Respiratory Medicine (2010) 104, 571e577

ava i lab le at www.sc ienced i rec t . com

j ourna l homepage : www.e lsev ier . com/ loca te / rmed

Deficiency of pulmonary Va24 Vb11 natural killerT cells in corticosteroid-naıve sarcoidosis patients

Peter Korosec*, Matija Rijavec, Mira Silar, Izidor Kern, Mitja Kosnik,Katarina Osolnik

University Clinic of Respiratory and Allergic Diseases, Golnik 36, 4204 Golnik, Slovenia

Received 17 June 2009; accepted 15 November 2009Available online 1 December 2009

KEYWORDSSarcoidosis;Bronchoalveolar lavage;Invariant natural killerT cells

* Corresponding author. Tel: þ386 4143.

E-mail address: peter.korosec@klin

0954-6111/$ - see front matter ª 200doi:10.1016/j.rmed.2009.11.008

Summary

Invariant natural killer T (NKT) cells might contribute to the amplified and prolonged T-cellimmune response that characterizes sarcoidosis. Therefore, we want to investigate thefrequency and distribution of pulmonary invariant NKT cells in corticosteroid-naıve patientswith sarcoidosis. We used multi-parameter flow cytometry with antibodies against CD3, CD4,CD8, CD14, CD19, CD45, CD16/56, TCR Va24, and TCR Vb11, on bronchoalveolar lavage fluid(BALF), to examine the frequency and distribution of pulmonary invariant NKT cells in 47 newlydiagnosed sarcoidosis patients and in 8 control subjects. The frequencies of BALF Va24 Vb11invariant NKT cells were significantly lower in patients with sarcoidosis in comparison tocontrol subjects. Moreover, lower invariant NKT cell frequencies in patients with sarcoidosissignificantly correlated with exaggerated BALF lymphocytosis and CD4 T cell responses. Thisstudy demonstrated a pulmonary deficiency in the frequency of a subset of T cells with immu-noregulatory function in patients with sarcoidosis.ª 2009 Elsevier Ltd. All rights reserved.

Introduction

Sarcoidosis is a multi-system disorder that predominantlyinvolves the lungs and is characterized by a Th1-biased,CD4-positive T cell response and granuloma formation.1

Despite progress in the characterization of the immuneprofile, little is known about the aetiology of sarcoidosisand how its response is generated. Recently, newly

2569 432; fax: þ386 4 2569

ika-golnik.si (P. Korosec).

9 Elsevier Ltd. All rights reserved

identified subsets of T cells with immunoregulatory func-tions, invariant natural killer T (invariant NKT) cells, haveshown to be deficient in the peripheral blood of sarcoidosispatients.2 NKT cells comprise a unique subgroup oflymphocytes that express features of both T and naturalkiller (NK) cells. These cells co-express T cell receptors(TCR) as well as markers associated with NK cells, such asCD56 and/or CD161.3 In humans, two major subsets ofnatural killer T cells have been described.4 The more widelystudied subset is invariant NKT cells, which express a highlyrestricted T cell receptor with an invariant Va24-Ja18 chainpaired with Vb11 and are dependent on the presentation ofthe glycolipid antigen through CD1d, a member of class I

.

572 P. Korosec et al.

non-polymorphic antigen presenting molecules.5 Thesecond class of NKT cells, CD1d- independent cells, expressa non-biased TCRa/b repertoire and are independent ofCD1d for their activity.6,7 Upon activation, NKT cells rapidlyproduce both Th1- and Th2-specific cytokines, notably IL-4and IFN-g, and exhibit both antigen-specific and NK-likecytolytic activities.4,6 Their cytokine secretion is deter-mined by the integration of different signals received in theperiphery.8 This rapid cytokine production is a manifesta-tion of innate-like immunity and provides NKT cells with theability to link the innate and adaptive immune responses.The dual regulatory and adjuvant functions of NKT cells arestill poorly understood. Invariant NKT cells might partici-pate in the innate immune response to promote antimi-crobial9 and anti-tumor immunity.10,11 Second, they canactively induce T cell tolerance12 and protect againstautoimmune disorders with increased CD4-positive Th1responses in preclinical models of type 1 diabetes,13,14

multiple sclerosis,15,16 and rheumatoid arthritis.17 Since,invariant NKT cells were also implicated in the regulation ofTh1-based response and deficiency of invariant NKT cellswas found in the peripheral blood of sarcoidosis patients,but inadequately evaluated in the lung,1 we examined thefrequency and distribution of pulmonary invariant NKT cellsin the bronchoalveolar lavage fluid (BALF) of large numberof corticosteroid-naıve sarcoidosis patients.

