Alternative pathway for the development of Vα14+ NKT cells directly from CD4–CD8– thymocytes that bypasses the CD4+CD8+ stage

Dashtsoodol, Nyambayar; Shigeura, Tomokuni; Aihara, Minako; Ozawa, Ritsuko; Kojo, Satoshi; Harada, Michishige; Endo, Takaho A; Watanabe, Takashi; Ohara, Osamu; Taniguchi, Masaru

doi:10.1038/ni.3668

Article
Published: 30 January 2017

Alternative pathway for the development of V_α14⁺ NKT cells directly from CD4^–CD8^– thymocytes that bypasses the CD4⁺CD8⁺ stage

Nyambayar Dashtsoodol^1,2,
Tomokuni Shigeura¹,
Minako Aihara¹,
Ritsuko Ozawa¹,
Satoshi Kojo¹,
Michishige Harada¹,
Takaho A Endo ORCID: orcid.org/0000-0002-2801-2463³,
Takashi Watanabe³,
Osamu Ohara³ &
…
Masaru Taniguchi¹

Nature Immunology volume 18, pages 274–282 (2017)Cite this article

5253 Accesses
53 Citations
106 Altmetric
Metrics details

Subjects

Abstract

Although invariant V_α14⁺ natural killer T cells (NKT cells) are thought to be generated from CD4⁺CD8⁺ double-positive (DP) thymocytes, the developmental origin of CD4^–CD8^– double-negative (DN) NKT cells still remains unresolved. Here we provide definitive genetic evidence obtained, through studies of mice with DP-stage-specific ablation of expression of the gene encoding the recombinase component RAG-2 (Rag2) and by a fate-mapping approach, that supports the proposal of the existence of an alternative developmental pathway through which a fraction of DN NKT cells with strong T-helper-type-1 (T_H1)-biased and cytotoxic characteristics develop from late DN-stage thymocytes, bypassing the DP stage. These findings provide new insight into understanding of the development of NKT cells and propose a role for timing of expression of the invariant T cell antigen receptor in determining the functional properties of NKT cells.

You have full access to this article via your institution.

Download PDF

Modulation of TCR signalling components occurs prior to positive selection and lineage commitment in iNKT cells

Article Open access 08 December 2021

A double-negative thymocyte-specific enhancer augments Notch1 signaling to direct early T cell progenitor expansion, lineage restriction and β-selection

Article 31 October 2022

Generation of hypoimmunogenic T cells from genetically engineered allogeneic human induced pluripotent stem cells

Article 17 May 2021

Main

Natural killer T cells (NKT cells) are characterized by the expression of a unique T cell antigen receptor (TCR) α-chain composed of a single invariant α-chain variable region 14–α-chain joining region 18 (V_α14J_α18) in mice¹ and V_α24J_α18 in humans^2,3, both of which recognize almost identical glycolipid-binding sites contributed mainly by the proximal J_α18 region⁴. Thus, both human NKT cells and mouse NKT cells recognize the same glycolipid ligands, such as α-galactosylceramide (α-GalCer), in conjunction with the monomorphic major histocompatibility complex (MHC)-like antigen-presenting molecule CD1d^5,6 and show cross-species reactivity⁷.

The majority of NKT cells are T-helper-type-1 (T_H1)-type NKT cells, which have important protective roles in responses to tumors or pathogens due to their very rapid and powerful innate-like responses. In this context, both human CD4^–CD8^– (double-negative (DN)) NKT cells and mouse DN NKT cells, in contrast to CD4⁺ NKT cells, have been shown to elicit strong antitumor activity and T_H1-type responses^8,9,10. However, those functional differences between DN NKT cells and CD4⁺ NKT cells cannot be adequately explained by the currently accepted CD4⁺CD8⁺ (double-positive (DP)) pathway model for the development of NKT cells.

The DP pathway model is based on experiments that demonstrated the generation of NKT cells by intrathymic transfer of DP cells from Cd1d^–/– mice into wild-type recipients¹¹, as well as by fate-mapping experiments designed to trace cells that developed through the DP stage through the use of reporter mice that express Cre recombinase from the Cd4 promoter (Cd4-Cre) or the gene encoding the transcription factor RORγt (Rorc-Cre), together with a transgene with a loxP-flanked stop cassette expressed from the ubiquitous Rosa26 gene, for the expression of green fluorescent protein (GFP), in which GFP⁺ NKT cells were detected but GFP^– NKT cells were not¹². According to this DP pathway model, the DP thymic precursor cells that have rearranged an invariant V_α14J_α18 chain are positively selected on CD1d to give rise to DN NKT cells as well as CD4⁺ NKT cells^13,14. However, it has also been shown that not all DN αβ T cells originate from DP-stage cells¹⁵ and also that the timing of TCR expression affects fate 'decisions'¹⁶. Therefore, questions about the influence of the timing of NKT cell TCR expression on the differentiation and function of DN NKT cells have remained unanswered.

To clarify the developmental origin and functional attributes of DN NKT cells, we reasoned that fate-mapping or conditional-gene-ablation approaches controlled by DP-stage-specific expression of Cre recombinase from the enhancer E8_III (E8_III-Cre) would be ideally suited for testing the possibility of generating NKT cells from the DN stage, because in this model, Cre is expressed specifically in DP-stage thymocytes without affecting cells in the DN stage^17,18. Through the use of this approach, we were able to demonstrate the existence of an alternative NKT cell developmental pathway, distinct from the DP pathway, that originated at the DN stage of thymic ontogeny. This DN pathway 'preferentially' gave rise to T_H1 NKT cells with strong cytotoxic potential and distinct distribution patterns in peripheral tissues. This adds new insight to the understanding of NKT cell development and suggests that the differentiation stage as well as the microenvironmental niche of precursor cells undergoing positive selection have an important role in determining their further developmental programs.

Results

Detection of NKT cells of DN-stage origin by fate mapping

To investigate whether NKT cells can be derived from DN-stage precursors through an alternative developmental pathway different from the DP pathway, we performed fate-mapping experiments by using mice bearing a reporter transgene with a loxP-flanked stop cassette expressed from the ubiquitous Rosa26 gene, for the expression of yellow fluorescent protein (YFP), controlled by the DP-stage-specific expression of E8_III-Cre (E8_III-Cre⁺Rosa26-YFP mice). In these E8_III-Cre⁺Rosa26-YFP mice, the expression of E8_III-Cre is regulated by the E8_III enhancer and promoter elements of the gene encoding the co-receptor CD8α, which restricts its expression exclusively to DP thymocytes^17,18. Therefore, the Cre-mediated deletion of the loxP-flanked stop cassette turns on the expression of the gene encoding the YFP reporter and thus all progeny of DP-stage thymocytes are labeled with YFP.

Flow cytometry of splenic TCRβ⁺ cells from E8_III-Cre⁺Rosa26-YFP mice showed almost completely uniform expression of the YFP reporter by CD4⁺ T cells and CD8⁺ T cells (Fig. 1a), which indicated that conventional CD4⁺ T cells and CD8⁺ T cells were the progeny of DP thymocytes. Although the majority of DN TCRβ⁺ splenocytes were YFP⁺ cells of DP-stage-thymocyte origin, we observed that ∼14% of DN TCRβ⁺ cells were YFP^– (Fig. 1a), which indicated that these YFP^– TCRβ⁺ cells were likely to have differentiated from DN-stage thymocytes and to have bypassed the DP stage during thymic development. Further analysis of this population through the use of staining with α-GalCer-loaded CD1d dimers, in which unloaded CD1d dimers were used to exclude non-specific staining, revealed that ∼14% of YFP^– DN TCRβ⁺ splenocytes were NKT cells (Fig. 1b). In comparison, the frequency of NKT cells among the YFP⁺ DN TCRβ⁺ fraction, which had developed through the DP stage, was ∼25% (Fig. 1b).

