Curcumin, the active constituent of turmeric, inhibits amyloid peptide-induced cytochemokine gene expression and CCR5-mediated chemotaxis of THP-1 monocytes by modulating early growth response-1 transcription factor

Journal of Neurochemistry, 2004, 91, 1199–1210
doi:10.1111/j.1471-4159.2004.02800.x
Curcumin, the active constituent of turmeric, inhibits
amyloid peptide-induced cytochemokine gene expression
and CCR5-mediated chemotaxis of THP-1 monocytes
by modulating early growth response-1 transcription factor
Ranjit K. Giri, Vikram Rajagopal and Vijay K. Kalra
Department of Biochemistry and Molecular Biology, University of Southern California, Keck School of Medicine, Los Angeles,
California, USA
Abstract
(MIP-1b, MCP-1 and IL-8) in both peripheral blood monocytes
Epidemiological studies show reduced risk of Alzheimer’s
and THP-1 cells. We found that curcumin inhibited Ab1)40-
disease (AD) among patients using non-steroidal in?amma-
induced MAP kinase activation and the phosphorylation of
tory drugs (NSAID) indicating the role of in?ammation in AD.
ERK-1/2 and its downstream target Elk-1. We observed that
Studies have shown a chronic CNS in?ammatory response
curcumin inhibited Ab1)40-induced expression of CCR5 but
associated with increased accumulation of amyloid peptide
not of CCR2b in THP-1 cells. This involved abrogation of
and activated microglia in AD. Our previous studies showed
Egr-1 DNA binding in the promoter of CCR5 by curcumin as
that interaction of Ab1)40 or ?brilar Ab1)42 caused activation of
determined by: (i) electrophoretic mobility shift assay,
nuclear transcription factor, early growth response-1 (Egr-1),
(ii) transfection studies with truncated CCR5 gene promoter
which resulted in increased expression of cytokines (TNF-a
constructs, and (iii) chromatin immunoprecipitation analysis.
and IL-1b) and chemokines (MIP-1b, MCP-1 and IL-8) in
Finally, curcumin inhibited chemotaxis of THP-1 monocytes in
monocytes. We determined whether curcumin, a natural
response to chemoattractant. The inhibition of Egr-1 by cur-
product known to have anti-in?ammatory properties, sup-
cumin may represent a potential therapeutic approach to
pressed Egr-1 activation and concomitant expression of
ameliorate the in?ammation and progression of AD.
cytochemokines. We show that curcumin (12.5–25 lM) sup-
Keywords: amyloid peptide, cytochemokines, early growth
presses the activation of Egr-1 DNA-binding activity in THP-1
response-1, curcumin, monocytes.
monocytic
cells.
Curcumin
abrogated
Ab1)40-induced
J. Neurochem. (2004) 91, 1199–1210.
expression of cytokines (TNF-a and IL-1b) and chemokines
Alzheimer’s disease (AD) is a neurodegenerative disorder,
the most frequent cause of loss of memory and cognitive
functions of the brain, which affects more than 5% of the
Received June 3, 2004; revised manuscript received July 31, 2004;
population over the age of 65 years. The disease is
accepted August 3, 2004.
characterized by increased deposition of amyloid-b (Ab)
Address correspondence and reprint requests to Vijay K. Kalra,
peptide and neuro?brilary tangles in the brain, senile plaques
Department of Biochemistry and Molecular Biology, HMR-611, USC
Keck School of Medicine, Los Angeles, CA 90033, USA.
around reactive microglia, and progressive loss of neurons in
E-mail: vkalra@usc.edu
the brain (Berg et al. 1993; Mattson and Rydel 1996). One
Abbreviations used: Ab, amyloid-b peptide; AD, Alzheimer’s disease;
?nds an increased presence of monocytes/macrophages in the
Egr-1, early growth response-1; EMSA, electrophoretic mobility shift
cerebral vessel wall and reactive or activated microglial cells
assay; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-terazolium bro-
in the adjacent parenchyma (Yamada et al. 1996; Maat-
mide; NSAID, non-steroidal in?ammatory drugs; PBM, peripheral blood
monocytes;
PBS,
phosphate-buffered
saline;
PMA,
4b-phorbol
Schieman et al. 1997; Uchihara et al. 1997; Wisniewski
12-myristate 13-acetate; PMSF, phenylmethanesulfonyl ?uoride; SDS,
et al. 1997). Studies (Eglitis and Mezey 1997) have shown
sodium dodecyl sulfate; SDS–PAGE, sodium dodecyl sulfate–poly-
that peripheral hematopoietic cells (e.g. monocytes) can cross
acrylamide gel electrophoresis.
Ó 2004 International Society for Neurochemistry, J. Neurochem. (2004) 91, 1199–1210
1199

