Topical Prostaglandin E Analog Restores Defective Dendritic Cell-Mediated Th17 Host Defense Against Methicillin-Resistant Staphylococcus Aureus in the Skin of Diabetic Mice

Abstract

People with diabetes are more prone to Staphylococcus aureus skin infection than healthy individuals. Control of S. aureus infection depends on dendritic cell (DC)-induced T-helper 17 (Th17)-mediated neutrophil recruitment and bacterial clearance. DC ingestion of infected apoptotic cells (IACs) drive prostaglandin E₂ (PGE₂) secretion to generate Th17 cells. We speculated that hyperglycemia inhibits skin DC migration to the lymph nodes and impairs the Th17 differentiation that accounts for poor skin host defense in diabetic mice. Diabetic mice showed increased skin lesion size and bacterial load and decreased PGE₂ secretion and Th17 cells compared with nondiabetic mice after methicillin-resistant S. aureus (MRSA) infection. Bone marrow-derived DCs (BMDCs) cultured in high glucose (25 mmol/L) exhibited decreased Ptges mRNA expression, PGE₂ production, lower CCR7-dependent DC migration, and diminished maturation after recognition of MRSA-IACs than BMDCs cultured in low glucose (5 mmol/L). Similar events were observed in DCs from diabetic mice infected with MRSA. Topical treatment of diabetic mice with the PGE analog misoprostol improved host defense against MRSA skin infection by restoring DC migration to draining lymph nodes, Th17 differentiation, and increased antimicrobial peptide expression. These findings identify a novel mechanism involved in poor skin host defense in diabetes and propose a targeted strategy to restore skin host defense in diabetes.

Figure 1

Uncontrolled MRSA skin infection in diabetes correlates with decreased skin PGE₂ and Th17 cells in the infected skin. A: Control and diabetic mice were infected subcutaneously with MRSA (5 × 10⁶ cfu). Lesion development was monitored every other day for 7 days. B: Representative images of the skin of diabetic and control mice at day 7 postinfection. C–J: Skin biopsy specimens were collected at day 7 postinfection, and isolated cells were examined for bacterial counts (C), expression of CD4 and IL-17A (D and E), and expression of TGF-β (T-bet) (F) and Foxp3 (G) by CD4⁺ lymphocytes. The lymphocyte population was gated on T-cell receptor-β–positive (TCRβ⁺) CD4⁺ cells. Levels of PGE₂ (H), IL-17A (I), and IL-10 (J) were determined by ELISA in skin biopsy homogenates. Data are mean ± SEM of samples from 5–10 mice. *P < 0.05 vs. control.

Figure 2

Impaired DC migration and Th17 generation in skin-draining lymph nodes (LN) of diabetic mice. A: Representative dot plots of the DC (CD11c⁺MHC II⁺) population and a histogram of the percentage of FarRed⁺ DCs in the brachial lymph nodes of control and diabetic mice 36 h postinfection. B and C: Percentage and number of migrating DCs (CD11c⁺MHC II⁺FarRed⁺) in the lymph nodes. D: Representative dot plots of the Th17 (CD4⁺IL-17A⁺) population in the brachial lymph nodes 36 h postinfection in control and diabetic mice. E and F: Percentage and numbers of IL-17A–producing CD4⁺ lymphocytes. Data are mean ± SEM of samples from 5–10 mice. *P < 0.05 vs. control.

Figure 3

Decreased CCR7 and MHC II expression by Langerin+ DCs in diabetic skin during MRSA infection. A and B: Representative dot plots of CD11c⁺CD11b⁺ and Langerin+ DCs in the skin of control and diabetic mice 18 h postinfection. C: Percentage of Langerin+ DCs in skin of control and diabetic mice 18 h postinfection. Percentage of Langerin+ DCs expressing MHC II (D), CD86 (F), or CCR7 (H and I) 18 h postinfection. MFI of staining of MHC II (E) and CD86 (G) in Langerin+ DCs from diabetic and control mice 18 h postinfection. J and K: Percentage of Langerin+ CCR7⁺ DCs and CCR7 MFI on Langerin+ DCs in the skin of naive control and diabetic mice or 18 h postinfection. Data are mean ± SEM of samples from 5–10 mice. *P < 0.05 vs. control; %P < 0.05 vs. diabetic naive skin; #P < 0.05 vs. infected control skin.

