Discovery of a Small Molecule PDI Inhibitor That Inhibits Reduction of HIV-1 Envelope Glycoprotein gp120

Article abstract of DOI:10.1021/cb100387r

Protein disulfide isomerase (PDI) is a promiscuous protein with multifunctional properties. PDI mediates proper protein folding by oxidation or isomerization and disrupts disulfide bonds by reduction. The entry of HIV-1 into cells is facilitated by the PDI-catalyzed reductive cleavage of disulfide bonds in gp120. PDI is regarded as a potential drug target because of its reduction activity. We screened a chemical library of natural products for PDI-specific inhibitors in a high-throughput fashion and identified the natural compound juniferdin as the most potent inhibitor of PDI. Derivatives of juniferdin were synthesized, with compound 13 showing inhibitory activities comparable to those of juniferdin but reduced cytotoxicity. Both juniferdin and compound 13 inhibited PDI reductase activity in a dose-dependent manner, with IC50 values of 156 and 167 nM, respectively. Our results also indicated that juniferdin and compound 13 exert their inhibitory activities specifically on PDI but do not significantly inhibit homologues of this protein family. Moreover, we found that both compounds can inhibit PDI-mediated reduction of HIV-1 envelope glycoprotein gp120.

Full text of DOI:10.1021/cb100387r

ARTICLES
pubs.acs.org/acschemicalbiology
Discovery of a Small Molecule PDI Inhibitor That Inhibits
Reduction of HIV-1 Envelope Glycoprotein gp120
Maola M. G. Khan,†,‡ Siro Simizu,†,§ Ngit Shin Lai,†,z Makoto Kawatani,†
Takeshi Shimizu,† and Hiroyuki Osada†,‡,*
†Chemical Library Validation Team, Chemical Biology Core Facility, Chemical Biology Department,
RIKEN Advanced Science Institute, Saitama, Japan
‡Graduate School of Science and Engineering, Saitama University, Saitama, Japan
§Department of Applied Chemistry, Faculty of Science and Technology, Keio University, Yokohama, Japan
S Supporting Information
b
ABSTRACT: Protein dis-
ulfide isomerase (PDI) is a
promiscuous protein with
multifunctional properties.
PDI mediates proper pro-
tein folding by oxidation
or isomerization and dis-
rupts disulfide bonds by re-
duction. The entry of HIV-1
into cells is facilitated by the
PDI-catalyzed reductive
cleavage of disulfide bonds
in gp120. PDI is regarded
as a potential drug target
because of its reduction
activity. We screened a
chemical library of natural
products for PDI-specific inhibitors in a high-throughput fashion and identified the natural compound juniferdin as the most potent
inhibitor of PDI. Derivatives of juniferdin were synthesized, with compound 13 showing inhibitory activities comparable to those of
juniferdin but reduced cytotoxicity. Both juniferdin and compound 13 inhibited PDI reductase activity in a dose-dependent manner,
with IC50 values of 156 and 167 nM, respectively. Our results also indicated that juniferdin and compound 13 exert their inhibitory
activities specifically on PDI but do not significantly inhibit homologues of this protein family. Moreover, we found that both
compounds can inhibit PDI-mediated reduction of HIV-1 envelope glycoprotein gp120.
rotein disulfide isomerase (PDI) is a 57-kDa oxidoreductase
of the thioredoxin superfamily that is expressed mainly in the
membrane reductive status, and nitric oxide signaling in mega-
karyocytes and platelets.8
P
endoplasmic reticulum (ER) of eukaryotic cells. In the ER, PDI
acts predominantly as an oxidase to form disulfide bonds in
nascent secretory proteins. Furthermore, it catalyzes the rear-
rangement of incorrect disulfide bonds through isomerase activ-
ity, thus mediating proper protein folding.1,2 PDI is also
expressed on the cell surfaces of lymphocytes, hepatocytes, and
platelets3,4 and, together with additional ER secretory proteins,
can be secreted to the cell surface through a regulatory pathway.
In endosomes and on the cell surface, PDI acts as a reductase to
cleave disulfide bonds in proteins.5-7 Surface-associated PDI is
involved in a wide range of functions, including the reduction of
disulfide bonds in diphtheria toxin, facilitating the entry of the
toxin into cells, the reduction of disulfide bonds in the ectodo-
main of human thyrotropin receptor, the maintenance of plasma
The functions of PDI are regulated by the redox state of its
active site CXXC motif, Cys-Gly-His-Cys in PDI. In this motif,
the two amino acids lying between the cysteine residues are
important in maintaining its redox potential. All PDI-catalyzed
reactions occur through thiol-disulfide exchanges between PDI
and neighboring proteins or peptides. During oxidation, PDI
with an oxidized CXXC motif reacts with the substrate dithiols of
proteins or peptides to form disulfide bonds. During reduction,
however, PDI with a cysteinyl dithiol CXXC motif reduces the
Received: August 18, 2010
Accepted: December 1, 2010
Published: December 01, 2010
r
2010 American Chemical Society
245
dx.doi.org/10.1021/cb100387r ACS Chem. Biol. 2011, 6, 245–251
|

ACS Chemical Biology
ARTICLES
disulfide bonds of substrate proteins or peptides, resulting in an
active-site disulfide in PDI.9
The reductase activity of surface-associated PDI is of immense
importance because of its involvement in HIV-1 fusogenic
events.10-14 PDI clusters in the vicinity of the HIV-1 primary
receptor CD4, which has separate binding sites for PDI and the
HIV-1 envelope glycoprotein gp120.15 The gp120 of HIV-1
binds to CD4, followed by the PDI-catalyzed reduction of at
least two of the nine disulfide bonds of gp120. Reduction of the
structure-stabilizing disulfide bonds of gp120 leads to major
conformational changes, enhancing the ability of gp120 to
interact with its co-receptors CCR5 and CXCR4.10,16 Subse-
quently, gp41, a viral glycoprotein that is noncovalently bound to
gp120, undergoes rearrangement into its fusogenic intermediates
and HIV-1 enters into the host cell.13,16 We have sought to
therapeutically target the PDI-mediated rearrangement of gp120
using small molecule inhibitors of PDI.
