B7H1/CD80 interaction augments PD-1-dependent T cell apoptosis and ameliorates graft-versus-host disease

Abstract

Interactions of B7H1 (programmed death ligand 1 [PD-L1]) with its two ligands, PD-1 and CD80, on T cells play a pivotal role in controlling T cell activation, proliferation, anergy, and apoptosis. However, the interactions between the two pathways remain unknown. Using an alloimmune response model of graft-versus-host disease (GVHD), we report in this study that: 1) Comparison of proliferation and apoptosis of wild-type (WT) and PD-1(-/-)CD4(+) conventional T (Tcon) cells in WT and B7H1(-/-) recipients revealed that B7H1/CD80 interaction per se augments T cell proliferation, and this interaction augments T cell apoptosis mediated by B7H1/PD-1 interaction. This observation was recapitulated in an in vitro MLR assay. 2) Specific blockade of the B7H1/CD80 axis by anti-B7H1 mAb reduces WT-alloreactive Tcon cell proliferation, IL-2 production, expression of PD-1, and apoptosis, resulting in worsening GVHD. In contrast, specific blockade of B7H1/CD80 interaction reduces donor PD-1(-/-) Tcon cell proliferation without an impact on apoptosis, resulting in ameliorating GVHD. 3) B7H1 fused to an Ig Fc domain (B7H1-Ig), when produced in vivo by hydrodynamic injection of B7H1-Ig plasmid, ameliorates GVHD by augmenting proliferation and apoptosis of WT- alloreactive Tcon cells. Conversely, B7H1-Ig treatment has no impact on apoptosis but augments PD-1(-/-) T cell proliferation and worsens GVHD. These results indicate that B7H1/CD80 interaction augments Tcon cell proliferation, IL-2 production, and expression of PD-1, which leads to increased apoptosis mediated by the B7H1/PD-1 pathway. Additionally, by engaging both PD-1 and CD80, B7H1-Ig can be a powerful therapeutic reagent for downregulating the T cell immune response.

Figure 1. Lack of host tissue expression of B7H1 results in reduced proliferation and apoptosis of alloreactive donor CD4+ Tcon cells and worsens GVHD

TBI-conditioned BALB/c or B7H1^−/− BALB/c recipients were transplanted with CD25⁻CD8⁻ -Splenocytes (2.5×10⁶) and TCD-BM (2.5×10⁶) from GFP-Foxp3 C57BL/6 donors. A & B: Percentage of body weight changes and survival after HCT (N=8). C: Yield of donor CD4+ T cells from recipient spleen and liver at day 3 after HCT (Mean ± SE, N=4). D: Three days after HCT, mononuclear cells from the spleen of the BrdU-treated recipients were stained for donor marker H-2k^b, CD4, Annexin V, DAPI, or BrdU. Gated donor CD4⁺ T cells are shown as Annexin V versus DAPI or CD4 versus BrdU. One representative is shown of 4 replicate experiments as well as the mean ± SE of percentage of BrdU⁺CD4⁺ or Annexin V⁺DAPI⁺ cells. E & G: Yield of donor CD4⁺ T cells from recipient spleen and liver at day 6 after HCT (Mean ± SE, N=4). F & H: Representative staining patterns and mean ± SE of percentage of BrdU⁺CD4⁺ T cells or Annexin+CD4+ T cells in the spleen and liver 6 days after HCT (N=4). *p<0.05,**p<0.01, ***p<0.001.

Figure 2. Lack of host tissue expression of B7H1 results in reduced proliferation of PD-1^−/− Tcon cells and amelioration of GVHD

TBI-conditioned BALB/c or B7H1^−/− BALB/c recipients were transplanted with CD25⁻CD8⁻ -Splenocytes (0.25×10⁶) and TCD-BM (2.5×10⁶) from PD-1^−/−C57BL/6 donors. A & B: Percentage of body weight changes and survival after HCT (N=12). C & E: Yield of donor CD4⁺ T cells in the spleen and liver 8 days after HCT. D & F: Mononuclear cells from the spleen or liver of the recipients were stained for donor CD4, Annexin V, DAPI, or BrdU. The mean (±SE) of percentage of Brdu⁺CD4⁺ T cells or Annexin⁺CD4⁺ T cells in the spleen or liver are shown (N= 4). Representative flow cytometry staining patterns show gated donor CD4⁺ T cells in Annexin V versus DAPI or CD4 versus BrdU. The mean (±SE) of percentage of Brdu⁺CD4⁺ T cells or annexinv⁺CD4⁺ T cells in the spleen or liver are also shown (N= 4). *p<0.05 **p<0.01.

