Original Article | Published:

SIRT1 RNAi knockdown induces apoptosis and senescence, inhibits invasion and enhances chemosensitivity in pancreatic cancer cells

Gene Therapy volume 18, pages 920928 (2011) | Download Citation


The NAD+-dependent deacetylase, sirtuin 1 (SIRT1), has been recently been suspected to have a role in tumorigenesis. We investigated the expression of SIRT1 in pancreatic cancer and the effect of SIRT1-targeted RNA interference (RNAi) on cell proliferation and tumor formation in a pancreatic cancer cell line, PANC1. The expression of SIRT1 was investigated in 49 specimens of pancreatic cancer and adjacent normal pancreatic tissues. SIRT1 was overexpressed in pancreatic cancer tissues at both the mRNA and protein levels, with increased SIRT1 positivity associated with tumors from patients over 60 years old, tumors larger than 4 cm, higher TNM (extent of tumor (T), the extent of spread to lymph nodes (N), and presence of distant metastasis (M)) stage or the presence of lymph node or hepatic metastases. The PANC-1 was stably transfected with a SIRT1 small hairpin RNA (shRNA) expression plasmid and compared with untransfected and PANC-1-negative RNAi cells. Proliferation of PANC-1–SIRT1–RNAi cells was significantly reduced, accompanied by increased rates of apoptosis, G1 arrest and senescence. Furthermore, FOXO3a expression was markedly upregulated in PANC-1–SIRT1–RNAi cells, but no significant difference in p53 expression was observed. The invasive ability of PANC-1–SIRT1–RNAi cells was markedly reduced in vitro, which was linked to increased E-cadherin and reduced-MMP expression. Additionally, PANC-1–SIRT1–RNAi cells had a significantly reduced capacity to form tumors in vivo compared with untransfected and PANC-1-negative RNAi cells. These results suggest that SIRT1 may promote cell proliferation and tumor formation in pancreatic cancer, and downregulation of SIRT1 using shRNA could provide a novel therapeutic treatment.


Pancreatic cancer is one of the most aggressive solid malignancies, the third most common gastrointestinal malignancy, the fifth leading cause of cancer mortality and is a leading cause of cancer-related deaths in the developed world.1 One key characteristic of pancreatic cancer is early systemic metastasis and extremely rapid local tumor progression.2 Resistance to existing chemotherapeutic agents occurs commonly in pancreatic cancer therapy,3 and therefore, there is an urgent need to elucidate the molecular mechanisms that underlie the malignant behavior of pancreatic cancer.

It has been recently suspected that sirtuin 1 (SIRT1) is involved in tumorigenesis.4, 5 SIRT1 is one of seven members of the sirtuin family of nicotinamide adenine dinucleotide (NAD+)-dependent class III histone deacetylases, with sequence homology to the enzymatic domain of yeast silent information regulator 2 (sir2).6, 7 SIRT1 has a large number of histone and non-histone substrates, which are involved in the regulation of metabolism, differentiation, proliferation, senescence, protein degradation and apoptosis.8 Overexpression of SIRT1 provides a cell survival advantage by inhibiting apoptosis and resisting senescence.9, 10 SIRT1 is a well-known regulator of cellular stress, induction of apoptosis and aging-related senescence, and as cancer is an age-related disease, it is hypothesized that SIRT1 may have an important role in cancer. SIRT1 deacetylates non-histone proteins, including various transcription factors such as p53,11 forkhead class O transcription factor (FOXO) family,12, 13 nuclear factor-κB,14 Bax/Bcl-2 (ref. 15) and Ku70,16 which are involved in growth regulation, stress response, apoptosis and cancer progression.

SIRT1 localizes to the promoters of several aberrantly-silenced tumor suppressor genes, whose DNA is hypermethylated, implying that SIRT1 is associated with the epigenetic hallmarks of cancer as SIRT1 can deacetylate histones.17 Furthermore, SIRT1 expression is elevated in human prostate cancer,18 acute myeloid leukemia19 and primary colon cancer,20 and overexpression of SIRT1 is associated with oncogenic transformation, tumor cell survival and resistance to therapy.21 In addition, a relationship between SIRT1 activity and expression of the tumor suppressor adhesion molecule, E-cadherin,22 is closely related to the metastatic potential of tumors.23 This has led to the hypothesis that SIRT1 may promote tumorgenesis and drug resistance and provide a potential target for cancer therapy.24, 25 It has also been shown that SIRT1 may inhibit proliferation and tumor formation. Wang and colleagues26 observed SIRT1 overexpression in approximately 25% of stage I/II/III colorectal adenocarcinomas, but this was only rarely present in advanced stage IV tumors; however, approximately 30% of these advanced adenocarcinomas showed lower levels of SIRT1 expression than normal tissues. Additionally, the knockdown of SIRT1 expression using RNA interference (RNAi) accelerated tumor xenograft formation in colon cancer cells, whereas overexpression of SIRT1 inhibited tumor formation. Herranz et al.27 reported that a three-fold increase in the expression of SIRT1 in mouse lines significantly protected hepatocytes from metabolic syndrome-associated hepatic cancer. These contradictory results imply that the role of SIRT1 is complex and may depend on tumor type.

As the incidence of pancreatic cancer increases exponentially with age, a mechanistic connection has been suggested between carcinogenesis and aging. Therefore, we speculated that SIRT1 may be involved in the carcinogenesis of pancreatic cancer. Apart from the regulation of aging and longevity in mammals, SIRT1 is also linked to survival of cancer cells and resistance to apoptosis,28, 29 and the increased expression of SIRT1 in breast cancer cells protects against genotoxic stress-induced apoptosis via the ras–mitogen-activated protein kinase pathway.30 Therefore, we hypothesized that the overexpression of SIRT1 in pancreatic cancer may promote chemoresistance and the decreased expression of SIRT1 could sensitize pancreatic cells to chemotherapy.

To determine the level of SIRT1 expression in pancreatic cancer cells, we compared tumor tissues and the surrounding normal tissues and investigated the association with various clinical parameters and SIRT1 expression. To explore the role of SIRT1 in pancreatic cancer cells further, a SIRT1–small hairpin RNA (shRNA) plasmid was transfected into the pancreatic cancer cell line (PANC-1) to knock down SIRT1 expression; cell proliferation, apoptosis, cell cycle arrest and senescence was then investigated. The invasive ability and chemosensitivity to 5-fluoruracil (5-FU) and Gemcitabine were evaluated. Furthermore, the expression of p53, FOXO3a, Bcl-2 and Bax were quantified to investigate the mechanisms involved in apoptosis and senescence, whereas E-cadherin, MMP-2 and MMP-9 proteins were analyzed to investigate cell invasion.


