OX40 costimulation turns off Foxp3+ Tregs

Abstract

OX40 is a recently identified T-cell costimulatory molecule that belongs to the TNF/TNFR superfamily. OX40 can be expressed by both activated T effector cells and Foxp3(+) Tregs. It is well known that OX40 delivers a potent costimulatory signal to T effector cells, but very little is known about the role of OX40 in regulating the suppressor properties of Foxp3(+) Tregs and the de novo generation of new inducible Foxp3(+) Tregs from T effector cells. In the present study, we found, by using a newly created foxp3gfp knockin model, that OX40 was dispensable for the genesis and suppressor functions of naturally arising CD4(+)Foxp3(+) Tregs, but stimulating OX40 on the Foxp3(+) Tregs abrogated their ability to suppress T effector cell proliferation, IFN-gamma production, and T effector cell-mediated allograft rejection. OX40 costimulation did not significantly affect proliferation and survival of the naturally arising Foxp3(+) Tregs, but profoundly inhibited Foxp3 gene expression. Importantly, OX40 costimulation to T effector cells prevented the induction of new inducible Foxp3(+) Tregs from T effector cells. Our study identified OX40 as a key negative regulator of Foxp3(+) Tregs and may have important clinical implications in models of transplantation and autoimmunity.

Figure 1

OX40 expression by CD4⁺Foxp3⁺ Tregs and CD4⁺Foxp3^— T effector cells. (A) Spleen cells from naive foxp3gfpKI mice were stained with cychrome-anti-CD4 and PE-anti-OX40. CD4⁺GFP(Foxp3)⁺ cells and CD4⁺GFP(Foxp3)⁻ cells were selectively gated and expression of OX40 on the Foxp3⁺ and Foxp3⁻ subsets was analyzed. A representative plot of 5 experiments is shown. (B) FACS sorted CD4⁺ GFP(Foxp3)⁻ T effector cells were stimulated in vitro with anti-CD3 and anti-CD28, and expression OX40 on the activated T cells was analyzed and shown. Naive CD4⁺ T effector cells were included as a control (the solid histogram). (C) The sorted CD4⁺ GFP(Foxp3)⁻ T effector cells were stimulated with anti-CD3/anti-CD28 plus TGF-β for 4 days, and the induction of GFP(Foxp3)⁺ cells was determined. Expression of OX40 on the GFP(Foxp3)⁻ T effector cells and the converted GFP(Foxp3)⁺ Tregs was shown. (D) CD4⁺GFP(Foxp3)⁻ T effector cells were cocultured with the natural CD4⁺GFP(Foxp3)⁺ Tregs or the converted CD4⁺GFP(Foxp3)⁺ Tregs. The cell mixture was stimulated with anti-CD3 plus APCs, and suppression of T effector cell proliferation was shown as mean (CPM ± SD) of triplicate assays. Representative data of 3 individual experiments are shown.

Figure 2

Role of OX40 in the genesis and suppressor functions of CD4⁺Foxp3⁺ Tregs. (A). Real-time RT-PCR analysis of Foxp3 gene transcripts in CD4⁺CD25⁺ Tregs sorted from wt C57BL/6 and OX40KO mice. Data shown are representative of 3 individual experiments. (B) Spleen (SP) and lymph node (LN) cells from wt foxp3gfpKI mice and OX40KO foxp3gfpKI mice were stained with cychrome–anti-CD4 and then compared for the presence of CD4⁺GFP(Foxp3)⁺ T cells by FACS. The data shown are representative data of 4 individual experiments. Numbers in quadrants are the relative percentage of cells in each region given by the flow cytometer. (C) CD4⁺GFP⁻ T effector cells sorted from foxp3gfpKI mice were mixed with CD4⁺GFP(Foxp3)⁺ Tregs from wt and OX40KO foxp3gfpKI mice at different ratios, and suppression of T effector cell proliferation was shown. The data shown are representative of 4 individual experiments. Error bars represent SD of triplicate assays.

Figure 3

Effect of OX40 costimulation on the induction of new CD4⁺Foxp3⁺ Tregs from T effector cells. (A) CD4⁺GFP(Foxp3)⁻ T effector cells were sorted from wt foxp3gfpKI mice and OX40KO foxp3gfpKI mice, The T effector cells were stimulated with anti-CD3 plus APCs in the presence or absence of TGF-β for 2 to 5 days. Induction of new GFP(Foxp3)⁺ T cells in the CD4⁺ fraction was determined by FACS. The dot plot shown is one of 3 individual experiments 4 days after the culture. Numbers in quadrants are the relative percentage of cells in each region given by the flow cytometer. (B) Induction of new GFP(Foxp3)⁺ Tregs calculated from 3 individual experiments. The conversion shown is the mean ± SD of 3 independent experiments at different time points. (C) Suppression of T effector cell proliferation by CD4⁺GFP(Foxp3)⁺ Tregs converted from either the wt or the OX40KO T effector cells. Data shown are representative of 3 independent experiments. Error bars represent SD of triplicate assays.

