Accepted Manuscript

Curcumin suppresses oncogenicity of human colon cancer cells by covalently

modifying the cysteine 67 residue of SIRT1

Yeon-Hwa Lee, Na-Young Song, Jinyoung Suh, Do-Hee Kim, Wonki Kim, Jihyae Ann,
Jeewoo Lee, Jeong-Heum Baek, Hye-Kyung Na, Young-Joon Surh

PII: S0304-3835(18)30369-0
DOI: 10.1016/j.canlet.2018.05.036
Reference: CAN 13920

To appear in: Cancer Letters

Received Date: 1 April 2018

Revised Date: 22 May 2018
Accepted Date: 23 May 2018

Please cite this article as: Y.-H. Lee, N.-Y. Song, J. Suh, D.-H. Kim, W. Kim, J. Ann, J. Lee, J.-H. Baek,
H.-K. Na, Y.-J. Surh, Curcumin suppresses oncogenicity of human colon cancer cells by covalently
modifying the cysteine 67 residue of SIRT1, Cancer Letters (2018), doi: 10.1016/j.canlet.2018.05.036.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Curcumin suppresses oncogenicity of human colon cancer cells by covalently modifying

the cysteine 67 residue of SIRT1

Yeon-Hwa Lee, Na-Young Song, Jinyoung Suh, Do-Hee Kim, Wonki Kim, Jihyae Ann,

Jeewoo Lee, Jeong-Heum Baek, Hye-Kyung Na, and Young-Joon Surh*

PT
ABSTRACT

RI
SIRT1, an NAD+-dependent histone/protein deacetylase, has diverse physiological

SC
actions. Recent studies have demonstrated that SIRT1 is overexpressed in colorectal cancer,

suggesting its oncogenic potential. However, the molecular mechanisms by which

U
overexpressed SIRT1 induces the progression of colorectal cancer and its inhibition remain
AN
largely unknown. Curcumin (diferuloymethane), a major component of the spice turmeric

derived from the plant Curcuma longa L., has been reported to exert chemopreventive and
M

anti-carcinogenic effects on colon carcinogenesis. In the present study, we found that

D

curcumin reduced the expression of SIRT1 protein without influencing its mRNA expression

in human colon cancer cells, suggesting posttranslational regulation of SIRT1 by this

TE

phytochemical. Notably, ubiquitination and subsequent proteasomal degradation of SIRT1

EP

were induced by curcumin treatment. Results of nano-LC-ESI-MS/MS revealed the direct

binding of curcumin to cysteine 67 of SIRT1. In line with this result, the protein stability and
C

clonogenicity of a mutant SIRT1 in which cysteine 67 was substituted by alanine were

AC

unaffected by curcumin. Taken together, these observations suggest that curcumin facilitates

the proteasomal degradation of oncogenic SIRT1 through covalent modification of SIRT1 at

the cysteine 67 residue.

 ACCEPTED MANUSCRIPT

1 (Revised manuscript_Clean copy)

3 Curcumin suppresses oncogenicity of human colon cancer cells by covalently modifying the

4 cysteine 67 residue of SIRT1

PT
5

RI
6 [Running title: Curcumin inactivates SIRT1 through direct interaction]

SC
8 Yeon-Hwa Leea, Na-Young Songa, Jinyoung Suha, Do-Hee Kima, Wonki Kima, Jihyae Anna,

9 Jeewoo Leea, Jeong-Heum Baekb, Hye-Kyung Nac, and Young-Joon Surha,d,e,*

10
U
AN
a
11 Tumor Microenvironment Global Core Research Center and Research Institute of

12 Pharmaceutical Sciences, College of Pharmacy, Seoul National University, Seoul 08826,

M

13 Republic of Korea.
D

b
14 Division of Colon and Rectal Surgery, Department of Surgery, Gachon University Gil Medical
TE

15 Center, Incheon 21565, Republic of Korea

c
16 Department of Food Science and Biotechnology, College of Knowledge-Based Services
EP

17 Engineering, Sungshin Women’s University, Seoul 01133, Republic of Korea

d
18 Cancer Research Institute, Seoul National University, Seoul 03080, Republic of Korea.
C

e
19 Department of Molecular Medicine and Biopharmaceutical Sciences, College of Pharmacy,
AC

20 Seoul National University, Seoul 08826, Republic of Korea.

21 _______________

22 *Corresponding author. E-mail address: surh@snu.ac.kr (Y.-J. Surh)

23

1
 ACCEPTED MANUSCRIPT

24

25 ABSTRACT

26 SIRT1, an NAD+-dependent histone/protein deacetylase, has diverse physiological actions.

27 Recent studies have demonstrated that SIRT1 is overexpressed in colorectal cancer, suggesting

PT
28 its oncogenic potential. However, the molecular mechanisms by which overexpressed SIRT1

RI
29 induces the progression of colorectal cancer and its inhibition remain largely unknown.

30 Curcumin (diferuloymethane), a major component of the spice turmeric derived from the plant

SC
31 Curcuma longa L., has been reported to exert chemopreventive and anti-carcinogenic effects on

32 colon carcinogenesis. In the present study, we found that curcumin reduced the expression of

33
U
SIRT1 protein without influencing its mRNA expression in human colon cancer cells, suggesting
AN
34 posttranslational regulation of SIRT1 by this phytochemical. Notably, ubiquitination and

35 subsequent proteasomal degradation of SIRT1 were induced by curcumin treatment. Results of

M

36 nano-LC-ESI-MS/MS revealed the direct binding of curcumin to cysteine 67 of SIRT1. In line

D

37 with this result, the protein stability and clonogenicity of a mutant SIRT1 in which cysteine 67
TE

38 was substituted by alanine were unaffected by curcumin. Taken together, these observations

39 suggest that curcumin facilitates the proteasomal degradation of oncogenic SIRT1 through
EP

40 covalent modification of SIRT1 at the cysteine 67 residue.

41
C

42 Keywords: Colorectal cancer; SIRT1; Curcumin; Tetrahydrocurcumin; Thiol modification

AC

43

44

45

46

2
 ACCEPTED MANUSCRIPT

47

48 1. Introduction

49 Colorectal cancer (CRC) is the most common malignancy worldwide. It is the third leading

50 cause of cancer-related death in both men and women in the United States [1]. It has been widely

PT
51 believed that frequent consumption of Western diet, especially red meat and saturated fats,

RI
52 increases the risk of CRC. In contrast, the incidence of CRC in India is relatively low.

53 Epidemiologic studies suggest that dietary phytochemicals, such as curcumin abundant in Indian

SC
54 food, may be responsible for the low risk of intestinal cancer in India [2].

55 Curcumin, a yellow pigment present in the rhizome of turmeric (Curcuma longa L.,

56
U
Zingiberaceae) used in preparing the curry powder, is one of the most extensively investigated
AN
57 phytochemicals in regard to chemopreventive and antitumorigenic potential. Curcumin

58 modulates the expression or activity of multiple proteins involved in cellular signal transduction
M

59 pathways. It has been shown to block the tumor necrosis factor-alpha-induced nuclear
D

60 translocation and DNA binding of NF-κB in human myelomonoblastic leukemia (ML-1a) cells
TE

61 [3]. Curcumin also suppressed the expression of the c-Fos subunit of AP-1 and its DNA binding

62 activity in HPV-18-positive HeLa cells [4].

EP

63 The SIRT protein family of NAD+-dependent class Ⅲ histone deacetylase consists of seven

64 isoforms (SIRT1-7). Among these, SIRT1 was first reported to extend the life span of budding
C

65 yeast [5]. Subsequent studies demonstrated diverse physiological actions of SIRT1, such as
AC

66 metabolic regulation, differentiation and stress response. However, the role of SIRT1 in

67 tumorigenesis is still controversial. It has been shown that SIRT1 exerts tumor suppressive

68 effects by deacetylating the lysine 310 residue of RelA/p65 subunit of NF-κB, a tumor promoting

69 transcription factor often overexpressed in cancer cells. This leads to decreased production of the

3
 ACCEPTED MANUSCRIPT

70 anti-apoptotic cIAP-2 gene, sensitizing human non-small cell lung cancer (NCI-H1299) cells to

71 apoptosis [6]. Moreover, SIRT1 reduces TGF-β-driven epithelial-to-mesenchymal transition of

72 V12H-Ras transformed human breast epithelial (HMLER) cells by deacetylating Smad4 and

73 repressing its transcriptional activity toward the MMP7. As a result, the epithelial marker E-

PT
74 cadherin is less susceptible to cleavage from the cell surface [7]. On the other hand, there are

RI
75 some reports suggesting that SIRT1 may act as a tumor promoter via deacetylation and

76 inactivation of the p53 suppressor protein. Thus, deacetylation of p53 at lysine 382 by SIRT1

SC
77 causes attenuation of p53-dependent apoptosis in response to DNA damage in human non-small

78 cell lung cancer (NCI-H1299) and human breast cancer (MCF-7L) cells [8]. Further, it has been

79
U
suggested that SIRT1 promotes the development of prostate cancer through deacetylation-
AN
80 dependent inactivation of forkhead box protein O1 (FOXO1), a transcription factor involved in

81 cell cycle arrest, apoptosis and detoxification of reactive oxygen species [9].
M

82 Our previous studies have revealed that SIRT1 is highly expressed in human colon tumor
D

83 tissues compared with adjacent normal tissues [10], which is implicated in the progression of
TE

84 CRC [11]. Here, we report that curcumin directly binds to SIRT1 and decreases its stability in

85 human colon cancer (HCT-116) cells, thereby suppressing its oncogenicity.

EP

86

87 2. Materials and methods

C

88 Reagents
AC

89 Curcumin was purchased from LKT Laboratories, Inc (St. Paul, MN, USA).

90 Tetrahydrocurcumin prepared by a conventional catalytic hydrogenation of curcumin was kindly

91 provided by Prof. Jeewoo Lee of Seoul National University. Dulbecco's Modified Eagle's

92 Medium (DMEM), Minimum Essential Medium (MEM), RPMI 1640 medium and fetal bovine

4
 ACCEPTED MANUSCRIPT

93 serum were purchased from Gibco-BRL (by Thermo Fisher Scientific Inc.; Waltham, MA, USA).