Figure 1: Demonstration of the DN-stage-thymocyte origin of NKT cells by a novel fate-mapping approach using E8_III-Cre⁺Rosa26-YFP mice, in which YFP expression is controlled by the DP-stage-specific expression of E8_III-Cre.

We then investigated the presence of YFP^– DN NKT cells in the thymus of E8_III-Cre⁺Rosa26-YFP mice (Fig. 1c). The results demonstrated the presence of YFP^– TCRβ⁺ cells; ∼21% of these were mature NKT cells, as indicated by staining with α-GalCer–CD1d dimers (Fig. 1c). Although the majority of thymic DN NKT cells were of DP-stage-thymocyte origin (the absolute number of DN NKT cells was ∼2 × 10⁵ YFP⁺ cells and ∼1 × 10⁴ YFP^– cells) (data not shown), we note that YFP^– NKT cells represented a readily detectable population.

To confirm the results obtained with the E8_III-Cre⁺Rosa26-YFP mice, we generated another, similar reporter model with Cd4-Cre (the widely used gold standard for fate-mapping experiments to trace cells that went through the DP stage) and determined whether the resultant Cd4-Cre⁺Rosa26-YFP mouse harbored YFP^– NKT cells or not. Analysis of Cd4-Cre⁺Rosa26-YFP mice demonstrated the presence of YFP^– cells among DN TCRβ⁺ splenocytes (Supplementary Fig. 1a) as well as YFP^– NKT cells (Supplementary Fig. 1b). This confirmed the results obtained with the E8_III-Cre⁺Rosa26-YFP mice. Together, the results of the fate-mapping experiments demonstrated that not all NKT cells were the progeny of DP thymocytes and further suggested the possibility that a fraction of the DN NKT cells developed directly from the thymic DN stage, which precedes the DP stage of thymic differentiation.

Conditional ablation of Rag2 in DP-stage thymocytes

To investigate the development of the NKT cells of DN-stage-thymocyte origin, we next used mice with conditional ablation of loxP-flanked Rag2 alleles (Rag2^fl/fl) specifically at the DP stage. For this purpose, we introduced the E8_III-Cre transgene into Rag2^fl/fl mice to generate E8_III-Cre⁺Rag2^fl/fl mice in which the DP-stage-specific expression of Cre would excise Rag2^fl/fl in DP thymocytes while leaving Rag2^fl/fl in DN thymocytes intact.

To confirm the DP-stage-specific conditional deletion of Rag2^fl/fl, we sorted DP and DN thymocyte fractions from E8_III-Cre⁺Rag2^fl/fl mice and their Rag2^fl/fl (Cre^–) littermates (which served as a control) (Fig. 2a), and assessed Rag2 expression by real-time quantitative RT-PCR (qPCR). Rag2 transcripts were selectively lower in abundance in DP thymocytes, while Rag2 expression in DN thymocytes was largely unaffected, in E8_III-Cre⁺Rag2^fl/fl mice relative to the expression of Rag2 in the control mice (Fig. 2b). Therefore, the E8_III-Cre⁺Rag2^fl/fl mouse was indeed a faithful experimental model of DN-stage-restricted Rag2 expression. Consequently, flow cytometry of thymocytes showed a major decrease in both the frequency and absolute number of CD4⁺ or CD8⁺ T cells in E8_III-Cre⁺Rag2^fl/fl mice relative to the abundance of these cells in the control mice, whereas cells in the DP fraction were slightly greater in abundance, by ∼10%, in E8_III-Cre⁺Rag2^fl/fl mice relative to their abundance in the control mice, but with no apparent change in the DN fraction (Fig. 2b,c), which suggested that the DP-stage-specific ablation of Rag2 resulted in a developmental block at the DP stage.

**Figure 2: DN-stage-restricted expression of *Rag2* in E8_III-Cre⁺*Rag2*^fl/fl mice.**

Detection of transcripts encoding V_α14J_α18 of DN-stage origin

To determine the effect of DP stage-specific RAG deletion on the rearrangement of an invariant V_α14J_α18 TCRα chain (encoded by Trav11 and Traj18, respectively), we sorted DP and DN thymocyte fractions from E8_III-Cre⁺Rag2^fl/fl mice, gating the cells as TCRβ^neg–loCD5^loCD69^– cells (Fig. 3a). We noted that the lower abundance of TCRβ^pos–hiCD5^hiCD69^hi thymocytes in E8_III-Cre⁺Rag2^fl/fl mice than in control mice (Fig. 3a and data not shown) seemed to suggest an efficient developmental block beyond the DP stage in E8_III-Cre⁺Rag2^fl/fl mice. qPCR analysis of these sorted thymocyte fractions revealed that DP-stage-specific ablation of Rag2 resulted in loss of Trav11Traj18 transcripts in the DP fraction, whereas these transcripts were readily detected in the DN thymocyte fraction (Fig. 3b).

**Figure 3: Detection of *Trav11Traj18* transcripts and NKT cells of DN-stage origin in the thymus of E8_III-Cre⁺*Rag2*^fl/fl mice.**

As another definitive way to demonstrate ongoing gene rearrangement of Trav11Traj18 in the DN thymocyte fraction, we sorted bulk DN thymocytes from E8_III-Cre⁺Rag2^fl/fl mice and performed high-throughput next-generation sequencing of Trav11Trac transcripts (which encode V_α14 and the α-chain constant region (C_α)), regardless of the J_α-region-encoding gene being used. A total of 56,056 sequences encoding V_α14J_α chains were obtained, and among 3,002 total out-of-frame sequences, 293 (∼9.8%) were Trav11Traj18 (Table 1). Of particular note, analysis of the out-of-frame sequences encoding V_α14J_α chains revealed diverse use of J_α-encoding genes, such as Traj6, Traj9, Traj11, Traj18, Traj21, Traj26, Traj27, Traj32, Traj33, Traj34, Traj35, Traj38, Traj39, Traj42, Traj47, Traj48, Traj49, Traj53, Traj56 and Traj57 (Supplementary Fig. 2); this provided definitive evidence of ongoing random TCR α-chain rearrangement in the DN fraction of E8_III-Cre⁺Rag2^fl/fl mice. In addition, among 52,703 total productive sequences, 47,841 (∼90.8%) were in-frame Trav11Traj18 sequences (Table 1). The predominance of in-frame sequences encoding the invariant V_α14J_α18 chain (Trav11Traj18) among all the in-frame sequences encoding V_α14J_α chains would be best explained by the presence of positively selected mature NKT cells among the sorted bulk population of DN thymocytes. The presence of out-of-frame Trav11Traj18 rearrangements unambiguously demonstrated that Trav11Traj18 rearrangements were initiated, at least in part, at the DN stage.

Table 1 Trav11Trac transcripts in DN thymocytes of E8_III-Cre⁺Rag2^fl/fl

Full size table

NKT cells that developed through an alternative pathway

Next we investigated the presence of mature NKT cells positive for α-GalCer-loaded CD1d dimers in the DN TCRβ⁺ thymocyte fraction of E8_III-Cre⁺Rag2^fl/fl mice. Flow cytometry of these DN thymocytes demonstrated the presence of mature NKT cells of DN-stage origin, at a frequency of ∼67% among the gated DN TCRβ⁺ thymocytes (Fig. 3c), among which the absolute number of DN NKT cells from E8_III-Cre⁺Rag2^fl/fl mice was ∼50% of that from their control littermates (Fig. 3d). Next we analyzed use of TCR V_β chains to investigate whether NKT cells of DN-stage-thymocyte origin were normally selected. The V_β repertoire of NKT cells of DN-stage-thymocyte origin was grossly normal and similar to that of NKT cells of DP-stage-thymocyte origin (Supplementary Fig. 3).

To further verify the results obtained with E8_III-Cre⁺Rag2^fl/fl mice, we made another model of conditional ablation of Rag2 at the DP stage by generating Cd4-Cre⁺Rag2^fl/fl mice. Analysis of sorted DP and DN thymic fractions showed that Trav11Traj18 transcripts were present in the DN thymocytes but not in DP thymocytes (Supplementary Fig. 4a), which suggested that the mature NKT cells detected in Cd4-Cre⁺Rag2^fl/fl mice (Supplementary Fig. 4b,c) probably represented NKT cells of DN-stage-thymocyte origin. These data confirmed the results obtained with E8_III-Cre⁺Rag2^fl/fl mice.