Due to the ubiquitous role of PDI, it is difficult to target a
specific function of the enzyme. Potential therapeutic agents
must be unable to permeate cell membranes to specifically inhibit
the reductase activity of PDI on the cell surface. If the inhibitor is
membrane-permeable, it will inhibit the oxidase function of PDI,
which is important for the proper protein folding. Several agents,
such as bacitracin, DTNB, pCMBS, PAO, aPAO, GSAO, diethyl-
stilbesterol, and estrone, have been described as PDI inhibitors,
although none has been used therapeutically.10,15,17 Bacitracin
was one of the first agents used as a PDI inhibitor, but its clinical
use is limited by nephrotoxicity,18 low membrane permeability,19
and PDI-independent effects.20 Low membrane permeability of
bacitracin is a limitation if its target is to block the oxidase activity
of PDI that is catalyzed in the ER. However, the reductive
function of PDI is not affected by membrane permeability as it
occurs mainly on the cell surface. DTNB and pCMBS react
nonspecifically with free thiols and are not specific for PDI. The
compounds aPAO and GSAO are blockers of the CXXC vicinal
cysteines, which are present in many PDI family members as well
as other proteins, making them highly toxic.21
In this study we employed high-throughput screening (HTS)
of RIKEN Natural Product Depository (NPDepo) compounds
to identify therapeutically potent PDI inhibitors. We synthesized
a series of derivatives of juniferdin, the most potent inhibitor
found in NPDepo, and studied the structure-activity relation-
ships of the derivatives. Furthermore, we examined the specificity
of the identified PDI inhibitors in terms of their function and
homologous protein targets and investigated the potential use of
these inhibitors in blocking the PDI-catalyzed reduction of
disulfide bonds in the HIV-1 envelope glycoprotein gp120, with
a view to using these compounds as inhibitors of HIV-1 entry.
Figure 1. Screening for PDI inhibitors. (a) Structure of the four leading
PDI inhibitors: (i) nordihydroguiaretic acid, NDGA; (ii) diethylstilbesterol,
DES; (iii) estrone; and (iv) juniferdin. (b) Effect of putative inhibitors on
PDI reductase activity. Compounds were assayed at a final concentration of
10 μM, with juniferdin showing the highest inhibitory activity.
nordihydroguiaretic acid, and juniferdin (Figure 1, panel a).
Estrone is a naturally occurring estrogen secreted by ovaries
during menopause,23 and diethylstilbestrol is an orally active
synthetic nonsteroidal estrogen.24 Both estrone and diethylstil-
besterol were found to inhibit the reduction activity of PDI.17
Although their mechanism of inhibition is still obscure, PDI
contains binding sites for estrogens and can act as a reservoir for
hormones.25 Thus, inhibition of PDI reductase activity by estrogens
is likely to occur through the binding of these hormones. The
putative PDI inhibitor nordihydroguiaretic acid has been described
as a potent antioxidant that may reduce cell damage 26 and inhibit
membrane trafficking.27 Juniferdin, a sesquiterpenoid isolated from
the plant Ferula junipernia, has been reported to possess estrogenic
activity.28 Hence, we compared the inhibitory potency of juniferdin
and previously described estrogenic compounds, 17-β estradiol and
estrone, 17 by the insulin reduction assay. Our results suggest both
17-β estradiol and estrone are weaker inhibitors than juniferdin
(Supplementary Figure 1, panel a and panel b). Though at higher
concentration (10 μM) they show almost similar inhibition, at
lower concentrations (1 and 0.1 μM) estrogenic compounds are
much less potent. We also checked the effect of juniferdin and 17-β
estradiol on MCF-7 proliferation. We found juniferdin has a very
low stimulatory effect on MCF-7 proliferation compared to 17-β
estradiol (Supplementary Figure 2). Moreover, a comparison of the
four PDI hit inhibitors, each at a concentration of 10 μM, showed
that juniferdin was the most potent (Figure 1, panel b). We
therefore selected juniferdin as the lead PDI inhibitor for further
analyses.
’ RESULTS AND DISCUSSION
Identification of Juniferdin as a PDI Inhibitor. We sought
to identify small molecules that would potently and specifically
block the reductase activity of cell surface PDI. We therefore
screened the compounds of NPDepo using a high-throughput
turbidometric assay system.22 PDI catalyzes the reduction of
insulin in the presence of dithiothreitol (DTT); the reduced
insulin chains aggregate, and turbidometry is monitored spectro-
photometrically at 620 nm.