Materials and methods

Patients

Forty-seven patients (mean age Z 47 years; range 26e72years; 24 women and 23 men; 5 smokers and 13 ex-smokers,Caucasians) were recruited from the University Clinic ofRespiratory and Allergic Diseases in a prospective way overa 2.5-year period. All patients were subjected to a diag-nostic work-up, which included posteroanterior chestradiography, optionally high-resolution CT, pulmonaryfunction tests, and fiberoptic bronchoscopy with biopsy andbronchoalveolar lavage (BAL) (Table 1). All patients hadnewly diagnosed, histologically confirmed pulmonarysarcoidosis, without previous corticosteroid treatment. The

Table 1 Bronchoalveolar lavage fluid (BALF) cell countsand results of lung function testing from control subjectsand study patients with sarcoidosis. Data are presented asmedian with range in parentheses.

BALF cell counts: Controls(N Z 8)

Sarcoidosis(N Z 47)

Total cells (�105cells/ml) 0.95 (0.6e5.3) 1.3 (0.3e4.2)Neutrophils (%) 2 (1e7) 3 (1e40)Eosinophils (%) 1 (1e2) 1 (1e5)Macrophages (%) 90.5 (82e95)* 55 (14e96)Lymphocytes (%) 6 (2e14)* 33 (3e77)Lung function: FEV1 (%) 99.5 (86e121)* 90.5 (22e136)FVC (%) 98.5 (78e119) 90 (33e122)DLCO (%) 90.5 (74e105) 86 (42e134)

* Significantly different from the values for patients withsarcoidosis, P< 0.04; ManneWhitney test.

diagnosis of sarcoidosis was in agreement with the state-ment adopted by the joint committees of ATS and ERS.18

Lofgren’s syndrome, which was defined as acute presenta-tion, with or without fever, with erythema nodosum andpolyarthralgia and bilateral hilar lymphadenopathy onchest radiography, was identified in 14 of those patients.Fifteen patients had Scadding chest X-ray stage 1 chestradiograph, 28 stage 2, three stages 3 and one stage 4radiograph at the time of diagnosis.19 All patients werefollowed up for 1 year according whether treatment wasrequired and whether there was need for long-term treat-ment. We also included 8 control subjects without anypulmonary morbidities (mean age Z 56 years; range 28e84years; 5 women and 3 men; 1 smokers and 1 ex-smokers,Caucasians). Both groups showed no significant differencesaccording to age, sex and smoking habit (t and Fischer’stest). All subjects gave their written consent and the studywas approved by the National Ethics Committee.

Bronchoalveolar lavage procedure

Bronchoalveolar lavage (BAL) was performed as previouslydescribed.7 Briefly, we instilled 7 aliquots of 20 ml and one of10 ml, up to a total volume of 150 ml of saline. After eachinstillation the aliquot was immediately recovered, pooledand thereafter proceed for cytological and flow cytometryanalysis. Cytospinpreparationswere stainedaccording to theMay-Grunwald Giemsa and Papanicolaou method. Differen-tial cell counts were performed by one observer blinded tothe clinical characteristics, counting 200 cells (Table 1).

Flow cytometry

Flow cytometry was performed as previously described7 onan FACSCalibur (BD Biosciences, San Jose, CA, USA) usingmAb against CD3, CD4, CD8, CD14, CD19, CD45, CD16/56,Va24 (clone 6B11) T cell receptor (TCR) (all from BDBiosciences), Vb11 TCR (Immunotech, Marseille, Fran-ce),20e22 and isotype-matched antibody controls (BDBiosciences) that were directly conjugated to either FITC,PE or PerCP. The BALF was first strained through a 70 mmcell strainer (BD Biosciences); the cells were centrifuged at460� g for 5 min, resuspended in Haemaccel (Behring-werke, Marburg, Germany), and incubated with therespective mAbs for 15 min, followed by lysing, washing,and fixation. Labeled cells were analyzed by CellQuestsoftware (BD Biosciences). Lymphocytes were distinguishedby the initial gating using characteristic lymphocyteforward/side-scatter properties and the gate was furtherconfirmed by CD14/CD45 staining. Invariant NKT cells(CD3þ Va24þ Vb11þ), NKT cells (CD3þ CD16/56þ), naturalkiller (NK) cells (CD3-CD16/56þ), helper and suppressor Tcells, and B lymphocytes were scored in percentages oflymphocytes.