SIRT1 is overexpressed in human pancreatic cancer tissues and associated with clinical parameters

The expression of SIRT1 in pancreatic cancer tissues was evaluated by immunohistochemical staining and quantitative real-time PCR (qPCR). The immunohistochemical results showed that the number of samples that expressed SIRT1 was significantly higher in pancreatic cancers than adjacent normal pancreatic tissue (75.51 versus 22.45%, P<0.05). qPCR demonstrated that the expression of SIRT1 mRNA in pancreatic cancer was significantly higher than in the adjacent normal pancreatic tissue (0.927±0.068 versus 0.355±0.079, P<0.01). Western blotting corroborated the increased level of SIRT1 protein expression in tumor tissues (Supplementary Figure 1). Analysis of the clinical data associated with each specimen indicated that SIRT1 positivity was associated with patients over 60 years old, tumors greater than 4 cm (length), higher TNM (extent of tumor (T), the extent of spread to lymph nodes (N), and presence of distant metastasis (M)) staging, or tumors with lymph node or hepatic metastases. In contrast, there was no statistical association with SIRT1 positivity and the degree of differentiation, vessel or nerve infiltration and the location of tumor within the pancreas (Table 1).

Table 1: Association of SIRT1 positivity in pancreatic cancer with patient and tumor characteristics

SIRT1-targeted shRNA significantly decreased SIRT1 expression

Three expression plasmids containing shRNA that were targeted against SIRT1 (SIRT1–RNAi-1, 2 or 3) were transfected into PANC-1 cells. In transient transfections, the SIRT1–RNAi-1 plasmid had the strongest inhibitory effect on SIRT1 expression at both the mRNA and protein levels (Figure 1a). Therefore, stable clones of both SIRT1–RNAi-1 and the negative-RNAi-transfected cells were selected with G418 via the finite deliquation method. Using flow cytometry, the purity of stable transfectants, identified by the expression of the plasmid green fluorescent protein using flow cytometry, was greater than 95%. Western blotting and qPCR demonstrated that the levels of SIRT1 mRNA and SIRT1 protein in PANC-1–SIRT1–RNAi cells were reduced by 98% and 86%, respectively, compared with untransfected PANC-1 cells (Figure 1b), and a non-significant effect on SIRT1 expression was observed in PANC-1-negative RNAi cells.

Figure 1
Figure 1

The effect of SIRT1–RNAi knockdown on SIRT1 expression and proliferation of pancreatic cancer cells. (a) PANC-1 cells were transfected with the SIRT1–RNAi (SIRT1–RNAi-1, SIRT1–RNAi-2 or SIRT1–RNAi-3), negative RNAi or positive RNAi plasmids. SIRT1 mRNA and protein expression in negative RNAi and positive RNAi transfected cells was not affected, compared with control cells. SIRT1 expression was significantly reduced by SIRT1–RNAi-1, -2 and -3. The SIRT1–RNAi-1 plasmid had greatest inhibition of SIRT1 mRNA and protein expression and was used for clone generation by stable transfection. (n=6, *P<0.01). (b) Stable transfectants expressing green fluorescent protein were identified by flow cytometry. The M1 ratio of cells expressing green fluorescent protein for PANC-1, PANC-1-negative RNAi and PANC-1–SIRT1–RNAi cells were 0.03, 97.55 and 98.58%, respectively. SIRT1 expression was quantified using qPCR and western blotting. SIRT1 expression in PANC-1 and PANC-1-negative RNAi cells was significantly higher than PANC-1–SIRT1–RNAi cells (*P<0.01). (c) Proliferation was quantified using the MTT assay and the doubling time of PANC-1–SIRT1–RNAi cells was significantly longer than PANC-1 and PANC-1-negative RNAi cells (n=6, *P<0.01).

Obvious proliferation inhibition of PANC-1–SIRT1–RNAi cells was observed

The 3-(4,5-dimethyl-2-thiazyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) assay was used to investigate the effect of SIRT1–RNAi on PANC-1 cell growth. There was no significant difference in the proliferation of PANC-1 cells and PANC-1-negative RNAi cells, which suggested that negative shRNA interference did not alter cell proliferation. Compared with PANC-1 and PANC-1-negative RNAi cells, the growth of PANC-1–SIRT1–RNAi cells was significantly reduced. (Figure 1c).

SIRT1-targeted shRNA induces cell cycle arrest and apoptosis in PANC-1 cells

To investigate the mechanism by which SIRT1 shRNA inhibited proliferation, we analyzed the cell cycle and apoptosis in PANC-1–SIRT1–RNAi cells using flow cytometry. No significant difference was observed between PANC-1 controls and cell transfected with PANC-1-negative RNAi; however, the G0/G1 distribution in PANC-1–SIRT1–RNAi cells significantly increased (50.54±3.58 versus 69.22±3.91%, P<0.05) and the ratio of S and G2/M phases decreased. In addition, the sub-G1 ratio in PANC-1–SIRT1–RNAi cells increased (8.74±1.22 to 17.41±2.18%, P<0.05), which represented cells with a hypodiploid genome because of DNA degradation during apoptosis (Figure 2a).

Figure 2
Figure 2

Cell cycle and apoptosis analysis of PANC-1 RNAi cells. (a) The cell cycle was analyzed by flow cytometry. Compared with PANC-1 (sub-G1: 8.74±1.22; G0/G1: 50.54±3.58; S: 10.39±1.67; G2/M: 30.72±2.42) and PANC-1-negative RNAi cells (sub-G1: 8.62±2.52; G0/G1: 56.78±1.16; S: 8.53±2.33; G2/M: 25.60±1.45), the cell cycle of PANC-1–SIRT1–RNAi (sub-G1: 17.41±2.18; G0/G1: 69.22±3.91; S: 7.40±1.59; G2/M: 6.39±0.97) cells was significantly blocked at G0/G1. (b) TUNEL assay indicated apoptosis in PANC-1–SIRT1–RNAi cells was significantly increased (n=5, *P <0.01). (c) Apoptosis regulating factors were quantified using qPCR and western blotting. FOXO3a and Bax were upregulated in PANC-1–SIRT1–RNAi cells, p53 was not affected and anti-apoptotic Bcl-2 was downregulated, compared with PANC-1 and PANC-1-negative RNAi cells (n=5, *P<0.01).

