Homotypic interactions mediated by Slamf1 and Slamf6 receptors control NKT cell lineage development

Abstract

Commitment to the T and natural killer T (NKT) cell lineages is determined during alphabeta T cell receptor (TCR)-mediated interactions of common precursors with ligand-expressing cells in the thymus. Whereas mainstream thymocyte precursors recognize major histocompatibility complex (MHC) ligands expressed by stromal cells, NKT cell precursors interact with CD1d ligands expressed by cortical thymocytes. Here, we demonstrated that such homotypic T-T interactions generated "second signals" mediated by the cooperative engagement of the homophilic receptors Slamf1 (SLAM) and Slamf6 (Ly108) and the downstream recruitment of the adaptor SLAM-associated protein (SAP) and the Src kinase Fyn, which are essential for the lineage expansion and differentiation of the NKT cell lineage. These receptor interactions were required during TCR engagement and therefore only occurred when selecting ligands were presented by thymocytes rather than epithelial cells, which do not express Slamf6 or Slamf1. Thus, the topography of NKT cell ligand recognition determines the availability of a cosignaling pathway that is essential for NKT cell lineage development.

Figure 1. NKT cell developmental arrest in SAP^-/- and Fyn^-/- mice

A. Conserved frequency of Vα14-Jα18 rearrangements in Fyn and SAP-deficient thymocytes. Quantitative RT-PCR of canonical Vα14-Jα18 rearrangements and Cα in sorted DP thymocytes from WT, Fyn^-/-, Jα18^-/-, CD1^-/- (left) and in DP thymocytes from CD1^-/-SAP^+/+ and CD1^-/- SAP^-/- mice (right). SAP^-/- mice were crossed onto a CD1^-/- background to eliminate any contamination of DP thymocyte by mature NKT cells. The values shown are ratios between Vα14-Jα18 and Cα transcripts. B. Vα14-Jα18 transgenic thymocytes on a Fyn^+/+, Fyn^-/- or SAP^-/- background were stained with CD1d-αGalCer tetramers and CD24 to enumerate the immature CD24^high and mature CD24^low stages. Cell frequencies are indicated in the corresponding gates. C. Left column, CD1d-αGalCer tetramer positive thymocytes were enriched by autoMACS from pools of thymi obtained from 2 week-old SAP^+/- and SAP^-/- littermates, Fyn^-/- and Jα18^-/- mice prior to FACS analysis with CD24. Numbers in the CD1d-αGalCer⁺ CD24^high and CD24^low gates represent absolute cell numbers recovered from 2 pooled thymi. Right column, gated CD1d-αGalCer⁺ CD24^high cells were analyzed for CD69 expression. Percentages are indicated over corresponding brackets. Similar results were obtained in three independent experiments. D. Stage of NKT cell developmental arrest in the Vα14 Tg SAP^-/- thymus. CD1d-αGalCer tetramer+ thymocytes were enriched using paramagnetic beads and submitted to FACS analysis as in 1C.

Figure 2. Expression pattern of SLAM family members on thymocytes and NKT cells

A. Double negative (DN), double positive (DP) and mature single positive (SP) thymocytes from B6 mice (or BALB/c for Ly9) were analyzed for expression levels of SLAM family receptors. DP are represented by a solid black line, SP a dashed dark grey line and DN by a light grey line as indicated. Shaded profiles represent isotype controls or staining of thymocytes lacking the corresponding gene or epitope. B. SLAM and Ly108 expression at different stages of NKT thymocyte development. Left panels, immature CD24^high; middle panels, mature CD24^lowCD44^highNK1.1^-; right panels, terminally differentiated CD24^lowCD44^highNK1.1⁺ cells in the thymus (black) and spleen (grey). The findings are summarized in the NKT cell developmental chart below the FACS panels. C. Immunohistochemical staining shows abundant SLAM and Ly108 expression in cortical and medullary cells (mostly T lineage) but not on the cortical epithelial cells. D. Flow cytometry analysis of thymic CD11b⁺ macrophages and CD11c⁺ dendritic cells (red) compared with DP thymocytes (green). The blue profile corresponds to a staining with biotin-conjugated antibody after incubation with excess unconjugated antibody (negative control).