Figure 4

Exogenous PGE₂ restores CD86 and CCR7 expression in BMDCs cultured in high glucose and incubated with MRSA-IACs. Representative histograms of expression (MFI) of MHC II (A), CD80 (B), CD86 (C), and CCR7 (D). Fold change in expression (MFI) of MHC II (E), CD80 (F), CD86 (G), and CCR7 (H) by BMDCs from control or diabetic mice after efferocytosis of MRSA-IACs compared with inactivated BMDCs from control or diabetic mice. The analyzed populations were gated on CD11c⁺CD11b⁺. Data are mean ± SEM of at least three independent experiments performed in triplicate. *P < 0.05 vs. control 5 mmol/L glucose; &P < 0.05 vs. control 25 mmol/L glucose; #P < 0.05 vs. diabetic 5 mmol/L glucose; %P < 0.05 vs. diabetic 25 mmol/L glucose.

Figure 5

Exogenous PGE₂ enhances BMDC migration in a CCR7-dependent manner after recognition of MRSA-IACs. A: Fold change of PGE₂ levels in the supernatants of BMDC + MRSA-IAC cocultures after 18 h determined by EIA. B: Migratory BMDCs using flow cytometry acquisition software. C: Representative images of migrating BMDCs. Data are mean ± SEM of at least three independent experiments performed in triplicate. *P < 0.05 control vs. 5 mmol/L glucose; $P < 0.05 control vs. 25 mmol/L glucose; #P < 0.05 diabetic control vs. diabetic 5 mmol/L glucose; %P < 0.05 diabetic control vs. diabetic 25 mmol/L glucose.

Figure 6

Misoprostol (Miso) improves DC migration and Th17 development in skin-draining lymph nodes in diabetes. A: Representative histograms of the percentage of migrating DCs in the brachial lymph nodes of MRSA-infected control or diabetic mice treated or not treated with misoprostol. B: Representative dot plots of the percentage of Foxp3- and IL-17A–expressing T lymphocytes in the brachial lymph nodes of control and diabetic mice treated or not treated with misoprostol. C: Percentage of migrating DCs (CD11c⁺FarRed⁺) in the brachial lymph nodes. The percentage of intracellular IL-17A– (D), Foxp3- (E), and interferon-γ (IFN-γ)–producing (F) T-cell receptor-β–positive CD4⁺ T lymphocytes determined by FACS analysis. Data are mean ± SEM of samples from 5–10 mice. *P < 0.05 vs. control; #P < 0.05 vs. untreated diabetic mice.

Figure 7

Topical misoprostol (Miso) decreases MRSA skin infection in diabetes. A: Lesion and abscess size are represented by affected area in mm³ as described in research design and methods. B: Bacterial load determination in the skin of diabetic mice treated or not treated with Miso at day 7 after infection. C: Representative dot plots of the percentage of Foxp3- and intracellular IL-17A–expressing T lymphocytes determined by FACS analysis. Percentage and number of IL-17A– (D and E), Foxp3- (F and G), and interferon-γ (IFN-γ)–expressing (H and I) T-cell receptor-β–positive (TCRβ⁺) CD4⁺ lymphocytes determined by FACS analysis. mRNA expression of β-defensin 1 (J), β-defensin 2 (K), β-defensin 3 (L), β-defensin 4 (M), β-defensin 5 (N), and Cramp (O). Data are mean ± SEM of samples from 5–10 mice. *P < 0.05 vs. control; #P < 0.05 vs. untreated diabetic mice.

Figure 8

Proposed model of PGE₂ regulation of DC migration and Th17 generation during MRSA skin infection in diabetic mice. During MRSA skin infection, PGE₂ is produced in the skin and enhances CCR7-dependent DC migration to lymph nodes and Th17 generation. Increased Th17 enhances defensin generation and bacterial clearance in infected nondiabetic mice. In diabetic mice, hyperglycemia decreases the Ptges mRNA expression that culminates in low PGE₂ production and deficient CCR7-dependent DC migration and Th17 generation in the lymph nodes, which impairs skin host defense. Adding back a topical PGE analog restores the DC/Th17 axis and improves host defense in diabetic mice. AA, arachidonic acid.