Altered Invariant Natural Killer T cell Subsets and its Functions in Patients with Oral Squamous Cell Carcinoma

Authors


Abstract

Invariant natural killer T (iNKT) cells are glycolipid-reactive T lymphocytes that share receptors and function with natural killer (NK) cells and reportedly play a pivotal role in various immune responses. However, iNKT cells are not well characterized in patients with oral squamous cell carcinoma (OSCC). We investigated the populations and functions of circulating iNKT (CD3+6B11+) cells from thirty-eight patients with OSCC and twenty-eight healthy donors by flow cytometry. Circulating iNKT cells were significantly lower (< 0.01) in patients as compared to those in healthy controls. Further, iNKT subsets revealed a marked decrease in CD4CD8 (double negative, DN) subset with concomitant increase in CD8+ subset in patients as compared to healthy controls (= 0.03 and < 0.01, respectively), whereas CD4+ subset was similarly distributed in both groups. The functional analysis demonstrated that residual iNKT cells from patients had impaired proliferative response to α-galactosylceramide (α-GalCer)-pulsed dendritic cells (DCs) and Th2-like cytokine profile. However, in vitro activation with α-GalCer-pulsed DCs restores IFN-γ expression and enhances antitumour activity to human cancer cells lines (SCC-4, KB and MCF7). It appears that the selectively enriched iNKT subsets and modulation of their function by specific ligand/agonist may be useful for cellular therapy in patients with OSCC. Further, reduced levels of iNKT cells and its DN subset may be used as potential prognostic factors for patients with OSCC.

Introduction

Oral cancer is one of the major threats to public health in the developed world and increasingly in the developing world [1]. An estimated 263,020 new cases and 127,654 deaths from oral cavity cancer occurred in 2008 worldwide. Its prevalence is particularly higher among men possibly due to tobacco habit. In south-central Asia, cancer of the oral cavity ranks among the four most common types of cancer. In India, the age-adjusted incidence rate of oral cancer is 7.5 per 100,000 populations [2]. Although environmental carcinogens (e.g. tobacco and alcohol) are major aetiological factors, impaired immune functions in patients with OSCC have been associated with increasing tumour load and their migration to the distant sites as well as poor prognosis [3, 4].

Natural killer T (NKT) cells are a small population of thymus-derived T cells that express natural killer (NK) cell lineage markers and possess functional properties of both T and NK cells [5, 6]. Type I NKT cells, often referred to as invariant natural killer T (iNKT) cells, express an invariant T cell receptor (TCR)-α chain (Vα24-Jα18 in humans and the homologous Vα14-Jα18 in mice) that is paired with semi-invariant TCR-β chain (Vβ11 in humans and Vβ2, Vβ7 or Vβ8.2 in mice). Invariant natural killer T cells recognize glycolipid antigen α-GalCer (KRN-7000) in the context of non-classical MHC-like molecule CD1d on the antigen-presenting cells (APCs) including dendritic cells (DCs) [7-10]. Upon activation, iNKT cells can rapidly produce large amounts of cytokines including IFN-γ, IL-4 and IL-13. Activated iNKT cells impart a strong antitumour activity through high levels of IFN-γ secretion [11]; hence, ex vivo-expanded iNKT cells have been used successfully in phase II clinical trial in patients with head and neck squamous cell carcinoma. The results showed tumour regression in five of 10 patients, while seven patients showed increased levels of iNKT at the tumour site [12].

There are reports indicating both qualitative and quantitative abnormalities in iNKT cells in patients with different malignancies such as melanoma, prostate, lung and breast cancer [13-16]. While reduced frequency of circulating iNKT cells predicted poor clinical outcome in patients with head and neck cancer [17], increased populations at tumour sites have been reportedly associated with improved prognosis in patients with head and neck carcinoma, neuroblastoma, colorectal carcinoma and leukaemia [17-20]. These data suggest a significant contribution of iNKT cells in the antitumour immune responses in humans. However, the scenario of iNKT cell subset [CD4+, CD4CD8 (double negative, DN) and CD8+] ratios that may play crucial role in cancer prognosis and outcome has been observed in only intrahepatic malignant tumours [21, 22], although interplay among the iNKT subsets in other cancers is not clear.

