WO2013124478A1

WO2013124478A1 - Symmetric dimethylarginine (sdma) modifies high density lipoprotein (hdl) to induce endothelial dysfunction

Info

Publication number: WO2013124478A1
Application number: PCT/EP2013/053670
Authority: WO
Inventors: Ferdinand Hermann Bahlmann; Danilo Fliser; Thimotheus SPEER
Original assignee: Universität des Saarlandes
Priority date: 2012-02-24
Filing date: 2013-02-25
Publication date: 2013-08-29

Abstract

Use of SDMA for determination of a HDL-SDMA level in blood or blood plasma and Use of SDMA for the evaluation of the benefit of an admin istration of a medicament for effecting an elevation of the HDL-level in patient with a HDL-level which should be increased or modified.

Description

Symmetric Dimethylarginine (SDMA) Modifies High Density Lipoprotein (HDL) to Induce Endothelial Dysfunction

The present invention pertains to a method of use of SDMA for determination of a HDL-SDMA complex level in blood or blood-plasma : a method for determining the responsiveness of patients suffering an elevated cardiovascular risk to a treatment with HDL elevating and/or modifying drugs.

Atherosclerotic cardiovascular disease and hypertension are highly prevalent in Western populations and are the leading cause of death. Endothelial injury and dysfunction (ED) represent a common link of all cardiovascular risk factors acting upon the vascular system to promote development and progression of atherosclerosis and hypertension. An important characteristic of ED is the dysbalance between reduced athero-protective endothelial nitric oxide (NO) availability and an increased production of reactive oxygen species (ROS), promoting vasoconstriction, inflammatory activation and impaired endothelial repair.

Interestingly, lipids and lipoproteins critical in atherogenesis are also involved in immunological processes of the innate and adaptive immune response. Arterial endothelial cells express signaling pattern-recognition receptors of the innate immune system, such as TLR-1 , TLR-2 and TLR-4. Innate immunity is mainly based on recognition of pathogens by pattern recognition receptors. Among these receptors, Toll-like receptors (TLR) have been studied extensively. TLR recognize a broad variety of exogenous pathogen-associated molecular pattern (PAMP) and endogenous danger-associated molecular pattern (DAMP). TLRs have been suggested to play a crucial rule in the pathogenesis of atherosclerotic diseases. Recent evidence suggests that known atherogenic lipids, i.e. oxidized phospholipids and oxidized low-density lipoprotein (LDL), can trigger a TLR-2 dependent response and apoptosis in macrophages. However, the mode of activation of en- dothelial TLRs in atherosclerosis and their relation to endothelial dysfunction remains to be characterized.

Based on a rather rigid picture of atherosclerosis, for a long time, LDL was thought to promote the development of ED and high-density lipoprotein (HDL) was thought to prevent the development of ED and atherosclerosis. However, recent evidence suggests that endothelial effects of HDL can be highly heterogeneous under several disease conditions (Charakida et al., 2009 JAMA 302, 1210- 1217; Sorrentino et al., 2010 Circulation 121, 110-122). Furthermore, in contrast to observational trials revealing a protective effect of high HDL plasma concentrations, a recent study documented, that genetic mechanisms raising HDL cholesterol plasma concentrations are not associated with a lower risk for myocardial infarction (Voight et al., 2012 Lancet doi : 10.1016/S0140- 6736(12)60312-2.). To determine whether modification of HDL may be mechanistically involved in these controversial findings, the inventors isolated HDL from patients with chronic kidney disease (CKD) as a population with a particular high risk for cardiovascular events and mortality and a high prevalence of ED and hypertension.

CKD is highly prevalent in Western populations and recent large epidemiological studies have suggested that up to 10% of the population is affected . In patients with CKD the risk of CV disease and death is substantially increased as compared to the general population, and most of these patients die before reaching the final stage of CKD. Moreover, even in an apparently healthy unselected population, mild impairment of kidney function, i.e. glomerular filtration rate (GFR) be- low 90 ml/min/1.73m², was identified as a strong independent predictor of CV mortality within the next 10 years. The underlying cause(s) remain poorly understood, however, endothelial dysfunction and an accelerated atherosclerosis likely play a pivotal role in particular in the early stages of CKD. As kidney function deteriorates, several substances normally excreted by the kidneys accumulate, and water-soluble metabolites of L-arginine such as asymmetric dimethylarginine (ADMA), an endogenous inhibitor of endothelial NO synthase (eNOS), have recently come into the focus of CV medicine. By inhibiting NO formation, ADMA causes endothelial dysfunction, vasoconstriction, and may increase blood pressure. Less attention has been paid to its structural isomer symmetric dimethy- larginine (SDMA). Although SDMA blood levels rise in parallel with declining GFR (SDMA is an excellent marker of kidney function (Kielstein et al 2006 Nephrol Dial Transplant. 21(9) : 2446-51), its pathophysiological role remains largely undefined and SDMA is thought to be functionally inactive. More recent studies have linked SDMA with increased oxidative stress and atherosclerosis and high- lighted its predictive value for cardiovascular Dyslipidemia in patients with CKD frequently involves low high density lipoprotein (HDL) cholesterol levels, leading to the question whether HDL may represent a therapeutic target in these patients. Speer et al report in Circulation 2011; 124: A11343 High Density Lipoprotein (HDL) from patients with early CKD exerts ad- verse endothelial effects about data which demonstrate that important endothe- lial-protective effects of HDL, such as stimulation of endothelial NO production and promotion of endothelial repair, are substantially altered in patients with CKD, already in the early stages of the disease. Focusing on HDL, the present invention discloses to the determination whether modification of HDL may alter its endothelial effects by modulation of the innate immune system.

