The Hedgehog inhibitor cyclopamine impairs the benefits of immunotherapy with activated T and NK lymphocytes derived from patients with advanced cancer

Abstract

Hedgehog (Hh) signaling is activated in various types of cancer and contributes to the progression, proliferation, and invasiveness of cancer cells. Many Hh inhibitors are undergoing clinical trial and show promise as anticancer drugs. Hh signaling is also induced in the activated T and NK (TNK) lymphocytes that are used in immunotherapy. Activated TNK lymphocyte therapy is anticipated to work well within a tumor’s hypoxic environment. However, most studies on the immunobiological functions of activated TNK lymphocytes have been conducted on healthy donor samples, under normoxic conditions. In the present study, we evaluated the effects of Hh inhibition and oxygen concentrations on the function of activated TNK lymphocytes derived from patients with advanced cancer. Proliferation, migration, surface NKG2D expression, and cytotoxicity were all significantly inhibited, and IFN-γ secretion was significantly increased upon Hh inhibitor treatment of activated TNK lymphocytes under hypoxic conditions in vitro. Tumors from mice injected with cyclopamine-treated activated TNK lymphocytes showed a significant increase in tumor size and had fewer apoptotic cells compared with the tumors in mice injected with control activated TNK lymphocytes. These results suggest that Hh signaling plays a pivotal role in activated TNK lymphocyte cell function. Combination therapy using Hh inhibitors and activated TNK lymphocytes derived from patients with advanced cancer may not be advantageous.

Introduction

The hedgehog (Hh) signaling pathway is essential for the growth, patterning, and morphogenesis of multiple organs [1, 2]. Recent studies have linked Hh pathway activation and tumor initiation [3]. Soon thereafter, it was reported that Hh signaling was activated in various types of cancers and contributed to the progression and invasiveness of cancer cells [4–6]. As a molecular target drug, Hh inhibitors are thought to be promising therapies for some cancers and are under clinical study [6–10]. Of the Hh inhibitors, vismodegib, a small-molecule inhibitor of Smoothened (Smo), was approved by the US Food and Drug Administration for the treatment of unresectable or metastatic basal cell carcinomas of the skin in January 2012 [11].

Immunotherapy using patient-isolated, activated T and NK (TNK) lymphocytes has recently been developed, and anticancer effects have been reported in many institutions [12–14]. Recently, it was also reported that Hh signaling is active and contributes to cell cycle progression, growth, activation, and cytokine production in peripheral CD4⁺ T cells and activated T cells [15–18]. In addition, the synergic effects of combination therapy including activated TNK lymphocytes and chemotherapeutics, such as temozolomide or paclitaxel, have also been demonstrated [19]. However, the effects of combination therapy, using molecular inhibitors instead of chemotherapy, and immunotherapy are still unclear. Therefore, knowing the benefits and drawbacks of the combination molecular inhibitor plus immunotherapy will be important for the advancement of cancer treatments.

It has been reported that the oxygen concentration in venous blood is 5.3 % and that in the lymphoid organs is 0.5 % [20]. In particular, deep in tumor environments where activated TNK lymphocytes are anticipated to work, oxygen concentrations are thought to be lower than 1.3 % [21]. Importantly, the activated T lymphocytes used for therapy are activated under normoxic conditions ex vivo and are injected intravenously and expected to immediately function in hypoxic conditions [22]. Recently, it was also reported that hypoxia-inducible factor (HIF) 1 − α, which is induced in low oxygen conditions, regulates some key functions in activated TNK lymphocytes [23–25]. However, most data concerning the immunobiological function of the activated TNK lymphocytes have been obtained from healthy donors and normoxic culture conditions, very unlike those in vivo. In addition, we have recently reported that monocyte-derived dendritic cells (DCs) from patients with advanced cancer were functionally impaired compared with healthy DCs [26]. Our goal is to improve the effect of immunotherapy and to better understand the functions of the activated TNK lymphocytes derived from patients with advanced cancer cultured in hypoxic, physiologically relevant, conditions.

