Author + information
- Received September 14, 2005
- Revision received March 27, 2006
- Accepted April 11, 2006
- Published online August 1, 2006.
- Stefano Taddei, MD⁎ (, )
- Daniele Versari, MD,
- Alessandro Cipriano, MD,
- Lorenzo Ghiadoni, MD, PhD,
- Fabio Galetta, MD,
- Ferdinando Franzoni, MD,
- Armando Magagna, MD,
- Agostino Virdis, MD and
- Antonio Salvetti, MD
- ↵⁎Reprint requests and correspondence:
Dr. Stefano Taddei, Department of Internal Medicine, University of Pisa, Via Roma, 67, 56100 Pisa, Italy.
Objectives We assessed the role of cytochrome P450 2C9 (CYP 2C9)-derived endothelium-derived hyperpolarizing factor (EDHF) in the forearm microcirculation of essential hypertensive patients (EH) by utilizing sulfaphenazole (SUL), a selective CYP 2C9 inhibitor.
Background In EH patients, EDHF acts as a compensatory pathway when nitric oxide (NO) availability is reduced. Cytochrome P450 2C9 is a possible source of EDHF.
Methods In 36 healthy subjects (normotensive [NT]) and 32 hypertensive patients (HT), we studied forearm blood flow (strain-gauge plethysmography) changes induced by intraarterial acetylcholine (ACH) and bradykinin (BDK), repeated during NG-monomethyl-L-arginine (L-NMMA) (100 μg/100 ml/min) or SUL (0.03 mg/100 ml/min). In HT, the effect of SUL on ACH and BDK was repeated during vitamin C (8 mg/100 ml/min). Sodium nitroprusside (SNP) was utilized as control.
Results In NT, vasodilation to ACH and BDK was blunted by L-NMMA and not changed by SUL. In contrast, in HT responses to ACH and BDK, reduced compared with NT, were resistant to L-NMMA. In these patients, SUL blunted vasodilation to ACH and to a greater extent the response to BDK. When retested with vitamin C, SUL was no longer effective on both endothelial agonists. In 2 final groups of normotensive control subjects, vasodilation to ACH or BDK residual to cyclooxygenase and L-NMMA blockade was further inhibited by simultaneous SUL infusion. Response to SNP, similar between NT and HT, was unaffected by SUL.
Conclusions Cytochrome P450 epoxygenase-derived EDHF acts as a partial compensatory mechanism to sustain endothelium-dependent vasodilation in HT, particularly the BDK-mediated response, when NO activity is impaired because of oxidative stress.
Essential hypertension is characterized by the presence of endothelial dysfunction, acting as a mechanism promoting atherosclerosis and cardiovascular events (1). One of the principal characteristics of endothelial dysfunction in hypertension is the presence of impaired response to endothelium-dependent agonists, above all acetylcholine (ACH) or bradykinin (BDK), as compared with that in healthy control subjects (2–6). This diminished relaxing response to ACH or BDK is, moreover, resistant to NG-monomethyl-L-arginine (L-NMMA), indicating the presence of compromised nitric oxide (NO) availability caused by oxidative stress (5,6). In these circumstances, other mediators account for the vascular relaxing effects of these compounds. Available evidence supports the possibility that endothelial factor(s) causing smooth muscle cells hyperpolarization (EDHF) act as a compensatory pathway responsible for endothelium-dependent vasodilation in presence of reduced NO availability (6).
One of the possible pathways for EDHF involves the activation of cytochrome P450 epoxygenase (CYP 2C), which is mainly expressed in endothelial cells (7). The role of this enzyme in producing hyperpolarization is dependent on the generation of metabolites of arachidonic acid (EET), which either initiate the endothelial cell hyperpolarization (8,9) or are released from endothelial cells to stimulate K+on vascular smooth muscle cells (10). CYP2C-derived EDHF has been identified in porcine coronary arteries (11) and human mammary arteries in vitro (12). In addition, selective blockade of CYP 2C can blunt NO- and prostacyclin-independent vasodilation to BDK in the forearm microcirculation in healthy humans (13). On the other hand, CYP 2C9 could be also involved in the generation of oxygen free radicals, thereby leading to NO breakdown (14). In line with this possibility, a CYP 2C9-dependent reduction in NO availability has been demonstrated in the forearm microcirculation of patients with coronary artery disease (15). Because essential hypertension is characterized by both NO-independent vasodilation, likely mediated by EDHFs production, and oxidative stress, the aim of the present study was to investigate the possible role of CYP 2C9 in modulating endothelial responses in essential hypertensive patients. To address this issue, we tested the effect of sulfaphenazole, a selective CYP 2C9 inhibitor, on vasodilation to ACH and BDK in the peripheral circulation of essential hypertensive patients.