Banach et al disclose in Med Sci Monit. 2009 Dec; 15(12) : MSl-5 : "Time for new indications for statins?" that the role of statins in the treatment and prevention of cardiovascular diseases, such as coronary artery disease, acute coronary syndromes, diabetes or stroke is well established. However, there are still some questions regarding the role of statins in patients with heart failure, hypertension, atrial fibrillation and chronic kidney disease (CKD). The review discusses the current evidence regarding the use of statins in patients with hypertension (with or without dyslipidemia) or CKD. They conclude that new large randomized clinical trials are needed to confirm the potentially beneficial positive results ob- tained so far.

Kiechl et al. disclose in Atherosclerosis 205 (2009) 261-265, that asymmetric and symmetric dimethylarginines are of similar predictive value for cardiovascular risk in the general population.

Summary of the invention

The present invention demonstrates that HDL from patients with mild to severe CKD (HDL^CKD), in contrast to HDL from healthy subjects, inhibits endothelial NO production and increases blood pressure. HDL^CKD suppresses endothelial repair and promotes endothelial inflammatory activation. The inventors hypothesized that ADMA, an endogenous inhibitor of eNOS that accumulates in the plasma in different clinical disease conditions (e.g . chronic kidney disease), may be involved in this process. Plasma was fractionated from healthy subjects and patients with reduced kidney function and ADMA was measured in the HDL fraction using a mass spectrometry approach (ESI-MS/MS). However, although ADMA serum levels were elevated in patients with chronic kidney diseases, they could not detect significant amounts of ADMA in the HDL fraction. Surprisingly, the inventors found instead SDMA, a structural isomer of ADMA, in the HDL fraction in a substantial amount, but remarkably only in samples obtained from patients with reduced kidney function. The inventors newly report that symmetric di- methylarginine (SDMA), which accumulates in CKD, is specifically associated to HDL^CKD and alters its endothelial effects, so that physiological HDL converts into "noxious" HDL. Interestingly, only HDL supplemented with SDMA inhibited endothelial NO production, while neither supplementation of HDL with ADMA nor sup- plementation of LDL with ADMA and SDMA caused any significant changes of the effects of the lipoproteins on endothelial NO release. Adverse endothelial effects of HDL^CKD or HDL^SDMA are prevented by blockade or knock-out of Toll-like- receptor(TLR)-2, known to be activated by microbial lipoproteins. Taken together, these data describe a novel pathomechanism leading to a HDL phenotype promoting hypertension and likely accelerating atherosclerosis even in incipient CKD, which may contribute to elevated cardiovascular morbidity and mortality.

The present invention is based on the result that SDMA associates to HDL under formation of a SDMA-HDL complex and that accumulation of SDMA in HDL leads to a dysfunctional HDL and thereby contributes to the increased CV morbidity and mortality in conditions with a minor impairment of kidney function.

Subject matter of the invention is a method of use of SDMA for determination of a HDL-SDMA level in blood or blood-plasma. In particular SDMA is used for the determination of a level of dysfunctional HDL-SDMA complex, or for the determination of a level of dysfunctional HDL associated SDMA leading to endothelial dysfunction.

The HDL-SDMA level can be measured by Tandem mass spectrometry (MS/MS), HPLC Amino acid profiling and/or ELISA.

According to the invention the determination of the HDL-SDMA level allows an evaluation of the benefit of an administration of a medicament for effecting an elevation of the HDL-level in a patient having a HDL-level which should be increased or HDL that should be modified by drug treatment, e.g . administering an agent which prevents association of SDMA to HDL but keeping essentially the vasoprotective effect of HDL or reestablishing that effect. Accordingly, the amount of SDMA associated with HDL is determined yielding a HDL-SDMA level and if the SDMA concentration associated with HDL (H DL-SDMA) is about _> 0.025 pmol / g protein the medicament for elevation of the HDL-level is not administered to the patient.

Brief Description of the Figures

Figure 1 : HDL from CKD patients inhibits endothelial NO production .

Figure 2 : ADMA and SDMA levels. Figure 3 : HDL supplemented with SDMA induces endothelial dysfunction and hypertension .

Figure 4 : Endothelial cell migration after incubation with HDL.

Figure 5 : Modified HDL induces endothelial inflammation and suppresses endothelial repair.

Figure 6 : Modified HDL activates TLR-2 to induce endothelial dysfunction and hypertension .

Detailed Description of the Invention

The invention demonstrates for the first time that HDL from CKD patients substantially inhibits endothelial nitric oxide production and endothelial repair and increases arterial blood pressure. Accumulation of SDMA in H DL in CKD has been identified as a culprit leading to dysfunctional HDL^CKD and the lipoprotein- sensitive endothelial TLR-2 as the receptor which mediates adverse endothelial effects of both HDL^CKD and SDMA-enriched HDL. These observations allow a new diagnostic evaluation of patients with low HDL levels as for example CKD patients based on this novel pathomechanism present already in the earliest stages of endothelial injury and dysfunction promoting hypertension and likely the acceleration of CV disease.

Endothelial dysfunction - an early step in the development of CV disease - is a predictor of death in the general population and in CKD patients. In healthy individuals, endothelial NO generation stimulated by H DL serves as an important stimulus to preserve endothelial function . Consistently, Spieker et al . could show that infusion of H DL improves endothelial dysfunction in hypercholesterolemic subjects by increasing NO bioavailability (Spieker et al . 2002 Circulation 105, 1399- 1402) . By contrast, the present invention discloses that HDL^CKD strongly inhibits endothelial NO production in a dose dependent manner. Importantly, this adverse effect was already present in H DL of patients with a minor impairment of kidney function as well as in HDL from children with an impaired kidney function, indicating that chronic kidney disease per se and not concomitant diseases are re- sponsible for these adverse effects of HDL on endothelial NO production . Moreover, infusion of HDL^CKD into mice led to a significant increase of blood pressure, supporting the notion that modified HDL in CKD patients could at least aggravate if not initiate hypertension . This observation might be of overriding clinical importance, since hypertension is a strong promoter of progression and mortality in CKD. In addition and in contrast to HDL^healthy which promotes the repair of vascular lesions; HDL^CKD inhibited the re-endothelialization of endothelial lesions both in vitro and in vivo. Finally, HDL^CKD was found to act as a pro-inflammatory rather than anti-inflammatory stimulus, in contrast to HDL^Healthy. The invention demonstrates the inhibition of endothelial NO production as a key principle for the pro-inflammatory effects of modified H DL in CKD. HDL^CKD may thereby aggravate vascular inflammation, which is a common feature of CKD.

