Modulation of Vα19 NKT cell immune responses by α-mannosyl ceramide derivatives consisting of a series of modified sphingosines



We have demonstrated that analogues of α-mannosyl ceramide (α-ManCer) consisting of a series of immunosuppressive 2-aminoalcohol derivatives in place of sphingosine promote a greater immune response from mouse invariant Vα19-Jα26 (AV19-AJ33) TCR-bearing NKT (Vα19 NKT) cells than α-ManCer itself. To further characterize the immune responses of Vα19 NKT cells to the α-ManCer analogues, cytokine production by the cells was examined in detail. We found that certain α-ManCer derivatives individually induced either Th1- or Th2-dominant cytokine production in culture. The Th1- or Th2-biased immune responses of Vα19 NKT cells were dependent on MHC class I-like MR1, since they were induced by coculture with the MR1 transfectants previously loaded with the glycolipids and were inhibited in the presence of anti-MR1 antiserum. Presumably, the recognition of the α-mannosyl residue of the α-ManCer analogues by the invariant TCR is individually modulated, depending on the altered interaction with the groove of the antigen-presenting MR1. Priming of the Vα19 invariant TCR-transgenic mice in vivo with these glycolipid derivatives resulted in the induction of the Th1- or Th2-biased immune responses. Thus, these α-ManCer derivatives are likely to be useful in immunotherapy for either Th1 or Th2 excess autoimmune diseases, modulating the function of Vα19 NKT cells.


α-galactosyl ceramide


α-mannosyl ceramide


mononuclear cell

Vα14 NKT cell:

NK1.1+ T cell bearing an invariant Vα14-Jα18 TCR α chain

Vα19 NKT cell:

NK1.1+ T cell bearing an invariant Vα19-Jα26 TCR α chain


Natural killer T (NKT) cells are defined as lymphocytes bearing both the common NK marker NK1.1, a product of a member of the NKR-P1 gene family, and TCR-CD3 complex 1. The major component of NKT cells (Vα14 NKT cell) 2, 3 is characterized by the expression of the invariant TCR α chain (mouse Vα14-Jα18, human Vα24-Jα 18), and is positively selected by the no polymorphic MHC class I-like CD1d molecule in association with ß2m 4, 5. Vα14 NKT cells are responsive to certain glycosphingolipids in the context of CD1d such as α-galactosyl ceramide (α-GalCer, 6) isolated from marine sponge 7, α-glucuronosyl and α-galacturonosyl ceramides from α-proteobacteria 8, 9, and intracellular lysosomal isoglobotriaosyl ceramide 10.

We have recently demonstrated the presence of a novel NK1.1+ T cell repertoire (designated as Vα19 NKT cell in this study) expressing the Vα19-Jα26 invariant TCR α chain 11 that was previously found in human, bovine, and TAP-deficient mouse peripheral blood cells by quantitative PCR analysis 12. The cells bearing the invariant Vα19-Jα26 TCR are absent in mice lacking the non-classical MHC class I molecule MR1, thus suggesting that they are positively selected by MR1 13. It is estimated that Vα19 NKT cells represent 1% of mononuclear cells (MNC) in the liver 11, thus they are a considerably large population as a lymphocyte clone. Localization of the invariant Vα19-Jα26 TCR+ cells in gut lamina propria is also reported 13. Similar to Vα14 NKT cells 14, Vα19 NKT cells immediately produce large amounts of both Th1- and Th2-promoting immunoregulatory cytokines in response to the engagement of the invariant TCR and thus are considered to have important roles in the regulation of the immune system (15 and Shimamura, M. et al., Characterization of a novel NKT cell repertoire expressing an invariant Vα19-Jα26 TCRα chain using the invariant TCR transgenic mice, Abstract #4 for the 2nd International Workshop on CD1 antigen presentation and NKT cells, Woods Hole 2002). Therefore, the search for specific antigens for Vα19 NKT cells is quite important in developing new therapies for various immunoregulatory disorders based on the functional modulation of the repertoire.

