High expression of Galectin-1 in pancreatic stellate cells plays a role in the development and maintenance of an immunosuppressive microenvironment in pancreatic cancer

Authors

  • Dong Tang,

    1. Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, People's Republic of China
    2. Jiangsu Province Academy of Clinical Medicine Institute of Tumor Biology, 300 Guangzhou Road, Nanjing 210029, People's Republic of China
    Search for more papers by this author
    • D.T. and Z.Y. contributed equally to this work

  • Zhongxu Yuan,

    1. Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, People's Republic of China
    2. Jiangsu Province Academy of Clinical Medicine Institute of Tumor Biology, 300 Guangzhou Road, Nanjing 210029, People's Republic of China
    Search for more papers by this author
    • D.T. and Z.Y. contributed equally to this work

  • Xiaofeng Xue,

    1. Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, People's Republic of China
    2. Jiangsu Province Academy of Clinical Medicine Institute of Tumor Biology, 300 Guangzhou Road, Nanjing 210029, People's Republic of China
    Search for more papers by this author
  • Zipeng Lu,

    1. Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, People's Republic of China
    2. Jiangsu Province Academy of Clinical Medicine Institute of Tumor Biology, 300 Guangzhou Road, Nanjing 210029, People's Republic of China
    Search for more papers by this author
  • Ye Zhang,

    1. Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, People's Republic of China
    2. Jiangsu Province Academy of Clinical Medicine Institute of Tumor Biology, 300 Guangzhou Road, Nanjing 210029, People's Republic of China
    Search for more papers by this author
  • Hui Wang,

    1. Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, People's Republic of China
    2. Jiangsu Province Academy of Clinical Medicine Institute of Tumor Biology, 300 Guangzhou Road, Nanjing 210029, People's Republic of China
    Search for more papers by this author
  • Minyong Chen,

    1. Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, People's Republic of China
    2. Jiangsu Province Academy of Clinical Medicine Institute of Tumor Biology, 300 Guangzhou Road, Nanjing 210029, People's Republic of China
    Search for more papers by this author
  • Yong An,

    1. Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, People's Republic of China
    2. Jiangsu Province Academy of Clinical Medicine Institute of Tumor Biology, 300 Guangzhou Road, Nanjing 210029, People's Republic of China
    Search for more papers by this author
  • Jishu Wei,

    1. Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, People's Republic of China
    2. Jiangsu Province Academy of Clinical Medicine Institute of Tumor Biology, 300 Guangzhou Road, Nanjing 210029, People's Republic of China
    Search for more papers by this author
  • Yi Zhu,

    1. Jiangsu Province Academy of Clinical Medicine Institute of Tumor Biology, 300 Guangzhou Road, Nanjing 210029, People's Republic of China
    Search for more papers by this author
  • Yi Miao,

    Corresponding author
    1. Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, People's Republic of China
    2. Jiangsu Province Academy of Clinical Medicine Institute of Tumor Biology, 300 Guangzhou Road, Nanjing 210029, People's Republic of China
    • Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, 300 Guangzhou Road, Nanjing 210029, People's Republic of China
    Search for more papers by this author
    • Tel: +8615805189988, Fax: +862583781992

  • Kuirong Jiang

    Corresponding author
    1. Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, People's Republic of China
    2. Jiangsu Province Academy of Clinical Medicine Institute of Tumor Biology, 300 Guangzhou Road, Nanjing 210029, People's Republic of China
    • Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, 300 Guangzhou Road, Nanjing 210029, People's Republic of China
    Search for more papers by this author
    • Tel: +8613951701663, Fax: +862583781992


Abstract

Galectin-1 is implicated in making tumor cells immune privileged, in part by regulating the survival of infiltrating T cells. Galectin-1 is strongly expressed in activated pancreatic stellate cells (PSCs); however, whether this is linked to tumor cell immune escape in pancreatic cancer is unknown. Galectin-1 was knocked down in PSCs isolated from pancreatic tissues using small interfering RNA (siRNA), or overexpressed using recombinant lentiviruses, and the PSCs were cocultured with T cells. CD3+, CD4+ and CD8+ T cell apoptosis was detected by flow cytometry; T cell IL-2, IL-4, IL-5 and INF-γ production levels were quantified using ELISA. Immunohistochemical analysis showed increased numbers of PSCs expressed Galectin-1 (p < 0.01) and pancreatic cancers had increased CD3+ T cell densities (p < 0.01) compared to normal pancreas or chronic pancreatitis samples. In coculture experiments, PSCs that overexpressed Galectin-1 induced apoptosis of CD4+ T cells (p < 0.01) and CD8+ T cells (p < 0.05) significantly, compared to normal PSCs. Knockdown of Galectin-1 in PSCs increased CD4+ T cell (p < 0.01) and CD8+ T cell viability (p < 0.05). Supernatants from T cells cocultured with PSCs that overexpressed Galectin-1 contained significantly increased levels of Th2 cytokines (IL-4 and IL-5, p < 0.01) and decreased Th1 cytokines (IL-2 and INF-γ, p < 0.01). However, the knockdown of PSC Galectin-1 had the opposite effect on Th1 and Th2 cytokine secretion. Our study suggests that the overexpression of Galectin-1 in PSCs induced T cell apoptosis and Th2 cytokine secretion, which may regulate PSC-dependent immunoprivilege in the pancreatic cancer microenvironment. Galectin-1 may provide a novel candidate target for pancreatic cancer immunotherapy.

