Characterization of human invariant natural killer T subsets in health and disease using a novel invariant natural killer T cell-clonotypic monoclonal antibody, 6B11

Introduction

CD1d-restricted T-cell receptor (TCR)-invariant natural killer T (iNKT) cells are a subset of CD3⁺ lymphocytes highly conserved through mammalian evolution and thought to subserve regulatory/effector functions in a wide variety of immune responses.^1–4 Indeed, murine iNKT cells have been shown to play crucial roles in antitumour, autoimmune, and antimicrobial responses.^5–7 These findings have encouraged further characterization of human iNKT cells and studies of their participation in diseases involving dysregulated immune functions. While murine iNKT cells have been subject to extensive investigation, less is known regarding human iNKT cells, particularly with respect to the influence of age, gender, racial background and environmental factors. Moreover, the contribution of these cells to human disease involving immune dysregulation remains controversial.^8–11

The evaluation of human iNKT cells is particularly complicated because of their low frequency (i.e. ranging from 0·01% to 1%) amongst peripheral blood T cells and a lack of reagents specific for iNKT cells.² As a result, human iNKT cells have largely been characterized using clones obtained following their expansion with α-galactosyl ceramide (α-GalCer) and long-term cell culture with autologous-irradiated mononuclear cells and cytokines (e.g. interleukin (IL)-2, IL-7);^12,13 however, this process may change some of the surface markers and functional patterns originally expressed by unmanipulated, resting human iNKT cells. More recently, TCR-based strategies utilizing combinations of semiselective antibodies and CD1d tetramers have been developed.^14,15 Despite these advances, difficult technical challenges remain when using these reagents in terms of defining specificity and sensitivity.¹⁵

In the TCR α chain of human iNKT cells, the Vα24 segment is joined with Jα18 in a germ-line configuration, resulting in an invariant CDR3 loop encoded by the mature TCR α chain.^16,17 This α chain pairs with a restricted range of randomly rearranged Vβ chains, with Vβ11 being the most prominent in humans.¹⁸ Therefore, iNKT cells have classically been identified using monoclonal antibodies (mAbs) against the Vα24 and Vβ11 chains^2,18. However, non-invariant and non-CD1d-restricted Vα24⁺ T cells can also pair with Vβ11, and contribute to the Vα24+/Vβ11⁺ subpopulation. This could lead to an overestimation of iNKT cell number, especially in individuals with decreased number of iNKT cells.^2,19

In previous studies, iNKT cells have largely been evaluated using the coexpression of Vα24 and natural killer (NK) cell markers as CD161 or CD56.^20–22 However, it is now clear that only a portion of iNKT cells express NK cell markers and this expression is associated with a late stage of cell differentiation.²³ Furthermore, CD161 and other NK markers are expressed on activated CD8⁺ T lymphocytes. As 15–25% of peripheral blood T cells express CD161, this marker may include non-CD1d-restricted non-invariant Vα24⁺ T cells.²⁴

The recent development of CD1d tetramers loaded with α-GalCer has significantly enhanced the specificity of iNKT cell evaluation.^14,25,26 However, CD1d restriction is not limited to the iNKT cell population, because CD1d restriction has been found on T cells expressing CD161 with diverse repertoires.^27,28 Using this approach, the identification is limited by the antigen loaded into the CD1d molecule, with background staining¹⁵ cross species variability and displacement of CD1d tetramers by combination with other monoclonal antibodies.^28,29

The TCR of iNKT cells represents a unique case of a highly conserved antigen receptor rearrangement and a novel alternative for specific detection of human iNKT cells. Toward this end, we recently generated a mAb against the conserved CDR3 region of the canonical Vα24Jα18 TCR rearrangement (clone 6B11).^19,30 The 6B11 mAb recognizes all T cells expressing the invariant TCR α chain, and can be used in combination with anti-Vα24, anti-Vβ11 or anti-CD3 with high specificity and sensitivity for iNKT cell detection.

In this study, we analysed the suitability of the mAb 6B11 for identification and quantification of iNKT cells in a large cohort of healthy individuals, an effort that involved both their phenotypic characterization as well the detection of activation-induced intracellular cytokines. In addition, as a model to determine the utility of the mAb 6B11 to study the frequency and phenotype of circulating iNKT cells in different human immune-mediated diseases, a cohort of type 1 diabetes patients, their relatives and a cohort of high-risk of diabetes individuals were evaluated.

Materials and methods

Results

The term ‘natural killer T cells (NKT)’ has widely been used to define a broad range of T-cell subsets. Currently, it is thought that iNKT cells define the major population of CD1d-restricted T cells that may or may not express NK markers.^2,13,32 There are at least two populations of CD1d-restricted NKT cells, which can be distinguished based on their TCR repertoire. The best-studied subpopulations are iNKT cells, also known as type-I NKT cells.³² Recently, we developed a mAb against the CDR3 loop of the invariant a chain of iNKT cell TCR, clone 6B11, which was utilized in this study to specifically identify and characterize human iNKT cells.