To further confirm the results of Cre-recombinase-mediated approaches, such as E8_III-Cre⁺Rosa26-YFP and E8_III-Cre⁺Rag2^fl/fl mice, we used mice with sequence encoding GFP knocked into Rorc (which encodes the transcription factor RORγt; called 'Rorc^–/–' here), which is a characterized mouse model that lacks Trav11Traj18 rearrangements in DP stage thymocytes because of the shortened lifespan of DP thymocytes^19,20. qPCR analysis of DN and DP thymocyte fractions sorted from Rorc^–/– mice revealed that Trav11Traj18 transcripts were present only in the DN thymocyte fraction but not in the DP thymocyte fraction (Supplementary Fig. 5a) and, in agreement with the presence of Trav11Traj18 transcripts, analysis of the DN thymic population revealed the presence of mature CD1d-dimer⁺ TCRβ⁺ NKT cells in Rorc^–/– mice (Supplementary Fig. 5b).

To further investigate the developmental time window for the initiation of Trav11Traj18 rearrangements in DN-stage thymocytes, we used mice deficient in the pre-TCRα-chain (Ptcra^–/–), a well-known mouse model that lacks a functional β-selection process during the T cell development beyond DN stage 3 of thymic ontogeny²¹. We observed a complete lack of expression of Trav11Traj18 transcripts in both DN thymic fractions and DP thymic fractions in Ptcra^–/– mice (Supplementary Fig. 6a), accompanied by the absence of mature NKT cells within the DN compartment (Supplementary Fig. 6b). These results suggested an absolute requirement for β-selection in the initiation of Trav11Traj18 rearrangements in the DN stage, which seemed to occur beyond DN stage 3 of development, presumably at DN stage 4, as shown previously²². Collectively, the results reported above provided evidence for the existence of an alternative developmental pathway of NKT cells from DN-stage precursor cells that was different from the DP developmental pathway.

T_H1-biased function of NKT cells of DN-stage origin

To investigate the properties of the NKT cells of DN-stage-thymocyte origin, we sorted DN thymic NKT cells from E8_III-Cre⁺Rag2^fl/fl mice and stimulated the cells for 48 h with antibody to the TCR invariant chain CD3 (anti-CD3) and antibody to the co-receptor CD28 (anti-CD28), then analyzed their ex vivo cytokine-secretion profiles. For reference purposes, we used the well-characterized wild-type C57BL/6 thymic NKT cells sorted on the basis of differences the stages in their expression of the marker NK1.1, whereby the NK1.1^– fraction consisted of CD44^lo ('stage 1') and CD44^hi ('stage 2') NKT cells, and the NK1.1⁺ fraction consisted of CD44^hiNK1.1⁺ ('stage 3') NKT cells.

The results showed that NKT cells from E8_III-Cre⁺Rag2^fl/fl mice produced substantial amounts of interferon-γ (IFN-γ) and secreted a little interleukin 4 (IL-4), but no IL-17A or IL-10, a cytokine profile that mirrored that of wild-type NK1.1⁺ stage 3 NKT cells (Fig. 4a). As expected, the NK1.1^– stage 1 and stage 2 NKT cells from wild-type mice had a robust capacity to secrete IL-4 and IL-17 after being activated (Fig. 4a). Of particular note, we observed that only wild-type NK1.1^– NKT cells secreted IL-10 after stimulation with anti-CD3 and anti-CD28 (Fig. 4a). These results demonstrated that the mature NKT cells of DN-stage-thymocyte origin closely resembled wild-type NK1.1⁺ stage 3 NKT cells in terms of their cytokine-secretion profiles, as well as in their surface CD24^loCD44^hiNK1.1⁺ phenotype (Supplementary Fig. 7).

**Figure 4: Cytokine-secretion patterns and gene-expression profiles of NKT cells of DN-stage origin from E8_III-Cre⁺*Rag2*^fl/fl mice.**

T_H1 gene-expression profile of NKT cells of DN-stage origin

To investigate the molecular basis of the T_H1-cell-biased cytokine production by NKT cells of DN-stage-thymocyte origin, we performed next-generation sequencing (RNA-seq) analysis of NKT cells from E8_III-Cre⁺Rag2^fl/fl mice ('NKT cells of DN-stage origin'), with wild-type C57BL/6 thymic NKT cell subsets sorted on the basis of differences among in the stages in their expression of the markers NK1.1 and CD44. Clustering analysis of genes with differences in expression by NKT cells in the three stages showed that the NKT cells of DN-stage-thymocyte origin were closer to wild-type stage 3 NKT cells (with a correlation coefficient of 0.986 at the total transcriptome level) than to wild-type stage 1 or stage 2 NKT cells(for which the correlation coefficients were 0.856 and 0.875, respectively) (Fig. 4b).

Moreover, a correlation between the NKT cells of DN-stage origin and wild-type stage 3 NKT cells was also observed when genes encoding products related to the T_H1 response were analyzed (Fig. 4c). For example, key T_H1-cell-related genes, such as Ifng, Gzmb, Gzma and Prf1, the transcription-factor-encoding genes Stat4, Tbx21 and Runx3, and the surface-molecule-encoding genes Fasl, Cd40lg, Infgr1, Slamf7, Il18rap, Il12rb2, Il7r, Il2rb, Nkg7 and Klrb1c, were upregulated in both the NKT cells of DN-stage origin and wild-type stage 3 NKT cells relative to their expression in stage 1 or stage 2 NKT cells, while expression of Id3, which encodes the T_H2-cell-related transcription factor Id3, was lower in the NKT cells of DN-stage origin and wild-type stage 3 NKT cells than in stage 1 or stage 2 NKT cells (Fig. 4c). Together these results demonstrated that NKT cells of DN-stage origin expressed key genes characteristic of T_H1 cells and cytotoxic effector cells in a fashion very similar to that of wild-type stage 3 NKT cells.

Distinct tissue distribution of NKT cells of DN-stage origin

Next we investigated the peripheral distribution of DN NKT cells of DN-stage origin from E8_III-Cre⁺Rag2^fl/fl mice, compared with that of NKT cells of DP-stage origin (i.e., YFP⁺ DN NKT cells from E8_III-Cre⁺Rosa26-YFP mice). For analysis of the DN NKT cell fraction in the spleen, we used a negative-selection approach via depletion of cells expressing the surface antigens CD4, CD8, CD19 and TER119. The results of subsequent flow cytometry revealed that while the DN NKT cells of DP-stage origin were readily detectable in the spleen, liver, lungs, mesenteric lymph nodes, gut lamina propria and visceral adipose tissue, the DN NKT cells of DN-stage origin were detected mainly in the liver, spleen and lungs but not in the mesenteric lymph nodes, lamina propria or adipose tissue (Fig. 5a,b). When we compared the absolute number of cells in these peripheral locations, we found that DN NKT cells of DP-stage origin were mainly in the spleen, while those of DN-stage origin were mainly in the liver (Fig. 5b), which suggested distinct tissue-distribution patterns for DN NKT cells of DN-stage origin and those of DP-stage origin.