Patients and methods

Patients

Thirty-eight patients with advanced cancer who underwent immunotherapy at the Fukuoka General Cancer Clinic (Fukuoka, Japan) were enrolled in this study. The patient profiles are summarized in Table 1. Written informed consent was obtained from all individuals.

Table 1.

Patient characteristics enrolled in this study

No.	Age	Gender	Primary organ	Metastasis
1	77	F	Pancreas
2	81	F	Colon(T)	Peritoneum
3	60	M	Rt. lung	Lung
4	69	M	Rt. lung	Bone
5	80	F	Pancreas	Peritoneum
6	71	M	Stomach	Lymph
7	43	M	Pancreas	Liver, peritoneum
8	64	F	Uterine body	Lung, bone
9	41	F	Breast	Peritoneum, lymph
10	58	M	Stomach	Lymph
11	69	M	Pancreas
12	64	M	Liver
13	64	F	Rectum	Liver, lung, bone
14	60	M	Rectum	Lung
15	60	M	Pancreas	Peritoneum
16	75	M	Prostate
17	81	F	Breast
18	82	F	Gall bladder	Liver
19	80	M	Gall bladder
20	64	F	Rectum	Liver, lung
21	69	F	Ovary	Peritoneum
22	74	F	Bile duct	Bone
23	68	F	Pancreas	Liver
24	60	M	Pharynx	Lymph, lung
25	73	M	Pancreas	Liver
26	54	M	Pancreas	Liver, peritoneum
27	58	M	Rectum	Liver, peritoneum
28	55	F	Bile duct	Liver, peritoneum
29	63	F	Stomach	Peritoneum
30	58	F	Pancreas	Lymph
31	57	M	Gall bladder	Liver
32	57	M	Bile duct	Liver, lymph
33	60	M	Lt. kidney	Rt. adrenal gland
34	61	M	Esophagus	Liver, lymph
35	67	F	Stomach	Peritoneum
36	83	F	Bladder
37	54	F	Lt. lung	Pleura
38	58	F	Rectum	Liver, lymph

Induction of activated TNK lymphocytes

Human peripheral blood mononuclear cells (PBMCs) were obtained by apheresis (HAEMONETICS Co, Stoughton, MA, USA) prior to immunotherapy and stored at −80 °C until used. Thawed PBMCs were washed by RPMI-1640 (Nipro, Osaka, Japan) three times and were cultured in RPMI-1640 supplemented with 0.5 % human serum, 100 μg/ml penicillin (Meijiseika, Tokyo, Japan), and 100 μg/ml streptomycin (Meijiseika) (hereafter referred to as RPMI medium). After overnight culture, the non-attached fraction was collected and used as resting lymphocytes in this study. Where noted, the thawed PBMCs were cultured in RPMI medium supplemented with 200 U/ml IL-2 (Primmune Inc. Kobe, Japan) in culture plates coated with 5 μg/ml anti-CD3 monoclonal antibody (OKT3, JANSSEN PHARMACEUTICAL K.K., Tokyo, Japan) for 7 days. Thereafter, the lymphocytes were transferred to oxygen-permeable culture bags and were cultured in RPMI medium with 200 U/ml IL-2 for an additional 7–14 days. Then, the lymphocytes were collected as activated TNK lymphocytes and were used clinically or experimentally. Induced activated TNK lymphocytes and resting lymphocytes were cultured under normoxic and hypoxic conditions for experiments. For normoxic conditions, cells were cultured in 5 % CO₂ and 95 % air. For hypoxic conditions, cells were cultured in 1 % O₂, 5 % CO₂, and 94 % N₂ using a multigas incubator (Sanyo, Tokyo, Japan). Cyclopamine, an Smo inhibitor purchased from Certificate of Analysis (North York, Canada), was diluted in 99 % ethanol.

