C1q as a unique player in angiogenesis with therapeutic implication in wound healing

Abstract

We have previously shown that C1q is expressed on endothelial cells (ECs) of newly formed decidual tissue. Here we demonstrate that C1q is deposited in wound-healing skin in the absence of C4 and C3 and that C1q mRNA is locally expressed as revealed by real-time PCR and in situ hybridization. C1q was found to induce permeability of the EC monolayer, to stimulate EC proliferation and migration, and to promote tube formation and sprouting of new vessels in a rat aortic ring assay. Using a murine model of wound healing we observed that vessel formation was defective in C1qa(-/-) mice and was restored to normal after local application of C1q. The mean vessel density of wound-healing tissue and the healed wound area were significantly increased in C1q-treated rats. On the basis of these results we suggest that C1q may represent a valuable therapeutic agent that can be used to treat chronic ulcers or other pathological conditions in which angiogenesis is impaired, such as myocardial ischemia.

Keywords: animal models; complement; vasculogenesis.

Fig. 1.

Immunohistochemical analysis of human dermal granulation tissue and intact skin. Complement deposition was analyzed using polyclonal antibodies to various complement components and revealed by the streptavidin–biotin–peroxidase complex method using aminoethyl carbazole as chromogenic substrate (red signal). Note the localization of C1q on the endothelium and stroma of the granulation tissue (A), whereas C1q was undetectable in intact skin (B). Note also that C1r is present only in tissue-infiltrating cells with histiocytic morphology (C), whereas C1s distribution resembles that of C1q and is expressed on vessels and infiltrating inflammatory cells (D). Both C4 (E) and C3 (F) were undetectable. Arrows indicate blood vessels. (Scale bars, 30 μm.)

Fig. 2.

Local expression of C1q mRNA in granulation tissue of wound-healing and intact skin. (A) Quantitative PCR analysis of RNA extracted from wound-healing and intact skin. The relative amount of the three chains of C1q was normalized with 18S expression and RNA extracted from blood-derived macrophages was used as calibrator. Data are presented as mean of three independent experiments ± SD; *P < 0.05, **P < 0.01 vs. intact skin. (B–E) In situ hybridization of wound-healing (B and C) and intact skin (D and E) sections for the C1q C chain mRNA expression using digoxigenin-labeled antisense (B and D) and sense (C and E) RNA probes followed by incubation with alkaline phosphatase-conjugated antidigoxigenin antibody and staining with Nitro blue tetrazolium and 5-Bromo-4-chloro 3-indolyl phosphate. Arrows indicate blood vessels. (Scale bars, 30 μm.)

Fig. 3.

In vitro effects of C1q on endothelial cells. (A) The permeabilizing activity of C1q (10 µg /mL), VEGF (20 ng/mL), or bradikinin (BK, 10⁻⁶ M) added to the upper chamber of the Transwell was evaluated measuring the amount of FITC-labeled BSA leaked through a monolayer of endothelial cells into the lower chamber. (B) Proliferation of endothelial cells was investigated counting the number of cells expressing the proliferation marker Ki-67 at various time intervals after addition of C1q (10 µg/mL) or VEGF (20 ng/mL). (C) Migration of endothelial cells using the scratch assay. A confluent monolayer of HUVEC was stripped in the middle with a pipette tip and the scratched area covered by cells that had migrated after exposure to C1q (10 µg/mL) or VEGF (20 ng/mL) or medium alone up to 24 h was measured. The continuous line in the scratched area represents the limit of cell migration. The dotted line indicates the border of the scratched area. (D and E) Migration of endothelial cells from the upper to the lower chamber of a Transwell system. C1q (10 μg/mL), VEGF (20 ng/mL), cC1q (0.5 μg/mL), or gC1q (0.5 μg/mL) was added to the lower chamber and serum-free medium was used as a control. Boiled C1q and C1q mixed with polymyxin B (50 µg/mL) were also analyzed to exclude the presence of LPS. Inhibition experiments were performed incubating the cells with F(ab)₂ anti-cC1qR or anti-gC1qR (40 μg/mL) for 30 min at 37 °C before addition of C1q. The cells were allowed to migrate up to 12 h. The data are expressed as mean of five independent experiments run in triplicate ± SD; *P < 0.05, **P < 0.01 vs. control.

Fig. 4.

In vitro and ex vivo angiogenesis induced by C1q. (A) Capillary-like tubules were allowed to form from endothelial cells exposed to C1q (10 µg/mL) or VEGF (20 ng/mL) in Matrigel. The tubules were visualized after staining with Phalloidin–Alexa Fluor 546. (Scale bars, 300 μm.) (B) Vessel number was counted using LAS software connected to Leica microscope. (C) Photomicrographs showing vessel sproutings formed after 6 and 9 d of incubation of aortic rings with C1q (10 µg/mL) or VEGF (20 ng/mL). (Scale bars, 500 μm.) (D) Photomicrograph at two magnifications (40× and 100×) of a microvessel sprout (indicated by the rectangular area in C) formed after 9 d of incubation with C1q showing endothelial cells stained in green with isolectin-B4 and pericytes stained in red with rabbit IgG anti-NG2. (E) Quantification of number and maximal length (millimeters) of vessels and number of branchings formed after incubation of aortic rings with C1q for 9 d. The data are expressed as mean of five independent experiments run in duplicate ± SD; *P < 0.05, **P < 0.01 vs. control.

Fig. 5.

Angiogenesis in a wound-healing model in C1qa^−/− and WT mice. Sections of paraffin-embedded skin of wound-healing samples were obtained from WT, C3^−/−, C1qa^−/−, and C1qa^−/− mice treated with a topical application of C1q (5 μg) 14 d after surgery and stained for the expression of CD34. Three animals for each group were used. The MVD was quantified on five different 200× microscopic fields of skin sections immunostained for CD34. (A) Representative pictures of CD34 immunostained sections showing the MVD in WT, C3^−/−, C1qa^−/−, and C1qa^−/− mice treated with C1q. Arrows indicate blood vessels present in the tissue. (Scale bars, 30 μm.) (B) Quantification of MVD in five fields of skin sections from WT, C3^−/−, C1qa^−/−, and C1q-treated C1qa^−/− mice. (C) Evaluation of wound closure expressed as percentage of the initial punch biopsy area. The data are presented as mean ± SD; *P < 0.05 vs. C1qa^−/−.

Fig. 6.

Angiogenesis in a rat wound-healing model. (A) Representative pictures of paraffin-embedded skin sections of wounds treated with local application of either C1q (5 μg) or VEGF (1 ng) or saline 14 d after surgery. The sections were stained with anti-CD34. Arrows indicate blood vessels present in the tissue (Scale bars, 30 μm.) (B) Sections obtained from five rats were also analyzed blindly for MVD in five different fields at 200× magnification. (C) Evaluation of wound closure expressed as percentage of the initial punch biopsy area. The data are presented as mean ± SD; *P < 0.05, **P < 0.01 vs. saline.