Figure 3. Individual contributions of SLAM and Ly108 to NKT cell development

Scatter plots showing absolute numbers (log₂ scale) of CD1d-αGalCer positive cells in individual thymi, spleens and livers of 3-5 week-old SLAM ^+/- and ^-/- littermates and Ly108 ^+/- and ^-/- littermates. Because of variations in liver lymphocyte recovery between experiments, liver NKT cell numbers are enumerated per 10⁵ lymphocytes. Liver NKT cells were not examined in SLAM ^-/- mice (ND, not determined). * denotes statistical significance (p<0.05) using an unpaired t test.

Figure 4. NKT cell development in competition chimeras

A. 1:1 mixtures of WT + SLAM^-/-, WT + Ly108^-/-, WT + CD48^-/-, 2B4^-/-+ Ly108^-/-, CD48^-/-+ Ly108^-/- bone marrow were injected into lethally irradiated Jα18^-/- hosts and chimeras were analyzed at 6-8 weeks. Absolute NKT cell numbers were calculated for each CD45 allele-marked compartment, then divided by the number of lymphocytes in the corresponding CD45 allele-marked fraction and multiplied by the total number of lymphocytes to adjust for differences between the ratio of reconstitution by the two bone marrows which ranged between 0.3 and 0.7. *, ** and *** denote statistical significance (p<0.05, p<0.01 and p<0.001 respectively) using a t test for paired comparisons (2 experiments are pooled as they showed similar results (ANOVA)). B. FACS analysis of MACS-enriched CD1d-αGalCer tetramer⁺ CD24^high and CD24^low cells originating from the wt and mutant (CD45 allele-marked) bone marrows. The chimeric distribution of the tetramer-negative thymocytes is also shown (bottom dot plots).

Figure 5. Design of the pseudo-double KO chimeras

Cell interaction schemes illustrating the functional double deficiency created in the CD1^-/- compartment of mixed chimeras of the SLAM^-/- + Ly108^-/-CD1^-/- (top) vs the single deficiency in the CD1^+/+ compartment (bottom). The Ly108^-/-CD1^-/- NKT precursors must interact with SLAM^-/- thymocytes, the sole source of CD1d ligands, functionally removing both SLAM and Ly108 signals during TCR engagement (top), whereas the SLAM^-/- NKT precursors must interact with SLAM^-/- thymocytes, creating a single KO situation (bottom).

Figure 6. NKT cell developmental block in pseudo-double KO chimeras

A. Mixed radiation bone marrow chimeras as indicated. FACS dot plots are gated on CD1d-αGC+ NKT lineage cells and show expression of the CD45.1 allelic marker in tetramer-positive and negative cells in different tissues, as indicated. B. Summary scatter plots show NKT cell numbers in different tissues and in CD45 allele-marked compartments of individual mixed chimeras. *, ** and *** denote statistical significance (p<0.05, p<0.01 and p<0.001 respectively) using a t test for paired comparisons of the two hemopoietic compartments in individual mixed chimeras. Results are pooled from 2-3 sets of chimeras showing similar results (ANOVA). C. Calculated ratios of NKT cells found in the two bone marrow-derived compartments of individual chimeras.

Appendix 1. SLAM locus analysis

A. Strain-specific antibody staining. Flow cytometric analysis of bone marrow with antibody clone 2B4 (anti CD244.2) and thymi with antibody clone 30C7 (anti Ly9) of indicated mouse strains. B. Expression of SNP rs31532197 downstream of Ly108. PCR amplification of genomic DNA from indicated strains using primers Forward - GCCTTTCATCTTGGGTTTCA Reverse - AGCTGGGGAAAGGTAAGGAG followed by BSSII digest.