Cell cycle analysis

Activated TNK lymphocytes were treated with the Hh signaling inhibitor, cyclopamine. Lymphocytes were fixed in ice-cold 75 % ethanol for at least 1 h. Cell pellets were washed twice with cold phosphate-buffered saline (PBS) and were incubated for 30 min at room temperature in 1 ml PBS containing 50 μg propidium iodide (Sigma-Aldrich, St. Louis, MO, USA), 0.1 % Triton X-100, 1 mM EDTA, and 0.5 mg ribonuclease A (Sigma-Aldrich). After staining, samples were analyzed using FACScan (BD Biosciences, San Jose, CA, USA) at 20,000 events per sample. Data from flow cytometry were analyzed with the ModFit LT software program (Verity Software House, Topsham, ME, USA) and the CellQuest software program (BD Bioscience).

Semiquantitative reverse transcription (RT) PCR

Total RNA was extracted by using High Pure RNA Isolation kit (Roche Diagnostics Gmbh, Mannheim, Germany). The sequences of the primers used were as follows: Gli1, forward 5′-TACATCAACTCCGGCCAATAGG-3′, reverse 5′-CGGCGGCTGACAGTATAGGCA-3′; and glyceraldehyde-3-phosphate dehydrogenase, forward 5′-CCACCCATGGCAAATTCCATGGCA-3′, reverse 5′-TCTAGACGGCAGGTCAGGTCCACC-3′. Amplification conditions comprised an initial denaturation step for 2 min at 95 °C followed by 30 cycles of 94 °C for 1 min, 58 °C for 1 min, and 72 °C for 2 min. RT-PCR products were separated on an ethidium bromide 2 % agarose gel and visualized with a Molecular Imager FX (Bio-Rad, Hercules, CA).

Western blotting

Whole-cell protein extraction was performed with M-PER Reagents (Pierce Biotechnology, Rockford, IL, USA) according to the manufacturer’s instructions. Protein concentration was determined using a NanoDrop 1000 (Thermo Scientific, Waltham, MA, USA), and whole-cell extracts (50 μg) were separated by electrophoresis on a sodium dodecyl sulfate polyacrylamide gel and transferred to Protran nitrocellulose membranes (Whatman GmbH, Dassel, Germany). Blots were then incubated with anti-Cdk4 (1:200, SC-601, Santa Cruz) and anti-α-tubulin (1:1000, Sigma-Aldrich Co., St Louis, MO, USA) primary antibodies overnight at 4 °C. Blots were then incubated in horseradish peroxidase-linked secondary antibody (Amersham Biosciences, Piscataway, NJ, USA) at room temperature for 1 h. Immunocomplexes were detected using an ECL plus Western Blotting Detection System (Amersham Biosciences) and visualized with a Molecular Imager FX (Bio-Rad). α-Tubulin was used as a protein loading control.

Migration assay

The migration of activated TNK lymphocytes was assessed based on the number of migrated cells using transwell inserts (pore size, 3.0 μm; BD Biosciences, Heidelberg, Germany). Cells were suspended in RPMI medium with 0.5 % serum. Then 1 × 10⁵ cells were added to the upper chamber and incubated for 5 h under normoxic and hypoxic conditions. After incubation, all cells that had migrated from the upper to the lower side of the filter were counted using a light microscope (BX50; Olympus, Tokyo, Japan). Lymphocyte migration was assessed in triplicate wells.

Migration was confirmed by time-lapse imaging. Images of activated lymphocytes treated or not with cyclopamine were acquired every 30 s for 15 min using LuminaVision (Mitani Corporation, Tokyo, Japan). Cells were identified in each image and tracked over time using Image-Pro Analyzer software (Nippon Roper KK, Tokyo, Japan).