Thirty-six healthy subjects and 32 matched essential hypertensive patients participated in the study (Table 1).Subjects with smoking history (more than 5 cigarettes per day), ethanol consumption (more than 60 g—half liter of wine—per day), hypercholesterolemia (total cholesterol >200 mg/dl), diabetes mellitus, cardiac and/or cerebrovascular ischemic vascular disease, impaired renal function, and other major pathologies were excluded. In accordance with institutional guidelines, all patients were aware of the investigational nature of the study and gave written consent to it.
Subjects were defined as normal according to the absence of familial history of essential hypertension and blood pressure (BP) values below 140/90 mm Hg (Table 1). Essential hypertensive patients were recruited from among newly diagnosed cases when they reported the presence of positive family history of essential hypertension and if supine arterial BP (after 10 min of rest), measured 3 times at 1-week intervals, was consistently found to be >140/90 mm Hg (Table 1). Secondary forms of hypertension were excluded by routine diagnostic procedures. Patients were enrolled if never treated (n = 10) or reporting a history of discontinued or ineffective pharmacologic antihypertensive treatment (n = 22).
Vascular reactivity was assessed by the perfused forearm technique. Briefly, the brachial artery was cannulated for drug infusion at systemically ineffective rates, and for intraarterial BP and heart rate monitoring. Forearm blood flow (FBF) was measured in both forearms (experimental and contralateral forearm) by strain-gauge venous plethysmography (3). Circulation to the hand was excluded 1 min before FBF measurement by inflating a pediatric cuff around the wrist at suprasystolic BP. Details concerning the method as performed in our laboratory, including sensitivity and reproducibility, have already been published (3,5).
Endothelium-dependent and endothelium-independent vasodilation
In 20 normotensive subjects and 20 essential hypertensive patients, endothelium-dependent forearm vasodilatation was evaluated by obtaining a dose-response curve to intraarterial ACH (cumulative increase in infusion rates: from 0.15 to 15 μg/100 ml of forearm tissue/min, 5’ each dose) and BDK (from 5 to 50 ng/100 ml tissue/min, 5’ each dose). In 8 of 20 normotensive subjects and 8 of 20 essential hypertensive patients, endothelium-independent vasodilatation to sodium nitroprusside (SNP) (from 1 to 4 μg/100 ml tissue/min, 5’ each dose), which acts directly on the smooth muscle cells, was also assessed.
Effects of L-NMMA and sulfaphenazole on vasodilation to ACH and BDK
To compare the relative contribution of NO and CYP 2C9-derived EDHF in endothelium-dependent vasodilation, in 12 of the previous series of 20 normotensive subjects and 12 of 20 hypertensive patients, the dose-response curves to ACH and BDK were performed during saline (0.2 ml/min), in the presence of L-NMMA (100 μg/100 ml tissue/min) or sulfaphenazole (0.03 μg/100 ml tissue/min). Infusions under L-NMMA were performed according to the NO-clamp technique, which requires SNP co-infusion (0.4 and 0.3 μg/100 ml tissue/min for 5’ in normotensive subjects and hypertensive patients, respectively) in order to neutralize the L-NMMA–induced vasoconstriction and restore baseline FBF. Details concerning the method as performed in our laboratory have already been published (16).
To exclude any possible role of CYP 2C9 in endothelium-independent vasodilation, in those groups of normotensive subjects and hypertensive patients assigned to SNP infusion, a dose-response curve to this compound was repeated in the presence of intra-arterial sulfaphenazole (0.03 μg/100 ml tissue/min).
Effect of simultaneous L-NMMA and sulfaphenazole administration in normotensive subjects
This series was designed to assess whether EDHF component of endothelium-dependent vasodilation can be unmasked under inhibition of NO availability. Acetylcholine and BDK were infused in 2 further groups of n = 8 normotensive subjects (each group) during saline (0.2 ml/min), in the presence of L-NMMA (100 μg/100 ml tissue/min) or sulfaphenazole (0.03 μg/100 ml tissue/min) and during both L-NMMA and sulfaphenazole. To avoid any possible influence of cyclooxygenase (COX)-derived vasoactive prostanoids, all subjects were given oral acetylsalicylic acid (1 g) 2 h before the study.
