You are here

Article Text

Article menu
Basic and translational research
Extended report
Profound invariant natural killer T-cell deficiency in inflammatory arthritis
  1. Susan J Tudhope1,
  2. Alexei von Delwig1,
  3. Jane Falconer1,
  4. Arthur Pratt1,
  5. Tom Woolridge2,
  6. Gillian Wilson2,
  7. John D Isaacs1,
  8. Wan-Fai Ng1
  1. 1Institute of Cellular Medicine, Musculoskeletal Research Group, Newcastle University, Newcastle upon Tyne, UK
  2. 2Musculoskeletal Directorate, Freeman Hospital, Newcastle upon Tyne, UK
  1. Correspondence to Dr Wan-Fai Ng, Musculoskeletal Research Group, Institute of Cellular Medicine, Newcastle University, Newcastle upon Tyne, NE2 4HH, UK; Wan-Fai.Ng{at}ncl.ac.uk

Abstract

Objectives Data from rodent models indicate that invariant natural killer T (iNKT) cells are key regulators of many immune responses including autoimmune arthritis, but their role in human diseases is unclear. The aims of this study are to determine whether iNKT cell frequency and function are altered in patients with rheumatoid arthritis (RA), and the clinical significance of such iNKT cell abnormalities.

Methods Peripheral blood iNKT cell frequency and proliferative response to an iNKT cell-specific agonist, α-galactosylceramide were measured in 46 RA patients (including 23 untreated, newly diagnosed patients), 22 healthy controls and 27 patients presenting with recent-onset joint pain. The relationship between iNKT cell frequency and clinical characteristics and the effects of immunosuppressive treatment was examined.

Results Compared with healthy controls, RA patients had a decreased frequency of peripheral blood iNKT cells (median 0.001% vs 0.021%, p<0.001) and the proliferative response of this subset to α-galactosylceramide was also diminished in the patient group (median fold-expansion 31 vs 121, p=0.037). These abnormalities preceded the initiation of disease-modifying or immunosuppressive therapy, whose effect was to increase the circulating iNKT cell frequency (p=0.037). Furthermore, iNKT cell frequency correlated inversely with the systemic inflammatory marker, C-reactive protein (p=0.008). Finally, in patients presenting with recent-onset joint symptoms, normal peripheral blood iNKT cell frequency predicted a non-inflammatory cause of joint pain.

Conclusion iNKT cell deficiency is present in patients with RA and other inflammatory arthropathy. Normal iNKT cell frequency predicts non-inflammatory causes of joint pain.

https://doi.org/10.1136/ard.2009.125849

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Data from rodent models implicate invariant natural killer T (iNKT) cells in the regulation of many immune responses including autoimmunity.1,–,5 Unlike conventional T cells, iNKT cells recognise glycolipid antigens presented by the non-polymorphic, MHC class I-like molecule, CD1d.6 iNKT cells use a restricted repertoire of T-cell receptor (TCR) α and β-chains (Vα24–Jα281 preferentially paired with Vβ11 in humans) and express surface receptors such as NK1.1 (CD161) that are normally found on natural killer cells. Upon activation with α-galactosylceramide, a synthetic glycolipid that activates iNKT cells in vitro and in vivo,7 8 iNKT cells upregulate the expression of CD40L and release both T-helper (Th)1 and Th2 cytokines.1,–,5 Several synthetic glycolipid analogues of α-galactosylceramide have been developed that can fine-tune the profile of cytokine production by iNKT cells, and have been shown to ameliorate some murine models of autoimmune arthritis.9,–,11

    To date, only limited data are available regarding the role of human iNKT cells in rheumatoid arthritis (RA). This is partly because of their relative rarity, they account for 0.05–0.1% of peripheral blood mononuclear cells (PBMC), rendering accurate enumeration problematical.12 13 Although three studies have indicated that peripheral blood iNKT cell frequency is reduced in RA patients14,–,16 conclusions were based on the enumeration of as few as 100 000 cells per sample in one study, and an unspecified number in the remaining two. Furthermore, the majority of the included patients were receiving immunosuppressive drugs, confounding the analyses. In addition, data from different murine models of autoimmune inflammatory arthritis suggest that iNKT cells could be protective during the induction phase but pathogenic during the effector phase of arthritis.17 18 Therefore, whether iNKT cell frequency and function differs between patients with early and established RA remains to be established.