Fluorescence-activated cell sorting (FACS) analysis

To analyze the expression of natural killer group 2, member D (NKG2D), CD3, CD4, CD8, CD16, and CD56, on activated TNK lymphocytes, the cells were incubated for 1 h with fluorescein isothiocyanate (FITC)-conjugated anti-CD3, CD4, and CD16 monoclonal antibodies (mAbs), or phycoerythrin (PE)-conjugated anti-NKG2D, CD8, and CD56 mAbs (BD Pharmingen, San Diego, CA, USA). In some experiments, pancreatic cancer cell lines, ASPC-1 and PANC-1, were stained with a PE-conjugated anti-MICA/B mAb (R&D system, Minneapolis, MN, USA). Mouse IgG was used as an isotype control (BD Pharmingen). For staining, cells were washed twice with PBS and incubated in PBS containing 3 % bovine serum albumin (Sigma, St. Louis, MO, USA) and 0.1 % NaN₃ (FACS buffer; Sigma) and the appropriate concentration of labeled mAb for 1 h at 4 °C. After washing with FACS buffer, the fluorescence intensity of the gated lymphocyte populations was measured using a FACSCalibur flow cytometer (BD Pharmingen) and analyzed with CELLQuest software (BD Pharmingen).

Cytotoxic assay

We modified an adherent target detachment assay described elsewhere [27] to measure the cytotoxicity of activated TNK lymphocytes. Target cells (ASPC-1 and PANC-1 cells, 5000 cells/well) were seeded in a 96-well flat bottom plate and were incubated for 24 h to allow adherence. TNK lymphocytes treated or not treated with 10 μM of cyclopamine were cultured for 3 days under normoxic and hypoxic conditions and washed three times to remove any remaining cyclopamine. Then, effector cells (activated TNK lymphocytes) at an effector/target (E/T) ratio of 20:1 were added to the culture. Target and effector cells were incubated for 5 h under normoxic and hypoxic conditions, respectively. To quantify viable adherent cells, a WST-8 reagent solution (Dojindo Laboratories, Kumamoto, Japan) was added to the washed wells, followed by incubation for 1 h at 37 °C. The absorbance at 450 nm was then measured using a microplate reader (ImmunoMini NJ-2300; Nalge Nunc International, Rochester, NY, USA).

Enzyme-linked immunosorbent assay (ELISA)

TNK lymphocytes treated or not treated with 10 μM of cyclopamine were cultured for 3 days under normoxic and hypoxic conditions and washed three times to remove any remaining cyclopamine. Activated TNK lymphocytes and ASPC-1 cells were then cocultured for 24 h (E/T ratio of 20:1) under normoxia and hypoxia, respectively. The supernatant from each well was collected, and the concentration of IFN-γ was measured using an ELISA kit according to the manufacturer’s instructions (Biosource, Carlsbad, CA, USA). The detection limit of the assay was 15.6 pg/ml.

Xenograft model

Xenograft studies were done as reported previously [19]. ASPC-1 cells (1 × 10⁶ cells) in 50 μl RPMI medium were injected subcutaneously into three BALB/c female nude mice (4–6 weeks old) in each group. All animals were obtained from the Charles River Laboratory (Wilmington, MA, USA) and maintained in standard conditions according to institutional guidelines. These animal experiments were approved by the Ethics Committee at Kyushu University (Inspection No. A23-051-3). Primary tumor size was measured every 2 days with calipers; approximate tumor volumes were determined using the formula 0.5ab², where a is the longer and b is the smaller of the two perpendicular diameters. After the tumors had grown to an average size of 10 mm³, 50 μl of saline or activated TNK lymphocytes (1 × 10⁶ cells) treated or not with 10 μM cyclopamine was injected near the tumors subcutaneously three times per week. All animals were euthanized 5 weeks after tumor cell injection.

TUNEL assay

Detection of apoptotic cells in the tumor from xenograft model mouse was performed using the In Situ Apoptosis Detection Kit (Takara Bio, Tokyo, Japan) according to the manufacturer’s protocol.

Statistical analysis

An unpaired two-tailed Student’s t test and one-way analysis of variance (ANOVA) followed by Bonferroni’s test were used for statistical analysis. A P value of <0.05 was considered significant.