Apoptotic cells were quantified using the TUNEL assay and positive staining nuclear staining was quantified (Figure 2b). SIRT1 interference significantly increased the apoptotic index of PANC-1–SIRT1–RNAi cells (64.91±13.48%, P<0.001) compared with PANC-1 and PANC-1-negative cells (7.53±2.33% and 9.14±4.49%, respectively).

We analyzed mRNA and protein expression of the apoptosis-associated factors, p53, FOXO3a, Bcl-2 and Bax. Compared with PANC-1 cells, FOXO3a mRNA expression was significantly upregulated in PANC-1–SIRT1–RNAi cells, but in PANC-1-negative RNAi cells. No significant difference was observed in p53 mRNA expression among the cell lines. The differences in p53 and FOXO3a expression observed at the mRNA level were consistent with results of the protein expression analysis. In PANC-1–SIRT1–RNAi cells, but not PANC-1-negative cells, a significant increase in Bax and a reduction in the level of Bcl-2 expression at the mRNA and protein levels were observed in comparison with PANC-1 cells (Figure 2c).

Knockdown of SIRT1 promotes senescence in PANC-1 cells

The senescence-associated β-galactosidase (SA-β-Gal) assay was used to investigate senescence, a mechanism of growth arrest independent of apoptosis. Compared with PANC-1 and PANC-1-negative RNAi cells, the number of senescent cells was significantly increased in PANC-1–SIRT1–RNAi cells (Figure 3).

Figure 3
Figure 3

SIRT1 knockdown promoted cell senescence in PANC-1–SIRT1–RNAi cells. Representative images of SA-β-Gal staining ( × 400). Senescent cells expressing SA-β-Gal form a local blue precipitate upon cleavage. The percentage of SA-β-Gal–positive cells in PANC-1, PANC-1-negative RNAi and PANC-1–SIRT1–RNAi cells were 8.91±1.44%, 10.07±1.38%, 47.23±3.91%, respectively (n=5, *P <0.01).

Downregulation of SIRT1 enhanced the chemosensitivity of PANC-1 cells

PANC-1, PANC-1-negative RNAi and PANC-1–SIRT1–RNAi cells were treated with 5-FU and Gemcitabine to determine their chemosensitivity. 5-FU reduced the growth of all cell lines in both, in a dose- and time-dependent manner. Compared with PANC-1 and PANC-1-negative RNAi cells, PANC-1–SIRT1–RNAi cells displayed increased sensitivity to 5-FU (P<0.01), especially at 50 μg ml−1, which is approximately the peak plasma level of 5-FU achieved during chemotherapy.31 Additionally, the chemosensitivity of PANC-1–SIRT1–RNAi cells to Gemcitabine was significantly enhanced, compared with PANC-1 cells (Figure 4).

Figure 4
Figure 4

SIRT1 knockdown increased chemosensitivity to 5-FU and Gemcitabine. Various concentrations of 5-FU and Gemcitabine were applied to PANC-1, PANC-1-negative RNAi and PANC-1–SIRT1–RNAi cells for 24, 48 and 72 h, and cell viability was determined using the MTT assay. PANC-1–SIRT1–RNAi cells showed increased sensitivity to 5-FU (a) or Gemcitabine (b) at different dosages. The 5-FU IC50s of PANC-1, PANC-1-negative RNAi and PANC-1–SIRT1–RNAi cells were 117.42±31.19, 104.54±29.22 and 51.37±18.94 μg ml−1, respectively. The Gemcitabine IC50s were 78.32±15.49, 72.68±20.11 and 21.47±7.53 μg ml−1, respectively (n=3, *P<0.01).

Knockdown of SIRT1 reduces the invasiveness of PANC-1 cells

Invasive ability was evaluated using the in vitro transwell migration assay. Compared with PANC-1 and PANC-1-negative RNAi cells, the number of PANC-1–SIRT1–RNAi cells that invaded the lower side of the membrane was significantly decreased (Figure 5a). To investigate the mechanism responsible for the reduced invasive ability caused by the downregulation of SIRT1, we quantified the protein expression levels of E-cadherin, MMP-2 and MMP-9. E-cadherin expression was significantly increased in PANC-1–SIRT1–RNAi cells, whereas MMP-2 and MMP-9 expression were significantly reduced. In addition, the activated fraction of MMP-9 was decreased in PANC-1–SIRT1–RNAi cells (Figure 5b).

Figure 5
Figure 5

The SIRT1 knockdown reduced the invasive ability of PANC-1 cells. (a) The in vitro transwell migration assay was performed and after 24 h, the filter was fixed, stained and the number of cells counted in 10 randomly selected fields. The invasion of PANC-1-negative RNAi and PANC-1–SIRT1–RNAi cells could be observed by fluorescence microscopy due to expression of the plasmid reporter gene, green fluorescent protein. The numbers of PANC-1, PANC-1-negative RNAi and PANC-1–SIRT1–RNAi cells were 59±13, 61±10 and 22±6, respectively, with a significant decrease observed in the invasive ability of PANC-1–SIRT1–RNAi cells (n=3, *P<0.01). (b) Expression of E-cadherin, MMP-2 and MMP-9 was measured by qPCR and western blotting. Compared with PANC-1 and PANC-negative RNAi cells, E-cadherin in PANC-1–SIRT1–RNAi cells was significantly increased and MMP-2, total MMP-9 and active MMP-9 were significantly reduced (n=3, *P<0.01).

PANC-1–SIRT1–RNAi cells have a reduced ability to form tumor xenografts in nude mice

The growth of PANC-1–SIRT1–RNAi cells was evaluated in nude mice to determine if decreased cell proliferation observed in the vitro assays would be reproduced in vivo. Compared with PANC-1 and PANC-1-negative RNAi cells, the growth rate and size of PANC-1–SIRT1–RNAi tumors were significantly reduced, with no significant differences observed between PANC-1 and PANC-1-negative RNAi tumors (Supplementary Figure 2 and Supplementary Table 1). To investigate the growth inhibition induced by the SIRT1 knockdown further, xenograft tumors were harvested to analyze SIRT1 expression. PANC-1–SIRT1–RNAi tumors had lower levels of SIRT1 expression, indicating that the SIRT1 shRNA plasmid construct had been retained in vivo over multiple cell divisions (Supplementary Figure 3A). TUNEL assays indicated significantly more apoptotic cells in PANC-1–SIRT1–RNAi tumors, compared with PANC-1 and PANC-1-negative RNAi tumors (Supplementary Figure 3B).