94 MG-132 was obtained from Enzo Life Sciences (Exeter, UK). CNBr-Sepharose 4B was supplied

95 by Amersham Biosciences (Buckinghamshire, UK). Matrigel® Growth Factor Reduced (GFR)

96 Basement Membrane Matrix was obtained from CORNING Incorporated (Bedford, MA, USA).

PT
97

RI
98 Cell culture

99 HCT-116 cells were maintained routinely in DMEM, while HCT-15, DLD-1 and MCF-7

SC
100 cells were grown in RPMI 1640 medium. CCD841CoN cells were maintained routinely in MEM

101 under humidified atmosphere of 5% CO2/95% air. Each medium contains 10% fetal bovine

102
U
serum and an 100 ng/ml antibiotics mixture. All cell lines were obtained from the American Type
AN
103 Culture Collection (ATCC).

104
M

105 Transient transfection with SIRT1 siRNA

D

106 Cells were plated in 60-mm dish and grown to 60% confluence. Scrambled or SIRT1 siRNA
TE

107 (20 nM) was transfected into cells with Lipofectamine® RNAiMAX (Invitrogen by Thermo

108 Fisher Scientific Inc.; Waltham, MA, USA) according to the manufacturer's instructions. The
EP

109 target sequences for human SIRT1 siRNA #1 were sense 5'-

110 ACUUUGCUGUAACCCUGUA(dTdT)-3' and antisense 5'-

C

111 UACAGGGUUACAGCAAAGU(dTdT)-3' and for human SIRT1 siRNA #2 were sense 5′-
AC

112 AGAGUUGCCACCCACACCU(dTdT)-3′, antisense 5′-

113 AGGUGUGGGUGGCAACUCU(dTdT)-3′. After 48 h of transfection, cells were harvested.

114 siRNA oligonucleotide targeting SIRT1 was purchased from Genolution (Seoul, South Korea).

115

5
 ACCEPTED MANUSCRIPT

116 Transient transfection of His-Ub, FLAG-SIRT1-WT and FLAG-SIRT1-C67A plasmid

117 Transient transfections with plasmid vector were performed using Lipofectamine® 2000

118 (Invitrogen by Thermo Fisher Scientific Inc.; Waltham, MA, USA) according to the instructions

119 supplied by the manufacturer. After 48 h of transfection, cells were harvested or treated with

PT
120 DMSO or curcumin according to purpose of the experiment.

RI
121

122 MTT assay

SC
123 For the MTT assay, cells were seeded in 48-well plates at a density of 0.5 × 104 cells/well.

124 After incubation for the indicated period of time, cells were treated with the MTT solution (1

125
U
mg/ml) for 2 h. The resultant formazan crystals formed in intact cells were dissolved in DMSO,
AN
126 and the absorbance at 570 nm was read using a microplate reader.

127
M

128 Wound healing assay

D

129 For the wound healing assay, cells were plated into Culture-insert (Ibidi; Martinsried,
TE

130 Germany) to make a 500 µm gap. After detaching the insert from a dish, cells were grown for the

131 indicated period of time and then photographed under the microscope.
EP

132

133 Anchorage-independent growth assay

C

134 For the anchorage-independent growth assay, 6-well plates or 60-mm dishes were pre-
AC

135 coated with 0.5% agarose as the bottom layer. On top of the bottom layer, cells mixed with 0.3%

136 agarose were seeded. Cells were maintained in semi-solid medium for 14 days, and colonies

137 were stained with 0.05% crystal violet. The number of colonies larger than 100 µm was counted

138 under a microscope.

6
 ACCEPTED MANUSCRIPT

139

140 Nude mouse xenograft assay

141 For the nude mouse xenograft assay, four-week-old male BALB/c nude mice (total 42) were

142 purchased from RaonBio, Inc (Yongin, South Korea) and were maintained under specific

PT
143 pathogen-free (SPF) conditions with a 12-h light/12-h dark cycle. After one week of acclimation

RI
144 period, 5 × 106 scrambled or SIRT1 siRNA transfected HCT-116 cells re-suspended in equal

145 volumes of phosphate-buffered saline (PBS) and matrigel (total volume of 200 µl) were

SC
146 subcutaneously injected into the flanks of mice to generate colon cancer xenograft tumors (n = 6

147 per group). For another xenograft assay, 5 × 106 HCT-116 cells were prepared in the same way as

148
U
described above and injected into the flanks of mice. When the tumor size reached approximately
AN
149 150 mm3, mice were randomized into control and treatment groups (50 mg/kg and 100 mg/kg of

150 curcumin or 50 mg/kg and 100 mg/kg of tetrahydrocurcumin). Curcumin or tetrahydrocurcumin

M

151 dissolved in corn oil was peritumorally administered to mice for a total of eight times during
D

152 three weeks (n = 6 per group). Tumor volume was regularly measured with digital calipers and
TE

153 calculated according to the formula, V= 0.5ab2, where ‘a’ is the longest and ‘b’ is the shortest

154 perpendicular diameters. After mice were killed, xenograft tumors were excised and fixed in
EP

155 formalin for further analysis. All experimental protocols for animal experiments were approved

156 by the Institutional Animal Care and Use Committee (IACUC) of Seoul National University
C

157 (authorization number: SNU-160912-1).

AC

158

159 Measurement of SIRT1 catalytic activity

160 SIRT1 deacetylase activity was assessed by using the fluorometic activity assay kit (Sigma-

161 Aldrich; St. Louis, MO, USA) according to the manufacturer's protocol. Samples were

7
 ACCEPTED MANUSCRIPT

162 transferred to the Costar 96-well dark plate, and the deacetylation-dependent fluorescent signal

163 was detected by using a fluorescent reader with 360 nm excitation and 460 nm emission

164 wavelengths.

165

PT
166 Western blot analysis

RI
167 Lysates from cells were separated by running through 6-10% sodium dodecyl sulfate

168 polyacrylamide gel (SDS-PAGE) and transferred to the polyvinylidene difluoride membranes

SC
169 according to the protocol described elsewhere [12]. The membranes were then incubated with

170 primary antibodies against SIRT1, His tag, actin (Santa Cruz Biotechnology, Inc.; Dallas, TX,

171
U
USA), ubiquitin (Life Technolgies by Thermo Fisher Scientific Inc.; Waltham, MA, USA),
AN
172 acetyl-p53, phospho-SIRT1, phospho-SAPK/JNK, SAPK/JNK (Cell Signaling Technology;

173 Danvers, MA, USA) and FLAG (Sigma-Aldrich; St. Louis, MO, USA). Blots were washed with
M

174 TBST and then probed with horseradish peroxidase-conjugated secondary antibodies (Pierce
D

175 Biotechnology; Rockford, IL, USA). The transferred proteins were visualized by using a Western
TE

176 blotting detection reagent kit (AbClon; Seoul, South Korea).

177
EP

178 Reverse transcription-polymerase chain reaction (RT-PCR) and real-time qPCR

179 Total RNA was isolated from HCT-116, DLD-1, HCT-15 and MCF-7 cells by using the
C

180 TRIzol® reagent (Invitrogen by Thermo Fisher Scientific Inc.; Waltham, MA, USA), and RT-
AC

181 PCR was performed according to the standard protocol. The primers used for the reverse

182 transcription-PCR are as follows (forward and reverse, respectively): for SIRT1, 5′-

183 GCAACATCTTATGATTGGCAC-3′ (position 620 to 640) and 5′-

184 AAATACCATCCCTTGACCTGAA-3′ (position 891 to 870) with a product size of 272 bp. For

8
 ACCEPTED MANUSCRIPT

185 GAPDH, 5′-GCATGGCCTTCCGTGTCCCC-3′ (position 857 to 876) and 5′-

186 CAATGCCAGCCCCAGCGTCA-3′ (position 1072 to 1053) with a product size of 216 bp. For

187 the real-time PCR, a RealHelix™ SYBR Green I qPCR kit (NanoHelix Co., Ltd.; Seoul, South

188 Korea) was used. The primers used for the real-time qPCR are as follows: for SIRT1, 5'-

PT
189 AGAAGAACCCATGGAGGATG-3' (position 1391 to 1410), and 5'-

RI
190 TCATCTCCATCAGTCCCAAA-3' (position 1504 to 1485) with a product size of 114 bp. For

191 RPL32, 5'-CTCTTTCCACGATGGCTTTG-3' (position 437 to 418) and 5'-

SC
192 GTCAAGGAGCTGGAAGTGCT-3' (position 338 to 357) with a product size of 100 bp. PCR

193 products were measured by 7500 Fast Real-Time PCR System (Applied Biosystems by Thermo

194
U
Fisher Scientific Inc.; Waltham, MA, USA) using fluorescent SYBR Green I dye.
AN
195

196 Immunofluorescent analysis

M

197 Human paraffin-embedded colon cancer tissue array with matched adjacent normal colon
D

198 tissue (US Biomax, Inc., cat. no. CO703; Rockville, MD, USA) was subjected to
TE

199 deparaffinization with xylene. Following antigen retrieval by heated citrate buffer, sections were

200 permeabilized and blocked according to the standard protocol. After overnight incubation at 4 °C
EP

201 with anti-SIRT1 antibody, the tissue sections were washed with PBS and then labeled with

202 TRITC-conjugated secondary antibody for 1 h at room temperature. The slides were then
C

203 analyzed under a fluorescent microscope.