Synthesis and Structure-Activity Relationship of Junifer-
din Derivatives. The potent PDI-inhibitory activity of junifer-
din prompted us to synthesize its derivatives to generate
inhibitors with improved potency and to examine structure-
activity relationships (SAR). The general approach we used
to synthesize these derivatives is shown in Supplementary
We screened about 10,000 NPDepo compounds in a high-
throughput manner. After three rounds of screening, we identi-
fied four putative PDI inhibitors: estrone, diethylstilbestrol,
246
dx.doi.org/10.1021/cb100387r |ACS Chem. Biol. 2011, 6, 245–251

ACS Chemical Biology
ARTICLES
Table 1. Potency (IC50) of PDI Inhibition by Juniferdin and
Its Derivatives 2-11
compound
R1
R2
R3
IC50 (μM)
1: juniferdin
OH
OMe
OH
OAc
OAc
H
OH
OH
OAc
OH
OAc
OH
OH
0.156
>10
2
3: juniferidin
0.453
0.627
0.693
4.96
4
5
6
7
F
0.958
>10
8
1-octyl
9
cyclooctyl
2.85
10
11
cyclododecyl
(1R)-menthyl
0.565
0.667
Table 2. Potency (IC50) of PDI Inhibition by Juniferdin
Derivatives 12-17
Figure 2. Dose-dependent inhibition of PDI by (a) juniferdin and (b)
compound 13. PDI reductive activity was measured in the presence or
absence of 10, 1, 0.1, or 0.01 μM juniferdin or compound 13 in a 100-μL
reaction volume containing 1.3 μM PDI, 0.5 mM DTT, 0.13 mM insulin,
100 mM sodium phosphate buffer, 2 mM EDTA. (c) IC50 curves for
juniferdin, compound 13, and compound 2. The IC50 values of
juniferdin and compound 13 were 156 and 167 nM, respectively.
found that the octyl 4-hydroxybenzoate (8) had no inhibitory
activity, whereas the inhibitory activity of the cyclooctyl deriva-
tive (9) was 18-fold lower than that of 1. Interestingly, derivatives
with substituted cyclododecyl (10) and (1R)-menthyl (11)
groups showed good inhibitory activity.
The most promising results came from SAR investigations of
compounds with a modified sesquiterpene ring (Table 2). Juni-
ferol, a dihydroxy derivative with a sesquiterpene 11-membered
ring system, showed no inhibitory activity. The 9,10-monoepoxide
(13) 1:1 stereoisomers showed inhibition similar to that of 1,
whereas the 2,3-monoepoxide (14) showed 15-fold lower inhibi-
tion, and the 2,3,9,10-diepoxide (15) was detrimental to activity. Of
the triols in the sesquiterpene ring, the 4,9,10-trihydroxy derivative
(16) showed no inhibition, whereas the stereoisomeric 4,9,10-
trihydroxy derivative (17) showed modest inhibition. Thus these
SAR studies revealed that both benzoate moiety and cyclic alcohol
were indispensable and a p-hydroxybenzoate group was important
for strong inhibition of PDI activity. It was also interesting that the
hydroxyl phenyl group was present on all four putative inhibitors of
PDI (Figure 1, panel a). In summary, compound 13 was the most
potent inhibitor of the derivatives we synthesized, whereas com-
pound 2 showed the least inhibition.
Scheme 1. We focused on derivatizing its R1 (the para position of
the benzoate), R2 (the C-4 hydroxyl group of the sesquiterpene
ring), and R3 (the sesquiterpene ring) groups. The structures of
these compounds are shown in Supplementary Table 1.
We found that the p-methoxybenzoate (2) derivative com-
pletely abolished the PDI-inhibitory activity of juniferdin, the
p-fluoro derivative (7) reduced its inhibitory activity 6-fold, and
the benzoate (6) without a p-hydroxyl group reduced its activity
30-fold (Table 1). In contrast, the monoacetates (3, 4) and the
diacetate (5) showed moderate activity (3- to 4-fold less than 1).
In investigating the importance of the sesquiterpene ring, we
247
dx.doi.org/10.1021/cb100387r |ACS Chem. Biol. 2011, 6, 245–251

ACS Chemical Biology
ARTICLES
Figure 4. Inhibition of PDI-mediated reduction of HIV-1III-B gp120 by
juniferdin and compound 13. ELISA plates were coated with recombi-
nant gp120 and incubated for 1-3 min with PBST, 0.35 μM PDI,
0.5 mM DTT, and 1 μM compounds. Free-SH groups were labeled by
NEM-biotin and exposed to streptavidine-ALP. All reactions were
normalized relative to DTT at each time point.
Figure 3. Specificity of juniferdin, compound 2, and compound 13.
Effects of 10 μM juniferdin, compound 2, and compound 13 on (a) PDI
oxidase activity, measured by the continuous RNase refolding assay, and
(b) PDI, ERp57, and ERp72 reductase activity, measured by the insulin
turbidometric assay method.
significant inhibition of PDI oxidase activity (Supplementary
Figure 3), thus confirming the reaction specificity.
To investigate the target specificity of juniferdin and com-
pound 13, they were tested against proteins similar to PDI in
structure and activity. PDI is a 17-member protein family that
belongs to the thioredoxin superfamily. In the PDI family, ERp57
and ERp72 show reductase activity similar to that of PDI and
share the same active site sequence, CGHC. Although PDI and
ERp57 each have two active sites, ERp72 contains three active
sites.30 We investigated whether juniferdin and compound 13
were specific inhibitors of PDI or if they could also inhibit the
reduction activity of ERp57 and ERp72, as determined using
insulin turbidometry assays. We found that, at 10 μM, both
juniferdin and compound 13 showed negligible inhibition
(∼10%) of ERp57 and ERp72 reduction activity (Figure 3, panel
b). Furthermore these PDI members contain thioredoxin-like
domains; we therefore evaluated whether juniferdin could inhibit
thioredoxin mediated insulin reduction. We found that juniferdin
did not inhibit thioredoxin reductase activity (Supplementary
Figure 4). In conclusion our results show that juniferdin and
compound 13 specifically inhibit the reduction activity of PDI.