Pancreatic ductal adenocarcinoma (PDAC) is an extremely aggressive malignancy, which is resistant to currently available systemic therapies. PDAC has one of the worst prognoses of all human cancers with incidence rates nearly equal to mortality rates.1 There is evidence that excessive desmoplasia play a crucial role in the aggressive behavior of pancreatic cancer,2 which impedes effective systemic treatments on a molecular level. Pancreatic stellate cells (PSCs) are stellate-shaped mesenchymal pancreatic cells, which have been identified as important regulators of desmoplasia in PDAC.3

In their quiescent state, PSCs can be identified by the presence of vitamin A-containing lipid droplets in the cytoplasm and the expression of desmin and glial-fibrillary-acidic protein (GFAP).4 In response to pancreatic damage or stress, PSC are transformed into an activated myofibroblast-like phenotype. Activated PSCs express α-smooth muscle actin (α-SMA), and synthesize excessive amounts of ECM proteins, including Collagen I and III, fibronectin and matrix-degrading enzymes such as MMPs.5–7 Activated PSCs have a variety of cell functions and can promote the proliferation, migration, invasion and metastasis of pancreatic cancer cells;1, 8, 9 however, the factors that PSCs secrete to advance pancreatic cancer progression remain largely unknown.

Galectin-1, a member of the galectin family of β-galactoside-binding proteins, is a homodimer of 14-kDa subunits possessing two β-galactoside-binding sites. It participates in a variety of biological functions including cell-cell and cell-matrix interactions and cell growth.10 Galectin-1 acts via both intracellular sugar-independent interactions with other proteins, and extracellular sugar-dependent autocrine or paracrine interactions with β-galactoside-containing glycoconjugates.11

It has become increasingly evident that Galectin-1 expression is dysregulated in cancer, which suggests that Galectin-1 may support the invasion and metastasis formation of cancer cells,12 promote tumor angiogenesis13 and protect tumors from host immune responses.14 Previous studies have shown that extracellular Galectin-1 can inhibit T-cell proliferation15 and induce apoptosis of activated T-cells,14 suggesting Galectin-1 may be involved in the mechanism of tumor immune evasion. Recently, it has been shown that Galectin-1 is strongly expressed in isolated culture-activated PSCs and induces chemokine production and proliferation.16, 17 Despite the fact that Galectin-1 is expressed in pancreatic cancer tissues,17 the functional role of endogenous Galectin-1 in pancreatic cancer has not been yet characterized.

The purpose of our study was to investigate the effects of altered Galectin-1 expression levels in primary PSCs on primary peripheral T cells. PSCs were isolated from resected fresh pancreatic tissue and Galectin-1 was knocked down using small interfering RNA (siRNA) or overexpressed using recombinant lentiviruses. PSCs that overexpress Galectin-1 induced T cell apoptosis and increased Th2 cytokine secretion, which may enhance their ability to escape immune surveillance by tumors. The knockdown of Galectin-1 in PSCs may provide a novel target for pancreatic cancer immunotherapy.

Abbreviations

GFAP: glial-fibrillary-acidic protein; PDAC: pancreatic ductal adenocarcinoma; PSCs: pancreatic stellate cells; qRT-PCR: quantitative reverse transcription-polymerase chain reaction; siRNA: small interfering RNA; α-SMA: α-smooth muscle actin; UICC: International Union against Cancer Classification

Material and Methods

Patients and pancreatic tissues

Pancreatic cancer tissue samples were obtained from 66 patients undergoing pancreaticoduodenectomy for pancreatic cancer and from 18 patients with chronic pancreatitis at the First Affiliated Hospital of Nanjing Medical University, China. The pancreatic cancer patients comprised 45 men and 21 women with a median age of 55 years (range, 37–83 years), and the chronic pancreatitis patients comprised 13 men and five women with a median age of 54.5 years (range, 27–71 years). The clinicopathological characteristics of the patients are described in Table 1. Ten normal pancreatic control tissue samples were obtained from patients undergoing partial pancreatic resections for bile duct or duodenal ampullary cancer. All patients provided informed consent for their participation in the study, which was approved by the Ethical Committee of Nanjing Medical University, China. The tissues adjacent to the specimens were evaluated histologically in all patients, according to the criteria of the World Health Organization. Tumor stages were assessed according to the International Union against Cancer Classification (UICC).