To address the possibility that exogenous NO from the SNP co-infusion utilized for the “NO clamp” could have effects on EDHF pathway, in an adjunctive group (n = 6) of healthy subjects we performed the identical experimental design by replacing SNP with papaverine (150 μg/100 ml tissue/min), a phosphodiesterase activity inhibitor.
Role of oxidative stress in determining the response to sulfaphenazole in essential hypertensive patients
These series were designed to assess whether oxygen-free radical production can modulate the sulfaphenazole sensitive pathway. Thus, in an adjunctive group of 6 essential hypertensive patients, the dose-response curves to ACH were performed during saline (0.2 ml/min) and in the presence of sulfaphenazole (0.03 μg/100 ml forearm tissue/min). These infusions were then repeated under intraarterial administration of vitamin C (8 mg/100 ml forearm tissue/min), an antioxidant (17). Finally, in a further group of 6 essential hypertensive patients, the same above-described protocol was repeated but replacing ACH with BDK.
In each series, the sequence of the agonists was randomized. Sulfaphenazole and L-NMMA were started 10 min before ACH or BDK and continued throughout. A 30-min washout was allowed between each dose-response curve, while a 60-min period was allowed when L-NMMA was infused. These periods of time were validated in adjunctive studies.
In order to confirm that in essential hypertension oxygen-free radical inactivation by vitamin C restores the NO availability, in a final group (n = 6) of hypertensive patients, the dose-response curves to ACH were performed during saline (0.2 ml/min) and in the presence of L-NMMA. These infusions were then repeated under intraarterial administration of vitamin C.
Data were analyzed in terms of changes in FBF. Because arterial BP did not change significantly during the FBF study, increments in FBF were taken as evidence of local vasodilation. Results are expressed as mean ± SD, except for the figures (results described as mean ± SEM). Differences between 2 means were compared by the Student ttest for paired or unpaired observations, as appropriate. Responses to ACH, BDK, and SNP were analyzed by analysis of variance for repeated measures, and Scheffé’s test was applied for multiple comparison testing. Differences were considered to be statistically significant when p was <0.05.
Acetylcholine (Farmigea S.p.A., Pisa, Italy), BDK, HCl, L-NMMA, papaverine, sulfaphenazole (Clinalfa AG, Läufelfingen, Switzerland), vitamin C (Bracco, Milan, Italy), and SNP (Malesci, Milan, Italy) were obtained from commercially available sources and diluted freshly to the desired concentration by adding normal saline. Sodium nitroprusside was dissolved in glucosate solution and protected from light by aluminum foil.
Age, gender, smoking history, and metabolic profile were similar, and within a normal range, between the 2 study groups, differing only in BP values (Table 1).
Effect of L-NMMA and sulfaphenazole on response to ACH, BDK, and SNP in hypertensive patients
The FBF increase induced by ACH and BDK was significantly reduced in hypertensive patients as compared with normotensive subjects (Table 2).In contrast, vasodilation to SNP was similar in both groups (Table 2, Fig. 1).In normotensive subjects, as expected, responses to ACH and BDK were significantly blunted by L-NMMA (Table 2, Fig. 1). In these subjects, sulfaphenazole administration, which did not significantly change basal FBF (data not shown), was devoid of effects on responses to ACH and BDK (Table 2, Fig. 1). Sulfaphenazole also did not change the response to SNP (Fig. 1).
In hypertensive patients, responses to ACH and BDK were resistant to L-NMMA (Table 2, Fig. 1). In contrast, in these patients, sulfaphenazole, while not modifying basal FBF (data not shown), significantly reduced the vasodilation induced by ACH and BDK (Table 2, Fig. 1).
It is worth noting that the degree of percent inhibition exerted by sulfaphenazole was greater on the response to BDK (26.9%) as compared with the response to ACH (15.1%) (Fig. 2).Vascular response to SNP was not affected by the CYP 2C9 inhibitor (Fig. 1).