Identification and characterization of human CD1d-restricted iNKT cells is generally based on the combinations of partially specific monoclonal antibodies (mAbs) that recognize invariant TCR-Vα24 and semi-invariant TCR-Vβ11 or large CD1d tetramer (fluorescent-labelled soluble CD1d molecule loaded with α-GalCer) [23]. However, these methods have some limitations and therefore may lead to an overestimation of iNKT cell number because CD1d-unrestricted and non-invariant Vα24+ TCR can also pair with Vβ11+ TCR and represent the Vα24+Vβ11+ subpopulation [6, 23]. In the present study, we have therefore used more specific 6B11 mAb that recognizes conserved CDR3 loop of the canonical Vα24-Jα18 TCR in combination with anti-CD3, which is considered to be an excellent tool for the analysis of iNKT cells [6, 23, 24]. In this study, we planned to enumerate circulating iNKT cells, their subset distribution and functional characterization in terms of proliferative response to α-GalCer-pulsed DCs in vitro, Th-like cytokine profile and antitumour activity. To the best of our knowledge, this is the first report on iNKT cells in patients with OSCC in the Indian population.

Materials and methods

Study population

This case–control study included 38 biopsy-proven patients with OSCC and 28 age- and gender-matched healthy volunteers as controls. Patients with any other chronic ailment, active disease or receiving chemotherapy or radiation treatment were excluded from the study. The staging of tumour was performed as per American Joint Cancer Committee (AJCC) criteria [25]. The study protocol was approved by the Ethics Committee of the All India Institute of Medical Sciences, New Delhi, and written informed consent was obtained from each study subject.

Isolation of peripheral blood mononuclear cells (PBMCs)

Peripheral blood was collected in heparinized tubes and diluted to 1:2 with fresh sterile phosphate-buffered saline (PBS; pH 7.2). Peripheral blood mononuclear cells were isolated by density gradient centrifugation using Histopaque-1077 (Sigma-Aldrich, St. Louis, MO, USA). Viability of the cells was determined by Trypan Blue dye exclusion test. The cells were aliquoted into three fractions, first one was used for immunophenotyping, second for proliferation assay and the third aliquot was used to generate DCs.

Immunophenotyping by flow cytometry

Immunophenotyping of iNKT cells was performed using following combinations of anti-human mAbs (BD Pharmingen): APC-conjugated anti-CD3, FITC-conjugated anti-iNKT (clone 6B11; mAb against the CDR3 loop of the α-chain from the invariant TCR), FITC-conjugated anti-CD161, PE-conjugated anti-CD4 and PE-conjugated anti-CD8 (Sigma-Aldrich). Briefly, 1 × 106 of freshly isolated PBMCs were stained with appropriate mAbs for 30 min at 4 °C in dark, and cells were washed and fixed with 2% paraformaldehyde, acquired in BD LSR II cytometer and analysed using FACSDiva software (BD Biosciences). Among CD3+-gated cells, fractions of iNKT cells and CD161+ NKT cells were analysed. Further CD4+, CD8+ and CD4CD8 (double negative; DN) iNKT subsets were analysed as fractions of CD3+6B11+ iNKT cells (Fig. 1A). Because both CD4 and CD8 mAbs were labelled with PE, the DN iNKT cell subset was calculated as follows:

display math
Figure 1.

Percentages of natural killer T (NKT) cell subsets. (A) Gating strategy used to enumerate CD3+CD161+ NKT, CD3+6B11+ (iNKT), CD4+ iNKT and CD8+ iNKT cells in the peripheral blood. (B) Representative dot plots showing NKT subsets in a randomly selected patient with oral squamous cell carcinoma (OSCC) and healthy control. (C) Comparison of iNKT cell percentages in early-stage patients (stages I & II; n = 10) and late-stage patients (stages III & IV; n = 28). Results are represented as mean ± SEM; P-value as calculated using two-tailed Mann–Whitney U-test.