The apoptosis, senescence and chemosensitivity were analyzed in PANC-1 cells transiently transfected with SIRT1–RNAi-3 plasmid and negative-RNAi

To further confirm that the effects of SIRT1–RNAi were not caused by inhibition of another gene apart from SIRT1, the apoptosis, senescence and chemosensitivity were repeatedly analyzed in PANC-1 cells transiently transfected with negative RNAi (PANC-1-negtive RNAi) and SIRT1–RNAi-3 plasmid (PANC-1–SIRT1–RNAi-3 cells), which also significantly downregulated the SIRT1 expression. Apoptotic cells were quantified with flow cytometry (Supplementary Figure 4A). PANC-1 cells transiently transfected with SIRT1–RNAi-3 significantly increased the apoptotic index (7.91±2.58%, P<0.05), compared with PANC-1 and PANC-1-negative RNAi cells (1.43±0.79% and 2.14±0.74%, respectively). Compared with PANC-1 and PANC-1-negative RNAi cells, the number of senescent cells was significantly increased in PANC-1–SIRT1–RNAi-3 cells (32.56±4.16 versus 6.37±2.34 and 8.29±3.38%, P<0.05; Supplementary Figure 4B).

To compare the chemosensitivity, PANC-1, PANC-1-negative RNAi and PANC-1–SIRT1–RNAi-3 cells were treated with different concentration of 5-FU and Gemcitabine, respectively. Compared with PANC-1 and PANC-1-negative RNAi cells, PANC-1–SIRT1–RNAi cells demonstrated significantly decreased IC50s of 5-FU (56.21±15.42 versus 104.38±28.46 and 96.32±34.57 μg ml−1, P<0.01) and Gemcitabine (26.43±8.12 versus 71.13±17.52 and 76.54±23.31 μg ml−1, P<0.01).


Immunohistochemical and western blot analyses indicated that SIRT1 positivity and the expression level of SIRT1 in pancreatic cancer tissues were both significantly higher than in adjacent normal pancreatic tissues. This provided preliminary evidence to suggest that SIRT1 is overexpressed in pancreatic cancer cells and may participate in its carcinogenesis. Although SIRT1 expression was increased in most pancreatic cancer samples compared with control tissue, patient-to-patient variability was observed, as SIRT1 was expressed at lower levels in surrounding normal tissue in some individuals. Other research has reported that SIRT1 mRNA was overexpressed in only 1 of 11 pancreatic cancers; however, interestingly, HADC 7, one of the HADC type I family, was significantly overexpressed in 9 out of the 11 cases.32 The lower frequency of SIRT1 overexpression compared with our results may be due to differences in the racial profile of the two patient populations, or differences in the tumor stages evaluated in each study (two cases of IIIIV stage of 11 patients, compared with 26 cases of IIIIV in our study).

An analysis of the clinical characteristics indicated that SIRT1 expression was significantly associated with increased patient age, tumor size, tumor stage and the presence of metastases, which further indicated that SIRT1 acts as a tumor promoter and facilitates the infiltration of pancreatic cancer.

Clinical research on SIRT1 has shown contradictory results. SIRT1 expression is associated with the poor prognosis of diffuse large B-cell lymphoma,33 breast carcinoma34 and gastric cancer,35 whereas other research has revealed that SIRT1 expression is associated with CIMP (CpG island methylator phenotype)-high and MSI (microsatellite instability)-high colon cancer, but is not related to colorectal cancer-specific overall survival.36 In ovarian epithelial tumors, serous carcinomas with a high FIGO stage were shown to express SIRT1 less frequently than lower stage serous carcinomas, and increased expression of SIRT1 correlated with increased overall survival.37 These contradictory results imply that SIRT1 may be a promoter or inhibitor of tumorigenesis, depending on the type and stage of tumor, which needs further research and large-scale clinical investigations.

As SIRT1 was overexpressed in pancreatic cancer and associated with clinical characteristics, we used SIRT1–shRNA to stably transfect PANC-1 cells to elucidate the role of SIRT1. MTT proliferation assays revealed that knockdown of SIRT1 in pancreatic cancer cells resulted in a significant inhibition of proliferation, which indicated that the overexpression of SIRT1 could favor pancreatic cancer cell growth, as other studies had reported for prostate cancer, colon cancer and Burkitt lymphoma cells.28, 38, 39, 40 Our hypothesis was further verified by our observation that xenograft growth of PANC-1–SIRT1–RNAi cells in nude mice was significantly reduced; however, this result was contrary to reports that SIRT1 overexpression inhibited cell proliferation and the formation of colon cancer and also protected hepatic cells from metabolic syndrome-associated hepatic cancer.26, 27

The discrepancies in the effect of SIRT1 between different cancers may be a result of the different tissue types and their mechanisms of tumorigenesis. Although SIRT1 overexpression can prevent metabolic syndrome-associated hepatic cancer, it cannot protect against the formation of 3-methylcholanthrene-induced fibrosarcomas. Due to the important role of SIRT1 in metabolism, it may be able to protect against metabolism-associated carcinogenic damage.27 Pancreatic cancers are frequently associated with mutation of the K-ras oncogene. Apart from promoting tumorigenesis, K-ras can also activate senescence-like growth arrest, which may inhibit tumor growth.41 As SIRT1 is involved in the regulation of apoptosis and senescence, it may be possible that SIRT1 overexpression in pancreatic cancer inhibits K-ras-induced senescence. In this study, the knockdown of SIRT1 by shRNA significantly increased senescence, which was consistent with the cell cycle analysis results of increased G1 DNA content and reduced S phase in PANC-1–SIRT1–RNAi cells, but not in PANC-1-negative RNAi transfected cells. These results imply that inhibition of tumor formation by SIRT1–RNAi may be due to a reduced ability of SIRT1 to prevent K-ras-induced senescence. Our results also demonstrated that SIRT1–RNAi increased sensitivity of PANC-1 cells to 5-FU and Gemcitabine. Although the results of Wang et al. showed that a reduction in SIRT1 enhanced the proliferation of colon cancer cells, it also sensitized tumor cells to chemotherapy agents, which was consistent with our results.27 This suggests that SIRT1 can protect pancreatic cancer cells from chemotherapy-induced apoptosis and SIRT1 knockdowns inhibit the protective effect of increasing chemosensitivity.