Statistical analyses

The distribution of data was tested by the ShapiroeWilkstest. Since the data were not normally distributed, weperformed a ManneWhitney. The strength of associationbetween invariant natural killer T cells and other variables

Invariant NKT cells in sarcoidosis 573

was obtained by the Spearman rank-order method. Proba-bility values of P< 0.05 were accepted as significant.Analyses were performed with GraphPad Prism 5.

Results

When specimens of BALF were examined for the presenceof CD4þ and CD8þ T cells, we observed a higher number ofCD4þ cells, with an increase in the CD4:CD8 ratio inpatients with sarcoidosis and a lower number of CD4þ cellswith a normal CD4:CD8 ratio in control subjects (Table 2).

In all study subjects, the majority of pulmonarylymphocytes were T cells, with a very low number of Blymphocytes, thus the frequencies of T cells and thelymphocyte counts were higher in sarcoidosis patients(Tables 1 and 2). The sarcoidosis group further showedlower macrophage counts (Table 1). There was no signifi-cant difference in the presence of BALF natural killer cellsand natural killer T cells, as detected by staining withmonoclonal antibodies to CD3 and CD16/56 (Table 2).

We then examined the BALF for the presence ofinvariant NKT cells, as detected by staining with mono-clonal antibodies to CD3, Va24 and Vb11 T cell receptors.We found a striking reduction in pulmonary lymphocytes co-expressing invariant T cell receptors in patients withsarcoidosis (Figs. 1 and 2). The median (range) percentageof lymphocytes staining for CD3, Va24 and Vb11 TCRwas 0.19% (0.001e2.9%) for patients with sarcoidosis. Thiswas significantly lower than in control subjects. (1.03%(0.13e1.4%); P Z 0.001; ManneWhitney test).

We also analyzed the frequencies of BALF invariant NKTcells in different clinical categories of our patients, dividedaccording to the mode of presentation (Lofgren’s syndromevs. other forms), and in patients with different chestradiograph stages. Invariant NKT cell frequencies werecomparably low in both clinical categories of sarcoidosispatients (median (range); Lofgren’s syndrome vs. otherforms; 0.21% (0.02e1.1%) vs. 0.19% (0.001e2.9%)) and indifferent radiograph stages (median (range); without vs.with parenchymal disease 0.15% (0.001e1.1%) vs. 0.23%(0.02e2.9%)). There was also no difference according tosmoking history (smokers and ex-smokers vs. non smokers;0.21% (0.001e1.9%) vs. 0.19% (0.02e2.9%)) or sex.

Table 2 T lymphocytes, B lymphocytes, NK and NKT cellsin BALF from study subjects. Data are presented as medianwith range in parentheses.

% of lymphocytes Controls(N Z 8)

Sarcoidosis(N Z 47)

T lymphocytes (CD3þ) 85 (76e97)* 92 (74e100)T helper (CD3þ CD4þ) 55 (38e80)* 77 (15e92)T suppressor (CD3þ CD8þ) 24.3 (13e52)* 13 (2e84)B lymphocytes (CD19þ) 0 (0e4) 1 (0e4)NK cells (CD3-CD16/56þ) 3 (2e7) 2 (1e13)NKT cells (CD3þ CD16/56þ) 3 (2e5) 2 (0e11)CD4/CD8 ratio 2.3 (0.8e6.2)* 6.3 (0.2e46)

*Significantly different from the values for patients withsarcoidosis, P< 0.03; ManneWhitney test.

Eighteen of 33 patients with non-Lofgren’s sarcoidosiswere treated with systemic steroid therapy. In thosepatients, we found slightly lower pre-treated frequenciesof invariant NKT cells than in patients without need forcorticosteroid treatment (0.11% (0.02e1.3%) vs. 0.31%(0.001e2.9%)). However, these differences did not reachstatistical significance. Eleven patients needed only oneperiod of treatment, lasting for a maximum of 1 year, but 7patients needed treatment with corticosteroids for morethan 1 period or long-term treatment lasting for more than1 year. Both subgroups showed comparable pre-treatedfrequencies of BALF invariant NKT cells.