**Figure 5: Peripheral tissue distribution of NKT cells of DN-stage origin from E8_III-Cre⁺*Rag2*^fl/fl mice and NKT cells of DP-stage origin from E8_III-Cre⁺Rosa26-YFP mice.**

In an attempt to explain the enhanced liver-homing capacity of NKT cells of DN-stage origin relative to their poor homing into peripheral lymphoid locations, we performed RNA-seq analysis of genes encoding lymphocyte-homing-related factors, such as chemokine receptors and integrins expressed on thymic and liver NKT cells of DN- or DP-stage origin. Transcriptome analysis revealed that liver NKT cells of DN-stage origin had high expression of Ccr5 and Cxcr3, which encode well-known liver-homing factors²³ (Fig. 5c). We also observed that liver NKT cells of DN-stage origin, in contrast to those of DP-stage origin, had high expression of the integrin-encoding genes Itga1 and Itga4 (Fig. 5c); the latter integrin is reported to have a role in the hepatic localization of lymphocytes via interaction with the adhesion molecule VCAM-1 (ref. 23). Of note, our results demonstrated that NKT cells of DN-stage origin, in contrast to those of DP-stage origin, had high expression of S1pr1 and S1pr5, which encode receptors for sphingosine 1-phosphate (Fig. 5c,d); the former receptor is involved in the egress of lymphocytes from the thymus and secondary lymphoid organs²⁴ and might have a role in homing to the liver as a result of differences in the gradient of sphingosine 1-phosphate, while the latter is required for the trafficking of NK cells²⁵.

In contrast to the results obtained for the expression of genes encoding liver-homing factors, NKT cells of DN-stage origin had lower expression of the chemokine-receptor-encoding genes Ccr6, Ccr7 and Ccr9 than that of their counterparts of DP-stage origin (Fig. 5e); these chemokine receptors are reported to have important roles in the trafficking of lymphocytes into lymph nodes and peripheral sites²⁴. Together these observations could provide some explanation of not only the greater liver-homing capacity of NKT cells of DN-stage origin but also to their inferior ability to home to peripheral lymphoid organs.

Superior cytotoxicity of NKT cells of DN-stage origin

Because liver NKT cells with a CD4^– phenotype are reported to elicit antitumor activity superior to that elicited by CD4⁺ NKT cells¹⁰, and also because of the substantial presence of NKT cells of DN-stage origin in the liver compartment, we compared the cytotoxic function of DN NKT cells of DN-stage origin with that of DN NKT cells of DP-stage origin. We first analyzed the expression of T_H1-cell-related genes and genes encoding cytotoxicity-related products in steady-state ex vivo cell samples; we compared DN NKT cells of DN-stage origin (from E8_III-Cre⁺Rag2^fl/fl mice) with those of DP-stage origin (i.e., YFP⁺ DN NKT cells from E8_III-Cre⁺Rosa26-YFP mice). qPCR analysis results showed higher expression of T_H1-cell-related genes, such as Ifng, Tbx21, Il12rb2, Cd40lg and Runx3, and of genes encoding cytotoxicity-related factors, such as Gzmb (which encodes granzyme B), Prf1 (which encodes perforin) and Fasl (which encodes the ligand for the death receptor Fas), in NKT cells of DN-stage origin than in their counterparts of DP-stage origin (Fig. 6a).

**Figure 6: Cytotoxic-lineage characteristics of liver NKT cells of DN-stage origin.**

Consistent with their gene-expression profiles, liver NKT cells of DN-stage origin that were activated with anti-CD3 and anti-CD28 showed greater cytotoxic activity in vitro against YAC-1 target cells (mouse lymphoma cells) than that of their counterparts of DP-stage origin (Fig. 6b). To further confirm the results obtained with E8_III-Cre⁺Rag2^fl/fl mice, we also assessed the cytotoxic activity of YFP^– thymic NKT cells of DN-stage origin and YFP⁺ thymic NKT cells of DP-stage origin (purified from E8_III-Cre⁺Rosa26-YFP mice). The results showed that the cytotoxic activity of YFP^– NKT cells seemed to be stronger than that of YFP⁺ NKT cells (Supplementary Fig. 8). Together these results demonstrated that the DN pathway 'preferentially' generated NKT cells with superior cytotoxic effector function.

Discussion

Here we identified a previously unknown NKT cell developmental pathway that originates from DN thymocytes and bypasses the DP stage. We have designated this the 'DN pathway', a pathway distinct from the currently accepted DP pathway model, which states that all NKT cells are generated at the DP stage during thymic development⁵. The DP pathway model is based mainly on the results of two types of experiments: intrathymic transfer of Cd1d^–/– DP thymocytes¹¹ or fate-mapping approach using Cd4-Cre or Rorc-Cre¹². However, despite the insights gained from such studies, there are several important points that need to be considered. First, the developmental origin of DN NKT cells cannot be formally addressed by the simple transfer of DP thymocytes alone. Second, it has been shown that Rorc expression^26,27 as well as Cd4-Cre activity^28,29,30 are detected as early as the thymic DN stage 3. Therefore, these fate-mapping experiments could have resulted in unanticipated labeling of the progeny of DN stage thymocytes, in addition to those of the DP stage.

To circumvent those problems, we chose to use E8_III-Cre for our fate mapping and for conditional ablation of Rag2. The experiments using E8_III-Cre⁺Rosa26-YFP mice and E8_III-Cre⁺Rag2^fl/fl mice unambiguously demonstrated that a fraction of DN NKT cells were able to develop directly from DN-stage thymocytes without entering into the DP stage. This suggested a scenario in which DN-stage precursor cells harboring rearranged Trav11 and Traj18 generated by random rearrangement events are positively selected on CD1d and develop into mature DN NKT cells.

TCR-sequencing experiments revealed the presence of diverse out-of-frame sequences encoding V_α14J_α chains, including out-of-frame Trav11Traj18 sequences, in the DN population of E8_III-Cre⁺Rag2^fl/fl mice, which provided compelling evidence of ongoing stochastic TCR α-chain rearrangements at the DN stage. Of note, we observed 'preferential' rearrangements of Trav11 in the DN thymocyte fraction with particular genes encoding J_α chains, such as Traj6, Traj18, Traj26, Traj32 and Traj57, on the basis of the frequency of use of J_α-encoding genes, whose patterns were quite similar to those observed in DP thymocytes³¹. Since about 10% (293 of 3,002) of Trav11-containing out-of-frame sequences were Trav11Traj18, these results indicated that the tertiary DNA structure between the Trav11 locus and loci encoding J_α during gene rearrangement events might contribute to form the Trav11Traj18 rearrangement at unexpectedly high frequency, as suggested previously³².

Despite the fact that both DN pathways and DP pathways contributed to the generation of DN NKT cells, the former pathway 'preferentially' gave rise to IFN-γ-producing T_H1 NKT cells with augmented cytotoxic function, whereby NKT cells of DN-stage origin demonstrated much stronger cytotoxicity against YAC-1 target cells and had higher expression of genes encoding cytotoxicity-related products, such as granzyme B, perforin and the transcription factor Runx3, than that of their counterparts of DP-stage origin. Such 'preferential' development of T_H1-type cells is probably a general attribute of unconventional T cells that develop as a result of early TCR rearrangements, as shown previously³³. A potential mechanism for the 'preferential' development of T_H1-biased NKT cells might be related to differences in the timing of cytokine–cytokine receptor interactions during positive selection mediated by the TCR–CD1d interaction. Consistent with that, it is known that the timing of TCR α-chain expression is critically important for the further functional fate determination of T cells¹⁶.

In addition, a published study has elegantly shown, through the use of MHC-null mice, that the developmental pathway for cytotoxic CD8⁺ T cells is not dependent on positive selection by cognate TCR–MHC interactions itself. Instead, signals emanating from the IL-7 receptor (IL-7R) can bypass TCR–MHC–mediated signaling and result in the differentiation of cytotoxic CD8⁺ T cells¹⁷. In this context, it has been reported that DN-stage thymocytes normally express IL-7R, downregulate its expression after differentiating into the DP stage and then re-express it as post-selection αβ T cells¹⁸. It has also been shown that IL-7R determines the fate of cytotoxic effector cells via induction of Runx3, which upregulates genes associated with cytotoxic cell lineages¹⁷. Although Runx3 acts synergistically with the transcription factor T-bet to enhance the transcription of T_H1-cell-related genes³⁴, it seems that T-bet, in contrast to Runx3, is induced mainly by TCR signaling. Concordantly, our gene-expression-profiling experiments revealed that the NKT cells of DN-stage origin had higher expression of genes encoding IL-7R (Il7r) and its downstream associated genes characteristic of cytotoxic cell lineages, such as Runx3, Gzmb, Prf1, Tbx21 and Il12rb2, than that of the NKT cells of DP-stage origin. Therefore, lineage commitment at the DN stage might have an important role in determining the functional characteristics of NKT cells.