Results

The Hh inhibitor cyclopamine decreased proliferation

To evaluate whether Hh signaling was active in activated TNK lymphocytes, Gli1 (a Hh target gene) mRNA levels were investigated. As shown in Fig. 1a, Gli1 mRNA was detected in the activated TNK lymphocytes grown in both normoxic and hypoxic conditions. Cyclopamine, a Hh inhibitor, decreased Gli1 expression in the lymphocytes, indicating efficacy. In contrast, Gli1 mRNA was not detected in resting lymphocytes. Next, the effects of cyclopamine on proliferation were evaluated. Cyclopamine treatment significantly decreased proliferation in activated TNK lymphocytes under hypoxic conditions in a dose-dependent manner (Fig. 1b) with cell viability near 95 % at all dosages. Having determined an effective concentration, we next added 10 μM of cyclopamine and measured proliferation under both normoxic and hypoxic conditions in both resting lymphocytes and activated TNK lymphocytes. In both resting and activated TNK lymphocytes, cyclopamine treatment significantly decreased proliferation independent of oxygen concentration, and hypoxia alone also reduced proliferation in control and cyclopamine-treated groups compared with normoxia. This effect was additive as the cyclopamine treatment in addition to hypoxia resulted in the largest decrease in proliferation (Fig. 1c). To assess the mechanism of the reduced proliferation, cell cycle was analyzed. The percent of cells in G0/G1 in the cyclopamine-treated activated TNK lymphocytes was significantly higher than that in non-treated activated TNK lymphocytes in both normoxic and hypoxic conditions (Fig. 1d). Consistent with this, Cdk4 expression in cyclopamine-treated activated TNK lymphocytes was lower than that of controls (Fig. 1e).

Fig. 1.

The Hh inhibitor cyclopamine decreases proliferation. a Gli1 mRNA expression in resting lymphocytes and activated TNK lymphocytes, treated or not with cyclopamine, was evaluated by PCR. b Activated TNK lymphocytes were seeded in 48-well plate in RPMI medium with or without 1, 10, or 50 μM of cyclopamine and were cultured for 3 days under hypoxic conditions. Cells were counted using a light microscope. C Resting lymphocytes (5 × 10⁵ cells) and activated TNK lymphocytes (2 × 10⁵ cells) were seeded in 48-well plate in RPMI medium with or without 10 μM of cyclopamine and were cultured for 10 days under normoxic (n) and hypoxic (H) conditions. Cells were counted using a light microscope. d Cell cycle analysis of activated TNK lymphocytes treated or not with 10 μM of cyclopamine under normoxia and hypoxia for 3 days was performed by FACS. e Cdk4 expression was evaluated by Western blot. Fifty micrograms of protein was loaded, and α-tubulin was used as a loading control. The graph shows mean ± SD. *P < 0.05

Hh inhibitor decreased random migration

Next, we observed the influence of cyclopamine on migration, another important function of activated TNK lymphocytes, which we analyzed in resting and activated TNK lymphocytes. Cyclopamine treatment did not affect nondirectional migration in resting lymphocytes in either normoxic or hypoxic conditions (Fig. 2a). Interestingly, the number of migrated activated TNK lymphocytes was 300 times higher than that of resting T cells in control conditions. Unlike the resting T cells, cyclopamine treatment significantly decreased activated TNK cell migration in both normoxic and hypoxic conditions (Fig. 2b). Nondirectional migration was also evaluated by time-lapse imaging. Images of resting lymphocytes, activated TNK lymphocytes under hypoxia and normoxia, and cyclopamine-treated activated TNK lymphocytes under hypoxia and normoxia were acquired every 30 s for 15 min (Fig. 2c). The velocities of the five types of lymphocytes in the presence or absence of cyclopamine under normoxia or hypoxia were calculated using Image-Pro Analyzer. The activated TNK lymphocytes treated with cyclopamine under hypoxia showed the significant reduction in mobility compared to those treated with cyclopamine under normoxia or control activated TNK lymphocytes under hypoxia (Fig. 2d).

Fig. 2.