PDI enzymatic activity depends on domain architecture and
the presence or absence of additional residues that may regulate
the pKa of the active site cysteines. PDI consists of four domains,
a, b, b0, and a0, with a linker region between b0 and a0 and a
C-terminal acidic region (Supplementary Figure 5). The a and a0
domains are catalytically active, while the b0 domain functions as
a peptide binding domain. Either a or a0 is required for oxidase
activity, and the linear combination of one of these with the
substrate binding domain b0 is required for simple isomerization.9
In contrast, all four of these domains are needed for complex
isomerization, causing major conformational changes, which can be
achieved through cycles of reduction and oxidation. Moreover, the
residues that influence the pKa values of the active site cysteines also
regulate PDI functions. The exact reasons behind the specificity of
juniferdin and compound 13 to PDI reductase activity are still
obscure. Presumably these inhibitors bind to specific residues or
domains that are required only for PDI reductase activity.
Dose-Dependent Inhibition of PDI Activity by Juniferdin
and Compound 13. When we assayed the dose-dependency of
juniferdin on PDI reductase activity at 10, 1, 0.1, or 0.01 μM,
using the insulin turbidometry assay, we found that juniferdin
inhibited PDI reductase dose-dependently (Figure 2, panel a)
and at 10 μM, juniferdin inhibited approximately 98% of the PDI
reductase. In addition, compound 13 showed dose-dependent
inhibition of PDI (Figure 2, panel b). Dose-response curves
showed that the IC50 values for juniferdin and compound 13
were 156 and 167 nM, respectively (Figure 2, panel c).
To date, juniferdin is the most potent inhibitor of PDI
reduction activity. Of the previously described PDI inhibitors,
bacitracin has been found to inhibit ∼95% PDI activity at
nonphysiological concentrations (3 mM), whereas the most
potent estrogen (17-β estradiol) showed 40-60% inhibition at
1 μM, and membrane-impermeable sulfhydryl blockers DTNB
and pCMBS completely inhibited PDI at 1 mM.5,15,17 Since
juniferdin and compound 13 exerted ∼98% PDI inhibition at
10 μM concentrations, they are 10- to 300-fold more potent than
the previously mentioned inhibitors (Supplementary Table 2).
Moreover, juniferdin and compound 13 inhibited only PDI,
whereas bacitracin, DTNB, and pCMBS inhibited many sulfhy-
dryl-group-containing enzymes.
Specific Inhibition of PDI Reductase Activity by Juniferdin
and Compound 13. To confirm the functional specificity of
juniferdin and compound 13, we examined their effects on the
oxidase activity of PDI. The oxidase activity was monitored by
measuring the continuous refolding of denatured RNase.29 PDI
catalyzes the oxidative renaturation of reduced, denatured RNase
in a glutathione redox buffer. The oxidized RNase hydrolyzes the
substrate cCMP to CMP, which can be measured at 284 nm. We
found that, at final concentrations of 10 μM, neither juniferdin
nor compound 13 had an effect on the oxidative folding activity
of PDI (Figure 3, panel a). We also measured the PDI-mediated
oxidation of reduced lysozyme. According to this assay, PDI
reactivates lysozyme in a glutathione redox buffer and in turn
lysozyme clears the suspension of Micrococcus lysodeikticus cell
wall. We found that juniferdin neither at 1 nor 10 μM showed
Juniferdin and Compound 13 Inhibit the PDI-Catalyzed
Reduction of gp120. Since the PDI-catalyzed reduction of
disulfides in gp120 has been shown to be crucial for HIV-1 entry
into cells,10,11,15 we examined whether juniferdin and compound
13 could inhibit the PDI-catalyzed reduction of gp120. HIV-1
gp120 was fixed onto ELISA plates, which were incubated with
248
dx.doi.org/10.1021/cb100387r |ACS Chem. Biol. 2011, 6, 245–251

ACS Chemical Biology
ARTICLES
the potential inhibitor and DTT in the presence or absence of
PDI for a fixed time, and NEM-biotin was used to specifically
react with the reduced thiols produced by the PDI. Increased
time-dependent formation of free thiols by PDI was observed, in
accordance with published results.31 However, we found that
both juniferdin and compound 13 significantly inhibited the
PDI-catalyzed reduction of gp120, while compound 2 showed no
inhibition (Figure 4). In conclusion, juniferdin and compound
13 inhibit the PDI-catalyzed reduction of HIV gp120, showing
that they may inhibit viral entry.
Several anti-PDI agents, such as DTNB, bacitracin, PAO, and anti-
PDI antibodies, have been used to inhibit the PDI-mediated reduc-
tion of gp120 and subsequent HIV infection. DTNB, bacitracin, and
anti-PDI antibody have been found to markedly inhibit HIV infection
of human lymphocytes,15 as have aPAO and AT3.10 Although initially
showing promising results, these agents could not be used therapeu-
tically due to their nonspecific activities and toxicities. The two
compounds we identified, juniferdin and compound 13, showed
specific inhibition of PDI reductase activity and did not inhibit the
other homologues of PDI. Hence, their ability to inhibit the PDI-
catalyzed reduction of HIV gp120 suggests that juniferdin and
compound 13 may be promising as inhibitors of viral entry.
cDNA was subcloned into the pcDNA3.1/Myc-His (þ) vector
(Invitrogen). A stable cell line expressing PDI was established by
transfecting pcDNA3.1-PDI into HT1080 cells using Effectence Trans-
fection Reagent (QIAGEN), followed by selection with 400 μg/mL
G418. Cells transfected with pcDNA3.1-PDI were designated HT1080-
32
PDI-MH cells. Human ERp57 and ERp72 cDNA
without ER
retention sequence (QEDL and KEEL, respectively) were subcloned
into the pcDNA3.1/Myc-His (þ) vector. Stable cell lines prepared by
transfecting HT1080 cells with clones plasmids were designated
HT1080-ERp57-MH and HT1080-ERp72-MH cells, respectively.
