benzaldehyde, 3.18 g (26.4 mmol) of
anhydrous MgSO4, and 25 mL of CH2Cl2 was stirred at room temperature
for 24 h and then filtered under vacuum through a fine porosity sintered
glass funnel. The white residue was washed with 10 mL of CH2Cl2. The
combined filtrates were concentrated by rotary evaporation, and the
resulting colorless oil was dried in vacuo. A solution of the residue in
10 mL of absolute ethanol was stirred under nitrogen as 0.45 g (12 mmol) of
NaBH4 was added in portions over 30 min. The reaction mixture was stirred
at room temperature overnight (at least 12 h), then was vacuum filtered
through a medium porosity sintered glass funnel, rinsing the reaction flask
and the residue with ∼10 mL of ethanol. The combined filtrates were
concentrated to dryness by rotary evaporation, and the residue was stirred
with 20 mL of water. The resulting, slightly cloudy mixture was extracted
with CH2Cl2 (3 ꢁ 10 mL). The combined CH2Cl2 extracts were dried
(Na2SO4) and concentrated by rotary evaporation, yielding 1.97 g (70%) of
40 as a white solid. 1H NMR (500 MHz, CDCl3/TMS) δ 7.71 (m, 2 H,
o-Ts), 7.32 (m, 2 H, o-Ph), 7.26 (m, 5 H, m,p-Ph, m-Ts), 3.70 (s, 2 H,
CH2Ph), 3.04 (t, 6 Hz, 2 H, CH2NH), 2.67 (t, 6 Hz, 2 H, CH2N), 2.41 (s, 3
H, ArCH3), 1.62 (quint, 6 Hz, 2 H, CCH2C). 13C NMR (125 MHz,
CDCl3) δ 143.2, 139.8, 137.3, 129.8, 128.7, 128.3, 127.3, 127.2, 54.0, 48.5,
43.8, 28.0, 21.7. IR (neat, cmꢀ1) 3043 (w), 2831 (w), 1598 (w), 1495 (w),
1477 (w), 1454 (w), 1324 (s), 1304 (m), 1286 (w), 1273 (w), 1185 (w),
1155 (s), 1089 (m), 1072 (s), 1043 (w), 1031 (w), 979 (m), 835 (m), 815
(s), 781 (m), 742 (s), 702 (s), 660 (s). MS m/z 319 (MH+). Anal. Calcd for
C17H24N2O4S: C, 63.72; H, 7.55; N, 8.74. Found: C, 63.78; H, 6.94; N, 8.85.
N-(4-Methoxybenzenesulfonyl)-N0-(p-toluenesulfonyl)bis-
(3-aminopropyl)benzylamine (41). A mixture of 0.80 g (2.5 mmol)
of 40, 0.40 g (3.8 mmol) of Na2CO3, 0.18 g (1.2 mmol) of NaI, 0.93 g
(3.0 mmol) of N-(3-bromopropyl)-4-methoxybenzenesulfonamide, and
9 mL of MeCN was stirred with heating under reflux for 24 h, then
cooled to room temperature and filtered through a fine porosity sintered
glass funnel, washing the residue with 10 mL of MeCN. The combined
filtrates were concentrated by rotary evaporation, and a solution of the
resulting residue in 15 mL of CH2Cl2 was washed with 3 ꢁ 5 mL of H2O,
dried (Na2SO4), and concentrated by rotary evaporation. The resulting
faintly yellow oil was dried in vacuo to give 1.34 g (98%) of 41. Alumina
column chromatography, eluting with CH2Cl2/EtOAc, gave 0.75 g
9-Cyclohexylmethyl-1,5-(4-methoxybenzenesulfonyl)-3-
methylene-1,5,9-triazacyclododecane (10). A solution of 0.21
g (0.39 mmol) of 3-methylene-1,5-bis(4-methoxybenzenesulfonyl)-
1,5,9-triazacyclododecane (9) and 0.29 g (2.6 mmol) of cyclohexane-