The Hh inhibitor cyclopamine decreased nondirectional migration. a and b Migration of resting (a) and activated (b) lymphocytes treated or not with 10 μM of cyclopamine was determined. The cells that migrated from the upper chamber to a lower chamber were counted by light microscope. c Representative pictures of the cell tracking of migrating resting lymphocytes and activated TNK lymphocytes treated or not with 10 μM of cyclopamine under hypoxia. Magnification is ×10. d The velocity of the indicated lymphocytes was calculated by Image-Pro Analyzer software. The graph shows mean ± SD. *P < 0.05

Hh inhibitor decreased NKG2D but not CD4 and CD8 expressions

NKG2D is a potent activating receptor expressed on natural killer (NK) cells and CD8 T cells, and the interaction of NKG2D with its ligand, MIC A/B, plays a pivotal role in the immune response in tumors [28]. We isolated and activated patient TNK lymphocytes to use in immunotherapy and characterized them in vitro. The activated TNK lymphocytes could be divided into two groups: one having CD3⁺ NKG2D⁺ T cells with CD16⁺ CD56⁺ NK cells (Fig. 3a, Case-1). The other mainly consisted of CD3⁺ NKG2D⁺ T cells (Fig. 3a, Case-2). First, we investigated NKG2D, CD4, and CD8 expressions on the resting lymphocytes. In resting lymphocytes, there were no significant differences in the percent of CD3⁺, NKG2D⁺, CD4⁺, CD8⁺, or CD4⁺ CD8⁺ populations between the cyclopamine-treated and control group nor between normoxic and hypoxic conditions (Fig. 3b). Next, we analyzed NKG2D, CD4, and CD8 expressions in the activated TNK lymphocytes. NKG2D and CD8 expressions in the activated TNK lymphocytes were increased compared with resting lymphocytes, while CD4 expression on the activated TNK lymphocytes was decreased compared with the resting lymphocytes (Fig. 3b and c). Cyclopamine treatment did not affect the percentage of CD4⁺, CD8⁺, and CD4⁺ CD8⁺ populations (Fig. 3c). The ratio of CD4 to CD8 lymphocytes in the activated TNK lymphocytes was decreased compared with the resting lymphocytes (Fig. 3d). Cyclopamine did not further affect the ratio of CD4 to CD8 lymphocytes neither in resting nor in activated TNK lymphocytes (Fig. 3d). The CD3⁺ NKG2D⁺ population in the activated TNK lymphocytes was significantly decreased upon cyclopamine treatment under both normoxic and hypoxic conditions (Fig. 3c). Interestingly, there was no significant difference in the percentage of CD3⁺ NKG2D⁺ population in activated lymphocytes between normoxic and hypoxic conditions in the presence of IL-2 (Fig. 3c). However, without IL-2, the percentage of CD3⁺ NKG2D⁺ population in the activated TNK lymphocytes decreased gradually in both normoxic and hypoxic conditions (Fig. 3e). The percentage of CD3⁺ NKG2D⁺ population in activated TNK lymphocytes cultured without IL-2 under hypoxic conditions was significantly higher than when cultured under normoxic conditions (Fig. 3f). Cyclopamine did not affect CD3⁺ NKG2D⁺ expression both in normoxia and in hypoxia (Fig. 3f).

Fig. 3.

The Hh inhibitor cyclopamine decreased NKG2D but not CD4 or CD8 expressions. a Two representative types of activated TNK lymphocytes used in this study. The surface expressions of NKG2D, CD3, CD56, and CD16 were investigated by FACS. b and c The percent of CD3⁺NKG2D⁺, CD4⁺, CD8⁺, and CD4⁺ CD8⁺ populations in resting (b) and activated (c) lymphocytes treated or not treated with 10 μM of cyclopamine for 3 days under normoxic (n) and hypoxic (H) conditions were analyzed by FACS. d The ratio of CD4⁺ to CD8⁺ was determined in resting and activated TNK lymphocytes under normoxia (n) and hypoxia (H). e Representative results showing NKG2D expression on activated TNK lymphocytes in the absence of IL-2 under normoxia and hypoxia. f NKG2D expression on activated TNK lymphocytes treated or not treated with 10 μM of cyclopamine cultured in the absence of IL-2 for 3 days. The graph shows mean ± SD. *P < 0.05

Hh inhibitor decreased cytotoxicity of activated TNK lymphocytes to pancreatic cancer cells in vitro and in vivo