    In this study, we have enumerated the peripheral blood iNKT cell frequency in 23 untreated, newly diagnosed RA patients, 23 RA patients receiving disease modifying therapies, 22 healthy controls and 27 patients presenting with recent onset of painful joints. The frequency of peripheral blood iNKT cells was determined by analysing 500 000 to over 1 million lymphocytes per sample, providing robust data (the statistical estimation of the minimum number of lymphocytes required for the analysis to generate robust data is provided in the Methods section). In addition, we investigated the relationship between peripheral blood iNKT cells frequency and inflammation, and the effects of immunosuppressive therapy. Finally, we explored whether iNKT cell analysis may improve the diagnostic algorithm of patients presenting with joint pain.

    Methods

    Subjects

    Forty-six RA patients and 27 patients presenting with recent-onset painful joints at a rheumatology clinic of a teaching hospital in the UK were recruited. Twenty-two healthy volunteers were recruited from the staff of Newcastle University. Ethical approval was obtained from Newcastle and North Tyneside research ethics committee. Tables 1 and 2 summarise the clinical features of the patients. Detailed clinical information on individual patients presenting to the early arthritis clinic is provided in supplementary table 1 (available online only).

    View this table:
    Table 1

    Clinical characteristics of the RA patients in this study

    View this table:
    Table 2

    Group characteristics of the patients presenting to the early arthritis clinic

    Culture medium, antibodies and reagents

    RPMI-1640 medium supplemented with l-glutamine, penicillin/streptomycin and 5% human AB serum (all from Sigma-Aldrich, Gillingham, Dorset, UK) was used. All cells were incubated at 37°C in a humidified incubator with 95% air and 5% carbon dioxide. The following antibodies were used in flow-cytometric analysis. Fluorescein isothiocyanate-conjugated antiVα24 (C15; Beckman Coulter, High Wycombe, UK), phycoerythrin-conjugated antiVβ11 (C21; Beckman Coulter), peridin chlorophyll protein-conjugated antiCD3 (Immunotech, Hamburg, Germany).

    Cell preparation

    Peripheral blood samples were obtained after informed consent. PBMC were isolated using density gradient centrifugation over lymphoprep.

    Flow-cytometric analysis

    Five million PBMC were incubated with flurochrome-conjugated antibodies for 30 min, washed and analysed with a BD LSRII 18-colour flow cytometer. Samples were gated on lymphocytes according to the forward and side scatter characteristics. A minimum of 500 000 lymphocyte-gated events was collected for each sample. The number of Vα24+Vβ11+CD3+ cells was determined and expressed as a percentage of lymphocytes.

    Estimation of absolute peripheral blood iNKT cell number

    Peripheral blood iNKT cell numbers in patients were estimated by multiplying the percentage of iNKT cells among lymphocytes by the ‘lymphocyte count’ (cell number/l, determined by a contemporaneous complete blood count measured in a clinical laboratory). As the complete blood count was not performed in healthy volunteers, the mean value (2.25×106/l) and the lower limit (1.5×106/l) of the normal ranges of lymphocyte count were used in the calculation.

    iNKT cell stimulation

    One million PBMC were stimulated with 100 ng/ml α-galactosylceramide (Alexis Biochemicals, Exeter, UK) for 12 days, which give optimal expansion of iNKT cells (unpublished observations). The percentages of iNKT cells of lymphocytes were enumerated before and after α-galactosylceramide stimulation.

    Statistical analysis

    A sample size calculation was undertaken to determine the minimum number of lymphocyte-gated events that was necessary to capture during flow-cytometric analysis. For this calculation an a-priori expected iNKT cell frequency of 0.1% was derived from the available published data on the frequency of peripheral blood iNKT cells of normal individuals,12 19,–,22 and an ability to detect differences of 0.01% was considered desirable. Based on 95% CI, the minimum number of lymphocyte-gated events required (assuming 100% specificity and sensitivity for the flurochrome-conjugated antibodies used) was calculated as 389 000, and so for the purposes of this study was set to 500 000. Regarding comparator group sample size (patients and controls), a power calculation demonstrated that 23 and 17 individuals were needed in each group to achieve 90% and 80% power, respectively, to detect a difference in the means between groups of 1 SD of the mean of the control group (p<0.05).

    Statistical analysis including descriptive statistics, group differences, treatment effect and correlation analyses was performed using GraphPad Prism software version 4. The statistical tests used for each analysis are specified individually in the text; p<0.05 was considered statistically significant.

