Ex vivo modulation of the Foxo1 phosphorylation state does not lead to dysfunction of T regulatory cells

Abstract

Peripheral regulatory CD4+ T cells (Treg cells) prevent maladaptive inflammatory responses to innocuous foreign antigens. Treg cell dysfunction has been linked to many inflammatory diseases, including allergic airway inflammation. Glucocorticoids that are used to treat allergic airway inflammation and asthma are thought to work in part by promoting Treg cell differentiation; patients who are refractory to these drugs have defective induction of anti-inflammatory Treg cells. Previous observations suggest that Treg cells deficient in the transcription factor FoxO1 are pro-inflammatory, and that FoxO1 activity is regulated by its phosphorylation status and nuclear localization. Here, we asked whether altering the phosphorylation state of FoxO1 through modulation of a regulatory phosphatase might affect Treg cell function. In a mouse model of house dust mite-induced allergic airway inflammation, we observed robust recruitment of Treg cells to the lungs and lymph nodes of diseased mice, without an apparent increase in the Treg cytokine interleukin-10 in the airways. Intriguingly, expression of PP2A, a serine/threonine phosphatase linked to the regulation of FoxO1 phosphorylation, was decreased in the mediastinal lymph nodes of HDM-treated mice, mirroring the decreased PP2A expression seen in peripheral blood monocytes of glucocorticoid-resistant asthmatic patients. When we asked whether modulation of PP2A activity alters Treg cell function via treatment with the PP2A inhibitor okadaic acid, we observed increased phosphorylation of FoxO1 and decreased nuclear localization. However, dysregulation of FoxO1 did not impair Treg cell differentiation ex vivo or cause Treg cells to adopt a pro-inflammatory phenotype. Moreover, inhibition of PP2A activity did not affect the suppressive function of Treg cells ex vivo. Collectively, these data suggest that modulation of the phosphorylation state of FoxO1 via PP2A inhibition does not modify Treg cell function ex vivo. Our data also highlight the caveat in using ex vivo assays of Treg cell differentiation and function, in that while these assays are useful, they may not fully recapitulate Treg cell phenotypes that are observed in vivo.

Fig 1. HDM potently induces allergic airway inflammation despite significant T regulatory cell recruitment.

(a) Allergic airway inflammation was induced with three separate priming and challenge doses of house dust mite (HDM) over 14 days. Disease severity was assessed at day 15. (b) Histological assessment of inflammation in lung sections of representative HDM-treated and PBS-treated mice. * in H&E images indicates inflammatory infiltrate and arrowheads in PAS images indicate mucus. (c) Eosinophils were quantified in the bronchoalveolar lavage (BAL) by flow cytometry (CD45⁺Cd11b⁺SiglecF⁺CD11c⁻). Treg (CD4⁺CD25⁺FoxP3⁺) Teff (CD4⁺CD25⁺FoxP3⁻) populations in the lung parenchyma (d) and mediastinal lymph nodes (MLN) (e) were quantified by flow cytometry. (f) IL-10 concentration in BAL was quantified by ELISA. (g) Schematic of PP2A-FoxO1 axis. (h) PP2Acα sub-unit transcript levels were measured by RT-PCR. Transcript levels in HDM-treated mice were normalized to PBS-treated controls. Numbers in the bars of c, d, e, f and h represent the number of animals analyzed. Significance in all panels was determined by two-tailed students’ T-test: *<0.05; **<0.01; ****<0.0001.

Fig 2. PP2A inhibition increases FoxO1 phosphorylation and cytosolic localization in CD4⁺ T cells.