The cytotoxicity of activated TNK lymphocytes in the hypoxic local tumor site is an important factor of immunotherapy. Thus, we evaluated the cytotoxicity of T lymphocytes treated with cyclopamine. Pancreatic cancer cell lines, ASPC-1 and Panc-1, which express MICA/B on their surface, were used as target cells (Fig. 4a). After coculture of ASPC-1 or Panc-1 cells with activated TNK lymphocytes for 5 h, the number of viable cells was measured. Cyclopamine-treated activated TNK lymphocytes were significantly impaired in their cytotoxicity toward ASPC-1 and Panc-1 cells versus controls under normoxia and hypoxia (Fig. 4b). To investigate the mechanism of the decreased cytotoxicity, the levels of IFN-γ in the supernatant after 1-day coculture of Panc-1 cells with activated TNK lymphocytes treated or not treated with cyclopamine were analyzed. Unexpectedly, IFN-γ secretion in the supernatant of the cyclopamine-treated group was significantly higher than that in controls under hypoxia, while it was significantly lower under normoxia (Fig. 4c). Next, the cytotoxicity of activated TNK lymphocytes treated with cyclopamine was examined in a xenograft model. Activated TNK lymphocytes were derived from Case-2 patients (Fig. 3a) whose activated TNK lymphocytes had few NK cells. The tumor size in mice injected with cyclopamine-treated activated TNK lymphocytes did not decrease as much as in mice injected with control activated TNK lymphocytes (Fig. 4d). When the tumor was evaluated for apoptosis using a TUNEL assay, the number of apoptotic cells in the tumors from mice injected with cyclopamine-treated activated T lymphocytes was significantly reduced compared with those injected with control activated T lymphocytes (Fig. 4e and f).

Fig. 4.

The Hh inhibitor cyclopamine decreased cytotoxicity to pancreatic cancer cells in vitro and in vivo. a The expression of MICA/B in ASPC-1 and Panc-1 cells was analyzed by FACS. Empty histogram, isotype control; filled histogram, MICA/B. b After activation, TNK lymphocytes treated or not treated with 10 μM of cyclopamine were cultured for 3 days under normoxic and hypoxic conditions and washed three times to eliminate excess cyclopamine, and activated TNK lymphocytes (effector cells, E) and ASPC1 or Panc1 (target cells, T) were cocultured for 5 h under hypoxic conditions (E/T ratio = 20:1). Viable cancer cells were measured using the WST-8 reagent. c After activation, the TNK lymphocytes treated or not with 10 μM of cyclopamine were cultured for 3 days under normoxic and hypoxic conditions and washed three times to eliminate excess cyclopamine, and activated TNK lymphocytes and Panc1 cells were cocultured for 24 h under hypoxic conditions (E/T ratio = 20:1). IFN-γ secretion in the supernatant was investigated by ELISA. d Mice bearing tumors were injected subcutaneously with 50 μl of saline, activated TNK lymphocytes (Ly-ct, 1 × 10⁶ cells), or activated TNK lymphocytes treated with 10 μM cyclopamine (Ly-cyc, 1 × 10⁶ cells) in 50 μl saline three times a week. Tumor volume was estimated at the indicated days. e Apoptotic tumor cells from xenograft model were examined by TUNEL staining. Apoptotic tumor cells were labeled with FITC. Representative pictures are shown. Original magnification was ×400. f The graph shows the ratio of the apoptotic cell number in Ly-cyc group to Ly-ct group in Fig. 4 e. The graph shows mean ± SD. *P < 0.05

Discussion

In the present study, clinically derived activated TNK lymphocytes were evaluated after Hh inhibition in normal and reduced oxygen conditions. The activated TNK lymphocytes used were NKG2D high and CD8 high compared with resting lymphocytes, suggesting that the lymphocytes were highly activated. After intravenous injection of activated TNK lymphocytes that had been induced ex vivo, the cells should migrate to and work within the hypoxic tumor environment. Thus, we focused mainly on the analysis of activated TNK lymphocytes under hypoxic conditions. In the present study, we demonstrated that Hh inhibition reduced functions such as proliferation, migration, surface NKG2D expression, and cytotoxicity in activated TNK lymphocytes derived from patients with advanced cancer when cultured in hypoxic conditions.