ADMA, a methylation product of L-arginine residues, is known as an endogenous eNOS inhibitor in patients with CKD and other diseases. Therefore the inventors hypothesized that ADMA may be involved in the process. However, although ADMA serum levels were elevated in patients with chronic kidney diseases, they could not detect significant amounts of ADMA in the H DL fraction . The inventors surprisingly found instead its structural isomer SDMA in the HDL fraction in a substantial amount, but only in samples obtained from patients with reduced kidney function . To address whether SDMA converts physiological HDL into "nox- ious" HDL, the inventors supplemented H DL and LDL from healthy donors with SDMA or ADMA and measured subsequently the effects on endothelial NO production

Interestingly, only H DL supplemented with SDMA inhibited endothelial NO production, while neither supplementation of HDL with ADMA nor supplementation of LDL with ADMA and SDMA caused any significant changes of the effects of the lipoproteins on endothelial NO release. The amount of SDMA in the HDL fraction after supplementation was comparable to that observed in HDL^CKD. Notably, SDMA alone, in the absence of HDL, did not significantly change endothelial NO production. Supplementation of HDL (HDL^SDMA) with increasing concentrations of SDMA dose dependently suppressed endothelial cell NO production. Similarly to the effect of HDL from patients with reduced kidney function, HDL^SDMA induced basal endothelial superoxide production These findings indicate that the association of SDMA in the HDL may be crucially involved in the deterioration of HDL's protective effects and its change into a pro-athergenic agent.

Collectively, the results on which the invention is based support the notion that the vascular effects of HDL are modified by binding of SDMA for example in the clinical setting of CKD. Such dysfunctional HDL turns into "noxious" HDL and may contribute to increased CV morbidity and mortality as it is observed even in patients with mild kidney impairment. TLR-2 and TLR-4 represent receptors of the innate immune system involved in recognition of pathogen-associated lipoproteins. Both, TLR-2 and TLR-4 are highly expressed on HAECs. The inventors measured endothelial NO production in HAECs stimulated with HDL^Healthy, HDL^CKD or HDL^SDMA in the presence or absence of specific neutralizing antibodies against TLR-2 and TLR-4. They observed that blocking of TLR-2 but not TLR-4 restored the effects of HDL^CKD and HDL^SDMA on endothelial NO production . Importantly, the neutralizing antibodies did not affect NO production in the presence of HDL^Healthy. Consistently, injection of HDL^CKD into Tlr2-/- mice, but not in Tlr4-/- mice, failed to increase ABP.

These findings have further impact: First, in general dysfunctional (lipo)proteins contribute to the CV disease burden of patients at increased risk of kidney injury as well as subjects at risk for reduced kidney function such as elderly. Second, in the general population, reduced HDL-cholesterol levels are associated with an increased risk for coronary disease. Low HDL-cholesterol levels remain predictive for CV events in patients treated with statins who have low LDL cholesterol lev- els. Several attempts have therefore been made to increase HDL-cholesterol levels in patients at high CV risk, for example by using CETP inhibitors. However, rising of the HDL level is counterproductive if SDMA is present in amounts yielding critical concentrations of SDMA-HDL complexes.

The invention is further described by means of the following non-limiting exam- pies. Examples

Participants

To examine the endothelial effects of H DL in a cardiovascular "high-risk" population, the investors isolated HDL from adults (n =45) and children (n = 22) with different degrees of chronic kidney dysfunction and age- and gender-matched healthy control subjects (n = 15 adults and n = 10 children . Baseline demographic and clinical characteristics of the study population are shown in Table 1 & 2.

HDL from CKD patients induces endothelial dysfunction and increases blood pressure NO production has been analysed in human aortic endothelial cells (HAEC) incubated with H DL isolated from adult patients with chronic kidney dysfunction (H DL^CKD) and from corresponding healthy subjects (HDL^Healthy) by electron spin resonance (ESR) spectroscopy. HDL^Healthy stimulated endothelial NO production, but in marked contrast, H DL^CKD strongly inhibited NO production in HAECs (Figure 1A) . Notably, H DL from patients with mildly reduced kidney function already substantially inhibited endothelial cell NO release (Figure 1A). Both effects were observed to be dose-dependent (Figure I B). To examine whether chronic kidney disease per se and not concomitant diseases are responsible for these adverse effects of H DL on endothelial NO production, H DL was isolated from children with an impaired kidney function and its effects on the endothelial NO production was measured . Importantly, HDL^CKD from children substantially inhibited endothelial NO production, whereas HDL^Healthy from children induced endothelial NO production (Figure 1C) .

A reduced endothelial NO production has been observed to increase blood pres- sure, the inventors therefore examined the differential effects of HDL on blood pressure to obtain further evidence for differential effects of HDL^CKD in vivo. l_l_DLHeaithy _{gs we}|| _gs j-jQj_ KD _were injected into mice and their blood pressure was measured before and 90 minutes after injection . DL^Healthy decreases systolic blood pressure (- 12 ± 2 mmHg), whereas HDL^CKD significantly increased blood pressure in vivo (+9 ± 2 mmHg, Figure Id) .