carboxaldehyde in 10 mL of absolute ethanol was stirred at room
temperature for 1 h. Then a solution of 87 mg (1.4 mmol) of NaBH3CN
in 5 mL of absolute ethanol was added dropwise over 10 min. The
resulting mixture was stirred at room temperature for 24 h and then
concentrated to dryness by rotary evaporation. A solution of the residue
in 10 mL of CH2Cl2 was washed with H2O (3 ꢁ 5 mL) followed by 5 mL
of saturated aqueous NaCl, then dried (Na2SO4). Concentration by
rotary evaporation followed by two silica column chromatograms,
eluting with CH2Cl2/MeOH, gave 0.17 g (71%) of 10 as a colorless,
viscous oil. 1H NMR (500 MHz, CDCl3/TMS) δ 7.72 (m, 4 H,
o-ArSO2), 7.00 (m, 4 H, m-ArSO2), 5.18 (s, 2 H, CdCH2), 3.88 (s,
6 H, OCH3), 3.79 (s, 4 H, H2, H4), 3.15 (t, 7 Hz, 4 H, H6, H12), 2.27 (t,
6 Hz, 4 H, H8, H10), 1.97 (d, 7 Hz, 2 H, CH2Cy), 1.62 (m, 10 H, H7,
H11, Cy), 1.25 (m, 1 H, Cy), 1.14 (m, 2 H, Cy), 0.70 (m, 2 H, Cy). 13C
NMR (125 MHz, CDCl3) δ 163.1, 138.2, 130.5, 129.5, 116.7, 114.5,
62.4, 55.8, 51.1, 50.6, 44.3, 36.1, 32.1, 27.0, 26.2, 24.7. IR (neat, cmꢀ1
)
2923 (w), 2848 (w), 2801 (w), 1596 (m), 1578 (w), 1497 (m), 1462
(w), 1334 (m), 1306 (m), 1258 (s), 1153 (s), 1092 (m), 1026 (m), 938
(m), 910 (m), 834 (m), 806 (m), 729 (s), 690 (m).
10 HCl. A solution of 0.13 g of 10 in 10 mL of 2 N HCl in methanol/
3
water was stirred at room temperature for 4 h. Concentration by rotary
evaporation and drying (0.1 mm) gave a residue that was triturated with
ether (5 ꢁ 10 mL) using sonication and dried in vacuo, giving 0.13 g
1
(93%) of 10 HCl as a white powder. H NMR (500 MHz, CDCl3/
3
TMS) δ 7.69 (m, 4 H, o-ArSO2), 7.02 (m, 4 H, m-ArSO2), 5.34 (s, 2 H,
CdCH2), 3.90 (s, 6 H, ArOCH3), 3.70 (m, 4 H, H2, H4), 3.40 (m, 2 H,
H8/H10), 3.19 (m, 4 H, H6/H8/H10/H12), 3.12 (m, 2 H, H6/H12),
2.77 (t, 7 Hz, 2 H, CH2Cy), 2.32 (m, 2 H, H7/H11), 1.97 (m, 4 H, H7/
H11, Cy), 1.80 (m, 3 H, Cy), 1.69 (m, 1 H, Cy), 1.2 (m, 5 H, Cy). IR
(neat, cmꢀ1) 3401 (w), 2930 (w), 2851 (w), 1596 (m), 1577 (w), 1498
(m), 1455 (w), 1336 (m), 1307 (w), 1262 (m), 1157 (s), 1092 (m),
1023 (w), 806 (w). MS m/z 620 (MH+). Anal. Calcd for C31H45-
1
N3O6S2 HCl: C, 56.73; H, 7.06; N, 6.40. Found: C, 56.92; H, 7.46;
(54%) of 41 as an amber viscous oil. H NMR (500 MHz, CDCl3/
3
N, 6.52.
TMS) δ 7.76 (m, 2 H, o-ArSO2), 7.70 (m, 2 H, o-Ts), 7.27 (m, 5 H, o,
p-Ph, m-Ts), 7.20 (m, 2 H, m-Ph), 6.96 (m, 2 H, m-ArSO2), 3.86 (s, 3 H,
ArOCH3), 3.46 (s, 2 H, CH2Ph), 2.92 (m, 4 H, CH2NH), 2.43 (t, 6 Hz,
N-(3-Aminopropyl)-p-toluenesulfonamide (11). A filtered
solution of 50.4 g (0.26 mol) of p-toluenesulfonyl chloride in 180 mL
of toluene was added dropwise over 6 h to a vigorously stirred solution of
65 mL (58 g, 0.78 mol) of 1,3-diaminopropane and 80 mL of toluene.