(a) Primary CD4⁺ T cells were negatively selected from the lymph nodes of 6–8 week old mice by magnetic cell sorting. CD4⁺ T cells were treated with the indicated dosages of okadaic acid for up to 48 hours in cultures supplemented with 1μg/mL anti-CD3. (b) Western blot analysis of FoxO1 phosphorylation at Thr24 and Akt at Ser 473. (c) Western blot analysis of FoxO1 localization in fractionated CD4 cells. Beta-actin and HDAC were used as cytoplasmic and nuclear controls respectively. (d) Viability of CD4⁺ T cells in culture after 48 hours with the indicated concentrations of okadaic acid. (e) Western blot analysis of in vitro phosphatase assay. Immunoprecipitated FoxO1 was treated with PP2A +/- 100nM okadaic acid. b, c and d are representative data from 3 individual experiments; e is representative of 2 experiments.

Fig 3. Okadaic acid increases phosphorylation and cytosolic localization of FoxO1 in T regulatory cells.

(a) Primary CD4⁺ T cells were isolated by negative magnetic selection from the lymph nodes of FoxP3-GFP mice. Cells were cultured on plates coated with anti-CD3 and anti-CD28 and treated with TGFβ to initiate Treg skewing. Okadaic acid was added after 24 hours in culture. After 72 hours in culture, cells were removed from TCR stimulus for 2 hours prior to protein isolation. (b) Treg differentiation was assessed at the end of the assay by GFP expression. (c) Representative western blot of FoxO1 phosphorylation at Thr24 in Treg cells. Data are representative of 3 independent experiments. (d) Representative western blot of FoxO1 in fractionated Treg cells. Beta actin and HDAC were used as cytosolic and nuclear controls respectively. Bar graph shows summary localization data from 3 experiments as measured by band density and normalized to cytosolic control.

Fig 4. Okadaic acid does not impair T regulatory cell differentiation in vitro.

(a) Naïve CD4 cells were isolated from the lymph nodes of FoxP3-GFP mice and differentiated in to Tregs for 72 hours. (b) GFP and (c) CD25 expression were assessed by flow cytometry after 72 hours of Treg skewing. Data were compiled from 3 independent experiments with a total n = 6 mice per group. (d) Primary CD4 cells were isolated from the lymph nodes of FoxP3 GFP mice and pre-treated with okadaic acid for 48 hours prior to Treg differentiation. GFP expression was assessed by flow cytometry at 24, 48 and 72 hours after the initiation of skewing. Data were compiled from 3 independent experiments with n = 6 mice per group. Statistical analysis of b and c was performed using a two-tailed students’ T-test–comparison of vehicle control and okadaic acid treated conditions: FoxP3 p = 0.7; CD25 p = 0.12. Statistical analysis of e was performed by 2-way ANOVA–p-interaction = 0.19.

Fig 5. Okadaic acid does not induce a pro-inflammatory phenotype or affect lymph node homing signatures in Treg cells.

Primary CD4⁺ T cells were differentiated into Tregs in the presence of okadaic acid for 72 hours. (a) After differentiation, IFNγ and IL-17 production and (b) CCR7 and CD62L expression were measured by flow cytometry. Flow panels of cytokine expression are representative data from 3 independent experiments. Surface marker expression was compiled from 2 independent experiments with n = 3 mice total. Data in b was analyzed by non-parametric students’ T-test: CD62L p = 0.5; CCR7 p = 0.77.

Fig 6. Okadaic acid does not impair T regulatory cell mediated suppression of CD4 effector cells.

(a) Teff and Tregs were isolated from the lymph nodes and spleen of FoxP3-GFP mice. Cells were co-cultured at indicated ratios in the presence of antigen presenting cells and 0.15 μg/mL anti-CD3 for 96 hours +/- okadaic acid. (b) CD4⁺ T cell proliferation in the presence of okadaic acid was measured by cell-trace violet dye dilution. Data is representative of 4 independent experiments. (c) Western blot of FoxO1 phosphorylation at Thr24 in CD4⁺ T cells treated with okadaic acid for 96 hours. Data is representative of 2 independent experiments. (d) Proliferation index of Teff cells in the presence of increasing numbers of Tregs was assessed by cell-trace violet dye dilution. Data is compiled from 4 independent experiments–statistical analysis was performed by 2-way ANOVA; p-interaction = 0.44.