Generation and maturation of DCs from PBMCs

Dendritic cells were generated from the PBMCs as previously described with minor modification [26]. Briefly, PBMCs were cultured in 6-well plate at a density of 2 × 106 cells/ml in RPMI-1640 medium supplemented with 5% (v/v) FBS, 1% sodium pyruvate, 100 U/ml penicillin G and 100 μg/ml streptomycin (all from HyClone Lab. Inc.) and incubated at 37 °C in 5% CO2 for 2 h for monocyte adherence. Non-adherent cells (typically ~70–75% of PBMCs) were removed by washing twice with the respective medium. The adherent cells were allowed to differentiate into immature DCs (iDCs) by culturing them in same media as mentioned above with 10% (v/v) heat-inactivated FBS. Recombinant human GM-CSF (20 ng/ml; BD Pharmingen) and recombinant human IL-4 (20 ng/ml; eBioscience, Inc., San Diego, CA, USA) were added to the culture on days 0, 2 and 4. On day 5–6 of culture, DCs were matured overnight by the addition of 1 μg/ml bacterial lipopolysaccharides (LPS; Sigma-Aldrich) in the culture media with or without 100 ng/ml of α-GalCer (KRN-7000; Biomol Int, LP). Phenotype of DCs was confirmed by flow cytometry using anti-CD14, anti-HLA-DR and anti-CD209 (DC-SIGN) mAbs labelled with appropriate fluorochromes (BD Pharmingen).

Proliferation assay by flow cytometry

The proliferation/expansion of iNKT cells with or without α-GalCer-pulsed DCs was performed from PBMCs of patients (n = 30) and healthy donors (n = 28) by previously described method [26]. Briefly, autologous PBMCs were cultured with or without α-GalCer-pulsed mature DCs at PBMCs/DC ratio of 10:1 in RPMI-1640 supplemented with 10% (v/v) heat-inactivated foetal bovine serum. Human recombinant IL-2 (50 U/ml; eBioscience, Inc.) was added on days 4 and 7 of the culture. On 10th day, cells were harvested, counted and analysed by flow cytometry.

Intracellular cytokine assay by flow cytometry

Intracellular staining for IFN-γ and IL-4 was performed as described earlier [4]. Briefly, freshly isolated PBMCs were cultured with α-GalCer-pulsed and unpulsed DCs for different time points, such as days 1, 3, 5 and 7. Because optimum expression of cytokines was observed on day 3, subsequent intracellular staining was performed at this time point. GolgiPlug™ (BD Biosciences) was added before 5 h of termination of the culture. After termination of the culture, cells were washed twice with cold staining buffer (PBS supplemented with 2% FBS and 0.1% NaN3) and labelled with cell surface mAbs for 30 min and washed twice with staining buffer. Cells were then fixed and permeabilized using Cytofix/Cytoperm™ Plus Kit (BD Biosciences, San Diego, CA, USA) and treated with anti-IFN-γ and anti-IL-4 mAbs, respectively. After washing twice with staining buffer, cells were acquired in BD LSR II, and these data were analysed using FACSDiva software (BD Biosciences).

Tumour growth inhibition assay

In this study, the tumour growth inhibition assay was performed on OSCC cell line, SCC-4 [American Type Culture Collection (ATCC), Manassas, VA] and KB cells [National Centre for Cell Science (NCCS), Pune, India]. Human breast cancer cell line, MCF7 (ATCC, Manassas, VA), was also used in parallel to observe whether activated iNKT cells impart cytotoxicity to other tumour cells non-specifically. Briefly, PBMCs from 18 patients with OSCC and 13 healthy donors were cultured in RPMI-1640 supplemented with 10% (v/v) heat-inactivated foetal bovine serum for 2 h. Non-adherent cells were aspirated and were cultured separately with IL-2 (50 U/ml) supplemented with or without 100 ng/ml of α-GalCer for 3 days. Both adherent and non-adherent cells were used as effector cells against the tumour targets. Tumour cells (104 cells/well) were incubated in 96-well plate for 4 h at 37 °C to form monolayer and further cultured with effector cells with different effector/target (E:T) ratios (4:1, 10:1 and 20:1) until the tumour cells in the control wells became confluent. The cells were washed twice with culture media by gentle aspiration, and then, CellTiter 96® Aqueous One Solution Reagent (Promega Corporation, Fitchburg, WI, USA) containing a novel tetrazolium compound [(3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulphophenyl)-2H tetrazolium, inner salt; MTS] and electron coupling reagent (phenazine ethosulphate; PES) was added to each well as per manufacturer's instruction. The absorbance (OD) of soluble formazan produced by reduction of MTS was recorded at 490 nm, and tumour growth inhibition was calculated as follows:

