Connective tissue growth factor (CTGF) is a novel cysteine-rich, secreted protein. Recently, we found that inhibition of the endogenous expression of CTGF by its antisense oligonucleotide and antisense RNA suppresses the proliferation and migration of vascular endothelial cells. In the present study, the following observations demonstrated the angiogenic function of CTGF in vitro and in vivo: (i) purified recombinant CTGF (rCTGF) promoted the adhesion, proliferation and migration of vascular endothelial cells in a dose-dependent manner under serum-free conditions, and these effects were inhibited by anti-CTGF antibodies; (ii) rCTGF markedly induced the tube formation of vascular endothelial cells, and this effect was stronger than that of basic fibroblast growth factor or vascular endothelial growth factor; (iii) application of rCTGF to the chicken chorioallantoic membrane resulted in a gross angiogenic response, and this effect was also inhibited by anti-CTGF antibodies. (iv) rCTGF injected with collagen gel into the backs of mice induced strong angiogenesis in vivo. These findings indicate that CTGF is a novel, potent angiogenesis factor which functions in multi-stages in this process.

Abstract-This study aimed to examine whether an elevated activity of heme oxygenase (HO)-1 in the tissue attenuates endothelial cell-leukocyte interactions microvessels in vivo. When rats were pretreated with an intraperitoneal injection of hemin, an HO-1 inducer, mesenteric tissues, including their microvessels, displayed a marked induction of HO-1 concurrent with an increase in plasma concentrations of bilirubin-IX␣, the product of HO-catalyzed degradation of protoheme IX. In these rats, oxidative stress such as superfusion with H 2 O 2 and ischemia-reperfusion of the tissues neither induced rolling nor exhibited adherent responses of leukocytes in venules. In contrast, the oxidative stresses evoked marked rolling and adhesion of leukocytes in the control rats without HO-1 induction. The HO-1 induction also downregulated leukocyte adhesion elicited by other pro-oxidant stimuli such as N -nitro-L-arginine methyl ester. The decreases in the oxidant-elicited leukocyte adhesive responses under HO-1-inducing conditions were restored by perfusion with zinc protoporphyrin-IX, an HO inhibitor, but not with copper protoporphyrin-IX, which did not inhibit the enzyme. Furthermore, the effects of zinc protoporphyrin-IX were repressed by superfusion with bilirubin or biliverdin at the micromolar level, but not by the same concentration of carbon monoxide, another product of HO. These results indicate that induction of the HO-1 activity serves as a potential stratagem to prevent oxidant-induced microvascular leukocyte adhesion through the action of bilirubin, a product of HO reaction. (Circ Res. 1999;85:663-671.)

Carbon monoxide (CO) derived from heme oxygenase has recently been shown to play a role in controlling hepatobiliary function, but intrahepatic distribution of the enzyme is unknown. We examined distribution of two kinds of the heme oxygenase isoforms (HO-1 and HO-2) in rat liver immunohistochemically using monoclonal antibodies. The results showed that distribution of the two isoforms had distinct topographic patterns: HO-1, an inducible isoform, was observed only in Kupffer cells, while HO-2, a constitutive form, distributed to parenchymal cells, but not to Kupffer cells. Both isoforms were undetectable in hepatic stellate cells and sinusoidal endothelial cells. Of the two isoforms, HO-2 in the parenchymal cell rather than HO-1 in the Kupffer cell, appears to play a major role in regulation of microvascular tone. In the perfused liver, administration of HbO2, a CO-trapping reagent that can diffuse across the fenestrated endothelium into the space of Disse, elicited a marked sinusoidal constriction, while administration of a liposome-encapsulated Hb that cannot enter the space had no effect on the microvascular tone. These results suggest that CO evolved by HO-2 in the parenchymal cells, and, released to the extrasinusoidal space, served as the physiological relaxant for hepatic sinusoids.

The adherence and emigration of leukocytes have been implicated as a rate-limiting step in the microvascular dysfunction associated with reperfusion of ischemic tissues. The objective of the present study was to define the relation between leukocyte-endothelial cell adhesion and albumin leakage in rat mesenteric venules exposed to ischemia and reperfusion (I/R). Leukocyte adherence and emigration as well as albumin extravasation were monitored in single post-capillary venules using intravital fluorescence microscopy. Ischemia (0, 10, 15, or 20 minutes) was induced by complete occlusion of the superior mesenteric artery, and all parameters were monitored for 30 minutes after reperfusion. The magnitude of the leukocyte adherence and emigration and albumin leakage elicited by I/R was positively correlated with the duration of ischemia. The albumin leakage response was also highly correlated with the number of adherent and emigrated leukocytes. Monoclonal antibodies against the adhesion glycoproteins CD18, CD11b, intercellular adhesion molecule-1 (ICAM-1) (at 10 and 30 minutes), and L-selectin (at 10 minutes), but not P- or E-selectin, reduced I/R-induced leukocyte adherence and emigration as well as albumin leakage. Platelet-leukocyte aggregates were formed in postischemic venules; the number of aggregates was reduced by antibodies against P-selectin, CD11b, CD18, and ICAM-1, but not E- or L-selectin. These results indicate that reperfusion-induced albumin leakage is tightly coupled to the adherence and emigration of leukocytes in postcapillary venules. This adhesion-dependent injury response is primarily mediated by CD11b/CD18 on activated neutrophils and ICAM-1 on venular endothelium and appears to require L-selectin-dependent leukocyte rolling.