Because a major fraction of pulmonary T cells in patientswith sarcoidosis were CD4þ T cells, we wanted to establishwhether there was any correlation between frequencies ofBALF invariant NKT cells and the predominance of CD4þ Tcells in the sarcoidosis group. First, we found a significantnegative correlation between BALF invariant NKT cellsfrequencies and the total cell count (P Z 0.001,R Z�0.48), lymphocyte count (P Z 0.01, R Z�0.37), andthe concentration of BALF lymphocytes (P< 0.0001,R Z�0.58; Spearman method) in the sarcoidosis group.Second, there was a significant negative correlationbetween invariant NKT cell frequencies and BALF concen-tration of T cells (P< 0.0001, R Z�0.58) and CD4þ T cells(P Z 0.0008, R Z�0.47; Spearman method). There wasalso a significant correlation with macrophage counts(P Z 0.003, R Z 0.43). However, there was no significantcorrelation between the invariant NKT cells and other BALFcells, other T, B or NK lymphocyte subgroups, pulmonaryfunction test results or age in patients with sarcoidosis.Moreover, we could not demonstrate any significant corre-lation between BALF invariant NKT cells and any variablesin the control group, such as BALF cell counts, T, B, or NKlymphocytes, pulmonary function test results or age.

The BALF invariant NKT cells frequencies were alsoevaluated by ROC curve analyzes (Fig. 3). The area undera ROC curve was 0.79 and the P value was 0.009. Theoptimal recalculated threshold was <0.97%, with thesensitivity of 87% and the specificity of 75%. ROC curveanalyzes were also performed for CD4:CD8 ratio (Fig. 3).The optimal CD4:CD8 threshold was >3.3, with the sensi-tivity of 70% and the specificity of 88%, and with area of0.79 and the P value of 0.01.

Discussion

In this work, we evaluated the frequencies of invariantVa24 Vb11 NKT cells in the bronchoalveolar lavage fluid ofsarcoidosis patients. Previous report of Ho et al.2 demon-strated the peripheral blood deficiency of invariant NKTcells in sarcoidosis patients. However, previous data werenot consistent at the pulmonary or granuloma level. Mem-pel et al.23 have reported an absence of invariant NKT cellsin cutaneous lesions of sarcoidosis patients. In contrast,Kobayashi et al.24 suggested accumulation of invariant NKTcells in granulomatous lesions. Ho et al.2 also demonstratedlow numbers of invariant NKT cells in BALF of sarcoidosispatients, but BALF were done only in limited number ofpatients and with no data in control subjects. Thussarcoidosis is a multi-system disorder that predominantly

Figure 1 Flow-cytometric analysis of invariant NKT (CD3þ Va24þ Vb11þ) in bronchoalveolar lavage fluid from two patients withLofgren’s syndrome (panel A), a patient with non-Lofgren sarcoidosis (panel B), and two control subjects (panel C). Cells werestained with mAbs against CD3, Va24þ and Vb11þ TCR. Dot plots were generated after first forward/side-scatter lymphocytegating and second CD3þ gating.

Controls Sarcoidosis

0%

1%

2%

3%

4%P=0.001

VF

LA

Bα

V4

2β

tn

ai

ra

vn

i1

1

)s

et

yc

oh

pm

yl

fo

%(

sl

le

cT

KN

Figure 2 Summary plot of invariant NKT (CD3þ Va24þVb11þ) cell frequencies in BALF of patients with sarcoidosisand control subjects. The horizontal solid bars indicate themedian value for each group.

574 P. Korosec et al.

involves the lungs, and therefore, we focused on the BALFpulmonary T cells. We demonstrated a major pulmonarydeficiency of Va24 Vb11 invariant NKT cells in the lungsairway lumen of newly identified, corticosteroid-naıvesarcoidosis patients with approximate 5 fold reduction incomparison to control subjects. Moreover, invariant NKTcell frequencies were in significant negative correlationwith the BALF lymphocytosis and with the amplified BALF Tcell and CD4þ T cell response.

