Programmed death one homolog maintains the pool size of regulatory T cells by promoting their differentiation and stability

Programmed death one homolog (PD-1H) is an immunoglobulin superfamily molecule and primarily acts as a coinhibitor in the initiation of T cell response to antigens. Here, we report that genetic ablation of PD-1H in mice blocks the differentiation of naive T cells to Foxp3+ inducible Treg cells (iTreg) with a significant decrease of iTreg in lymphoid organs. This effect of PD-1H is highly specific for iTreg because both naturally generated iTreg in gut-related tissues and in vitro induced iTreg by TGF-β were decreased whereas the genesis of natural Treg (nTreg) remains normal. The suppressive function of both iTreg and nTreg, however, is not affected by the loss of PD-1H. In addition to decreased production, PD-1H deficient iTreg could also rapidly convert to CD4+ T helper 1 or T helper 17 cells in an inflammatory environment. Our results indicate that PD-1H is required for maintenance of iTreg pool size by promoting its differentiation and preventing its conversion to other CD4+ T cell subsets. These findings may have important implications for manipulating Tregs to control inflammation.


Supplementary Figure 2: PD-1H regulates differentiation of iTreg cells in a lymphopenic environment.
(A) Naïve CD4 + CD62L hi T cells from WT mice (CD45.2) and PD-1H KO mice (CD45.1) were mixed at ratio of 1:1 and a total of 2 million cells were transferred into Rag1 KO mice (n=4); Foxp3's upregulation was determined 20 days later. This figure shows the change in ratio of WT and KO CD4 + T cells before and after transfer. The frequency of Foxp3 + cells was determined by intracellular staining. (B) Absolute numbers of CD25 + Foxp3 + T cells in the indicated organs from the Rag1 KO mice were counted. (C) Cytokine production of transferred naïve CD4 + T cells was analysed. Cells from the indicated organs were activated in vitro for 4 hours in the presence of PMA/Ionomycin/BFA and then analysed by intracellular staining. Data shown are representative of 2 independent experiments. .

Supplementary Figure 3: PD-1H plays a redundant role on the generation and suppressive function of nTreg cells. (A)
The percentage of CD25 + Foxp3 + nTreg in the spleen and LN was determined by Flow cytometry using intracellular staining for Foxp3 and cell surface staining of CD25. Six weeks old littermates from WT and PD-1H KO mice (n=5) were used. The absolute numbers of Foxp3 + CD25 + T cells in the spleen and LN were counted. (B) The expression of CD25, GITR, Lag-3, CTLA-4, ICOS and PD-1 on the nTreg from WT or PD-1H KO Foxp3(GFP) mice were determined by flow cytometry gating on the CD4 + Foxp3(GFP + ) cells. (C) Foxp3(GFP + ) nTreg from WT or PD-1H KO Foxp3(GFP) mice were sorted and co-cultured at various Treg/Teff ratios with CFSE-labelled CD8 + Teff cells in the presence of mitomycin C treated spleen cells plus anti-CD3 for 3 days. The CFSE dilution was determined by flow cytometry. Data shown are representative of at least 3 independent experiments.

Supplementary Figure 4: PD-1H determines the pool size of Treg cells in bone marrow chimeric mice. (A)
A total of 10 million mixed bone marrow cells from CD45.1 WT mice and CD45.2 mice were transferred into sub-lethally irradiated mice (CD45.1/CD45.2). Mice were analysed 10 weeks after reconstruction. The gating strategy for the WT and KO T cells was shown. (B) Foxp3 frequency in the indicated organs was determined by intracellular staining. (C) Foxp3 frequency in the spleen were summarized, and the absolute number of Foxp3 + cell and Foxp3cells were shown. Data shown are one of two independent experiments.

Supplementary Figure 6: Analysis of CD4 + Th1 and Th17 cells in recipients upon transferring with WT or PD-1H KO iTreg cells in the EAE model. (A) Naïve T cells from WT or PD-1H KO Foxp3
(GFP) mice were differentiated into iTreg cells in vitro as described above. IFN-γ and IL-4 neutralizing mAb were added to the cultures to promote the generation of iTreg. Foxp3 (GFP + ) cells were sorted and the purity was assessed by FACS. (B,C) In the EAE model after the transfer of iTreg cells, recipient CD45.1 + CD4 + T cells were re-stimulated ex vivo by PMA/Ionomycin/BFA. IFN-γ + or IL-17 + cells of recipient CD4 + T cells were determined using intracellular staining, gated on CD45.2 -CD4 + T cells. Data from a representative pair of mice were shown in (B) and a graphic plot of the data was shown from a group of 4 or 5 mice (C). Data shown were representative of at least 3 independent experiments.

Supplementary Figure 7: PD-1H deficient iTreg cells fail to retain their suppressive function and Foxp3 expression. (A)
The weight of Rag1 KO host mice (n=5 per group) at various times after the transfer of CD25 -CD45RB hi CD45.1 T cells alone or together with CD45.2 + Foxp3(GFP + ) iTreg cells from WT or KO mice (As described in Methods). Weight changes of recipient mice after transfers were normalized to their initial weight before the transfer. *P<0.05, (two-way ABOVA test). (B) H&E staining of colon tissues (Scale Bar: 100 μm) and clinical score (C) from the Rag1 KO host mice after inducing colitis (week 10, n=5).