Table 1. Patient clinicopathological characteristics
inline image

Cells and culture conditions

Human PSCs were isolated from fresh normal pancreas tissue using a Nycodenz gradient density gradient centrifugation, as detailed previously,18 and cells were maintained as previously described.19 The PSC cell type was confirmed by immunohistochemical staining for desmin, α-SMA and stellate or spindle-shaped cell morphology.1, 20 Cells from passage numbers 2 to 5 were used for all assays. Coculture experiments were performed as previously described;14, 21 briefly, peripheral blood mononuclear cells from healthy donors were isolated through a Ficoll-Hypaque gradient centrifugation and separated using a Nylon Fiber Column (Wako Pure Chemical Industries, Osaka, Japan) according to the manufacturer's instructions. Cells were activated for 72 hr with 5 μg/mL PHA in RPMI media that contained 10% FCS, and then cultured with 20 ng/mL IL-2. IL-2 was removed 24 hr before the apoptosis assays. Activated T cells (1 × 106) were then cocultured with monolayers (1 × 106 cells) of naive PSCs, or PSCs with either Galectin-1 overexpression or knocked down Galectin-1, scrambled shRNA expression, or recombinant Galectin-1 (1 μg/mL) for 24 hr at 37°C. After coculture, all T cells were harvested and sequentially evaluated by flow cytometry analysis, and the supernatant were harvested for cytokine secretion analysis by ELISA.

Quantitative reverse transcription-polymerase chain reaction

Total RNA was extracted from cultured cells using Trizol reagent (Invitrogen, Beijing, China) according to the manufacturer's instructions. Quantitative reverse transcription-polymerase chain reaction (qRT-PCR) was performed using a SYBR® Premix Ex Taq™ Reverse Transcription-PCR kit (TaKaRa, Shiga, Japan) and the ABI PRISM® 7500 Fast Real-Time PCR System (Applied Biosystems, Carlsbad, CA). The sequences of the gene specific primers designed for human Galectin-1 and β-actin were purchased from Invitrogen. The primers for Galectin-1 were 5′-GAGGTGGC TCCTGACGCTAA-3′ (Forward) and 5′-CCTTGCTGTTGCACACGATG-3′ (Reverse). The primers for β-actin were 5′-AGAAAATCTGGCACCACACC-3′ (Forward) and 5′-TAGCACAGCCTGGATAGCAA-3′ (Reverse). Each reaction mixture was incubated at 42°C for 30 min for reverse transcription to synthesize first-strand cDNA from total RNA using the gene-specific primer. q-PCR was initiated by incubation at 95°C for 30 sec predenaturation, followed by 40 cycles of 95°C for 5 sec, 60°C for 34 sec. The expression level of Galectin-1 was calculated using a standard curve constructed with total RNA from normal PSCs, normalized to β-actin internal control, and expressed as a ratio of Galectin-1 to β-actin expression. All samples were run in triplicate.

Immunohistochemical staining and evaluation

Pancreatic tissue samples were fixed by immersion in 4% paraformaldehyde overnight at 4°C and then embedded in regular paraffin wax and cut into 4-μm sections. Hematoxylin and eosin (H&E) staining and immunohistochemical analyses were performed as previously described.22, 23 For immunohistochemistry, tissue sections were deparaffinized and rehydrated in PBS. After antigen retrieval with target retrieval solution, endogenous peroxidase activity was blocked by incubation in 0.3% hydrogen peroxide, and sections were then blocked in 10% normal goat serum. Sections were incubated with mouse monoclonal anti-Galectin-1 (1:200; sc-166618, Santa Cruz Biotechnology, Santa Cruz, CA), anti-Desmin (1:200; sc-70961, Santa Cruz Biotechnology), anti-CD3 (1:200; sc-52382, Santa Cruz Biotechnology), anti-α-SMA antibodies (1:200; MA1-37027; Thermo, Fremont, CA) and anti-Cytokeratin 19 antibody (1:200; MA106329;ABR) overnight at 4°C, followed by a biotinylated rabbit anti-mouse IgG antibody (1:200; Vector Laboratories, Burlingame, CA). Finally, samples were incubated with peroxidase-conjugated streptavidin (Boster, Wuhan, China). Color was developed by incubating the slides for several minutes with diaminobenzidine and nuclei were counterstained with hematoxylin. For negative controls, the primary antibody was replaced with nonspecific rabbit IgG. The results of immunohistochemical staining were interpreted by two experienced pathologists and the mean density of staining was calculated using the ImagePro Plus 6.0 software (ImagePro, Bethesda, MD).

Western blotting analysis

Western blotting was performed as previously described.24 Briefly, cells were lysed in SDS buffer and 100 μg total cellular protein was separated on 10% or 10–20% gradient SDS-polyacrylamide gels, transferred to PVDF, and incubated with mouse anti-Galectin-1 antibody (1:500) overnight at 4°C. After incubation with peroxidase-conjugated rabbit anti-mouse IgG antibody (Cell Signaling Technologies, Beverly, MA), proteins were visualized using ECL (GE Healthcare, Chalfont St. Giles, UK). α-Tubulin was used as the loading control.

