Background
iNKT cells are a minor subpopulation of the human peripheral blood T cell repertoire. They are members of the innate lymphocyte subset, which includes natural killer cells (NK) and γ/δT cells. Both mouse and human iNKT cells, unlike non-variant NKT, express an invariant TCR and recognise lipid antigens in association with the HLA class Ib molecule CD1d. Human iNKT cells express a combination of Vα24Vβ11 [
1] and in mouse Vα14Vβ2,7 or 8 [
2]. The precise physiological role of iNKT cells is uncertain but they are involved in the activation of dendritic cells (DC) in response to microbial lipid antigens and thereby provide a link between innate and adaptive immune responses to infection [
3‐
5]. They are also thought to recognise self-lipid antigens presented on CD1d and hence to play a role in tolerance induction and the suppression of autoimmunity [
6‐
8]. Conversely, iNKT cells are involved in cancer immunosurveillance, recognising altered self lipid antigens expressed by malignant cells and eliminating them at an early stage of transformation [
9].
iNKT cells are universally responsive to αGalCer, a glycosphingolipid antigen derived from the marine sponge
Agelas mauritanius, or to its synthetic analogue KRN7000 [
10]. αGalCer binds to the hydrophobic groove of CD1d and activates all iNKT cells by means of TCR recognition [
9]. Responding cells proliferate and characteristically release both Th1 and Th2 cytokines.
αGalCer has been reported to promote DC maturation and subsequent antigen-specific T cell responses via selective engagement of iNKT cells. DC maturation is then enhanced via CD40L expression by activated iNKT cells and via their release of IFNγ resulting in IL-12 production and HLA upregulation by DC in a positive feed back manner with consequent promotion of antigen presentation [
9].
This adjuvant effect is thought to underlie the observations made in a number of mouse-models that αGalCer promotes specific anti-tumour immunity and clearance of established tumours [
11‐
13]. In turn this has led to the proposed use of iNKT cell ligands like αGalCer as adjuvants for human tumour immunotherapy [
12,
13].
However, iNKT cells are reduced in both number and activity in patients with certain malignancies [
14‐
19]. Moreover,
in-vivo, vaccination with soluble αGalCer leads to only transient activation of iNKT cells followed by long-term unresponsiveness [
20]. Thus the optimal use of αGalCer-based immunotherapy for cancer patients is envisaged to involve infusing αGalCer
in-vitro-expanded autologous iNKT cells followed by αGalCer-pulsed, tumour antigen-loaded DCs.
Quantitative defects in iNKT cells are predictive of progression in certain autoimmune diseases. For example, iNKT cells are reduced in diabetes-prone NOD mice and increasing iNKT cell numbers by adoptive transfer [
21] or via the introduction of a
Vα14-Jβ18 transgene suppresses subsequent disease progression [
22].
All iNKT cells expand when cultured
in-vitro with αGalCer and IL-2 in a CD1d-restricted manner [
23] and
in-vivo following administration of αGalCer-pulsed DCs [
24]. However, the expansion ratio of human iNKT cells is known to be highly variable between individuals [
25‐
27].
In this study we have evaluated the iNKT expansion profiles of a panel of twenty five healthy human donors to assess the degree of individual variability on stimulation with various CD1d ligands. We also sought to define some of the factors that may influence such donor variation.
Methods
Patient samples
All donor blood samples were obtained from Queen Elizabeth Hospital, Edgbaston, Birmingham UK, following patient consent and local Ethics Approval.
Antibodies and flow cytometric analysis
Human iNKT cells were identified using a novel monoclonal antibody (clone 6B11) specific for the CDR3 loop of the human Vα24Jα18 TCR alpha chain which has been previously reported to specifically identify iNKT cells [
28,
29] referred to subsequently as 6B11. Other antibodies comprise CD3-APC, CD4-PerCP and CD8α-FITC (BD Biosciences). For surface staining, cells were washed with staining buffer (PBS + 2% FCS), and incubated for 30 mins, on ice in the dark with the relevant antibody combinations. Cells were then washed twice with staining buffer and analysed by flow cytometry (BD Coulter). Data was collected using a FACS Calibur (BD Biosciences) and analysed by "Flowjo" flow cytometry software (Tree Star, Inc).