Effect of simultaneous L-NMMA and sulfaphenazole administration in normotensive subjects
As in the previous series of experiments, sulfaphenazole did not change the response to ACH, whereas L-NMMA infusion significantly blunted the vasodilator effect of the agonist (Table 3,Fig. 3).When ACH was finally repeated in the presence of simultaneous sulfaphenazole and L-NMMA, we observed an inhibition of vasodilation to ACH (Table 3, Fig. 3), which was far greater than that observed with L-NMMA alone. Similar data were observed when ACH was replaced by BDK. Indeed, the vasodilation to BDK was unaffected by sulfaphenazole, blunted by L-NMMA, and further reduced by concomitant sulfaphenazole and L-NMMA (Table 3, Fig. 3).
Results obtained with papaverine were identical to those seen with SNP. Indeed, sulfaphenazole did not change the response to ACH, and L-NMMA blunted the response to ACH. When ACH was repeated with simultaneous sulfaphenazole and L-NMMA, a further inhibition of response to ACH was seen (data not shown).
Effect of vitamin C on sulfaphenazole-induced inhibition of vasodilation to ACH and BDK in hypertensive patients
In this group of hypertensive patients, the vasodilation to ACH was inhibited by sulfaphenazole and increased by vitamin C (Fig. 4).Of importance, sulfaphenazole, when retested under vitamin C, no longer inhibited the response to ACH (Figs. 2 and 4). Similarly, vasodilation to BDK was blunted by sulfaphenazole and potentiated by vitamin C (Fig. 4). When co-infused with the antioxidant, sulfaphenazole was no longer effective (Figs. 2 and 4).
In the final group of hypertensive patients, the vasodilation to ACH (saline: from 3.3 ± 0.2 ml/100 ml/min to 16.2 ± 2.8 ml/100 ml/min, +391%) was resistant to L-NMMA (baseline: 3.3 ± 0.4 ml/100 ml/min; L-NMMA: 2.1 ± 0.4 ml/100 ml/min; L-NMMA + SNP: 3.3 ± 0.4 ml/100 ml/min; L-NMMA + SNP and ACH: 16.0 ± 3.5 ml/100 ml/min, +384%). Under vitamin C, response to ACH was significantly improved (from 3.3 ± 0.4 ml/100 ml/min to 22.0 ± 3.9 ml/100 ml/min, +567%; p < 0.001 vs. ACH alone), and the inhibiting effect of L-NMMA on ACH was restored (baseline: 3.3 ± 0.2 ml/100 ml/min; vitamin C + L-NMMA: 2.0 ± 0.3 ml/100 ml/min; vitamin C + L-NMMA and SNP: 3.3 ± 0.3 ml/100 ml/min; vitamin C + L-NMMA, SNP, and ACH: 13.1 ± 1.8 ml/100 ml/min, +296%).
In both normotensive subjects and hypertensive patients, contralateral FBF did not significantly change throughout the study (data not shown).
The present results confirm that vasodilation to ACH and BDK, 2 specific endothelial agonists that stimulate G-protein–coupled receptors, but not to SNP, a dilator acting directly on smooth muscle cells, was blunted in hypertensive patients as compared with control subjects, confirming the presence of endothelial dysfunction in the peripheral circulation of essential hypertensive patients (2–6). In addition, while in healthy subjects the vasodilator responses to ACH and BDK were blunted by L-NMMA, an inhibitor of NO synthase, indicating preserved NO availability, in essential hypertensive patients vasodilation elicited by endothelial agonists was found to be resistant to NO inhibition, demonstrating the presence of impaired NO availability (5,6). Another relevant finding is that, in the same study subgroups, sulfaphenazole, a highly selective CYP 2C9 inhibitor, blunted the vasodilation elicited by ACH and BDK in hypertensive patients only, whereas it was ineffective in healthy subjects. Furthermore, the degree of inhibition exerted by the CYP 2C9 inhibitor was greater on BDK- as compared with ACH-induced vasodilation. Finally, the inhibitory effect of sulfaphenazole on endothelial agonists was specific because the CYP 2C9 inhibitor did not change the response to SNP.