Quantification of Galectin-1 using Enzyme Linked Immunosorbent Assay (ELISA)

PSC culture supernatants and PSC and T cell coculture supernatants were harvested and stored at −80°C. The level of Galectin-1, IL-2, IL-4, IL-5 and INF-γ were determined using a commercial ELISA kit (Boster, Wuhan, China). The ELISA detection sensitivities were ≥156 pg/mL for Galectin-1, ≥15.6 pg/mL for IL-2, IL-5 and INF-γ, and ≥7.8 pg/mL for IL-4.

Flow cytometry analysis

T cells were obtained from coculture supernatants, suspended in PBS containing 1% fetal bovine serum at a density of 1 × 105 cells per 100 μL and incubated with 20 μL CD3-FITC, CD8-PE-CY5 or CD4-FITC (BD Pharmingen, Franklin Lakes, NJ) at room temperature in the dark for 20 min. Each samples was centrifuged for 4 min at 450×g and resuspended in 250 μL 1% FCS containing 0.1% NaN3 (FACS-buffer). Next, 5 μL annexin V-APC (Bender MedSystems, San Diego, CA) was added, mixed gently and incubated for 15 min at room temperature in the dark. Labeled cells were analyzed with a BD FACSCalibur flow cytometer (Beckman Coulter, Fullerton, CA). FSC/SSC-gating was performed for the evaluation of T cells.

Preparation and transduction of recombinant lentiviruses

The plasmids used in preparation of recombinant lentiviruses have been described previously.25, 26 The hGalectin-1 gene fragment was excised from a human cDNA library and cloned into pHAGE-CMV-MCS-IZsGreen between the BamHI and XhoI restriction enzyme sites (Fig. 3a). Galectin-1 specific oligonucleotides21, 27 (Galectin-1 RNAi #1, 5′-CCGGGCTGCCAGATGGATACGAATCTCGAGATTCGTATCCATCTGGCAGCTTTTTTG-3′; Galectin-1 RNAi #2, 5′-CCGGCAGCAACCTGAATCTCAAACTCGAGTTTGAGATTCAGGTTGCTGTTTTTTG-3′) or the scrambled oligonucleotide (SCR, 5′-CCTAAGGTTAAGTCGCCCTCGCTCGAGCGAGGGCGACTTAACCTTAGG-3′) were ligated into the pLKO.1-puro vector (Fig. 3b). To produce recombinant lentiviruses, 293FT cells (Invitrogen) were cultured according to the manufacturer's instructions and cotransfected using Lipofectamine™ 2000 reagent (Invitrogen) with 6 μg pHAGE-CMV-hGal-1-MCS-IZsGreen, 4.5 μg psPAX2 and 1.5 μg pMD2.G for hGalectin-1 overexpressing lentivirus, and 13 μg pLKO.1-U6-shRNA-puro, 5 μg pVSVG and 7.5 μg delta8.91 for the hGalectin-1 knockdown lentivirus. After 16 hr, the media were changed and supernatants containing the recombinant lentiviruses were harvested 48 hr later, filtered through 0.45 μm filters and used for transduction experiments.

For viral infection, PSCs were incubated for 6 hr with virus supernatant diluted in culture medium supplemented with 8 μg/mL polybrene. After infection, the Galectin-1-overexpressing PSCs were selected by expression of GFP, using a FACSCalibur flow cytometer (Fig. 3c), and PSCs expressing the Galectin-1 shRNA knockdown plasmid were selected by culture in media containing 2.0 μg/μL puromycin. Galectin-1 expression was confirmed in whole-cell extracts using Western blotting and qRT-PCR, and in supernatants using ELISA before cultured subclones were used in experiments.

Statistical analysis

Values are expressed as the mean ± standard deviation. All experiments were repeated three times. One way ANOVA and t-tests were performed using SPSS 13.0 to compare differences between groups. All p-values were two-sided, and p-values ≤0.05 were considered to be statistically significant.

Results

Pancreatic cancer stromal cells express both α-SMA and Galectin-1

To identify the location of Galectin-1 in pancreatic cancer tissues, we performed immunohistochemical staining on serial sections of pancreatic cancer tissues for Galectin-1 and α-SMA, which is a marker for activated PSCs. The majority of stromal cells surrounding cancer cells expressed α-SMA (Figs. 1a–1c), and Galectin-1 was only expressed in α-SMA-positive regions, which suggests that activated PSCs express Galectin-1 (Figs. 1d–1f). We observed that the staining intensity of Galectin-1 protein expression significantly increased gradually from normal pancreas to chronic pancreatitis to PDAC (Figs. 1d–1f, and 1n,p < 0.01). Stromal cells that expressed Galectin-1 were scattered in tissue samples of chronic pancreatitis, but in pancreatic cancer samples were shown to surround the tumor, forming a tight fibrotic barrier. These findings indicate that PSCs that express Galectin-1 may play an important role for pancreatic cancer cells.

