CD31 Expression Defines Heterogeneity of Donor Plasmacytoid Dendritic Cell Populations That Home to the Thymus: Implications for Control of GvHD

Introduction: Plasmacytoid dendritic cells (pDC) are known to possess tolerogenic properties in allogeneic bone marrow transplantation (allo-BMT), but the relationship between pDC lineage and their modulation of graft-versus-host disease (GvHD) has not been defined. Recently, we have shown that murine treatment with the pleiotropic cytokine FMS-like Tyrosine Kinase 3 Ligand (Flt3L) increases pDC content in the marrow, and that allo-BMT of Flt3L-treated marrow (FBM) grafts leads to higher overall survival and reduced GvHD in lethally irradiated hosts. Furthermore, FBM pDC have increased potency in preventing GvHD on a cell-by-cell basis (Hassan, EBMT, 2018). Interestingly, FBM pDC expressed CD11b, a myeloid-specific marker that is not expressed on currently defined pDC. Lineage ontogeny of pDC can be tracked based upon exclusive expression of CD31 and Ly6C on myeloid-progenitor derived pDC, while pDC from lymphoid progenitors express RAG1 and Siglec H. We hypothesized that treatment of murine bone marrow donors with Flt3L, which enhances their GvHD-reducing activity, modifies the distribution of lineage-specific precursors of pDC in bone marrow, and that lineage-associated pDC will have distinct biological activity in allo-BMT.

Methods and Results: To assess the effect of Flt3L on BM pDC, we treated C57BL/6 mice with Flt3L (CDX-301, 300 ug/kg) on days -1 and -4 relative to bone marrow harvest. Using flow cytometric analysis, we show that FBM pDC over-express myeloid-specific markers CD31, Ly6C and CD11b, but down-regulate the lymphoid-specific marker Siglec H (Figure 1). Moreover, we observed similar phenotypic trends for myeloid-specific markers in pDC from the marrow of RAG1KO mice, confirming the existence of a unique myeloid-derived population of pDC.

To determine the effect of pDC linage on their gene expression, we performed Illumina RNA Sequencing on human-derived pDC from Flt3L-mobilized peripheral blood and from human bone marrow. Interestingly, we found that Flt3L-treated pDC overexpress PRSS16, which is known to be exclusively expressed on cortical thymic epithelial cells. Furthermore, using GFP-transgenic mice as a source of donor pDC in B6 -> B10.BR allo-BMT, we show via confocal microscopy that donor pDC selectively home to the recipient thymus (Figure 2). Moreover, flow cytometric analysis revealed that homing of FBM pDC to the thymus is not impaired, in spite of decreased CCR9 expression on FBM pDC in comparison to BM pDC. Because it is currently believed that CCR9 expression on pDC is necessary for their homing to the thymus, these observations denote that FBM pDC use a CCR9-independent homing strategy.

Since CD31 expression is upregulated in FBM pDC with enhanced immune-regulatory activity in allo-BMT models, these data suggest that local expression of CD31 by donor pDC in the thymic microenvironment may be relevant to their ability to favorably regulate thymopoiesis and limit the development of alloreactive T cells.

Conclusions: Our data poses a link between pDC lineage and control of post-transplant GvHD, possibly via regulation of positive and negative selection of donor-derived T cells in the thymus. Specifically, we identify a CD31+ myeloid-specific population of pDC that emerges in the marrow upon treatment with Flt3L and exhibits a genetic profile that resembles thymic epithelial cells as a candidate for adoptive cell therapy to prevent GvHD.

View largeDownload slide

View largeDownload slide

Close modal