In addition to their functional bias, NKT cells of DN-stage origin had a peripheral distribution pattern different from that of NKT cells of DP-stage origin. NKT cells of DN-stage origin were present mainly in the liver, a result that could be explained in part by their higher expression of genes encoding liver-homing factors than that of NKT cells of DP-stage origin. In contrast, the NKT cells of DP-stage origin were present mainly in the spleen but also in mesenteric lymph nodes, lamina propria and adipose tissues, and they had high expression of genes encoding the homing factors responsible for peripheral localization, such as CCR6, CCR7 and CCR9. These results raise questions about whether NKT cells generated by distinct pathways have specific physiological functions.

As for the biological importance of the DN pathway, we speculate that NKT cells of DN-stage origin, with their potent T_H1-cell-skewed functions, might mediate a potent adjuvant activity that is essential for the clonal expansion of effector cells that mediate effective protection against pathogens^35,36,37 and thus is important for survival of the species. In fact, when we investigated the rate of nonsynonymous mutations versus that of synonymous mutations in Trav11, relative to that of other genes encoding V_α regions (such as Trav1), in different species, we observed positive selection only in Trav11 (not in genes encoding other Vа regions) at a particular species divergence, such as Apodemus speciosus to Mus species or Rattus to Mesocricetus auratus. This would suggest a phylogenic importance for NKT cells in host protection³⁸.

In conclusion, we have identified an alternative DN pathway of NKT cell development that bypasses the DP pathway. We propose that the acquisition of diverse functional characteristics by NKT cells might be dependent on the timing of TCR expression as well as on the cytokine receptor signaling in precursor cells undergoing positive selection. We also propose that the developmental environments or specific niches provide an appropriate cytokine milieu required for the differentiation of these cells.

Methods

Mice.

C57BL/6J mice were purchased from Charles River Laboratories or CLEA Japan. Rag2^fl/fl mice³⁹ were from The Jackson Laboratory. Ptcra^–/–, Rorc^–/–, Rosa26-YFP, Cd4-Cre and E8_III-Cre mice (all of C57BL/6 background) have been described previously^12,17,21,40. All mice were kept under specific-pathogen-free conditions and were used at 8–16 weeks of age unless indicated otherwise. All experiments were in accordance with protocols approved by the RIKEN Animal Care and Use Committee.

Cell preparation.

Single cells from designated organs were prepared as reported^41,42. Anti–mouse CD4 and CD8 microbeads and AutoMACS were used according to the manufacturer's instructions (Miltenyi Biotec) to prepare DN thymocytes and splenocytes. Where indicated, anti–mouse CD19 and TER119 microbeads (Miltenyi Biotec) were also used in addition to anti–mouse CD4 and CD8 microbeads for negative selection by magnetic-activated cell separation.

Flow cytometry and cell sorting.

for flow cytometry, monoclonal antibodies (mAbs) to the following were conjugated to FITC, PE, PerCP-Cy5.5, PE-Cy7, APC, APC-eFluor780 and BV421 (all mAbs here used at a concentration of 1 μl per 1 million cells): CD24 (BioLegend, clone M1/69), CD25 (BD Biosciences, clone PC61), CD44 (eBioscience, clone IM7), CD117 (eBioscience, clone 2B8), CD45R/B220 (BioLegend, clone RA3-6B2), NK1.1 (BD Biosciences, clone PK136) and TCRβ (BD Biosciences, clone H57-597), as well as α-galactosylceramide-loaded CD1d dimer (BD Biosciences). The panel of FITC-labeled mAbs to mouse TCR V_β was purchased from BD Biosciences (catalog number 557004; 20 μl used per test, as recommended by the manufacturer). Cell staining was performed after blockade with anti-FcR (BD Bioscience, clone 2.4G2). Forward light-scatter gating and staining with the membrane-impermeable DNA-intercalating dye 7-AAD were used to gate out doublets and dead cells. Samples were analyzed on a FACSCanto II or FACSAria III (BD Biosciences), and data were analyzed with FlowJo (Tree Star). Cells were sorted using a FACSAria III with post-sort purities above 99%.

Cytokine measurement.

NKT cells (2 × 10⁴) sorted by flow cytometry were stimulated with plate-bound mAb to CD3 (BD Biosciences, clone 145-2C11) and soluble mAb to CD28 (BD Biosciences, clone 37.51) at 10 μg/ml. Supernatants were collected after 48 h of incubation. IFN-γ, IL-4 and IL-17A were measured by a LEGENDPlex-bead-based immunoassay (BioLegend) according to the manufacturer's instructions. Data were acquired using a FACSCanto II (BD Biosciences) and were analyzed with LEGENDPlex software (BioLegend).

Quantitative real-time PCR.

RNA was prepared from sorted cells using RNeasy Plus Micro or RNeasy Plus Mini Kit (Qiagen) with a gDNA eliminator column treatment step. cDNA was synthesized using Superscript VILO cDNA Synthesis Kit (Life Technologies). Quantitative real-time PCR (qPCR) was performed using the ABI PRISM 7900HT system (Applied Biosystems) with a FastStart Universal Probe Master mix (Roche), or with the LightCycler 480 (Roche Applied Science) with a LightCycler 480 Probes Master mix (Roche). Relative gene expression was calculated with the 2^–ΔΔCt method. The detection of Trav11Traj18 transcripts was performed as described previously¹¹. The following primer pairs and probes were used for Universal Probe Library (UPL)- or TaqMan-probe-based quantitative PCR assays: Rag2 (forward) 5′-TGCCAAAATAAGAAAGAGTATTTCAC-3′, (reverse) 5′-GGGACATTTTTGATTGTGAATAGG-3′, UPL probe 4; Traj18Trac (forward) 5′-GGACTCAGCTGATTGTCATACCT-3′, (reverse) 5′-AGACCGAGGATCTTTTAACTGGT-3′, UPL probe 62; Gzmb (forward) 5′-GATGAAGATCCTCCTGCTACTGC-3′, (reverse) 5′-GTAAGGCCATGTAGGGTCGAGA-3′, probe (FAM-TAMRA) 5′-CCCGATGATCTCCCCTGCCTTTGTCCT-3′; Prf1 (forward) 5′-TGAGAAGACCTATCAGGACCAGTA-3′, (reverse) 5′-AGTCAAGGTGGAGTGGAGGTT-3′, probe (FAM-TAMRA) 5′-ACCAGGCGAAAACTGTACATGCGACACTC-3′; Fasl (forward) 5′-TCACCAACCAAAGCCTTAAAGTATC-3′, (reverse) 5′-AGAGGGATGGACCTTGAGTGG-3′, probe (FAM-TAMRA) 5′-AAATGGGCCACACTCCTCGGCTCTTT-3′; Cd40lg (forward) 5′-GCACACGTTGTAAGCGAAGC-3′, (reverse) 5′-CCCATTTTCAAGCATTACCAAGTTG-3′, probe (FAM-TAMRA) 5′-TGCAGCATCCGTTCTACAGTGGGCCA-3′. The following gene-expression assays (all from Applied Biosystems) were used: Ifng (Mm01168134_m1), Tbx21 (Mm00450960_m1), Runx3 (Mm00490666_m1), Il12rb2 (Mm00434200_m1) and Gapdh (Mm99999915_g1).

Next-generation sequencing (RNA-seq).