AC

204

205 Immunocytochemical analysis

206 Cells grown on coverslips were prepared as described previously [13]. Cells were then

207 blocked with PBS with Tween-20 containing 5% bovine serum albumin for 1 h at room

9
 ACCEPTED MANUSCRIPT

208 temperature and incubated with an antibody against SIRT1 overnight. After rinse with PBS, cells

209 were labeled with FITC-conjugated secondary antibody for 1 h at room temperature. Finally,

210 cells were further stained with propidium iodide (PI; Molecular Probes®; Eugene, OR, USA) to

211 visualize the nuclei. The slides were then analyzed under a fluorescent microscope.

PT
212

RI
213 Preparation of cytosolic and nuclear proteins

214 Confluent cells in 100 mm dishes were treated with curcumin and then harvested. Pellets

SC
215 were suspended in hypotonic lysis buffer to obtain cytosolic extract, and then resulting pellets

216 were suspended in hypertonic lysis buffer to collect the nuclear extract according to the

217
U
previously reported procedure [14]. Each separately obtained supernatant containing cytosolic or
AN
218 nuclear proteins was stored at -70°C after determination of the protein concentration.

219
M

220 Immunoprecipitation
D

221 For ubiquitination of SIRT1, cells were transfected with FLAG-SIRT1 and His-ubiquitin
TE

222 plasmids and incubated with MG-132 (20 µM) and curcumin (10 µM) for 3 h. Lysates from cells

223 were prepared as described earlier [14]. After resolving the samples by SDS-PAGE, membranes
EP

224 were incubated with primary antibody to His tag.

225
C

226 Preparation of curcumin- or tetrahydrocurcumin-Sepharose 4B beads

AC

227 For the pull down assay, curcumin- or tetrahydrocurcumin-conjugated Sepharose 4B beads

228 were prepared according to the protocol described by Urusova et al [15]. Proteins bound to the

229 beads were analyzed by immunoblotting with anti-SIRT1 or FLAG antibody.

230

10
 ACCEPTED MANUSCRIPT

231 Analysis by nano-LC-ESI-MS/MS

232 Recombinant SIRT1 protein (Active Motif, cat. no. 31340; Carlsbad, CA, USA) at 50 nM in

233 a final volume of 50 µl 20 mM Tris-HCl (pH, 8.0) containing 100 mM KCl, 1 mM dithiothreitol

234 (DTT), 0.2 mM EDTA and 20% glycerol was incubated for 2 h at room temperature in the

PT
235 presence of DMSO or curcumin (10 µM). The products of in-gel digestion of recombinant SIRT1

RI
236 pre-incubated with or without curcumin were processed according to procedure as previously

237 described [16]. The acquired LC-ESI-MS/MS fragment spectra were searched in the

SC
238 BioWorksBrowserTM (version Rev. 3.3.1 SP1; Thermo Fisher Scientific Inc., San Jose, CA,

239 USA) with the SEQUEST search engines against the data in FASTA format generated from

240
U
SIRT1 (NCBI accession number NM_012238) in National Center for Biotechnology Information
AN
241 (http://www.ncbi.nlm.nih.gov/).

242
M

243 Site-directed mutagenesis

D

244 Mutations in wild-type SIRT1 were introduced by site-directed mutagenesis using a

TE

245 QuikChange® site-directed mutagenesis kit (Stratagene; Cedar Creek, TX, USA) according to the

246 manufacturer's protocol. Mutant strand synthesis reaction was performed by denaturation at 96°C
EP

247 for 45 sec, annealing at 60°C for 45 sec and extension at 72°C for 8 min with 30 cycles using 5.4

248 kb plasmid template encoding FLAG-SIRT1. Generally,

C

249 extension time of 1 min/kb is recommended depending on the length of plasmid template and
AC

250 extension time of 8 minutes was set as an optimal condition. The following complementary

251 primer pairs were used: Cys67Ala: sense, 5′-

252 GCGGCGGCCAGGGGCGCCCCGGGTGCGGCGGCG-3′; antisense, 5′-CGC

253 CGCCGGTCCCCGCGGGGCCCACGCCGCCGC-3′. The DNA sequences of all plasmids were

11
 ACCEPTED MANUSCRIPT

254 verified by sequencing (Cosmo Genetech, Seoul, South Korea).

255

256 Statistical analysis

257 The statistical significance was determined using the Student's t-test. Differences between

PT
258 groups were considered statistically significant if p < 0.05.

RI
259

260 3. Results

SC
261 SIRT1 is overexpressed in human CRC cell lines as well as human colon tumor tissues.

262 To determine the protein expression of SIRT1 in human CRC tissues compared with

263
U
adjacent normal colon tissues, immunofluorescent analysis was performed on human CRC tissue
AN
264 microarray. The data revealed that 27 out of 35 samples (77.14%) showed upregulation of SIRT1

265 in human colon tumor tissues compared with that of corresponding adjacent normal colon tissues,
M

266 and representative images of immunofluorescent staining of SIRT1 are shown in Fig. 1A.
D

267 Likewise, the protein levels of SIRT1 were upregulated in the majority of CRC cell lines we
TE

268 examined compared with normal human colon epithelial (CCD841CoN) cells (Fig. 1B). To

269 verify the role of SIRT1 in viability and migration of colon cancer cells, SIRT1 was silenced in
EP

270 human CRC cell lines by transfection with SIRT1-specific small interfering RNA (siRNA). As

271 illustrated in Fig. 1C and D, knockdown of SIRT1 led to the decreased viability and migration
C

272 capacity of CRC cell lines.

AC

273

274 Knockdown of SIRT1 attenuates tumor growth in vivo.

275 To further evaluate the effect of SIRT1 on tumor formation in vivo, a xenograft mouse

276 model was employed using subcutaneous injections of control or SIRT1-silenced cells. The

12
 ACCEPTED MANUSCRIPT

277 representative image of the excised tumors at three weeks after inoculation is presented in Fig.

278 2A. The result indicated that the SIRT1 siRNA group exhibited a lower tumor growth rate than

279 the scrambled siRNA group (Fig. 2B). Silencing of SIRT1 resulted in a significant decrease of

280 the tumor volume [mean ± SD (mm3); 410.29 ± 297.343 (scrambled siRNA) vs 94.63 ± 40.37

PT
281 (SIRT1 siRNA)] (Fig. 2C) and the weight [mean ± SD (mg); 333.33 ± 287.52 (scrambled siRNA)

RI
282 vs 66.67 ± 51.64 (SIRT1 siRNA)] (Fig. 2D) as measured at the time of excision. As depicted in

283 Fig. 2E, the hematoxylin and eosin (H&E) stained sections of resected tumor tissues showed that

SC
284 polygonal or spindle-shaped tumor cells with hyperchromatic nuclei were less frequent in

285 SIRT1-silenced xenograft tumor compared with those observed in control xenograft tumor.

286
U
Immunohistochemical analysis also showed a relatively weak brown nuclear staining of
AN
287 proliferating cell nuclear antigen (PCNA) in SIRT1-silenced xenograft tumor, indicative of

288 reduced cellular proliferation (Fig. 2F). Moreover, knockdown of SIRT1 was associated with
M

289 decreased protein levels of other cellular proliferation indices, such as cyclin D1 and c-Myc (Fig.
D

290 2G). Thus, the results from the xenograft mouse model parallel those observed in the in vitro
TE

291 assays shown in Fig. 1C and D, corroborating the oncogenic function of SIRT1 in the

292 progression of colon cancer.

EP

293

294
C

295 Curcumin reduces the expression of SIRT1 protein without influencing its mRNA
AC

296 expression in colon cancer cells.

297 To investigate the ability of curcumin to regulate SIRT1 in colon cancer cells, we first

298 examined whether curcumin could affect the protein expression of SIRT1. When HCT-116 cells

299 were treated with curcumin (1 or 10 µM) for 3 h, the protein level of SIRT1 decreased

13
 ACCEPTED MANUSCRIPT

300 significantly at the higher concentration (Fig. 3A). Consistent with this finding, catalytic activity

301 of SIRT1 was gradually diminished upon curcumin treatment (Fig. 3B). However, curcumin had

302 no significant effect on the protein expression or the catalytic activity of SIRT1 in normal human

303 colon epithelial (CCD841CoN) cells (Supplementary Fig. 1A and B, respectively). When HCT-

PT
304 116 cells were exposed to 10 µM curcumin for various time periods, there was the time-

RI
305 dependent inhibition of SIRT1 protein expression as measured by Western blot analysis (Fig. 3C).

306 However, the expression of its mRNA transcript remained unchanged by curcumin treatment as

SC
307 revealed by reverse transcription-PCR analysis (Fig. 3D with the quantitative data in

308 Supplementary Fig. 1C). This finding was verified by Real-time qPCR analysis

309
U
(Supplementary Fig. 1D). Similar observations were made in other human CRC cell lines as
AN
310 well as human breast cancer MCF-7 cells (Supplementary Fig. 1E). Immunoblot (Fig. 3E) and

311 immunocytochemical (Fig. 3F) analyses further revealed that curcumin markedly reduced the
M

312 cytosolic accumulation of SIRT1.

D

313
TE

314 Curcumin stimulates ubiquitin-dependent proteasomal degradation, thereby decreasing the

315 protein stability of SIRT1.

EP

316 As curcumin decreased the protein expression of SIRT1 without affecting its mRNA

317 expression, we postulated that curcumin could decrease the SIRT1 protein stability. To test this
C

318 supposition, HCT-116 cells were pre-incubated with cycloheximide (CHX), a protein synthesis
AC

319 inhibitor, prior to exposure to curcumin. Treatment of HCT-116 cells with curcumin significantly

320 shortened the half-life of SIRT1 (Fig. 4A), suggesting that curcumin destabilizes SIRT1 by

321 enhancing its degradation rather than inhibiting de novo synthesis. Notably, the inhibitory effect

322 of curcumin on SIRT1 protein accumulation was blunted by a proteasome-specific inhibitor MG-

14
 ACCEPTED MANUSCRIPT

323 132 (Fig. 4B). MG-132 alone also enhanced the basal level of SIRT1 protein, indicating that

324 proteasomes are required for both steady-state and curcumin-induced degradation of SIRT1.