Several recent studies were undertaken with the aim ofclarifying the aetiopathology of sarcoidosis. These studiesprovided evidence that patients with sarcoidosis have anexaggerated Th1-biased, CD4-positive T cell response whichmight includes different subsets of CD4 T cells from CD4 Tcells lacking CD2825 to AV2S3þ CD4 T cells.26 Theyhypothesized that formation of organized immune granu-lomas might be a response to undefined antigen; however,what causes cells of the immune system to behave in thisamplified manner is unknown. Ho et al. recently suggested

0.0 0.2 0.4 0.6 0.8 1.0

0.0

0.2

0.4

0.6

0.8

1.0

iNKT < 0.97%sensitivity 87%specificity 75%

ROC curve for BALF iNKT cells

False positive rate

et

ar

evi

tis

op

eu

rT

0.0 0.2 0.4 0.6 0.8 1.0

0.0

0.2

0.4

0.6

0.8

1.0

CD4:CD8 > 3.3sensitivity 70%specificity 88%

ROC curve for BALF CD4:CD8 ratio

False positive rate

et

ar

evi

tis

op

eu

rT

Figure 3 ROC curve analysis of BALF invariant NKT cells and CD4:CD8 ratio. The area under a ROC curve for was 0.79 and theP value was �0.01. The optimal threshold was <0.97% for invariant NKT cells was and >3.3 for CD4:CD8 ratio.

Invariant NKT cells in sarcoidosis 575

that loss of an immunoregulatory subset of T cells, likeinvariant NKT cells, could contribute to the lack of controlof CD4-positive T cells.2 This idea is supported by studieswith animal models, which has shown that invariant NKTcells protect against disorders with increased CD4-positiveTh1 responses.13e17 The most persistent data are in thenonobese diabetic (NOD) mouse strain that spontaneouslydevelops autoimmune diabetes and exhibits both quanti-tative and qualitative invariant NKT cell defects.13,14 Thisidea was further supported by the findings that reconsti-tution of invariant NKT cells prevented chronic immunedamage to the pancreatic islet cells and the developmentof diabetes.13,14 Furthermore, there is evidence that indi-viduals predisposed to diabetes mellitus have dysfunctionof invariant NKT cells.27

The possible deficiency of invariant NKT cells insarcoidosis patients opens several important clinical ques-tions. First, is there any correlation between invariant NKTcells and the clinical course and/or therapeutic conse-quences of sarcoidosis? It is well-described that patientswith Lofgren’s syndrome have a better prognosis than dopatients with other manifestations of the disease.28 More-over, Lofgren’s syndrome is characterized by a defined startand end of immune activity, as more than 80% of patientshave complete disease resolution within 1e2 years.Consequently, only 1 (7%) of our Lofgren patients hadrequired corticosteroid treatment. By contrast, patientswho present with classical pulmonary infiltrates tend tohave a longer clinical course, and indeed, 18 (53%) our non-Lofgren patients required corticosteroid treatment. Wedemonstrated that the frequencies of invariant NKT cellswere comparably low in both phenotypes in differentradiograph stages at the starts of the disease. All patientswere clinically followed up to 1 year after the inclusion intothe study, and the patients who required corticosteroidtreatment tended to have lower initial frequencies ofinvariant NKT cells. However, those differences did notreach statistical significance, and pre-treated deficiencywas comparable between patients who needed only oneperiod of treatment, lasting for at most 1 year, and patientswho needed treatment with corticosteroid for more than 1period or long-term treatment lasting for more than 1 year.The major question that remains unanswered is whathappens to invariant NKT cells during the course of thedisease and do those cells influence the course of diseaseby limiting the expansion of antigen-driven T cells.

Unfortunately, none of the current studies follows upinvariant NKT cells through the course of disease, and theseobservations are only a single time point of the diseaseprocess.

It is noteworthy that sarcoidosis patients do not sharepulmonary invariant NKT abnormalities with other intersti-tial lung diseases, like hypersensitive pneumonitis, or withpuzzling observations in obstructive lung diseases. A ROCcurve analyse even suggests that evaluation of BALFinvariant NKT cell frequencies might be helpful in thedifferential diagnosis. We have recently shown significantlyincreased levels of BALF CD8þ CD56þ NKT cells that did notexpress the invariant T cell receptor in hypersensitivepneumonitis, but no such differences were observed insarcoidosis patients.7 We confirmed these recent observa-tions also in this study, as both sarcoidosis and controlsubjects have normal BALF CD56þ NKT numbers. Similardata were observed in two other previous reports.29,30

Despite initial identification of NKT cells by invariant TCRachain and NK surface antigens, CD56 and/or CD161, it isnow evident that NKT cells comprise a heterogeneoussubgroup with functionally distinct subpopulations thatexpress different patterns of a/b T cell receptors and NKcell antigens.3e7