We also measured the expression of FOXO3a, p53, Bax and Bcl-2, which have key roles in the regulation of cell apoptosis. FOXO3a, p53 and Bax may promote separate apoptotic pathways within the same cell line.42, 43, 44 Our study showed that the proapoptotic genes, FOXO3a and Bax, were upregulated, whereas the anti-apoptotic gene Bcl-2 was significantly reduced in PANC-1–SIRT1–RNAi cells. These results are in line with reports that SIRT1 can deacetylate FOXO3a and Bax, thus attenuating FOXO- and Bax-induced apoptosis and potentiating FOXO- and Bax-induced cell-cycle arrest.12, 45 Additionally, SIRT1 can activate Bcl-2 during the apoptotic process,8 and thus, the downregulation of SIRT1 could lead to a decrease in Bcl-2. Although there are remarkable similarities in function between p53 and FOXO3a,46 PANC-1 cells carry a mutant loss-of-function p53 gene;47 therefore, our finding that p53 expression was not altered indicates that the involvement of SIRT1 in apoptosis may be p53-independent in PANC-1 cells.

In addition to proliferation, invasiveness is an important characteristic of cancer cells. The invasive ability was significantly decreased in PANC-1–SIRT1–RNAi cells. SIRT1 is a repressor of E-cadherin, which acts by binding to the E-cadherin promoter and forming deacetylase repressor complexes after transcription.22 In this study, E-cadherin in PANC-1 and PANC-1-negative cells was reduced, compared with PANC-1–SIRT1–RNAi cells. In addition, PANC-1 and PANC-1-negative RNAi cells expressed and secreted high levels of MMP-2 and MMP-9, which were significantly reduced in PANC-1–SIRT1–RNAi cells. This suggested that the knockdown of SIRT1 could reduce the metastatic ability of tumors by increasing E-cadherin and reducing MMP expression.

In summary, this research indicates that SIRT1 may act as a promoter of pancreatic cancer tumorigenesis, and SIRT1-targeted RNAi could provide a novel molecular therapeutic intervention. Our results and research in different cancers suggest that the role of SIRT1 may depend on the type of tumor, stage of tumor and the mechanism of tumorigenesis. Therefore, further investigation into SIRT1 expression and its association with patient survival and response to treatment is critical if SIRT1-targeted RNAi treatments are to be applied to different tumor types. Additionally, as SIRT1 is a regulator of both apoptosis and senescence, the SIRT1 RNAi may have some adverse effect on normal cell, such as enhancing the senescence. Thus, though some research had reported that SIRT1-targeted RNAi does not affect normal cell lines,40 the effects of SIRT1-targeted RNAi in normal human cells should be thoroughly evaluated in case of potential adverse effects.

Materials and methods


Rabbit anti-SIRT1, anti-Bax, anti-Bcl-2 polyclonal antibodies and mouse anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) monoclonal antibody were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Rabbit anti-FOXO3a, MMP-2 and MMP-9 polyclonal antibodies were purchased from Abcam Biotechnology (Cambridge, UK). Rabbit anti-E-cadherin polyclonal antibody was purchased from Cell Signaling Technology (Danvers, MA, USA). Goat anti-rabbit/mouse horseradish peroxidase (HRP)-conjugated secondary antibodies were purchased from Boster Company (Wuhan, China). ECL reagent was obtained from Thermo Scientific (Pittsburgh, PA, USA). Nude mice were obtained from the Experimental Animal Center, Institute of Organ Transplantation, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China. Propidium iodide, RNase A and MTT were purchased from Sigma-Aldrich (St Louis, MO, USA).

Patients and tumor tissues

Human pancreatic cancer tissues and the corresponding non-tumor tissues were acquired from 49 patients at the Pancreatic Disease Institute, Union Hospital, Wuhan, China between 2008 and 2010, including 31 males and 18 females, aged between 32 and 80 years old, with an average age of 54.75±12.05 years. None of the patients received chemotherapy or radiotherapy before surgical excision. The diagnosis of pancreatic cancer was confirmed by pathology. Immediately after surgical removal, tissue samples were either snap-frozen in liquid nitrogen for mRNA extraction or fixed in 10% buffered formalin solution and paraffin-embedded for histological analysis.

Immunohistochemical staining

Sections of 4 μm of each paraffin-embedded tissue were mounted onto silane-treated slides, deparaffinized with xylene and rehydrated through graded alcohols into distilled water, treated with 3% H2O2 for 15 min to quench endogenous peroxidase activity and blocked with 1% preimmune goat serum for 30 min. After overnight incubation at 4 °C with rabbit polyclonal antibody to SIRT1 (1:100), slides were washed with Tris-buffered saline containing 0.05% Tween 20, treated with rabbit anti-goat horseradish peroxidase-labeled secondary antibody (1:200) for 1 h and incubated for 5 min with 3, 3-diaminobenzidine substrate to visualize the antigen–antibody complex, and then lightly counterstained with hematoxylin. Negative control sections were stained with only secondary antibody. Two experienced pathologists independently analyzed SIRT1 staining while blinded to the clinicopathological data and clinical outcomes of the patients. Cases with >30% positive cells in a section were considered positive for SIRT1 expression.

Cell culture

PANC-1 was purchased from the ATCC (Manassas, VA, USA). Cells were grown in RPMI-1640 medium supplemented with 10% heat-inactivated fetal bovine serum, 1% L-glutamine, 100 U ml−1 penicillin and 100 U ml−1 streptomycin in a humidified atmosphere containing 5% CO2 at 37 °C. PANC-1 cells were grown in a monolayer culture using 25-cm2 culture flasks and were periodically detached from the flask surface using 0.25% trypsin–ethylene diamine tetraacetic acid solution. All media and supplements were obtained from Gibco Life Technologies Ltd (Paisley, UK). Cells in the logarithmic phase of growth were used for all studies described.