325 Curcumin-induced proteasomal degradation of SIRT1 was associated with its ubiquitination (Fig.

326 4C). Moreover, the ubiquitination of exogenous SIRT1 was also enhanced by curcumin in HCT-

PT
327 116 cells (Fig. 4D).

RI
328

329 The α,β-unsaturated carbonyl group of curcumin plays an important role in reducing the

SC
330 stability of SIRT1 as well as growth of HCT-116 tumor xenografts.

331 Several studies have indicated that anti-inflammatory and anti-carcinogenic properties of

332
U
curcumin are attributed to its electrophilic α,β-unsaturated carbonyl moiety [17-19]. This
AN
333 prompted us to compare curcumin (Fig. 5A, left) and its non-electrophilic analogue

334 tetrahydrocurcumin (Fig. 5A, right) for their effects on SIRT1 stability. While curcumin
M

335 markedly reduced the expression level of SIRT1 protein, tetrahydrocurcumin had no effect (Fig.
D

336 5B). In line with this observation, tetrahydrocutcumin barely induced ubiquitination of SIRT1
TE

337 (Fig. 5C). Curcumin has been reported to react directly with intracellular proteins through its

338 α,β-unsaturated carbonyl group [20-23]. A pull-down experiment using curcumin- or

EP

339 tetrahydrocurcumin-conjugated Sepharose 4B beads showed that curcumin bound to SIRT1, but

340 tetrahydrocurcumin only weakly interacted with SIRT1 (Fig. 5D). These findings suggest that
C

341 the α,β-unsaturated carbonyl group of curcumin is essential for its direct interaction with SIRT1,
AC

342 which appeared to facilitate proteasomal degradation. In parallel with these results, curcumin

343 apparently diminished the anchorage-independent growth (Fig. 5E) and migration

344 (Supplementary Fig. 2A) of HCT-116 cells, whereas tetrahydrocurcumin was not effective. To

345 explore the effects of curcumin and its non-electrophilic analogue on tumor growth in vivo, mice

15
 ACCEPTED MANUSCRIPT

346 bearing HCT-116 xenografts were treated with peritumoral injection (50 or 100 mg/kg) of each

347 compound for a total of eight times during three weeks. As depicted in Fig. 5F, curcumin is

348 much more effective than tetrahydrocurcumin in retarding tumor growth. Injection of 50 or 100

349 mg/kg of curcumin exhibited significant reduction in the tumor volume [mean ± SD (mm3);

PT
350 533.12 ± 100.23 (vehicle), 305.52 ± 147.17 (curcumin, 50 mg/kg), 125.50 ± 49.67 (curcumin,

RI
351 100 mg/kg), 500.18 ± 99.45 (tetrahydrocurcumin, 50 mg/kg), 501.23 ± 123.43

352 (tetrahydrocurcumin, 100 mg/kg)] (Fig. 5G) and the weight [mean ± SD; 716.67 ± 75.28

SC
353 (vehicle) 350.17 ± 176.01 (curcumin, 50 mg/kg), 200.33 ± 89.00 (curcumin, 100 mg/kg), 616.67

354 ± 160.21 (tetrahydrocurcumin, 50 mg/kg), 600.50 ± 89.45 (tetrahydrocurcumin, 100 mg/kg)]

355
U
(Fig. 5H). Histological examination of tumor tissues from the mice treated with curcumin
AN
356 exhibited necrotic cell death (Fig. 5I, upper panels) whereas those from mice treated with an

357 100 mg/kg dose of tetrahydrocurcumin showed slight reduction in hyperchromatic tumor cells
M

358 (Supplementary Fig. 2B). Concomitantly, immunohistochemical staining of SIRT1 in the

D

359 xenograft tumors from the mice treated with curcumin showed a relatively weak intensity
TE

360 compared with those of control or tetrahydrocurcumin group (Fig. 5I, lower panels). The

361 inhibitory effect of curcumin on SIRT1 protein expression in HCT-116 tumor xenografts was
EP

362 verified by Western blot analysis (Fig. 5J with the quantitative data in Supplementary Fig. 2C).

363
C

364 Cysteine 67 of SIRT1 is a putative target for its covalent modification by curcumin.
AC

365 Nucleophilic cysteine thiols are readily attacked by the reactive electrophilic α,β-

366 unsaturated carbonyl group in the Michael addition reaction [24, 25]. To verify the possible

367 involvement of thiol modification of a specific cysteine residue(s) of SIRT1 by curcumin, HCT-

368 116 cells were treated with curcumin in the presence or absence of the thiol reducing agent DTT

16
 ACCEPTED MANUSCRIPT

369 or N-acetyl-L-cysteine (NAC). The result indicated that both DTT and NAC dampened

370 curcumin-mediated suppression of SIRT1 (Supplementary Fig. 3A and B, respectively). These

371 observations imply that one or more of SIRT1 cysteine residue(s) is/are susceptible to covalent

372 modification by curcumin, which facilitates proteasomal degradation of SIRT1. To further

PT
373 investigate the modifiable residue(s) of SIRT1 by curcumin, recombinant human SIRT1 was

RI
374 incubated with curcumin followed by nano-LC-ESI-MS/MS analysis. The data revealed

375 modification of the cysteine 67 residue by curcumin (Fig. 6A). A peak of the 755.28 Da

SC
376 molecular mass was detected only in curcumin-treated sample in addition to a peak of 386.15 Da

377 in the control. The difference in masses between the curcumin treated sample and the DMSO

378
U
treated sample was 368.38, which corresponds to the molecular weight of curcumin. To verify
AN
379 the significance of cysteine 67 in interaction between curcumin and SIRT1, cysteine to alanine

380 substitution mutant (SIRT1-C67A) was generated by site-directed mutagenesis. The resulting
M

381 mutant SIRT1 construct was transfected into HCT-116 cells. Subcellular localization (Fig. 6B)
D

382 and deacetylase activity (Fig. 6C and Supplementary Fig. 3C) of SIRT1 were not affected by
TE

383 cysteine 67 to alanine substitution. As shown in Fig. 6D, curcumin failed to bind to SIRT1

384 mutated at cysteine 67. Consistent with a result in HCT-116 cells, the same mutation (SIRT1-
EP

385 C67A) in DLD-1 cells also hampered the covalent interaction between curcumin and SIRT1

386 (Supplementary Fig. 3D).

C

387
AC

388 Cysteine 67 modification by curcumin induces conformational changes in SIRT1 that

389 would affect its protein stability.

390 We next examined the involvement of cysteine 67 in curcumin-induced destabilization of

391 SIRT1 protein. When the protein level of SIRT1 was measured, HCT-116 cells transfected with

17
 ACCEPTED MANUSCRIPT

392 SIRT1-WT showed decreased stability of SIRT1 upon curcumin treatment. However, the protein

393 level of SIRT1 in HCT-116 cells harbouring mutant SIRT1 (SIRT1-C67A) was not diminished in

394 the presence of curcumin (Fig. 7A). Consistent with this finding, curcumin barely induced the

395 ubiquitination of SIRT1 in HCT-116 cells expressing SIRT1-C67A (Fig. 7B). The above results

PT
396 indicate that binding of curcumin to cysteine 67 of SIRT1 is important in the ubiquitination-

RI
397 dependent degradation of SIRT1. In parallel with these findings, anchorage-independent growth

398 of cells expressing cysteine 67 mutant SIRT1 was affected by curcumin to a lesser extent than

SC
399 cells expressing SIRT1-WT (Fig 7C). Similar result was also seen in the wound healing assay

400 (Supplementary Fig. 4).

401
U
AN
402 4. Discussion

403 Meta-analysis of gastrointestinal cancers revealed that the expression of SIRT1 was not
M

404 correlated with worse overall survival in CRC [26]. However, another meta-analysis showed that
D

405 the expression of SIRT1 is associated with the poor prognosis in CRC patients [27]. While there
TE

406 is some discrepancy regarding the role of SIRT1 in tumor progression, recent studies have

407 revealed that SIRT1 is abnormally overexpressed in CRC tissues [11]. Notably, the elevated
EP

408 levels of SIRT1 are significantly associated with not only tumor invasion and lymph node

409 metastasis [28-30], but higher grades of malignancy and mutations of KRAS and BRAF [31],
C

410 suggesting the prognostic utility of SIRT1.

AC

411 In a phase IIa cancer prevention trial with participants with rectal aberrant crypt foci

412 (ACF), daily administration (p.o.) of 4 g curcumin for 30 days significantly reduced ACF

413 formation after 30 days of daily oral administration, resulting in 7.3 ± 8.1 ng/ml (approximately

414 20 nM) venous plasma concentration of curcumin [32]. In a phase I clinical study, a daily oral

18
 ACCEPTED MANUSCRIPT

415 consumption of 3.6 g curcumin for up to 4 months resulted in a plasma concentration of 11.1 nM

416 and also decreased serum levels of prostaglandin E2 in the CRC patients [33]. In another phase Ⅰ

417 clinical trial, oral consumption of 4-8 g of curcumin for 3 months resulted in a peak plasma level

418 of 0.51-1.77 µM in volunteers with high risk conditions including intestinal metaplasia [34].

PT
419 Such low bioavailability of curcumin is attributable to its poor absorption from the gut.

RI
420 Curcumin in cultured cells also undergoes rapid metabolism. Thus, only less than 2% of

421 curcumin was detected after 3 h incubation in human colon cancer (Caco-2) cells [35].

SC
422 It has been reported that curcumin can activate or upregulate SIRT1 in some cases [36-39].