Purification of PDI, ERp57, and ERp72. Recombinant PDI,
ERp57, and ERp72 were purified as described previously 33 with slight
modifications. Briefly, cells overexpressing PDI, ERp57, or ERp72 were
grown to confluence, washed three times with serum-free DMEM, and
incubated in serum-free media for 24 h. The conditioned media were
collected and incubated with Ni-NTA agarose (Qiagen) for 2 h at 4 °C.
The Ni-NTA agarose was washed thrice with sodium phosphate buffer,
and Ni-NTA agarose-bound PDI, ERp57, or ERp72 was eluted with 500
mmol/L imidazole in sodium phosphate buffer and electrophoresed on
SDS-polyacrylamide gels. The gels were stained with Coomassie Bril-
liant Blue G-250. Western blotting with anti-Myc (sc-40, Santa Cruz
Biotechnology) antibody was performed as described.33
Measurement of PDI Reductase Activity. Reductase activity
was assayed by measuring the PDI-catalyzed reduction of insulin in the
presence of DTT, thus measuring the aggregation of reduced insulin
chains at 650 nm.34 The incubation mixture contained 100 mM sodium
phosphate buffer pH 7.0, 0.5 mM DTT, 2 mM EDTA, 0.13 mM bovine
insulin (Sigma), and 1.3 μM purified human PDI.
Since these compounds showed promising therapeutic value, we
investigated whether they possess any cytotoxic activity. Using the
WST-8 assay, we measured the effects of juniferdin and compound
13 on the viability of HeLa, HepG2, HL-60, HT1080, and K562
cell lines. We found that compound 13 was less cytotoxic than
juniferdin against all cell lines (Supplementary Table 3).
High-Throughput Screening for PDI Inhibitors. A high-
throughput screening (HTS) system was used to search for PDI
inhibitors. For HTS, 8664 compounds at four concentrations each,
3.33, 1.111, 0.33, and 0.11 mg/mL, were collected from NPDepo,
RIKEN 35 in 96-well plates. To each well was added a reaction mixture
containing 2 mM EDTA, 0.13 mM insulin, 10 μM compounds, 1.3 μM
PD,I and 0.5 mM DTT in 100 mM sodium phosphate buffer, using a
Biomek 2000 liquid handling system (Beckman). After about 30 min,
absorption was measured spectrophotometrically at 650 nm (Wallac
1420 ARVO, PerkinElmer), for 2 h at 5 min intervals. Following the first
step of HTS, using a robotic system, 201 compounds were selected as
primary hits. The second step was performed manually to minimize
machine error and artifacts. In the first and second steps, compounds
were assayed at 0.33 mg/mL. After the second step 24 compounds were
selected for the next round. In the third step compounds of different
concentrations were used, and finally four putative PDI inhibitors were
identified.
Measurement of PDI Oxidase Activity. Reduced and dena-
tured RNase A was prepared as described.36 Ten milligrams of native
bovine RNase A (Sigma) was incubated in 2 mL of 6 M guanidine HCl in
0.1 M Tris-acetate (pH8.0), 2 mM EDTA, and 0.15 M DTT. The
reduced and denatured RNase A was separated from DTT and
guanidine-HCl using Sephadex G-25 equilibrated with 0.1% (v/v) acetic
acid. PDI oxidase activity was measured using a continuous RNaseA
refolding assay,29,37 in which PDI catalyzes the oxidative renaturation of
RNase A in a glutathione redox buffer, and the oxidized RNase A then
hydrolyzes the substrate cytidine 20-30-cyclin monophosphate (cCMP),
monitored spectrophotometrically at 284 nm. Typical assays contained
100 mM sodium phosphate buffer, pH 7.5, 2 mM GSH, 0.2 mM GSSG,
8 μM reduced RNaseA, 1.4 μM PDI, and 4.5 mM cCMP. The reaction
mixture was equilibrated at 25 °C for 15 min before adding the substrate
cCMP. Absorbance was measured at 284 nm for 1 h at 5 min intervals.
Synthesis of Juniferdin Derivatives. All compounds were
prepared using the procedure shown in Supplementary Scheme 1. All
moisture-sensitive reactions were performed under a nitrogen atmo-
sphere with dehydrated solvents under anhydrous conditions. Juniferdin
Summary and Implications. Using HTS we identified juni-
ferdin as a novel small molecule PDI inhibitor. Using juniferdin
as a lead compound, we synthesized a series of derivatives, with
SAR assays showing that compound 13 had inhibitory activity
similar to that of juniferdin. Both inhibitors specifically blocked
the reductase activity of PDI, a selective feature that can be
therapeutically harnessed in situations where reduction-depen-
dent protein conformational changes occur on the cell surface.