The self-antigens presented by MR1 have not been identified 16. The discovery of α-GalCer as a stimulant for Vα14 NKT cells prompted us to investigate artificial glycosphingolipids as agonists for Vα19 NKT cells. We found that α-mannosyl ceramide (α-ManCer) was the best stimulus for Vα19 NKT cells among a series of the synthetic α-glycosyl ceramides with a naturally occurring monosaccharide 15. This glycolipid was presented by MR1 and caused Vα19 NKT cells to secrete both IL-4 and IFN-γ. Furthermore, we have recently found immunopromotive activity toward Vα19 NKT cells in modified α-ManCer consisting of a series of derivatives of an immunosuppressive antibiotic ISP-I 17 in place of the sphingosine unit 18; the activity was more intensive than that of the parental α-ManCer. Hence, we have continuously made efforts to characterize the synthetic α-ManCer derivatives for finding stimulants capable of modulating the function of Vα19 NKT cells.


α-ManCer derivatives induce either Th1 or Th2 dominant Vα-19 NKT cell responses

The α-ManCer derivatives characterized in this study are listed with conventional abbreviations in Fig. 1. They were tested for potency to induce Th1 or Th2-dominant immune responses from Vα19 NKT cells. Liver mononuclear cells (MNC) isolated from invariant Vα19-Jα26 TCR transgenic (Vα19 Tg) mice with the TCR α–/– background (in which Vα19 NKT cells are the sole component of NKT cells) and C57BL/6 mice as a control (among which Vα14 NKT cells represent the largest proportion) were cultured in the presence of the glycolipids. The amount of IL-4 and IFN-γ secreted into the supernatants was determined (Fig. 2A and B). α-ManCer but not α-GalCer analogues more or less enhanced the production of both IL-4 and IFN-γ by Vα19 Tg+ TCR α–/– but not C57BL/6 cells. These results were in accord with the report 18 that the proliferation and IL-2 production of Vα19 NKT cells were induced in the presence of α-ManCer derivatives in the culture medium. The IL-4 production by Vα19 Tg+ cells predominated on day 1 of culture, whereas the IFN- γ production reached maximum on day 2 of culture (Fig. 2B). This profile of cytokine production is similar to the profile observed in Vα19 Tg+ cells upon TCR engagement with immobilized anti-CD3 antibody. Depletion of the NK1.1+ or TCRαß+ population reduced the responsiveness of the responder cells (Fig. 2C). Thus, it is strongly suggested that the potential to respond to the α-ManCer derivatives was confined to Vα19 NKT cells. Interestingly, the relative intensity of IL-4 to IFN-γ secretion by Vα19 NKT cells was dependent on the chemical structure of the stimulator. α-ManCer with either a 2-hydroxymethyl group (Man2HMC16, Man2HMC24), or a 4-phenyl group (Man4PhC16) more intensively induced both IL-4 and IFN-γ production than the α-ManCer without any substitutions in the sphingosine portion (ManC16) or 3-hydroxy α-ManCer (Man3OHC16). On the other hand, the α-ManCer with both 2-hydroxymethyl and 4-phenyl groups (Man2HM4PhC16) induced less IL-4 in Vα19 NKT cells, and the cytokine production was biased to IFN-γ (Fig. 2A and B). To demonstrate more clearly the cytokine profile induced with the individual α-ManCer analogues, the fold-increase in IL-4 production by Vα19Tg+TCRα–/– liver MNC on day 1 versus that in IFN-γ production on day 2 in each culture was plotted in Fig. 2D. This profile strongly suggests that manipulation of the sphingosine portion of α-ManCer alters the interaction between invariant Vα19 TCR and the α-mannosyl residue in the glycolipids, resulting in the modulation of the immune responses of Vα19 NKT cells. More detailed cytokine profiles obtained from the culture of hepatic Vα19 NKT cells with representative α-ManCer analogues were examined (Fig. 3). Man4PhC16 induced production of both proinflammatory (IFN-γ, IL-12, IL-17) and Th2-promoting (IL-4, IL-5, IL-10) cytokines. Man2HM4PhC16 promoted proinflammatory whereas Man2HMC16 enhanced Th2-biased cytokine secretion. These results further support that modified α-ManCer are capable of modulating immune responses of Vα19 NKT cells.