Transfection of SIRT1 shRNA expression plasmids

SIRT1 (NM_012238) shRNA were designed by GeneChem RNA Technologies (GeneChem Co. Ltd, Shanghai, China), and the target gene plasmids, negative and positive control plasmids containing green fluorescent protein were also synthesized by GeneChem. The target sequences were as follows: 5′-CCTTCTGTTCGGTGATGAA-3′ (487–505, 19 bp, SIRT1–RNAi-1); 5′-CCATTCTTCAAGTTTGCAA-3′ (974–992, 19 bp, SIRT1–RNAi-2); 5′-TGAAGTGCCTCAGATATTA-3′ (1424–1442, 19 bp, SIRT1–RNAi-3); 5′-CGTACGCGGAATACTTCA-3′(19 bp, negative RNAi); 5′-CTGAAGACCTGAAGACA AT-3′ (19 bp, GAPDH-positive RNAi). The plasmids were transfected into PANC-1 cells using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. Knockdown was confirmed by qPCR and western blotting 48 h after transfection. The target gene plasmids, with the most effective knockdown, and the negative-RNAi plasmid were stably transfected into PANC-1 cells and monoclones were selected using the finite deliquation method into 96-well plates in media supplemented with 1000 μg ml−1 G418 (Gibco-BRL, Gaithersburg, MD, USA) for 2 months, after which time they were maintained with 600 μg ml−1 G418. The purity of stable transfectants was evaluated by detecting green fluorescent protein expression using a FACS Calibur flow cytometer (BD Biosciences, San Jose, CA, USA).

Cell proliferation assay

The effect of SIRT1 on proliferation was demonstrated using MTT assay. Briefly, 200 cells per well were incubated in 96-well plates for 1 to 10 days, 20 μl of 5 mg ml−1 MTT reagent was added to each well, incubated at 37 °C for 4 h; the supernatant was removed and 150 μl of dimethyl sulfoxide (Sigma-Aldrich, Carlsbad, CA, USA) was used to dissolve the resultant formazan crystals. Absorption values were read at 570 nm using a multiscanner autoreader (Dynatech MR 5000, Chantilly, VA, USA). The MTT assay was repeated three times with six replicates.

Cell cycle and apoptosis assay with flow cytometry and TUNEL assay

Cells for cell cycle analysis were washed in cold phosphate-buffered saline (PBS), fixed in cold ethanol (70%) for 30 min at 4 °C, washed twice with cold PBS and stained with 50 μg ml−1 propidium iodide in the presence of 25 μg ml−1 of RNase A and analyzed using flow cytometery. Data from 10 000 events per sample were collected and calculated with ModFit LT software (Verity Software House, Topsham, ME, USA). The experiment was performed in duplicates and repeated twice.

Cells for apoptosis analysis were collected and washed twice in cold PBS and resuspended in 100 μl of annexin-binding buffer. And then, 5 μl of AnnexinV-FITC (Invitrogen) conjugate and 2 μl of propidium iodide (1 mg ml−1) were added and this suspension was incubated for 15 min at room temperature. The samples were then further diluted with 400 μl of annexin-binding buffer. Cells were identified by a FACS Calibur flow cytometer (BD Biosciences). The in situ Cell Death Detection Kit from Roche (Roche, South San Francisco, CA, USA) was used to perform TUNEL assay according to the manufacturer's instructions, and cells were visualized with light microscopy. The apoptotic index was calculated as the number of apoptotic cells/total number cells × 100%.

Cell senescence assay

Cellular senescence is characterized by induction of SA-β-Gal48 and SA-β-Gal staining was carried out as previously described.49 Freshly adherent cells were fixed in 4% paraformaldehyde at room temperature, washed in PBS and frozen at −80 °C in optimum cutting temperature compound. Cells were immersed in SA-β-Gal staining solution at 37 °C and viewed under bright field microscopy at room temperature. The quantification of SA-β-Gal-positive cells was performed by counting cells at 10 random fields per dish and assessing the percentage of SA-β-Gal-positive cells from at least 100 cells per field.

Cell viability assay

The chemosensitivity was evaluated using the MTT assay. Cells were plated at 5000 cells per well in 96-well plates. Twenty-four hours after plating, cells were incubated at various concentrations of 5-FU (Sigma-Aldrich) or Gemcitabine (Eli Lilly and Co., Suresnes Cedex, France) for 24, 48 or 72 h and the MTT assay was performed. The survival rate of cells was calculated as cell viability (%)=A570 (experiment)/A570 (control group) × 100%. Each concentration had six replicate wells and the experiment was repeated three times.

In vitro invasion assay

The invasive assay was performed as previously described using transwell cell culture chambers (8 μM pore size polycarbonate membrane, Costar, Cambridge, MA, USA).50 Briefly, membranes were coated with Matrigel (BD Biosciences, Bedford, MA, USA). Cells were resuspended at 1 × 106/ml in RPMI-1640 supplemented with 0.5% fetal bovine serum and 100 μl cell suspension was loaded into the upper chamber, and the lower chamber was loaded with 600 μl of RPMI-1640 with 10% fetal bovine serum. After 24 h incubation at 37 °C, the filter was fixed in 4% paraformaldehyde and stained with hematoxylin (Sigma-Aldrich). The cells on the upper side of the filter were wiped off with a cotton swab and the number that had invaded to the undersurface of the membrane was counted in 10 randomly selected microscopic fields. Each assay was performed in triplicate.

Quantitative real-time PCR (qPCR)

Total RNA from PANC-1, PANC-1-negative RNAi and PANC-1–SIRT1–RNAi cells was isolated with Trizol (Invitrogen), treated with DNAse 1 and reverse-transcribed to cDNA using the Enhanced avian HS RT-PCR kit (Sigma-Aldrich) according to the manufacturer's protocol. Real-time PCR was performed to quantify human SIRT1, FOXO3a, p53, Bax, Bcl-2, E-cadherin, MMP-2 and MMP-9, relative to GAPDH, with three replicates of each cDNA sample. PCR reactions consisted of 1 μl of cDNA, 1 μl of SYBR Green PCR I (Qiagen, Tokyo, Japan), 5 μl of 10 × buffer, 1.6 μl (10 mM) of primers, 7 μl of MgCl2 (25 mM), 0.5 μl of Taq DNA polymerase (1 U μl−1 TOYOBO, Tokyo, Japan), 1 μl of deoxyribonucleotide triphosphate and 33 μl distilled water. After denaturation of the enzyme for 2 min at 94 °C, the PCR was performed for 45 cycles, at 94 °C for 30 s, 57 °C for 30 s, and 72 °C for 30 s. Fluorometric PCR was performed using the FTC-2000 system (FUNGLYN, Shanghai, China) and relative expression was determined relative to GAPDH, using the 2−ΔΔCT method51 to determine crossing points.