RNA was prepared with the RNeasy Plus Micro Kit (Qiagen). RNA-seq library samples were prepared using a TruSeq DNA sample Prep Kit v2 (Illumina) or ThruPLEX DNA-seq Kit (Rubicon Genomics) and a SMARTer Ultra Low RNA Kit for Illumina (Clontech). The sequencing was carried out using HiSeq1500 equipment (Illumina). The sequence reads were mapped to mouse genome (NCBI version 37) using TopHat2 version 2.0.8 and botwie2 version 2.1.0 with default parameters, and gene annotation was provided by NCBI. The transcript abundances were estimated using Cufflinks (version 2.1.1). Cufflinks was run with the same reference annotation with TopHat2 to generate FPKM (fragments per kilobase per million mapped reads) values for known gene models.

TCR sequencing.

DN thymocytes obtained from E8_III-Cre⁺Rag2^fl/fl mice and sorted by flow cytometry were lysed and RNA was extracted using an RNeasy kit (Qiagen). cDNA was prepared with Superscript First-Strand SuperMix (Life Technologies). PCR to amplify the Trav11Trac transcripts was done as described previously⁴³. PCR products were purified and sequenced on a MiSeq system with a MiSeq Reagent Kit v3-600 (Illumina). Mouse genes encoding J_α regions were analyzed using IMGT/HighV-QUEST from the IMGT (international ImMunoGeneTics information system) database⁴⁴.

Cytotoxicity assay.

Cytotoxic killing of YAC-1 target cells was performed with purified NKT cells (2 × 10⁴), which were activated overnight with anti-CD3 (BD Biosciences, clone 145-2C11) and anti-28 (BD Biosciences, clone 37.51) (10 μg/ml each) in complete RPMI-1640 medium supplemented with mouse IL-2 (10 ng/ml, R&D). Target cells before incubation with effector cells were labeled with CellTrace Violet (Thermo Fisher Scientific) and, after 4 h of co-culture, were analyzed using the Annexin V staining kit according to the manufacturer′s instructions (BioLegend). The percentage of specific cytotoxicity was calculated by the following formula: % cytotoxicity = (target-spontaneous / maximum-spontaneous) × 100.

Statistical analysis.

Two-tailed, unpaired t-tests were done using PRISM 6 software (GraphPad). P < 0.05 was considered statistically significant.

Data availability.

Raw and processed RNA-seq data that support the findings of this study have been deposited in the NCBI Gene Expression Omnibus (GEO) database with the accession codes GSE53150 and GSE89532.

Accession codes

Primary accessions

Gene Expression Omnibus

References

Taniguchi, M., Harada, M., Kojo, S., Nakayama, T. & Wakao, H. The regulatory role of Vα14 NKT cells in innate and acquired immune response. Annu. Rev. Immunol. 21, 483–513 (2003).
Article CAS PubMed Google Scholar
Lantz, O. & Bendelac, A. An invariant T cell receptor α chain is used by a unique subset of major histocompatibility complex class I-specific CD4⁺ and CD4^–8^– T cells in mice and humans. J. Exp. Med. 180, 1097–1106 (1994).
Article CAS PubMed Google Scholar
Dellabona, P., Padovan, E., Casorati, G., Brockhaus, M. & Lanzavecchia, A. An invariant Vα24-αQ/Vβ11 T cell receptor is expressed in all individuals by clonally expanded CD4-8- T cells. J. Exp. Med. 180, 1171–1176 (1994).
Article CAS PubMed Google Scholar
Borg, N.A. et al. CD1d-lipid-antigen recognition by the semi-invariant NKT T-cell receptor. Nature 448, 44–49 (2007).
Article CAS PubMed Google Scholar
Bendelac, A., Savage, P.B. & Teyton, L. The biology of NKT cells. Annu. Rev. Immunol. 25, 297–336 (2007).
Article CAS PubMed Google Scholar
Salio, M., Silk, J.D., Jones, E.Y. & Cerundolo, V. Biology of CD1- and MR1-restricted T cells. Annu. Rev. Immunol. 32, 323–366 (2014).
Article CAS PubMed Google Scholar
Brossay, L. et al. CD1d-mediated recognition of an α-galactosylceramide by natural killer T cells is highly conserved through mammalian evolution. J. Exp. Med. 188, 1521–1528 (1998).
Article CAS PubMed PubMed Central Google Scholar
Lee, P.T., Benlagha, K., Teyton, L. & Bendelac, A. Distinct functional lineages of human Vα24 natural killer T cells. J. Exp. Med. 195, 637–641 (2002).
Article CAS PubMed PubMed Central Google Scholar
Gumperz, J.E., Miyake, S., Yamamura, T. & Brenner, M.B. Functionally distinct subsets of CD1d-restricted natural killer T cells revealed by CD1d tetramer staining. J. Exp. Med. 195, 625–636 (2002).
Article CAS PubMed PubMed Central Google Scholar
Crowe, N.Y. et al. Differential antitumor immunity mediated by NKT cell subsets in vivo. J. Exp. Med. 202, 1279–1288 (2005).
Article CAS PubMed PubMed Central Google Scholar
Gapin, L., Matsuda, J.L., Surh, C.D. & Kronenberg, M. NKT cells derive from double-positive thymocytes that are positively selected by CD1d. Nat. Immunol. 2, 971–978 (2001).
Article CAS PubMed Google Scholar
Egawa, T. et al. Genetic evidence supporting selection of the Vα14i NKT cell lineage from double-positive thymocyte precursors. Immunity 22, 705–716 (2005).
Article CAS PubMed Google Scholar
Pellicci, D.G. et al. A natural killer T (NKT) cell developmental pathway involving a thymus-dependent NK1.1^–CD4⁺ CD1d-dependent precursor stage. J. Exp. Med. 195, 835–844 (2002).
Article CAS PubMed PubMed Central Google Scholar
Benlagha, K., Wei, D.G., Veiga, J., Teyton, L. & Bendelac, A. Characterization of the early stages of thymic NKT cell development. J. Exp. Med. 202, 485–492 (2005).
Article CAS PubMed PubMed Central Google Scholar
Aifantis, I. et al. The E delta enhancer controls the generation of CD4^–CD8^– αβTCR-expressing T cells that can give rise to different lineages of αβ T cells. J. Exp. Med. 203, 1543–1550 (2006).
Article CAS PubMed PubMed Central Google Scholar
Baldwin, T.A., Sandau, M.M., Jameson, S.C. & Hogquist, K.A. The timing of TCR α expression critically influences T cell development and selection. J. Exp. Med. 202, 111–121 (2005).
Article CAS PubMed PubMed Central Google Scholar
Park, J.H. et al. Signaling by intrathymic cytokines, not T cell antigen receptors, specifies CD8 lineage choice and promotes the differentiation of cytotoxic-lineage T cells. Nat. Immunol. 11, 257–264 (2010).
Article CAS PubMed PubMed Central Google Scholar
McCaughtry, T.M. et al. Conditional deletion of cytokine receptor chains reveals that IL-7 and IL-15 specify CD8 cytotoxic lineage fate in the thymus. J. Exp. Med. 209, 2263–2276 (2012).
CAS PubMed PubMed Central Google Scholar
Sun, Z. et al. Requirement for RORγ in thymocyte survival and lymphoid organ development. Science 288, 2369–2373 (2000).
Article CAS PubMed Google Scholar
Guo, J. et al. Regulation of the TCRα repertoire by the survival window of CD4⁺CD8⁺ thymocytes. Nat. Immunol. 3, 469–476 (2002).
Article CAS PubMed Google Scholar
Fehling, H.J., Krotkova, A., Saint-Ruf, C. & von Boehmer, H. Crucial role of the pre-T-cell receptor α gene in development of αβ but not γδ T cells. Nature 375, 795–798 (1995).
Article CAS PubMed Google Scholar
Dashtsoodol, N., Watarai, H., Sakata, S. & Taniguchi, M. Identification of CD4^–CD8^– double-negative natural killer T cell precursors in the thymus. PLoS One 3, e3688 (2008).
Article CAS PubMed PubMed Central Google Scholar
Crispe, I.N. Migration of lymphocytes into hepatic sinusoids. J. Hepatol. 57, 218–220 (2012).
Article PubMed Google Scholar
Cyster, J.G. Chemokines, sphingosine-1-phosphate, and cell migration in secondary lymphoid organs. Annu. Rev. Immunol. 23, 127–159 (2005).
Article CAS PubMed Google Scholar
Jenne, C.N. et al. T-bet-dependent S1P5 expression in NK cells promotes egress from lymph nodes and bone marrow. J. Exp. Med. 206, 2469–2481 (2009).
Article CAS PubMed PubMed Central Google Scholar
Villey, I., de Chasseval, R. & de Villartay, J.P. RORγT, a thymus-specific isoform of the orphan nuclear receptor RORγ / TOR, is up-regulated by signaling through the pre-T cell receptor and binds to the TEA promoter. Eur. J. Immunol. 29, 4072–4080 (1999).
Article CAS PubMed Google Scholar
He, Y.W., Deftos, M.L., Ojala, E.W. & Bevan, M.J. RORγt, a novel isoform of an orphan receptor, negatively regulates Fas ligand expression and IL-2 production in T cells. Immunity 9, 797–806 (1998).
Article CAS PubMed PubMed Central Google Scholar
Wolfer, A. et al. Inactivation of Notch 1 in immature thymocytes does not perturb CD4 or CD8 T cell development. Nat. Immunol. 2, 235–241 (2001).
Article CAS PubMed Google Scholar
Laky, K. & Fowlkes, B.J. Presenilins regulate αβ T cell development by modulating TCR signaling. J. Exp. Med. 204, 2115–2129 (2007).
Article CAS PubMed PubMed Central Google Scholar
Kastner, P. et al. Bcl11b represses a mature T-cell gene expression program in immature CD4⁺CD8⁺ thymocytes. Eur. J. Immunol. 40, 2143–2154 (2010).
Article CAS PubMed PubMed Central Google Scholar
Bedel, R. et al. Lower TCR repertoire diversity in Traj18-deficient mice. Nat. Immunol. 13, 705–706 (2012).
Article CAS PubMed PubMed Central Google Scholar
Genolet, R., Stevenson, B.J., Farinelli, L., Osterås, M. & Luescher, I.F. Highly diverse TCRα chain repertoire of pre-immune CD8⁺ T cells reveals new insights in gene recombination. EMBO J. 31, 1666–1678 (2012).
Article CAS PubMed PubMed Central Google Scholar
Hendricks, D.W. & Fink, P.J. Uneven colonization of the lymphoid periphery by T cells that undergo early TCRα rearrangements. J. Immunol. 182, 4267–4274 (2009).
Article CAS PubMed Google Scholar
Djuretic, I.M. et al. Transcription factors T-bet and Runx3 cooperate to activate Ifng and silence Il4 in T helper type 1 cells. Nat. Immunol. 8, 145–153 (2007).
Article CAS PubMed Google Scholar
Brigl, M., Bry, L., Kent, S.C., Gumperz, J.E. & Brenner, M.B. Mechanism of CD1d-restricted natural killer T cell activation during microbial infection. Nat. Immunol. 4, 1230–1237 (2003).
Article CAS PubMed Google Scholar
Kinjo, Y. et al. Recognition of bacterial glycosphingolipids by natural killer T cells. Nature 434, 520–525 (2005).
Article CAS PubMed Google Scholar
Shimizu, K. et al. Invariant NKT cells induce plasmacytoid dendritic cell (DC) cross-talk with conventional DCs for efficient memory CD8+ T cell induction. J. Immunol. 190, 5609–5619 (2013).
Article CAS PubMed Google Scholar
Kanno, R. et al. Evolution of Vα14 TCR gene family in mice. in Genetics in Wild Mice (eds. Moriwaki, K., Shiroishi, T. & Yonekawa, H.), 193–202 (Japan Scientific Societies Press, Karger, Tokyo, Basel, 1994).
Hao, Z. & Rajewsky, K. Homeostasis of peripheral B cells in the absence of B cell influx from the bone marrow. J. Exp. Med. 194, 1151–1164 (2001).
Article CAS PubMed PubMed Central Google Scholar
Srinivas, S. et al. Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev. Biol. 1, 4 (2001).
Article CAS PubMed PubMed Central Google Scholar
Watarai, H., Nakagawa, R., Omori-Miyake, M., Dashtsoodol, N. & Taniguchi, M. Methods for detection, isolation and culture of mouse and human invariant NKT cells. Nat. Protoc. 3, 70–78 (2008).
Article CAS PubMed Google Scholar
Moro, K., Ealey, K.N., Kabata, H. & Koyasu, S. Isolation and analysis of group 2 innate lymphoid cells in mice. Nat. Protoc. 10, 792–806 (2015).
Article CAS PubMed Google Scholar
Dashtsoodol, N. et al. Generation of novel Traj18-deficient mice lacking Vα14 natural killer T cells with an undisturbed T cell receptor α-chain repertoire. PLoS One 11, e0153347 (2016).
Article CAS PubMed PubMed Central Google Scholar
Lefranc, M.P. IMGT, the International ImMunoGeneTics Information System for Immunoinformatics: methods for querying IMGT databases, tools, and web resources in the context of immunoinformatics. Mol. Biotechnol. 40, 101–111 (2008).
Article CAS PubMed Google Scholar

