iNKT cells play important roles in immune regulation by bridging the innate and acquired immune systems. The functions of iNKT cells have been investigated in mice lacking the Traj18 gene segment that were generated by traditional embryonic stem cell technology, but these animals contain a biased T cell receptor (TCR) repertoire that might affect immune responses. To circumvent this confounding factor, we have generated a new strain of iNKT cell-deficient mice by deleting the Traj18 locus using CRISPR/Cas9 technology, and these animals contain an unbiased TCR repertoire. We employed these mice to investigate the contribution of iNKT cells to metabolic disease and found a pathogenic role of these cells in obesity-associated insulin-resistance. The new Traj18-deficient mouse strain will assist in studies of iNKT cell biology.

Invariant natural killer T (iNKT) cells are characterized by their expression of a semi-invariant T cell receptor (TCR), Vα14 Jα18 (Trav11-Traj18) paired with Vβ8.2 (Trbv13-2), Vβ7 (Trbv29), or Vβ2 (Trbv1) in mice and Vα24 Jα18/Vβ11 (TRAV10-TRAJ18/TRBV25-1) in humans. These TCRs recognize glycolipid antigens such as α-galactosylceramide (α-GalCer) presented by the monomorphic major histocompatibility complex (MHC) class I-like molecule, CD1d. iNKT cells are potent immune regulators due to their rapid and massive production of a wide range of cytokines upon stimulation, such as IFN-γ, GM-CSF, IL-4, IL-13, IL-17A, and IL-101,2. This feature enables iNKT cells to bridge the innate and acquired immune systems, and to participate in immune responses during a variety of conditions, including infection, autoimmunity, allergy, tumorigenesis, as well as obesity and metabolic disease3,4,5.

NKT cell-deficient mouse strains, including Traj18-deficient (Traj18−/−) mice that selectively lack iNKT cells, and CD1d-deficient (Cd1d−/−) mice that lack both iNKT (also called type 1 NKT) cells and variant NKT (vNKT or type 2 NKT) cells, have greatly facilitated studies on NKT cells. Investigators have employed Traj18−/− mice to study iNKT cell functions, and have compared immune responses in Traj18−/− and Cd1d−/− mice to investigate type 2 NKT cell functions. However, the Traj18−/− mouse strain that was generated using traditional embryonic stem cell/gene targeting technology and has been widely used in iNKT cell studies6, was reported to contain an impaired TCR repertoire diversity, due to the PGK-Neor cassette that was employed to replace the Traj18 allele and might have inadvertently caused alterations in TCR gene transcription and rearrangement7. These findings therefore call into question prior studies that have employed these Traj18−/− animals to reach conclusions about iNKT cell functions.

iNKT cells are enriched in human and murine adipose tissue8. However, functional analyses from different research groups using models of NKT cell-deficiency (Cd1d−/−or Traj18−/− mice) have reached divergent conclusions9,10,11. It is possible that the lower TCR diversity of the Traj18−/− mice used in some of the studies could potentially contribute to these divergent results.

Recently, three new Traj18−/− mouse lines have been established by different research groups. Two lines were generated by Cre/lox technology12,13, and a third was generated by transcription activator-like effector nuclease (TALEN) methodology14. Each of these Traj18−/− mouse lines was shown to contain a selective deletion of the Traj18 locus and iNKT cell-deficiency, in the absence of a biased TCR repertoire. Concerning iNKT cell functions, these animals have thus far been employed to investigate the role of iNKT cells in allergen-induced pulmonary inflammation12 and α-GalCer-mediated suppression of tumor metastases13. Additional studies are needed to reassess iNKT cell functions and those of type 2 NKT cells with these novel Traj18−/− mouse lines and Cd1d−/− mice.

Here we have successfully generated new Traj18−/− mouse lines by using the Clustered Regularly Interspaced Short Palindromic Repeat (CRISPR)/Cas9 system. Both FACS analysis and cytokine production results confirmed the lack of iNKT cells in the newly generated Traj18−/− mouse strains in both C57BL/6 J (B6) and BALB/cAJcl (BALB/c) backgrounds. Analysis of the TCRα repertoire confirmed that these Traj18−/− mice harbor an undisturbed TCRα repertoire. Using this new mouse strain on the B6 background, we re-assessed the contribution of iNKT cells to obesity-associated metabolic disease, and found that obese Traj18−/− mice show reduced weight gain and ameliorated metabolic parameters, thus indicating a pathological role of iNKT cells in the development of obesity-associated disorders.