Western blotting analysis

Whole cell extracts were prepared in a lysis buffer (50 mM Tris-HCl, pH 8.0; 150 mM NaCl; 1 mM ethylene diamine tetraacetic acid; 1% Triton-X 100; 1 mg ml−1 aprotinin; 1 mg ml−1 leupeptin and 100 μg ml−1 phenylmethylsulfonyl fluoride). Protein content was determined using the BCA Protein Assay (Pierce, Rockford, IL, USA) and 40 μg protein was separated on 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels under denaturing conditions, transferred onto nitrocellulose (Millipore, Billerica, MA, USA), blocked with 5% non-fat milk powder in Tris-buffered saline for 1 h and incubated with the relevant primary antibodies overnight at 4 °C (SIRT1, 1:800; FOXO3a, 1:500; p53, 1:500; Bax, 1:400; Bcl-2, 1:200; E-cadherin, 1:1000; MMP-2 1: 1000; MMP-9 1:1 000; GAPDH: 1:400). After washing with Tris-buffered saline containing 0.1% Tween 20, the membrane was incubated with secondary HRP-conjugated antibodies (Pierce), visualized using ECL (Pierce) and exposed to a Kodak X-OMAT film (Sigma-Aldrich). Band intensities were quantified relative to GAPDH using the Alpha DigiDoc 1201 (Alpha Innotech, San Leandro, CA, USA).

Implantation of cells into nude mice

Enzymatically dissociated cells were washed in PBS and maintained at a temperature of 4 °C until subcutaneous injection of 1 × 107 cells in the dorsum of 4-week-old female nude mice. Mice were maintained in a specific pathogen-free room under constant temperature and humidity. Seven weeks after injection, mice were killed by cervical dislocation, and the presence of each tumor nodule was confirmed by the In Vivo Multispectral Imaging System (Carestream Health, Rochester, NY, USA) and necropsy. Tumor volumes were calculated with the formula: ((length+width)/2)3 × 0.5236.52 Tumors were fixed in 4% paraformaldehyde, embedded in paraffin or optimal cutting temperature compound and 4 μm sections were cut and stained with hematoxylin–eosin. SIRT1 protein expression in primary tumors was determined using immunohistochemistry and western blotting. Cryosectioned and apoptotic cells were detected using the TUNEL assay and cell senescence was quantified using SA-β-Gal staining, as previously described.

Statistical analysis

Results are expressed as mean±s.d. Differences between groups were evaluated using the Student's t-test and Fisher's exact test. P-values less than 0.05 were accepted as significant.


  1. 1.

    , , , , , et al. Cancer statistics, 2008. CA Cancer J Clin 2008; 58: 71–96.

  2. 2.

    , , , , . Invasion and metastasis in pancreatic cancer. Mol Cancer 2003; 2: 14.

  3. 3.

    , , , , , . Microarray analysis of gene expression profile of multidrug resistance in pancreatic cancer. Chin Med J (Engl) 2007; 120: 1743–1752.

  4. 4.

    , . Histone deacetylases and cancer. Oncogene 2007; 26: 5420–5432.

  5. 5.

    , . Sirtuins: critical regulators at the crossroads between cancer and aging. Oncogene 2007; 26: 5489–5504.

  6. 6.

    , , , , , . The absence of SIR2alpha protein has no effect on global gene silencing in mouse embryonic stem cells. Mol Cancer Res 2003; 1: 402–409.

  7. 7.

    . The Sir 2 family of protein deacetylases. Curr Opin Chem Biol 2005; 9: 431–440.

  8. 8.

    , , , . Transcriptional regulation of neuronal genes and its effect on neural functions: NAD-dependent histone deacetylase SIRT1 (Sir2alpha). J Pharmacol Sci 2005; 98: 200–204.

  9. 9.

    . Sirtuins as potential targets for metabolic syndrome. Nature 2006; 444: 868–874.

  10. 10.

    , . Mammalian sirtuins—emerging roles in physiology, aging, and calorie restriction. Genes Dev 2006; 20: 2913–2921.

  11. 11.

    , , , , , et al. hSIR2(SIRT1) functions as an NAD-dependent p53 deacetylase. Cell 2001; 107: 149–159.

  12. 12.

    , , , , , et al. Stress-dependent regulation of FOXO transcription factors by the SIRT1 deacetylase. Science 2004; 303: 2011–2015.

  13. 13.

    , , , , , et al. Mammalian SIRT1 represses forkhead transcription factors. Cell 2004; 116: 551–563.

  14. 14.

    , . Sirtuins in mammals: insights into their biological function. Biochem J 2007; 404: 1–13.

  15. 15.

    , , , , , . Involvement of mitochondria and caspase pathways in N-demethyl-clarithromycin-induced apoptosis in human cervical cancer HeLa cell. Acta Pharmacol Sin 2006; 27: 1622–1629.

  16. 16.

    , , , , , et al. Calorie restriction promotes mammalian cell survival by inducing the SIRT1 deacetylase. Science 2004; 305: 390–392.

  17. 17.

    , . The fundamental role of epigenetic events in cancer. Nat Rev Genet 2002; 3: 415–428.

  18. 18.

    , , , , , et al. SIRT1 is significantly elevated in mouse and human prostate cancer. Cancer Res 2007; 67: 6612–6618.

  19. 19.

    , , , , , et al. Histone deacetylases in acute myeloid leukaemia show a distinctive pattern of expression that changes selectively in response to deacetylase inhibitors. Leukemia 2005; 19: 1751–1759.

  20. 20.

    , , , , , et al. Function of the SIRT1 protein deacetylase in cancer. Biotechnol J 2007; 2: 1360–1368.

  21. 21.

    , , . Hyaluronan-mediated CD44 interaction with p300 and SIRT1 regulates beta-catenin signaling and NFkappaB-specific transcription activity leading to MDR1 and Bcl-xL gene expression and chemoresistance in breast tumor cells. J Biol Chem 2009; 284: 2657–2671.

  22. 22.

    , , . E-cadherin transcriptional down-regulation by epigenetic and microRNA-200 family alterations is related to mesenchymal and drug-resistant phenotypes in human breast cancer cells. Int J Cancer 2010; 126: 2575–2583.

  23. 23.

    , , , , , et al. The significances of lymph node micrometastasis and its correlation with E-cadherin expression in pT1-T3N0 gastric adenocarcinoma. J Surg Oncol 2008; 97: 125–130.

  24. 24.