display math

Statistical analysis

Statistical analysis was performed using unpaired t-test with Welch correction or Mann–Whitney U-test (two-tailed) to compare the cell populations between two groups, Wilcoxon signed rank test to compare treated and untreated groups (after confirmation of Gaussian distributions; normality test was checked in each case) and Spearman's nonparametric test to determine correlation coefficient. Results are presented as mean ± standard error of mean (SEM), and P-value < 0.05 was considered to be significant. All statistical tests were performed using GraphPad Software, version 5 (La Jolla, CA, USA).

Results

Clinicopathological features

The clinicopathological features of patients with OSCC are shown in Table 1. Age of the patients ranged from 30 to 70 years with mean ± SEM 51.8 ± 1.64 years. Of the total 38 patients, 76.3% were men, while women comprises only 23.7%. The most common site of tumour was buccal mucosa (52.6%), tongue (23.7%) and followed by alveolus (21.1%). Lip was the rarely affected site (2.6%). Majority of patients (63.2%) presented with larger (T3+T4) tumours, while 36.8% cases presented with smaller (T1+T2) tumours. Cervical lymph node was involved in 44.7% cases. Clinical staging revealed that most of the patients (73.7%) studied were in the advanced stage of the disease.

Table 1. Clinicopathological features of patients with oral squamous cell carcinoma (OSCC)
FeaturesValues
Patients with OSCC (n = 38)Healthy controls (n = 28)
Age (yr)
Range30–7023–66
Mean ± SEM51.8 ± 1.6447.73 ± 1.33
GenderNo. (%) 
Male29 (76.3)21 (75.0)
Female9 (23.7)7 (25.0)
Site
Buccal mucosa20 (52.6) 
Tongue9 (23.7) 
Alveolus8 (21.1) 
Lip1 (2.6) 
Tumour size
T1 + T214 (36.8) 
T3 + T424 (63.2) 
Lymph node involvement
N021 (55.3) 
N+17 (44.7) 
Clinical stage
Early stage (I & II)10 (26.3) 
Late stage (III & IV)28 (73.7) 

Frequency of iNKT cells and its subsets

The percentages of both iNKT and CD3+CD161+ NKT cells were found to be significantly lower (0.24% versus 0.55%, < 0.01 and 13.96% versus 30.24%, < 0.01, respectively) in the circulation of patients with OSCC as compared to healthy donors (Fig. 1B; Table 2). Further, intragroup comparison showed significantly reduced (0.18% versus 0.41%, = 0.01; Fig. 1C) circulating iNKT cells in the late-stage patients (III and IV) as compared to patients with early-stage oral cancer (I and II). The subset analysis revealed significantly increased CD8+ iNKT subset (43.27% versus 25.37%, < 0.01) and significantly reduced DN subset (20.95% versus 34.30%, = 0.03) in patients with OSCC as compared to healthy controls. However, no significant change was observed in CD4+ iNKT subset in patients (Fig. 1B; Table 2).