Figure 1.

A list of the glycosyl ceramide derivatives characterized in the present study. Glycosphingolipids modified with a 2-hydroxymethyl, 3-hydroxyl, or 4-octylphenyl group are represented as 2HM, 3OH or 4Ph.

Figure 2.

Immune responses of Vα19 NKT cells in culture elicited by α-ManCer derivatives. (A) Liver MNC from Vα19 Tg+ TCRα–/– and C57BL/6 mice were cultured with the addition of glycolipids dissolved in DMSO (2 μg/mL). After 2 days, the immune responses were monitored by measuring the concentrations of IL-4 and IFN-γ in the culture fluid. The filled bars represent the culture of Vα19Tg+TCRα–/– cells, whereas the open bars show the results of C57BL/6 cells. Abbreviations of glycolipids are listed in the legend to Fig. 1. The average of the results obtained from independent eight experiments is indicated. The p values in Dunnett's multiple comparison post-test are calculated in comparison with the control. * p<0.01; ** p<0.05. (B) Dose-dependent activation of Vα19 NKT cells by α-ManCer derivatives in culture. Liver MNC from Vα19 Tg+ TCRα–/– mice were cultured with the indicated dose of glycolipids. After 1 day, the culture fluid was exchanged with the fresh medium with glycolipids. The concentration of cytokines in the culture supernatants (0∼1 day, 1∼2 day) were determined. One of the three independent experiments giving essentially the same profiles of cytokine production is shown. (C) Determination of the cell populations in the Tg liver responding to the α-ManCer analogues. Liver MNC prepared from Vα19Tg TCRα–/– mice were depleted of NK1.1+ or TCR αß+ cells as described in Materials and methods. Cells were cultured with the glycolipids, and the concentration of IL-4 and IFN-γ in the supernatants was determined by ELISA. The average of triplicate cultures in one of the independent three experiments is shown. (D) Modulation of immune responses of Vα19 NKT cells by α-ManCer derivatives. Liver MNC from Vα19 Tg+ TCRα–/– mice were cultured with α-ManCer derivatives as indicated in (A). The concentrations of IL-4 on day 1 of culture and the IFN-γ on day 2 of culture are plotted. Results are shown as the fold-increase relative to the control cultures with the vehicle. Large squares represent the fold-increases in cytokine production on the average.

Figure 3.

Either Th1- or Th2-dominant cytokine production by Vα19 NKT cells depending on the presence of the α-ManCer derivatives. Liver MNC prepared from Vα19 Tg+ TCRα–/– mice were cultured with the indicated α-ManCer derivative (2 μg/mL). Cytokine production in the culture supernatants on days 1 and 2 was determined by ELISA. One of the three experiments with essentially the similar profiles is shown. The p values in Dunnett's multiple comparison post-test are calculated in comparison with the control (cytokine levels in culture with vehicle). * p<0.01; ** p< 0.05.

Immune responses of Vα19 NKT cells primed in vivo with α-ManCer derivatives

The immune responses of Vα19 NKT cells specifically induced by the α-ManCer derivatives were also observed when they were primed in vivo with the glycolipids. Spleen cells from Vα19Tg+TCRα–/– and C57BL/6 mice injected 90 min previously with the glycolipids were cultured and cytokines secreted into the supernatants were determined (Fig. 4). Vα19Tg+TCRα–/– splenocytes produced IL-4 and IFN-γ in a similar fashion to those stimulated in vitro. They produced increased amount of both IL-4 and IFN-γ with Man4PhC16. Relatively enhanced IL-4 production was observed when the cells were primed with Man2HMC16, whereas IFN-γ production predominated when they were stimulated with Man2HM4PhC16. On the other hand, C57BL/6 cells apparently displayed less responsiveness to these α-ManCer analogues presumably due to the lower frequency of Vα19 NKT cells in the spleen. Thus, α-ManCer derivatives injected into mice possibly target Vα19 NKT cells and promote either Th1- or Th2-dominant immune responses.

