mPGES-1-Mediated Production of PGE2 and EP4 Receptor Sensing Regulate T Cell Colonic Inflammation

Abstract

PGE₂ is a lipid mediator of the initiation and resolution phases of inflammation, as well as a regulator of immune system responses to inflammatory events. PGE₂ is produced and sensed by T cells, and autocrine or paracrine PGE₂ can affect T cell phenotype and function. In this study, we use a T cell-dependent model of colitis to evaluate the role of PGE₂ on pathological outcome and T-cell phenotypes. CD4⁺ T effector cells either deficient in mPGES-1 or the PGE₂ receptor EP4 are less colitogenic. Absence of T cell autocrine mPGES1-dependent PGE₂ reduces colitogenicity in association with an increase in CD4⁺RORγt⁺ cells in the lamina propria. In contrast, recipient mice deficient in mPGES-1 exhibit more severe colitis that corresponds with a reduced capacity to generate FoxP3⁺ T cells, especially in mesenteric lymph nodes. Thus, our research defines how mPGES-1-driven production of PGE₂ by different cell types in distinct intestinal locations impacts T cell function during colitis. We conclude that PGE₂ has profound effects on T cell phenotype that are dependent on the microenvironment.

Keywords: IBD–inflammatory bowel diseases; PGE2; T cell; Th17 & Tregs cells; Th17 activation; Treg = regulatory T cell; colitis; inflammation immunomodulation.

Figure 1

Basal production of PGE₂ in the colon is regulated by COX-2 and mPGES-1 with significant contribution from cells of the adaptive immune system. (A) Colon explants from WT (blue), mPGES-1^−/− (red), Rag1^−/− (black) or Rag1^−/− × mPGES-1^−/− (red/black) mice were cultured for 12 h and their supernatants were analyzed for basal PGE₂ production. The specific COX2 inhibitor NS-398 was added to evaluate the differential contribution of COX2 and COX1 to the PGH₂ pool prior to PGE₂ biosynthesis. (B) PGE₂ production in colon explants from untreated WT mice (“Untreated,” blue circles) or Rag1^−/− mice undergoing WT Teff cell-driven colitis at week 10 post-transfer (“Colitic,” black squares). ** indicates a P value < 0.05 using a 2-tailed heteroscedastic Student's T-test. (C) Immunohistochemistry of COX2 (upper row) and mPGES-1 (lower row) in the colon of mice that transferred with WT Teff cells (CD4⁺CD25⁻CD45RB^hi). Scale bar is 100 μm long. Arrowheads 1 and 2 point to localization within the LP, while arrowhead 3 indicates localization in the brush border. ** Significant at P < 0.05 and *** at P < 0.01 in a one-way ANOVA with Welch's correction.

Figure 2

mPGES-1-deficiency in T effector cells protects against colitis. (A) Weight loss in Rag1^−/− mice that received transfer of 1 × 10⁶ CD4⁺CD25⁻CD45RB^hi T cells (Teff) from WT or mPGES-1^−/− donors. The dashed lines correspond to mice that received WT Teff and WT or mPGES-1^−/− CD4⁺CD25⁺ (Treg) co-transfers. (B) Colon pathology scores from cohorts receiving transfers of WT or mPGES-1^−/− Teff cells. Flow cytometry analysis of the (C) mesenteric lymph nodes (mLN) and (D) colon lamina propria (cLP) CD4⁺ populations at the end of the experiment (week 10), with representative dot plots indicating intracellular expression of RORγt and FoxP3 and graphs below indicating summarized results from 4 experiments. ** indicates a significant difference with P < 0.05 using a two-tailed heteroscedastic Student's T-test between the WT and mPGES-1^−/− Teff donor groups.

