Regarding how important is the prevention of cardiovascular events in RA, the understanding of the mechanisms associated to endothelial dysfunction, as well as the identification of new relevant treatments are of particular importance. The present study investigated for the first time the effect of a curative treatment with an arginase inhibitor (nor-NOHA) on vascular function of AIA rats. Our results demonstrated that this treatment improves endothelial function via an increase in NOS activity and EDHF production, a decrease in COX-2, TX synthase and PGI2 synthase activities as well as a decrease in O2-. production. However, our data did not argue for a role of arginase in the chronic phase of inflammatory joint damages in the AIA model.
In accordance with previous studies [3
], our data showed impaired endothelial function in AIA rats. To ascertain that the abnormal response of vessels from AIA rats to Ach was not due to decreased response of vascular smooth muscle cells (VSMCs) to NO, we demonstrated that the relaxing effect of the NO donor SNP is not impaired in AIA rats, in agreement with recent studies in rats [4
] as well as in RA patients [36
]. Likewise, to eliminate the contribution of an abnormal response of VSMCs to endothelium-derived vasoconstrictive mediators, we demonstrated that the effects of NE, ANG-II and ET-1 are unaltered in AIA rats. To understand the mechanisms involved in the effects of nor-NOHA, and because arginase and endothelial NOS share the same substrate L-arginine, aortic rings were first incubated with the non-selective NOS inhibitor L-NAME. Our data showed that the treatment with nor-NOHA significantly restored the decreased NOS activity observed in AIA rats. This result argues for the contribution of increased vascular arginase activity to NO deficiency in AIA rats, and demonstrated that in vivo
arginase inhibition restores the equilibrium between NOS and arginase pathways, as previously identified in animal models of cardiovascular diseases, such as hypertension [10
], atherosclerosis [13
] or ageing [14
]. Another new finding is the lack of effect of the selective iNOS inhibitor 1400 W on endothelial dysfunction in AIA. Data on vascular iNOS expression in AIA are controversial since the expression was found to be unchanged [4
] or increased [35
]. Our data showing the lack of effect of 1400 W on Ach-induced vasodilation, whatever the group of rats, did not argue for a contribution of the iNOS pathway in the impaired endothelial function associated to RA. This result contrasts with that of a study conducted in a small cohort of RA patients suggesting a deleterious role of increased vascular iNOS expression in vascular dysfunction [36
]. However, in this study, the iNOS inhibitor used was aminoguanidine, which has a wide range of pharmacological actions besides iNOS inhibition including antioxidant effects [39
], and it cannot be excluded that this latter property accounted for its vascular effect. Collectively, these data emphasize the need for further studies to define the role of iNOS in RA-associated endothelial dysfunction.
The COX pathway plays an important role in endothelial function. COX-1 is expressed constitutively and is usually abundant in all animal and human endothelial cells, whereas endothelial COX-2 is induced mainly during inflammatory response [40
]. Physiologically, COX synthesizes both vasorelaxant prostanoids, such as PGI2
and vasoconstrictive prostanoids, such as TXA2
. On the basis of the effects observed for indometacin and the preferential COX-2 inhibitor NS-398, the present study demonstrates for the first time that COX-2 activation contributes to endothelial dysfunction in aorta of AIA rats. Our result is consistent with the results of a recent study showing that vascular COX-2 expression was overexpressed in rabbits with both chronic AIA and atherosclerosis [41
]. Although the contribution of COX-2 and TX synthase overactivation to endothelial dysfunction has been already demonstrated in animal models of cardiovascular diseases [40
], the negative role of PGI2
is more surprising. In fact, our results confirm the janus face of PGI2
and suggest that, as described in spontaneously hypertensive rats (SHR) [42
can induce vasoconstriction in AIA rats. Importantly, nor-NOHA treatment totally normalized the COX-pathway dysfunction in AIA rats. These data are in accordance with a previous study conducted in SHR in which nor-NOHA treatment was able to improve the COX-dependent component of Ach by decreasing vascular COX-2 expression [11
Despite the ongoing debate of the molecular identity and signalling pathways, the contribution of endothelium-derived hyperpolarizing factor(s) (EDHFs) to the endothelium-dependent relaxation is also considered as an important feature of normal endothelium function [31
]. EDHF has been demonstrated unequivocally in various blood vessels from different species, including human [43
]. The acronym 'EDHF' is applied to a factor inducing vascular relaxation in the presence of COX plus NOS inhibitors and which is inhibited by charybdotoxin + apamin but unaffected by iberiotoxin + apamin. Animal studies have identified several mediators that might act as EDHF, such as K+
, cytochrome P450 metabolites, lipoxygenase products, NO itself, H2
, cyclic adenosine monophosphate, C-type natriuretic peptide and electrical coupling through myoendothelial gap junctions [43
]. Whatever the mediator, EDHF induces a potassium-mediated event associated to a reduction in intracellular K+
in vascular smooth muscle [45
]. Experimental evidence indicates that EDHF action is of critical importance for endothelial function when NO production is compromised [31
]. Our data showed for the first time that EDHF production is impaired in AIA rats, suggesting that the EDHF-mediated compensatory dilator system is lacking in RA. Of interest, our data reported that arginase inhibition restored the EDHF contribution to that of control rats. Because of the existence of a cross-talk between NO and EDHF pathways [31
], it can be hypothesized that the improvement of NO production by nor-NOHA contributes to the recovery of EDHF production in AIA treated rats. However, albeit non-elucidated, a direct effect of nor-NOHA on EDHF production cannot be excluded.