For example, PDI-catalyzed reduction of disulfide bonds in
gp120 is a prerequisite for HIV-1 entry into host cells (Sup-
plementary Figure 6). We found that the PDI inhibitors juni-
ferdin and compound 13 inhibited the PDI-catalyzed reduction
of the disulfide bonds in gp120. These specific PDI inhibitors
may be promising antiviral therapeutic agents. Moreover, by
inhibiting viral entry into cells, these inhibitors may act at the
starting point of a disease cycle, thereby quenching disease
progression at the earliest possible stages. Future confirmation
of the antiviral properties in vivo of these compounds may
provide a solid basis for their therapeutic development.
The promiscuity of PDI and its activity in reducing disulfide
bonds make this enzyme a prime target for therapeutic agents.
Reduction of disulfide bonds by surface-associated PDI is important
for the entry of virus and toxins into the cells. For example, the
cholera and diphtheria toxins contain disulfide bonds cleaved by
PDI, thus facilitating the passage of the toxins. Functionally specific
PDI inhibitors such as juniferdin and compound 13 may therefore
largely prevent the maturation of toxins that are dependent on
disulfide bond cleavage. Thus, selective inhibitors of PDI are likely to
find use in a broad range of therapeutic applications.
’ METHODS
Construction of Plasmids and Establishment of Stable Cell
Lines. The ER retention signal sequence (KDEL)-deleted human PDI
249
dx.doi.org/10.1021/cb100387r |ACS Chem. Biol. 2011, 6, 245–251

ACS Chemical Biology
ARTICLES
was purchased from Namiki Shoji Co. Ltd. All reagents and solvents
were purchased from commercial suppliers and used without further
purification. All reactions were monitored by thin-layer chromatography
(TLC) on 0.25 mm Merck silica gel plates (60 F254) using UV light, a
color reaction with 10% ethanolic phosphomolybdic acid, and heat as
detecting methods. All compounds were purified by preparative TLC
using 0.5 mm Merck silica gel plates (60 F254) or by HPLC using Sensyu
Pak PEGASIL Silica 120-5 (200 mm ꢀ 250 mm) (Sensyu Scientific Co.,
Ltd.). The structure and purity of all compounds were confirmed by 1H
NMR spectra recorded on a JEOL JNM-ECP-500 MHz spectrometer in
CDCl3. Treatment of juniferdin (1) with acetic anhydride-pyridine in
CH2Cl2 at RT resulted in the monoacetate 4 as the major product, along
with the monoacetate 3 (juniferidin) and diacetate 5. The p-methoxy
benzoate 2 was synthesized by methylation of 1 with trimethylsilyldia-
zomethane (TMSCHN2). Juniferdin was hydrolyzed with 1 N NaOH
in MeOH at 60 °C to yield juniferol 12. Oxidation of 1 with
m-chloroperbenzoic acid (MCPBA) yielded the 9,10-monoepoxide 13
(1:1 stereoisomers) and the diepoxide 15 (4:1 stereoisomers), whereas
oxidation of 1 with tert-butylhydroperoxide in the presence of vanadyl
acetyacetonate yielded the 2,3-monoepoxide 14 (single stereoisomer)
and the diepoxide 15 (4:1 stereoisomers). Oxidation of 1 with osmium
tetroxide yielded the triol 16 as the major product together with a small
amount of the stereoisomer 17. Benzoate 6 and p-fluorobenzoate 7 were
prepared by esterification of jufiferol 12 with the corresponding acid
chloride, and the p-hydroxybenzoates 8-11 were prepared by ester-
ification of p-hydroxybenzoic acid with the corresponding alcohol and
N-(3-dimethylaminopropyl)-N0-ethylcarbodiimide (EDCI). The struc-
ture of each compound was confirmed by examination of spectra data.
The spectra data have been included in Supporting Information.
Inhibition of PDI-Mediated Reduction of HIV-1 gp120.
PDI-mediated reduction of HIV-1IIIB gp120 was assayed as described,
with slight modifications.31 Recombinant HIV-1IIIB gp120 produced in
CHO cells was purchased from ImmunoDiagnostics, and 96-well Maxi-
Sorp ELISA plates (Nunc) were coated with 4 μg/mL gp120 diluted in
carbonate buffer, pH 9.6 at 4 °C overnight. The plates were washed three
times with PBST (phosphate -buffered saline, pH 7.4, plus 0.05% Tween-20)
and the PDI-mediated reduction of gp120 was determined by incubating the
wells with PBST in the presence or absence of 0.35 μM PDI, 0.5 mM DTT,
and 10 μM compounds for 1-3 min. The plates were washed four times with
PBST and incubated for 1 h at RT with 2 μM NEM-Biotin (Sigma), which
reacts specifically with thiol groups. The wells were again washed four times
with PBST and incubated for 1 h at RT with alkaline phosphatase conjugated
streptavidine (MABTECH). After 4 washes with PBST, the plates were
incubated in the dark for 15 min with p-nitrophenyl phosphate (Sigma), and
absorbance was measured at 405 nm using Wallac 1420 ARVO.
Present Addresses
zInstitute for Research in Molecular Medicine (INFORMM),
Universiti Sains Malaysia, 11800 Minden, Pulau Pinang, Malaysia.
’ ACKNOWLEDGMENT
We would like to thank K. Wierzba, S. Takahashi, T. Saito, and
Y. Kondoh for their valuable suggestions; R. Nakazawa for DNA
sequencing (RIKEN); and E. Sanada and Y. Fukushima for
technical assistance. This study was supported in part by a
Grant-in-Aid from the Ministry of Education, Culture, Sports,
Science, and Technology of Japan (MEXT).