Much more controversial are data about invariant NKTcells in asthma. Akbari et al.20 demonstrated highlyincreased number of invariant NKT cells (about 60% of CD3þCD4þ cells) in the lungs of patients with asthma, anda requirement for invariant natural killer T cells was indi-cated in the development of allergen-induced airwayhyperreactivity in mice.31 These findings have been chal-lenged by Vijayanand et al.21 and Thomas et al.,22 whoshowed significantly lower invariant NKT cells counts inasthma (fewer than 2% of T cells). Thomas suggested thatthis discrepancy may result from different gating strate-gies, as in BALF, it is very important to gate using charac-teristic lymphocyte forward-scatter (size) and side-scatter(granularity) to exclude large granular cells that might stainnonspecifically. Nevertheless, the critical question is whatthe ‘‘normal’’ frequency of BALF invariant NKT cells is. Forexample Akbari et al.20 and Mutalithas et al.32 includedsarcoidosis patients as a control subjects, and these wasalso the case for recent communication of Akbari group.33

On the other hand Vijayanand21 and Thomas22 did notinclude any control BALF samples. However, both Vijaya-nand and Thomas, which used methodologically

576 P. Korosec et al.

comparably setting as in our report, showed comparableBALF invariant NKT asthma count as our values in controlsubject.

The deficiency of invariant NKT cells in sarcoidosispatients is unexplained. The possible cause might includean absence or reduction due to genetic predisposition;cellular relocation might not be the case, as deficiency wasobserved in different compartments. Baev et al.34 recentlycharacterize one of the mechanisms that drive NKT celldifferentiation and demonstrate that this mechanism isimpaired in NOD mice that lack regulatory invariant NKTcell function. NOD mice carry a genetic defect of theSlamf1 gene that is associated with reduced signalinglymphocyte activation molecule (SLAM) expression ondouble-positive thymocytes and altered invariant NKT celldevelopment in the thymus.34,35 Second, a recent study byCantorna et al.36 on a KO mice model showed thatexpression of vitamin D receptor is required for normaldevelopment and function of invariant NKT cells. Appar-ently, the active form of vitamin D is produced at the sitesof the granulomatous reaction, and it was suggested thatvitamin D receptor gene polymorphism might be alsoa genetic risk factor for sarcoidosis.37

In summary, this study demonstrated a pulmonarydefect in the frequency of a subset of T cells with immu-noregulatory function in patients with sarcoidosis. More-over, lower invariant NKT cell frequencies in BALF ofpatients with sarcoidosis significantly correlated toelevated BALF lymphocytosis and T and CD4 T cell expan-sion. Therefore, our data suggest that this deficiency mightcontribute to the amplified immune response.

Conflict of interest

The authors declare that they do not have any conflict ofinterest related to this manuscipt.

References

1. Agostini C, Meneghin A, Semenzato G. T-lymphocytes andcytokines in sarcoidosis. Curr Opin Pulm Med. 2002 Sep;8(5):435e40.

2. Ho LP, Urban BC, Thickett DR, Davies RJ, McMichael AJ.Deficiency of a subset of T-cells with immunoregulatoryproperties in sarcoidosis. Lancet 2005 Mar 19e25;365(9464):1062e72.

3. Norris S, Doherty DG, Collins C, McEntee G, Traynor O,Hegarty JE, O’Farrelly C. Natural T cells in the human liver:cytotoxic lymphocytes with dual T cell and natural killer cellphenotype and function are phenotypically heterogenous andinclude Valpha24-JalphaQ and gammadelta T cell receptorbearing cells. Hum Immunol 1999 Jan;60(1):20e31.

4. Kronenberg M, Gapin L. The unconventional lifestyle of NKTcells. Nat Rev Immunol 2002 Aug;2(8):557e68.

5. Matsuda JL, Gapin L, Fazilleau N, Warren K, Naidenko OV,Kronenberg M. Natural killer T cells reactive to a singleglycolipid exhibit a highly diverse T cell receptor beta reper-toire and small clone size. Proc Natl Acad Sci U S A 2001 Oct 23;98(22):12636e41.

6. Wajchman HJ, Pierce CW, Varma VA, Issa MM, Petros J,Dombrowski KE. Ex vivo expansion of CD8þ CD56þ and CD8þCD56-natural killer T cells specific for MUC1 mucin. Cancer Res2004 Feb 1;64(3):1171e80.