---

1200
R. K. Giri et al.
the blood–brain barrier and undergo differentiation into
Ab-infusion induced spatial memory de?cits and Ab deposits
microglial cells in the brain. It has also been shown (Fiala
in Sprauge–Dawley rats (Frautschy et al. 2001). Curcumin
et al. 1998; Giri et al. 2000, 2002) that both soluble and
has also been shown to protect against Ab-induced injury to
?brilar form of Ab augment the transmigration of monocytes
neuronal cells (Park and Kim 2002).
across monolayer of cultured brain endothelial cells derived
Our results indicate that Ab1)40-induced gene expression
either from normal or AD individuals.
of speci?c cytokines (TNF-a and IL-1b), chemokines (MCP-
Studies have shown that non-steroidal anti-in?ammatory
1, IL-8 and MIP-1b) and chemokine receptor (CCR5) in
drugs (NSAID) reduce the incidence and progression of AD
THP-1 monocytes is abrogated by curcumin. Moreover, we
(Mackenzie 1996; Combs et al. 2000). These studies thus
show that curcumin inhibits Ab- induced Egr-1 DNA-
support the notion that in?ammation plays a role in the
binding activity in these monocytic cells. Furthermore, we
pathogenesis of AD (Akiyama et al. 2000). Activated micro-
show that Ab-induced CCR5 expression in THP-1 mono-
glia, like activated macrophages, have been shown to generate
cytes, which plays a role in chemotaxis in response to
in?ammatory molecules, such as cytokines (TNF-a and
b-chemokines (MIP-1b), is abrogated in response to curcu-
IL-1b), chemokines (MCP-1), C-reactive protein and comple-
min. To our knowledge, this is the ?rst report showing that
ment components (McGeer et al. 1993, 2000; Bradt et al.
inhibition of Egr-1 activation (among other transcription
1998; Combs et al. 2001). Our recent studies (Giri et al. 2003)
factors), by curcumin, a pharmacological agent, can block
show that amyloid peptides, both soluble Ab1)40 and ?brilar
Ab-mediated in?ammatory response in monocytes.
Ab1)42, at physiological concentrations, as found in the plasma
of AD individuals (Kuo et al. 1999), show increase in the gene
expression of speci?c cytokines (TNF-a and IL-1b) and
Materials and methods
chemokines (MCP-1, IL-8 and MIP-1b) in THP-1 monocytes
and peripheral blood monocytes. We also showed that amyloid
Amyloid peptides and their ?brillation state
peptide-induced expression of these cytokines and chemok-
Human amyloid peptides (Ab1)40 and Ab1)42) were custom
ines in monocytes was regulated by activation of transcription
synthesized, puri?ed and characterized by amino acid analysis and
laser desorption spectrophotometry as described earlier (Giri et al.
factor AP-1 and Egr-1 (Giri et al. 2003). Moreover, amyloid
2003). The non-?brilar form of Ab
peptide-induced expression of selective cytokines (TNF-a and
1)40 was prepared by dissolving
it in dimethylsulfoxide at a concentration of 2 mg/mL or freshly
IL-1b) and chemokines (MCP-1, IL-8 and MIP-1b) in THP-1
prepared in endotoxin-free water. The absence of ?brilar forms in
monocytes was abrogated by small inhibitory RNA duplexes
this preparation was con?rmed by a thio?avin T ?uorescence assay
(siRNA) for early growth response-1 (Egr-1) mRNA (Giri
and far-UV CD spectra (Giri et al. 2003). These peptide solutions
et al. 2003). These studies suggested that inhibition of Egr-1
were negative for endotoxin (< 10 pg/mL), as determined by
by siRNA for Egr-1 may represent a potential therapeutic
Limulus lysate test (Giri et al. 2003). Ab1)40 when freshly prepared
target to ameliorate the in?ammation in AD.
in water was monomeric, although it showed a small amount
We sought to identify pharmacological agent(s) that could
( 10%) of the dimeric form when kept for 7 days, as analyzed by
block Egr-1-mediated cytokine and chemokine expression,
electrophoresis on native gel followed by western blotting with an
and at the same time be effective and safe for use in humans.
antibody to Ab1)40. Ab1)42 (2 mg/mL) when dissolved in water and
kept at 37°C for 7 days showed ?brilar content.
Studies (Pendurthi et al. 1997; Pendurthi and Rao 2000)
have shown that curcumin (diferuloylmethane), a major
Reagents
biological active component of turmeric (Curcuma longa),
Curcumin as curcuminoid was purchased from Sigma Chemical
inhibits phorbol-ester (4b-phorbol 12-myristate 13-acetate;
Company (St Louis, MO, USA).
PMA)-induced activation of Egr-1, AP-1 and NF-jB in
Anti-phospho-p42/44 MAPK (E10: monoclonal) was purchased
endothelial cells. Turmeric is used as a curry spice and herbal
from Cell Signaling Inc. (Beverly, MA, USA). Rabbit anti-ERK-1
medicine in India for the treatment of a number of
(SC-93), anti-phospho-Elk-1 (SC-8406: monoclonal), rabbit anti-
in?ammatory conditions, cancer and other diseases (Ammon
Egr-1 (SC-110X), goat anti-SP-1 (SC-59X) and secondary antibod-
and Wahl 1991; Aggarwal et al. 2003; Bharti et al. 2003).
ies conjugated to horseradish peroxidase were purchased from Santa
Epidemiological studies in India, where turmeric is used
Cruz Biotechnology (Santa Cruz, CA, USA). MIP-1a and MIP-1b
routinely, show that the incidence of AD between the ages of
recombinant proteins were purchased from R&D Systems Inc.
(Minneapolis, MN, USA). Custom-made multiprobe templates for
70 and 79 years is  4.4-fold less than that seen in the USA
TNF-a, IL-1b, RANTES, MIP-1b, MCP-1 and IL-8, and the house
(Ganguli et al. 2000). These studies are supported in animal
keeping genes L-32 and GAPDH were obtained from Pharmingen
models, wherein Lim et al. (2001) showed that administra-
(San Diego, CA, USA). All other reagents, unless otherwise
tion of dietary curcumin to amyloid transgenic mice (APPS),
speci?ed, were purchased from Sigma.
which display age-related neuritic plaques (Hsiao et al.
1996) and age-related memory de?cits (Chapman et al.
Cell culture and isolation of peripheral blood monocytes
1999), resulted in the reduction of plaque burden. In related
The THP-1 monocytic cell line obtained from ATCC (Manassas,
studies,
it
has
been
shown
that
curcumin
prevents
VA,
USA)
was
cultured
in
RPMI-1640
containing
10%
Ó 2004 International Society for Neurochemistry, J. Neurochem. (2004) 91, 1199–1210