The effect of HDL was explored on endothelial cell superoxide production in HAECs and it was observed that HDL^CKD strongly induced endothelial superoxide production . In contrast, treatment with H DL^Healthy did not change basal endothelial superoxide levels (Figure le) . SDMA but not ADMA is associated to HDL and inhibits endothelial NO production

It was hypothesized that ADMA, an endogenous inhibitor of eNOS that accumulates in the plasma in different clinical disease conditions (e.g. chronic kidney disease), may be involved in this process. Plasma was fractionated from healthy subjects and patients with reduced kidney function and measured ADMA in the HDL fraction using a mass spectrometry approach (ESI-MS/MS). However, although ADMA serum levels were elevated in patients with chronic kidney diseases (Figure 2), no significant amounts of ADMA could be detected in the HDL fraction. Instead, its structural isomer SDMA was found in the HDL fraction in a substantial amount, but only in samples obtained from patients with reduced kidney function (Figure 3a).

To address whether SDMA converts physiological HDL into "noxious" HDL, the inventors supplemented HDL and LDL from healthy donors with SDMA or ADMA and measured subsequently the effects on endothelial NO production. Interestingly, only HDL supplemented with SDMA inhibited endothelial NO production, while neither supplementation of HDL with ADMA nor supplementation of LDL with ADMA and SDMA caused any significant changes of the effects of the lipoproteins on endothelial NO release. The amount of SDMA in the HDL fraction af- ter supplementation was comparable to that observed in HDL^CKD. Notably, SDMA alone, in the absence of HDL, did not significantly change endothelial NO production (Figure 3B). Supplementation of HDL (HDL^SDMA) with increasing concentrations of SDMA dose dependently suppressed endothelial cell NO production (Figure 3C). Similarly to the effect of HDL from patients with reduced kidney func- tion, HDL^SDMA induced basal endothelial superoxide production (Figure 3D).

Furthermore, reconstituted HDL (rHDL) was supplemented consisting of Apo- Al : POPC: cholesterol in a molar ratio of 1 : 100 : 10 with and without SDMA and measured its effect on endothelial cell NO production was measured. Supplementation of rHDL with SDMA reduced endothelial NO production (Figure 3E). More- over, after supplementation with SDMA, apolipoprotein Al (Apo-Al) inhibited endothelial NO production, whereas Apo-Al without SDMA did not significantly affect endothelial NO production (Figure 3E). This indicates that SDMA may associate with Apo-Al, the major apolipoprotein of HDL. HDL from healthy subjects is known to stimulate endothelial NO production by Akt dependent eNOS phosphorylation via activation of endothelial scavenger receptor B-I (SR10 BI) or sphingosine-1 phosphate receptors. Here, they found - in contrast to HDL^Healthy - that HDL^CKD as well as HDL^SDMA significantly reduced phosphorylation of Akt (Ser473) leading to a reduced eNOS-activating phosphorylation (Serl l77) and a significantly enhanced eNOS-inhibiting phosphorylation (Thr495) (Figure 3F-I). Therefore, it was hypothesized that such modified HDL changes its affinity to these endothelial receptors mediating the protective endothelial effects of HDL^Healthy.

HDL from CKD patients inhibits endothelial repair mechanisms

Several studies previously demonstrated that reduced NO bioavailability along with enhanced superoxide production may also impair other endothelial functions, such as endothelial repair after injury and its anti-inflammatory capacity. To elucidate whether modified HDL also affects these endothelial properties by inhibiting endothelial NO production, the inventors assessed the effect of modified HDL on endothelial repair and inflammation.

In an endothelial wound healing assay, HDL^Healthy strongly stimulated endothelial cell migration, whereas HDL^CKD inhibited endothelial migration (Figure 4 A+B). In line with these in vitro observations, HDL^CKD and HDL^SDMA inhibited endothelial repair in vivo in a carotid artery injury model (Figure 5A + B). This was in marked contrast to HDL^Healthy, which promoted endothelial repair. In Tlr2-/- mice, we observed that HDL^CKD regained its capacity to promote endothelial repair (Figure 5C). These findings underscore the important role of TLR- 2 to mediate adverse endothelial effects of HDL^CKD and HDL^SDMA.

HDL from CKD patients induces inflammatory endothelial activation

The effect of HDL was examined on endothelial mononuclear cell (MNC) adhesion. It was observed that HDL^CKD increased endothelial pro-inflammatory activation as determined by a higher amount of adhering mononuclear cells to a TNFa- stimulated endothelial monolayer, which was in contrast to HDL^Healthy (Figure 5 D+E). Accordingly, HDL^CKD and HDL^SDMA increased endothelial VCAM-1 expression, while HDL^Healthy significantly reduced VCAM-1 expression on endothelial cells (Figure 5F). Next, the effect of NO on TNFa induced VCAM-1 expression in the presence of HDL was analysed . Inhibition of eNOS through L-NAME reduced the ^ ^ anti-inflammatory effects of HDL^Healthy by enhancing endothelial VCAM- 1 expression, whereas supplementation of NO using a solid NO-donor, diethylenetria- mine/nitric oxide (DetaNO), restored the anti-inflammatory function of the endothelium by reducing endothelial VCAM- 1 expression (Figure 5F) . Taken together, these results clearly suggest that a reduced endothelial NO bioavailability induced by HDL^CKD and HDL^SDMA represents a major mechanism explaining adverse endothelial effects of modified HDL.

Adverse endothelial effects of HDL from CKD patients are mediated by TLR-2 TLR-2 and TLR-4 represent receptors of the innate immune system involved in recognition of pathogen-associated lipoproteins. Both, TLR-2 and TLR-4 are highly expressed on HAECs (Figure S5A+ B) . Endothelial NO production was measured in HAECs stimulated with H DL^Healthy, HDL^CKD or HDL^SDMA in the presence or absence of specific neutralizing antibodies against TLR-2 and TLR-4. It was observed that blocking of TLR-2 but not TLR-4 restored the effects of H DL^CKD and I_I _D I_SDMA _{on enc}j₀thelial NO production (Figure 6C) . Importantly, the neutralizing antibodies did not affect NO production in the presence of H DL^Healthy. Consistently, injection of HDL^CKD into Tlr2-/- mice, but not in Tlr4-/- mice, failed to increase ABP (Figure 6D) .