’ ABBREVIATIONS USED
PDI,protein disulfide isomerase; HIV,human immunodeficiency
virus; gp,glycoprotein; DTNB,5,50-dithiobis(2-nitorbenzoic
acid); PAO,phenylarsine oxide; aPAO,p-amino PAO; GSAO,
4-(N-(S-glutathionylacetyl)aminophenylarsenoxide; pCMBS,
p-chloromercuriphenylsulfonate
’ REFERENCES
(1) Freedman, R. B. (1989) Protein disulfide isomerase: multiple roles in
the modification of nascent secretory proteins. Cell 57, 1069–1072.
(2) Frand, A. R., Cuozzo, J. W., and Kaiser, C. A. (2000) Pathways
for protein disulphide bond formation. Trends Cell Biol. 10, 203–210.
(3) Tager, M., Kroning, H., Thiel, U., and Ansorge, S. (1997)
Membrane-bound proteindisulfide isomerase (PDI) is involved in
regulation of surface expression of thiols and drug sensitivity of
B-CLL cells. Exp. Hematol. 25, 601–607.
(4) Jordan, P. A., and Gibbins, J. M. (2006) Extracellular disulfide
exchange and the regulation of cellular function. Antioxid. Redox
Signaling 8, 312–324.
(5) Mandel, R., Ryser, H. J., Ghani, F., Wu, M., and Peak, D. (1993)
Inhibition of a reductive function of the plasma membrane by bacitracin
and antibodies against protein disulfide-isomerase. Proc. Natl. Acad. Sci.
U.S.A. 90, 4112–4116.
(6) Zhang, L. M., St., Croix, C., Cao, R., Wasserloos, K., Watkins,
S. C., Stevens, T., Li, S., Tyurin, V., Kagan, V. E., and Pitt, B. R. (2006)
Cell-surface protein disulfide isomerase is required for transnitrosation
of metallothionein by S-nitroso-albumin in intact rat pulmonary vascular
endothelial cells. Exp. Biol. Med. (Maywood, NJ, U. S.) 231, 1507–1515.
(7) Ramachandran, N., Root, P., Jiang, X. M., Hogg, P. J., and Mutus,
B. (2001) Mechanism of transfer of NO from extracellular S-nitro-
sothiols into the cytosol by cell-surface protein disulfide isomerase. Proc.
Natl. Acad. Sci. U.S.A. 98, 9539–9544.
(8) Turano, C., Coppari, S., Altieri, F., and Ferraro, A. (2002)
Proteins of the PDI family: unpredicted non-ER locations and functions.
J. Cell Physiol. 193, 154–163.
Statistical Analysis. PDI activity in the presence of compound was
determined by the formula, PDI activity (%) = (OD[PDIþDTTþcompound]
-
ODDTT)/(OD[PDIþDTT] - ODDTT) ꢀ 100. OD values were taken from
the exponential phase (95 min) of the reaction. For the calculation of IC50
values, compounds from 10 μM to 0.1 nM were used. The IC50 curves were
analyzed by using the equation “log(inhibitor) vs normalized response-
variable slope” of nonlinear regression to determine the IC50 values. The
errors bars indicate mean ( SE of three replicates. All graphs were drawn
using GraphPad Prism 5 software.
(9) Freedman, R. B., Klappa, P., and Ruddock, L. W. (2002) Protein
disulfide isomerases exploit synergy between catalytic and specific
binding domains. EMBO Rep. 3, 136–140.
(10) Gallina, A., Hanley, T. M., Mandel, R., Trahey, M., Broder, C. C.,
Viglianti, G. A., and Ryser, H. J. (2002) Inhibitors of protein-disulfide
isomerase prevent cleavage of disulfide bonds in receptor-bound glyco-
protein 120 and prevent HIV-1 entry. J. Biol. Chem. 277, 50579–50588.
(11) Barbouche, R., Miquelis, R., Jones, I. M., and Fenouillet, E.
(2003) Protein-disulfide isomerase-mediated reduction of two disulfide
bonds of HIV envelope glycoprotein 120 occurs post-CXCR4 binding
and is required for fusion. J. Biol. Chem. 278, 3131–3136.
(12) Markovic, I., Stantchev, T. S., Fields, K. H., Tiffany, L. J., Tomic,
M., Weiss, C. D., Broder, C. C., Strebel, K., and Clouse, K. A. (2004) Thiol/
disulfide exchange is a prerequisite for CXCR4-tropic HIV-1 envelope-
mediated T-cell fusion during viral entry. Blood 103, 1586–1594.
(13) Berger, E. A., Murphy, P. M., and Farber, J. M. (1999)
Chemokine receptors as HIV-1 coreceptors: roles in viral entry, tropism,
and disease. Annu. Rev. Immunol. 17, 657–700.
’ ASSOCIATED CONTENT
S
Supporting Information. This material is available free
b
of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*Phone: 81-48-467-9542. Fax: 81-48-462-4669. E-mail: hisyo@
riken.jp.
250
dx.doi.org/10.1021/cb100387r |ACS Chem. Biol. 2011, 6, 245–251

ACS Chemical Biology
ARTICLES
(14) Pierson, T. C., Doms, R. W., and Pohlmann, S. (2004)
Prospects of HIV-1 entry inhibitors as novel therapeutics. Rev. Med.
Virol. 14, 255–270.
(34) Lundstrom, J., and Holmgren, A. (1990) Protein disulfide-
isomerase is a substrate for thioredoxin reductase and has thioredoxin-
like activity. J. Biol. Chem. 265, 9114–9120.