423 Curcumin has been reported to exert preventive effects against pathogenesis of several disorders

424
U
by activating SIRT1 in cells or tissues of noncancerous origin. When normal isolated rat hearts
AN
425 were exposed to ischemia reperfusion (IR) injury, the expression of SIRT1 decreased. However,

426 curcumin pretreatment protects the SIRT1 from IR-induced loss in isolated hearts [36]. In
M

427 another study, the level of SIRT1 in cultured neurons isolated from healthy Sprague–Dawley rats
D

428 was substantially decreased by Aβ25–35, a neurotoxic peptide accumulated in the brain of patients
TE

429 with Alzheimer’s disease. Curcumin pretreatment attenuated the loss of cellular accumulation of

430 SIRT1, resulting in enhanced neuronal survival against Aβ25–35-induced toxicity [37]. Curcumin
EP

431 induced upregulation of SIRT1 in vascular smooth muscle (VSMC) cells, which was speculated

432 to account for the anti-aging effect of curcumin in these cells [39]. The above findings suggest
C

433 involvement of SIRT1 activation in health benefits of curcumin. Interestingly, even in malignant
AC

434 cells, activation of SIRT1 contributes to anticarcinogenic effect of curcumin. For instance,

435 curcumin induces ATM-mediated apoptotic signaling through activation of SIRT1 in head and

436 neck squamous cell carcinomas (FaDu and Cal27) cells [38].

437 In contrast to the aforementioned findings, we found that the protein levels of SIRT1

19
 ACCEPTED MANUSCRIPT

438 declined upon treatment with curcumin in 3 different human CRC cell lines (HCT-116, DLD-1

439 and HCT-15). Interestingly, the SIRT1 protein level in normal colonic epithelial cells was not

440 affected by curcumin. Besides differences in amounts and subcellular localization of SIRT1

441 between normal and cancer cells [10, 40], there might be some distinguishable factors, such as

PT
442 intracellular redox environment surrounding the SIRT1 protein that may confer the differential

RI
443 effects of curcumin on SIRT1 protein level or activity.

444 Increased protein expression of SIRT1 in CRC and hepatocellular carcinoma is not

SC
445 attributable to elevated mRNA expression of SIRT1, suggesting that overexpression of SIRT1 is

446 correlated with its increased protein stability [41, 42]. Nonetheless, much less is known about

447
U
regulation of SIRT1 protein stability. A very few studies have suggested that deubiquitination [43]
AN
448 and phosphorylation of SIRT1 at serine 27 and 47 are associated with the enhanced SIRT1

449 protein stability [44, 45]. Substantial differences between curcumin and its non-electrophilic
M

450 analogue tetrahydrocurcumin in their effects on SIRT1 accumulation in HCT-116 cells lead us to
D

451 speculate that curcumin-induced destabilization of SIRT1 protein is attributable to its

TE

452 electrophilic nature. We note that curcumin is more effective than tetrahydrocurcumin in

453 reducing tumor growth in a murine xenograft model, highlighting the unique structural feature of
EP

454 curcumin essential for its biological activities.

455 Curcumin has two electrophilic α,β-unsaturated carbonyl groups, and hence can act as a
C

456 Michael reaction acceptor. There are several reports suggesting that curcumin can directly
AC

457 modify cysteine thiols of cellular proteins by Michael addition, thereby modulating their

458 functions [21, 23, 46]. SIRT1 contains 19 cysteine residues, and some of which are susceptible to

459 redox modification. It has been reported that the cysteine 67 residue of SIRT1 is directly

460 modified by S-nitrosoglutathione formed from nitric oxide and glutathione, which attenuates

20
 ACCEPTED MANUSCRIPT

461 activation of SIRT1 by resveratrol [47]. In addition, cysteine 371 and 374 residues present in the

462 catalytic domain of SIRT1 are targeted for reduction by redox factor-1 (Ref-1). Ref-1, as a

463 cellular reductant, stimulates SIRT1 activity by keeping SIRT1 in the reduced form [48]. It has

464 also been shown that the cysteine 482 residue of SIRT1 is covalently modified by 4-hydroxy-2-

PT
465 nonenal (4-HNE), a reactive hydroxyalkenal produced by lipid peroxidation. Carbonylation of

RI
466 SIRT1 by 4-HNE results in inhibition of deacetylase activity and protein expression of SIRT1

467 [49].

SC
468 In the current study, nano-LC-ESI-MS/MS analysis revealed that curcumin directly binds to

469 the cysteine 67 residue located in the N-terminal domain of SIRT1 which functions as a linker in

470
U
interacting with its substrate proteins [50, 51]. Our results show that cysteine 67 to alanine
AN
471 substitution renders HCT-116 colon cancer cells insensitive to curcumin-mediated inhibition of

472 their clonogenicity and proteasomal degradation of SIRT1. Thus, it is plausible that covalent
M

473 thiol modification at the cysteine 67 residue contributes to destabilization of SIRT1 by curcumin.
D

474 As replacement of the cysteine 67 residue by alanine hampers direct interaction of SIRT1 with
TE

475 curcumin, it is likely that cysteine 67 is a bona fide binding site of curcumin. We speculate that

476 direct binding of curcumin to cysteine 67 might cause conformational changes of SIRT1, leading
EP

477 to the protrusion of lysine residues that can be ubiquitinated. This allows E3 ubiquitin ligases to

478 have better access to the SIRT1. Unfortunately, the crystallographic structure of the full-length
C

479 human SIRT1 has not been solved yet, so the alteration in the structure of SIRT1 complexed with
AC

480 curcumin cannot be verified at this moment.

481 In conclusion, curcumin that contains the electrophilic α,β-unsaturated carbonyl moiety can

482 directly bind to SIRT1 protein, preferentially at the cysteine 67 residue. The resulting structural

483 modification facilitates the ubiquitin-dependent proteasomal degradation of oncogenic SIRT1

21
 ACCEPTED MANUSCRIPT

484 overexpressed in colon cancer cells (Fig. 8). This may account for suppression of CRC

485 progression by curcumin.

486

487 Conflict of interest

PT
488 The authors declare no conflict of interest.

RI
489

490 Acknowledgements

SC
491 This study was supported by the Global Core Research Center (GCRC) grant (No. 2011-

492 0030001) from the National Research Foundation (NRF) of Republic of Korea.

U
AN
493

494 References
M

495 [1] R.L. Siegel, K.D. Miller, A. Jemal, Cancer statistics, 2016, CA Cancer J Clin, 66 (2016) 7-30.

496 [2] K.M. Mohandas, D.C. Desai, Epidemiology of digestive tract cancers in India. V. Large and
D

497 small bowel, Indian J Gastroenterol, 18 (1999) 118-121.

TE

498 [3] S. Singh, B.B. Aggarwal, Activation of transcription factor NF-κB is suppressed by curcumin

499 (diferuloylmethane) [corrected], J Biol Chem, 270 (1995) 24995-25000.

EP

500 [4] B.K. Prusty, B.C. Das, Constitutive activation of transcription factor AP-1 in cervical cancer
C

501 and suppression of human papillomavirus (HPV) transcription and AP-1 activity in HeLa cells
AC

502 by curcumin, Int J Cancer, 113 (2005) 951-960.

503 [5] M. Kaeberlein, M. McVey, L. Guarente, The SIR2/3/4 complex and SIR2 alone promote

504 longevity in Saccharomyces cerevisiae by two different mechanisms, Genes Dev, 13 (1999)

505 2570-2580.

506 [6] F. Yeung, J.E. Hoberg, C.S. Ramsey, M.D. Keller, D.R. Jones, R.A. Frye, M.W. Mayo,
22
 ACCEPTED MANUSCRIPT

507 Modulation of NF-κB-dependent transcription and cell survival by the SIRT1 deacetylase,

508 EMBO J, 23 (2004) 2369-2380.

509 [7] P. Simic, E.O. Williams, E.L. Bell, J.J. Gong, M. Bonkowski, L. Guarente, SIRT1 suppresses

510 the epithelial-to-mesenchymal transition in cancer metastasis and organ fibrosis, Cell Rep, 3

PT
511 (2013) 1175-1186.

RI
512 [8] H. Vaziri, S.K. Dessain, E. Ng Eaton, S.I. Imai, R.A. Frye, T.K. Pandita, L. Guarente, R.A.

513 Weinberg, hSIR2(SIRT1) functions as an NAD-dependent p53 deacetylase, Cell, 107 (2001)

SC
514 149-159.

515 [9] B. Jung-Hynes, M. Nihal, W. Zhong, N. Ahmad, Role of sirtuin histone deacetylase SIRT1 in

516
U
prostate cancer. A target for prostate cancer management via its inhibition?, J Biol Chem, 284
AN
517 (2009) 3823-3832.

518 [10] N.Y. Song, Y.J. Surh, Janus-faced role of SIRT1 in tumorigenesis, Ann N Y Acad Sci, 1271
M

519 (2012) 10-19.

D

520 [11] N.Y. Song, Y.H. Lee, H.K. Na, J.H. Baek, Y.J. Surh, Leptin induces SIRT1 expression
TE

521 through activation of NF-E2-related factor 2: Implications for obesity-associated colon

522 carcinogenesis, Biochem Pharmacol, 153 (2018) 282-291.

EP

523 [12] D.H. Kim, N.Y. Song, E.H. Kim, H.K. Na, Y. Joe, H.T. Chung, Y.J. Surh, 15-deoxy-D12,14-

524 prostaglandin J2 induces p53 expression through Nrf2-mediated upregulation of heme

C

525 oxygenase-1 in human breast cancer cells, Free radical research, 48 (2014) 1018-1027.
AC

526 [13] N.Y. Song, H.K. Na, J.H. Baek, Y.J. Surh, Docosahexaenoic acid inhibits insulin-induced

527 activation of sterol regulatory-element binding protein 1 and cyclooxygenase-2 expression

528 through upregulation of SIRT1 in human colon epithelial cells, Biochem Pharmacol, 92 (2014)

529 142-148.

23
 ACCEPTED MANUSCRIPT

530 [14] D.H. Kim, E.H. Kim, H.K. Na, Y. Sun, Y.J. Surh, 15-deoxy-D12,14-prostaglandin J2 stabilizes,

531 but functionally inactivates p53 by binding to the cysteine 277 residue, Oncogene, 29 (2010)

532 2560-2576.