Figure 3. Specific blockade of B7H1/B7.1 interaction in the presence of PD-1 reduces proliferation and apoptosis of alloreactive donor CD4+ Tcon cells and worsens GVHD

TBI-conditioned BALB/c recipients were transplanted with CD25-CD8- splenocytes (2.5×10⁶) and TCD-BM (2.5×10⁶) from GFP-Foxp3 C57BL/6 donors. Five days after transplantation, recipients were treated daily with anti-B7H1 (43H12) that specifically blocks B7H1-CD80 interaction or Rat IgG at a dose of 200 μg/mouse for 2 days. Donor CD4+ T cell yield, proliferation with BrdU incorporation, and apoptosis in the spleen and liver were measured 24 hours after first anti-B7H1 (43H12) injection. A & B: Percentage of body weight changes and survival after HCT (N=10). C & E: Yield of donor CD4+ T cells in the spleen and liver 6 days after HCT. D & F: Mononuclear cells from the spleen or liver of the recipients were stained for donor CD4, Annexin V, DAPI, or BrdU. The mean (±SE) of percentage of Brdu⁺CD4⁺ T cells or annexin⁺CD4⁺ T cells in the spleen or liver are shown (N= 4). G&H: Donor CD4⁺ T cells from the spleen of BALB/c recipients given rat IgG or anti-B7H1 (43H12) antibodies were sorted 6 days after HCT, which is 24 hours after first anti-B7H1 injection. Total RNA and protein were isolated and expression level of Caspase 3 and Bcl-xl were analyzed with real time PCR (Panel G) and western blot (Panel H). I: Total RNA was isolated from sorted donor CD4+ T cells from the spleen of BALB/c recipients given rat IgG or anti-B7H1 (43H12) antibodies. And expression level of PD-1 was measured with real time PCR. The expression levels relative to GAPDH is shown. Mean ± SE from three replicate experiments is shown. J: PD-1 expression on donor CD4+ T cells was measured with Flow Cytometry. Gated donor CD4+ T cells were shown in histogram of PD-1. The filled gray area represents negative control, the dashed or solid lines respectively represent T cells from recipients treated with anti-B7H1 or control rat IgG. The mean (±SE) of fluorescence of PD-1 is shown; N=4. *p<0.05,**p<0.01, ***p<0.001.

Figure 4. Specific blockade of B7H1/CD80 interaction in the presence of PD-1 reduces IL-2 and increases IL-10 production by activated donor CD4+ Tcon cells

TBI-conditioned BALB/c recipients were given HCT and treated with anti-B7H1 as described in Figure 3. 24 hours after antibody blockade spleens were harvested and the percentage of IL-2⁺, IL-10⁺, IFN-γ⁺, and TNF-α⁺ cells among donor CD4+ Tcon cells were analyzed by intracellular staining. A: Gated donor CD4+ T cells are shown as CD4 versus IL-2, IL-10, IFN-γ or TNF-α. One representative of 4 replicate experiments and the mean (± SE) of percentage of IL-2⁺, IL-10⁺, IFN-γ⁺, and TNF-α⁺ cells among donor CD4⁺ T cells are shown. B: mRNA levels of IL-2, IL-10, IFN-γ and TNF-α in sorted donor CD4⁺ T cells 24 hours after rat IgG or B7H1 (43H12) treatment were evaluated with real time PCR. The expression levels relative to GAPDH is shown. Mean ± SE from three replicate experiment is shown. C: Serum levels of IL-2, IL-10, IFN-γ, and TNF-α in rat IgG treatment group and B7H1 (43H12) treatment group were measured with ELISA. The mean (± SE) of IL-2, IL-10, IFN-γ, TNF-α concentration in the serum are shown (N=4). *p<0.05,**p<0.01, ***p<0.001.