Download references

Acknowledgements

We thank A. Singer (US National Institutes of Health) for E8_III-Cre mice; F. Constantini (Columbia University) for Rosa26-YFP mice; D. Littman (New York University) for Rorc^–/– mice; P. Burrows for critical reading of this manuscript; T. Hasegawa for support with animal facility services; and N. Takeuchi for secretarial assistance. Supported by the Japan Society for the Promotion of Science (KAKENHI grants 24790490 to N.D. and 23229005 to M.T.).

Author information

Authors and Affiliations

Laboratory for Immune Regulation, RIKEN Center for Integrative Medical Sciences, Yokohama, Japan
Nyambayar Dashtsoodol, Tomokuni Shigeura, Minako Aihara, Ritsuko Ozawa, Satoshi Kojo, Michishige Harada & Masaru Taniguchi
Core Research Laboratory, Mongolian National University of Medical Sciences, Ulaanbaatar, Mongolia
Nyambayar Dashtsoodol
Laboratory for Integrative Genomics, RIKEN Center for Integrative Medical Sciences, Yokohama, Japan
Takaho A Endo, Takashi Watanabe & Osamu Ohara

Authors

Nyambayar Dashtsoodol
View author publications
Search author on:PubMed Google Scholar
Tomokuni Shigeura
View author publications
Search author on:PubMed Google Scholar
Minako Aihara
View author publications
Search author on:PubMed Google Scholar
Ritsuko Ozawa
View author publications
Search author on:PubMed Google Scholar
Satoshi Kojo
View author publications
Search author on:PubMed Google Scholar
Michishige Harada
View author publications
Search author on:PubMed Google Scholar
Takaho A Endo
View author publications
Search author on:PubMed Google Scholar
Takashi Watanabe
View author publications
Search author on:PubMed Google Scholar
Osamu Ohara
View author publications
Search author on:PubMed Google Scholar
Masaru Taniguchi
View author publications
Search author on:PubMed Google Scholar

Contributions

N.D. conceived of the study, performed experiments, analyzed data and wrote the paper; T.S., M.A. and R.O. assisted with experiments; S.K. and M.H. discussed experiments; T.A.E. did bioinformatics analyses; T.W. and O.O. performed next-generation sequencing experiments; M.T. directed the study and wrote the paper.

Corresponding authors

Correspondence to Nyambayar Dashtsoodol or Masaru Taniguchi.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Integrated supplementary information

Supplementary Figure 1 Confirmation of the DN-stage origin of NKT cells with the Cd4-Cre⁺Rosa26-YFP mouse.