Figure legends

Figure 1 : HDL from CKD patients inhibits endothelial NO production

(a) Effect of HDL (50 pg/mL) from healthy subjects and patients with different stages of kidney disease (CKD-V°, CKD-III-IV°, CKD-II°) on endothelial NO production determined by ESR spectroscopy (n = 15 per group) .

(b) Dose dependent effect of H DL from healthy subjects and patients with end stage kidney disease (CKD-V°) on endothelial NO production ( lhr, n = 3-5 per group) .

(c) NO production in HAEC determined by ESR spectroscopy after incubation with HDL (50 pg/rnl, 1 hr) from children with kidney dysfunction and healthy children as control (n =9- 13 per group) . (d) Effect of HDL injection on systolic blood pressure in mice. Systolic blood pressure before and 90 minutes after injection of HDL ( 15 mg/kg) from healthy subjects or patients with CKD (n = 5-6 per group) .

(e) Superoxide production in HAEC determined by ESR spectroscopy after in- cubation with HDL (50 pg/rnl, 1 hr, n =6 per group).

All data are presented as mean ± SEM .

Figure 2 : ADMA and SDMA levels

(a) ADMA and SDMA levels in the serum healthy subjects and patients with kidney dysfunction determined by ESI-5 MS/MS (n = 10 per group) .

(b) Representative mass chromatograms obtained for the standard mixture of SDMA, ADMA and ADMA-D7. Following transitions were used for quantification : m/z 259 > 227 (SDMA), m/z 259 > 214 (ADMA) and m/z 266 > 77 (ADMA-D7). The concentrations in this mixture were 4 pmol/L for SDMA, ADMA and ADMA-D7, respectively. We obtained the following concentrations : SDMA 3.973 pmol/L, ADMA 3.892 prnol/L

(c-d) ESI-MS/MS product ion spectra pattern obtained for (C) ADMA and (D) SDMA. Specific fragments used for quantifications are marked with arrows (ADMA 214 m/z, SDMA 228 m/z).

(e-f) Representative mass chromatograms obtained from plasma (E) of a healthy subject and (F) a patient with kidney dysfunction .

Figure 3 : HDL supplemented with SDMA induces endothelial dysfunction and hypertension (A) ADMA and SDMA levels associated to HDL from healthy subjects and patients with kidney dysfunction determined by ESI-MS/MS (n = 10 per group) . (B) Effect of SDMA alone (4 μΜ), healthy HDL (50 pg/ml) or LDL (100 pg/ml) supplemented with or without SDMA or ADMA (4 μΜ equivalent to 0.5 prnol/g lipoprotein) on endothelial NO production (n = 3-6 per group) .

(C) Effect of H DL^Healthy supplemented with different concentrations of SDMA (0.5 prnol/g HDL protein) on endothelial NO production (n = 3 per group) .

(D) Effect of H DL (50 pg/ml, 1 hr) supplemented with SDMA (0.5 pmol/g HDL protein) on endothelial superoxide production (n = 6 per group) . (E) Effect of rHDL (50 Mg/ml) or Apo-Al (25 Mg/ml) supplemented with SDMA (0.5 pmol/g protein) on endothelial NO production (n=4-6 per group).

(F) Phosphorylation of Akt at Ser473, eNOS activating phosphorylation at Serl l77 and eNOS inhibiting phosphorylation at Thr495 determined by western blot analysis in HAEC incubated with HDL (50 Mg/ml) for 10 min as indicated (blot as representative of at least 3 independent experiments).

(G) Quantification of Akt phosphorylation at Ser473 in HAEC incubated with HDL (50 Mg/ml) for 10 min normalized to total Akt expression (n = 3-6).

(H) Quantification of eNOS-activating phosphorylation at Serl l77 in HAEC incubated with HDL (50 M9/^m for 10 min normalized to total eNOS expression (n = 3-6 per group).

(I) Quantification of eNOS-inhibiting phosphorylation at Thr495 in HAEC incubated with HDL (50 M9/^m for 10 min normalized to total eNOS expression (n = 3-6 per group).

All data are presented as mean ± SEM .

Figure 4:

(A) Endothelial cell migration after incubation with HDL (50 g/ml, 24 hr) in a scratch assay (n=4-8 per group).

(B) Representative micrographs of endothelial cell migration in a scratch assay after incubation with HDL.

Figure 5 : Modified HDL induces endothelial inflammation and suppresses endo- thelial repair

(A) Reendothelialized area at day 3 after carotid injury and injection of HDL^Healthy, HDL^CKD or HDL^SDMA (15 mg/kg) in nude mice. PBS treated animals serve as control (n = 7-13 per group).

(B) Representative photographs of carotid arteries after Evans Blue staining. (C) Reendothelialized area at day 3 after carotid injury in Tlr2-/- mice with injection of HDL^Healthy, HDL^CKD or HDL^SDMA (15 mg/kg). PBS treated animals serve as control (n = 6 per group).

(D) Adhesion of mononuclear cells (MNC) to TNF-a (5 ng/ml) treated endothelial cells after incubation with HDL^Healthy or HDL^CKD (50 Mg/ml). MNC are stained with Dil and HAEC with DAPI (n = 6 per group). (E) Representative micrographs of high-power fields of 6 independent experi^¬ ments obtained by fluorescence microscopy.

(F) Effect of HDL^Healthy, HDL^CKD or HDL^SDMA (50pg/ml) with or without co- incubation with LNAME (0.3 mM) and HDL^CKD or HDL^SDMA with or without co- incubation with DetaNO (1 mM) on TNFa-induced (5 ng/ml) endothelial

VCAM-1 expression (n = 3-6 per group).