Activated PSCs express Galectin-1 and confine CD3+ T cells to the tumor mesenchyme

As Galectin-1 is known to contribute to T cell homeostasis,28, 29 CD3 was detected using immunohistochemistry in serial sections of samples of PDAC, chronic pancreatitis and normal pancreas. The intensity of CD3 staining in PDAC was significantly higher than in chronic pancreatitis and normal pancreas tissues (Figs. 1g1i, and 1o,p < 0.01). Similar to the pattern of Galectin-1 expression, CD3+ T cells were scattered in chronic pancreatitis and normal pancreas tissues (Figs. 1h and 1i), and surrounded the tumor cells in PDAC (Fig. 1g). In PDAC, CD3+ T cells surrounded the tumor cells with few infiltrating T cells in the parenchyma of the tumor. Galectin-1 staining was located in the mesenchymal cell population, which was composed of activated PSCs, around the tumor foci (Figs. 1d and 1g). These results suggest that CD3+ T cells were abnormally increased around the tumor foci, and the surrounding activated PSCs with high expression levels of Galectin-1 might comprise an immunological barrier.

Figure 1.

Characterization of activated PSC α-SMA, Galectin-1 and CD3 expression in 66 cases of pancreatic cancer, 18 cases of chronic pancreatitis and 10 normal pancreatic control tissue samples. (a) Activated PSCs expressing α-SMA located adjacent to, and surrounding cancer cells; (b) activated PSC expressing α-SMA distributed diffusely within regions of fibrosis in chronic pancreatitis; (c) sparsely distributed α-SMA expression in normal pancreatic tissue; (d) strong Galectin-1 expression in activated PSCs which appear to create a barricade between tumor and stromal tissue in PDAC; (e) Galectin-1 expression colocalized with activated PSCs in chronic pancreatitis and (f) weak Galectin-1 expression in normal pancreas; (g) CD3 is expressed around PDAC tumor foci; (h) diffuse expression of CD3 is observed in chronic pancreatitis; (i) CD3 is sporadically expressed in the normal pancreas; (j–l) H&E staining of samples of pancreatic cancer, chronic pancreatitis and normal pancreas. The arrows indicate tissue damage in chronic pancreatitis; (m) CK19 immunohistochemical staining for pancreatic cancer cells. Original magnification for images (a–i, m) × 100 and images (j–l) × 200; And (n) quantification of the mean Galectin-1 expression intensity in different pancreatic tissues. Δp < 0.05 vs. NP, *p < 0.01 vs. NP, #p < 0.01 vs. CP. (o) Mean expression intensity of CD3 in different pancreatic tissues. *p < 0.01 vs. NP, #p < 0.01 vs. CP. NP: normal pancreas; CP: chronic pancreatitis; PC: pancreatic cancer. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Activated PSCs express Galectin-1

Human PSCs were isolated from the normal pancreas of patients undergoing surgery for bile duct cancer. In freshly isolated primary human PSC cultures (2 days after culture), which were positive for desmin and negative for α-SMA expression (Figs. 2a and 2b),7 weak Galectin-1 expression was detected using immunofluorescent staining (Fig. 2c); however, in activated PSCs cultured for 7 days, the PSC activation marker, α-SMA, was expressed (Fig. 2d), and strong cytoplasmic Galectin-1 expression was also observed (Fig. 2f).

Figure 2.

Primary human PSCs were isolated from the normal pancreatic tissues of patients undergoing resections for bile duct cancer or duodenal ampullary cancer. Immunohistochemical staining of desmin (a), α-SMA (b) and Galectin-1 (c) in unactivated PSCs. Expression of the activation markers α-SMA (d), negative control of α-SMA (e) and Galectin-1 (f) in activated PSC cultures was confirmed using immunohistochemistry. The unactivated PSCs were obtained on day two and the activated PSCs were obtained on day 7 of primary culture. In all images, immunohistochemical staining was indicated using the brown diaminobenzidine color reaction and sections were counterstained with hematoxylin. Original magnifications, ×200. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

In vitro characterization of Galectin-1 overexpression and knockdown in PSCs

To obtain PSCs with different expression levels of Galectin-1 and imitate the levels of protein in different pancreatic conditions, we performed lentiviral hGalectin-1 and shRNA-Galectin-1 transduction in PSCs. PSCs that overexpressed hGalectin-1 (Galectin-1-PSC) were purified by screening for GFP using flow cytometry, and PSCs that contained the shRNA-Galectin-1 knockdown plasmids (shRNA-Galectin-1-PSC#1 and #2) were selected with puromycin. qRT-PCR was performed to examine Galectin-1 expression in naive PSCs (Naive), overexpression plasmid control GFP-PSCs (GFP), overexpressing Galectin-1-PSC (Over), shRNA scrambled control-PSC (Scr), and knockdown shRNA-Galectin-1-PSC #1 (Sh-1) and shRNA-Galectin-1-PSC #2 (Sh-2). Galectin-1 mRNA expression in Galectin-1-PSC was significantly greater than plasmid control transduced GFP-PSCs or naive PSCs (Fig. 3g). Galectin-1 mRNA expression in shRNA-Galectin-1-PSC #1 and #2 cells was reduced by more than 85% compared to control scramble-PSCs or naive PSCs (Fig. 3g). These changes were confirmed at the protein level using Western blotting; Galectin-1 expression increased in Galectin-1-PSCs and was reduced in shRNA-Galectin-1-PSC #1 and #2 cells compared to naive PSCs (Fig. 3e). To determine the secretion of Galectin-1 by different PSC lines, the concentration of Galectin-1 in the supernatants were quantified by an ELISA. Galectin-1-PSC secreted a larger quantity Galectin-1 than naive PSCs, while shRNA-Galectin-1-PSC #1 and #2 did not secrete detectable levels of Galectin-1 (Fig. 3h).