The resulting white suspension was stirred at room temperature over-
night (12ꢀ17 h). The white precipitate was collected by vacuum
filtration, washed with 20 mL of cold toluene, and dried (0.1 mm)
overnight, yielding 67 g of crude product as a white solid, mp
95ꢀ135 °C. This solid was stirred for approximately 0.5 h with 1 L of
1:1 (v/v) methanol/water. Then the resulting suspension was filtered
(vacuum). The residue consisted of 4.08 g of 1,3-bis(p-toluenesulfona-
mido)propane, mp 140ꢀ147 °C (lit.38 147ꢀ149 °C). The filtrate was
stored at 4 °C for 3 h, then filtered to give an additional 0.43 g of 1,3-
bis(p-toluenesulfonamido)propane, mp 143ꢀ146 °C. The filtrate was
concentrated by boiling to a volume of about 600 mL, then stored at 4 °C
for 3 h. The resulting precipitate was collected by vacuum filtration,
washed with water, and dried (0.1 mm). The filtrate was stored at 4 °C
for 3 h and filtered to obtain a second crop. A total yield of 28.7 g (48%)
of 11 was obtained as beige crystals, mp 114ꢀ117 °C (lit. 112ꢀ
114 °C,39 113ꢀ115 °C38). 1H NMR (400 MHz, CD3OD) δ 7.73 (m, 2
H, o-Ts), 7.38 (m, 2 H, m-Ts),2.88(t, 7Hz,2H,CH2NH),2.63(t,7Hz, 2H,
CH2NH2), 2.42 (s, 3H, ArCH3), 1.59 (quint, 7 Hz, 2 H, CCH2C). 13C NMR
(125 MHz, CD3OD) δ 144.7, 139.0, 130.9, 128.2, 41.7, 39.7, 33.5, 21.6.
4 H, CH2N), 2.42 (s, 3 H, ArCH3), 1.65 (quint, 6 Hz, 4 H, CCH2C). 13
C
NMR (125 MHz, CDCl3) δ 163.0, 143.4, 138.1, 137.2, 131.7, 129.9,
129.41, 129.35, 128.7, 127.6, 127.3, 114.4, 58.9, 55.8, 52.1, 42.40, 42.38,
26.2, 21.7. IR (neat, cmꢀ1) 3279 (w), 2946 (w), 1597 (m), 1579 (w),
1497 (m), 1453 (w), 1324 (m), 1260 (m), 1155 (s), 1094 (m), 1026
(w), 962 (w), 910 (w), 835 (m), 816 (w), 736 (m), 700 (w), 665 (m).
41 HCl. A solution of 0.13 g of 41 in 10 mL of 2 N HCl in methanol/
3
water was stirred at room temperature for 4 h then concentrated by
rotary evaporation. The residue was dried (0.1 mm), then triturated with
ether (5 ꢁ 10 mL) using sonication, giving 0.12 g (86%) of 41 HCl as a
3
beige powder. 1H NMR (500 MHz, CD3OD) δ 7.78 (m, 2 H, o-ArSO2),
7.73 (m, 2 H, o-Ts), 7.52 (m, 5 H, Ph), 7.40 (m, 2 H, m-Ts), 7.09 (m, 2
H, m-ArSO2), 4.36 (s, 2 H, CH2Ph), 3.87 (s, 3 H, ArOCH3), 3.22 (m, 4
H, CH2N+), 2.91 (m, 4 H, CH2NH), 2.43 (s, 3 H, ArCH3), 1.96 (m, 4 H,
CCH2C). IR (neat, cmꢀ1) 3075 (w), 2933 (w), 2844 (w), 1596 (w),
1499 (w), 1322 (m), 1260 (m), 1156 (s), 1091 (m), 1027 (w), 926 (w),
831 (m), 806 (m), 755 (m), 734 (m), 704 (m), 689 (m). MS m/z 546
(MH+). Anal. Calcd for C27H35N3O5S2 HCl H2O: C, 54.03; H, 6.38;
3
3
N, 7.00. Found: C, 54.16; H, 6.71; N, 7.17.