---

Curcumin inhibits Ab-induced cytochemokines
1201
heat-inactivated fetal calf serum as described previously (Giri et al.
Technology). Horseradish peroxidase-conjugated secondary anti-
2003). On the day of the experiment THP-1 cells (1 · 106 cells/mL)
bodies were used to develop the membrane and visualization of
were cultured in serum-free RPMI-1640 for 4–6 h. Peripheral blood
bands was performed using Supersignal chemiluminescent substrate
monocytes (PBM) were isolated from blood collected in EDTA as
(Pierce Biotechnology, Rockford, IL, USA). Blots were stripped and
the anticoagulant as previously described (Giri et al. 2003).
reprobed using a 1:1000 dilution of antibodies against the
antip42/44 antibody to normalize the protein loading. The intensity
Cell viability assay
of bands was quanti?ed utilizing Alpha Imager 2000 gel documen-
Brie?y, THP-1 (5000 cells/well) were incubated in duplicate, in 96-
tation system.
well plates, in the absence and presence of Ab1)40 peptide (125 nM)
for 1 h, followed by incubation with curcumin at the indicated
Preparation of nuclear extracts
concentrations for 4 h in a ?nal volume of 0.1 mL. Then 25 lL
Nuclear extracts were prepared from THP-1 cells as described
of
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-terazolium
bromide
previously (Giri et al. 2003). Brie?y, 5 · 106 cells were washed
(MTT) solution [5 mg/mL in phosphate-buffered saline (PBS)]
with ice-cold PBS, resuspended in 400 lL of cell lysis buffer
was added to each well. The contents were incubated at 37°C for
[10 mM HEPES at pH 7.9, 100 mM KCl, 1.5 mM MgCl2, 0.1 mM
2 h. We added 0.1 mL of extraction buffer [20% sodium dodecyl
EGTA, 0.5 mM dithiothreitol, 0.5 mM phenyl methanesulfonyl
sulfate (SDS) in 50% dimethyl formamide) to each well and wells
?uoride, 0.5% Nonidet p-40 and 1 lL/mL of protease inhibitor
were incubated for an additional 24 h. Optical density was measured
cocktail (Calbiochem, La Jolla, CA, USA)], swelled on ice for
at 590 nm. Percentage viability was calculated compared with
30 min followed by vigorous vortex mixing for 5–10 s. A nuclear
untreated control (100%).
pellet was obtained by centrifugation of the homogenate at 10 000 g
for 30 s. The nuclear pellet was resuspended in 50 lL of nuclear
RNase protection assay
extraction buffer (10 mM HEPES at pH 7.9, 1.5 mM MgCl2,
THP-1 monocytes were treated with Ab1)40 peptide for various
420 mM NaCl, 0.1 mM EGTA, 0.5 mM dithiothreitol, 5% glycerol,
times and total RNA was isolated with TRIzol reagent (Invitrogen,
0.5 mM PMSF and 1 lL/mL of protease inhibitor cocktail).
Carlsbad, CA, USA). RNase protection assays were performed on
Contents were mixed intermittently for 60 min. The nuclear extract
total RNA extracted from THP-1 cells using custom-made multi-
was obtained by centrifuging at 10 000 g for 10 min at 4°C.
probe templates for TNF-a, IL-1b, RANTES, MIP-1b, MCP-1,
IL-8, CCR2a and CCR5, and the housekeeping genes L-32 and
Electrophoretic mobility shift assay (EMSA) for transcription
GAPDH (Pharmingen, San Diego, CA, USA). Brie?y, templates
factors Egr-1
were labeled with [a-32P] UTP using T7 RNA polymerase according
The oligonucleotide used as probes were as follows: Egr-1, 5¢-GGA-
to the manufacturer’s protocol. RNA (10 lg) was hybridized with
TCCAGCGGGGGCGAGCGGGGGCGA-3¢ and 3¢-CCTAGGTC-
32P-labeled template probe (8 · 105 c.p.m.) for 12–16 h at 56°C.
GCCCCCGCTCGCCCCCGCT-5¢, which were synthesized at Nor-
The contents were treated with RNase mixture (Pharmingen)
ris Cancer Center Microchemical core facility at USC. Probes were
followed by phenol–chloroform extraction as previously described
5¢-end labeled with 100 lCi of [c-32P] ATP using T4-polynucleotide
(Giri et al. 2003). Protected mRNA hybrids were resolved on a 6%
kinase. The labeled single-stranded sense oligonucleotide probe was
denaturing polyacrylamide-sequencing gel and exposed to X-ray
mixed with labeled antisense probe and incubated at 65°C for 5 min
?lm for 24 h. The intensity of bands corresponding to TNF-a, MIP-
followed by annealing at room temperature (25°C) for 15 min.
1b, IL-1b, MCP-1, IL-8, CCR2a, CCR5, RANTES, L-32 and
The DNA-binding reaction mixture contained nuclear proteins
GAPDH were analyzed using an Alpha Imager 2000 gel documen-
(2–4 lg),
32P-labeled double-stranded oligonucleotide probe
tation system (San Leandro, CA, USA). Values were expressed as
( 50 000 c.p.m.) and 2 lg of poly(dI–dC). To demonstrate
relative expression of mRNA normalized to the mean of L-32 and
speci?city of DNA–protein interaction, a 50-fold excess of
GAPDH mRNA.
unlabeled double-stranded oligonucleotide probe was added. In
supershift assays, nuclear extracts were pre-incubated for 20 min at
Western blot analysis
room temperature with 2 lg of antibody to either Egr-1 or SP-1,
For western blot analysis, THP-1 cells were cultured in RPMI-1640
prior to the addition of radiolabeled probe. The DNA–protein
medium containing 10% FBS for 3–4 days. On the day of the
complex was then size fractionated from the free DNA probe by
experiment, cells were pelleted and resuspended at 1 · 106cells/mL
electrophoresis in a 4% non-denaturing polyacrylamide gel. The gel
in serum-free RPMI-1640 and incubated for an additional 3 h prior
was dried and exposed to X-ray ?lm.
to treatment with Ab1)40 peptide (125 nM). Where indicated, THP-1
monocytes were incubated with curcumin or pharmacological
Transient transfection of THP-1 cells and luciferase activity
inhibitors for 30 min prior to Ab1)40 treatment. The medium was
assay
aspirated and cells were lysed in RIPA buffer [1 ·
PBS, 1%
The ?re?y luciferase reporter gene plasmids of CCR5 promoter
Nonidet p-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM sodium
(PA-3) used were kindly provided by Dr Sunil Ahuja (San Antonio,
orthovanadate, 10 lg/mL phenylmethanesulfonyl ?uoride (PMSF)
TX, USA). Mummidi et al. (1997) previously described their
and 1 lL/mL of protease inhibitor cocktail). 10 lg of proteins were
preparation and features. THP-1 cells (2–3 · 106 cells/well) were
size fractionated in 10% sodium dodecyl sulfate–polyacrylamide gel
cultivated in six-well chambers. The reporter gene constructs were
electrophoresis (SDS–PAGE) gel and transferred to nitrocellulose
transiently transfected in THP-1 cells by using Lipofectamine
membrane (Bio-Rad, Hercules, CA, USA). Blots were probed with
reagent (Invitrogen). Transfection ef?ciency was normalized by
1:1000 dilution of antiphospho-p42/44 antibody (Cell Signaling
cotransfecting
THP-1
cells
with
CCR5
promoter-luciferase
Ó 2004 International Society for Neurochemistry, J. Neurochem. (2004) 91, 1199–1210