Figure 4.

Priming of Vα19 NKT cells in vivo with α-ManCer derivatives. Spleen cells from Vα19 Tg+ TCRα–/– and C57BL/6 mice previously injected intravenously with glycolipids (20 μg/animal) or vehicle (DMSO) in 200 μL PBS via the tail were cultured for the period indicated. Culture supernatants were harvested and tested for production of cytokines at the indicated time points. The thick line with dots and the fine line in each panel represent the cytokine production in the culture with glycolipids and vehicle, respectively. Three independent experiments were performed and essentially the same profiles were obtained.

MR1-restricted stimulation of Vα19 Tg+ cells with the α-ManCer derivatives

MHC restriction of the immune responses by Vα19 NKT cells to the α-ManCer derivatives was examined. Affinity of α-ManCer derivatives to immobilized MR1/ß2m proteins were suggested by the binding assay shown in Fig. 5. The binding of α-ManCer derivatives to the plastic well previously coated with MR1/ ß2m proteins was indirectly detected by the subsequent binding of Con A to the glycolipids. The binding of α-ManCer derivatives to MR1 was further supported by the observation that Vα19 Tg+ cells were activated to some degree when they were cultured in the plastic wells pre-coated with MR1/ ß2m proteins followed by α -ManCer derivatives (Supporting Information online).

Figure 5.

Binding of α-ManCer analogues to immobilized MR1. (A) The MR1/ ß2m proteins expressed in E. coli proteins were analyzed by SDS polyacrylamide gel electrophoresis. (B) Binding of α-ManCer analogues to immobilized MR1. MR1/ ß2m proteins (3 μg/mL) were immobilized on a 96-well plate. After washing, α-ManCer analogues were added to the wells and incubated. The glycolipids bound to the MR1 proteins were detected with peroxidase-conjugated Con A. Results are shown as difference in color development (OD 450 nm) from the well added by α-GalCer. One representative experiment of four is shown. Closed symbols represent the binding of peroxidase-conjugated Con A to the well previously coated with the MR1/ ß2m proteins, whereas open symbols represent the binding to the well without coating with the proteins.

Next, MR1-restricted immune responses in culture by Vα19 Tg+ cells were tested (Fig. 6). Liver MNC from Vα19 Tg+ TCRα–/– and C57BL/6 mice were cultured with the α-glycosphingolipid analogues in the presence of anti-MR1 antiserum or pre-immune serum, and the responses were analyzed (Fig. 6C). The immune responses by Vα19 NKT cells to the α-ManCer derivatives were drastically reduced in the presence of anti-MR1 antiserum in culture. On the other hand, the antiserum had no effects on the responses of C57BL/6 liver MNC (where CD1d-restricted Vα14 NKT cells represent about 30% of population) toward α-GalCer (GalC24).

Figure 6.

Stimulation of Vα19 NKT cells with α-ManCer derivatives in the context of MR1. (A)Preparation of MR1-transfectants. MR1 cDNA tagged with FLAG was expressed in Raji cells. Cell lysates were analyzed by Western blotting. A 40-kDa band was specifically stained with the partially purified anti-MR1 antiserum for the MR1-Raji cells corresponding to the band stained with anti-FLAG antibody. (B) Expression of MR1 by MR1-transfected cells. MR1- or mock-transfected Raji cells were stained with anti-MR1 antiserum in the presence or absence of the partial MR1 peptide used as an antigen or pre-immune serum peptide and analysed by FACS. (C) Inhibition of the immune responses in the presence of anti-MR1 antiserum. Liver MNC prepared from Vα19 TCR Tg+ TCR α–/– or C57BL/6 mice were stimulated with α-ManCer derivatives for 2 days in the presence of anti-MR1 antiserum or pre-immune serum. Cytokines were measured by ELISA. The average of triplicate culture in one of the representative three experiments is shown. (D) Stimulation of Vα19 NKT cells with α-ManCer derivatives presented by MR1 transfectants. T-lineage cells (CD5+ cells) were enriched from liver MNC of Vα19 TCR Tg+ TCR α–/– or ß2m–/– mice. They were stimulated with MR1-transfected or mock-transfected Raji cells, previously loaded with α-ManCer derivatives. Immune responses were determined on day 2 of culture by measuring the secretion of IL-4 and IFN-γ. The average of triplicate cultures in one of the representative three experiments is shown.