9-Benzyl-1-(4-methoxybenzenesulfonyl)-3-methylene-5-(p-
toluenesulfonyl)-1,5,9-triazacyclododecane (42). A mixture of
5719
dx.doi.org/10.1021/jm2002603 |J. Med. Chem. 2011, 54, 5712–5721

Journal of Medicinal Chemistry
ARTICLE
0.30 g (0.55 mmol) of 42, 0.40 g (1.4 mmol) of 2-methylene 1,3-
propanebis(tert-butylcarbonate), 15 mg (0.034 mmol) of dppb, 16 mg
(0.017 mmol) of tris(dibenzylideneacetone)dipalladium(0), and 35 mL of
anhydrous MeCN was stirred and heated under reflux for 24 h, cooled to
room temperature, and concentrated to dryness by rotary evaporation. A
solution of the residue in 26 mL of CH2Cl2 was washed with saturated
aqueous NaHCO3 (2 ꢁ 9 mL) and saturated aqueous NaCl (9 mL), dried
(2) Janeway, C. A., Jr. The role of CD4 in T-cell activation: accessory
molecule or co-receptor? Immunol. Today 1989, 10, 234–238.
(3) Ellmeier, W.; Sawada, S.; Littman, D. R. The regulation of CD4
and CD8 coreceptor gene expression during T cell development. Annu.
Rev. Immunol. 1999, 17, 523–554.
(4) Nepom, G. T. Therapy of autoimmune diseases: clinical trials
and new biologics. Curr. Opin. Immunol. 2002, 14, 812–815.
(5) Taylor, P. C. Antibody therapy for rheumatoid arthritis. Curr.
Opin. Pharmacol. 2003, 3, 323–328.
(6) Liu, E.; Moriyama, H.; Paronen, J.; Abiru, N.; Miao, D.; Yu, L.;
Taylor, R. M.; Eisenbarth, G. S. Nondepleting anti-CD4 monoclonal
antibody prevents diabetes and blocks induction of insulin autoantibo-
dies following insulin peptide B:9-23 immunization in the NOD mouse.
J. Autoimmun. 2003, 21, 213–219.
(Na2SO4), and concentrated to dryness by rotary evaporation. 42 HCl. A
3
solution of crude 42 in 10 mL of 2 N HCl in methanol/water was stirred at
room temperature for 4 h, then concentrated by rotary evaporation. The
residue was dried (0.1 mm), then triturated with ether (5 ꢁ 15 mL) using
sonication, giving 0.33 g (92%) of 42 HCl as a cream-colored powder.
3
42 HCl was further purified by adding ether dropwise to a solution in
3
CH2Cl2 until cloudiness was observed. The precipitated solid was collected
(7) Hanns-Martin, L. T-cell-activation inhibitors in rheumatoid
arthritis. BioDrugs 2003, 17, 263–270.
by filtration and dried in vacuo, giving 0.20 g (56%) of 42 HCl H2O as a
3
3
1
cream-colored solid. H NMR (500 MHz, CD3OD) δ 7.75 (m, 2 H,
o-ArSO2), 7.70 (m, 2 H, o-Ts), 7.55 (m, 5 H, Ph), 7.46 (m, 2 H, m-Ts), 7.14
(m, 2 H, m-ArSO2), 5.41 (s, 1 H, CdCH2), 5.40 (s, 1 H, CdCH2), 4.38
(s, 2 H, CH2Ph), 3.90 (s, 3 H, ArOCH3), 3.76 (s, 4 H, H2, H4), 3.45 (bs, 2
H, H8/H10), 3.35 (bs, 2 H, H8/H10), 3.19 (m, 4 H, H6, H12), 2.46 (s, 3
H, ArCH3), 2.06 (bs, 4 H, H7, H11). IR (neat, cmꢀ1) 2923 (w), 2395 (w),
1596 (w), 1498 (w), 1457 (w), 1333 (m), 1306 (m), 1261 (m), 1154 (s),
1106 (m), 1091 (m), 1020 (m), 924 (w), 897 (m), 835 (m), 806 (s), 783
(m), 736 (s), 686 (s). MS m/z 598 (MH+). Anal. Calcd for
C31H39N3O5S2 HCl H2O: C, 57.08; H, 6.49; N, 6.44. Found: C, 57.31;
(8) Kim, Y. H.; Douvic, M.; Obitz, E.; Gniadecki, R.; Iversen, L.;
Osterborg, A.; Whittaker, S.; Illidge, T. M.; Schwarz, T.; Kaufmann, R.;
Cooper, K.; Knudsen, K. M.; Lisby, S.; Baadsgaard, O.; Knox, S. J.