    Results

    Peripheral blood iNKT cell frequency and number are markedly reduced in patients with untreated early RA as well as established disease

    We enumerated the frequency of peripheral blood iNKT cells in 46 RA patients and 22 healthy controls using flow cytometry. Of the 46 RA patients, 23 were taking immunosuppressive treatment and 23 were untreated, newly diagnosed patients. iNKT cells were defined by the co-expression of Vα24 and Vβ11 TCR α and β-chains, respectively (figure 1A). We found that the frequency of iNKT cells was more than 15-fold lower among patients with RA (medians and interquartile ranges (IQR) were 0.021% and 0.0068–0.0870% for controls and 0.001% and 0.0004–0.0026% for RA patients, respectively, p<0.001, two-tailed Mann–Whitney U test). Furthermore, newly diagnosed RA patients who were not on immunosuppressive therapy also had a reduced frequency of circulating iNKT cells (median 0.0007% (untreated, early RA) vs 0.0016% (established RA), p=0.41, figure 1B). As the frequencies of iNKT cells in early and established RA are both profoundly diminished, the analysis of more lymphocyte-gated events may be required to detect any significant difference in iNKT cell frequencies between these two groups. The absolute number of peripheral blood iNKT cells in RA patients was also significantly reduced (see supplementary figure 1, available online only).

    Figure 1
    Figure 1

    Peripheral blood invariant natural killer T (iNKT) cell frequency is markedly reduced in rheumatoid arthritis (RA) patients. (A) The gating strategy (indicated by the circle within the flow-cytometric plots) used in the enumeration of peripheral blood iNKT cell frequency: The letter and the number at the top left corner of the plot indicate the gate (A for lymphocytes, B for CD3+ lymphocytes and C for iNKT cells) and the number of events within that gate, respectively. Representative flow-cytometric plots of a healthy control (upper panel) and a RA patient (lower panel) is shown. (B) The frequency of peripheral blood iNKT cells as a percentage of lymphocytes was shown for all RA patients (all RA, filled diamonds), established RA patients receiving immunosuppressive therapy (RA, filled squares), untreated, newly diagnosed RA patients (ERA, filled triangles) and healthy controls (Ctrl, filled inverted triangles). ***p<0.001. Horizontal lines represent median values of the groups.

    Impaired proliferative response to α-galactosylceramide, an iNKT-specific agonist, among PBMC of RA patients

    To determine whether the functional response of peripheral blood iNKT cells from RA patients is also impaired, we compared the responses of peripheral blood iNKT cells from 20 RA patients and 15 healthy volunteers after stimulation with α-galactosylceramide for 12 days (figure 2). The percentage of iNKT cell among PBMC increased by a median of 121 (IQR 54–228) times for healthy controls, compared with 31 (IQR 7–65) times for RA patients (p=0.037, two-tailed Mann–Whitney U test). There was no correlation between α-galactosylceramide responsiveness and the baseline frequency of iNKT cells (r=–0.2, p=0.16; two-tailed Spearman's).

    Figure 2
    Figure 2

    The proliferative response to α-galactosylceramide was impaired in peripheral blood mononuclear cells (PBMC) from rheumatoid arthritis (RA) patients. (A) PBMC from RA patients (diamonds) and healthy controls (inverted triangles) were stimulated with α-galactosylceramide for 12 days and the percentage of invariant natural killer T (iNKT) cells before (filled symbols) and after (unfilled symbols) stimulation is shown. Horizontal lines represent the median values of the groups. (B) Fold expansion is defined by the ratio of the percentage of iNKT cells after and before stimulation with α-galactosylceramide. *p<0.05. Dotted lines represent the interquartile range of the fold expansion for healthy controls.

    It has been suggested that RA patients can be classified according to their iNKT cell responsiveness to α-galactosylceramide.14 15 In order to determine whether α-galactosylceramide non-responsiveness is more common among RA patients or is affected by immunosuppressive treatment or disease stage, we used the 25th percentile of the fold expansion of iNKT cells from our healthy controls as the cut-off value for α-galactosylceramide responsiveness. Using this approach, 75% of RA patients were non-responders compared with 20% of healthy controls. There was no significant difference in α-galactosylceramide responsiveness between patients with established RA (six non-responders out of eight) and early untreated RA (nine non-responders out of 12).