MR1 restriction was further supported by the immune responses of Vα19 Tg+ cells toward MR1-transfectants. CD5+ cells were prepared from the liver MNC of Vα19 Tg+ TCRα–/– or ß2m–/– mice to enrich T-lineage cells, and they were stimulated with the MR1-transfected or non-transfected cells of a B lymphoma line (Raji) 19 previously loaded with α-ManCer derivatives (Fig. 6D). The CD5+ cells from Vα19 Tg+ TCRα–/– but not ß2m–/– livers responded to the stimulation with the MR1-transfectants treated with the α-ManCer derivatives.

Taken together, it is strongly suggested by these findings that Vα19 NKT cells respond to α-ManCer derivatives that are presented by non-classical MHC class I-like MR1 molecules.


In this study, we demonstrated that TCR engagement of Vα19 NKT cells was induced by the stimulation with the modified α-ManCer that were presented by MR1-transfectants (Fig. 6) or immobilized MR1/ ß2m proteins (Supporting Information). Either Th1- or Th2-dominant immune responses of Vα19 NKT cells were brought by the administration of α-ManCer analogues not only in vitro but also in vivo (Fig. 4); thus the potentials as an immune modulator were found in the glycolipids.

The structural modification of the sphingosine moiety of the α-ManCer analogues could alter the interaction between the antigen-presenting MHC molecule, MR1. It has recently been reported that 2-hydroxymethyl 4-phenyl sphingosine (FTY720) and the related compounds mimic sphingosine-1-phosphate and work as an agonist for sphingosine-1-phosphate receptor 20. This finding suggests that the modified α-ManCer as well as the natural α-ManCer are capable of being presented by MR1. The altered interaction between the modified α-ManCer and MR1 could presumably influence the spatial location of the α-mannosyl residue to be recognized by the invariant Vα19 TCR and eventually result in the modulation of the immune responses of Vα19 NKT cells. Induction of Th2-biased immune responses of Vα14 NKT cells with the α-GalCer consisting of a short sphingosine base has been reported 21. It is proposed in the report that the sporadic stimulation of the invariant Vα14 TCR with the galactose residue of the modified α-GalCer causes insufficient transcription of c-Rel, which is responsible for the expression of Th1 cytokines such as IFN-γ 22. Provided that this speculation on the immune responses of Vα14 NKT cells is applicable to those of Vα19 NKT cells, introduction of 4-phenyl or 2-hydroxymethyl group into the sphingosine portion of α-ManCer is suggested to stabilize the interaction with MR1 and sustain the stimulation of the invariant Vα19 TCR with the modified α-ManCer, because the modified α-ManCer induced more IFN-γ production in Vα19 NKT cells than α-ManCer without substitution. In fact, the affinity of Man2HM4PhC16 or Man4PhC16 to immobilized MR1 proteins was suggested to be larger than that of ManC16 by the binding assay (Fig. 5). However, the supposed binding affinity of the glycolipids to the MR1 antigen-presenting groove did not necessarily parallel with the activity to induce cytokine production, especially IL-4 production, by Vα19 NKT cells. For example, Man2HMC16 induced Vα19NKT cells to produce more IL-4 than Man2HM4PhC16 or ManC16. The special location of the α-mannosyl residue of the glycolipids, which is influenced by the interaction between the glycolipid and MR1, may also be important to determine the antigenic activity in addition to the stability of the interaction between them.