Clinical efficacy of zanolimumab (HuMax-CD4): two phase 2 studies
in refractory cutaneous T-cell lymphoma. Blood 2007, 109 (11),
4655–4662.
(9) Dalgleish, A. G.; Beverly, P. C.; Clapham, P. R.; Crawford, D. H.;
Greaves, M. F.; Weiss, R. A. The CD4 (T4) antigen is an essential component
of the receptor for the AIDS retrovirus. Nature 1984, 312, 763–767.
(10) Klatzmann, D.; Champagne, E.; Chamaret, S.; Gruest, J.;
Guetard, D.; Hercend, T.; Gluckman, J. C.; Montagnier, L. T-lympho-
cyte T4 molecule behaves as the receptor for human retrovirus LAV.
Nature 1984, 312, 767–768.
3
3
H, 6.84; N, 6.40.
’ ASSOCIATED CONTENT
(11) Freeman, M. M.; Seaman, M. S.; Rits-Volloch, S.; Hong, X.;
Kao, C. Y.; Ho, D. D.; Chen, B. Crystal structure of HIV-1 primary
receptor CD4 in complex with a potent antiviral antibody. Structure
2010, 18 (12), 1632–1641.
(12) Matthias, L. J.; Azimi, I.; Tabrett, C. A.; Hogg, P. J. Reduced
monomeric CD4 is the preferred receptor for HIV. J. Biol. Chem. 2010,
285 (52), 40793–40799.
S
Supporting Information. Detailed synthetic procedures
b
and characterization data (1H and 13C NMR, IR, MS, and
combustion microanalysis) for compounds 12ꢀ39. This material
is available free of charge via the Internet at http://pubs.acs.org.
(13) Veiga, A. S.; Santos, N. C.; Castanho, M. A. R. B. An insight on
the leading HIV entry inhibitors. Recent Pat. Anti-Infect. Drug Discovery
2006, 1, 67–73.
(14) Estꢀe, J. A.; Telenti, A. HIV entry inhibitors. Lancet 2007,
370, 319–328.
’ AUTHOR INFORMATION
Corresponding Author
*Phone: 775-784-1842. Fax: 775-784-6804. E-mail: twb@unr.edu.
(15) DeClercq, E. The design of drugs for HIV and HCV. Nat. Rev.
Drug Discovery 2007, 6, 1001–1018.
’ ACKNOWLEDGMENT
(16) Qian, K.; Morris-Natschke, S. L.; Lee, K.-H. HIV entry in-
hibitors and their potential in HIV Therapy. Med. Res. Rev. 2008, 2009
(29), 369–393.
(17) Reimann, K. A.; Khunkhun, R.; Lin, W.; Gordon, W.; Fung, M.
A humanized, nondepleting anti-CD4 antibody that blocks virus entry
inhibits virus replication in rhesus monkeys chronically infected with
simian immunodeficiency virus. AIDS Res. Hum. Retroviruses 2002,
18, 747–755.
(18) Kuritzkes, D. R.; Jacobsen, J.; Powderly, W. G.; Godofsky, E.;
DeJesus, E.; Haas, F.; Reimann, K. A.; Larson, J. L.; Yarbough, P. O.;
Curt, V.; Shanahan, W. R., Jr. Antiretroviral activity of the anti-CD4
monoclonal antibody TNX-355 in patients infected with HIV type 1.
J. Infect. Dis. 2004, 189, 286–291.
This work was supported by the K.U. Leuven (GOA No. 10/
014 and No. PF/10/018), the FWO (No. G.485.08), the
CHAARM Project of the European Commission, and the
Dormeur Investment Service Ltd. We are grateful to B. Provin-
ciael, S. Claes, and E. Fonteyn for excellent technical assistance.