(15) Ryser, H. J., Levy, E. M., Mandel, R., and DiSciullo, G. J. (1994)
Inhibition of human immunodeficiency virus infection by agents that
interfere with thiol-disulfide interchange upon virus-receptor interac-
tion. Proc. Natl. Acad. Sci. U.S.A. 91, 4559–4563.
(16) Ryser, H. J., and Fluckiger, R. (2005) Progress in targeting
HIV-1 entry. Drug Discovery Today 10, 1085–1094.
(35) Osada, H. (2010) Introduction of new tools for chemical
biology research on microbial metabolites. Biosci. Biotechnol. Biochem.
74, 1135–1140.
(36) Pigiet, V. P., and Schuster, B. J. (1986) Thioredoxin-catalyzed
refolding of disulfide-containing proteins. Proc. Natl. Acad. Sci. U.S.A. 83,
7643–7647.
(17) Tsibris, J. C., Hunt, L. T., Ballejo, G., Barker, W. C., Toney, L. J.,
and Spellacy, W. N. (1989) Selective inhibition of protein disulfide
isomerase by estrogens. J. Biol. Chem. 264, 13967–13970.
(18) Wang, E. J., Snyder, R. D., Fielden, M. R., Smith, R. J., and Gu,
Y. Z. (2008) Validation of putative genomic biomarkers of nephrotoxi-
city in rats. Toxicology 246, 91–100.
(37) Hillson, D. A., Lambert, N., and Freedman, R. B. (1984)
Formation and isomerization of disulfide bonds in proteins: protein
disulfide-isomerase. Methods Enzymol. 107, 281–294.
(19) Godin, B., and Touitou, E. (2004) Mechanism of bacitracin
permeation enhancement through the skin and cellular membranes from
an ethosomal carrier. J. Controlled Release 94, 365–379.
(20) Mou, Y., Ni, H., and Wilkins, J. A. (1998) The selective
inhibition of beta 1 and beta 7 integrin-mediated lymphocyte adhesion
by bacitracin. J. Immunol. 161, 6323–6329.
(21) Donoghue, N., Yam, P. T., Jiang, X. M., and Hogg, P. J. (2000)
Presence of closely spaced protein thiols on the surface of mammalian
cells. Protein Sci. 9, 2436–2445.
(22) Smith, A. M., Chan, J., Oksenberg, D., Urfer, R., Wexler, D. S.,
Ow, A., Gao, L., McAlorum, A., and Huang, S. G. (2004) A high-
throughput turbidometric assay for screening inhibitors of protein
disulfide isomerase activity. J. Biomol. Screening 9, 614–620.
(23) Nelson, L. R., and Bulun, S. E. (2001) Estrogen production and
action. J. Am. Acad. Dermatol. 45, S116–124.
(24) Dodds, E. C., Goldberg, L., Lawson, W., and Robinson, R.
(1938) Estrogenic activity of certain synthetic compounds. Nature 141,
247–248.
(25) Primm, T. P., and Gilbert, H. F. (2001) Hormone binding by
protein disulfide isomerase, a high capacity hormone reservoir of the
endoplasmic reticulum. J. Biol. Chem. 276, 281–286.
(26) Arteaga, S., Andrade-Cetto, A., and Cardenas, R. (2005) Larrea
tridentata (Creosote bush), an abundant plant of Mexican and
US-American deserts and its metabolite nordihydroguaiaretic acid.
J. Ethnopharmacol. 98, 231–239.
(27) Nakamura, M., Nakazawa, J., Usui, T., Osada, H., Kono, Y., and
Takatsuki, A. (2003) Nordihydroguaiaretic acid, of a new family of
microtubule-stabilizing agents, shows effects differentiated from pacli-
taxel. Biosci. Biotechnol. Biochem. 67, 151–157.
(28) Nazrullaev, S. S., Saidkhodzhaev, A. I., Akhmedkhodzhaeva,
K. S., Syrov, V. N., Rasulev, B. F., and Khushbaktova, Z. A. (2008)
Estrogen activity of terpenoids from plants of the genus Ferula. Chem.
Nat. Compd. 44, 572–577.
(29) Lyles, M. M., and Gilbert, H. F. (1991) Catalysis of the
oxidative folding of ribonuclease A by protein disulfide isomerase:
dependence of the rate on the composition of the redox buffer.
Biochemistry 30, 613–619.
(30) Maattanen, P., Kozlov, G., Gehring, K., and Thomas, D. Y.
(2006) ERp57 and PDI: multifunctional protein disulfide isomerases
with similar domain architectures but differing substrate-partner asso-
ciations. Biochem. Cell. Biol. 84, 881–889.
(31) Auwerx, J., Isacsson, O., Soderlund, J., Balzarini, J., Johansson,
M., and Lundberg, M. (2009) Human glutaredoxin-1 catalyzes the
reduction of HIV-1 gp120 and CD4 disulfides and its inhibition reduces
HIV-1 replication. Int. J. Biochem. Cell Biol. 41, 1269–1275.
(32) Miyazaki, I., Simizu, S., Ichimiya, H., Kawatani, M., and
Osada, H. (2008) Robust and systematic drug screening method
using chemical arrays and the protein library: identification of novel
inhibitors of carbonic anhydrase II. Biosci. Biotechnol. Biochem. 72,
2739–2749.
(33) Simizu, S., Suzuki, T., Muroi, M., Lai, N. S., Takagi, S., Dohmae,
N., and Osada, H. (2007) Involvement of disulfide bond formation in
the activation of heparanase. Cancer Res. 67, 7841–7849.
251
dx.doi.org/10.1021/cb100387r |ACS Chem. Biol. 2011, 6, 245–251