Antigen nonspecific polyclonal activation of Vα19 cells in culture with immobilized anti-CD3 antibody induced the cytokine production, where the ratio of IL-4/IFN-γ was similar to that when Vα19 cells were stimulated with Man4PhC16 (data not shown). Based on this standard, the cytokine releases of Vα19 NKT cells induced with ManC16, Man3OHC16, Man2HMC16 and Man2HMC24 are assigned to be Th2 biased, whereas those with Man2HM4PhC16 are Th1 biased (Fig. 2D). To our knowledge, the present finding on Man2HM4PhC16 is the first model for the induction of Th1-dominant immunity by administration of glycolipid activators.

Vα19 NKT cells sorted from MNC of Vα19 TCR Tg+ TCRα–/– livers had potentials to secrete both IL-4 and IFN-γ following TCR engagement. Nevertheless, it is possible that cytokine production found in the cells from Vα19 Tg mice in part arose not only from the Vα19 NKT cells but also from the bystander cells. In fact, intracellular IFN-γ production in NK1.1+ TCRand NK1.1 TCR+ cells as well as NK1.1+ TCR+ cells in the Vα19 Tg cells were found following TCR engagement with anti-CD3 antibody (submitted for publication). Similarly, cells in the other lineage such as dendritic cells in the culture may contribute to producing IL-12, for instance, in a secondary manner. However, it should be noted that the specific TCR engagement of Vα19 NKT cells with the modified α-ManCer triggers the following characteristic cytokine production of Vα19 Tg cells in vitro and in vivo that will contribute to the regulation of Th1/Th2 homeostasis.

Vα19 and Vα14 NKT cells are suggested to share roles in the regulation of the immune system despite being subjected to the controls of independent MHC restriction and antigen specificity. Presumably, these two repertoires are individually involved in certain immune regulatory functions. Recently, localization of Vα19 but not Vα14 invariant TCR α chain-bearing cells in gut lamina propria was reported 13. In addition, we found that they accumulated in the lesions of patients suffering from multiple sclerosis and autoimmune inflammatory neuropathy 23. We also found that over-generation of invariant Vα19 TCR+ cells inhibited the induction of experimental autoimmune encephalomyelitis, a mouse model of multiple sclerosis 24. These findings suggest the possible functional sharing among these subsets. Hence, a series of the α-ManCer derivatives will be important as a prospective immunotherapeutic reagent selectively targeting Vα19 NKT cells for various autoimmune diseases in either Th1 or Th2 excess.

Materials and methods

Synthetic glycolipids

α-ManCer consisting of 3-hydroxy sphingosine (Man3OHC16, Fig. 1) was synthesized as previously described 15. Other synthetic glycolipids consisting of immunosuppressant FTY720 and its related compounds listed in Fig. 1 were synthesized as described 18.


C57BL/6 mice and ß2m-deficient (ß2m–/–) mice with the C57BL/6 genetic background were obtained from Sankyo Service (Tokyo, Japan) and Jackson Laboratory (Bar Harbor, ME). TCR Cα-deficient mice 25, originally with the 129 genetic background that were backcrossed with C57BL/6 mice for ten times, were a gift of Dr. H. Ishikawa (Keio University).

Vα19-Jα26 invariant TCR Tg mice with the TCR Cα-deficient background were established as described previously (submitted for publication, 15). In brief, a Vα19-Jα26 invariant TCR gene segment amplified from the genomic DNA of hybridoma NB403 11 was linked to the TCR α chain promoter, constant, and enhancer regions. The transgene thus produced was introduced into the fertilized eggs of TCR α–/– mice and a Tg line was established. The Tg+ cells preferentially develop as NKT cells in the Tg mice. The Tg+, NK1.1+ population accounts for 30% of total MNC in the Tg liver.