Funding from the National Science Foundation (Grnat CHE-
0521191) in support of NMR spectrometers in the Department
of Chemistry, University of Nevada, is also gratefully acknowledged.
’ ABBREVIATIONS USED
ACE-Cl, R-chloroethyl chloroformate; AIDS, acquired immune
deficiency syndrome; CADA, cyclotriazadisulfonamide; CD4,
cluster of differentiation 4; dan, 5-dimethylaminonaphthalene-
(19) Schols, D. HIV co-receptors as targets for antiviral therapy.
Curr. Top. Med. Chem. 2004, 4, 883–893.
(20) Kazmierski, W. M.; Gudmundsson, K. S.; Piscitelli, S. C. Small
Molecule CCR5 and CXCR4-Based Viral Entry Inhibitors for Anti-HIV
Therapy Currently in Development. In Annual Reports in Medicinal
Chemistry; Macor, J. E., Ed.; Academic Press: Amsterdam, 2007; Vol. 42,
pp 301ꢀ320.
1-yl; dppb, 1,4-bis(diphenylphosphino)butane; CHO CD4-YFP,
3
Chinese hamster ovary/CD4/yellow fluorescent protein; ER,
endoplasmic reticulum;HIV, human immunodeficiency virus; SIV,
simian immunodeficiency virus
(21) Conti, L.; Varano, B.; Gauzzi, M. C.; Matarrese, P.; Federico,
M.; Malorni, W.; Belardelli, F.; Gessani, S. Impairment of human
immunodeficiency virus type 1 (HIV-1) entry into Jurkat T cells by
constitutive expression of the HIV-1 vpr protein: role of CD4 down-
modulation. J. Virol. 2000, 74, 10207–10211.
’ REFERENCES
(1) Littman, D. R. The structure of the CD4 and CD8 genes. Annu.
Rev. Immunol. 1987, 5, 561–584.
5720
dx.doi.org/10.1021/jm2002603 |J. Med. Chem. 2011, 54, 5712–5721

Journal of Medicinal Chemistry
ARTICLE
(22) Nokta, M. A.; Li, X. D.; Nichols, J.; Mallen, M.; Pou, A.;
Asmuth, D.; Pollard, R. B. Chemokine/CD4 receptor density ratios
correlate with HIV replication in lymph node and peripheral blood
of HIV-infected individuals. AIDS Res. Hum. Retroviruses 2001, 15,
161–169.
(23) Xiang, J.; McLinden, J. H.; Rydze, R. A.; Chang, Q.; Kaufman,
T. M.; Klinzman, D.; Stapleton, J. T. Viruses within the flaviviridae
decrease CD4 expression and inhibit HIV replication in human CD4+
cells. J. Immunol. 2009, 183, 7860–7869.
(39) Antonopoulou, G.; Magrioti, V.; Stephens, D.; Constantinou-
Kokotou, V.; Dennis, E. A.; Kokotos, G. Synthesis of 2-oxoamides based
on sulfonamide analogs of γ-amino acids and their activity on phospho-
lipase A2. J. Pept. Sci. 2008, 14, 1111–1120.
(40) Saab, N. H.; West, E. E.; Bieszk, N. C.; Prews, C. V.; Mank,
A. R.; Casero, R. A., Jr.; Woster, P. M. Synthesis and evaluation of
unsymmetrically substituted polyamine analogues as modulators of
human spermidine/spermine-N-acetyltransferase (SSAT) and as poten-
tial antitumor agents. J. Med. Chem. 1993, 36, 2998–3004.
(24) Lisco, A.; Grivel, J. C.; Biancotto, A.; Vanpouille, C.; Origgi, F.;
Malnati, M. S.; Schols, D.; Lusso, P.; Margolis, L. B. Viral interactions in
human lymphoid tissue: human herpesvirus 7 suppresses the replication
of CCR5-tropic human immunodeficiency virus type 1 via CD4
modulation. J. Virol. 2007, 81, 708–717.