    Preserved peripheral blood iNKT cell frequency distinguishes non-inflammatory joint pain from RA in an early arthritis clinic setting

    We next investigated whether peripheral blood iNKT cell frequency can distinguish RA from other diagnoses. We studied 27 patients attending an early arthritis clinic with less than 12 months of joint symptoms. Their clinical characteristics, initial diagnosis and final diagnosis (after a follow-up period of up to 12 months) are summarised in table 2. Final diagnoses were RA (nine), osteoarthritis (six), non-inflammatory arthralgia not caused by osteoarthritis (six), self-limiting inflammatory arthropathy (one) and other inflammatory arthropathy (one sarcoid, one gout, one psoriatic arthropathy and two seronegative arthritis). Figure 3 illustrates the peripheral blood iNKT cell frequency according to the final diagnosis. Patients with RA had a significantly lower peripheral blood iNKT cell frequency compared with patients with non-inflammatory joint pain not caused by osteoarthritis (p<0.001, two-tailed Mann–Whitney U test), who were similar to healthy controls. Interestingly, patients with other inflammatory arthropathies and osteoarthritis also had reduced circulating iNKT cell frequencies.

    Figure 3
    Figure 3

    Normal peripheral blood invariant natural killer T (iNKT) frequency predicts against rheumatoid arthritis (RA) as a diagnosis for patients presenting with recent-onset painful joints. The percentages of iNKT cells in peripheral blood lymphocytes from 27 patients presenting with recent-onset joint pain was determined by flow cytometry. The results were categorised according to the final diagnosis: RA (triangles), other inflammatory arthropathy (IA, diamonds), osteoarthritis (OA, circles), non-inflammatory joint pain not due to osteoarthritis (non-IA, not OA, crosses). The peripheral blood iNKT cell frequency of healthy controls (Ctrl, inverted triangles) was included for comparison. Horizontal lines represent the median values of the groups. ***p<0.001. NS, non-statistically significant.

    Peripheral blood iNKT cell frequency increases following immunosuppressive treatment in RA patients

    Our finding that peripheral blood iNKT cell frequencies were lower in untreated RA patients prompted us to study the effects of therapy. We enumerated the frequency of peripheral blood iNKT cells in seven newly diagnosed RA patients and on five further occasions after commencement of methotrexate at weeks 2, 4, 6, 8 and 13. The iNKT cell frequency increased in all seven patients as early as 2 weeks into treatment (p=0.037, two-tailed paired t-test, figure 4A). Despite variability during the study, the increase in iNKT cell frequency was sustained in all but one patient at the 13-week endpoint (p=0.01, see supplementary figure 2, available online only). Absolute numbers of iNKT cells increased in parallel (figure 4A) but there was no obvious link to clinical response defined using either the disease activity score in 28 joints, C-reactive protein (CRP) or erythrocyte sedimentation rate (ESR) (data not shown).

    Figure 4
    Figure 4

    The frequency and number of invariant natural killer T (iNKT) cells in peripheral blood increases following the initiation of methotrexate and correlated inversely with C-reactive protein (CRP). (A) The percentages (filled squares) and number (unfilled squares) of iNKT cells in the peripheral blood of seven rheumatoid arthritis patients were enumerated before and after treatment with methotrexate for 2 weeks. *p<0.05. (B) Serum concentration of CRP was plotted against the peripheral blood frequency of iNKT cells for all patients in this study with contemporaneous CRP results (60 out of 61 patients). The scale of the x-axis was split to allow better visualisation of individual data points.

    Peripheral blood iNKT cell frequency correlates inversely with CRP

    As the peripheral blood iNKT cell frequency was also reduced in patients with other inflammatory arthropathies, we sought a link to systemic inflammation, using CRP and ESR. We found that peripheral blood iNKT cell frequency was inversely correlated to CRP (n=60, figure 4B). The association was moderate (r=–0.34, two-tailed Spearman's) but statistically highly significant (p=0.008). The correlation was stronger if patients receiving immunosuppressive treatment were excluded (n=40, r=–0.44, p=0.0045, two-tailed Spearman's), suggesting that the effects of these drugs on iNKT cell levels is not mediated solely via a reduction in systemic inflammation. There was no correlation between peripheral blood iNKT cell frequency and ESR, however (data not shown). Similar results were obtained when absolute iNKT cell numbers were used in the statistical analyses (data not shown). There was also no significant correlation between the responsiveness to α-galactosylceramide and either CRP or ESR (p=0.62 and 0.16, respectively, two-tailed Spearman's).