All the experiments using mice were reviewed and approved by the experimental animal committee of Mitsubishi Kagaku Institute of Life Sciences.

Bioassay of the synthesized glycolipids in culture

MNC used as responders were prepared from mouse livers by density gradient centrifugation using Percoll (Pharmacia, Uppsala, Sweden) as described previously 26. Liver MNC from indicated strains of mice (2∼4 months of age) were cultured in DMEM supplemented with 10% FCS, 100 U/mL penicillin, 50 μg/mL streptomycin and 5 X 10–5 M 2-mercaptoethanol (106 cells/200 μL) in the presence of glycolipids. Glycolipids were dissolved in DMSO and were added to the culture at the final concentration of 2 μg/mL. Each culture included 1/200 v/v of DMSO (used as the vehicle). Concentrations of cytokines in the culture supernatants were determined by ELISA using ELISA kits (PharMingen, San Diego, CA). The error bars indicate the standard deviation. Statistical analysis (Dunnett's multiple comparison post-test) was applied to some results of the immune responses in comparison with the control (the immune responses in culture only with the vehicle).

In some experiments, liver MNC were treated with biotinylated anti-NK1.1 mAb (PK136, PharMingen) or anti-TCR Cß (H57–597, PharMingen) and then with magnetic beads coated with streptavidin (Dynal Biotech, Oslo, Norway) to remove NK1.1+ or TCR Cß+ cells before culture. The percentage of NK1.1+ or TCRαß+ cells after treatments with magnetic beads was less than 3%. In other cases, liver MNC were treated with magnetic beads coated with anti-B220 antibody (Dynal Biotech). After removal of B220+ cells, the remaining cells were then treated with magnetic beads coated with anti-CD5 antibody (Miltenyi Biotech, Gladbach, Germany). The CD5+ cells were enriched by MACS and were used as responders. The percentages of TCRαß+ cells in the CD5+ and CD5 fractions were in the range of 91–96 and 1–4%, respectively.


MR1-transfectants were prepared from cells of a human Burkitt's B lymphoma line, Raji (ATCC) 19. Mouse MR1 A cDNA 27 was amplified from genomic DNA of C57BL/6 spleen cells using RT-PCR kits (Takara, Tokyo, Japan).The following PCR primers were used for: (5′-MR1), 5′-ATGATGCTCCTGGTTACCTGG-3′; (FLAG-3′-MR1), 5′-CTACTTGTCATCGTCATCCTTGTAGTC(FLAG)-AGAGGGAGAGCTTCCCTCAT-3′.

The PCR product was cloned into pGEM-Teasy (Promega, Madison, WI). The MR1 cDNA was recombined into a eukaryotic expression vector, pCXN 28 (provided by Dr. J. Miyazaki, Osaka University). The expression vector was transfected into Raji cells. The transfectants were selected in the culture medium containing G418 (1 mg/mL) for 1 month. The expression of FLAG (Asp-Tyr-Lys-Asp-Asp-Asp)-MR1 in the transfectants was analyzed by Western blot using the combination of anti-FLAG antibody (Stratagene, La Jolla, CA) and HRP-labeled anti-mouse immunoglobulin (Sigma, St. Lewis, MO) or anti-MR1 antiserum (described below) and HRP-labeled anti-rabbit immunoglobulin (Sigma).

Generation of anti-MR1 antiserum

Rabbits were immunized with a keyhole limpet hemocyanin-conjugated polypeptide corresponding to the α2 domain of mouse MR1 (residue 139–161) 27 with Freund's complete adjuvant (Sigma). Anti-MR1 antiserum was concentrated by affinity chromatography using Sepharose 4B (Pharmacia) conjugated with the antigen peptide.