533 [15] D.V. Urusova, J.H. Shim, D.J. Kim, S.K. Jung, T.A. Zykova, A. Carper, A.M. Bode, Z.

PT
534 Dong, Epigallocatechin-gallate suppresses tumorigenesis by directly targeting Pin1, Cancer Prev

RI
535 Res (Phila), 4 (2011) 1366-1377.

536 [16] S. Kim, H.G. Lee, S.A. Park, J.K. Kundu, Y.S. Keum, Y.N. Cha, H.K. Na, Y.J. Surh, Keap1

SC
537 Cysteine 288 as a Potential Target for Diallyl Trisulfide-Induced Nrf2 Activation, Plos One, 9

538 (2014).

539
U
[17] M.T. Huang, W. Ma, Y.P. Lu, R.L. Chang, C. Fisher, P.S. Manchand, H.L. Newmark, A.H.
AN
540 Conney, Effects of curcumin, demethoxycurcumin, bisdemethoxycurcumin and

541 tetrahydrocurcumin on 12-O-tetradecanoylphorbol-13-acetate-induced tumor promotion,

M

542 Carcinogenesis, 16 (1995) 2493-2497.

D

543 [18] M.H. Pan, S.Y. Lin-Shiau, J.K. Lin, Comparative studies on the suppression of nitric oxide
TE

544 synthase by curcumin and its hydrogenated metabolites through down-regulation of I kappa B

545 kinase and NF kappa B activation in macrophages, Biochem Pharmacol, 60 (2000) 1665-1676.
EP

546 [19] H.O. Pae, S.O. Jeong, H.S. Kim, S.H. Kim, Y.S. Song, S.K. Kim, K.Y. Chai, H.T. Chung,

547 Dimethoxycurcumin, a synthetic curcumin analogue with higher metabolic stability, inhibits NO
C

548 production, inducible NO synthase expression and NF-κB activation in RAW264.7 macrophages
AC

549 activated with LPS, Mol Nutr Food Res, 52 (2008) 1082-1091.

550 [20] S. Mathews, M.N.A. Rao, Interaction of Curcumin with Glutathione, International Journal

551 of Pharmaceutics, 76 (1991) 257-259.

552 [21] J. Fang, J. Lu, A. Holmgren, Thioredoxin reductase is irreversibly modified by curcumin: a

24
 ACCEPTED MANUSCRIPT

553 novel molecular mechanism for its anticancer activity, J Biol Chem, 280 (2005) 25284-25290.

554 [22] M.G. Marcu, Y.J. Jung, S. Lee, E.J. Chung, M.J. Lee, J. Trepel, L. Neckers, Curcumin is an

555 inhibitor of p300 histone acetylatransferase, Med Chem, 2 (2006) 169-174.

556 [23] M. Niu, S. Wu, L. Mao, Y. Yang, CRM1 is a cellular target of curcumin: new insights for the

PT
557 myriad of biological effects of an ancient spice, Traffic, 14 (2013) 1042-1052.

RI
558 [24] B. Movassagh, P. Shaygan, Michael addition of thiols to α,β-unsaturated carbonyl

559 compounds under solvent-free conditions, Arkivoc, (2006) 130-137.

SC
560 [25] D.P. Nair, M. Podgorski, S. Chatani, T. Gong, W.X. Xi, C.R. Fenoli, C.N. Bowman, The

561 Thiol-Michael Addition Click Reaction: A Powerful and Widely Used Tool in Materials

562
U
Chemistry, Chemistry of Materials, 26 (2014) 724-744.
AN
563 [26] S.J. Wu, J.H. Jiang, J. Liu, X.H. Wang, Y. Gan, Y.F. Tang, Meta-analysis of SIRT1

564 expression as a prognostic marker for overall survival in gastrointestinal cancer, Oncotarget, 8
M

565 (2017) 62589-62599.

D

566 [27] G. Zu, A. Ji, T. Zhou, N. Che, Clinicopathological significance of SIRT1 expression in
TE

567 colorectal cancer: A systematic review and meta analysis, Int J Surg, 26 (2016) 32-37.

568 [28] L. Lv, Z. Shen, J. Zhang, H. Zhang, J. Dong, Y. Yan, F. Liu, K. Jiang, Y. Ye, S. Wang,
EP

569 Clinicopathological significance of SIRT1 expression in colorectal adenocarcinoma, Med Oncol,

570 31 (2014) 965.

C

571 [29] D.F. Yu, S.J. Jiang, Z.P. Pan, W.D. Cheng, W.J. Zhang, X.K. Yao, Y.C. Li, Y.Z. Lun,
AC

572 Expression and clinical significance of Sirt1 in colorectal cancer, Oncol Lett, 11 (2016) 1167-

573 1172.

574 [30] X. Chen, K. Sun, S. Jiao, N. Cai, X. Zhao, H. Zou, Y. Xie, Z. Wang, M. Zhong, L. Wei, High

575 levels of SIRT1 expression enhance tumorigenesis and associate with a poor prognosis of

25
 ACCEPTED MANUSCRIPT

576 colorectal carcinoma patients, Sci Rep, 4 (2014) 7481.

577 [31] L. Kriegl, M. Vieth, T. Kirchner, A. Menssen, Up-regulation of c-MYC and SIRT1

578 expression correlates with malignant transformation in the serrated route to colorectal cancer,

579 Oncotarget, 3 (2012) 1182-1193.

PT
580 [32] R.E. Carroll, R.V. Benya, D.K. Turgeon, S. Vareed, M. Neuman, L. Rodriguez, M. Kakarala,

RI
581 P.M. Carpenter, C. McLaren, F.L. Meyskens, Jr., D.E. Brenner, Phase IIa clinical trial of

582 curcumin for the prevention of colorectal neoplasia, Cancer Prev Res (Phila), 4 (2011) 354-364.

SC
583 [33] R.A. Sharma, S.A. Euden, S.L. Platton, D.N. Cooke, A. Shafayat, H.R. Hewitt, T.H.

584 Marczylo, B. Morgan, D. Hemingway, S.M. Plummer, M. Pirmohamed, A.J. Gescher, W.P.

585
U
Steward, Phase I clinical trial of oral curcumin: biomarkers of systemic activity and compliance,
AN
586 Clin Cancer Res, 10 (2004) 6847-6854.

587 [34] A.L. Cheng, C.H. Hsu, J.K. Lin, M.M. Hsu, Y.F. Ho, T.S. Shen, J.Y. Ko, J.T. Lin, B.R. Lin,
M

588 W. Ming-Shiang, H.S. Yu, S.H. Jee, G.S. Chen, T.M. Chen, C.A. Chen, M.K. Lai, Y.S. Pu, M.H.
D

589 Pan, Y.J. Wang, C.C. Tsai, C.Y. Hsieh, Phase I clinical trial of curcumin, a chemopreventive
TE

590 agent, in patients with high-risk or pre-malignant lesions, Anticancer Res, 21 (2001) 2895-2900.

591 [35] J.S. Dempe, R.K. Scheerle, E. Pfeiffer, M. Metzler, Metabolism and permeability of
EP

592 curcumin in cultured Caco-2 cells, Mol Nutr Food Res, 57 (2013) 1543-1549.

593 [36] Y. Yang, W. Duan, Y. Lin, W. Yi, Z. Liang, J. Yan, N. Wang, C. Deng, S. Zhang, Y. Li, W.
C

594 Chen, S. Yu, D. Yi, Z. Jin, SIRT1 activation by curcumin pretreatment attenuates mitochondrial
AC

595 oxidative damage induced by myocardial ischemia reperfusion injury, Free Radic Biol Med, 65

596 (2013) 667-679.

597 [37] Q. Sun, N. Jia, W. Wang, H. Jin, J. Xu, H. Hu, Activation of SIRT1 by curcumin blocks the

598 neurotoxicity of amyloid-beta25-35 in rat cortical neurons, Biochem Biophys Res Commun, 448

26
 ACCEPTED MANUSCRIPT

599 (2014) 89-94.

600 [38] A. Hu, J.J. Huang, R.L. Li, Z.Y. Lu, J.L. Duan, W.H. Xu, X.P. Chen, J.P. Fan, Curcumin as

601 therapeutics for the treatment of head and neck squamous cell carcinoma by activating SIRT1,

602 Sci Rep, 5 (2015).

PT
603 [39] W. Grabowska, M. Suszek, M. Wnuk, A. Lewinska, E. Wasiak, E. Sikora, A. Bielak-

RI
604 Zmijewska, Curcumin elevates sirtuin level but does not postpone in vitro senescence of human

605 cells building the vasculature, Oncotarget, 7 (2016) 19201-19213.

SC
606 [40] V. Byles, L.K. Chmilewski, J. Wang, L. Zhu, L.W. Forman, D.V. Faller, Y. Dai, Aberrant

607 cytoplasm localization and protein stability of SIRT1 is regulated by PI3K/IGF-1R signaling in

608
U
human cancer cells, Int J Biol Sci, 6 (2010) 599-612.
AN
609 [41] X. Zhang, S. Chen, M. Cheng, F. Cao, Y. Cheng, The expression and correlation of SIRT1

610 and Phospho-SIRT1 in colorectal cancer, Int J Clin Exp Med, 8 (2015) 809-817.
M

611 [42] H.C. Chen, Y.M. Jeng, R.H. Yuan, H.C. Hsu, Y.L. Chen, SIRT1 promotes tumorigenesis and
D

612 resistance to chemotherapy in hepatocellular carcinoma and its expression predicts poor
TE

613 prognosis, Ann Surg Oncol, 19 (2012) 2011-2019.