(a) YFP reporter expression within gated viable B220^-TCRβ⁺ splenic CD4⁺, CD8⁺, and DN fractions obtained from Cd4-Cre⁺Rosa26-YFP mice. Numbers on flow cytometry plots show the frequency of cells within indicated gates. (b) Detection of CD1d dimer⁺ NKT cells within gated YFP-negative and YFP-positive TCRβ⁺ DN splenocytes from Cd4-Cre⁺Rosa26-YFP mice. Prior to the cell staining, DN splenocytes were purified by AutoMACS using anti-CD4 and anti-CD8 microbeads. Numbers on flow cytometry plots show the percentage and absolute number of CD1d dimer⁺ NKT cells within gated YFP-negative and YFP-positive DN TCRβ⁺ cells. Results shown are from two mice done in one experiment.

Supplementary Figure 2 TCR α-chain joining-region gene-segment usage among out-of-frame Trav11-containing sequences detected in DN thymocytes from E8_III-Cre⁺Rag2^fl/fl mice.

High-throughput sequencing of Trav11Trac transcripts was carried out using the MiSeq system, where cDNA was prepared from sorted CD4^-CD8^- DN thymocytes from E8_III-Cre⁺Rag2^fl/fl mice. Bars depict the frequency (mean ± SEM, n = 3 biological replicates) of Traj gene segment usage among 3,002 out-of-frame Trav11 (CDR3) sequences detected when a total of 55,056 Trav11⁺ sequences were analyzed as shown in Table 1. The data are from one sequencing experiment.

Supplementary Figure 3 The TCR V_β repertoire of NKT cells of DN-stage-thymocyte origin.

Flow cytometry analysis of TCR Vβ chain usage analysis of thymic CD1d dimer⁺CD3ɛ⁺ DN NKT cells from E8_III-Cre⁺Rag2^fl/fl (black bars) and littermate control mice (gray bars). Bars represent mean percentages ± SEM of three independent experiments with at least two mice per group.

Supplementary Figure 4 Detection of Trav11Traj18 transcripts and DN NKT cells in the thymus of Cd4-Cre⁺Rag2^fl/fl mice.

(a) Viable 7-AAD^-B220^-CD1d dimer^-TCRβ^-/lo DP and DN thymocyte fractions that were electronically gated as CD5^loCD69^- cells were sorted from Cd4-Cre⁺Rag2^fl/fl mice, and Trav11Traj18 mRNA expression was determined by qPCR analysis. Relative gene expression (mean ± SEM, n = 3) with Gapdh used as internal control. Results are representative of two experiments, (n.d.) not detected. (b) Detection of CD1d dimer⁺ DN NKT cells in the thymus, and (c) in the liver from Cd4-Cre⁺Rag2^fl/fl mice, where littermate mice were used as controls. MACS depletion of CD4⁺, CD8⁺ and DP cells with magnetic microbeads was done prior to the cell staining. Numbers on flow cytometry plots show the percentage and the absolute number of CD1d dimer⁺TCRβ⁺ NKT cells among electronically gated DN cells. Results shown are from one experiment with two mice per genotype analyzed.

Supplementary Figure 5 Detection of Trav11Traj18 transcripts and mature DN NKT cells in the DN thymocyte fraction of Rorc^–/– mice.

(a) Trav11Traj18 mRNA expression in sorted CD4⁺CD8⁺ DP and CD4^-CD8^- DN thymic fractions from Rorc^-/- mice, as determined by qPCR. Bars show relative expression (mean ± SEM), where Gapdh was used as an internal control. Results are representative from three independent experiments with combined total of three to six mice per group, (n.d.) not detected. (b) Detection of CD1d dimer⁺ mature NKT cells within DN thymocytes of Rorc^-/- mice. Prior to the cell staining, DN thymocytes were purified with AutoMACS by using anti-CD4 and anti-CD8 microbeads. Numbers on plots indicate the percentage and absolute number (mean ± SEM, n=6 mice) of CD1d dimer⁺ NKT cells within gated DN TCRβ⁺ cells. Unloaded CD1d staining of DN thymocytes from B6 mice was used as a staining control. Results shown are representative from three independent experiments with two mice per group.

Supplementary Figure 6 Lack of Trav11Traj 18 transcripts as well as mature DN NKT cells in the thymus of Ptcra^–/– mice.

(a) The qPCR analysis of Trav11Traj18 mRNA expression in sorted thymic CD4⁺CD8⁺ DP, and CD44⁺CD25^- DN1, CD44⁺CD25⁺ DN2, CD44^-CD25⁺ DN3, CD44^-CD25^- DN4 fractions from Ptcra^-/- mice. Bars show relative expression (mean ± SEM, n = 3 biological replicates), where Gapdh was used as an internal control. Results are representative from three independent experiments, (n.d.) not detected. (b) Detection of CD1d dimer⁺ NKT cells within DN thymocytes from B6 and Ptcra^-/- mice. Numbers on flow cytometry plots indicate the percentage of CD1d dimer⁺TCRβ⁺ NKT cells within electronically gated DN cells. Prior to the cell staining, DN thymocytes were purified with AutoMACS by using anti-CD4 and anti-CD8 microbeads. Results shown are representative from three independent experiments with two mice per group.

Supplementary Figure 7 Surface phenotype of NKT cells of DN-stage-thymus origin is similar to that of B6 wild-type stage 3 NKT cells.

Viable 7-AAD^-B220^-CD1d dimer⁺TCRβ⁺ DN thymic NKT cells from B6 and E8_III-Cre⁺Rag2^fl/fl mice were analyzed for the expression of CD24, CD44, and NK1.1 surface antigens by flow cytometry. Numbers on the histograms show the percentage of CD24^lo DN NKT cells, and numbers on flow cytometry plots represent the percentage of DN NKT cells within the indicated quadrant gates. Results shown are representative from three independent experiments, n = 3 mice per genotype.

Supplementary Figure 8 Cytotoxic killing assay with YFP^– and YFP⁺ NKT cells from E8_III-Cre⁺Rosa26-YFP mice.

FACS-sorted YFP-negative and YFP-positive thymic DN NKT cells from E8_III-Cre⁺Rosa26-YFP mice were activated with anti-CD3 and anti-CD28 overnight, and then were co-cultured with CellTrace^TM Violet labeled YAC-1 target cells at a E/T ratio of 5:1 for 4 hrs. Samples then were FACS analyzed using annexin V and 7-aminoactinomycin D. Bars depict the mean specific cell lysis percentage ± SEM of triplicate samples done in one experiment, and were calculated by the following formula: % cytotoxicity = (target-spontaneous/maximum-spontaneous) ×100.

Supplementary information

Supplementary Text and Figures

Supplementary Figures 1–8 (PDF 622 kb)

Rights and permissions

Reprints and permissions

About this article

Cite this article

Dashtsoodol, N., Shigeura, T., Aihara, M. et al. Alternative pathway for the development of V_α14⁺ NKT cells directly from CD4^–CD8^– thymocytes that bypasses the CD4⁺CD8⁺ stage. Nat Immunol 18, 274–282 (2017). https://doi.org/10.1038/ni.3668

Download citation

Received: 30 September 2016
Accepted: 19 December 2016
Published: 30 January 2017
Issue date: March 2017
DOI: https://doi.org/10.1038/ni.3668

This article is cited by

Loss of Zfp335 triggers cGAS/STING-dependent apoptosis of post-β selection thymocytes
- Jeremy J. Ratiu
- William E. Barclay
- Yuan Zhuang
Nature Communications (2022)
Modulation of TCR signalling components occurs prior to positive selection and lineage commitment in iNKT cells
- Xuyen T. Dinh
- Dragana Stanley
- Margaret A. Jordan
Scientific Reports (2021)
Thymic iNKT cell differentiation at single-cell resolution
- Ke Wang
- Weijia Zhao
- Qing Ge
Cellular & Molecular Immunology (2021)
Liver X receptors regulate natural killer T cell population and antitumor activity in the liver of mice
- Kaori Endo-Umeda
- Hiroyuki Nakashima
- Makoto Makishima
Scientific Reports (2021)
Liver metastases
- Diamantis I. Tsilimigras
- Pnina Brodt
- Timothy M. Pawlik
Nature Reviews Disease Primers (2021)