(25) Vermeire, K.; Bell, T. W.; Choi, H. J.; Jin, Q.; Samala, M. F.;
Sodoma, A.; De Clercq, E.; Schols, D. The anti-HIV potency of
cyclotriazadisulfonamide analogs is directly correlated with their ability
to down-modulate the CD4 receptor. Mol. Pharmacol. 2003, 63,
203–210.
(26) Beaumier, C. M.; Harris, L. D.; Goldstein, S.; Klatt, N. R.;
Whitted, S.; McGinty, J.; Apetrei, C.; Pandrea, I.; Hirsch, V. M.;
Brenchley, J. M. CD4 downregulation by memory CD4+ T cells
in vivo renders African green monkeys resistant to progressive SIVagm
infection. Nat. Med. 2009, 15, 879–885.
(27) Vermeire, K.; Zhang, Y.; Princen, K.; Hatse, S.; Samala, M.;
Dey, K.; Choi, H.-J.; Ahn, Y.; Sodoma, A.; Snoeck, R.; Andrei, G.;
De Clercq, E.; Bell, T. W.; Schols, D. CADA inhibits human immuno-
deficiency virus and human herpesvirus 7 replication by down-modula-
tion of the cellular CD4 receptor. Virology 2002, 302, 342–353.
(28) Vermeire, K.; Princen, K.; Hatse, S.; De Clercq, E.; Dey, K.;
Bell, T. W.; Schols, D. CADA, a novel CD4-targeted HIV inhibitor, is
synergistic with various anti-HIV drugs in vitro. AIDS Res. Hum. Retro-
viruses 2004, 18, 2115–2125.
(29) Vermeire, K.; Schols, D.; Bell, T. W.; Marsh, M. A small-
molecule inhibitor of CD4 protein translocation down-modulates the
primary HIV receptor. Manuscript under revision.
(30) Bell, T. W.; Anugu, S.; Bailey, P.; Catalano, V. J.; Dey, K.; Drew,
M. G.; Duffy, N. H.; Jin, Q.; Samala, M. F.; Sodoma, A.; Welch, W. H.;
Schols, D.; Vermeire, K. Synthesis and structureꢀactivity relationship
studies of CD4 down-modulating cyclotriazadisulfonamide (CADA)
analogues. J. Med. Chem. 2006, 49, 1291–1312.
(31) Vermeire, K.; Lisco, A.; Grivel, J.-C.; Scarbrough, E.; Dey, K.;
Duffy, N.; Margolis, L.; Bell, T. W.; Schols, D. Design and cellular kinetics of
dansyl-labeled CADA derivatives with anti-HIV and CD4 receptor down-
modulating activity. Biochem. Pharmacol. 2007, 74, 566–578.
(32) Richman, J.; Atkins, T. J. Nitrogen analogs of crown ethers.
J. Am. Chem. Soc. 1974, 96, 2268–2270.
(41) Yokoshima, S.; Ueda, T.; Kobayashi, S.; Sato, A.; Kuboyama, T.;
Tokuyama, H.; Fukuyama, T. Stereocontrolled total synthesis of (+)-
vinblastine. J. Am. Chem. Soc. 2002, 124, 2137–2139.
(42) Osborne, A. R.; Rapoport, T. A.; van den Berg, B. Protein
translocation by the Sec61/SecY channel. Annu. Rev. Cell Dev. Biol. 2005,
No. 21, 529–550.
(43) Garrison, J. L.; Kunkel, E. J.; Hegde, R. S.; Taunton, J. A
substrate-specific inhibitor of protein translocation into the endoplasmic
reticulum. Nature 2005, 436, 287–289.
(44) Besemer, J.; Harant, H.; Wang, S.; Oberhauser, B.; Marquardt,
K.; Foster, C. A.; Schreiner, E. P.; de Vries, J. E.; Dascher-Nadel, C.;
Lindley, I. J. D. Selective inhibition of cotranslational translocation of
vascular cell adhesion molecule. Nature 2005, 436, 290–293.
(45) Williams, D. H.; Stephens, E.; OBrien, D. P.; Zhou, M. Under-
standing noncovalent interactions: ligand binding energy and catalytic
efficiency from ligand-induced reductions in motion within receptors
and enzymes. Angew. Chem., Int. Ed. 2004, 43, 6596–6616.
