Decreased macrophage number and activation lead to reduced lymphatic vessel formation and contribute to impaired diabetic wound healing

Abstract

Impaired wound healing is a common complication of diabetes. Although it is well known that both macrophages and blood vessels are critical to wound repair, the role of wound-associated lymphatic vessels has not been well investigated. We report that both the presence of activated macrophages and the formation of lymphatic vessels are rate-limiting to the healing of diabetic wounds. We have previously shown that macrophages contribute to the lymphatic vessels that form during the acute phase of corneal wound healing. We now demonstrate that this is a general phenomenon; cells that co-stain for the macrophage marker F4/80 and the lymphatic markers LYVE-1 (lymphatic vascular endothelium hyaluronate receptor) and podoplanin contribute to lymphatic vessels in full-thickness wounds. LYVE-1-positive lymphatic vessels and CD31-positive blood vessels were significantly reduced in corneal wound healing in diabetic mice (db/db) (P < 0.02) compared with control (db/+) mice. Glucose treatment of control macrophages led to the down-regulation of the lymphatic-specific receptor VEGFR3 and its ligands, vascular endothelial growth factor-C and -D (VEGF-C, -D). Interleukin-1beta stimulation rescued diabetic macrophage function; application of interleukin-1beta-treated db/db-derived macrophages to wounds in db/db mice induced lymphatic vessel formation and accelerated wound healing. These observations suggest a potential therapeutic approach for healing wounds in diabetic patients.

Figures

Figure 1

Contribution of macrophages to lymphatic vessel formation in normal wound healing. Sections of granulation tissue 5, 10, and 14 days after wounding stained for F4/80 (green), a macrophage marker; LYVE-1 (red), a marker of lymphatic endothelium; and TOPRO3 (blue), a nuclear marker. Results reveal vessel-like structures composed of cells that are double positive for markers of macrophages and lymphatic cells. Scale bar = 40 μm. The small window in D14/OVERLAY image shows a higher magnification of a F4/80 and LYVE-1 double-positive structure (white arrowhead). Scale bar = 16 μm.

Figure 2

Angiogenesis and lymphangiogenesis in the corneal suture model assay in db/db mice. a: A schematic illustrating the method of the suture placement on cornea. b: Quantification of lymphangiogenesis and hemangiogenesis in the corneal suture model assay in wild-type C57BL/6 mice. c: Hemangiogenesis (CD31; green) and lymphangiogenesis (LYVE-1; orange) in the db/db and db/+ (control) mouse corneal 7 days after suture placement. Graphs display quantification of lymphatic and blood vessel area.

Figure 3

Characterization of PECs from db/db and db/+ mice. a: Thioglycollate-induced macrophages were collected from the peritoneal cavity of normal 8-week-old male db/db and db/+ mice and cell numbers compared (n = 4). b: Real-time RT-PCR analysis of VEGFR3, VEGF-C, and VEGF-A relative expression (RE) = 2DCT; by PECs from db/db and db/+ mice. c: RT-PCR analysis of VEGFR3 and VEGF-C after IL-1β (20 ng/ml) stimulation on PECs from db/db mice.

Figure 4

Effect of glucose and IL-1β on control macrophages. Real-time RT-PCR analysis of VEGFR3, VEGF-C, and VEGF-A of db/+ mouse PECs cultured in high glucose (30 mmol/L) for 72 hours (n = 6) relative expression (RE) = 2DCT and in the presence or absence of IL-1β. The statistical analysis was assessed by analysis of variance post hoc analysis (Fisher’s PLSD).

Figure 5

Effect of IL-1β on tube formation by macrophages. a: Schematic of the tube formation assay. b: Flow cytometric analysis of cultured macrophages isolated from bone marrow and cultured in L929 cell condition medium. c: Quantification of cluster and tube-like structure formation in Matrigel (18 mm2) of db/+ peritoneal macrophages (1 × 106 cell/ml) plated into Matrigel and grown for 24 hours in high glucose alone or in high glucose with IL-1β. d: Photomicrograph of Matrigel assay in c. Statistical analysis was performed using analysis of variance post hoc analysis (Fisher’s PLSD).

Figure 6

Wound healing in db/db mice. a: Photomicrographs of wounds in db/db mice 7 and 14 days after administration of the indicated macrophages (1 × 106 cells), including saline (no macrophages), db/db macrophages, db/db macrophages treated with IL-1β, or untreated db/+ macrophages. b: Quantification of newly formed granulation tissue over wounds at days 7 and 14. The area of the wounds at day 7 that received the IL-1β-treated macrophages was significantly smaller than those that received nontreated db/db macrophages or db/+ macrophages. At day 14, there was a significant difference between the area of the wounds that received IL-1β-treated db/db macrophages and untreated db/db macrophages but no difference between IL-1β-treated db/db macrophages or db/+ macrophages. Statistical analysis was performed by analysis of variance post hoc analysis (Fisher’s PLSD).

Figure 7

Histological assessment of wounds in db/db mice treated with macrophages. a1 through a8: Photomicrograph of H&E-stained paraffin sections. a-5: Wounds that received IL-1-treated macrophages display many inflammatory cells in their granulation. a-6: Arrows indicate cord-like structures in granulation tissue that received IL-1β-treated-macrophages. b: Histological scores at days 7 and 14 after receiving the indicated macrophages. There was more extensive granulation tissue at day 7 in wounds that had received IL-1β-treated macrophages than in wounds that had received the untreated db/db macrophages or the db/+ macrophages. At day 14, there was a significant difference between the wounds that had received IL-1β-treated db/db macrophages versus untreated db/db macrophages but no difference between IL-1β-treated db/db macrophages and db/+ macrophages. The statistical analysis was performed using analysis of variance post hoc analysis (Fisher’s PLSD).

Figure 8

Effect of IL-1β treatment on formation of cord-like structures in the granulation tissue. a: Immunofluorescence images at day 7 after treatment with untreated db/db macrophages (LYVE-1; orange) and IL-1β-stimulated db/db macrophages (LYVE-1; orange). b: Quantification of cord-like structures (three or more cells) (1 mm2). Cells were counted under the fluorescence microscope following DAPI staining (data not shown).

Figure 9

Effect of IL-1β-treated macrophages on formation of lymphatic structures. Confocal microscopic images of wounds treated with IL-1β-stimulated db/db macrophages. a: Small vessel-like structures stained with LYVE-1 (orange) and F4/80 (green) at the edge of the granulation tissue (blue is nuclear staining with TOPRO-3). The arrow indicates the lumen that was constituted by cells that stained for both F4/80 (green) and LYVE-1 (red). Scale bar = 20 μm. b: Area outside of the granulation tissue stained with LYVE-1 (orange) and F4/80 (green) (blue is nuclear staining with TOPRO-3). V, vessel; L, lymphatic vessel. Note that some lymphatic vessels did not stain with F4/80 (arrowhead), whereas others stained for both F4/80 and LYVE-1 (arrow). Scale bar = 40 μm. c: A vessel-like structure in the granulation tissue stained with podoplanin (orange) and F4/80 (green). Scale bar = 40 μm.

Figure 10

The contribution of exogenously administered macrophages to lymphatic vessels in wound in db/db mice. Macrophages were isolated from wild-type C57BL/6 mice, prelabeled with DiLDL, and injected into excisional wounds created in diabetic mice. Granulation tissue was examined 7 days later. Top: DiI-prelabeled macrophages stained with LYVE-1 located in granulation tissue 7 days after transplantation. Bottom: DiI-prelabeled macrophages stained with antisera against podoplanin in granulation tissue 7 days after transplantation.