Figure 3.

Schematic representations of the control and Galectin-1-overexpressing/knockdown lentiviral constructs and the characterization of human PSCs in vitro. (a) Schematic representations of the control and Galectin-1-overexpressing pHAGE lentiviral constructs, encoding hGalectin-1 connected via an internal ribosomal entry site (IRES) to a bicistronic GFP reporter gene; (b) Schematic representations of the control and pLKO.1 lentiviral constructs, encoding a shRNA-Galectin-1 and puromycin resistance gene, which were used to knockdown Galectin-1 expression; (c and d) GFP expressing human PSCs transduced with the pHAGE lentivirus were selected using flow cytometry and observed by epifluorescent and light microscopy, which confirmed the presence of the Galectin-1-overexpressing lentivirus in all cells; (e and f) typical Western blot which indicates the increased hGalectin-1 expression in Gal-1-PSCs transduced with the Galectin-1 overexpression lentivirus, and reduced hGalectin-1 expression in shRNA-Gal-1-PSCs transduced with the Galectin-1 knockdown construct #1 or #2, compared to naive and the control lentivirus transduced PSCs; (g) quantitative analyses of the differentiated phenotype of human PSCs. Increased hGalectin-1 mRNA expression was observed in the Gal-1-PSCs, and reduced hGalectin-1 mRNA expression was observed in the shRNA-Gal-1-PSCs; (h) Galectin-1 expression in PSC supernatants was quantified using ELISA and increased in Gal-1-PSCs and reduced in shRNA-Gal-1-PSC supernatants. Key: Naive untransduced PSC (Naive), and PSCs transduced with GFP control (GFP), Galectin-1 overexpression construct (Over), scrambled shRNA control (Scr), shRNA-Galectin-1 #1 (Sh-1), shRNA-Galectin-1 #2 (Sh-2). *p < 0.01, #p > 0.05 (vs. naive PSCs). All experiments were repeated three times. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Endogenous Galectin-1 expression in PSCs contributed to T cell-induced immunosuppression

With the purpose of investigating the mechanism responsible for PSC specific expression of Galectin-1 in tumors, we assessed the functional effects on tumor infiltrating cells of PSCs that expressed different levels of Galectin-1. Activated T cells were cocultured with PSCs that overexpressed Galectin-1, or PSCs in which Galectin-1 expression had been knocked down. Twenty-four hours later, the T cell populations were assessed by double-color annexin-V staining and CD3 flow cytometry. Coculture with PSCs that overexpressed Galectin-1 induced significant apoptosis in CD3+ T cells, compared to naive PSCs coculture; increased numbers of viable CD3+ T cells were obtained after coculturing with PSCs with Galectin-1 knockdowns (Fig. 4a).

Figure 4.

Effect of PSC Galectin-1 expression levels on T cell apoptosis in coculture. T cells were cocultured with PSCs for 24 hr, isolated from the supernatant, labeled with CD3-FITC, CD8-PE-CY5 or CD4-FITC and annexin V-APC and analyzed using flow cytometry. (a) Galectin-1-overexpressing PSCs significantly increased apoptosis in CD3+ T cells (*p < 0.05, **p < 0.01, #p > 0.05 vs. naive PSCs.), and 50 mM β-lactose inhibited this effect; (b) Galectin-1-overexpressing PSCs significantly increased apoptosis in CD4+ T cells (**p < 0.001), and 50 mM β-lactose inhibited this effect; and (c) Galectin-1-overexpressing PSCs significantly increased apoptosis in CD8+ T cells (*p < 0.05), and 50 mM β-lactose inhibited this effect. rGal-1: recombinant Galectin-1. *p < 0.05, **p < 0.01, #p > 0.05 vs. naive PSCs. All experiments were repeated three times. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

β-lactose, a competitive inhibitor of Galectin-1, completely blocked the ability of endogenous Galectin-1 to induce T cell apoptosis, confirming the effect was specifically induced by Galectin-1. These studies suggest that endogenous Galectin-1 expressed by PSCs decreases the viability of infiltrating activated T cells in tumors. Next, we investigated if the effect of Galectin-1 secreted by PSCs depended on T cells subtype, using triple-color annexin-V, CD3 and CD4 or CD8 flow cytometry. Compared to T cells cocultured with naive PSCs, Galectin-1-overexpressing PSCs induced significant levels of apoptosis in both CD4+ and CD8+ T cells. Galectin-1 knockdown PSC cocultures had an increased number of viable CD4+ and CD8+ T cells compared to naive PSCs cocultures (Figs. 4b and 4c). Taken together, these results suggest that endogenous Galectin-1 secreted by PSCs selectively reduced the viability of CD4+ and CD8+ T cells.