Table 2. Mean percentages of circulating Natural killer T (NKT) cell subsets in patients with oral squamous cell carcinoma (OSCC) and healthy controls
NKT subsetsPatients with OSCC Mean ± SEM (%) RangeHealthy controls Mean ± SEM (%) RangeP-value
  1. ns, not significant.

iNKT cells (CD3+6B11+)0.24 ± 0.04 (0.01% to 0.87%) (n = 38)0.55 ± 0.09 (0.02% to 2.01%) (n = 28)<0.01
CD3+CD161+ NKT cells13.96 ± 1.61 (8.02% to 18.21%) (n = 7)30.24 ± 4.45 (19.12% to 57.20%) (= 8)<0.01
CD4+ iNKT subsets37.77 ± 3.56 (12.01% to 60.00%) (n = 18)41.73 ± 3.79 (18.75% to 55.79%) (n = 10)ns
CD8+ iNKT subsets43.27 ± 3.74 (14.71% to 76.23%) (n = 18)25.37 ± 3.69 (11.16% to 54.36%) (n = 10)<0.01
CD4CD8 iNKT subsets20.95 ± 3.39 (4.56% to 45.36%) (n = 18)34.30 ± 4.52 (15.29% to 62.99%) (n = 10)0.03

Next, to evaluate the overall relevance of iNKT cells in oral cancer, the statistical correlations between total circulating iNKT cells and their subsets were calculated. The computed correlation coefficient showed strong inverse relation of CD4+ iNKT subset in healthy donors (R = −0.8545, < 0.01) but weak correlation in patients, while DN iNKT cells showed positive correlation in both patients and healthy donors (R = 0.5627, = 0.01 and R = 0.3333, ns, respectively). However, CD8+ iNKT subset showed inverse correlation (R = −0.4819, = 0.04) in patients but not in the healthy donors (Fig. 2A).

Figure 2.

Inter-relation between various iNKT cell subsets. (A) Correlation between the percentages of total iNKT cells and its various subsets from peripheral blood mononuclear cells (PBMCs) of healthy donors (n = 10) and patients with oral squamous cell carcinoma (OSCC) (n = 18). R is nonparametric Spearman's correlation coefficient. (B) The study groups were divided into subgroups with high and low total iNKT percentages with reference to their mean values. Percentages of CD4+, CD8+ and DN iNKT subsets were compared between these subgroups using two-tailed unpaired t-test with Welch correction or two-tailed Mann–Whitney U-test. Results are presented as mean ± SEM; ns = not significant.

These data were further analysed by classifying the study subjects into two groups according to their respective mean frequencies of iNKT cells. Invariant natural killer T cell frequencies below their mean value are ranked as low iNKT group and vice versa (Fig. 2B). Healthy donors with high level of iNKT cells showed a lower proportion of CD4+ iNKT cell subsets and higher proportion of DN and CD8+ iNKT cell subsets than those with low level of iNKT cells (= 0.02, ns and ns, respectively), while OSCC patients with high level of iNKT cells showed a lower proportion of CD4+ and CD8+ iNKT cells and higher proportion of DN iNKT cells than those with low level of iNKT cells (ns, = 0.04 and = 0.03, respectively).

Proliferative response to α-GalCer-pulsed DCs

Proliferation/survival of iNKT cells was studied in a 10-day culture of PBMCs from both healthy controls (n = 28) and patients with OSCC (n = 30) treated with α-GalCer-pulsed and unpulsed DCs and recombinant IL-2. We found significantly higher levels of iNKT cells in both OSCC patients and healthy donors' PBMCs in response to α-GalCer-pulsed DCs as compared to that with unpulsed DCs. While in the culture with unpulsed DCs, the mean frequencies of iNKT cells were 0.21 ± 0.04% in patients with OSCC and 0.54 ± 0.09% in healthy donors, in α-GalCer-treated culture, the respective population increased to 0.30 ± 0.06% in patients and 1.04 ± 0.17% in healthy donors, respectively. These representative data are shown in Fig. 3A. The percentage increment of iNKT cells in response to α-GalCer-pulsed DCs is shown in Fig. 3B. A significantly reduced increment in iNKT cells was observed in patients as compared to the healthy donors (< 0.01). While in healthy donors, the relative increment in iNKT cells was 93.94 ± 14.36% in response to α-GalCer-pulsed DCs, in patients with OSCC, the increment was only 46.89 ± 5.43%, suggesting almost 50% reduced proliferation/survival in patients with OSCC.