Previous reports suggested that the overproduction of O2-.
contributes to endothelial dysfunction in AIA rats [4
]. Consistent with these findings, the present study demonstrates the improvement of ACh-induced vasorelaxation with a SOD mimetic. In the AIA model, previous data reported that statins [5
] as well as ANG-II receptor antagonists [6
] were able to decrease aortic O2-.
overproduction. In the present study, we provide evidence that this beneficial effect can also be obtained with an arginase inhibitor. It is worth noting that recent reports reported that statins inhibit arginase activity [46
] and that blockade of ANG-II AT1 receptors prevents the ANG II-induced elevation of arginase activity and expression [48
]. Although speculative, we hypothesize that arginase could be the downstream common effector of these two types of drugs. Because the deficiency of L-arginine causes a NOS uncoupling leading to O2-.
], the beneficial effect of arginase inhibition on O2-.
production might be due to the decrease in vascular eNOS uncoupling, as recently demonstrated in the vasculature of aged rats treated with the arginase inhibitor 2S-amino-6-boronohexanoic acid [49
]. Additionally, since COX-2 was found to contribute to O2-
], the nor-NOHA-induced decrease in COX-2 activity might also contribute to this effect. Moreover, our results confirmed the deleterious impact of increased NADPH oxidase activity to endothelial function in AIA [4
] and demonstrated that nor-NOHA reverses this phenomenon. Whether arginase inhibition can directly reduce NADPH oxidase activity/expression is currently unknown but a recent study showing that NADPH oxidase inhibition reduced arginase upregulation in retinal cells [50
] suggests an interplay between the two enzymes.
Although there is now ample evidence that RA is associated with endothelial dysfunction [1
], several issues remain unresolved concerning the potential contribution of disease activity and classical cardiovascular risk factors. In addition, whether endothelial dysfunction occurs before the onset of RA or is a consequence of the disease is still a matter of debate. The presence of endothelial dysfunction in the very early stages of RA was recently demonstrated and was not explained by differences in disease activity and inflammatory markers [51
]. Conversely, no impairment of endothelial function was observed in patients with < 7 years' disease duration whereas longer disease duration (> 14 years) was associated with severe endothelial dysfunction [52
]. Others showed that endothelial dysfunction reflects enhanced inflammatory disease activity [38
]. Moreover, on the observation that elevated inflammatory molecules are associated with increased risk of cardiovascular diseases in the general population, it has been speculated that RA-related inflammation might contribute to endothelial dysfunction [1
]. However, as emphasized in a recent review [1
], there is surprisingly no direct evidence supporting such an association between systemic inflammation and vascular dysfunction in RA patients. Therefore, two hypotheses might be formulated to explain how a treatment can improve endothelial function in RA. First, the benefits might be secondary to decreased systemic and/or vascular inflammation and disease activity. Second, the treatment may act directly on endothelial homeostasis independently of inflammation or disease activity. In our study, we showed that nor-NOHA treatment did not modify disease severity assessed by clinical, histological and radiological parameters, whereas it fully reversed endothelial dysfunction in AIA rats. These results demonstrate that the reduction of endothelial dysfunction is possible even though articular parameters are not improved and that endothelial dysfunction is not the consequence of the disease, at least in the chronic phase of the AIA model. These data are in keeping with a recent report showing that an angiotensin-converting enzyme (ACE) inhibitor improved endothelial dysfunction without any change in disease activity in RA patients [54
]. In this study, treatment with ramipril markedly improved endothelial function, although it modestly reduced circulating cytokines, indicating that other mechanisms than the reduction of vascular inflammation are likely involved in the beneficial effects of the ACE inhibitor. Likewise, our data showing that plasma IL-6 levels did not correlate with endothelial dysfunction strongly suggest that the benefits provided by nor-NOHA are related to the direct modulation of endothelium-derived vasorelaxant pathways rather than an anti-inflammatory effect. In addition, our data also suggest that arginase is not upregulated at the articular level and/or not involved in the evolution of the disease between the Day 13 and the Day 34 post-immunization, that is, during the chronic phase of AIA. These data are in agreement with a recent study conducted on arthritic mice showing that arginase gene expression in synovial tissue increased during acute inflammation but not during chronic inflammation [55
]. Further studies will be necessary to explore the role of arginase at the articular level in RA.
With the aim to identify clinically relevant peripheral markers of endothelial dysfunction in RA, we measured plasma VEGF levels in control and AIA rats. Although increased VEGF level is considered as a marker of endothelial dysfunction in hypercholesterolemia, atherosclerosis or hypertension [28
], whether VEGF is a marker of endothelial dysfunction in RA is not known. Previous studies conducted in RA patients reported elevated VEGF levels in serum, which were related to angiogenesis of synovial pannus but not with endothelial dysfunction [56
]. In the present study, we showed that VEGF levels are increased in AIA rats, and that nor-NOHA treatment decreased VEGF levels, whereas arthritis severity was unchanged. Importantly, we identified the existence of a negative correlation between VEGF levels and endothelial function, suggesting for the first time that high plasma VEGF levels might reflect endothelial dysfunction in RA. The mechanisms explaining the decrease in VEGF levels after arginase inhibition were not determined in the present study but might rely on decreased IL-6 production since VEGF synthesis is IL-6 dependent [57