All data are presented as mean ± SEM.

Figure 6: Modified HDL activates TLR-2 to induce endothelial dysfunction and hy- pertension

(A) Representative histogram of TLR-2 surface expression on HAEC determined by flow cytometry.

(B) Representative histogram of TLR-4 surface expression on HAEC determined by flow cytometry.

(C) Effect of TLR-2 and TLR-4 inhibition using a blocking antibody (10 pg/rnl, each) or an isotype control antibody (10 pg/rnl) on endothelial NO produc^¬ tion after incubation with HDL^Healthy, HDL^CKD and HDL^SDMA (0.5 prnol/g SDMA, 50 pg/rnl HDL, n = 3-5 per group).

(D) Δ ABP in Tlr2-/- and Tlr4-/- mice 90 min after i.v. injection of HDL^Healthy and HDL^CKD (15 mg/kg HDL, n = 3-6 per group).

Tables

Table 1: Baseline characteristics of study participants

Healthy CKD-

CKD-II° CKD-V°

Controls III/IV° P

N = 15 N = 15

Age (years) 60 ± 8 60 ± 11 67 ± 12 64 ± 11 0.153

26.7 ± 29.1 ± 30.1 ± 28.1 ±

BMI (kg/m²) 0.177

2.8 3.9 5.5 5.5

Male (%) 55.6 57.9 53.3 66.7 0.902

Diabetes mellitus - 15.8 26.7 40.0 0.386 (%)

HbAlc (%) 5.8 ± 0.3 5.8 ± 0.6 6.4 ± 1.2 5.4 ± 0.6 0.018 CAD (%) - 10.5 33.3 46.7 0.012

Hypertension (%) - 94.7 93.3 93.3 0.000

MAP (mmHg) 101 ± 7 103 ± 10 96 ± 12 98 ± 13 0.157

CRP (mg/dl) 1.9 ± 1.6 3.3 ± 2.7 3.6 ± 3.3 6.3 ± 6.1 0.035

Total cholesterol

191 ± 11 203 ± 47 185 ± 45 169 ± 57 0.116 (mg/dL)

Triglycerides 170 ±

122 ± 73 174 ± 91 156 ± 72 0,646 (mg/dL) 143

LDL cholesterol

118 ± 28 123 ± 41 107 ± 35 96 ± 27 0.046 (mg/dL)

HDL cholesterol

53 ± 16 49 ± 13 45 ± 9 36 ± 7 0.001 (mg/dL)

Creatinine (mg/dl) 0.8 ± 0.1 1.0 ± 0.2 2.3 ± 0.8 8.3 ± 3.9 0.000 eGFR, MDRD

102 ± 14 71 ± 14 32 ± 20 7 ± 3 0.000 (mL/min/1.73 m²)

Urea (mg/dl) 37 ± 12 44 ± 13 100 ± 31 136 ± 52 0.000

Current Medication

ACE Inhibitors - 42.1 20.0 26.7 0.019 (%)

ATI blocker (%) - 47.4 20.0 20.0 0.023

Beta Blockers - 68.4 73.3 73.3 0.001 (%)

Ca- Antagonists - 26.3 20.0 20.0 0.386 (%)

Diuretics (%) - 73.7 93.3 60.0 0.000

Statin therapy - 52.6 80.0 40.0 0.000 (%) Table 2 : Baseline characteristics of children study participants

Healthy CKD-

CKD-V°

Controls III/IV°

N = 15

N = 10 N = 15

Age (years) 13 ± 4 13 ± 2 13 ± 4 0.700

22.9 ± 23.3

BMI (kg/m²) 22 0.460

2.8 4.6

Male (%) 48 64 60 0.880

Diabetes mellitus

0 0 0

(%)

CAD (%) 0 0 0

Hypertension (%) 0 15 11 0.000

CRP (mg/dl) 0.8 ± 1.4 1.1 ± 1.0 4.3 ± 2.9 0.042

Total cholesterol

127 ± 27 143 ± 73 159 ± 81 0.065 (mg/dL)

Triglycerides 147

70 ± 149 44 0.880 (mg/dL) 105

LDL cholesterol

50 ± 27 66 ± 35 77 ± 46 0,063 (mg/dL)

HDL cholesterol

73 ± 19 81 ± 39 93 ± 85 0,072 (mg/dL)

eGFR, MDRD

112 ± 18 23 ± 12 8 ± 13 0.000 (mL/min/1.73 m²)

Current Medication

ACE Inhibitors

7.7

(%)

ATI blocker (%)

Beta Blockers

(%)

Ca- Antagonists

7.7 11.1

(%)

Diuretics (%)

Statin therapy

(%) Methods

Study design and participants

The cross-sectional study enrolled adults and children with CKD (KDOQI Stage CKD-II°, CKD-III°/IV° and CKD-V°). The study was conducted with approval of the local ethics committee (Ethikkommission der Arztekammer Saarland, Germany, Ethikvotum 06/10; Research Ethics Committee, Great Ormond Street Hospital, London, UK) and informed consent was obtained from all participants. Baseline characteristics are shown in Table 1 +2.

Lipoprotein Isolation

Lipoproteins were isolated from fresh, fasting plasma by density gradient ultra- centrifugation (HDL: density 1.063 to 1.21 g/cm³, LDL: density 1.006-1.063 g/cm³). Potassium bromide was used to adjust the density. Lipoprotein concentrations used in the present study were based on protein content, which was determined by Bradford assay. Purity of each lipoprotein preparation was assessed by SDS-PAGE and subsequent Coomassie Blue staining of the gel . Lipid-free human plasma Apo-Al was further purified from delipidated HDL

Preparation of reconstituted HDL (rHDL)

Reconstituted HDL (rHDL) comprising apoAl, POPC, and cholesterol was prepared by the sodium cholate dialysis method using an apoA- I/POPC/cholesterol molar ratio of 1 : 100 : 10.

