Capillary morphogenesis gene 2 (CMG2) mediates growth factor-induced angiogenesis by regulating endothelial cell chemotaxis.

Anthrax protective antigen (PA) is a potent inhibitor of pathological angiogenesis with an unknown mechanism. In anthrax intoxication, PA interacts with capillary morphogenesis gene 2 (CMG2) and tumor endothelial marker 8 (TEM8). Here, we show that CMG2 mediates the antiangiogenic effects of PA and is required for growth-factor-induced chemotaxis. Using specific inhibitors of CMG2 and TEM8 interaction with natural ligand, as well as mice with the CMG2 or TEM8 transmembrane and intracellular domains disrupted, we demonstrate that inhibiting CMG2, but not TEM8 reduces growth-factor-induced angiogenesis in the cornea. Furthermore, the antiangiogenic effect of PA was abolished when the CMG2, but not the TEM8, gene was disrupted. Binding experiments demonstrated a broad ligand specificity for CMG2 among extracellular matrix (ECM) proteins. Ex vivo experiments demonstrated that CMG2 (but not TEM8) is required for PA activity in human dermal microvascular endothelial cell (HMVEC-d) network formation assays. Remarkably, blocking CMG2-ligand binding with PA or CRISPR knockout abolishes endothelial cell chemotaxis but not chemokinesis in microfluidic migration assays. These effects are phenocopied by Rho inhibition. Because CMG2 mediates the chemotactic response of endothelial cells to peptide growth factors in an ECM-dependent fashion, CMG2 is well-placed to integrate growth factor and ECM signals. Thus, CMG2 targeting is a novel way to inhibit angiogenesis.

Nonsurgical mouse model of endometriosis-associated pain that responds to clinically active drugs.

Endometriosis is an estrogen-dependent inflammatory disease that affects approximately 10% of women. Debilitating pelvic or abdominal pain is one of its major clinical features. Current animal models of endometriosis-associated pain require surgery either to implant tissue or to remove the ovaries. Moreover, existing models do not induce spontaneous pain, which is the primary symptom of patients with chronic pain, including endometriosis. A lack of models that accurately recapitulate the disease phenotype must contribute to the high failure rate of clinical trials for analgesic drugs directed at chronic pain, including those for endometriosis. We set out to establish a murine model of endometriosis-associated pain. Endometriosis was induced nonsurgically by injecting a dissociated uterine horn into a recipient mouse. The induced lesions exhibited histological features that resemble human lesions along with an increase in proinflammatory cytokines and recruitment of immune cells. We also observed the presence of calcitonin gene-related peptide-, TRPA1-, and TRPV1-expressing nerve fibers in the lesions. This model induced mechanical allodynia, spontaneous abdominal pain, and changes in thermal selection behavior that indicate discomfort. These behavioral changes were reduced by drugs used clinically for endometriosis, specifically letrozole (aromatase inhibitor) and danazol (androgen). Endometriosis also induced neuronal changes as evidenced by activation of the NF-κB signaling pathway in TRPA1- and TRPV1-expressing dorsal root ganglion neurons. In conclusion, we have established a model of endometriosis-associated pain that responds to clinically active drugs and can, therefore, be used to identify novel therapies.