Endogenous Galectin-1 contributes to immune privilege by skewing the Th1/Th2 cytokine balance

As Galectin-1 is considered to be a novel regulator of immune cell homeostasis in cancer,14, 28 we investigated whether Galectin-1 is involved in CD4+ T cell function. Th1/Th2 cytokines were quantified in the supernatants of activated T cells from PSC cocultures. The supernatants from T cells cocultured with Galectin-1-overexpressing PSCs contained significantly more Th2 cytokines (IL-4 and IL-5), but markedly decreased Th1 cytokines (IL-2 and INF-γ), and this effect was specifically blocked by β-lactose (Fig. 5). However, the knockdown of PSC Galectin-1 had the opposite effect on Th1 and Th2 cytokine secretion (Fig. 5). These results further suggest that the overexpression of Galectin-1 promoted immune privilege by decreasing Th1-type cytokine production and promoting Th2-type cytokine production, and the knockdown of Galectin-1 can reverse this effect on T cell Th1/Th2 production.

Figure 5.

Effect of PSC Galectin-1 expression levels on T cell Th1/Th2 cytokine secretion in coculture. PSCs and T cells were cocultured for 24 hr, and the IL-2, IL-4, IL-5 and INF-γ contents were quantified using ELISA. Galectin-1-overexpressing PSCs significantly reduced the secretion of (a) IL-2 and (b) INF-γ from T cells (*p < 0.01) and significantly increased (c) IL-4 and (d) IL-5 secretion from T cells (*p < 0.01). *p < 0.05, **p < 0.01, #p > 0.05 vs. naive PSCs. All experiments were repeated three times.

Discussion

In our study, we have described the colocalization of activated PSCs and significant Galectin-1 expression in PDAC and also identified a mechanism by which expression of Galectin-1 in PSCs may promote tumor immunosuppression in pancreatic cancer. Immunohistochemical analysis indicated Galectin-1 was expressed by activated PSCs and α-SMA-positive myofibroblasts, which surround pancreatic cancer cells. Activated PSCs create a barrier between CD3+ T cells and tumor cells, which suggests that PSCs that express Galectin-1 can prevent CD3+ T cells from infiltrating the pancreatic tumor.30 Using functional in vitro assays, we showed lentivirus-driven overexpression of Galectin-1 in activated PSCs significantly decreases CD3+ T cell viability, and skewed the cytokine secretion balance towards a Th2 immune response. Taken together, these data directly suggest that endogenous Galectin-1 expressed by activated PSCs plays a role in the development and maintenance of the unique pancreatic cancer Th2 immunosuppressive microenvironment.

PSCs have been identified as a principal source of the excessive extracellular matrix observed in chronic pancreatitis5 and PDAC.3 On their activation by various autocrine or paracrine oxidative stresses, quiescent PSCs transform into an activated myofibroblast-like phenotype and produce excessive amounts of growth factor, cytokine and extracellular matrix proteins,16, 31 which promote proliferation, migration, and invasion of pancreatic cancer cells.1, 9, 20 Strikingly, PSCs have also been detected in metastatic pancreatic cancer foci in the livers of nude mice, which suggests that PSCs can migrate and establish a potentially favorable microenvironment for metastatic cells at distant sites.8 It is well known that PSCs in the pancreatic cancer microenvironment can promote tumor growth and invasion, protect tumor cells from apoptosis, and potentially create a barrier to the delivery of therapeutic compounds.32–34 We have shown that the activation of PSCs in vitro is associated with increased Galectin-1 expression, and as Galectin-1 can modulate PSC function and activation,16, 35 it is likely to play other roles in PSC biology.

Although the role of PSCs in pancreatic cancer progression has been extensively discussed, the role of PSCs in the escape from immune surveillance observed in pancreatic cancer has not been investigated. Galectin-1, which is expressed in activated PSCs, has been previously shown to regulate the selective immunoregulation of T-cell fate,36 which partly explains the host immune responses commonly observed in pancreatic cancer. In our study, Galectin-1-overexpressing PSCs decreased T cell viability, significantly decreased IL-2, INF-γ Th1 cytokine secretion, and increased IL-4, IL-5 Th2 cytokine secretion in vitro. Furthermore, the knockdown of Galectin-1 in PSCs had had the opposite effect on T cell viability and Th1/Th2 cytokine secretion. In murine model experiments, Galectin-1-treated transplanted mice showed reduced production of the Th1 cytokines, IL-2 and INF-γ, compared to animals treated with vehicle alone.37 The therapeutic effect of Galectin-1-DC was accompanied by increased percentage of apoptotic T cells and reduced number of IFN-γ-secreting CD4+ T cells in pancreatic lymph nodes.38 The Galectin-1 knockdown has been shown to partially restore the proliferation of CD4+ and CD8+ T cells.39 Additionally, the inhibition of Galectin-1 function could increase the growth and activation of CD8+ CTLs and their ability to kill cancer cells.40 The ability of Galectin-1 to selectively skew cytokine secretion may be linked to the observation that Th1 cells express a repertoire of cell surface glycans, to which Galectin-1 may bind and induce subsequent cell death; whereas, Th2 cells may be protected from binding of Galectin-1 due to differential cell surface glycoprotein sialylation.21, 41 In pancreatic cancer, the presence of tumor infiltrating lymphocytes with high Th2/Th1 ratios is indicative of a poor prognosis.42 The ability of activated PSCs that express Galectin-1 to influence the cytokine profile suggests that Galectin-1 plays a role in tumor immunosuppression and prognosis by modulating T cell responses in the pancreatic cancer microenvironment.