7. Korosec P, Osolnik K, Kern I, Silar M, Mohorcic K, Kosnik M.Expansion of pulmonary CD8þ CD56þ natural killer T-cells inhypersensitivity pneumonitis. Chest 2007 Oct;132(4):1291e7.

8. Kadowaki N, Antonenko S, Ho S, Rissoan MC, Soumelis V,Porcelli SA, Lanier LL, Liu YJ. Distinct cytokine profiles ofneonatal natural killer T cells after expansion with subsets ofdendritic cells. J Exp Med 2001;193:1221e6.

9. Brigl M, Bry L, Kent SC, Gumperz JE, Brenner MB. Mechanism ofCD1d-restricted natural killer T cell activation during microbialinfection. Nat Immunol 2003;4:1230e7.

10. Crowe NY, Coquet JM, Berzins SP, Kyparissoudis K, Keating R,Pellicci DG, Hayakawa Y, Godfrey DI, Smyth MJ. Differentialantitumor immunity mediated by NKT cell subsets in vivo. J ExpMed 2005;202:1279e88.

11. Shimizu K, Goto A, Fukui M, Taniguchi M, Fujii S. Tumor cellsloaded with _-galactosylceramide induce innate NKT and NKcell-dependent resistance to tumor implantation in mice.J Immunol 2007;178:2853e61.

12. Sonoda KH, Exley M, Snapper S, Balk SP, Stein-Streilein J. CD1-reactive natural killer T cells are required for development ofsystemic tolerance through an immune-privileged site. J ExpMed 1999;190:1215e26.

13. Sharif S, Arreaza GA, Zucker P, Mi QS, Sondhi J, Naidenko OV,Kronenberg M, Koezuka Y, Delovitch TL, Gombert JM, et al.Activation of natural killer T cells by alpha-galactosylceramidetreatment prevents the onset and recurrence of autoimmunetype 1 diabetes. Nat Med 2001;7:1057e62.

14. Hong S, Wilson MT, Serizawa I, Wu L, Singh N, Naidenko OV,Miura T, Haba T, Scherer DC, Wei J, et al. The natural killer T-cell ligand-galactosylceramide prevents autoimmune diabetesin non-obese diabetic mice. Nat Med 2001;7:1052e6.

15. Miyamoto K, Miyake S, Yamamura T. A synthetic glycolipidprevents autoimmune encephalomyelitis by inducing TH2 biasof natural killer T cells. Nature 2001;413:531e4.

16. Jahng AW, Maricic I, Pedersen B, Burdin N, Naidenko O,Kronenberg M, Koezuka Y, Kumar V. Activation of natural killerT cells potentiates or prevents experimental autoimmuneencephalomyelitis. J Exp Med 2001;194:1789e99.

17. Kaieda S, Tomi C, Oki S, Yamamura T, Miyake S. Activation ofinvariant natural killer T cells by synthetic glycolipid ligandssuppresses autoantibody induced arthritis. Arthritis Rheum2007;56:1836e45.

18. ATS/ERS/WASOG Committee. Statement on sarcoidosis. AmJ Respir Crit Care Med 1999;160:736e55.

19. Scadding JG. Prognosis of intrathoracic sarcoidosis in England.Br Med J 1961;4:1165e72.

20. Akbari O, Faul JL, Hoyte EG, Berry GJ, Wahlstrom J,Kronenberg M, DeKruyff RH, Umetsu DT. CD4þ invariant T-cell-receptorþ natural killer T cells in bronchial asthma. N EnglJ Med. 2006 Mar 16;354(11):1117e29.

21. Vijayanand P, Seumois G, Pickard C, Powell RM, Angco G,Sammut D, Gadola SD, Friedmann PS, Djukanovic R. Invariantnatural killer T cells in asthma and chronic obstructivepulmonary disease. N Engl J Med. 2007 Apr 5;356(14):1410e22.

22. Thomas SY, Lilly CM, Luster AD. Invariant natural killer T cellsin bronchial asthma. N Engl J Med. 2006 Jun 15;354(24):2613e6.

23. Mempel M, Flageul B, Suarez F, Ronet C, Dubertret L,Kourilsky P, Gachelin G, Musette P. Comparison of the T cellpatterns in leprous and cutaneous sarcoid granulomas. Pres-ence of Valpha24-invariant natural killer T cells in T-cell-reactive leprosy together with a highly biased T cell receptorValpha repertoire. Am J Pathol 2000 Aug;157(2):509e23.