    Discussion

    In this study, we have demonstrated a numerical and functional deficiency of peripheral blood iNKT cells in patients with RA. More importantly, normal peripheral blood iNKT cell frequency strongly predicted a non-inflammatory cause for patients presenting with joint pain. We also found that peripheral blood iNKT cell frequency correlated inversely with CRP and increased upon the initiation of immunosuppressive therapies.

    The rarity of iNKT cells among peripheral blood lymphocytes in humans means that their accurate enumeration can only be achieved by analysing a sufficient number of cells. Before embarking on this study we estimated the minimum number of cells required to give robust data. Recently, Montoya and colleagues13 enumerated the frequency of peripheral blood iNKT cells in 90 healthy individuals (aged 15–52 years) and reported a mean frequency of 0.055±0.054% of lymphocytes, similar to healthy control values in the current study. These investigators also analysed at least 500 000 ‘lymphocyte’-gated events.

    We defined iNKT cells by the co-expression of Vα24 and Vβ11 TCR α and β-chains. It has been suggested that α-galactosylceramide/CD1d tetramers provide the most specific tool for iNKT cell identification and that non-canonical, CD1d-restricted, α-galactosylceramide-specific T cells have been described,22 23 but several investigators have demonstrated independently that the frequency of Vα24+Vβ11+CD3+ T cells, even at low numbers, corresponds well with the iNKT cell frequency determined by CD1d tetramers.13 24 Furthermore, non-canonical, α-galactosylceramide/CD1d-specific T cells are detectable largely after the expansion of PBMC with α-galactosylceramide rather than in ex vivo PBMC.22 It should also be noted that other CD1d-restricted (but not α-galactosylceramide-specific) T cells (also termed type II NKT cells) have been described.25,–,27 However, the functional capacities and significance of these other NKT cell subsets remain poorly understood.

    The frequency of circulating iNKT cells has been reported to fall with age at a rate of 0.3–0.5% per year,28,–,30 although not in all studies.16 As the RA group was 20 years older than the control group, age could account for a two to threefold decrease in iNKT cell frequency in RA patients. However, we observed a greater than 15-fold reduction, suggesting that age alone was not responsible for the difference. Immunosuppressive therapies such as methotrexate might be expected to reduce iNKT cell frequency, but levels actually increased following treatment.

    The above observations suggest a true reduction of peripheral blood iNKT cell frequency in RA and in other inflammatory joint disease, potentially explained partly by a negative correlation between iNKT cell frequency and CRP. Indeed, our data suggest that a normal iNKT cell frequency may identify patients with non-inflammatory or self-limiting joint pain. Furthermore, it is noteworthy that two patients with undifferentiated arthritis who were subsequently diagnosed with inflammatory arthritis had normal CRP but reduced peripheral blood iNKT cell frequency at presentation. Perhaps surprisingly, peripheral blood iNKT cell frequency was also reduced in patients with osteoarthritis, which is traditionally considered a non-inflammatory condition. In this regard, our findings appear to point to the increasingly recognised role played by inflammation in osteoarthritis pathogenesis.31,–,33 Our data thus suggest that whereas a preserved circulating iNKT cell frequency predicts non-inflammatory joint pain, a reduced peripheral blood iNKT cell frequency does not implicate RA, but could be associated with other inflammatory joint disease or osteoarthritis (but the use of anti-cyclic citrullinated peptide antibodies and CRP may be useful to differentiate these diagnoses). Therefore, a consideration of the potential additive contribution of peripheral blood iNKT cell enumeration in the ongoing evaluation of diagnostic or prognostic algorithms for the assessment of patients presenting with joint pain is warranted.

    Although it is tempting to speculate a pathogenic role of iNKT cell deficiency in RA and other inflammatory arthropathies, the dissection of any mechanistic link requires further study. A mundane explanation for the reduced circulating iNKT cells in RA patients is their preferential accumulation in the inflamed joints. However, although iNKT cells were found in inflamed joints, the frequency was not increased compared with peripheral blood.14 15 We have enumerated the iNKT cell frequency in two paired synovial fluid and blood samples of RA patients with similar results (unpublished observations). Another possible explanation is that iNKT cell frequency is suppressed by systemic inflammation.