Figure 3.

Proliferation capacity of iNKT cells in response to α-GalCer-pulsed dendritic cells (DCs). (A) Representative dot plots of peripheral blood mononuclear cells from both healthy donors and patients with oral squamous cell carcinoma (OSCC) showing proliferation of iNKT cells in response to α-GalCer-pulsed/unpulsed DCs. (B) Percentage increment in iNKT cell in total PBMCs of healthy controls (n = 28) and patients with OSCC (n = 30) in response to α-GalCer-pulsed DCs taking unpulsed DCs as reference. Results are presented as mean ± SEM; P-value was calculated using two-tailed unpaired t-test with Welch correction.

IFN-γ- and IL-4-producing iNKT cells

Next, we determined the effect of α-GalCer-pulsed DCs on cytokine-producing ability and particularly the Th1/2-like biasing in the iNKT cells by intracellular staining after a 3-day coculture of PBMCs and α-GalCer-pulsed DCs from both healthy controls (n = 19) and patients with OSCC (n = 27). The cytokine analysis was performed on CD3+6B11+-gated INKT cells (as shown in Fig 1A). Fig. 4 shows the frequency of IFN-γ (11.62 ± 1.45%)- and IL-4 (21.53 ± 2.49%)-producing iNKT cells in patients and healthy donors (IFN-γ, 14.53 ± 2.35%; IL-4, 29.68 ± 5.03%). The frequency of IFN-γ-producing iNKT cells remained lower as compared to the IL-4-producing iNKT cells in both patients with OSCC and healthy donors, suggesting a Th2 bias in circulating iNKT cells. Although α-GalCer treatment led to significant expansion of both IFN-γ- and IL-4-producing iNKT cells (Figs. 4A, B), the expansion of IFN-γ-producing iNKT cells was relatively higher as compared to that of the IL-4-producing cells, both in patients and in healthy donors. The ratio of IFN-γ- and IL-4-expressing cells tended to be lower in patients with OSCC as compared to healthy donors both at basal level and after activation, and the difference was statistically not significant (Fig. 4C).

Figure 4.

Th-like cytokine profiling of CD3+6B11+ iNKT cells in healthy controls and patients with oral squamous cell carcinoma (OSCC) and its modulation by α-GalCer-pulsed dendritic cells (DCs). (A, B) Frequency of 6B11+IFN-γ+ iNKT and 6B11+IL-4+ iNKT cells in healthy controls (n = 19) and patients with OSCC (n = 27) peripheral blood mononuclear cells with/without α-GalCer-pulsed DCs activation. (C) IFN-γ+/IL-4+ iNKT cell ratio showing more recovery of IFN-γ-producing iNKT cells after activation with α-GalCer-pulsed DCs. Results are represented mean ± SEM; ns, not significant.

Tumour growth inhibition

As iNKT cells have been reportedly contribute to antitumour immunity by the cytotoxic/growth inhibition of tumour cells, we examined whether antitumour activities of these cells are compromised in patients with OSCC. The results are represented in Fig. 5. All experiments were carried out independently in triplicates with five wells for control in each set. While unpulsed PBMCs from healthy controls showed tumour growth inhibition of up to 56.03%, 53.76% and 58.08%, unpulsed PBMCs from patients with OSCC showed 42.78%, 36.18% and 39.19% inhibition (KB, SCC-4 and MCF7, respectively; E:T = 20:1). Addition of α-GalCer to the system led to enhanced antitumour activities of PBMCs from both healthy controls (75.2%, 72.28% and 70.37%) and patients with OSCC (50.29%, 52.54% and 47.72%) against KB, SCC-4 and MCF7, respectively (E:T = 20:1) (Figs. 5A–C). As α-GalCer specifically activates iNKT cells, observed increment in tumour growth inhibition by α-GalCer-treated PBMCs may be solely attributed to iNKT cells. Noticeably, non-adherent and adherent fractions alone did not exhibit antitumour response, possible reason may be the absence of potent APCs and iNKT cells, respectively.

Figure 5.