Discussion

A variety of methods have been used to identify iNKT cells. In the current study, we found no significant differences in the frequency of total iNKT cells and their subsets in peripheral blood comparing 6B11, human CD1d tetramers and anti-Vα24/anti-Vβ11 mAbs. It is possible that noninvariant or non-CD1d-restricted Vα24⁺ cells can also pair with Vβ11 and increase the detectable population of Vα24⁺/Vβ11⁺ cells. Thus, the population defined by the conventional Vα24/Vβ11 staining may include a proportion of non-iNKT cells. Staining with 6B11 (alone or in combination) is specific for iNKT cells as this mAb recognizes the CDR3 loop of the invariant α chain from CD1d-restricted iNKT cells.

CD1d tetramer staining alone invariably includes 0·01–0·05% non-Vα24 cells;⁴¹ even coexpression of Vα24 and Vβ11 does not completely overlap with CD1d-tetramer reactivity in humans, and could be complicated by competition between CD1d tetramers and anti-TCR mAbs (e.g. anti-Vα24).^28,29 Concurrent preincubation with anti-Vα24 has been used to block tetramer binding to iNKT cells, and thus identify non-invariant CD1d-restricted populations.^14,28,29 In other reports, the frequency of iNKT cells detected in PBMC using 6B11 was approximately 95% of the tetramer reactive CD1-restricted T cells.¹⁹ Additional studies on NKT cells in humans have focused on the Vα24⁺ population,^11,44 the CD3⁺ CD56⁺ cells⁴⁵ or the CD3⁺ CD161⁺ cells⁴⁶ from peripheral blood or pathology tissues. As the accurate detection of iNKT cells is the first step in their analysis, we suggest that the best phenotypic definition of iNKT cells should be based on the invariant TCR α chain rearrangement (Vα24Jα18), specifically detected by the 6B11 mAb. We propose that 6B11, alone or in combination with anti-Vα24, anti-Vβ11 or anti-CD3 is an excellent tool for analysis of human iNKT cells.

Our results confirm previous reports regarding the heterogeneity of iNKT cell frequency in peripheral blood from healthy humans.^11,22,41 However, in comparison with the study by Sandberg et al. we observed a greater range of iNKT cell frequency determined as the percentage of lymphocytes (0·01% to 0·92%), with no significant difference in that frequency between females and males.²² There was a modest trend towards increased frequency of iNKT cells in female subjects. The differences between these results and those of Sandberg et al. are likely the result of the use of anti-Vα24 and anti-CD161 in their study; when using these mAbs, there was a large population of non-iNKT cells that expressed Vα24 and CD161, while CD161 can be detected on up to 20% of circulating lymphocytes.^2,24

The factors that regulate the size of the circulating iNKT cell pool in vivo are poorly understood. In previous studies, the frequency of peripheral blood iNKT cells was generally stable over time.^11,16 In the cohort studied here, longitudinal analysis of iNKT cell frequency in peripheral blood from 12 healthy individuals was stable over at least a 12-month period. However, in one other report, it was found that an environmental factor (an unidentified respiratory infection) modified the frequency of circulating iNKT cells in a normal donor.²² Additionally, human iNKT cells are thought to fluctuate in the peripheral blood of patients with a variety of organ-specific and systemic autoimmune conditions and cancer.^44,47–50

Human iNKT cells have long been known to segregate into CD4⁺ and CD4^– subsets, with distinct phenotypic and functional characteristics;¹⁴ however, there are few descriptions evaluating subsets of iNKT cells that express CD8.^14,51,52 Using different combinations of mAbs, four well-defined iNKT-cell subsets were noted based on TCR coreceptor expression. In contrast to previous studies²² where the CD4⁺ subset was predominant (59·5%), a similar frequency of CD4⁺ and DN iNKT cells was noted. In addition, as we defined in a preliminary report³⁴ there was a direct inverse relationship between the percentage of CD4⁺ and total iNKT cell frequency. The different mAbs used for iNKT cell detection may partially explain these differences.

An expansion of the CD4^– subset of peripheral blood iNKT cells was reported for healthy individuals when the frequency of total iNKT cells was above 0·1%.²² Our results show that when iNKT cells are higher than 0·16%, the DN cells were predominant while the CD4⁺ subset predominated in individuals with total iNKT cells below this percentage. Consistent with other studies, the frequency of iNKT cell subsets appears to remain stable in healthy individuals. This is in contrast to cancer and the limited, but controversial results in autoimmunity.^{10,47–49,53–55} The expansion of CD4^– iNKT cells seems to be extra-thymic and driven by enhanced proliferation of CD4^– relative to CD4⁺ iNKT cells.^56,57 We recently found that Toll-like receptor-9-dependent signalling resulted in the preferential expansion and survival of CD4^– iNKT cells.⁵⁸ Given the association of type 1 diabetes development and recurrent enteroviral infections, we speculate that the expansion of DN iNKT cells seen in this study in patients at risk for type 1 diabetes might be driven by such a mechanism. This is of potential interest because DN iNKT cells are thought to be proinflammatory and effectors of tumour and viral surveillance^1,14,59–61 and the CD4⁺ subset has been strongly associated with immunoregulatory function of these cells. Moreover, the selective loss of CD4⁺ iNKT cells in non-obese deiabetic mice resulted in accelerated diabetes.⁶⁹ It should be noted that the findings on iNKT cell frequency in patients at risk have been controversial. Perhaps this is related to the reagents and methodologies used.^{10,37,42,55,57} However, further longitudinal population and effector studies will be needed to evaluate the relationship of this finding to type 1 diabetes.