    Our data demonstrated a rise in iNKT cell frequency and number following treatment with methotrexate. There was no correlation between either baseline values or the rise in iNKT cell frequency and clinical response. A small sample size is a possible explanation for this negative finding, but we have recently extended our analysis to include 16 of the 23 early RA patients, using CRP and ESR values at 3 months as outcome measures with similar results (unpublished observation). Interestingly, a very recent report demonstrating that peripheral iNKT cell frequency recovered after rituximab treatment and low iNKT cell frequency was associated with active disease34 appears to corroborate our findings.

    We showed that poor responsiveness to α-galactosylceramide is common among RA patients, although responsiveness did not correlate with any clinical parameters. Several reasons may account for this lack of association. First, α-galactosylceramide is a specific agonist but not a natural ligand for iNKT cells. Its action is more akin to that of a mitogen for T cells and it is possible that testing the responses to a physiological ligand may be more informative. Second, we have only reported the proliferative responses to α-galactosylceramide. Linsen et al15 reported that there was a bias of IFN-γ-producing Vα24+ cells from RA patients compared with healthy controls after two rounds of α-galactosylceramide stimulation over a period of 2 weeks. However, there was no correlation between the cytokine response and the responder status and disease parameters.15 In this study, we found no difference in the percentages of IFN-γ+, IL-4+, IL-10+ or IL-17+ iNKT cells in α-galactosylceramide-expanded PBMC samples from six RA patients and four healthy controls (see supplementary figure 3, available online only), the number of iNKT cells remained small even after α-galactosylceramide stimulation in some samples, lessening the reliability of the data. Third, the impaired α-galactosylceramide responses may be linked with reduced antigen presentation although antigen-presenting cells from α-galactosylceramide non-responder RA patients are equally competent in presenting α-galactosylceramide to iNKT cells.14 Indeed, we found no difference in CD1d expression on CD14+ monocytes or CD19+ B cells between RA patients and healthy controls (unpublished observations).

    Several subsets of human iNKT cells with distinct profiles of cytokine production have been described. For instance, CD4+ iNKT cells produce both Th1 and Th2 cytokines, the CD8+ and CD4–CD8– subsets produce predominantly Th1 cytokines,12 35 whereas the NK1.1ℒ subset is linked with IL-17 production.36 In this study, although the expression of co-receptors CD4 and CD8 was measured, due to the very low numbers of iNKT cells in RA patients (median 8, mean 20, range 0–189), such an analysis would be unreliable and further studies are needed to investigate the role of different iNKT cell subsets in RA.

    In conclusion, we have demonstrated that peripheral blood iNKT cells are impaired both numerically and functionally in RA patients. In addition, our observations that the peripheral iNKT cell frequency correlated inversely to CRP and that normal iNKT cell frequency predicted a self-limiting or non-inflammatory cause of joint pain suggests that iNKT cells could play a role in the pathogenesis of RA and may be a useful additional biomarker in the diagnosis of RA and other inflammatory joint diseases.

    Acknowledgments

    The authors thank all the patients and volunteers who took part in this study and the medical and nursing staff who assisted with the blood sample collection. The authors are grateful for the assistance of Dr Tom Chadwick, clinical trial statistician, Newcastle University, in sample size calculation.