It is conceivable that, in these experimental conditions, CYP 2C9-induced vasodilation could be mediated by EDHF production, as indirectly supported by previous evidence. Bradykinin-mediated, NO- and prostacyclin-independent subcutaneous resistance artery vasodilation in healthy humans can be suppressed by high intravascular K+concentrations, suggesting the involvement of an EDHF (18). Moreover, a CYP 2C isoform is involved in the ACH-induced EDHF generation in hamster gracilis muscle (11,19). In contrast, the present study does not allow identification of the chemical nature of the EDHF released by CYP 2C9 stimulation. In this regard, in vitro evidence demonstrates that this enzyme can generate both 11,12- and 14,15-EET (7). Taken together, these results indicate that, while under healthy conditions, endothelium-dependent vasodilatation seems to be mainly dependent on NO production; in essential hypertensive patients it is instead related to alternative pathways, including production of CYP 2C9-derived EDHF, which is activated to compensate for the impaired NO availability. This possibility is further reinforced by the study evaluating the effect of sulfaphenazole on vasodilation to ACH or BDK residual to L-NMMA and COX blockade in healthy subjects. In this experimental condition of decreased NO availability, sulfaphenazole, while proving to be ineffective when tested in control conditions, produced a further decrease in the response to both endothelial agonists. These results are in perfect agreement with available evidence indicating that, in the forearm microcirculation of healthy subjects, miconazole, a non-selective CYP inhibitor, is ineffective on vasodilation to BDK in baseline conditions (13). However, after NO synthase and COX inhibition by L-NMMA and aspirin, BDK-mediated vasodilation was suppressed by miconazole, confirming that the EDHF contribution plays a primary role particularly in conditions characterized by deranged NO availability (13). Therefore, these series of observations suggest that, in the peripheral human circulation, a CYP 2C9-derived pathway may operate as a rapid compensatory mechanism for decreased NO activity. In this scenario, we cannot exclude that BP levels might be influenced by the magnitude of EDHF production. However, the relatively small number of hypertensive patients recruited in our study was characterized by a narrow range of BP values. Therefore, in such conditions, the expected relationship between BP and response to sulfaphenazole would not have been put in evidence. This important issue awaits clarification in future studies.
The cross-talk between NO and CYP pathways is further reinforced by the demonstration that neither L-NMMA nor sulphafenazole alone affect the increase in skeletal smooth muscle blood flow induced by dynamic exercise (20). Only the combination of the 2 inhibitors is able to inhibit the flow increase, thereby demonstrating that an interaction between the 2 systems appears to exist, with EDHF-dependent vasodilation taking over when NO availability is compromised.
The hypothetical cross-talk between NO and the EDHF pathway is not, however, in agreement with the possibility that exogenous NO derived from SNP utilized during the “NO clamp” could interfere with the effect of sulfaphenazole on response to ACH. However, in adjunctive experiments, papaverine, a phosphodiesterase activity inhibitor, provided superimposable results. This apparent contrast might be explained by the low dose of the compound utilized during the “NO clamp” technique. Indeed, such concentration, able to increase the FBF around 65%, cannot release enough exogenous NO to affect the EDHF pathway. It is worth noting that the NO released by ACH accounts for a greater amount of vasodilation (around 600%). However, because our experimental conditions do not allow us to investigate these intracellular mechanisms, we recognize our hypothesis as a speculative interpretation of our findings.
Further confirmation for this hypothesis comes from the study with intrabrachial vitamin C. In agreement with earlier evidence, the antioxidant increases the response to both ACH and BDK in essential hypertensive patients (5,6) and restores the sensitivity to L-NMMA, confirming that oxidative stress may be the main mechanism leading to impaired endothelium-dependent vasodilation by NO breakdown in essential hypertension. However, the relevant finding is that, under vitamin C administration, sulfaphenazole proved no longer effective in blunting the response to ACH and BDK. These results seem to indicate that, in essential hypertension, CYP 2C9-derived EDHF acts as a partial compensatory mechanism to sustain endothelial function when NO activity is impaired through the presence of oxidative stress. Similar findings were obtained in essential hypertensive patients, by preventing BDK-induced smooth muscle cell hyperpolarization with ouabain, a Na+K+/ATPase inhibitor (6). The present results extend this previous information by identifying the nature of at least 1 of the EDHFs involved.
The fact that the EDHF-mediated response becomes evident only in the presence of impaired NO availability is in line with experimental evidence indicating that physiological concentration of endothelial NO is sufficient to dampen EDHF generation (21). In microsomes and endothelial cells, NO donors can inhibit CYP activity (7). Different mechanisms can be proposed to account for the NO-induced inactivation of the CYP enzyme, including release of heme from CYP proteins, inhibition of CYP protein synthesis, or reaction between NO and superoxide anions to form peroxynitrite, which, in turn, may affect CYP expression or function (7).