---

1202
R. K. Giri et al.
constructs (10 lg/well) and 0.5 lg of renilla luciferase vector (pRL-
non-migrated cells. This was followed by the addition of PBS
CMV; Promega, Madison, WI, USA). Alternatively, THP-1 cells
containing 2 mM EDTA to the upper chamber and incubation at 4°C
cotransfected with 10 lg of the promoter less vector pGL3-Basic
for 15 min. Cells that had migrated into the lower compartment of the
(Promega) and 0.5 lg of renilla luciferase vector (pRL-CMV) were
Boyden chamber were counted in ?ve microscopic high-power ?elds
used as a negative control. After 2 days of transfection, the cells
(40 ·) utilizing an Olympus IMT-2 microscope. Where indicated,
were pelleted, washed in Dulbecco’s PBS and lysed in 1· passive
THP-1 cells were pretreated with Ab1)40 in the presence and absence
lysis buffer (Promega). The protein concentration in the cell lysates
of curcumin for 4 h, washed with serum-free medium and used
was determined by using the Bradford method. The ?re?y and
directly in the chemotaxis assay. Each sample was tested in triplicate.
renilla luciferase activities in the lysates were determined according
to the manufacturer’s instructions (Dual-Luciferase Reporter Assay
Statistical analysis
System, Promega) utilizing a luminometer (Berthold Technologies
Statistical analysis of the responses obtained from control and
USA, Oakridge, TN, USA). The relative luciferase activity in each
Ab-treated monocytes were carried out by one-way analysis of
sample was determined as follows: X ¼ Fire?y luciferase activity of
variance (ANOVA) utilizing INSTAT 2 (Graphpad, San Diego, CA, USA)
CCR5 promoter construct divided by renilla luciferase activity of
software program. The effects of curcumin on Ab-induced responses
pRL-CMV construct; Y ¼ Fire?y luciferase activity of promoter
were analyzed by comparing the response of monocytes in the
less vector pGL3-Basic divided by renilla luciferase activity of pRL-
presence and absence of inhibitor. Student’s t-test was used for
CMV vector; Z ¼ X‚Y and Relative luciferase activity is expressed
multiple comparisons. Values of p < 0.05 were considered signi?cant.
as Z ‚ lg of protein in the lysate sample.
Chromatin immunoprecipitation assay
Results
THP-1 cells (5 · 106 cells) were serum starved for 6 h followed by
treatment with Ab1)40 for the indicated time. Chromatin immuno-
Curcumin reduces Ab-induced cytokine and chemokine
precipitation analysis was performed as described previously
expression in PBM and THP-1 monocytic cells
(Reddy et al. 2003). Brie?y, after stimulation with Ab, cells were
Because our recent studies (Giri et al. 2000) showed that
washed with PBS and cross-linked with 1% formaldehyde at room
temperature for 10 min. Cells were lysed, sonicated and superna-
nanomolar concentrations (125 nM) of both Ab1)40 and
tants were recovered by centrifugation of lysate at 15 000 g for
Ab1)42 were effective in mediating the transmigration of
10 min at 4°C. The supernatant was diluted 4-fold in a dilution
monocytes across a monolayer of cultured human brain
buffer (1% Triton X-100, 2 mM EDTA, 150 mM NaCl and 20 mM
endothelial cells (Giri et al. 2000, 2002) and submicromolar
Tris–HCl, pH 8.1) followed by the addition of 2 lg of sheared
concentrations of amyloid peptide have been observed in
salmon sperm DNA, 2.5 lg of pre-immune serum and 20 lL of
plasma of AD subjects (Kuo et al. 1999), we studied the
protein A–Sepharose (50% slurry). The contents were kept at 4°C
effect of Ab over this submicromolar range. We have
for 2 h. The precleared supernatant was immunoprecipitated by
previously shown (Giri et al. 2003) that both Ab1)40 and
adding antibody (2 lg/mL) to either Egr-1 or SP-1, 2 lg of sheared
Ab1)42 at 125 nM caused an increase in mRNA expression of
salmon sperm DNA and 20 lL of protein A–Sepharose (50%
TNF-a, MIP-1b, IL-1b, MCP-1 and IL-8 in THP-1 mono-
slurry) and incubated at 4°C for 12–16 h. After several washings,
cytes and human PBM, thus we studied the effect of
the protein was digested with proteinase K (10 lg/mL) for 1 h. The
cross-linking between DNA and protein was reversed by incubating
curcumin at this dose of amyloid peptide. Because both non-
the immunoprecipitate at 65°C overnight. DNA was phenol–
?brilar Ab1)40 and ?brilar Ab1)42 were equally effective in
chloroform extracted, ethanol precipitated, air dried and dissolved
increasing mRNA expression of these aforementioned cyto-
in 50 lL of TE buffer (10 mM Tris–HCl, pH 8.0 and 1 mM EDTA).
chemokines, we used Ab1)40 in the studies described here.
Five microliters of DNA sample was subjected to polymerase chain
As shown in Fig. 1(a), 12.5–100 lM curcumin reduced
reaction (PCR) ampli?cation utilizing primers (5¢-CCAGCAGCAT
Ab1)40-mediated mRNA expression of TNF-a, IL-1b, MIP-
GACTGCAGTT- 3¢,
forward
primer;
5¢-GCTAATTGCT
1b, MCP-1 and IL-8 as determined by RNase protection
GGTGCTTGGAG- 3¢ reverse primer) corresponding to the promo-
assay. Under these conditions the mRNA expression of
ter region of CCR5 (from )847 to )603 respective to the
RANTES remained unchanged. Similarly, the ?brilar form of
transcription start site).
Ab1)42-mediated (125 nM) cytochemokine (TNF-a, IL-1b,
MIP-1b, MCP-1 and IL-8) mRNA expression (Fig. 1b) was
Chemotaxis assay
Chemotaxis was assayed in 96-well plates (Neuro Probe Inc.,
suppressed by curcumin (5, 12.5 and 25 lM). At 5 lM
Gaithersburg, MD, USA) with Transwell inserts of 5-lm pore size.
curcumin, the inhibition in cytochemokine expression was
Brie?y, THP-1 monocytes were washed and resuspended in serum-
modest. However, at a higher concentration of curcumin
free RPMI-1640 medium and 1 · 105 cells/50 lL were then loaded
(12.5–25 lM) there was almost complete abrogation of
onto insert of the Boyden chamber. Chemotaxis medium (30 lL of
Ab-induced cytochemokines expression. It is pertinent to
serum–free RPMI-1640 medium containing indicated amounts of
note that curcumin (6.25–25 lM) did not affect the viability
chemokines) was placed in the bottom compartment. After 2 h of
of THP-1 cells signi?cantly as determined by MTT assay
incubation at 37°C in a 5% CO2 incubator, cells were scraped from
(Fig. 1c) and Trypan Blue exclusion (Fig. 1d). However,
the upper chamber and washed with PBS (100 lL) to remove
Ó 2004 International Society for Neurochemistry, J. Neurochem. (2004) 91, 1199–1210