Two Traj18 gene-targeted single guide RNAs (called Traj18_sgRNA1 and Traj18_sgRNA2) (Supplemental Fig. 1a) were designed to target the Traj18 gene segment. We first validated whether the sgRNAs could recognize and cleave the Traj18 target sequence using an in vitro system, as described previously15. In brief, the targeted genome segment of the Traj18 locus (Supplemental Fig. 1b), including sgRNA target sequence, was inserted between the split-EGFP (enhanced green fluorescent protein) fragments that share 400 bp of DNA sequence, under control of the CAG promoter (pCAG-EGxnFP-target) and used as a reporter plasmid. We co-transfected pCAG-EGxnFP-target and pCAG-T3-hCas9-pA with or without pU6-sgRNA (Supplemental Fig. 1c) into HEK293T cells and the levels of reconstituted EGFP expression were evaluated by fluorescence microscopy (Supplemental Fig. 1d) and flow cytometry (Supplemental Fig. 1e) 48 hrs after transfection. Both Traj18_sgRNA1 and Traj18_sgRNA2 worked effectively, as revealed by EGFP expression in approximately 40% of the transfected cells.

Following validation of sgRNAs in HEK293T cells, we proceeded to generate Traj18 gene-targeted mutant mice by zygote injection. sgRNA and hCas9 mRNA were placed under the phage T3 promoter followed by in vitro transcription using T3 RNA polymerase (Supplemental Fig. 2a) and injected into the pronuclei of fertilized eggs of B6 mice. Pups derived from these fertilized eggs were genotyped by sequence analysis. Eight out of 11 mice from the Traj18_sgRNA1 (Supplemental Fig. 2b, Supplemental Table 1) and 10 out of 17 mice from the Traj18_sgRNA2 (Supplemental Fig. 2c, Supplemental Table 1) contained a partial deletion in the Traj18 locus. We selected three founder mice and established four new strains with a Traj18-partial deletion, Traj18−/− (1-1 L), Traj18−/− (1-1 S) and Traj18−/− (1-2) derived from Traj18_sgRNA1, and Traj18−/− (2-1) derived from Traj18_sgRNA2.

We compared the TCRα repertoire diversity in sorted pre-selection double-positive (DP) thymocytes (TCRβlow CD4+ CD8+ CD69) from Traj18−/− (1-1 L), Traj18−/− (1-1 S), Traj18−/− (1-2), Traj18−/− (2-1) and wild-type (WT) B6 mice. We performed PCR amplification of Trav11 (encoded Vα14) that contains iNKT-TCRα, or Trav14 (encoded Vα2), the most frequently used TCRα in αββT cells, by using a specific forward primer for each Vα encoding sequence and a reverse primer for the sequence encoding the TCRα constant region (Cα). The products were purified and subjected to next-generation sequencing analysis. All four Traj18-partial deletion mouse lines harbored similar Traj gene segments as WT B6 mice, except for Traj18 (Fig. 1a). Selective deficiency in Traj18 was confirmed in Traj18−/− (1-1 L), Traj18−/− (1-2), and Traj18−/− (2-1) lines, and we found a very low percentage of Traj18 usage in Traj18−/− (1-1 S) mice (Fig. 1b). We further examined iNKT cells in the thymus by staining with α-GalCer-loaded CD1d-dimers, and confirmed the absence of iNKT cells in each of the four lines (Fig. 1c). These results demonstrated that the newly generated Traj18-partial deletion mouse lines lacked iNKT cells and harbored an undisturbed TCRα repertoire, fulfilling the criteria of iNKT cell-deficient mice.

Figure 1
Figure 1

Generation of Traj18 −/− mouse lines by CRISPR/Cas9 technology. (a) TCRα repertoire diversity analyzed by next generation sequencing. TCRβlow CD4+ CD8+ CD69 thymocytes from WT B6, Traj18 −/− (1-1 L), Traj18 −/− (1-1 S), Traj18 −/− (1-2), and Traj18 −/− (2-1) were sorted. Trav11-Trac or Trav14-Trac PCR products were prepared and subjected to next-generation sequencing analysis. The graphs show percentages of productive Traj gene segment rearrangements. Data represents mean ± SD of three biologically independent samples per group. (b) Traj18 gene segment usage in Trav11-Trac or Trav14-Trac transcripts analyzed by next-generation sequencing. (c) Frequencies of iNKT cells (TCRβ+, α-GalCer/CD1d dimer+) in total thymocytes isolated from WT B6, Traj18 −/− (1-1 L), Traj18 −/− (1-1 S), Traj18 −/− (1-2), and Traj18 −/− (2-1) mice were analyzed by flow cytometry. Numbers represent the percentage of iNKT cells in the respective gates.