Consistent with the previous results [15–18], the Hh inhibitor cyclopamine decreased TNK cell proliferation under both normoxic and hypoxic conditions (Fig. 1c), and an increase in G0/G1 phase was observed in the cyclopamine-treated group (Fig. 1d). Previously, we have also shown that pancreatic cancer cells treated with cyclopamine had decreased cell proliferation owing to G0/G1 arrest as well [29]. G0/G1 arrest may contribute to the cyclopamine-induced decrease in TNK cells. Interestingly, there was no significant difference in NKG2D expression between normoxic and hypoxic culture conditions in the presence of IL-2 (Fig. 3c). In vivo, the IL-2 levels are thought to be low, and our results showed that in the absence of IL-2, NKG2D expression on activated TNK lymphocytes was significantly higher in hypoxic conditions compared with normoxic conditions (Fig. 3f). Because many important functions such as proliferation and migration were inhibited under hypoxia in the activated TNK lymphocytes, negative feedback may come into play and the low oxygen may maintain NKG2D expression. Cyclopamine also affected the cytotoxicity of the activated TNK lymphocytes in hypoxic conditions (Fig. 4b). One of the reasons for the decreased cytotoxicity in cyclopamine-treated TNK lymphocytes may be the decrease in surface NKG2D expression. In spite of the decreased cytotoxicity, IFN-γ secretion increased in cyclopamine-treated activated TNK lymphocytes under hypoxia (Fig. 4c). Consistent with our result, Schwinn et al. showed that IFN-γ downregulates NKG2D ligand expression and impairs NKG2D-mediated cytolysis [30]. Although our cytotoxic assay was performed with a 5-h incubation time after cyclopamine washout, and IFN-γ treatment for 5 h did not cause a significant decrease in MICA/B expression in ASPC-1 or Panc-1 cells (data not shown), IFN-γ might inhibit cytotoxic effects in this short term. Cytotoxicity was also confirmed in vivo. The cytotoxic effects were likely mediated by CD3⁺ NKG2D⁺ T lymphocytes because the activated lymphocytes used in the mouse study mainly consisted of CD3⁺ NKG2D⁺ T lymphocytes, not CD16⁺ CD56⁺ NK lymphocytes.

A schematic figure of our results is shown in Fig. 5. We demonstrated a decrease in function of activated TNK lymphocytes under hypoxic conditions by Hh inhibitor treatment. Cancer cells themselves showed decreased invasiveness and proliferation with cyclopamine treatment under hypoxic conditions [29, 31]. Therefore, cyclopamine showed the opposite effect in these cancer cells than in activated TNK lymphocytes under hypoxic conditions. Previously, we reported that cyclopamine treatment caused reduced chemosensitivity to 5-FU or gemcitabine under hypoxia in pancreatic cancer [29]. The present study suggests that combination therapy with Hh inhibitors and activated TNK lymphocytes may not have a synergic effect. However, we should keep in mind that clinically, there may be complicated interactions between oxygen concentration, Hh signaling, cancer cells, and activated TNK lymphocytes. Alternatively, because the Hh inhibitor negatively affects the function of activated TNK lymphocytes, activated TNK lymphocytes may be injected after a washout period of Hh inhibitor.

Fig. 5.

A schematic representation of our results. As shown previously, invasiveness and proliferation of cancer cells themselves decreased when treated with the Hh inhibitor cyclopamine under hypoxic conditions. However, cyclopamine decreased induction, proliferation, migration, NKG2D expression, and cytotoxicity of activated TNK lymphocytes under hypoxia

In conclusion, our results suggest that Hh signaling plays a pivotal role in the maintenance of the functions of activated TNK lymphocytes derived from patients with advanced cancer. We should exorcize caution when using combination therapy with Hh inhibitors and activated TNK lymphocytes derived from patients with advanced cancer.

Abbreviations

Hh

Hedgehog

PBMC

Peripheral blood mononuclear cell