    References

    1. ↵
      1. Wilson SB,
      2. Delovitch TL
      . Janus-like role of regulatory iNKT cells in autoimmune disease and tumour immunity. Nat Rev Immunol 2003;3:211–22.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Cerundolo V,
      2. Silk JD,
      3. Masri SH,
      4. et al
      . Harnessing invariant NKT cells in vaccination strategies. Nat Rev Immunol 2009;9:28–38.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Bendelac A,
      2. Savage PB,
      3. Teyton L
      . The biology of NKT cells. Annu Rev Immunol 2007;25:297–336.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Taniguchi M,
      2. Harada M,
      3. Kojo S,
      4. et al
      . The regulatory role of Valpha14 NKT cells in innate and acquired immune response. Annu Rev Immunol 2003;21:483–513.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Hammond KJ,
      2. Kronenberg M
      . Natural killer T cells: natural or unnatural regulators of autoimmunity? Curr Opin Immunol 2003;15:683–9.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Bendelac A,
      2. Lantz O,
      3. Quimby ME,
      4. et al
      . CD1 recognition by mouse NK1+ T lymphocytes. Science 1995;268:863–5.
      OpenUrlAbstract/FREE Full Text
    7. ↵
      1. Kawano T,
      2. Cui J,
      3. Koezuka Y,
      4. et al
      . CD1d-restricted and TCR-mediated activation of valpha14 NKT cells by glycosylceramides. Science 1997;278:1626–9.
      OpenUrlAbstract/FREE Full Text
    8. ↵
      1. Brossay L,
      2. Chioda M,
      3. Burdin N,
      4. et al
      . CD1d-mediated recognition of an alpha-galactosylceramide by natural killer T cells is highly conserved through mammalian evolution. J Exp Med 1998;188:1521–8.
      OpenUrlAbstract/FREE Full Text
    9. ↵
      1. Chiba A,
      2. Oki S,
      3. Miyamoto K,
      4. et al
      . Suppression of collagen-induced arthritis by natural killer T cell activation with OCH, a sphingosine-truncated analog of alpha-galactosylceramide. Arthritis Rheum 2004;50:305–13.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Kaieda S,
      2. Tomi C,
      3. Oki S,
      4. et al
      . Activation of invariant natural killer T cells by synthetic glycolipid ligands suppresses autoantibody-induced arthritis. Arthritis Rheum 2007;56:1836–45.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Coppieters K,
      2. Van Beneden K,
      3. Jacques P,
      4. et al
      . A single early activation of invariant NK T cells confers long-term protection against collagen-induced arthritis in a ligand-specific manner. J Immunol 2007;179:2300–9.
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Lee PT,
      2. Benlagha K,
      3. Teyton L,
      4. et al
      . Distinct functional lineages of human V(alpha)24 natural killer T cells. J Exp Med 2002;195:637–41.
      OpenUrlAbstract/FREE Full Text
    13. ↵
      1. Montoya CJ,
      2. Pollard D,
      3. Martinson J,
      4. et al
      . Characterization of human invariant natural killer T subsets in health and disease using a novel invariant natural killer T cell-clonotypic monoclonal antibody, 6B11. Immunology 2007;122:1–14.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Kojo S,
      2. Adachi Y,
      3. Keino H,
      4. et al
      . Dysfunction of T cell receptor AV24AJ18+, BV11+ double-negative regulatory natural killer T cells in autoimmune diseases. Arthritis Rheum 2001;44:1127–38.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Linsen L,
      2. Thewissen M,
      3. Baeten K,
      4. et al
      . Peripheral blood but not synovial fluid natural killer T cells are biased towards a Th1-like phenotype in rheumatoid arthritis. Arthritis Res Ther 2005;7:R493–502.
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      1. van der Vliet HJ,
      2. von Blomberg BM,
      3. Nishi N,
      4. et al
      . Circulating V(alpha24+) Vbeta11+ NKT cell numbers are decreased in a wide variety of diseases that are characterized by autoreactive tissue damage. Clin Immunol 2001;100:144–8.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Coppieters K,
      2. Dewint P,
      3. Van Beneden K,
      4. et al
      . NKT cells: manipulable managers of joint inflammation. Rheumatology (Oxford) 2007;46:565–71.
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Kim HY,
      2. Kim HJ,
      3. Min HS,
      4. et al
      . NKT cells promote antibody-induced joint inflammation by suppressing transforming growth factor beta1 production. J Exp Med 2005;201:41–7.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Sandberg JK,
      2. Bhardwaj N,
      3. Nixon DF
      . Dominant effector memory characteristics, capacity for dynamic adaptive expansion, and sex bias in the innate Valpha24 NKT cell compartment. Eur J Immunol 2003;33:588–96.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Lucas M,