---

Curcumin inhibits Ab-induced cytochemokines
1203
(a)
Fig. 1 Curcumin inhibits both Ab1)40- and Ab1)42-mediated mRNA
Curcumin(µM) -
- 12.5 25 50 100 25
expression of cytokines and chemokines in THP-1 monocytes. THP-1
A?
-
+
+
+
+ +
-
1-40 (125 nM)
cells were treated with (a) Ab1)40 (125 nM) and (b) Ab1)42 (125 nM) for
2 h in the absence and presence of curcumin (12.5–100 lM). RNA
TNF-?
(10 lg) was subjected to a RNase protection assay as described in
IL-1?
Materials and Methods. The autoradiogram shows the protected
bands of each gene (TNF-a, IL-1b, MIP-1b, MCP-1, IL-8 and RAN-
MIP-1?
TES). The data are normalized to means of L-32 and GAPDH signal
(housekeeping genes). Data are representative of two separate
MCP-1
experiments. The cell viability of THP-1 cells was measured by (c)
MTT assay and (d) the Trypan Blue exclusion method.
IL-8
RANTES
50–100 lM curcumin was toxic causing a > 75% reduction in
cell viability. Because 12.5–25 lM curcumin was optimal in
L32
inhibiting mRNA expression of TNF-a, IL-1b, MIP-1b,
GAPDH
MCP-1 and IL-8, we used this concentration for the studies
described here. We determined whether curcumin (25 lM)
(b)
inhibited amyloid peptide-induced cytochemokine expres-
-
Curcumin (
-
5
12.5
25
µM)
sion in PBM. As shown in Fig. 2, Ab1)40 (125 nM) increased
A? 1-42 (125 nM)
-
+
+
+
+
the mRNA expression of cytokines (TNF-a and IL-1b) and
chemokines (MIP-1b, MCP-1 and IL-8) in PBM, whereas the
TNF-?
expression of RANTES remained unchanged as described
MIP-1?
previously (Giri et al. 2003). Curcumin (25 lM) inhibited
> 90% mRNA expression of TNF-a, IL-1b, MIP-1b, MCP-
IL-1?
1and IL-8 induced by Ab1)40 (125 nM) (Fig. 2). Because
MCP-1
curcumin showed a similar inhibition pro?le for amyloid
peptide-induced cytochemokine gene expression in both
IL-8
THP-1 cells and PBM, we utilized THP-1 monocytic cells as
a model system for subsequent studies.
RANTES
L32
Curcumin inhibits activation of ERK-1/2 and Elk-1, and
GAPDH
expression of Egr-1
Our previous studies (Giri et al. 2003) have shown that
(c)
120
Ab1)40 (125 nM) causes cellular signaling in THP-1 mono-
cytes leading to downstream activation of members of the
100
MAPK family, namely ERKs (ERK-1/ERK-2), but not of
80
p38MAP kinase. As shown in Fig. 3(a), Ab1)40 (125 nM)
i
ability
60
increased phosphorylation of both ERK-1 and ERK-2, which
40
was abrogated to the basal level when THP-1 cells were
Cell V
pretreated with curcumin (25 lM). Moreover, the phosphory-
20
(%) MTT Utilization)
lation of Elk-1, mediated by the activation of ERK, was
0
inhibited by > 90% by curcumin at a dose of 25 lM, whereas
A? 1-40 (125 nM) -
+
+
+
+
+
+
-
the effect was less ( 40%) in the presence of 12.5 lM
Curcumin(µM)
-
- 6.2512.5 25 50 100 50
curcumin (Fig. 3b). Our previous studies (Giri et al. 2003)
(d)
120
have shown that Ab-induced activation of ERKs and Elk-1
e
v
resulted in activation of the transcription factor Egr-1, thus
100
we examined whether curcumin affected Egr-1 protein
80
expression. As shown in Fig. 3(b), 12.5 lM curcumin
60
reduced Egr-1 protein by  80% and at a dose of 25 lM
iability (%)
completely reduced Egr-1 protein to the basal levels.
40
Cell V
rypan Blue Exclusi 20
T
Curcumin inhibits Ab-mediated activation of
0
transcription factor Egr-1
Ab 1-40 (125 nM) -
+
+
+
+
+
+
-
Our previous studies (Giri et al. 2003) have shown that
Curcumin(µM) -
-
6.25 12.5 25 50 100 50
Ab1)40 (125 nM) caused activation of transcription factor
Ó 2004 International Society for Neurochemistry, J. Neurochem. (2004) 91, 1199–1210