Endothelial cell culture

Human aortic endothelial cells (HAEC) were obtained from Clonetics^® and cultured in Endothelial Cell Growth Medium-2 (Clonetics^®, Lonza, Verviers, Belgium) supplemented with 10 % Fetal Calf Serum (FCS, Gibco, Invitrogen, Basel, Swit- zerland). Before the experiments, cells were starved in Endothelial Basal Medium (Lonza, Verviers, Belgium) with 0.5 % FCS overnight. Cells were used within passage 4 to 6.

ESR spectroscopy analysis of nitric oxide (NO) production of HDL in HAEC

NO production was measured by ESR spectroscopy analysis with the use of the spin-trap colloid Fe(DETC)₂ as described previously (Sorrentino et al ., 2010 Circulation 121, 110-122). In brief, HAECs (150,000 per well) were cultured on 6 well plates for 24 hour in endothelial cell growth medium- 2 and then incubated with H DL (50 pg/mL) for 1 hour. Cells were than washed twice with ice cold Krebs-Hepes buffer (KHB) . Then, 900 pL of KHB and 300 pL of colloid Fe(DETC)₂ (final concentration 285 pM) was added to each sample and incubated at 37°C for 60 min . ESR spectra were recorded using a e-scan ESR spectrometer (Bruker, Karlsruhe, Germany) . ESR instrumental settings were as follows : center-field (B0) 3280G, sweep 198G, microwave power 4db, amplitude modulation 8G, 4096 points resolution, sweep time 120s and number of scans 4. Signals were quantified by measuring the total amplitude after correction of baseline and subtracting background signals. Incubations with colloid Fe(DETC)2 alone were used to correct for background signals. The mean value of two different samples of each patients / healthy subject was used for further analysis.

ESR spectroscopy analysis of superoxide production in HAEC

Endothelial superoxide was measured as described previously (Sorrentino et al . , 2010 Circulation 121, 110- 122) using ESR spectroscopy and the spin trap 1- hydroxy-3-methoxycarbonyl-2,2,5,5-tetramethylpyrrolidine (CM H) . Cells were stimulated with the isolated lipoproteins or Pam3CSK4, respectively, for 1 hr.

Western blot analysis

Protein expression was determined by Western blot analysis. Cells were lysed in lysis buffer (50 mmol/l Tris pH 7.5, 150 mmol/l NaCI, 1 mmol/l EDTA, 0.5 % NP- 40) supplemented with protease and phosphates inhibitors (10 pg/ml Aprotinin, 10 pg/ml Leupeptin, 0.1 mmol/l Na₃V0₄, 1 mM NaF, 1 mmol/l PMSF) . Protein concentration was determined by the Bradford assay. 30 pg of protein were loaded per lane, resolved by 10 % SDS-PAGE, transferred to a PVDF membrane (Millipore, Billerica, MA, USA) by semidry transfer. Antibodies to human eNOS, phosphorylated eNOS at Serl l77 and phosphorylated eNOS at Thr495 were purchased from BD Transduction Laboratories (BD biosciences, Heidelberg, Germany) and used in a dilution of 1 : 2000. Antibody to human VCAM- 1 (R&D systems, Abingdon, U K) was used in a dilution of 1 : 2000. GAPDH (Millipore) was used as loading control . Determination of methylarginines using HPLC-ESI-MS/MS

EDTA-plasma or lipoprotein solution (50 pi) was supplemented with 20 pi internal standards solution (460 pmol/L 13C6-Arg and 45 pmol/L D7-ADMA both from Cambridge Isotope Laboratories, Andover, USA). To precipitate the proteins 200 pi of mixture methanol/acetonitrile / 0.1 M aqueous zinc sulphate (80% + 10% + 10%, v/v/v) (Sigma) was added . After centrifugation at 16,000 x g for 10 min at 10°C, the supernatant was collected and dried under nitrogen at 25°C. The dried sample was derivatised with 100 μΙ butanol solution containing 3 M HCI (Regis technologies, Socochim Lausanne) at 65 °C for 20 min . The derivatised samples were dissolved in 100 pL aqueous 0.2% trifluoroacetic acid (v/v) solution.

For the HPLC separation a Rheos 2000 pump with a degasser Rheos DSO-LC (Flux Instruments, Basel, Switzerland) and Betasil Phenyl-Hexyl (Symmetry, 3 pm, 3 x 100 mm) (Thermo Fisher Scientific) was used . The flow rate was 300 pL/min with a gradient elution of solution A (0.2% aqueous trifluoroacetic acid, v/v) and solution B (acetonitril) . The percentage of organic modifier (B) was changed linearly as follows : 1 min, 5%; 5 min, 50%; 2.5 min isocratic, 50%; 2.5 min, 5%. The injection volume was 10 pL and the total analysis run time was 10 min . Mass spectrometric analyses were performed using a Thermo Finnigan TSQ 7000 triple quadrupole mass spectrometer equipped with an electrospray ion source (ESI) operating in positive mode. Observed multiple-reaction monitoring (M RM) transitions were : m/z 259 > 228 for SDMA, m/z 259 > 214 for ADMA, and m/z 266 > 77 for its IS (ADMA-D7) . Data collection and analysis were done with Thermo Xcalibur software package, revision 1.2 (Figures 2 B- F) .