Another issue that can be addressed by the present results is the finding that the degree of inhibition exerted by sulphafenazole on BDK is greater than that exerted on ACH. This result is not easily interpretable. There is evidence indicating that an increase in calcium influx, induced either as a result of endothelial surface B2 receptor stimulation by BDK or through muscarinic receptor activation by ACH, stimulates endothelial phospholipase A2 to release arachidonic acid, which can, in turn, be metabolized by CYP enzymes (in addition to COX and lipoxygenase) (22). Thus, it is conceivable that these 2 agonists, which, however, act through different signal transduction pathways, might exert a different degree of CYP 2C9 activation, despite a similar vasodilating effect. On the other hand, available evidence in essential hypertension demonstrates that while BDK is sensitive to ouabain (6,23), ACH is resistant to the Na+K+/ATPase inhibitor in baseline conditions, although a certain effect can be unmasked after adequate stimulation with insulin (23). Therefore, it cannot be ruled out that ACH-induced NO-independent vasodilation could be greatly dependent on pathways different from CYP 2C9-mediated activation.
Finally, the present results do not support the possibility that, in essential hypertensive patients, CYP 2C9 might be a relevant source of oxidative stress, as demonstrated in patients with coronary artery disease (15). This discrepancy could be explained by the different degree of cardiovascular risk characterizing these 2 study populations. Because it is demonstrated that the degree of endothelial dysfunction is directly related to total cardiovascular risk (24,25), it is conceivable that CYP 2C9 shifts from production of relaxing EDHF to a role as a major source of oxidative stress. Such a shift would represent 1 of the pathogenetic events associated with the increase in cardiovascular risk.
In conclusion, the results outlined in this paper provide further information concerning the mechanisms responsible for impaired endothelium-dependent vasodilation in essential hypertension. Because it is becoming increasingly clear that endothelial dysfunction is an independent promoter of atherosclerosis and cardiovascular events (1), it is very likely that, in the near future, the determination of endothelium-dependent relaxation could be utilized for quantification of cardiovascular risk. In addition, restoration of endothelial function might become an adjunctive target of antihypertensive therapy. Thus, identification of the pathways that characterize endothelial dysfunction can increase knowledge on the pathophysiology of atherosclerotic disease and provide additional potential for the development of specific strategies to improve endothelial dysfunction.
- Abbreviations and Acronyms
- blood pressure
- CYP 2C
- cytochrome P450 epoxygenase
- endothelium-derived hyperpolarizing factor
- epoxyeicosatrienoic acid
- forearm blood flow
- nitric oxide
- sodium nitroprusside
- Received September 14, 2005.
- Revision received March 27, 2006.
- Accepted April 11, 2006.
- American College of Cardiology Foundation
- Panza J.A.,
- García C.E.,
- Kilcoyne C.M.,
- Quyyumi A.,
- Cannon R.O. III.
- Taddei S.,
- Virdis A.,
- Ghiadoni L.,
- Magagna A.,
- Salvetti A.
- Taddei S.,
- Virdis A.,
- Taddei S.,
- et al.
- Campbell W.B.,
- Gebremedhin D.,
- Pratt P.F.,
- Harder D.R.
- Archer S.L.,
- Gragasin F.S.,
- Wu X.,
- et al.
- Halcox J.P.,
- Narayanan S.,
- Cramer-Joyce L.,
- Mincemoyer R.,
- Quyyumi A.A.
- Fleming I.,
- Michaelis U.R.,
- Bredenkotter D.,
- et al.
- Fichtlscherer S.,
- Dimmeler S.,
- Breuer S.,
- Busse R.,
- Zeiher A.M.,
- Fleming I.
- Virdis A.,
- Taddei S.,
- Ghiadoni L.,
- et al.
- Frei B.,
- England L.,
- Ames B.N.
- Coats P.,
- Johnston F.,
- MacDonald J.,
- McMurray J.J.V.,
- Hillier C.
- Bolz S.B.,
- Fisslthaler B.,
- Pieperhoff S.,
- et al.
- Bauersachs J.,
- Popp R.,
- Hecker M.,
- Sauer E.,
- Fleming I.,
- Busse R.
- Feletou M.,
- Vanhoutte P.M.
- Taddei S.,
- Virdis A.,
- Mattei P.,
- Natali A.,
- Ferrannini E.,
- Salvetti A.
- Vita J.A.,
- Treasure C.B.,
- Nabel E.G.,
- et al.
- Benjamin E.J.,
- Larson M.G.,
- Keyes M.J.,
- et al.