---

1204
R. K. Giri et al.
(a)
Curcumin (25 µM)
-
-
+
Curcumin (25 µM)
-
+
-
A? 1-40 (125 nM)
-
+
-
A? 1-40 (125 nM)
+
+
-
pERK-1/2
TNF-?
ERK-1/2
MIP-1?
(b)
Curcumin (µM)
0
0
12.5
25
25
IL-1?
A? 1-40 (125 nM) -
+
+
+
-
pElk-1
MCP-1
Egr-1
NS
IL-8
Fig. 3 Curcumin inhibits Ab1)40-mediated phosphorylation of ERK-1/2
and Elk-1, and protein levels of Egr-1 in THP-1 monocytes. THP-1
RANTES
monocytes (5 · 106 cells) were pre-incubated with curcumin (25 lM)
for 30 min. Cells were then treated with Ab1)40 (125 nM) for 30 min.
Cell lysates were subjected to 10% SDS–PAGE followed by western
CCR-2b
blotting. (a) Blots were probed with antiphospho-ERK-1/2 antibody. To
normalize for protein loading, membranes were stripped and reprobed
with anti-ERK-1/2. (b) Curcumin inhibits the phosphorylation of Elk-1
protein and expression of Egr-1 protein in THP-1 monocytes. THP-1
L-32
cells were incubated in the absence and presence of curcumin for
30 min followed by treatment with Ab1)40 (125 nM) for 30 min. Nuclear
extracts (5 lg) were resolved in 10% SDS–PAGE followed by western
GAPDH
blot analysis using antibodies against phosphorylated Elk-1 (upper).
Membranes were stripped and reprobed using antibodies against Egr-
PBM
1 (b). The lower panel shows band (NS, non-speci?c), which was used
as a control to normalize the protein loading. The data are represen-
Fig. 2 Curcumin inhibits Ab1)40-mediated cytokine and chemokine
tative of three separate experiments.
mRNA expression in PBM. PBM (1.5 · 106 cells) were pre-incubated
with curcumin (25 lM) for 30 min followed by treatment with Ab1)40
(125 nM) for 2 h. Cytokine and chemokine mRNA expression were
previously described (Giri et al. 2003). THP-1 cells, which
analyzed by RNase protection assay analysis as described in Fig. 1.
were pretreated with curcumin (12.5 and 25 lM) for 30 min
The data are representative of three separate experiments.
prior to treatment with Ab1)40 (125 nM), did not show
activation of Egr-1 DNA-binding activity (Fig. 4, lanes 6
Egr-1 and AP-1, but not of CREB and NF-kB in THP-1
and 7). As shown in Fig. 4, lane 8, curcumin reduced basal
monocytes. Moreover, studies showed that transfection of
Egr-1 DNA-binding activity compared with THP-1 cells not
THP-1 monocytes with Egr-1 siRNA abrogated Ab-induced
treated with Ab1)40 (125 nM) (Fig. 4, lane 2).
mRNA expression of TNF-a, IL-1b, MIP-1b, MCP-1 and
IL-8. Pendurthi and Rao (2000) have shown that curcumin
Curcumin inhibits Ab-induced mRNA expression of
inhibits PMA and serum-induced activation of Egr-1 in
CCR5 in THP-1 monocytes
endothelial cells and ?broblasts. Thus, we determined whether
We previously (Giri et al. 2003) observed that Ab-induced
curcumin affected Ab-mediated activation of transcription
expression of MIP-1b and its cognate receptor CCR5 in THP-
factor Egr-1. As shown in Fig. 4, Ab1)40 (125 nM) caused a
1 monocytes. Moreover, these studies showed that CCR5
time-dependent activation of Egr-1 DNA-binding activity as
expressed on monocytic cells participated in the chemotaxis
determined by EMSA. At 60 min there was optimal Egr-1
of monocytes in response to chemoattractant MIP1-b and
DNA-binding activity. The Egr-1 signal was > 90% reduced in
RANTES. Thus, we examined whether curcumin inhibited
the presence of excess unlabeled Egr-1 probe (Fig. 4, lane 9).
CCR5 mRNA expression in THP-1 cells. As shown in
Furthermore, antibody to Egr-1 caused supershift of the band
Fig. 5(a), curcumin (12.5–50 lM) reduced CCR5 mRNA
corresponding to Egr-1 (Fig. 4, lane 10). Antibody to SP-1
expression in a dose-dependent manner. As shown, we see
failed to supershift the Egr-1 band (Fig. 4, lane 11) as
more than one transcript of CCR5 in Ab-treated THP-1 cells
Ó 2004 International Society for Neurochemistry, J. Neurochem. (2004) 91, 1199–1210

---

Curcumin inhibits Ab-induced cytochemokines
1205
in agreement with the data of Mummidi et al. (2000).
Anti-SP-1 Ab.
-
-
- -
-
-
-
-
- -
+
Anti-Egr-1 Ab. -
-
-
- - - - - -
+
-
However, curcumin (12.5–50 lM) did not affect mRNA
Cold Probe
-
-
-
-
-
-
-
- + - -
expression of CCR2b. Similarly, curcumin (25 lM) com-
Curcumin (µM) -
-
- -
- 12.5 25 25 - - -