Tumour growth inhibition by α-GalCer-activated PBMCs. Along with whole PBMCs, plastic-adherent and non-adherent fractions of PBMCs from both healthy controls (n = 13) and patients with oral squamous cell carcinoma (OSCC) (n = 18) were treated with IL-2 and α-GalCer independently and were used as effector cells against tumour targets (A) KB, (B) SCC-4 and (C) MCF7 cell lines in different E:T ratios. The results are shown as mean ± SEM.

Discussion

The present study for the first time has demonstrated the altered population and functional dynamics of circulating iNKT cells in patients with OSCC. We found that the frequency of peripheral blood iNKT cell was significantly decreased in patients as compared to the healthy controls. Moreover, late-stage tumours were associated with further reduction in iNKT cells as compared to those with the early-stage tumours. We also found reduced frequency of CD3+CD161+ NKT cell in patients as compared to healthy donors. Such impairment of CD3+6B11+ iNKT and CD3+CD161+ NKT cells together may impart impaired tumour surveillance in oral cancer patients with OSCC. However, there are previous reports that suggested heterogeneity in iNKT cell frequency in patients with cancer as well as normal donors [23, 27].

Among the iNKT subsets, we found a significant decrease in DN subset with concomitant increase in CD8+ subset in patients with OSCC. CD4+ iNKT cells are producers of more balanced mixture of both Th1 and Th2 cytokines, while DN and CD8+ iNKT cells, in contrast, are strict producers of Th1 cytokine, IFN-γ [27, 28]. Thus, it appears that DN and CD8+ iNKT cell subsets being Th1-biased may play a protective role in antitumour immunity, and loss of DN subset in patients with OSCC may provide immunosuppressive environment for the development and progression of tumour. Although, the reason for relatively higher frequency of CD8+ iNKT cell subset in patients with OSCC is difficult to explain, these cells may be transiently activated due to tumour load. Being predominantly pro-inflammatory in nature, it may further restrict tumour growth. However, one study reported that CD8+ iNKT subset acts at a later phase of immune response and suppresses the expansion of antigen-specific T cells by killing the CD1d-expressing APCs, possibly to control overspill of an immune response and thus behaves like regulatory T cells [29]. Therefore, the role of CD8+ iNKT subset seems to be elusive and requires further elucidations. It seems reasonable that in case with cancer not only the total number of iNKT cells but mutual balance between its different subsets may be important for antitumour immunity.

Several reasons can be assigned for the reduction in peripheral blood iNKT cells in patients with cancer, such as cell death, impaired proliferation and migration to or accumulation at the tumour site [21, 30]. Moreover, tumour microenvironment may reduce CD1d expression in myeloid DCs, which in turn results to the defective iNKT cell population (because this lineage of DCs is responsible for activating iNKT cells via CD1d) [31, 32]. A reduced frequency of myeloid DCs and HLA-DR expression was observed in patients with squamous cell carcinoma of head and neck resulting in changes in the different T cell subsets [33]. We have earlier shown attenuation of T cell subsets in patients with OSCC [3, 4]. As we observed significant reduction in these cells in late-stage tumours, a low number of circulating iNKT cells may represent a risk factor for the development of malignancies [34] rather than due to the presence of the tumour as suggested earlier [35] and may be useful as a prognostic marker for patients with OSCC.

Varying degree of impairment of immune response has been reported previously in patients with different cancers [14, 36] including those with OSCC [3, 4]. Unlike conventional CD4+ and CD8+ T cells, iNKT cells are activated by endogenous or exogenous glycolipid antigens and possess remarkable capacity to produce various immunoregulatory cytokines such as IFN-γ, IL-4, IL-10 and IL-13 as well as a variety of cell-death-inducing effector molecules [37]. Therefore, we investigated proliferative/survival response of iNKT cells from patients with OSCC in vitro to α-GalCer-pulsed DCs. iNKT cells of patients showed reduced proliferation and survival as compared to the healthy donors as reported earlier in patients with various malignancies [36]. It has been suggested that CD3+ T cells play a major role in reduced proliferation and survival of iNKT cells [36] by producing some immunosuppressive factors such as prostaglandin E2 (PGE-2) and TGF-β, which were shown to be overexpressed in patients with cancer. These factors can modify the functions as well as differentiation of immune cells [38] including iNKT cells, resulting in their poor proliferative or survival response.