Glycolipids
αGalCer (C26:0) and PI-3 (C20:2) were synthesised within GSB's laboratory by PAI, as previously described [
30]. The glycolipids were dissolved in DMSO at a concentration of 100 μg/ml, and diluted into culture medium to the required final concentration.
iNKT cell proliferation assays to CD1d ligands
Peripheral blood mononuclear cells (PBMC) were isolated from buffy coats by Ficoll density gradient centrifugation. Cells were cultured in RPMI 1640 + L-glutamine (2 mM) + 10% human serum (Lonza) + recombinant human IL-2 (IL-2) 100 u/ml (Peprotec) in the presence of αGalCer (C26:0) at a final concentration of either: 50, 100 or 500 ng/ml or with PI-3 (C20:2) at 100 ng/ml, for a 14-day culture period. Cell cultures were fed by replacement of half volume media twice weekly with fresh RPMI supplemented with 1% penicillin and streptomycin, 10% human serum and IL-2 (200 u/ml). No additional glycolipid was added during feeding of cells. The percentage of iNKT cells within donor populations were assessed by flow cytometry performed on days 0 (to allow comparison of peripheral iNKT frequency), 7, 10 and 14 days post-culture. Co-expression of 6B11 (an antibody specific to the Vα24Jα18 iNKT TCR) and CD3 were used to identify the iNKT population in each culture [
31]. For confirmation of iNKT proliferation, PBMC were stained with CFSE dissolved in DMSO for 10 mins prior to incubation with αGalCer/IL-2 and proliferation of iNKT was analysed by dilution of the CFSE signal/6B11
+ staining on day 14 post culture. For analysis of subpopulations of iNKT, cells were surface stained with 6B11-PE, CD3-APC, CD4-PerCP and CD8α-FITC.
Analysis of effects of age, gender, and peripheral iNKT cell frequency
To assess whether donor gender and/or age affected iNKT expansion, PBMC were isolated from donor blood samples and iNKT cells allowed to expand in culture with αGalCer/IL-2 over 14 days for flow cytometric assessment, as previously described. Peripheral iNKT levels were assessed by analysing the number of 6B11+ cells within the CD3+ population of the PBMC of each donor examined in the study on day 0.
Effects of adding exogenous IL-4 to 'poor' and 'strong' responder' iNKT cultures
PBMC were cultured with αGalCer and IL-2 (100 u/ml), from the blood of identified 'poor' and 'strong' responder donors as previously described. Exogenous recombinant human IL-4 (IL-4) (10 ng/ml, Peprotec) was added to the PBMC cultures at day 0 and the percentage of iNKT cells present at day 14 post-culture assessed by flow cytometry. IL-4 was added throughout the 14 day culture period being added to fresh media (20 ng/ml) prior to feeding of cells, as previously described.
Statistical Analysis
Data was tested for normal distribution (Shapiro Wilk). For data with normal distribution Students T tests were carried out (Excel spreadsheets). For data with non-normal distribution a non-parametric Man Whitney U test (SPSS) was performed, which does not assume normal distribution. P-values < 0.05 were considered as significant (95% confidence). All error bars represent the standard error of the mean.
Conclusion
We detected marked but reproducible donor-dependent differences in the expansion of blood iNKT cells in response to in-vitro culture with αGalCer and IL-2. Individual donors could be classified either as 'strong' or 'poor' responders in terms of their relative iNKT expansion efficiencies. The differences in the two response phenotypes could not be explained by peripheral levels of iNKT, the age or sex of the donor, or to the type or concentration of the CD1d ligand used. However, reduced iNKT proliferative responses were observed in older donors but not to a significant degree. Expansion in-vitro of iNKT cells derived from 'poor' responder donors was augmented by addition of exogenous IL-4 to the cell cultures. The results suggest that individual donor iNKT cell response phenotypes may be associated with inherent early differential production of IL-4 and/or other Th2 cytokines within the cell cultures.
The inherent differences observed in iNKT cell responses between individuals, could potentially influence their susceptibility to diseases in which iNKT cells are implicated, including autoimmune diseases, malignancies and infections. The results also suggest that the efficacy of immunotherapy with αGalCer/DC-based vaccines may depend upon the recipients' inherent iNKT cell response phenotype. Reconstitution of cancer patients exhibiting a 'poor' iNKT response phenotype with cytokine-induced, in-vitro expanded, autologous iNKT cells may then help to boost their subsequent response to immunotherapy with αGalCer/DC-based vaccines.
Authors' contributions
JEC carried out all the experiments. αGalCer and PI-3 were synthesised by PAI in GSB laboratory. Experimental design and manuscript preparation was performed by: JEC, MM, SMC, CRW, DAL and DHA. All authors have read and approved the manuscript.