614 [43] Z. Lin, H. Yang, Q. Kong, J. Li, S.M. Lee, B. Gao, H. Dong, J. Wei, J. Song, D.D. Zhang, D.
EP

615 Fang, USP22 antagonizes p53 transcriptional activation by deubiquitinating Sirt1 to suppress cell

616 apoptosis and is required for mouse embryonic development, Mol Cell, 46 (2012) 484-494.
C

617 [44] L. Wen, Z. Chen, F. Zhang, X. Cui, W. Sun, G.G. Geary, Y. Wang, D.A. Johnson, Y. Zhu, S.
AC

618 Chien, J.Y. Shyy, Ca2+/calmodulin-dependent protein kinase kinase beta phosphorylation of

619 Sirtuin 1 in endothelium is atheroprotective, Proc Natl Acad Sci U S A, 110 (2013) E2420-2427.

620 [45] J. Ford, S. Ahmed, S. Allison, M. Jiang, J. Milner, JNK2-dependent regulation of SIRT1

621 protein stability, Cell Cycle, 7 (2008) 3091-3097.

27
 ACCEPTED MANUSCRIPT

622 [46] A.T. Dinkova-Kostova, M.A. Massiah, R.E. Bozak, R.J. Hicks, P. Talalay, Potency of

623 Michael reaction acceptors as inducers of enzymes that protect against carcinogenesis depends

624 on their reactivity with sulfhydryl groups, Proc Natl Acad Sci U S A, 98 (2001) 3404-3409.

625 [47] R.S. Zee, C.B. Yoo, D.R. Pimentel, D.H. Perlman, J.R. Burgoyne, X. Hou, M.E. McComb,

PT
626 C.E. Costello, R.A. Cohen, M.M. Bachschmid, Redox regulation of sirtuin-1 by S-glutathiolation,

RI
627 Antioxid Redox Signal, 13 (2010) 1023-1032.

628 [48] S.B. Jung, C.S. Kim, Y.R. Kim, A. Naqvi, T. Yamamori, S. Kumar, A. Kumar, K. Irani,

SC
629 Redox factor-1 activates endothelial SIRTUIN1 through reduction of conserved cysteine

630 sulfhydryls in its deacetylase domain, PLoS One, 8 (2013) e65415.

631
U
[49] S. Caito, S. Rajendrasozhan, S. Cook, S. Chung, H.W. Yao, A.E. Friedman, P.S. Brookes, I.
AN
632 Rahman, SIRT1 is a redox-sensitive deacetylase that is post-translationally modified by oxidants

633 and carbonyl stress, Faseb Journal, 24 (2010) 3145-3159.

M

634 [50] F. Ghisays, C.S. Brace, S.M. Yackly, H.J. Kwon, K.F. Mills, E. Kashentseva, I.P. Dmitriev,
D

635 D.T. Curiel, S.I. Imai, T. Ellenberger, The N-Terminal Domain of SIRT1 Is a Positive Regulator
TE

636 of Endogenous SIRT1-Dependent Deacetylation and Transcriptional Outputs, Cell Rep, (2015).

637 [51] M. Pan, H. Yuan, M. Brent, E.C. Ding, R. Marmorstein, SIRT1 contains N- and C-terminal
EP

638 regions that potentiate deacetylase activity, J Biol Chem, 287 (2012) 2468-2476.

639
C

640
AC

641

642

643

644 Figure Legends

28
 ACCEPTED MANUSCRIPT

645 Figure 1. Constitutively elevated expression of SIRT1 in several human CRC cell lines as

646 well as colon adenocarcinoma tissues. (A) Immunofluorescent staining of samples from a colon

647 cancer tissue array with matched adjacent normal colon tissues was performed using anti-SIRT1

648 antibody. Surrounding normal colon tissues is further away from 1.5 cm of the rim of tumor,

PT
649 which was determined by a pathologist under microscope to be “normal”. The fluorescent

RI
650 intensity was analyzed by the image processing program Image J and results are shown as the

651 mean ± S.D. of thirty-five samples, **p < 0.01. Images of H&E stained tissue sections were

SC
652 provided by US Biomax Inc. Each scale bar represents 200 µm. (B) The basal levels of SIRT1 in

653 different human CRC cell lines (HCT-116, HCT-15 and DLD-1) as well as a normal human

654
U
colon epithelial CCD841CoN cells were assessed by Western blot analysis. (C) Human CRC cell
AN
655 lines (HCT-116, HCT-15, and DLD-1) were transiently transfected with scrambled or SIRT1

656 siRNA for the indicated period of time. The cell viability was examined by the MTT assay. The
M

657 results are shown as the mean ± S.D. of triplicates, *p < 0.05, ***p < 0.001. (D) For the wound
D

658 healing assay, human CRC cell lines were transiently transfected with scrambled or SIRT1
TE

659 siRNA for 48 h and were then plated into culture-insert (Ibidi). After 48-h incubation, cell

660 migration was assessed by measuring the wound size under a microscope. Each scale bar
EP

661 represents 500 µm. The values are presented as the mean ± S.D. of three independent

662 experiments, *p < 0.05, ***p < 0.001.

C

663
AC

664 Figure 2. Effects of SIRT1 knock-down on tumor growth in a human colon cancer

665 xenograft model in athymic nude mice. (A) HCT-116 cells were transfected with scrambled or

666 SIRT1 siRNA and 5 × 106 cells were injected subcutaneously into the flanks of BALB/c nude

667 mice. Photograph of the xenograft tumors excised from mice at the end of the experiment (day

29
 ACCEPTED MANUSCRIPT

668 21). (B) The tumor volume was determined by using the formula, V= 0.5 × longest diameter ×

669 (shortest diameter)2. The results are shown as the mean ± S.D. of six xenografts for each group,

670 *p < 0.05, ***p < 0.001. (C) The tumor volume of the mice at the end of the experiment. The

671 data are presented as the mean ± S.D., *p < 0.05. (D) The tumor weight of the mice at the end of

PT
672 the experiment. The results are expressed as the mean ± S.D., *p < 0.05. (E) Histological features

RI
673 were evaluated by H&E staining in sections from resected xenograft (magnification, × 10). Each

674 scale bar represents 200 µm. (F) Immunohistochemical analysis was performed with sections of

SC
675 excised xenograft using anti-PCNA and SIRT1 antibodies. Each scale bar represents 200 µm. (G)

676 The protein levels of SIRT1, cyclin D1, PCNA and c-Myc in resected tumor tissues were

677
U
determined by immunoblot analysis. The values are presented as the mean ± S.D. of six samples,
AN
678 **p < 0.01, ***p < 0.001.

679
M

680 Figure 3. Effects of curcumin on the expression and activity of SIRT1 in cultured human
D

681 colon cancer HCT-116 cells. (A) HCT-116 cells were treated with two different concentrations
TE

682 (1 or 10 µM) of curcumin for 3 h. The protein level of SIRT1 was measured by Western blot

683 analysis. The results are shown as the mean ± S.D. of three independent experiments, **p < 0.01.
EP

684 (B) Following treatment of HCT-116 cells with curcumin (1 or 10 µM) for 3 h, SIRT1

685 deacetylase activity was measured by using the fluorometric assay kit as described in Materials
C

686 and Methods. The data are presented as the mean ± S.D. of three independent experiments, **p <
AC

687 0.01. (C, D) HCT-116 cells were treated with curcumin (10 µM) for the indicated durations or

688 with two different concentrations (1 or 10 µM) of curcumin for 3 h. The protein (C) and mRNA

689 (D) levels of SIRT1 were measured by Western blot and reverse transcription-PCR analyses,

690 respectively. (E) Following exposure of HCT-116 cells to curcumin (10 µM) for 3 h, both

30
 ACCEPTED MANUSCRIPT

691 cytosolic and nuclear extracts were prepared and subjected to immunoblot analysis. α-Tubulin

692 and Sp1 were used as cytosolic and nuclear markers, respectively. (F) Immunocytochemical

693 analysis was performed using anti-SIRT1 antibody after the treatment of HCT-116 cells with

694 curcumin (10 µM) for 3 h. Nuclei were identified by propidium iodide (PI) staining. The scale

PT
695 bar represents 100 µm.

RI
696

697 Figure 4. Stimulation of ubiquitin-dependent proteasomal degradation of SIRT1 by

SC
698 curcumin. (A) HCT-116 cells were treated with cycloheximide (CHX; 71.07 µM) or vehicle for

699 indicated periods in the absence or presence of curcumin (10 µM). The relative protein levels of

700
U
SIRT1 remaining at different time points were assessed. The results are shown as the mean ± S.D.
AN
701 of three independent experiments, **p < 0.01, ***p < 0.001. (B) HCT-116 cells were exposed to

702 curcumin (10 µM) alone or in combination with a proteasome inhibitor MG-132 (20 µM), and
M

703 the protein level of SIRT1 was examined by Western blot analysis. The values are presented as
D

704 the mean ± S.D. of three independent experiments, *p < 0.05, ***p < 0.001. (C) Following
TE

705 treatment of HCT-116 cells with MG-132 (20 µM) and curcumin (10 µM) for 3 h, the

706 ubiquitination of endogenous SIRT1 was measured by immunoprecipitation. (D) HCT-116 cells
EP

707 were co-transfected with FLAG-SIRT1 and His-ubiquitin for 48 h, followed by MG-132 (20 µM)

708 and curcumin (10 µM) treatment for 3 h. The ubiquitination of exogenous SIRT1 was measured
C

709 by immunoprecipitation.
AC

710

711 Figure 5. Comparison of the effects of curcumin and tetrahydrocurcumin on the stability of

712 SIRT1, colon cancer cell transformation and the growth of xenograft tumors. (A) Chemical

713 structures of curcumin and its non-electrophilic analogue, tetrahydrocurcumin. Unlike curcumin,

31
 ACCEPTED MANUSCRIPT

714 tetrahydrocurcumin lacks the conjugated bonds in the central seven-carbon chain. (B) Following

715 treatment of HCT-116 cells with curcumin (10 µM) or tetrahydrocurcumin (10 µM), the protein

716 level of SIRT1 was examined by Western blot analysis. The data are presented as the mean ± S.D.