(46) Whitesides, G. M.; Krishnamurthy, V. M. Designing ligands to
bind proteins. Q. Rev. Biophys. 2005, 38 (4), 385–395.
(47) Williams, D. H.; Stephens, E.; Zhou, M. Ligand binding energy
and catalytic efficiency from improved packing within receptors and
enzymes. J. Mol. Biol. 2003, 329, 389–399.
(48) Baum, B.; Muley, L.; Smolinski, M.; Heine, A.; Hangauer, D.;
Klebe, G. Non-additivity of functional group contributions in pro-
teinꢀligand binding: a comprehensive study by crystallography and
isothermal titration calorimetry. J. Mol. Biol. 2010, 397, 1042–1054.
(49) Hilgeroth, A.; Molnur, J.; De Clercq, E. Using molecular
symmetry to form new drugs: hydroxymethyl-substituted 3,9-diazate-
traasteranes as the first class of symmetric MDR modulators. Angew.
Chem., Int. Ed. 2002, 41, 2623–3625.
(50) Sanmartín, C.; Font, M.; Palop, J. A. Molecular symmetry: a
structural property frequently present in new cytotoxic and proapoptotic
drugs. Mini-Rev. Med. Chem. 2006, 6, 639–650.
(51) Lemm, J. A.; OBoyle, D., II; Liu, M.; Nower, P. T.; Colonno, R.;
Deshpande, M. S.; Snyder, L. B.; Martin, S. W.; St. Laurent, D. R.;
Serrano-Wu, M. H.; Romine, J. L.; Meanwell, N. A; Gao, M. Identifica-
tion of hepatitis C virus NS5A inhibitors. J. Virol. 2010, 84, 482–491.
(52) Gao, M.; Nettles, R. E.; Belema, M.; Snyder, L. B.; Nguyen,
V. N.; Fridell, R. A.; Serrano-Wu, M. H.; Langley, D. R.; Sun, J.-H.;
OBoyle, D. R., II; Lemm, J. A.; Wang, C.; Knipe, J. O.; Chien, C.;
Colonno, R. J.; Grasela, D. M.; Meanwell, N. A.; Hamann, L. G.
Chemical genetics strategy identifies an HCV NS5A inhibitor with a
potent clinical effect. Nature 2010, 465, 96–102.
(33) Atkins, T. J.; Richman, J. E.; Oettle, W. F. Macrocyclic Poly-
amines: 1,4,7,10,13,16-Hexaazacyclooctadecane.; Wiley: New York, 1988;
pp 652ꢀ662.
(34) Ali, R.; Anugu, S.; Khan, C.; Demillo, V. G.; Goulinet-Mateo, F.;
Lamprecht, K.; Scarbrough, E.; Bell, T. W. Palladium-catalyzed synthesis
of 1,5,9-triazacyclododecanes by macrocyclization of disulfonamides
with carbonate esters of 2-methylene-1,3-dihydroxypropane. Manu-
script in preparation.
(53) Bell, T. W. Drugs for hepatitis C: unlocking a new mechanism
of action. ChemMedChem 2010, 210, 1663–1665.
(35) Olofsen, R. A.; Abott, D. E. Tests of a piperidino mask for the
protection of functionalized carbon sites in multistep syntheses. J. Org.
Chem. 1984, 49, 2795–2799.
(36) Olofsen, R. A.; Martz., J. T.; Senet, J.-P.; Piteau, M.; Malfroot, T.
A new reagent for the selective, high-yield N-dealkylation of tertiary
amines: improved synthesis of naltrexone and nalbuphine. J. Org. Chem.
1984, 49, 2081–2082.
(37) Demillo, V. Unsymmetrical Cyclotriazadisulfonamide (CADA)
Compounds: Synthesis and CD4 Down-Modulation Potency. Ph.D.
Dissertation, University of Nevada, Reno, NV, 2011.
(38) Pokhvisneva, G. V.; Luk’yanov, O. A. Synthesis of 2-substituted
1-tosyl-3(1-tosyliminoalkyl)imidazolidines and hexahydropyrimidines.
Russ. Chem. Bull., Int. Ed. 2006, 55, 903–906.
5721
dx.doi.org/10.1021/jm2002603 |J. Med. Chem. 2011, 54, 5712–5721