Figure 4

Stimulation of OX40 on CD4⁺GFP(Foxp3)⁺ Tregs abrogates their suppressor functions. (A) CD4⁺GFP(Foxp3)⁻ T effector cells sorted from wt and OX40KO foxp3gfpKI mice were stimulated with anti-CD3 plus wt APCs. T effector cell proliferation in the presence or absence of wt CD4⁺GFP(Foxp3)⁺ Tregs at different Tregs to T effector ratios was shown. Data shown are mean (CPM ± SD) of triplicate assays. (B) OX40 deficient CD4⁺GFP(Foxp3)⁻ T effector cells were stimulated with anti-CD3 plus wt APCs or OX40Ltg APCs in the presence or absence of wt CD4⁺GFP(Foxp3)⁺ Tregs. Cell proliferation was determined 3 days later by 3H-TdR uptake. Data shown are mean (CPM ± SD) of triplicate assays. (C) OX40 deficient CD4⁺GFP(Foxp3)⁻ T effector cells were stimulated with anti-CD3 plus OX40Ltg APCs. In these cultures, graded numbers of wt or OX40KO CD4⁺GFP(Foxp3)⁺ Tregs were added, and cell proliferation was determined 3 days later by 3H-TdR uptake. Data shown are mean (CPM ± SD) of triplicate assays. In all the suppression assays, representative data of 3 independent experiments are shown.

Figure 5

Suppression of IFN-γ production by CD4⁺GFP(Foxp3)⁺ Tregs with or without OX40 stimulation. (A) CD4⁺GFP(Foxp3)⁻ T effector cells sorted from OX40KO foxp3gfpKI mice were stimulated with anti-CD3 plus wt APCs or OX40Ltg APCs, IFN-γ production by the T effector cells in the presence or absence of wt CD4⁺GFP(Foxp3)⁺ Tregs was analyzed by ELISPOT assay. Data shown are mean (± SD) of 3 experiments. (B) OX40 deficient CD4⁺GFP(Foxp3)⁻ T effector cells were stimulated with anti-CD3 plus OX40Ltg APCs. In these cultures, wt or OX40 deficient CD4⁺GFP(Foxp3)⁺ Tregs were added as indicated, and suppression of IFN-γ production was shown. Data shown are mean (± SD) of 3 independent experiments. (C) CD4⁺GFP(Foxp3)⁻ T effector cells from OX40KO foxp3gfpKI mice were transferred into syngeneic Rag^−/− hosts (5 × 10⁵ cells/mouse), groups of host mice were also transferred with equal number of wt CD4⁺GFP(Foxp3)⁺ Tregs. The host mice were then grafted with DBA/2 skin grafts and treated with an agonist anti-OX40 mAb, and skin allograft survival was shown. (*) P < .05

Figure 6

Effect OX40 stimulation on survival and Foxp3 expression of CD4⁺GFP(Foxp3)⁺ Tregs. (A) Sorted CD4⁺GFP(Foxp3)⁺ Tregs were stimulated with anti-CD3 plus wt and OX40Ltg APCs. Cells were stained with PE-annexin V 24 hours later and analyzed by FACS. Freshly prepared Foxp3⁺ Tregs stained with PE-annexin V were included as a control. The annexin V profile in the GFP⁺ fraction was shown. (B) Sorted CD4⁺GFP(Foxp3)⁺ Tregs were stimulated with anti-CD3 and wt or OX40Ltg APCs. Foxp3 gene transcripts were analyzed 4 days later by real-time PCR. Data shown are mean A.U. of 4 experiments in each group. Error bars represent the SD of triplicate assays. (C) Sorted CD4⁺GFP(Foxp3)⁺ Tregs were stimulated with anti-CD3 and wt or OX40Ltg APCs, levels of GFP(Foxp3) expression in the CD4⁺ fraction were determined by FACS 4 days later and shown. The plot shown is one of 3 experiments. (A,C) Numbers are the relative percentage of cells in each region given by the flow cytometer.

Figure 7

Effect of OX40 stimulation to T effector cells on the induction of new Foxp3⁺ Tregs.(A) CD4⁺GFP(Foxp3)⁻ T effector cells were sorted from foxp3gfpKI mice and stimulated in vitro with anti-CD3 plus wt APCs, OX40Ltg APCs or OX40L^−/− APCs in the presence or absence of TGF-β, and induction of GFP(Foxp3)⁺ cells was determined 4 days later by gating onto the CD4⁺ fraction. The FACS plot shown is the representative data of 4 individual experiments. (B) The experiments were set up as described in panel A, and the image shown was captured by confocal microscopy 4 days after the culture using a Nikon Eclipse 80i system equipped with E-max software (Nikon Instruments, Melville, NY) (40×/0.75 NA oil immersion lens). Cells were labeled with PE-antimouse CD4. (C) CD4⁺CD25⁻ T effector cells were sorted from congenic CD90.1 mice and adoptively transferred into wt C57BL/6 and OX40Ltg mice (CD90.2). The host mice were treated with DST and anti-CD154. Induction of Foxp3 expression in the CD90.1⁺ fraction in the host spleen was determined by intracellular staining for the Foxp3 protein 6 days later. Data shown are representative of 3 experiments. (A,C) Numbers are the relative percentage of cells in each region given by the flow cytometer.