Several other phenotypic and functional markers have been used to characterize iNKT cells. The studies are particularly extensive for mice while much less is known about human iNKT cells. iNKT cell subsets are distinct from each other in their expression of surface molecules and receptors, suggesting that they might be targeted to different tissues and meant to exert different immune functions.^{14,19,21,22,62} The expression of 15 surface markers on resting peripheral blood iNKT cells were evaluated to determine the physiological characteristics of these cells, avoiding methods that may alter their expression patterns (e.g. expansion with cytokines and CD1d-restricted molecules such as α-GalCer). As a whole population, circulating iNKT cells have a variable expression of NK markers, with a low percentage bearing CD16 and CD56, and CD161 being the most frequently expressed molecule. In contrast, the majority of iNKT cells have a memory-effector phenotype because a high percentage of them expressed CD45RO and CD28, and a low percentage expressed CD62L. The expression of activation markers (e.g. CD25, CD38, CD69, CD154 and HLA-DR) was low, suggesting that peripheral blood iNKT cells are physiologically programmed to migrate to inflamed peripheral tissues where they become activated. A similar phenotypic pattern was also described for peripheral blood Vα24⁺/Vβ11⁺ cells from healthy controls²² and patients chronically infected with hepatitis C virus;^11,44 however, hepatic iNKT cells from those patients were highly activated (high expression of CD69), suggesting that the inflammatory environment in this organ provided the signals required to activate iNKT cells.

Examination of expression of these markers on iNKT subsets revealed group-specific differences. Interestingly, CD25 was mainly expressed on CD4⁺ iNKT cells and the 6B11⁺ CD4+ cells represent 6 ± 4% of the CD4⁺ CD25⁺ T cells in PBMC. Previous studies reported a similar tendency²² but in some studies the CD4⁺ CD25⁺ iNKT cells were found at higher percentages:⁴¹ 10% to 80% versus 7·2% ± 4·5 in this report. These differences might be caused by the exposure to activation signals dependent on the incubation with PMA/ionomycin or αGalCer-loaded tetramers, or in differences is study populations. An important part of the CD4⁺ CD25⁺ T cell pool are Treg cells⁶³ and considering the proposed immunoregulatory role of iNKT cells, it will be important to clarify the relationship of CD4⁺ CD25⁺ iNKT cells to Treg cells.

CD4⁺ and CD8⁺ iNKT expressed similar amounts of CD95 (8%), but expression was negligible on DN iNKT cells, suggesting different roles of these subsets in the control of immune reactions. Both DN and CD8⁺ iNKT cells expressed significantly more CD161 than CD4⁺ iNKT cells, consistent with previous reports for the DN subset.^13,14,64 No differences in the expression of CD56 and CD62L by resting peripheral blood iNKT cell subsets were noted. Other investigators observed that CD56 was expressed by CD4^– cells, while CD4⁺ iNKT cells were CD62L⁺ and CD4^– iNKT cells were CD62L negative.⁶⁵ These differences are likely to be caused, in part, by study subjects and reagents used.

iNKT cells were further characterized functionally using a combination of intracellular staining for cytokines with surface staining for the canonical TCR (6B11). The preferential production of IL-4 by CD4⁺ and similar production of proinflammatory cytokines by CD4⁺, DN, and now extended to CD8⁺ iNKT cells, was similar to previously reported studies.^14,41 In in vitro studies of cell lines and clones, both CD4⁺ and DN iNKT cells displayed a Th0-like profile and were capable of simultaneously producing both Th1 and Th2 cytokines.^13,43,66 The reasons for the discrepancies between ex vivo and in vitro studies remain unclear but are likely caused by changes induced by in vitro expansion of iNKT cell clones or cell lines.

Taken together, these results support the use of 6B11 for studies of iNKT cells, identifying at least four peripheral blood iNKT cell subsets based on classic T-cell coreceptor expression. Moreover, CD8⁺ iNKT cells are functionally very similar to DN iNKT cells.⁵² Considering that the frequency of these subsets is heterogeneous with an inverse relation between the CD4⁺ and DN subsets, that CD4⁺ iNKT cells display a Th0 phenotype while CD8⁺ and DN cells follow a Th1 pattern, and that the frequency of these last subsets is correlated with an inflammatory history, it seems reasonable to propose that the predominance of one or more of the iNKT-cell subsets may influence the susceptibility or resistant to immune-mediated diseases.^{42,48–50,67,68}