Full size image

Since each of the four novel Traj18−/− mouse lines lacked thymic iNKT cells and contained a TCRα repertoire similar to WT mice, we next selected the Traj18−/− (1-1 L) line for further experiments.

Frequencies of iNKT cells in the spleen and liver from WT B6 and Traj18−/− (1-1 L) mice were analyzed by flow cytometry, which revealed iNKT cell-deficiency in Traj18−/− (1-1 L) mice (Fig. 2a). Analysis of developmental stages of thymocytes revealed no difference between Traj18−/− (1-1 L) and WT B6 mice (Supplemental Fig. 3a). We also analyzed the frequencies of T cells with specific functions such as type 2 NKT cells, regulatory T cells (Treg) and mucosal-associated invariant T (MAIT) cells in the thymus, resulting in no differences between Traj18−/− (1-1 L) and WT B6 mice (Supplemental Fig. 3b–d). Similarly, no differences were observed in the frequencies of other immune cell types in the spleen, including CD4/8 αβT cells, γδT cells, B cells, NK cells, conventional dendritic cells, plasmacytoid dendritic cells, macrophages, and granulocytes in the spleen of Traj18−/− (1-1 L) and WT B6 mice (Supplemental Fig. 4a–f), indicating that the development of these immune cells was not affected by the deletion of Traj18 in the present mouse strain.

Figure 2
Figure 2

Traj18 −/− (1-1 L) mice lack iNKT cells and fail to respond to α-GalCer stimulation. (a) Total thymocytes, splenocytes and liver MNCs isolated from WT B6 or Traj18 −/− (1-1 L) mice were analyzed by flow cytometry. Numbers represent the frequencies of iNKT cells (TCRβ+, α-GalCer/CD1d dimer+) in the respective gates. (b) Total lymphocytes isolated from thymus (1 × 106), spleen (1 × 106), and liver (0.5 × 106) of WT B6 or Traj18 −/− (1-1 L) mice were stimulated with α-GalCer (0, 1 or 10 ng/mL), and the supernatants were collected 48 hrs post-stimulation. Cytokine levels of IFN-γ, GM-CSF, IL-4, IL-10, IL-13, and IL-17A were quantified by CBA. Data represents mean ± SD of each group (n = 3 per group). N.D., not detected.

Full size image

Levels of cytokines GM-CSF, IFN-γ, IL-4, IL-10, IL-13 and IL-17A, produced by total thymocytes, splenocytes and liver mononuclear cells (LMNCs) were measured after α-GalCer (0, 1, or 10 ng/mL) stimulation for 48 hrs, which revealed absence of cytokine production in Traj18−/− (1-1 L) mice (Fig. 2b). Similar loss of reactivity against α-GalCer was observed in other three lines, Traj18−/− (1-1 S), Traj18−/− (1-2) and Traj18−/− (2-1) (data not shown).

Because some functional studies on iNKT cells require animals on a genetic background distinct from the B6 strain, we also generated a Traj18−/− (1-1 L) BALB/c mouse line by backcrossing, and confirmed the absence of iNKT cells and cytokine production in response to α-GalCer stimulation (Supplemental Fig. 5a,b).

In summary, our data show that both B6 and BALB/c background Traj18−/− (1-1 L) strains are selectively iNKT cell-deficient.

NKT cells are one of cell types that reside in adipose tissue. However, divergent findings for the metabolic role of iNKT cells have been reported in studies using the previously generated Traj18−/− mouse strain. Therefore, we re-investigated the contribution of iNKT cells to the development of obesity induced by a high-fat diet (HFD) using our novel Traj18−/− mouse strain on the B6 background. We first investigated the frequencies of type 2 NKT cells, Treg, MAIT cells, γδ T cells, M0/1/2 macrophages in Traj18−/− (1-1 L) in the steady state, because these immune cells are considered to be effector and regulatory cells in metabolic disorders. It was found that no differences were observed in frequencies of these immune cells between WT B6 and Traj18−/− (1-1 L) (Supplementary Fig. 6a–e).

WT B6, Traj18−/− (1-1 L), and Cd1d−/− male mice were fed with a HFD or a normal chow diet (ND) starting from 8 weeks of age. For WT B6 and Traj18−/− (1-1 L) mice receiving ND, similar weight curves were observed during the 84-day feeding period. All mouse strains receiving…