Figure 5. Specific blockade B7H1/CD80 interaction in the absence of PD-1 reduces proliferation without impact on activation-induced apoptosis of alloreactive donor CD4⁺ Tcon cells and ameliorates acute GVHD

TBI-conditioned BALB/c recipients were transplanted with CD25⁻CD8⁻ splenocytes (2.5×10⁶) and TCD-BM (2.5×10⁶) from PD1^−/−C57BL/6 donors. Five days after transplantation, recipients were treated with anti-B7H1 (43H12) that specifically blocks B7H1/CD80 interaction or Rat IgG. Donor CD4⁺ T cell yield, proliferation with BrdU incorporation, and apoptosis in the spleen and liver were measured 8 days after HCT. A & B: Percentage of body weight changes and survival after HCT (N=8). C & E: Yield of donor CD4⁺ T cells in the spleen and liver 8 days after HCT. D & F: Mononuclear cells from the spleen or liver of the recipients were stained for donor CD4, Annexin V, DAPI, or BrdU. The mean (±SE) of percentage of Brdu⁺CD4⁺ T cells or Annexin⁺CD4⁺ T cells in the spleen or liver are shown (N= 4). Representative flow cytometry staining patterns shows gated donor CD4⁺ T cells in Annexin V versus DAPI or CD4 versus BrdU. The mean (±SE) of percentage of Brdu⁺CD4⁺ T cells or annexin⁺CD4⁺ T cells in the spleen or liver are also shown (N= 4). *p<0.05, ***p<0.001.

Figure 6. Anti-B7H1 treated PD-1^−/− alloreactive CD4⁺ Tcon cells have reduced production of IL-2 and TNF-α, but no significant changes in their IL-10 or IFN-γ production

TBI-conditioned BALB/c recipients were given HCT and anti-B7H1 treatment as described in Fig.4. Three days after antibody blockade spleens were harvested and the percentage of IL-2⁺, IL-10⁺, IFN-γ⁺, and TNF-α⁺ cells among donor CD4⁺ Tcon cells were analyzed by intracellular staining. A: Gated donor CD4⁺ T cells are shown as CD4 versus IL-2, IL-10, IFN-γ or TNF-α. One representative of 4 replicated intracellular staining experiments, the mean (± SE) of percentage of IL-2⁺, IL-10⁺, IFN-γ⁺, and TNF-α⁺ cells among donor CD4⁺ T cells are shown. B: Serum levels of IL-2, IL-10, IFN-γ, and TNF-α in rat IgG treatment group and B7H1 (43H12) treatment group were measured with ELISA. The mean (± SE) of IL-2, IL-10, IFN-γ, and TNF-α concentration in the serum are shown (N=4). *p<0.05,**p<0.01, ***p<0.001.

Figure 7. B7H1/CD80 interaction augments alloreactive CD4⁺ Tcon cell proliferation and apoptosis in the presence of PD-1 but augment proliferation only in the absence of PD-1 in an in vitro mixed lymphocyte reaction assay

Sorted CD25⁻CD4⁺ Tcon cells from WT or PD-1^−/− donor-type C57BL/6 mice were labeled with CFSE. The Tcon cells (0.1×10⁶) were co-cultured with CD11c⁺ DC cells (0.1×10⁶) from WT or B7H1^−/− BALB/c mice. Three days after culture, cells were harvested and stained for CD4, Annexin V and PD-1. Concentration of IL-2 in the supernatant were also measured. A: Gated WT CD4+ Tcon cells are shown as CFSE versus annexin V. B: PD-1 expression level of CD4+ Tcon cells and IL-2 concentration in supernatant. C: Gated PD−/− CD4+ Tcon cells are shown as histogram of CFSE or DAPI versus Annexin V. One representative flow cytometry pattern is shown of 3 independent experiments with triplicate wells. Mean ± SE of 9 samples in each group is shown. *p<0.05, **p<0.01, ***p<0.001

Figure 8. B7H1-Ig secreted by hepatocytes after hydrodynamic injection of B7H1-plasmid ameliorates acute GVHD by augmenting apoptosis of donor CD4⁺ Tcon cells in the presence of PD-1

