Junctional Adhesion Molecule 2 Represents a Novel Subset of Hematopoietic Stem Cells Poised for T Lymphopoiesis

Phenotypically well-characterized hematopoietic stem cells (HSCs) still represent a heterogeneous pool of primitive cells regarding to their functionality. In particular, different lineage potential of HSCs have been considered as one of key features of the HSC heterogeneity. The lineage output of HSCs is often coupled with cell cycle status or long-term reconstitution potential, however molecular mechanisms of the mutuality are unclear and other type of the regulation may exist. In addition, prospective isolation of such HSCs biased towards specific lineage(s) is still problematic, as many of categorizations highly rely on retrospective information, e.g. transplantation assay. Although several markers have been reported to be able to subdivide HSCs into subcategories, exploration of additional markers will allow us understanding further molecular mechanisms of HSC regulations including activation and lineage choice.

Here, we show that cell surface expression of Junctional adhesion molecule 2 (Jam2) represents higher reconstitution capacity of HSCs and the T cell potential. Flow cytometry analyses revealed that a subset of CD150⁺CD48^-KSL cells in mouse bone marrow (BM) were positive for Jam2 (Jam2⁺HSC, 36.6 ±13.0 %), while other Jam family member Jam1 (F11r) was expressed on all HSCs and Jam3 was not detected. To examine functional differences of Jam2⁺ and Jam2^-HSCs, 30 cells were separately transplanted into lethally irradiated mice. Peripheral blood analyses revealed that Jam2⁺HSCs reconstituted more efficiently than Jam2^-HSCs (77.5 ±15.9 and 51.7 ±29.3 %, respectively). In case of transplantation using 5 cells, the frequency of reconstituted mice was higher in Jam2⁺HSCs (7 in 11) compared to Jam2-HSCs (4 in 11), indicating that Jam2⁺ population is more enriched for functional HSCs. The expression of Jam2 on HSC is reversible, but not hierarchical, as both Jam2⁺ and Jam2^-HSCs reconstituted opposite population in the BM.Lineage analyses revealed that Jam2⁺HSCs have a greater potential in lymphoid cell reconstitution, particularly T cells, whereas the chimerism in myeloid cells was not significantly different from Jam2^-HSCs. This tendency of higher contribution to the T cell development was even more pronounced in the secondary transplantation experiments, where the contribution of Jam2⁺HSCs in T cells was close to 100 %. Of note, most of Jam2⁺HSCs were in a dormant state, suggesting that the T cell (or lymphoid) potential of Jam2⁺HSCs is independent of cell cycle progression.

Jam2 has been reported to interact with Jam1, which mediates the Notch signaling (Kobayashi et al., Nature, 2014). Competitive co-culture of Jam2⁺ vs Jam2^-HSCs on OP9-DL1 showed that Jam2⁺HSCs dominated the T cell production, whereas no difference was seen in B cell production upon OP9 co-culture. Since Jam2 positivity correlates to T cell potential, we asked if altered T lymphopoiesis environment affects the cell surface Jam2 expression. Comparison of C57BL/6, NOD, NOD-Scid and NOD-Scid Il2rγ KO (NSG) mice showed that HSCs of NSG mice have significantly higher frequency of Jam2⁺HSCs, suggesting that cell surface Jam2 expression might be regulated by specific cytokine(s) binding to IL2Rγ.

Our findings suggest Jam2 is a new marker for a subset of HSCs that preferentially generate T cells. In addition, this work uncouples the lineage choice and cell cycle status, which proposes a novel model to the lineage-determining machineries. Since efficient and immediate generation of T cells in transplantation therapy is important to minimize infectious risks, understanding the molecular basis of the Jam-Notch cooperation would contribute to establish safer and more efficient treatment.