Figure 3

Deficiency in CD4-intrinsic mPGES-1 impairs Teff CD4⁺ cell expansion but enhances Treg localization and RORγt expression in the colonic lamina propria. (A,B) Rag1^−/− recipient mice received a co-transfer of a 1:1 mix of CD45.1⁺ WT (blue) and CD45.2⁺ mPGES-1^−/− (red) Teff cells. Flow cytometric analysis of the (A) mLN and (B) cLP CD4⁺ populations, with representative dot plots indicating intracellular expression of CD45.1 or CD45.2 congenic marker expression together with RORγt. In the cLP plot (B), the shaded box indicates a unique RORγt^hi population of mPGES-1 deficient cells in the cLP. Graphs on the right indicate the proportions and total numbers for each group. (C) Co-transfer of either CD45.1⁺ WT Treg with CD45.2⁺ mPGES-1^−/− Teff cells or CD45.2⁺ mPGES-1^−−/− Treg cells with CD45.1⁺ WT Teff into Rag1^−/− recipients. Transfers were always performed with a 2:1 Teff:Treg ratio. In the cLP, mPGES-1^−/− CD4⁺ T cells are able to acquire higher RORγt expression than WT cells (shaded boxes). These CD4⁺RORγt^hi cells arise from both mPGES-1^−/− Teff cells and mature Treg cells. Graphs on the bottom show the proportions of WT or mPGES-1^−/− Treg cells that are either RORγt⁻ or RORγt⁺ in the mLN or the cLP. **P < 0.05 using in a one-way ANOVA with Welch's correction.

Figure 4

EP4 deficient CD4⁺ T effector cells have severely blunted colitogenic potential due to impaired proliferative capacity. (A) Weight loss in Rag1^−/− mice that received transfer of Teff cells from EP4^fl/fl (EP4^WT) or CD4^Cre × EP4^fl/fl (EP4^ΔCD4) donor mice. (B) Colon pathology scores from both cohorts. Flow cytometry analysis of the (C) mLN and (D) cLP CD4⁺ populations at the end of the experiment (week 10), indicating intracellular expression of RORγt and FoxP3 in summarized results from 4 experiments. (E) Colon images detailing lamina propria T cell infiltrates in both groups, with magnified inserts on the right-hand side. Blue = DAPI, Gray = CD3ε, Green = RORγt, and Red = FoxP3. **P < 0.01 using in a one-way ANOVA with Welch's correction.

Figure 5

Paracrine mPGES-1-deficiency in non-lymphoid cells facilitates colitis by inhibiting generation of CD4⁺FoxP3⁺ cells. (A) Weight loss in Rag1^−/− or Rag1^−/− x mPGES-1^−/− mice that received a transfer of WT Teff donor cells. (B) Segregated colon pathology scores from both cohorts. Flow cytometry analysis of (C) mLN and (D) cLP CD4⁺ populations at the end of the experiment (10 weeks), indicating intracellular expression of RORγt and FoxP3 in summarized results from 3 experiments. (E) Fluorescence microscopy analysis of colon sections denoting CD4⁺ cell infiltrates. Blue = DAPI, Gray = CD3ε, Green = RORγt, and Red = FoxP3. **P < 0.01 and *P < 0.05 using in a one-way ANOVA with Welch's correction.

Figure 6

PGE₂ controls phosphorylation of STAT3 in colon-infiltrating T cells. Individual colons from the 3 different groups of colitic mice indicated (1 = WT Teff → Rag1^−/−, 2 = EP4^ΔCD4 Teff → Rag1^−/−, 3 = WT Teff → Rag1^−/− x mPGES-1^−/−) were processed and stained for fluorescence microscopy evaluation DAPI, CD3ε, RORγt, FoxP3, and pSTAT3. (A) Quantification of total CD3ε⁺ and CD3ε⁻ cells in 15 regions/colon containing a total of 15–30K cells. (B) Quantification of total CD3ε⁺FoxP3⁺ and CD3ε⁺RORγt⁺ cells in the same groups. (C) Representative colon sections depicting co-localization of RORγt (green), pSTAT3 (red) and CD3 (gray) in the upper row, or of FoxP3 (green), pSTAT3 (red) with CD3 (gray) in the lower row in the same 3 groups. (D) Summary of the quantification of pSTAT3⁺ cells within the indicated CD3ε⁺ subsets (single RORγt⁺, RORγt⁻FoxP3⁻ or single FoxP3⁺). *P < 0.05 and **P < 0.01 using a one-way ANOVA with Welch's correction.