B7H1^−/− BALB/c mice were given hydrodynamic injection of B7H1- plasmid at a concentration of 10μg/ml in PBS at the volume of 10% body weight. Hepatocytes and serum of the treated mice were harvested at indicated time points. A: relative mRNA expression level of B7H1 plasmid in hepatocytes (N=4). B: Serum levels of B7H1-Ig measured with ELISA (N=4). C & D: Five days after hydrodynamic injection of B7H1 or control plasmid, the B7H1^−/− BALB/c were conditioned with TBI and transplanted with CD25⁻CD8⁻ splenocytes (2.5×10⁶) and TCD-BM (2.5×10⁶) from GFP-Foxp3 C57BL/6 donor. Recipients were monitored for bodyweight changes and survival (N=8). Six days after HCT, yield of donor CD4⁺ T cells in the spleen and liver as well as donor CD4⁺ T cell proliferation with BrdU labeling and apoptosis were measured as described above. E-H: 6 days after HCT, mononuclear cells from the spleen or liver of the recipients were stained for H-2K^b, CD4, Annexin V, DAPI, or BrdU. Panels E&G shows yield of donor CD4⁺ T cells from recipients spleen and liver at day 6 after HCT (N=4). Panels F&H: shows mean (±SE) of percentage of donor Brdu⁺CD4⁺ T cells among donor CD4⁺T cells in the spleen or liver and mean (±SE) of percentage of donor annexin⁺CD4⁺ T cells among donor CD4⁺ T cells in the spleen or liver (N=4). *p<0.05,**p<0.01, ***p<0.001.

Figure 9. In the absence of PD-1, B7H1-Ig produced by hepatocytes after hydrodynamic injection of B7H1-plasmid augmentes alloreactive CD4⁺ Tcon cell proliferation without impact on apoptosis

B7H1^−/− BALB/c mice were given hydrodynamic injection of B7H1- plasmid as described in Figure 6. Five days after hydrodynamic injection of B7H1 or control plasmid, the B7H1^−/− BALB/c were conditioned with TBI and transplanted with CD25⁻CD8⁻ splenocytes (0.25×10⁶) and TCD-BM (2.5×10⁶) from PD1^−/− C57BL/6 donors. Recipients were monitored for bodyweight changes and survival. 8 days after HCT, yield of donor CD4⁺ T cells in the spleen and liver as well as donor CD4⁺ T cell proliferation with BrdU labeling and apoptosis were measured described above. A & B: percentage of survival and body weight changes (N=8). C-F: 6 days after HCT, mononuclear cells from the spleen or liver of the recipients were stained for H-2K^b, CD4, Annexin V, DAPI, or BrdU. C&E: Yield of donor CD4⁺T cells from recipients spleen and liver at 6 days after HCT (N=4). D&F: Representative flow cytometry staining patterns shown as Annexin V versus DAPI or CD4 versus Brdu after gating on donor CD4⁺ T cells, as well as Mean (±SE) of percentage of donor Brdu⁺CD4⁺ T cells among donor CD4⁺ T cells in the spleen or liver. *p<0.05,**p<0.01.

Figure 10. Impact of B7H1/CD80 interaction on T cell proliferation, PD-1-dependent T cell apoptosis, and GVHD severity

A: Activated alloreactive Tcon cells upregulate expression of PD-1 and CD80 that interact with B7H1 expressed by host tissue cells (e.g. APCs) after allogeneic HCT. The interactions cannot prevent induction of GVHD. B: Blockade of B7H1/CD80 interaction reduces alloreactive Tcon cell proliferation and apoptosis, and reduction of apoptosis outweighs reduction of proliferation, resulting in expansion of alloreactive Tcon cells and worsening GVHD. C: In the absence of PD-1 (i.e. lack of B7H1/PD-1 interaction), B7H1/CD80 interaction per se augments alloreactive Tcon cell proliferation with no impact on apoptosis, resulting in expansion of alloreactive T cells and worsening GVHD.D: In the absence of PD-1, blockade of B7H1/CD80 interaction reduces Tcon cell proliferation and reduces Tcon expansion and ameliorates GVHD.