We further analysed Th-like cytokine profile by intracellular staining of these cells with anti-IFN-γ and anti-IL-4 mAbs. The frequencies of IL-4-expressing cells were higher, whereas IFN-γ expression was significantly reduced both in patients and in healthy individuals. Treatment with α-GalCer-pulsed DCs although recovered IFN-γ-expressing cells more, iNKT cells per se showed Th2 bias only. This may be due to reduced frequency of circulating iNKT cells in patients, particularly due to significant loss of DN iNKT cells. We have earlier reported a Th2 bias in patients with OSCC [4, 39]. It appears therefore that polarization of Th2 cytokine expression is not only due to CD4+ iNKT cell subset but also because of loss of IFN-γ and gain of IL-4-expressing iNKT cells in patients with OSCC. Loss of IFN-γ-producing iNKT cells has also been reported in patients with prostate cancer [14]. Presumably, activation of iNKT cells with agonist ligand may shift cytokine balance from immunosuppressive Th2 to protective Th1 type.

Invariant natural killer T cells express a wide variety of cell-death-inducing effector molecules [37] and have been shown to kill tumour cells in vitro [13, 36]. We therefore evaluated whether iNKT cells from patients with OSCC show cytotoxic activity against human OSCC and breast cancer (MCF7) cell lines and whether this ability is altered as compared to healthy controls. Indeed, we found that cytotoxic potential of iNKT cells from patients with OSCC is compromised as compared to that of the healthy controls. Our results suggested that compromised antitumour activities of OSCC patients' PBMCs may be attributed to quantitative and qualitative abnormalities of iNKT cells such as reduced mean frequency, proliferation and survival and production of IFN-γ. However, PBMCs from patients with OSCC showed significant cytotoxicity against SCC-4, KB and MCF7 cell lines when pulsed with α-GalCer as compared to the unpulsed PBMCs. In our study, unpulsed PBMCs from patients with OSCC also showed more than 50% growth inhibition of tumour cells, suggesting that presence of T, NK, and iNKT cells confers about 50% cytotoxicity to tumour targets, while treatment with α-GalCer further enhances it possibly through the activation of iNKT cells. In concordance with our results, a previous study has shown cytotoxicity of Vα24+ iNKT cells from patients with seven different cancers to histiocytic lymphoma (U937) cell line [36]. Although we have not used purified iNKT cells, but since we demonstrated that α-GalCer has facilitated the proliferation and survival of iNKT cells in the total PBMCs, it indicates that this enhanced cytotoxicity may be due to selectively expanded and survived iNKT population. Interestingly, the PBMCs from patients with OSCC showed similar degree of tumour growth inhibition/cytotoxicity against KB and MCF7 cells, suggesting that iNKT cells, similar to NK cells, may impart cytotoxic effect across different tumour cells irrespective of their origin.

Results of the present study suggest abnormalities in iNKT cell population and their proliferative and survival response to α-GalCer-pulsed DCs and cytokine expression in advanced-stage tumours, which appears to be related to growth and malignancies of tumours in patients with OSCC. However, treatment with α-GalCer-pulsed DCs tended to restore IFN-γ-expressing iNKT cells more than IL-4-expressing cells and increase their antitumour activity. It may be presumed that the restoration of iNKT cell population and function via their agonist and appropriate APCs may enhance the antitumour potential of these cells for cell-based vaccination strategies in patients with OSCC. Reduced frequency of iNKT cells and its DN subset may be used as prognostic factors for patient with OSCC.

Acknowledgment

This work was supported by a grant from the Department of Biotechnology (DBT), Ministry of Science and Technology, Government of India to SND. AKS is a recipient of research fellowship from DBT.

Conflict of interest

Authors or funding agency have no conflict of interest.

Ancillary