24. Kobayashi S, Kaneko Y, Seino K, Yamada Y, Motohashi S,Koike J, Sugaya K, Kuriyama T, Asano S, Tsuda T, Wakao H,Harada M, Kojo S, Nakayama T, Taniguchi M. Impaired IFN-gamma production of Valpha24 NKT cells in non-remittingsarcoidosis. Int Immunol 2004 Feb;16(2):215e22.

Invariant NKT cells in sarcoidosis 577

25. Roberts SD, Kohli LL, Wood KL, Wilkes DS, Knox KS. CD4þ CD28-T cells are expanded in sarcoidosis. Sarcoidosis Vasc DiffuseLung Dis 2005 Mar;22(1):13e9.

26. Grunewald J, Wahlstrom J, Berlin M, Wigzell H, Eklund A,Olerup O. Lung restricted T cell receptor AV2S3þ CD4þ T cellexpansions in sarcoidosis patients with a shared HLA-DRbetachain conformation. Thorax 2002 Apr;57(4):348e52.

27. Wilson SB, Kent SC, Patton KT, Orban T, Jackson RA, Exley M,Porcelli S, Schatz DA, Atkinson MA, Balk SP, Strominger JL,Hafler DA. Extreme Th1 bias of invariant Valpha24JalphaQ Tcells in type 1 diabetes. Nature 1998 Jan 8;391(6663):177e81.

28. Baughman RP, Lower EE. Treatment of sarcoidosis with corti-costeroids: who is going to relapse and why? Sarcoidosis VascDiffuse Lung Dis. 1998 Mar;15(1):19e20.

29. Katchar K, Soderstrom K, Wahlstrom J, Eklund A, Grunewald J.Characterisation of natural killer cells and CD56þ T-cells insarcoidosis patients. Eur Respir J. 2005 Jul;26(1):77e85.

30. Papakosta D, Kyriazis G, Gioulekas D, Kontakiotis T, Polyzoni T,Bouros D, Patakas D. Variations in alveolar cell populations,lymphocyte subsets and NK-cells in different stages of activepulmonary sarcoidosis. Sarcoidosis Vasc Diffuse Lung Dis. 2005Mar;22(1):21e6.

31. Akbari O, Stock P, Meyer E, Kronenberg M, Sidobre S,Nakayama T, Taniguchi M, Grusby MJ, DeKruyff RH, Umetsu DT.Essential role of NKT cells producing IL-4 and IL-13 in the

development of allergen-induced airway hyperreactivity. NatMed. 2003 May;9(5):582e8.

32. Mutalithas K, Croudace J, Guillen C, Siddiqui S, Thickett D,Wardlaw A, Lammas D, Brightling CJ. Bronchoalveolar lavageinvariant natural killer T cells are not increased in asthma.J Allergy Clin Immunol 2007 May;119(5):1274e6.

33. Matangkasombut P, Marigowda G, Ervine A, Idris L,Pichavant M, Kim HY, Yasumi T, Wilson SB, DeKruyff RH,Faul JL, Israel E, Akbari O, Umetsu DT. Natural killer T cells inthe lungs of patients with asthma. J Allergy Clin Immunol 2009May;123(5):1181e5.

34. Baev DV, Caielli S, Ronchi F, Coccia M, Facciotti F, Nichols KE,Falcone M. Impaired SLAMeSLAM homotypic interactionbetween invariant NKT cells and dendritic cells affects differ-entiation of IL-4/IL-10-secreting NKT2 cells in nonobese dia-betic mice. J Immunol. 2008 Jul 15;181(2):869e77.

35. Jordan MA, Fletcher JM, Pellicci D, Baxter AG. Slamf1, the NKTcell control gene Nkt1. J Immunol 2007;178:1618e27.

36. Cantorna MT, Yu S, Bruce D. The vitamin D receptor is requiredfor iNKT cell development. Proc Natl Acad Sci U S A 2008 Apr 1;105(13):5207e12.

37. Niimi T, Tomita H, Sato S, Kawaguchi H, Akita K, Maeda H,Sugiura Y, Ueda R. Vitamin D receptor gene polymorphism inpatients with sarcoidosis. Am J Respir Crit Care Med 1999 Oct;160(4):1107e9.