717 of three independent experiments, ***p < 0.001. (C) The ubiquitination of SIRT1 was measured

PT
718 by immunoprecipitation after treatment of HCT-116 cells with curcumin (10 µM) or the same

RI
719 concentration of tetrahydrocurcumin. (D) Whole-cell lysates from HCT-116 cells were incubated

720 overnight at 4 °C with Sepharose 4B beads conjugated to curcumin or tetrahydrocurcumin and

SC
721 were then pulled down by centrifugation. Covalent interaction between SIRT1 and curcumin or

722 tetrahydrocurcumin was assessed by immunoblot analysis using an antibody against SIRT1. (E)

723
U
For the anchorage-independent growth assay, HCT-116 cells were grown in soft agar for 14 days
AN
724 treated with 10 µM each of curcumin or tetrahydrocurcumin every other day and were then

725 stained with 0.05% crystal violet. The number of colonies larger than 100 µm in diameter was
M

726 counted. Each scale bar represents 200 µm. The results are shown as the mean ± S.D. of three
D

727 independent experiments performed, **p < 0.01, ***p < 0.001. (F) Photograph of the xenograft
TE

728 tumors from the mice treated with curcumin or tetrahydrocurcumin at the time of excision. The

729 treatment of mice and other experimental details are described in Materiald and Methods. (G)
EP

730 The tumor volume of the mice at the end of the experiment. The results are shown as the mean ±

731 S.D. of six xenografts for each group, *p < 0.05, ***p < 0.001. (H) The tumor weight of the mice
C

732 at the end of the experiment. The data are presented as the mean ± S.D. of six samples for each
AC

733 group, *p < 0.05, ***p < 0.001. (I) Histological morphology was assessed by H&E staining

734 (upper panels), and the protein level of SIRT1 was examined by immunohistochemical staining

735 (lower panels) in sections from resected xenograft (magnification, × 10). The asterisk indicates

736 the necrotic area. Each scale bar represents 200 µm. (J) The protein expression of SIRT1 in

32
 ACCEPTED MANUSCRIPT

737 excised tumor tissues was determined by immunoblot analysis.

738

739 Figure 6. The cysteine 67 residue of SIRT1 as a binding site of curcumin. (A) Recombinant

740 human SIRT1 (5 µg) was mixed with curcumin (10 µM) or vehicle (DMSO) for 1 h at room

PT
741 temperature and subjected to nano-LC-ESI-MS/MS. Data were analyzed using

RI
742 BioWorksBrowserTM as described in Materials and Methods. (B) HCT-116 cells were transiently

743 transfected with FLAG-tagged wild-type (SIRT1-WT) or mutant-SIRT1 (SIRT1-C67A). Both

SC
744 cytosolic and nuclear extracts were then prepared and subjected to Western blot analysis. α-

745 Tubulin and Lamin B were used as cytosolic and nuclear markers, respectively. (C) HCT-116

746
U
cells were transiently transfected with wild-type (SIRT1-WT) or mutant-SIRT1 (SIRT1-C67A),
AN
747 and the protein level of Ac-p53Lys382 was determined by immunoblot analysis. (D) HCT-116 cells

748 were transiently transfected with the FLAG-tagged plasmids expressing wild-type SIRT1
M

749 (SIRT1-WT) or mutant-SIRT1 (SIRT1-C67A). Whole-cell lysates were incubated with

D

750 curcumin-Sepharose 4B beads. The SIRT1 bound to the curcumin-Sepharose 4B beads was
TE

751 pulled down by centrifugation and analyzed by Western blot analysis using anti-FLAG antibody.

752
EP

753 Figure 7. Suppression of curcumin-induced ubiquitin-dependent proteasomal degradation

754 of SIRT1 and anchorage-independent growth in HCT-116 cells by mutation of SIRT1

C

755 cysteine 67. (A) HCT-116 cells were treated with curcumin (10 µM) and cycloheximide (CHX;
AC

756 71.07 µM) for indicated time periods following transfection with wild-type (SIRT1-WT) or

757 mutant-SIRT1 (SIRT1-C67A), and the protein level of SIRT1 remaining at each time point was

758 determined by immunoblot analysis. The results are shown as the mean ± S.D. of three

759 independent experiments, **p < 0.01, ***p < 0.001. (B) HCT-116 cells were transiently

33
 ACCEPTED MANUSCRIPT

760 transfected with the His-ubiquitin plasmids and FLAG-tagged wild-type SIRT1 (SIRT1-WT) or

761 mutant-SIRT1 (SIRT1-C67A). Following treatment of HCT-116 cells with MG-132 (20 µM) and

762 curcumin (10 µM) for the indicated durations, the ubiquitination of SIRT1 was measured by

763 immunoprecipitation. (C) HCT-116 cells were transiently transfected with wild-type or mutant

PT
764 SIRT1. Cells were then treated with curcumin (10 µM) for the anchorage-independent growth

RI
765 assay as described in Materials and Methods. Each scale bar represents 200 µm. The data are

766 presented as the mean ± S.D. of three independent experiments, **p < 0.01, ***p < 0.001.

SC
767

768 Figure 8. A proposed mechanism underlying curcumin-induced destabilization of oncogenic

769
U
SIRT1. Curcumin bearing the α,β-unsaturated carbonyl moiety can directly bind to SIRT1
AN
770 protein, preferentially at the cysteine 67 residue. The resulting conformational alteration of the

771 SIRT1 structure is speculated to facilitate its degradation through the ubiquitin-proteasome
M

772 system.
D

773
TE

774

775
EP

776 Supplementary Figure Legends

777 Supplementary Figure 1. (A) CCD841CoN cells were treated with 1 or 10 µM of curcumin for
C

778 3 h. The protein level of SIRT1 was measured by Western blot analysis. (B) Following treatment
AC

779 of CCD841CoN cells with curcumin (1 or 10 µM) for 3 h, SIRT1 deacetylase activity was

780 measured by using the fluorometric assay kit. (C, D) HCT-116 cells were treated with curcumin

781 (10 µM) for the indicated time periods or with 1 or 10 µM for 3 h. The mRNA levels of SIRT1

782 were measured by reverse transcription-PCR (C) and real-time qPCR (D) analyses using

34
 ACCEPTED MANUSCRIPT

783 GAPDH and RPL32, respectively as an equal loading control for normalization. (E) DLD-1,

784 HCT-15 and MCF-7 cells were treated with curcumin (10 µM) for the indicated durations. The

785 protein and mRNA levels of SIRT1 were measured by Western blot and reverse transcription-

786 PCR analyses, respectively.

PT
787

RI
788 Supplementary Figure 2. (A) HCT-116 cells were exposed to curcumin (10 µM) or

789 tetrahydrocurcumin (10 µM) for 48 h, and the cell migration was assessed. Each scale bar

SC
790 represents 500 µm. The values are resented as the mean ± S.D. of three independent experiments,

791 *p < 0.05. (B) Higher magnification images corresponding to the Fig. 5I. The asterisk indicates

792
U
the necrotic area. Each scale bar represents 300 µm. (C) The quantitative data of the protein
AN
793 expression of SIRT1 in the xenograft tumors from the mice treated with curcumin or

794 tetrahydrocurcumin. The values are presented as the mean ± S.D. of six samples, **p < 0.01,
M

795 ***p < 0.001.

D

796
TE

797 Supplementary Figure 3. (A, B) HCT-116 cells were exposed to curcumin (10 µM) in the

798 presence or absence of a thiol reducing agent DTT (500 µM) (A) or NAC (10 mM) (B). The
EP

799 results are shown as the mean ± S.D. of three independent experiments, **p < 0.01, ***p < 0.001.

800 (C) HCT-116 cells were transiently transfected with wild-type (SIRT1-WT) or mutant-SIRT1
C

801 (SIRT1-C67A), and the deacetylase activity of SIRT1 was measured by using the fluorometric
AC

802 assay kit. (D) DLD-1 cells were transiently transfected with the FLAG-tagged plasmids

803 expressing wild-type SIRT1 (SIRT1-WT) or mutant-SIRT1 (SIRT1-C67A). Following incubation

804 of whole-cell lysates from DLD-1 cells with curcumin-Sepharose 4B beads, the SIRT1 bound to

805 the curcumin-Sepharose 4B beads was pulled down by centrifugation and analyzed by

35
 ACCEPTED MANUSCRIPT

806 immunoblot analysis using anti-FLAG antibody.

807

808 Supplementary Figure 4. HCT-116 cells were transiently transfected with wild-type (SIRT1-

809 WT) or mutant-SIRT1 (SIRT1-C67A). Cells were then treated with curcumin (10 µM) or vehicle

PT
810 for the wound healing assay as described in Materials and Methods. Each scale bar represents

RI
811 500 µm. The values are presented as the mean ± S.D. of triplicates, ***p < 0.001.

812

U SC
AN
M
D
TE
C EP
AC

36
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

Curcumin suppresses oncogenicity of human colon cancer cells by covalently modifying

the cysteine 67 residue of SIRT1

Yeon-Hwa Lee, Na-Young Song, Jinyoung Suh, Do-Hee Kim, Wonki Kim, Jihyae Ann,

Jeewoo Lee, Jeong-Heum Baek, Hye-Kyung Na, and Young-Joon Surh*

PT
RI
Highlights

SC
• SIRT1 is overexpressed in colorectal cancer tissues compared with normal colon tissues.

U
• SIRT1 knockdown attenuates viability and migration of human colon cancer cells
AN
• Silencing of SIRT1 suppressed the growth of HCT-116 derived tumor xenografts.

• Curcumin inhibits migration and growth of HCT-116 cells.

M

• Curcumin covalently modifies SIRT1 on the cysteine 67 residue.

D

• Curcumin stimulates ubiquitin-dependent proteasomal degradation of SIRT1.

TE
C EP
AC