Thrombospondin-1/CD47 blockade following ischemia-reperfusion injury is tissue protective

Abstract

Background: Nitric oxide has prosurvival effects that can limit ischemia-reperfusion injuries. However, the matrix glycoprotein thrombospondin-1 is induced following ischemia-reperfusion injury and limits nitric oxide signaling by engaging its cell surface receptor CD47. In this article, the authors examine whether postinjury blocking of this inhibitory signal can protect from ischemia-reperfusion injury in a rat flap model.

Methods: A total of 40 tissue flaps were created in rats based solely on the deep inferior epigastric vessels. Microvascular clamps were used to create 45 minutes of ischemia time to the flaps. The flaps were then treated using a monoclonal antibody to CD47 or an isotype-matched control immunoglobulin G1 5 or 30 minutes after clamp removal. Twenty-four or 72 hours postoperatively, the necrotic area of the flap was determined, and serum, deep inferior epigastric vessels, and flaps were harvested for analysis from five rats in each respective group.

Results: Treatment with a CD47 antibody 5 minutes after reperfusion significantly reduces flap necrosis compared with immunoglobulin G1 control (9 percent versus 43 percent; p < 0.01). The protective effect is even more dramatic when treatment is delayed until 30 minutes after reperfusion (10 percent versus 88 percent for control; p < 0.01). Markers of neutrophil and endothelial cell activation along with total leukocytes are reduced in CD47 antibody-treated flaps, as are tissue malondialdehyde levels. Levels of cyclic guanosine monophosphate are elevated 72 hours postoperatively in the CD47 antibody-treated deep inferior epigastric vessels versus the control flaps.

Conclusions: Therapies targeting the thrombospondin-1 receptor CD47 offer potential for increasing tissue survival in ischemia-reperfusion injuries. The ability to protect when given after ischemia-reperfusion injury enables a broader clinical applicability.

Figure 1. Tissue protective effects of CD47 blockade treatment given post I/R injury

Myocutaneous rat island flaps measuring 2 × 6 cm were created on sex- and age-matched F344/NCr rats weighing between 250 and 300g. Deep inferior epigastric vessels (DIEV) were isolated and then clamped for 45 minutes. Five or 30 minutes post-reperfusion, flaps were treated with either an IgG₁ isotype control antibody or rat-specific CD47 antibody (10 μg in 1 ml of sterile PBS) by local subcutaneous injection within the full area of the flap. Postoperatively 72 hours, rats were re-anesthetized, images were obtained, and the area of flap necrosis was determined. Representative images are presented (A). Percentage of flap necrosis was determined as previously described and results represent the mean ± SD of 5 rats in group of treatment 5 minutes post-reperfusion (B) or treatment 30 minutes post-reperfusion (C). Sham surgeries were performed as previously described and results represent the mean ± SD of 5 rats

Figure 2. cGMP and levels are increased in flap vessels treated with CD47 antibody

DIEV were harvested 72 hours post flap creation and treatment with either IgG₁ isotype control antibody or CD47 antibody 30 minutes post reperfusion and compared to DIEV subjected only to flap creation without treatment. Levels of cGMP were analyzed by immunoassay and were increased in those DIEV treated with CD47 antibody (* p<0.01). The results of duplicate samples are presented as the mean ± SD for 5 rats in each group.

Figure 3. Circulating levels of serum IFN-γ are greatly decreased from baseline following CD47 suppression given post I/R injury

Rat serum was obtained intra-abdominally from the inferior vena cava 24 hours post-reperfusion immediately prior to euthanasia. Serum levels of IFN-γ from rats given either an IgG₁ isotype or CD47 antibody 30 minutes post reperfusion were determined by immunoassay and compared to normal levels of rat IFN-γ. Results represent percentage increase or decrease from baseline IFN-γ levels and are the mean ± SD of 5 rats in each group

Figure 4. CD47 blockade reduces lipid peroxidation in flap tissue post I/R injury

Levels of malondialdehyde (MDA), which is a stable product of lipid peroxidation, were determined from tissue samples taken from the distal area of each flap. Flaps were subjected to either an IgG₁ isotype control antibody or CD47 antibody 30 minutes post reperfusion, and flap tissue from each group was harvested at 24 or 72 hours after antibody delivery. Tissue samples were also taken from the same area in animals not subjected to I/R injury and used as a normal control. MDA levels were much higher in the IgG₁ isotype control antibody when compared to normal tissue (* p<0.01). However, levels of MDA were greatly decreased after treatment with CD47 antibody 72 hours post I/R (# p<0.001). MDA levels represent the mean ± SD of 5 rats in each treatment group.

Figure 5. Histopathological evaluation of rat flaps show reduction of inflammatory markers after treatment with CD47 antibody

After 45 minutes of flap ischemia and 30 minutes of reperfusion prior to treatment with either an IgG₁ isotype control antibody or CD47 antibody, age- and sex-matched rats were killed 72 hours post-operatively, and their flap tissue was excised, processed, and stained with hematoxylin and eosin (A; original magnification 4X, inset 20X). Images are representative of tissue flaps from 5 animals of each treatment group. Leukocyte cell infiltrate was determined in 10 high-power fields from each rat tissue section. Flap tissue subjected to the same conditions but harvested 24 hours post-operatively was examined by immunohistochemistry for the presence of elastase, a neutrophil activation marker. Using a rat-specific monoclonal antibody against elastase, staining was performed on both the IgG₁ isotype control antibody or CD47 antibody treated flaps (B; original magnification 10X. inset 20X). Images are representative of tissue flaps from 5 animals of each treatment group.