Murine cerebral malaria is associated with increased expression of eCAMs as well as leukocyte and platelet accumulation in brain vessels leading to microvascular damage and BBB breakdown [8, 30]. Here we show that exogenous NO supplementation attenuated, although it did not prevent, brain vascular inflammation (eCAM expression, leukocyte and platelet accumulation) and its consequences such as vascular leakage. We have previously shown that exogenous NO supplementation ameliorated pial blood flow, vasoconstriction and the occurrence of brain hemorrhages . Overall, these results indicate that low NO bioavailability has a role in the microvascular dysfunction observed during murine CM  and that exogenous NO can help to protect the vasculature, attenuating endothelial damage and dysfunction. The only partial effects of DPTA-NO treatment on inflammatory markers and on brain hemodynamics suggest that NO supplementation provides a subtle beneficial effect on CM pathogenesis, which for a significant number of animals can be enough to prevent death by CM, rather than an unmistakable modification of disease history. In addition, the overlap between treated and untreated mice in several of the measured parameters might suggest that NO-mediated mechanisms other than the ones reported here could be responsible for the clinical effect.
Studies of eCAM expression in human and murine CM consistently show an increased expression of ICAM-1, whereas the results for other eCAMs are variable. In cases of fatal human CM, the picture that emerges is of marked increases in ICAM-1, significant but less pronounced increases in VCAM-1 and E-selectin and no changes in P-selectin expression in the brain [31–33]. In murine CM, high expression of ICAM-1 in the brain is consistently observed [12, 34, 35] and VCAM-1 expression has been shown to be also markedly increased in brain and retinal vessels [12, 35]. P-selectin was shown to be increased whereas no change in E-selectin was reported on day 6 of infection [12, 36]. Similarly, in the present study, we found a marked increase in ICAM-1 expression in the brain on day 6 of infection, VCAM-1 and P-selectin had more modest increases and E-selectin expression was not significantly different in uninfected and saline-treated mice. ICAM-1, indeed, seems to play a relevant role in CM, as not only its expression in brain is consistently markedly increased, but also ICAM-1 gene knockout mice are protected from CM development [37, 38]. The roles of VCAM-1, E- and P-selectin are less clear. Increased expression of E- and P-selectin seems to be less intense in the brain as compared to other organs and tissues during PbA infection and also sepsis [12, 39]. Whether the lower magnitude of expression of selectins in the brain is relevant for CM development is debatable. P-selectin gene knockout did not change leukocyte accumulation in the brain during murine CM, and although causing a short delay in mortality it did not prevent it .
Low NO bioavailability plays a role in murine CM pathogenesis  and it can help to explain the increased expression of eCAMs in brain vessels with the consequent rolling and adherence of leukocytes during PbA infection. Indeed, impaired NO production by NO synthase (NOS) inhibitors such as L-NAME, or by NOS gene knockout, has been shown to increase eCAM expression in vivo and in vitro [19–21, 25]. Increased eCAM expression as well as vascular inflammation and platelet activation are also observed in hemolytic pathologies such as sickle cell crisis, in which cell-free hemoglobin acts as a potent NO scavenger limiting NO bioavailability . Increased eCAM expression is followed by increased leukocyte rolling and adherence , and NO supplementation with NO donors can inhibit eCAM expression and leukocyte migration [26, 40]. Here we show that exogenous NO indeed caused a significant downregulation of ICAM-1 and P-selectin, but not VCAM-1, expression in the brain of PbA-infected mice and, therefore, this may help to explain the protective effect of NO on murine CM. It is noteworthy, however, that exogenous NO decreased but did not actually prevent ICAM-1 expression, and did not significantly decrease VCAM-1 expression in PbA-infected mice. Accordingly, it decreased but did not prevent leukocyte and platelet accumulation in the brain on day 6, although it did prevent vascular leakage. While low NO bioavailability is an important component of endothelial dysfunction, it is not the sole cause of vascular inflammation in murine CM, which is the result of a complex cascade of events, including early activation of T cells  and production and release of high levels of several inflammatory cytokines . The partial inhibitory effect of NO supplementation on the inflammatory process during PbA infection is in agreement with our previous study showing that exogenous NO supplementation with DPTA-NO attenuated but did not inhibit the brain microcirculatory complications in PbA infection .
In saline-treated PbA-infected mice, the venular sites with higher accumulation of leukocytes presented higher platelet counts. Because there were also direct correlations between the number of adherent leukocytes/platelets and venular diameters, this finding indicates that the differential quantitative accumulation was due basically to the larger available area for binding in larger vessels rather than to eventual differential expression of eCAMs in different vascular beds, although the latter was not directly assessed and cannot be ruled out. This is in accordance with previous findings by Sun and colleagues regarding platelet adherence during murine CM . These results indicate that, during PbA infection, cerebral venules of different sizes in the range analyzed are similarly activated and similarly receptive to leukocyte and platelet adherence. DPTA-NO treatment, however, seems to differentially affect venules according to their size, as it had little effect on leukocyte accumulation in smaller (20-30 μm) venules and marked effect in larger (30-50 μm) ones. This effect is important not only in the sense that decreased inflammation may help preserve vascular integrity, but also because it decreases leukocyte-induced increases in vascular resistance, therefore increasing venous return and potentially improving blood flow. In our previous study , DPTA-NO sustained superior hemodynamics compared to saline, which in part is due to reduced vascular resistance due to geometrical and inflammatory changes. The results of these studies strongly support the view that NO therapy with a stable and long-acting molecule may constitute a useful therapeutic approach to partially decrease receptor-ligands interactions involved in CM cell adhesion and related hemodynamic complications. This may be true as well for mediators such as erythropoietin, which enhances endothelium-dependent vasodilatation mediated by NO in rodent cerebral vessels  and prevents vascular inflammation  and has also been shown to decrease inflammatory markers and partially protect mice against CM development .
Several studies have shown that murine CM is associated with BBB breakdown with consequent vascular leakage [29, 35]. Here we studied the vascular leakage at the level of individual pial blood vessels and show that arterioles and venules were similarly affected in PbA-infected mice leading to increased albumin leakage to the brain tissue. Because no correlation was found between leakage and leukocyte/platelet adherence or vessel diameters in venules, and because leakage occurred in arterioles where leukocyte/platelet adherence was minimal, these results indicate that the endothelial dysfunction leading to increased permeability in the brain vessels is not dependent on close contact with inflammatory cells. In our study in no case the studied sites presented evidence of vessel wall rupture, which would lead to massive leakage and hemorrhage. Therefore, although no correlation was found between sites of leakage and of leukocyte/platelet accumulation, we were not able to establish any relation between the latter and more severe vascular damage. Inhibition of NO production can result in increased vascular permeability  and therefore the low NO bioavailability during PbA infection can help to explain increased vascular leakage. Exogenous NO treatment was indeed sufficient to prevent vascular leakage in pial vessels. This is again in accordance with our previous data showing marked protection against brain hemorrhages provided by DPTA-NO treatment, despite only partial effect on pial hemodynamics .
In summary, exogenous NO supplementation significantly decreases brain vascular inflammation during PbA infection by decreasing the expression of eCAMs and consequently decreasing recruitment of leukocytes and platelets, which eventually results in decreased vascular dysfunction and damage.