    , , , , . Control of multidrug resistance gene mdr1 and cancer resistance to chemotherapy by the longevity gene sirt1. Cancer Res 2005; 65: 10183–10187.

  25. 25.

    . SIRT1: tumor promoter or tumor suppressor? Med Hypotheses 2006; 67: 341–344.

  26. 26.

    , , , , , et al. SirT1 is an inhibitor of proliferation and tumor formation in colon cancer. J Biol Chem 2009; 284: 18210–18217.

  27. 27.

    , , , , , et al. Sirt1 improves healthy ageing and protects from metabolic syndrome-associated cancer. Nat Commun 2010; 1: 1–8.

  28. 28.

    , . Role of p53 in the anti-proliferative effects of Sirt1 inhibition in prostate cancer cells. Cell Cycle 2009; 8: 1478–1483.

  29. 29.

    , , . Cancer-specific functions of SIRT1 enable human epithelial cancer cell growth and survival. Cancer Res 2005; 65: 10457–10463.

  30. 30.

    , , , , , et al. Sirt1 inhibitor, Sirtinol, induces senescence-like growth arrest with attenuated Ras-MAPK signaling in human cancer cells. Oncogene 2006; 25: 176–185.

  31. 31.

    , , , , , et al. Overexpression of neuropilin-1 promotes constitutive MAPK signalling and chemoresistance in pancreatic cancer cells. Br J Cancer 2005; 93: 233–241.

  32. 32.

    , , , , , et al. High histone deacetylase 7 (HDAC7) expression is significantly associated with adenocarcinomas of the pancreas. Ann Surg Oncol 2008; 15: 2318–2328.

  33. 33.

    , , , , , et al. SIRT1 expression is associated with poor prognosis of diffuse large B-cell lymphoma. Am J Surg Pathol 2008; 32: 1523–1531.

  34. 34.

    , , , , , et al. Expression of DBC1 and SIRT1 is associated with poor prognosis for breast carcinoma. Hum Pathol 2011; 42: 204–213.

  35. 35.

    , , , , , et al. Expression of DBC1 and SIRT1 is associated with poor prognosis of gastric carcinoma. Clin Cancer Res 2009; 15: 4453–4459.

  36. 36.

    , , , , , et al. SIRT1 histone deacetylase expression is associated with microsatellite instability and CpG island methylator phenotype in colorectal cancer. Mod Pathol 2009; 22: 922–932.

  37. 37.

    , , , , , et al. Expression and prognostic significance of SIRT1 in ovarian epithelial tumours. Pathology 2009; 41: 366–371.

  38. 38.

    , , , , , et al. Antitumor activity of a small-molecule inhibitor of human silent information regulator 2 enzymes. Cancer Res 2006; 66: 4368–4377.

  39. 39.

    , , , . Role of sirtuin histone deacetylase SIRT1 in prostate cancer. A target for prostate cancer management via its inhibition? J Biol Chem 2009; 284: 3823–3832.

  40. 40.

    , , . Cancer-specific functions of SIRT1 enable human epithelial cancer cell growth and survival. Cancer Res 2005; 65: 10457–10463.

  41. 41.

    , , , , , . Differential oncogenic Ras signaling and senescence in tumor cells. Cell Cycle 2004; 3: 1201–1207.

  42. 42.

    , , , , , . Phosphorylation of forkhead transcription factors by erythropoietin and stem cell factor prevents acetylation and their interaction with coactivator p300 in erythroid progenitor cells. Oncogene 2002; 21: 1556–1562.

  43. 43.

    . p53, the cellular gatekeeper for growth and division. Cell 1997; 88: 323–331.

  44. 44.

    . Double identity for proteins of the Bcl-2 family. Nature 1997; 387: 773–776.

  45. 45.

    , , , , , et al. Bax-inhibiting peptide protects cells from polyglutamine toxicity caused by Ku70 acetylation. Cell Death Differ 2007; 14: 2058–2067.

  46. 46.

    , , . The ins and outs of FoxO shuttling: mechanisms of FoxO translocation and transcriptional regulation. Biochem J 2004; 380(Part 2): 297–309.

  47. 47.

    , , . Caspase-3 activation downstream from reactive oxygen species in heat-induced apoptosis of pancreatic carcinoma cells carrying a mutant p53 gene. Pancreas 2001; 22: 255–260.

  48. 48.

    , , , . Senescence-associated (beta)-galactosidase reflects an increase in lysosomal mass during replicative ageing of human endothelial cells. J Cell Sci 2000; 113(Part 20): 3613–3622.

  49. 49.

    , , , , , et al. A biomarker that identifies senescent human cells in culture and in aging skin in vivo. Proc Natl Acad Sci USA 1995; 92: 9363–9367.

  50. 50.

    , , , , . Autocrine motility factor signaling enhances pancreatic cancer metastasis. Clin Cancer Res 2004; 10: 7775–7784.

  51. 51.

    , . Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001; 25: 402–408.

  52. 52.

    , , , , , et al. SirT1 is an inhibitor of proliferation and tumor formation in colon cancer. J Biol Chem 2009; 284: 18210–18217.

Download references


This study was funded by grants from the National Science Foundation Committee (NSFC) of China (Grant number: 30600594 and 30972900). Grant sponsor: National Natural Science Foundation of China; Grant number 30600594 and 30972900.

Author information


  1. Pancreatic Disease Institute, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China

    • G Zhao
    • , J Cui
    • , J-G Zhang
    • , Q Qin
    • , Q Chen
    • , T Yin
    • , S-C Deng
    • , Y Liu
    • , L Liu
    • , B Wang
    •  & C-Y Wang
  2. Department of General Surgery, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China

    • K Tian
    •  & G-B Wang


  1. Search for G Zhao in:

  2. Search for J Cui in:

  3. Search for J-G Zhang in:

  4. Search for Q Qin in:

  5. Search for Q Chen in:

  6. Search for T Yin in:

  7. Search for S-C Deng in:

  8. Search for Y Liu in:

  9. Search for L Liu in:

  10. Search for B Wang in:

  11. Search for K Tian in:

  12. Search for G-B Wang in:

  13. Search for C-Y Wang in:

Competing interests

The authors declare no conflict of interest.

Corresponding authors

Correspondence to G Zhao or C-Y Wang.

Supplementary information

About this article

Publication history







Supplementary Information accompanies the paper on Gene Therapy website (http://www.nature.com/gt)

Further reading