pletely inhibited CCR5 mRNA expression in Ab-treated
A? 1-40 (125 nM) - - + + + + + - + + +
PBM, although expression of CCR2b remained unaffected
Time (min) -
0
15 30 60 60 60 60 60 60 60
(Fig. 5b). These data indicate that the effect of curcumin is
SS Egr-1
speci?c in inhibiting the expression of CCR5 receptor in both
NS
Egr-1
PBM and THP-1monocytic cells.
Curcumin inhibits functional Egr-1 binding site in CCR5
promoter
As shown in Fig. 4, curcumin inhibited Ab-induced Egr-1
DNA
binding
using
5¢-GGATCCAGCGGGGGCGAG-
CGGGGGCGA-3¢ as the bona ?de Egr-1 consensus sequence
for EMSA analysis. Because we previously (Giri et al. 2003)
observed
that
Egr-1
siRNA
abrogated
Ab-induced
CCR5 expression and human CCR5 promoter (Mummidi
et al. 1997) contains GCGGGGGTG, at positions )702 to
)694, a potential Egr-1 putative binding site, we utilized
oligonucleotides (upper strand, 5¢-GTCCCTATATGGGG-
CGGGGGTGGGGGTGTCT-3¢) as the putative Egr-1 con-
sensus sequence in CCR5 promoter ()715 to )685) for
Fig. 4 Effect of curcumin on Ab1)40-mediated Egr-1 DNA-binding
EMSA analysis. As shown in Fig. 6, Ab1)40 (125 nM)
activities in nuclear extracts of THP-1 cells by gel shift assay. THP-1
caused a time (15–60 min) dependent increase in Egr-1 DNA
cells were pre-incubated with curcumin, where indicated, for 30 min,
binding. Egr-1 DNA binding was optimal at 60 min (Fig. 6,
followed by treatment with Ab1)40 (125 nM) for the indicated times.
lane 5). Curcumin at a dose of 12.5 and 25 lM completely
Nuclear extracts were prepared for EMSA using the oligonucleotide
abrogated Ab-induced Egr-1 DNA binding (Fig. 6, lanes 6
probe for Egr-1. Where indicated, a 50-fold excess of unlabeled probe
and 7). As shown in Fig. 6, lane 9, excess unlabeled Egr-1
was added to the nuclear extract 10 min before addition of the
radiolabeled probe. In the supershift assay, nuclear extracts were pre-
probe completely reduced the signal. Moreover, antibodies to
incubated with antibody (2 lg) to either Egr-1 or SP-1. The data are
Egr-1 supershifted the Egr-1 band (Fig. 6, lane 10). As a
representative of three independent experiments. SS, supershifted
negative control antibody to SP-1 failed to supershift the
band in the presence of antibody. NS, non-speci?c band.
Egr-1 band (Fig. 6, lane 11).
Curcumin inhibits Ab-induced CCR5 promoter activity
in THP-1 monocytes
(a)
(b)
We have observed that Ab induces CCR5 mRNA expression
Curcumin (µM)
-
- 12.5 25 50 50
- - 25
at the transcriptional level by transfecting THP-1 cells with
A?
-
+
+
+
+
-
- + +
the luciferase-reporter construct containing the CCR5 pro-
1-40 (125nM)
moter region (from )1976 to +33) coupled to the 5¢-end of
CCR 5
the luciferase reporter gene, designated as PA-1 (kindly
CCR 2b
provided by Dr Sunil Ahuja) (Mummidi et al. 1997). To
delineate the promoter region in CCR5 that was activated by
L 32
Ab, we performed transient transfection of THP-1 cells with
GAPDH
a series of 5¢ deletion constructs [PA-2 construct, which
PBM
THP-1
contains CCR5 promoter region ()1358 to +33); PA-3
construct ()731 to +33), which contains a putative Egr-1
Fig. 5 Curcumin inhibits amyloid peptide-induced mRNA expression
binding site and a SP-1 cis acting element and PA-4 construct
of CCR5 in THP-1 monocytes and PBM. (a) THP-1 cells and (b) PBM
()412 to +33) which contains a SP-1 binding site (data not
were treated with Ab1)40 (125 nM) for 2 h in the absence and presence
shown)]. As shown in Fig. 7(a), we observed that THP-1
of curcumin. RNA (10 lg) was subjected to the RNase protection
cells transfected with the PA-3 construct showed optimal
assay as described in Materials and Methods. The autoradiogram
(15-fold) increase in luciferase activity in response to Ab.
shows the protected bands of CCR5, CCR2b, L-32 and GAPDH
Furthermore,
curcumin
(25 lM)
completely
reduced
genes. Data are representative of three independent experiments. The
broad CCR5 band can be seen as two transcripts at a lower level of
luciferase activity in THP-1 cells transfected with CCR5
exposure of the autoradiogram.
promoter deletion construct PA-3.
Ó 2004 International Society for Neurochemistry, J. Neurochem. (2004) 91, 1199–1210