Endothelial mononuclear cell adhesion assay

HAECs were seeded in 24 well plates and grown until confluency. After serum withdrawal overnight, they were stimulated with TN Fa (5 ng/ml, 4 h) and H DL (50 pg/ml, 3 h). Peripheral blood mononuclear cells (PB-MNCs) were separated from peripheral whole blood of a healthy human volunteer by Ficoll density gradient centrifugation . PB-M NCs were labelled with Dil according to the manufacturer's protocol (Vybrant™ Cell-Labeling Solutions, Molecular Probes, Invitrogen) . Afterwards, lxl0⁶/well of the labelled M NC were added to the HAECs. After a 4 h incubation period, non-adherent PB-M NCs were carefully removed by washing with phosphate buffered saline (PBS) . The integrity of the endothelial monolayer was confirmed by staining with 4',6-diamidino-2-phenylindole (DAPI, Vec- tashield, Reactolab, Servion, Switzerland). Adherent Dil-labelled PB-M NCs were counted in 4 randomly selected high-power fields using a fluorescent microscope (DM-IRB, Leica) connected to a digital imaging system (Spot-RT; Diagnostic In- strument/Visitron Systems).

Inhibition of endothelial TLR-2 and TLR-4 by blocking antibodies

HAEC were incubated with the blocking antibodies against TLR-2 and TLR-4 (10 pg/rnl each, InvivoGen, San Diego, CA, USA) for lh. Afterwards they were stimulated with HDL and immediately used for experiments.

HDL supplementation with SDMA

HDL from different healthy donors was isolated as described above. After dialysis with 10 mM ammoniumhydrogencarbonate buffer, HDL was incubated with SDMA (Sigma) in different concentrations for 16 hours at 4 °C on a shaker. Afterwards, HDL was again dialyzed extensively to remove not-HDL bound SDMA. HDL bound SDMA concentration was quantified by HPLC/ESI/MS-MS analysis. This procedure yielded SDMA concentrations in the HDL fraction, which were comparable to those measured in HDL of CKD V° patients.

In vitro endothelial cell scratch assay

In vitro scratch assay was used to assess endothelial cell migration as described previously⁵³. HAEC were grown until confluency in 24 well plates in EGM-2 medium supplemented with 10 % FCS. For the assay, medium was changed to EBM containing 0.5 % FCS and 50 pg/rnl HDL was added to each well . Two parallel linear scratches were performed with a 200 μΙ pipette tip and pictures were taken at 4 different positions per well by phase contrast microscopy immediately after scraping and 24 hours later at the same positions. Newly closed distance was calculated by subtracting the width of the scratch after 24 hrs from the width at time point 0. Each experiment was performed in triplicates and mean was calcu- lated.

Effect of HDL on endothelial repair in vivo

The study was approved by the animal ethics committee of the Universitat des Saarlandes and conforms with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Pub. No. 85-23, revised 1996). Male NRMI^nu/nu athymic nude mice, aged 7 to 10 weeks, were anaesthetized with ketamine (100 mg/kg IP) and xylazine (5 mg/kg IP). Carotid artery electric injury was performed as described previously (Sorrentino et al ., 2010 Circulation 121, 110-122). In brief, the left common carotid artery was in- jured with a- bipolar microregulator (ICC50, ERBE-Elektromedizin GmbH, Tue- bingen, Germany) . An electric current of 2 W was applied for 2 seconds to each millimeter of carotid artery over a total length of exactly 4 mm with the use of a size marker parallel to the carotid artery. 15 mg/kg of HDL were injected in vol- ume of 250 μΙ 3 hours after carotid injury via tail vein injection with a 30-gauge needle. Three days after carotid injury, endothelial regeneration was evaluated by staining denuded areas with 50 μΙ of solution containing 5% Evans blue dye via tail vein injection as described previously (Sorrentino et al ., 2010 Circulation 121, 110- 122) . The reendothelialized area was calculated as difference between the blue-stained area and the injured area by computer-assisted morphometric analysis. This model has been shown to allow accurate quantification of reendo- thelialization . H DL from each CKD/healthy subject were injected into 2 nude mice, and mean values of reendothelialized area were used for analysis.

Blood pressure measurements The study was approved by the animal ethics committee of the Universitat des Saarlandes. Systolic blood pressure was measured in conscious animals with the use of an occlusive tail-cuff plethysmograph attached to a pneumatic pulse transducer (BP-2000 BP analysis system, Visitech Systems, Apex, NC) . Eight mice from each group underwent repeated measurements at 37°C before and 90 minutes after injection of HDL ( 15mg/kg) .

TLR2^V" & TLR4^~/~ mice

TLR2^V" and TLR4^V" were obtained from Jackson Laboratory. Flow cytometric detection of TLR expression

TLR expression was assessed by flow-cytometry using appropriate antibodies targeting TLR- 2 (clone : 11G7), TLR-4 (clone : HTA125) (BD Pharmingen and Immunokontact), TLR- 1 (clone : H2G2) and TLR-6 (both from Invivogen) .

Statistics

All data are expressed as mean ± SEM . Statistical comparisons were made by one-way analysis of variance or the nonparametric Kruskal Wallis test and a P- value <0.05 was considered statistically significant. Bonferroni adjustment was performed in the comparisons of the groups. All analyses were performed with GraphPad Prism (Version 4.0, GraphPad Software, Inc. ).

Claims

Use of SDMA for a determination of a HDL-SDMA level in blood or blood- plasma.

Use according to claim 1 for the determination of a level of an endothelial dysfunctional HDL-SDMA complex.

Use according to claim 2 for the determination of the level of the endothelial dysfunctional HDL associated SDMA.

Use according to at least one of the claims 1 to 3 wherein the HDL-SDMA level is measured by Tandem mass spectrometry (MS/MS), HPLC Amino acid profiling and/or ELISA.

Use of SDMA for the evaluation of the benefit of an administration of a medicament for effecting an elevation of the HDL-level in a patient with a HDL-level which should be increased or modified.

Use according to claim 3 wherein the amount of SDMA associated with HDL is determined yielding a HDL-SDMA level and if the SDMA concentration associated with HDL (HDL-SDMA) is about > 0.025 prnol/g protein, the medicament for elevation of the HDL-level is not administered to the patient.