Cytometric analysis of MR1-transfectant cells

MR1-transfectant cells were pretreated with anti-CD16 mAb (LNK16, Dainippon Pharma, Tokyo, Japan). After immunostaining of the cells with rabbit anti-MR1 antiserum, they were treated with an FITC-labeled second antibody (donkey IgG F(ab’)2 fragment against rabbit IgG, Jackson Laboratory). Mouse cells were pretreated with anti-Fc receptor antibody (2.4G2), then stained with anti-TCR Cß antibody (H57–597) or anti-NK1.1 antibody (PK136). The antibodies were obtained from Becton Dickinson (San Jose, CA). The stained cells were analyzed on a flow cytometer (FACScan, Becton Dickinson) equipped with Cell Quest software.

Stimulation of Vα19 Tg+ cells with MR1-transfectants

MR1- and mock-transfected Raji cells (1 × 105) were incubated in DMEM with glycolipids (5 μg/mL) for 18 h and washed with DMEM twice. After irradiation (3000 rad), they were cocultured with liver MNC (1 × 106) in DMEM (200 μL) for 2 days. Immune responses by the liver MNC in the mixed lymphocyte reactions were monitored by measuring concentrations of cytokines in the culture fluid by ELISA.

Stimulation of Vα19 Tg+ cells in vivo

Stimulation of lymphocytes in vivo was performed as reported 29. Vα19 Tg+ TCR α–/– or C57BL/6 mice (8-20 weeks of age) were intravenously injected with glycolipids (20 μg/200 μL PBS) instead of anti-CD3 antibody. Spleens were removed from mice 90 min after the injection. MNC were immediately prepared from them by density gradient centrifugation using lymphosepar II (IBL, Gunma, Japan, d = 1.090). They were cultured in DMEM (107 cells /mL). Cytokine concentration in the culture fluid was determined by ELISA.

Expression of MR1 proteins

MR1 A and ß2m cDNA were amplified from C57BL/6 spleen cells [27] using RT-PCR kits (Takara, Tokyo, Japan). MR1 cDNA with a sequence for a BirA site 30 was cloned into pRSET expression vector (Invitrogen, Carlsbad, CA). Similarly, ß2m cDNA was cloned into pET28a vector (Novagen, San Diego, CA). The proteins were expressed in BL21(DE3) codon plus RIL host cells (Stratagene, La Jolla, CA). The expressed proteins were dissolved in 6 M guanidine hydrochloride solution. MR1 proteins were labeled with biotin using a labeling kit (Roche, Mannheim, Germany), and purified by avidin gel chromatography (Pierce Biotechnology, Rockford, IL). The proteins were refolded as described 31. MR1/ ß2m complex yielded 35- and 12-kDa bands in SDS PAGE analysis after staining with Coomassie brilliant blue (more than 95% purity).

Binding assay

MR1 proteins in PBS (3 μg/mL, 100 μL) were incubated in a 96-well plate for overnight at 4°C. The plate was washed with PBS and blocked with the assay diluent solution for ELISA (Becton Dickinson) for 2 h at room temperature. After washing with PBS, the plate was incubated with the solution of modified α-ManCer or α-GalCer (33, 10 and 3 μg/mL in PBS including 0.1% Tween 20, 50 μL) for 2 h at room temperature. The plate was washed with PBS, then peroxidase-conjugated Con A (Sigma, 0.5 μg/mL in the assay diluent, 50 μL) was added to the plate. It was incubated for 30 min at room temperature. It was washed with PBS including 0.1% Tween 20 seven times, and was added by tetramethylbenzidine liquid substrate (Sigma). The color development was measured by an ELISA reader (Model 550, Bio-Rad, Hercules, CA).


The authors thank Dr. O. Kanie for advice. They thank Dr. H. Ishikawa and J. Miyazaki for providing the TCR Cα-deficient mice and the expression vector. They also thank Ms. R. Fujii and Ms. H. Watanabe, for help with experiments, Mr. S. Kamijo and his group for taking care of the mice, and Ms. Y. Murakami for secretarial assistance. This work was supported by a grant from the Ministry of Health, Welfare and Labor, Japan.


  1. 1


  2. 2


  3. 3


  4. 4


  5. 5


  6. 6