---

1206
R. K. Giri et al.
(a)
Anti-SP-1 Ab.
-
-
- - - - - - -
- +
Curcumin (25µM) + A?
Anti-Egr-1 Ab.
+
1-40
-
- - -
- - - - -
-
Cold Probe -
-
- - - - -
- +
- -
Curcumin (µM) - - - - - 12.5 25 25 - - -
A-3
A? 1-40 (125nM)
P
A? 1-40 (125 nM) - - + + + + + - + + +
Time (min) -
0
15 30 60 60 60 60 60
60 60
None
SS Egr-1
0
2
4
6
8
10
12
14
16
18
Relative Luciferase Activity
Egr-1
(b)
Curcumin (µM)
-
0
0 12.5 25
A? 1-40 (125nM) M
-
+
+
+
300 bp
Egr-1
200 bp
100 bp
300 bp
SP-1
200 bp
100 bp
Fig. 7 Effect of curcumin on CCR5 promoter. (a) CCR5 promoter
constructs (PA-3) and pCMV renilla luciferase construct were
cotransfected into THP-1 monocytes. After 2 days post-transfection,
cells were washed with serum-free media. Where indicated, cells were
Fig. 6 Curcumin inhibits putative Egr-1 binding to CCR5 promoter in
pre-incubated with curcumin (25 lM) and then treated with Ab1)40.
nuclear extracts of THP-1 cells as determined by EMSA. THP-1 cells
Cells were pelleted, lysates prepared and luciferase activity deter-
were pre-incubated with curcumin for 30 min prior to treatment with
mined by dual luciferase assay kit (see Materials and methods). Data
Ab1)40 (125 nM) for various times (15–60 min). Nuclear extracts were
are presented as relative luciferase activity as described in Materials
prepared and incubated with 32P-labeled oligonucleotide probe for
and Methods (n ¼ 3, mean ± SD). Results are expressed as the
putative Egr-1 binding site in CCR5 promoter. Where indicated a
percentage of luciferase activity relative to untreated cells. (b) Cur-
50-fold excess of unlabeled probe was added to the nuclear extract
cumin reduces Ab1)40-induced Egr-1 binding to native chromatin of
10 min before addition of the radiolabeled probe. In the supershift
THP-1 cells as demonstrated by chromatin immunoprecipitation
assay, nuclear extracts were incubated with either antibody to Egr-1
assay. Nucleotides () 847 to )603) in CCR5 promoter containing a
(2 lg) or SP-1 (2 lg) for 20 min before addition of the radiolabeled
putative Egr-1 binding element and a known SP-1 binding element
probe. The data are representative of three independent experiments.
were utilized for the chromatin immunoprecipitation assay. Soluble
SS, supershifted band in the presence of Egr-1 antibody.
chromatin was prepare