The role of tumor immunosuppression during PDAC progression has been well studied. Pancreatic cancer cells express a diverse range of cancer-associated antigens, which can potentially be recognized by T cells; these are detected in the bone marrow and peripheral blood of all PDAC patients.43 Pancreatic cancer-associated stromal cells, such as activated PSCs and infiltrating macrophages produce high concentrations of immunosuppressive cytokines,44, 45 and as these stromal cells are recruited and activated in PDAC, a predominantly immunosuppressive microenvironment develops. Our observation of a higher T-cell density in PDAC, compared to chronic pancreatitis, despite the immunosuppressive environment of the primary cancer, may be due to significant antigen production by pancreatic tumors. This is particularly the case in later stage tumors, which reach the bone marrow and stimulate effector T cell differentiation and their release into the circulation.43 Fibrotic tissue composed of activated PSCs may play an important role in the mechanisms of immune surveillance escape in PDAC. We have shown that the expression of endogenous Galectin-1 at high levels in activated PSCs can modulate immune cell function by controlling the proliferation and apoptosis of effector T cells. Galectin-1 can also block T cell activation and induce T cell death,28, 29, 46 and the ability of Galectin-1 to skew the Th1/Th2 cytokine balance in our coculture experiments may also contribute to PDAC immune privilege.21, 47 These combined effects of Galectin-1 on T cell function may explain the low levels of infiltrating T cells in pancreatic carcinomas.30 Our new findings, taken together with evidence from previous research, indicates that the ability of PDAC to escape immune surveillance is not due to altered T cell specific immunity; instead, the tumor cells are surrounded by an activated PSC barrier, which express high levels of Galectin-1 that can inhibit T-cell activity and maintain an in situ anergic state.

Human galectins are involved in a variety of biological and pathological processes, which include cell adhesion, apoptosis, differentiation, immune regulation and tumor evasion.48 Galectin-3, another member of the beta-galactoside-binding lectin family, has a similar role in the modulation of immune responses as Galectin-1.48 T cells play an important role in cancer immunosurveillance and tumor destruction. However, tumor cells alter immune responses by modulating immune cells through antigen stimulation and immunoregulatory cytokines, and as a result, the effector T cells become anergic.49 This anergic state is correlated with the absence of colocalization of the T cell receptors (TCR) and the CD8 coreceptors, which could be recognized by Galectin-3. This suggests that exracellular Galectin-3 decreases the TCR mobility on the surface of anergic T lymphocytes.50 Thus, Galectin-3 may function as an immune regulator to inhibit T cell immune responses and promote tumor growth, thus providing a new mechanism for tumor immune tolerance. The difference of Galectin-1 and Galectin-3 in pancreatic cancer is that Galectin-1 exhibited in fibroblasts and extracellular matrix cells (mainly localized in activated PSCs) around the cancer mass, whereas Galectin-3 is strongly present in most pancreatic cancer cells.17 Therefore, Galectin-1 secreted by activated PSCs around the tumor foci and Galectin-3 secreted by cancer cells may work together to maintain the immunosuppressive microenvironment found in pancreatic cancer.

Our study has provided further information on the role of the Galectin-1 function; however, the role of the high levels of Galectin-1 expression in activated PSCs has not yet been fully elucidated, and is of interest due to the ability of PSCs to regulate immune privilege in pancreatic cancer. We suggest that Galectin-1 may represent a novel potential therapeutic target to enhance immune surveillance in pancreatic cancer. The inhibition of Galectin-1 expression could promote host antitumor responses in primary pancreatic cancer by improving tumor T cell infiltration, reducing T cell apoptosis and increasing immune surveillance. A better understanding of the mechanisms that regulate Galectin-1 expression in PSCs is required before future therapeutic strategies could be developed.

In conclusion, our study indicates that Galectin-1 expression in activated PSCs regulates selective immune privilege in pancreatic cancer, and the inhibition of Galectin-1 may provide a novel therapeutic target to promote the antitumor immune response in primary pancreatic cancer.

Acknowledgements

The authors thank Prof. Lu Chun (Department of Microbiology and Immunology, Nanjing Medical University, Nanjing 210029, P.R. China) for kindly providing the lentiviral packaging system consisting of pHAGE-CMV-MCS-IZsGreen, psPAX2 and pMD2.G.

Ancillary