      2. Gadola S,
      3. Meier U,
      4. et al
      . Frequency and phenotype of circulating Valpha24/Vbeta11 double-positive natural killer T cells during hepatitis C virus infection. J Virol 2003;77:2251–7.
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Kukreja A,
      2. Cost G,
      3. Marker J,
      4. et al
      . Multiple immuno-regulatory defects in type-1 diabetes. J Clin Invest 2002;109:131–40.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      1. Gadola SD,
      2. Dulphy N,
      3. Salio M,
      4. et al
      . Valpha24-JalphaQ-independent, CD1d-restricted recognition of alpha-galactosylceramide by human CD4(+) and CD8alphabeta(+) T lymphocytes. J Immunol 2002;168:5514–20.
      OpenUrlAbstract/FREE Full Text
    23. ↵
      1. Lee PT,
      2. Putnam A,
      3. Benlagha K,
      4. et al
      . Testing the NKT cell hypothesis of human IDDM pathogenesis. J Clin Invest 2002;110:793–800.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Araki M,
      2. Kondo T,
      3. Gumperz JE,
      4. et al
      . Th2 bias of CD4+ NKT cells derived from multiple sclerosis in remission. Int Immunol 2003;15:279–88.
      OpenUrlAbstract/FREE Full Text
    25. ↵
      1. Park SH,
      2. Weiss A,
      3. Benlagha K,
      4. et al
      . The mouse CD1d-restricted repertoire is dominated by a few autoreactive T cell receptor families. J Exp Med 2001;193:893–904.
      OpenUrlAbstract/FREE Full Text
    26. ↵
      1. Godfrey DI,
      2. MacDonald HR,
      3. Kronenberg M,
      4. et al
      . NKT cells: what's in a name? Nat Rev Immunol 2004;4:231–7.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      1. Terabe M,
      2. Swann J,
      3. Ambrosino E,
      4. et al
      . A nonclassical non-Valpha14Jalpha18 CD1d-restricted (type II) NKT cell is sufficient for down-regulation of tumor immunosurveillance. J Exp Med 2005;202:1627–33.
      OpenUrlAbstract/FREE Full Text
    28. ↵
      1. Peralbo E,
      2. DelaRosa O,
      3. Gayoso I,
      4. et al
      . Decreased frequency and proliferative response of invariant Valpha24Vbeta11 natural killer T (iNKT) cells in healthy elderly. Biogerontology 2006;7:483–92.
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Jing Y,
      2. Gravenstein S,
      3. Chaganty NR,
      4. et al
      . Aging is associated with a rapid decline in frequency, alterations in subset composition, and enhanced Th2 response in CD1d-restricted NKT cells from human peripheral blood. Exp Gerontol 2007;42:719–32.
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. Molling JW,
      2. Kölgen W,
      3. van der Vliet HJ,
      4. et al
      . Peripheral blood IFN-gamma-secreting Valpha24+Vbeta11+ NKT cell numbers are decreased in cancer patients independent of tumor type or tumor load. Int J Cancer 2005;116:87–93.
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. Benjamin M,
      2. McGonagle D
      . Histopathologic changes at “synovio-entheseal complexes” suggesting a novel mechanism for synovitis in osteoarthritis and spondylarthritis. Arthritis Rheum 2007;56:3601–9.
      OpenUrlCrossRefPubMedWeb of Science
    32. ↵
      1. Haywood L,
      2. McWilliams DF,
      3. Pearson CI,
      4. et al
      . Inflammation and angiogenesis in osteoarthritis. Arthritis Rheum 2003;48:2173–7.
      OpenUrlCrossRefPubMedWeb of Science
    33. ↵
      1. Brooks P
      . Inflammation as an important feature of osteoarthritis. Bull WHO 2003;81:689–90.
      OpenUrlPubMedWeb of Science
    34. ↵
      1. Parietti V,
      2. Chifflot H,
      3. Sibilia J,
      4. et al
      . Rituximab treatment overcomes reduction of regulatory iNKT cells in patients with rheumatoid arthritis. Clin Immunol 2010;134:331–9.
      OpenUrlCrossRefPubMedWeb of Science
    35. ↵
      1. Gumperz JE,
      2. Miyake S,
      3. Yamamura T,
      4. et al
      . Functionally distinct subsets of CD1d-restricted natural killer T cells revealed by CD1d tetramer staining. J Exp Med 2002;195:625–36.
      OpenUrlAbstract/FREE Full Text
    36. ↵
      1. Michel ML,
      2. Keller AC,
      3. Paget C,
      4. et al
      . Identification of an IL-17-producing NK1.1(neg) iNKT cell population involved in airway neutrophilia. J Exp Med 2007;204:995–1001.
      OpenUrlAbstract/FREE Full Text

    Supplementary materials

    Footnotes

    • Funding This work was supported by the JGW Patteson Foundation Trust, c/o Sintons Solicitors, Barrack Road, Newcastle upon Tyne, NE4 6DB, UK. WFN and AP received salary support from the Arthritis Research Campaign, Copeman House, St Mary's Court, Chesterfield, S41 7TD, UK.

    • Competing interests None.

    • Ethics approval This study was conducted with the approval of the Newcastle and North Tyneside research ethics committee.

    • Provenance and peer review Not commissioned; externally peer reviewed.